CORNELL UNIVERSITY. THE THE GIFT OF ROSWELL P. FLOWER FOR THE USE OF THE N. Y. STATE VETERINARY COLLEGE " X897 8394-1 Cornell University Library QP 171.L43 Problems in animal metabolism :a course 3 1924 000 107 668 Cornell University Library '- The original of tiiis book is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924000107668 PROBLEMS IN ANIMAL METABOLISM A COURSE OF LECTURES GIVEN IN THE PHYSIOLOGICAL LABORATORY OF THE LONDON UNIVERSITY AT SOUTH KENSINGTON IN THE SUMMER TERM, 1904 '''''''' I: ^ \ - Vj BY J: B. LEATHES LECTUllER ON PHYSIOLOGY IN THE MEDICAL SCHOOL OF ST THOMAs's HOSPITAL PHILADELPHIA -- > U . ; P. BLAKISTOK'S SON AND CO. L (. V < \ 1012 WALNUT STREET f, 1906 , , \ n Printed in Great Britain 111 PREFACE The subject chosen for these lectures was, I think, sufficiently indicated in the title under which they were announced. The underlying motive in current investigations is frequently the solution of problems which are not expressly formulated. Such problems are certainly not ripe for dogmatic treatment. But the lectures in the University Laboratory at South Kensington are intended for those students of medicine who have already had the opportunity of learning the essential facts in physiology. They are planned to foster in those studying the practice of medicine an interest in the efforts of others who study the theory, and to keep alive the faith that among the theories of to-day lie the foundations of the practice of to-morrow. In these circumstances it appears permissible or even desirable to enunciate problems the solution of which may still be remote, although in the attempt contentious hypotheses must figure as prominently as facts that have been positively ascertained. The idea of the South Kensington lectures seems to admit the possibility of the written versions appearing as essays, fugitive it may be, but at the time serious, and justified by the experi- ence that the impulse to honest study, the most that a teacher can hope to impart, comes as often from an insight into the aims that direct and sustain the labours of others, as from the most lucid exposition and the most judicious selection of accredited dogmas. As an essay, therefore, these pages were intended to point out that the study of animal metabolism is daily becoming less vi PREFACE the study of the sum total of chemical change in the body, and more the study of the individual chemical reactions, the items that go to form the final sum. It can only be when this tendency has worked itself out, that we shall know how far the hopes are justified, that the power to cope with disease will be increased by a more exact knowledge of what it means regarded as a chemical problem. At present it is difficult to form any conception 6f the future progress of medicine, to which the chemical development of physiology and pathology should not contribute more than it, or perhaps any other development, has contributed in the past. And if practical medicine appears at the present time to owe but little to such studies, it may be that the glimpses of a new outlook in the study of metabolism, which are yearly becoming more frequent, are the first indications of a path that is to lead us far. Disease is a disturbance in the normal balance of concurrent chemical changes, and the balance is in the first instance disturbed because some one or other of the indispensable chemical reactions is either precipitated or checked. It is difficult to believe that the chances of our being able to restore the balance will not be improved the more we learn of the nature of each of these concurrent changes. To understand the structure of the body, it has been necessary to dissect it ; to understand the composite whole of metabolism, it must be analysed, and each particular reaction studied by itself and in its relations to the whole. This is beginning to be done, and that is why there are many who look forward to the work of the coming years along these lines with hope and confidence. CONTENTS I. Introductory : Physiological Chemistry and Meta BOLISM . . . II. The Assimilation and Synthesis of Carbohydrates III. The Catabolism of Carbohydrates IV. The Assimilation and Synthesis of Fat V. The Catabolism of Fat VI. The Assimilation and Synthesis of Proteids VII. The Catabolism of Proteids VIII. The Metabolism of Cyclic Formations . Index ...... I 20 so 72 97 122 144 165 199 ABBREVIATIONS Am. Ch. Jl. . American Chemical Journal. Am. Jl. Phys. . American Journal of Physiology. Ann. . . Annalen der Chemie und Pharmacie. A. P. I. . . Annales de rinsUtut Pasteur. A. St. P. . Archives des Sciences Biologiques de St Pitersbourg. B. . . . Berichte der deutschen Chemischen Gesellschaft. B. Cbl. . . Biochemisches Centralblatt. B. k. W. . Berliner klinische Wochenschrift. Cbl. f. k. M. . Centralblatt fiir klinische Medizin. C. R. S. B. . Comptes Rendus de la Socidte Biologique. D. A. f. k. M. Deutsches Archiv fiir klinische Medizin. D. M. W. . Deutsche Medizinische Wochenschrift. D.-R. A. . Dubois-Reymond's Archi-v fUr Physiologie. Ergeb. , . Ergebnisse der Physiologie. H. B. . . Hofmeisier's Beitrdge zur Chemischen Physiologie und Pathologie. H,-S. Z. . . Hoppe-Seyler's Zeitschrift fiir Physiologische Chemie. J. C. S. . . Journal of the Chemical Society. Jl. of Phys. . Journal of Physiology. J. Pr. Ch. . Journal fiir Praktische Chemie. M.f. Ch. . MoTMtshefte fiir Chemie. M.J. . . MalfsJahresberichtefiirThierchemie. M. M. W. . Miinchener Medizinische Wochenschrift. Pfi. A. . . Pflilger's Archiv fiir die Gesammte Physiologie. S. A. . . Schmiedeberg's Archiv fiir Pharmakologie und Experi- m.entelle Pathologie. Sk. A. . . Skandinavisches Archiv fiir Physiologie. V. A. . . Virchow's Archiv fiir Pathologie und Pathologische Anatomie. Z. f. B. . . Zeitschrift fiir Biologie. Z.f. k. M. . Zeitschrift fur klinische Medizin. viil PROBLEMS IN ANIMAL METABOLISM LECTURE I INTRODUCTORY: PHYSIOLOGICAL CHEMISTRY AND METABOLISM Metabolism is a very simple phrase, admirable in its compre- hensiveness ; it covers all the chemical changes in living organisms which constitute their life, the changes by which their food is assimilated and becomes part of them, the changes which it undergoes while it shares their life, and finally those by which it is returned to the condition of inanimate matter. Gathered together under this one phrase are some of the most intricate and inaccessible of natural phenomena. It implies also, and gently insists on the idea, that all the phenomena of life are at bottom chemical reactions. When a muscle twitches no less, than when a gland secretes, it is not too much to say that when we are moved to tears or laughter it is chemical reactions that are the underlying causes to which ultimate analysis must lead us. When it is possible to give an adequate account of animal metabolism in this sense, it is clear that physiological chemistry will have done its work and be an extinct science. But it is not necessary to point out that this ideal goal lies far beyond the horizon. And yet there must be few branches of science in which during recent decades more vigorous activity has been displayed A 2 INTRODUCTORY [iect. or in which what has been achieved holds out greater promise for the near future. Especially in the past twenty years the number and the energy of workers in physiological chemistry has increased from year to year, and the progress effected by their work has become more and more significant. At the same time it is remarkable that very much of what we value most in the contributions made during this period to physiological chemistry has not been immediately concerned with the problems of metabolism. It is rather in the elucida- tion of the nature and constitution of the material in which the processes of life are carried out, than in the interpretation of the chemical changes themselves, that progress has been conspicuous. A very brief reference to some of the advances made during the last twenty-five years, which are and always will be landmarks in the history, of this branch of science, will suffice to illustrate this fact, and at the same time show how it is that while the chemical chapters in physiological treatises yearly grow in length, the chapters on metabolism have to be left compara- tively little changed. In no subject is the progress effected greater or more familiar than in the chemistry of the carbohydrates. We have learnt that the simple sugars of which all carbohydrates are composed are either aldehydes or ketones of polyatomic alcohols, and may have as they occur in nature either five or six carbon atoms in chain : that they have therefore three or four unsym- metrical carbon atoms, each of which, that is to say, is combined with four atoms or groups of atoms that are all ■ different from each other : that consequently the different possible spatial arrangements of these atoms or groups of atoms about each unsymmetrical carbon atom give rise to differences between sugars in other respects identical : galactose and glucose, for instance, are both aldehydes, both hexoses, but owing to the different spatial configuration of the parts of the molecule, they are endowed with properties by which they may be readily dis- tinguished from each other. The comprehension of these fundamental properties of the sugars has led to the determina- tion of the chemical relationships between almost all the I.] CHEMISTRY OF CARBOHYDRATES 3 different sugars that occur in living organisms, to the synthesis of most of them, and the conversion of many of them in vitro into others. The work principally of Emil Fischer has taught us exactly where in the molecule each atom must stand, what is formed if it changes places with one of the other groups in the molecule, exactly why it is that there are so many varieties of sugar in nature with the same empirical and even constitutional formula, which of all the conceivable dispositions of the same atoms and groups of atoms do not occur in nature, and how many of these too may be built up. The chemistry of the carbohydrates has become a problem in geometrical permuta- tions, and almost all the possible permutations have been identi- fied, and most of these synthesised. And at the same time we have learnt that the geometrical character of chemical problems so conspicuous among the carbohydrates must be appreciated among compounds of other kinds for which as yet we have no such definite data. No less remarkable has been the development of the chemistry of proteids. The very first things we learn about proteids — the colour reactions — which were for so long nothing but empirical tests, are now rich in interest, significance, and associations. In the xanthoproteic reaction we recognise the facility with which aromatic radicals form nitro-compounds, and the well-known chromophoric properties of — NOg groups. Millon's reaction, which is a general reaction for mono-oxy- aromatic compounds provided that they do not contain — NOg groups, reveals the presence of tyrosine. But far more interest- ing than these two are those known as the biuret reaction, and the reaction of Adamkiewicz. Many other substances besides biuret and proteids are known to give a pink tinge to alkaline solutions of copper. H. Schiff studied the nature of these substances, and having determined their common characteristics, made the following generalisation : all substances that contain two \nh. 4 INTRODUCTORY [lect, groups united by the carbon atoms either directly to each other or to the same carbon or nitrogen atoms, give the biuret reaction. Oxamide, CO . NHj CO . NH2 therefore, and malonamide, \c—cu^—cf HaN/ \NH2 behave in this respect similarly to biuret itself, which is ^C— NH— Cf H2N/ \NH2 Inferences as to the constitution of proteid molecules were indi- cated by these observations. But the same. reaction is given by a basic condensation product of glycocoll, amido-acetic acid, which was first described many years ago by Curtius, the exact constitution of which has only recently been determined. And this compound is not covered by Schiff's generalisation. A; number of other condensation products of glycocoll and other amido acids, to which Fischer has given the name of peptides, also give the reaction. The general formula for these pep- tides is : ^^C_Ce°/H In the simplest of these bodies, glycylglycine, the value of R is H, and of R^ is — CHg . COOH. These values do not confer on the compound the property of giving the biuret reaction ; but certain other values do. In this way the relationship between these compounds on the one hand, and oxamide and the other substances which by Schiff's rule give the biuret reaction on the other hand, becomes clear, as also does the sort of expansion which this rule must undergo in order to include the poly- I.] CHEMISTRY OF PROTEIDS 5 peptides which give this reaction. In the expanding process, however, the rule loses its value ; the determining conditions and the exceptions become too complicated to be included in a single simple sentence. But the significance of what we have learnt about the biuret reaction in its bearing on the constitution of proteid molecules has become very great. For these molecules, when broken up by the hydrolytic action of acids and alkalies, or of water itself at high tempera- tures or in the presence of enzymes, yield large quantities of amido acids, and must therefore be very largely composed of amido-acid radicals. And under these conditions the polypep- tides also are broken up into their constituent amido acids. It is fair to argue, therefore, that proteids and polypeptides give the biuret reaction, because they both contain amido acids combined in the same way. Fischer's polypeptides which give the biuret reaction are not proteids, but they are very closely related to that large part of all proteid molecules to which the biuret reaction which all proteids give is due. The first step towards the synthesis of proteids has been taken, and at the same time the meaning of the most universal and important of all the proteid reactions, as well as the constitution of the greater part of the molecule of all proteids, has been elucidated. The reaction of Adamkiewicz also has an interesting story. For a long time it was reckoned of little account ; it was merely an empirical test, a trick ; and not reliable as that. For, carried out according to the original prescriptions, with glacial acetic acid and sulphuric acid, it was found to be capricious, and fre- quently failed. This capriciousness was shown by Hopkins to be due to the fact that glacial acetic acid had nothing to do with the reaction: when glacial acetic acid gave the test, it did so because it contained as an impurity an oxidation product, glyoxylic acid. Glyoxylic acid is not always present in acetic acid, and hence the uncertainty of the test. Carried out with glyoxylic acid itself, the reaction is exceedingly delicate, and quite certain. Hopkins, moreover, showed what it was in the proteids that caused them to give the reaction, and isolated as a crystalline substance a new component of the proteid 6 INTRODUCTORY [lect. molecule which proved of the greatest interest. Not only is it the cause of the Adamkiewicz reaction, it is the substance produced in the course of tryptic digestion of proteids which reacts with bromine in the long-familiar tryptophane or pro-» teinochromogen reaction ; hence the name tryptophane given by Hopkins to the new substance. It is probably the one and only source of all the familiar derivatives of indol which are formed from proteids by the action of bacteria in the intestine and elsewhere, and therefore of the indican in the urine ; it is the source of the indigo in the indigo plant; and since it is from tryptophane that in dogs and certain other animals, a substance found in the urine, kynurenic acid, which is a quino- line derivative, is foi-med, it has even been suggested that it is from the tryptophane group in proteids that all the important vegetable alkaloids that are based on a quinoline or pyridine ring are formed in the course of the proteid metabolism of the plants in which they occur. As soon, therefore, as we learn anything about proteids we learn how much of what is interesting and important in our knowledge is the result of the most recent work. But -this is of course only a small part of what the period under considera- tion has brought to light on this subject. The study of the action of boiling mineral acids on proteids was begun by Liebig, and in his laboratory led to the discovery among the resulting products of leucine, tyrosine, glycocoll, aspartic and glutamic acids. But as the first simple methods of separating these products from the mixtures in which they were obtained con- sisted in direct crystallisation or precipitation with metallic salts, only those products were isolated which were either particularly abundant or particularly insoluble; And the further investigation of the constitution of proteids along these lines by Schiitzenberger and others led to little that was definite. Within the last twenty years, however, on two occa^ sions a new irhpulse has been given to this study, which has in each case led to the most important results. In the first place, the. use of phosphotungstic acid as a re- agent for the precipitation of organic bases led to Drechsel's 1.] CHEMISTRY OF PROTEIDS 7 discovery of lysine as a constituent of proteids, and this to the proof given by Hedin that arginine is formed in considerable quantities in the hydrolysis of most proteid.s. The earliest fruitful use of this reagent in physiological chemistry was in the work of Schultze, who first described arginine as' present in young seedlings ; but the significance of this discovery was not at that time fully understood. Subsequently, in addition to lysine and arginine a third widely distributed basic con- stituent of proteids was added to the list by Hedin and Kossel," who independently found histidint in proteids of various kinds. More recently still, Emil Fischer has applied to the investi- gation of the cleavage products of proteids a number of methods not previously made use of for this purpose ; first of all by Converting the mono-amido acids into their ethyl esters, and distilling these in vacuo, he obtained these acids in a pure state, and then for the separation of the different components of each fraction he has devised new methods which have already given excellent results ; the most important of these consist in the use of phenyl isocyanate and naphthalene sulphonyl chloride. In this way several new amido acids have been brought to light among the products of proteid hydrolysis, and in many cases, too, others more familiar where they had not previously been found. Alanine, phenyl-alanine, amido-valerianic acid, serine; the pyrrholidine carboxylic acids, should now receive recognition as constituents of proteid molecules just as much as the sub- stances discovered by Liebig and his followers, leucine, tyrosine, glycocoll, and the rest. And these methods for isolating the mono-amido acids are likely to yield still further results of importance, while the detailed study of the combinations of the amido acids which is being carried out by Fischer is not likely to stop till still more important knowledge of the con- stitution of proteids has been acquired. If we take into account those proteids as well which are of a more complex structure, in which besides a typical proteid component, a component of quite a different nature is found, the progress that has been made in the comprehension of these prosthetic groups, as they have been called, the nucleic acids 8 INTRODUCTORY [lect. haematin, and the carbohydrate derivatives which combine with proteids, has kept pace with the development of the chemistry of proteids themselves. Especially in the case of the nucleic acids is this conspicuous. Almost all that is known of these sub- stances is due to the work of the last two decades, and much of this is so fresh and familiar that it is sufficient only to refer to it. In the chemistry of the glyco-proteids, the first and still the most important work was that of Schmiedeberg, who published his account of the carbohydrate derivatives obtained from carti- lage barely fifteen years ago. It has led to many other additions to the knowledge of this class of compound which are a great advance on the notions current twenty years ago. And what Schmiedeberg did for the glyco-proteids Nencki did for haematine; the work initiated by him seems likely to lead before many years pass to the synthesis of this most important pigment. Under these main headings and a few minor ones, among which the determination of the constitution of adrenaline would find a place, the principal achievements in what has been the most active and successful period in the history of physiological chemistry would probably almost all be included. - It is not, therefore, the difficulties connected with metabolism that have been removed. It is rather that it is possible now to have clearer views as to the composition and constitution of the material in which the chemical processes of life are carried out, and not that the chemical processes themselves are clearer for what has lately been accomplished in physiological chemistry. The knowledge that has been gained may be looked on as a sign that the time is at last approaching when we may hope to see some of the problems solved. To set about the study of reactions when only vague ideas about the nature of the reagents are possible, is not likely to lead to the desired results. Much of the discussion of problems in metabolism must have been in the past, and must even now be premature speculation. We may think already that the old disputes over the conver- sion of proteids, carbohydrates, and fats into each other were sometimes entered upon with little more regard for than know- I.] THE CHEMICAL POWERS OF CELLS 9 ledge of the chemical nature of the substances dealt with, or of the character of the changes propounded ; and attempts to handle sych questions now are still not likely to escape a similar judgment in the future. Nevertheless the strictly logical order has been departed from in some investigations that have led to valuable results in the past, and we still must wait for the rational elucidation of facts in metabolism that were proved by physiologists long ago. The formation of fat from starch by animals is accepted as proved, no less than the conversion of sugar into carbonic acid and water. Such facts we are accustomed to handle with familiarity, but it cannot be said that the chemistry of either of these changes, or of many others of equally fundamental importance, is much better understood now than thirty years ago. In the subject with which physiological chemistry deals there is much that has been built up solidly and substantially, as a wing in the building of organic chemistry ; and it is the work of this kind that has raised our hope in recent years. But there is much which is still merely of the nature of scaffolding, and it is a question yet whether the permanent structure that is to replace this scaffolding can be built of the same material and in the same style as the permanent parts already in existence. It has often been pointed out that no laboratory can rival an animal cell. Changes that we know take place in living cells can often be effected also in the laboratory. But when this is the case, the methods of the laboratory are far too violent, and generally too impotent as well, for it to be possible to suppose that they reproduce the methods by which the living cell works. In those cases in which it is sometimes thought that the steps of a reaction carried out in living organisms are known, these steps may be laboriously followed in the laboratory, but only with the use of agencies, and at a cost of material, which show the difference between the conditions of the two procedures. Carbonic acid can be reduced to carbonic oxide by passing it over red-hot charcoal, and this gas acting on dry potassium hydrate is converted into formic acid. Formate of lime on dry 10 INTRODUCTORY [lect. distillation yields small quantities of formic aldehyde, and formic aldehyde under the influence of lime-water will condense to form sugars, among which are small quantities of a sugar a-acrose, from which the sugars that occur in Nature can be prepared. These sugars are believed to be formed in the parts of plants which contain chlorophyll from carbonic acid, also by way of carbonic oxide and formic aldehyde, under the action of sunlight and the active substances that exist in the living plant. But if the stages in the reaction are the same under the two sets of conditions, there is nothing else that is. From carbohydrates animals certainly, and plants probably, make fats or oils, a transformation at which we can only stand aghast; or stiarch and sugar are oxidised smoothly and quietly at a low tempera- ture, at which, when protected from contact with living matter,, they show no tendency whatever to oxidation. The compara- tively simple changes involved in condensation and hydrolysis present just as striking differences between the methods of the laboratory and of the cell. Hippuric acid is very readily obtained by allowing benzoyl chloride to act on amido-acetic acid in the presence of caustic alkali, but in the dog's kidney it is made no less easily, without the use of either the acid chloride or caustic alkali, from the simple acids or their salts, Benzoic acid and glycocoll can, it is true, be made to condense directly in vitro, but for this they must be heated together in the dry state in a sealed tube at a temperature of i6o° C. for twelve hours. Urea has been synthesised artificially from the same substances as those from which, according to the commonly accepted theory of Schmiedeberg, it is made in the liver, ammonium carbonate or carbamate. But to prepare it from the former Drechsel employed either the alternating current given by an induction coil and interrupter worked by several Grove cells, or a constant current in the presence of platinum black : from the carbamate he obtained it by heating solutions in sealed tubes to 135° C. And in every case, in spite of the energetic measures, the yield was small. In the body the yield is almost quantitative, and the conditions are the ordinary normal conditions of life. For the hydrolytic changes brought I.] THE CHEMICAL POWERS OF CELLS 11 about constantly under these same conditions b)' living cells, we have to use high temperatures and strong acids or alkalies. Compare, too, the procedure by which taurine has lately been obtained from cysteine with the chemistry of the cells. The mercaptan group is oxidised to the sulphonic acid with bromine, and the product heated with water in a sealed tube for four hours at 240° C. in order to remove carbonic acid. Another recent instance is the decomposition of arginine. By boiling this substance with baryta water, Schultze obtained 20 per cent, of the theoretical amount of urea, and Hedin, in the same way, from 20 g. of the silver salt was unable to get enough urea to prepare a pure specimen. The cells of the liver and other organs, acting on arginine, make a clean cleavage into urea and ornithine : in one experiment Dakin and Kossel found that 5 c.c. of the fluid expressed from crushed liver cells completely converted 5 g. of arginine into its components in five minutes. Wherever we turn, in the reactions carried out in metabolism agencies appear at work of an entirely different character from those of which we have command in the laboratory. And even when the last secrets of the constitution of proteids have been laid bare, and when proteids complete in every detail have been synthesised, this supreme achievement of organic chemistry will still be merely introductory to the physiological chemistry which is to give us an insight into and teach us how to control the reactions that lie behind the phenomena of life. There was a time when physiological questions were debated without regard to facts that could be ascertained by the study of anatomy. This was succeeded by a period in which the study of physiology was temporarily identified with the study of anatomy. Then when all that anatomy could contribute had been learnt, it was found that the work of the physiologist, as we now understand it, was only beginning. So, too, there has been a time when the problems of physiological chemistry have been approached with but little knowledge of the nature of its material ; a period when this branch of physiological study began to appear as nothing but a branch of organic chemistry : 12 INTRODUCTORY [l but sooner or later it must be recognised that there is work here for physiology to do which it cannot get done elsewhere, any more than the whole of physiology could be learnt by dissections. This time may not have come. There is certainly very much that has yet to be learnt that is of the nature of pure chemistry. But there are certain conceptions which have had a growing influence in recent years, and which promise to be of great significance in the interpretation of physiological processes. Foremost among these is the idea of the part played in such processes by enzymes. In 1897, Buchner first found that a substance can be obtained from the living yeast plant, solu- tions of which, though entirely free from living organisms of any kind, cause sugar to be resolved into alcohol and carbonic acid, just as the living yeast cells do. The inference is that, since this substance is found in the plant, when the plant ferments sugar it does so because it contains this substance, and not because it is alive. The difficulties commented on above, of tracing any similarity between the conditions under which chemical changes are effected in the laboratory, and those under which the same changes are brought about in living organisms, led to the assumption that life was a cause of chemical change. Vital activity, the action of some property of living things which is not to be defined in any other way, and is not shared by anything that is not alive, was the only explanation forthcoming to meet this difficulty. The explana- tion was merely a verbal one, but there was no other. And up to the time of Buchner's discovery, the fermentation of sugar by yeast was always cited as the typical instance of fermenta- tion due to the vital activity of an organism, an organised ferment ; and this kind of fermentation was kept strictly distinct from fermentations brought about by enzymes or unorganised ferments — chemical substances, that is, that are active as ferments where there is no life. There were two totally distinct causes of fermentation : life, and enzymes. When, therefore, what had always been regarded as the crucial instance of the ferment I.] ENZYMES AND VITAL ACTIVITY 13 action of life was found to be nothing of the kind, there was a movement about the foundations of physiological- belief When yeast ferments sugar, it is a typical enzyme that is at work ; the only difference between this enzyme and the typical " unorganised ferments " is the accidental one that the latter are secreted by cells so as to act in external media removed from the cells, whereas this acts within the cells, does not leave them, and is therefore not to be found elsewhere, unless special measures are taken.to crush the cells and express their contents. The surrender of this point need not involve, and has not involved, the surrender of the whole position. It is only one among myriads of changes which we are in the habit of attributing to vital activity. But it has made more obvious than it was before, that in regarding life as the cause of the chemical reactions underlying the phenomena of life, we are indulging in a verbal illusion. It has called to mind, too, the fact that this is not the first or only instance in which important functions in metabolism are served by enzymes that are not excreted from the cells, but act within them. More than forty years before this, Claude Bernard had ascribed to an enzyme acting in the liver cells the conversion of stored glycogen into sugar for distribution by the blood to other parts of the body. And a suggestion made about the same time by Ludwig has also been recalled. Physiological chemistry, he said, may some day prove to be a chapter in the chemistry of catalytic action. It is remarkable that since Buchner's discovery, and in part, no doubt, as a result of the stimulus it gave, a large number of other intracellular enzymes have been discovered, so that our conception of the use made of enzymes in Nature has had to be very considerably extended. The first enzymes to be detected and studied, those of the digestive secretions, form a class physiologically, but not essentially, distinct from the .others. They are discharged from the cells, and serve the cells by bringing about changes in the fluid which is in contact with the cells. The ease with which this fluid could be removed from the neighbourhood of the cells that produce it, accounts for the fact that the enzymes of this class were for so long the only 14 INTRODUCTORY [lect. substances generally recognised as having the properties of enzymes, and also for the fact that it has not been easy to dissociate the idea of an enzyme from that of a secretion. In addition to these there are enzymes which are to be found only intimately associated with the cell substance and confined within the Hmits of the cell. Among these there are some the action of which has long been known, because the products of the changes they set up are formed in larger quantities than are necessary for the internal economy of the cells themselves. Besides the enzyme in yeast that causes alcoholic fermentation, the enzyme that causes lactic acid fermentation, in the Bacillus acidi ladici, has been isolated.^ The other fermentations of the same class which have not been proved to be due to enzymes would include, to name the most familiar, acetic and butyric fermentations of sugar and the ammoniacal fermentation of urea. In the sugar fermentations there is apparently an extraordinary extravagance in the wanton destruction of the food stuff of the cells. It is partially broken down by them, and the products are turned out in large quantities in their environment. It is common to speak of these substances, alcohol especially, as the excreta of the cells. But this is clearly incorrect. In the formation of alcohol from sugar only 4 per cent, of the energy of the sugar is liberated, and practically none when lactic acid is formed. These substances are intermediate, not final products of metabolism. The formation of an inter- mediate product in such large excess over the needs of the organism is a most remarkable phenomenon in the economy of the species, intelligible conceivably as a protective device. The over-production of a special substance, to which the species producing it is less sensitive than its rivals in the struggle for existence, might serve to keep these rivals away from the feeding ground. Milk that has once been effectively occupied by the lactic acid bacillus is a preserve in which poachers do not prosper. If these fermentations are all of them the work of intra- 1 Herzog, H.-S. Z. 37, 381, 1903 ; Buchner and Meisenheimer, B. 36, 634, 1903. I.] INTRACELLULAR ENZYMES IT, cellular enzymes, as is known to be the case in the instances of alcoholic and lactic acid fermentation,- then these enzymes constitute a second physiologically distinct group, an intra-' cellular group characterised by the over-production of a meta- bolic product' The other intracellular enzymes, distinguished from these by the absence of this characteristic, have been the latest to draw upon themselves the attention of biologists. Not only are they confined to the interior of the cells, but the products of their action are made use of entirely in the internal economy of the cells ; they are produced in quantities that can be dealt with in the cell, and there is ordinarily no over- production. It is the enzymes of this physiological group that are of special importance in the interpretation of metabolic processes. In addition to hydrolytic reactions in the cells, in which glycogen and the disaccharides, fats or proteids, are resolved into their components, there is a great variety of other reactions which have been ascribed to intracellular enzymes. The oxidation of aromatic aldehydes, the oxidation of purine bases, the oxidation of tyrosine ; the removal of amido groups from amido acids, and of urea from arginine ; the decomposition of peroxides, which is sometimes regarded as the means by which oxidation changes are restricted to the appropriate parts of the body, and excluded, for instance, from the blood ; the decomposition of uric acid, of glucose, and of other substances ; all these and many more have been attributed to the operation of enzymes within the cells. It is very possible that some of these attributions may not be confirmed. But the extension of the idea of enzyme action into all phases of metabolic activity has been to effect a revolution in physiological chemistry. There is an outlook even beyond the limits, to which it is hoped that organic chemistry may carry its study of the composition of the material in which these changes are carried out. The question rises at once : what is the precise value of the ' In such a group would be included the enzyme that hydrolyses glycogen in the liver, and others familiar in the physiology of mammalian animals. 16 INTRODUCTORY [i.ect. new conception of the part played by enzymes in animal meta- bolism ? Even supposing that enzymes are to account for all that is remarkable in biological chemistry, are we any better off than when it was all accounted for by the vital activities of the cells ? . If we have no exact idea of what an enzyme is and how it works, it may appear as if we were again trying to satisfy curiosity with an explanation that is verbal and nothing more. It is true that, though the attention of many workers in many countries has been concentrated on the problems of enzyme action for years, very few of the questions on this subject that seem most urgent are finally settled. But a very suggestive summarisation of the present tendencies is contained in Ostwald's definition of cata- lytic action, which was intended to cover, and include as a special instance of catalytic action, the action of enzymes. A catalytic agent is a substance that alters the velocity with which a chemical change is brought about. This is simplicity itself: but what does it imply ? A change in the velocity of reactions can also be brought about by a change in the temperature to which the reagents are exposed. It has been found that every rise of io° in the temperature increases the velocity of most reactions from two to threefold, which means that a reaction which at o° C. lasts a year may at ioo° C. be completed in less than seven minutes. A reaction which lasts a year, we are apt to regard as one that does not take place. If a catalytic agent effects a change in the velocity of a reaction, which is of this order of magnitude, it may appear to bring about a reaction that does not otherwise occur. We say that pepsine' and hydrochloric acid cause proteolysis. But proteids are hydrolysed by water alone, if the temperature is high enough, at i6o° C. for instance. If the rate at which this hydrolysis is effected at i6o° C. is double that at which it would be effected at 150° C. and so on for every 10° lower, then in six hours at 160° as much hydrolysis would be attained as in three years at 40° C. But in six hours at 40° C. , pepsine and hydrochloric acid can accomplish more still, so in this case the acceleration of hydrolysis effected by the enzyme appears to be greater than that due to a rise of temperature amounting to 1 20° C. I.] DEFINITION OF CATALYSIS 17 If enzymes produce changes in the velocity of reactions of this order of magnitude, then the fact that reactions occur in living organisms which do not occur in vitro may have to be differently stated; the reactions may occur in vitro, but too slowly to be detected. The definition implies, therefore, that the peculiarities in the behaviour of substances in living organisms where enzymes are at work may be due not to the endowment of matter with properties differing in kind from the properties it has in inanimate forms ; the substances do not react differently, but with a different velocity. The general properties of substances in vitro and in vivo are the same, and the general laws of reaction in vitro are applicable to reactions in vivo, if we allow for the operation of catalysis. And it is well known, of course, that the phenomena of catalysis are not restricted to the products of life. But it also implies that where differences in the nature of reactions appear to exist in comparing animate and inanimate nature, they may be due to the acceleration of one line of reaction to the exclusion of others — a selective catalysis. Yields in the laboratory are seldom theoretical : in addition to the main line of reaction there are other side lines or branch lines into which the reaction tends to be diverted, and in an extreme case there may be no detectable yield from the main line of reaction. If we suppose that the main line is brought under the influence of a catalytic agent, while the side lines are not, then the yield may become to all intents theoretical. Of if a side line is so influenced, practically the whole reaction may be guided into this line ; and if the side reaction is one that in vitro occurs only to a very limited extent, so that it altogether escapes detection, then the whole course of chemical change may appear to be altered past recognition, and a reaction of which in vitro we know nothing be the characteristic feature of the behaviour of the same substances in the presence of the catalytic agent. It must be remembered, too, that the definition includes the idea that the change in velocity may have a negative sign, and that changes that we are in the habit of regarding as inevitable may B U INTRODUCTORY [lfxt. in the sphere .of action of negative catalysts be avoided altogether. This conception of catalytic action, then, suggests that much if not all of what has hitherto appeared to separate biological chemistry from the chemistry of the laboratory may ultimately disappear. The methods of physical chemistry may, when developed and applied to biological questions, supply what analytical and synthetic methods can never furnish. Something more than this may be implied in Ostwald's generalisation. A reaction is accelerated by a rise of tempera- ture because the activity of molecular movements is increased as the temperature is raised, and the increased activity of molecular movements brings about more frequent encounters of the react- ing substances. The more frequiently the reacting molecules encounter one another, the greater the chance of their reacting on one another. But the frequency with which such encounters take place may also be increased by increasing the number of molecules in a given space. If two solutions completely mixed react with a certain velocity on each other when the solution is homogeneous, of equal concentration in all its parts, the addition of a third substance, if it affects the homogeneity of the solution, must increase the concentration of dissolved molecules in those parts of the system into which they congregate. The velocity of the reaction will be determined by the concentration of the most concentrated portions of the solution, provided the rate of diffusion is not too slow. Solutions in which fine suspended particles are present are not homogeneous. Owing to adsorp- tion, the concentration of dissolved molecules is different near the surface of the suspended particles from what it is elsewhere. Some instances of catalytic action appear to be due to the dis- turbance of the homogeneity of solutions by adsorption. Platinum decomposes hydrogen peroxide, and the amount of hydrogen peroxide decomposed varies, not with the mass of the platinum, .but with the amount of platinum surface. Platinum black produces far more change than platinum foil, and far more active than platinum black are the so-called colloidal solutions of platinum obtained by setting up an electric arc between I.] ADSORPTION AND CATALYSIS 19 platinum electrodes under water. If the catalytic action of platinum is due to the peculiar adsorptive properties of platinum, as Faraday first suggested, that of other catalytic agents may also be due to surface properties ; and enzymes in particular, in virtue of the fact that they are colloidal substances, may disturb the homogeneity of solutions in which they are suspended, and owe their properties in part to the alteration of the concentration of the solution in the neighbourhood of their surfaces. Whether the whole of enzyme action is to be accounted for by adsorption phenomena, it is not necessary for our argument in this connection to discuss. There are other conceptions which physical chemistry is at present making use of in order to interpret the phenomena of catalysis ; but in these conceptions, too, surface phenomena play a large part, and so they apply to the colloidal enzymes no less than to metallic ca,talysts. For our present purposes it is sufiRcient to indicate the directions in which we may look for illumination of the mysteries of meta- bolic processes. Physical chemistry applied to biological problems promises to supplement the analytical and synthetic work of the past twenty-five years, and to do as much for the elucidation of those problems as organic chemistry has done and is doing to determine the nature of the material with which in the physiology of metabolism we are concerned. Physical chemistry is engaged in the study of the nature of colloidal solutions and their properties, and also in that of the phenomena of catalysis and their interpretation. Till we have learnt all that can be taught us by the results of this study, it is undesir- able and unnecessary that we should hopelessly resign ourselves to the belief that the puzzles of animal metabolism must for ever remain unsolved. LECTURE II THE ASSIMILATION AND SYNTHESIS OF CARBOHYDRATES One of the most interesting of all the chemical reactions that have been studied in biology is that by which the synthesis of carbohydrates is carried out in plants from the carbon dioxide in the air.. In this synthesis, which is the work of the pigment chlorophyll, the first stage must necessarily be a reduction. The suggestion, made by Bayer many years ago, that formic aldehyde was the product formed in this reduction process, and that the carbohydrate synthesis took place by condensation of a number of molecules of formic aldehyde into one large molecule, has never been completely proved to be correct. But the condensation which the theory supposes to take place is a general reaction with aldehydes, known as the aldol condensation, from the product of condensation of two molecules of acetic aldehyde, aldol — /.O .OH .0 \h ' \h ' \h And it is well known that substances with the constitution and properties of sugar have actually been synthesised m vitro from formic aldehyde by the help of the condensing action of alkalies. Six molecules of formic aldehyde condensing by aldol con- densation — ^O /OH /OH /OH /OH /OH /D 6H.Cf^ = H.C< . C< . C< . C< . C/ . QZ \H \H \H \h \h \H \h — would clearly produce a hexose sugar. 20 LECT. ii] SYNTHESIS OF SUGAR IN ANIMALS 21 That the carbohydrates, free or combined, found in animals are ever formed by such a fundamental synthesis as this, has never been contended. Sugar and compounds such as starch, from which sugar is easily formed by changes that are among the most familiar in physiological chemistry, are so abundant in the food of animals, that the bulk at any rate of carbohydrates found in animals may certainly be traced to the substances of this nature taken into the body ready-made in the food. But it should be remembered that, if the course of the reaction by which plants synthesise carbohydrate from the carbonic acid of the air really corresponds to the scheme propounded by Bayer, the two stages in this synthesis, the reduction and the condensa- tion, are of very different character and significance. Aldol condensation is an exothermic, almost an isothermic reaction — 2 g-. mol Acetic aldehyde \ I ^ S- ™ol- Aldol = S5i cal. / \ = 546.8 cal. But extrinsic sources of energy are necessary for the conversion of carbonic acid into formic aldehyde, and that stage of the reaction is from all points of view a most exceptional one. But, given the formic aldehyde, all that follows is comparatively comprehensible and commonplace. And if in animal physiology we are driven to the conclusion that a true synthesis of carbo- hydrate does occur in the animal body, there is no peculiar difficulty in supposing that it might occur by the same kind of condensation as Bayer has hypothesised for the synthesis of starch in plants. The carbohydrate which is found free and uncombined in the animal body is in one or other of three forms : glucose, lactose, glycogen. Glucose, the most widely distributed of all the sugars in Nature, we may for the moment suppose to be provided in sufficient quantity in the food, either as free glucose, or in the form of disaccharides or starch, which in digestion are hydrolysed, and so give rise to glucose. But glycogen, a polymeric anhydride of glucose found only in animals, must be synthetically produced, even in such animals as take only animal food, since even the glycogen taken as food is hydrolysed 22 ASSIMILATION OF CARBOHYDRATES [i-ect. in digestion. It is well known that animals fed on glucose or starch are found to contain much more glycogen than animals that are given no carbohydrate in their food, and the condensa- tion of glucose to form glycogen is the most familiar of all the carbohydrate syntheses that take place in the animal body, and one that is admitted on all hands. The strict proof of its occurrence was furnished by the experiments of Otto, in Voit's laboratory, on fowls and rabbits. A rabbit was kept without food for four days, then given 80 g. of glucose, and eight hours later killed. In the liver alone more than 9 g. of glycogen was found, or nearly 17 per cent, of the whole weight of the liver, and nearly as much more in the rest of the body. The largest amount of glycogen that has been recorded in the liver of rabbits starved for a similar period is less than half a gramme for animals of the weight of the one in Otto's experiment.^ The liver alone, therefore, in this case contained nearly 9 g. of glycogen more than could be expected, unless the animal could make glycogen from glucose. And that the glycogen was not derived from proteid, rendered available for this purpose by the large supply of sugar, was quite clear ; for the nitrogen excreted during the experiment corresponded only to about 5 g. of proteid, and it is impossible to obtain from 5 g. of proteid 9 g. of glycogen under any hypothesis.^ How this synthesis is brought about we do not know. The fact that the reversed change, from glycogen to glucose, is carried out by an enzyme or enzymes obtained from the same organ, the liver, suggests that it may be a case of reversible enzyme action. The enzymes that hydrolyse glycogen which are found in the saliva and pancreatic juice take the hydrolysis as far only as maltose, and the conversion of maltose into glucose is effected in digestion by a different enzyme, maltase. But the two stages into which the conversion of glycogen into glucose can under these conditions be resolved involve, so far as we know, very closely related reactions, and one of these, the action of maltase, is of all enzyme actions the one in which ' Kiilz, Beitrcige zur Kenntniss lies Glycogens, Marburg, 1891. 2 YdA, Z.f. B. 28, 245, 1892. II.] FORMATION OF GLYCOGEN 23 reversibility has been most completely established.^ If' this reversibility of enzyme action is to be extended to the action of other enzymes than those for which it has been proved, then "one of the first reactions to be included under such a generalisation would naturally be that with which we are now concerned. The enzyme that hydrolyses glycogen, we must in that case suppose, does not do so completely. Its action comes to a stop at the equilibrium point, when a certain small portion of the glycogen is left unattacked. If less than this small amount is present within the sphere of action of the enzyme, then some of the glucose must be converted into glycogen. And if the glycogen as soon as it is formed is removed from the sphere of action of the enzyme, either by entering at once into chemical combina- tions, in which it is no longer subject to the operation of the enzyme, or in any other way, then so long as this is the case the enzyme would continue to produce glycogen, and no glycogen would be hydrolysed. As soon as the -glycogen, on the other hand, was set at liberty again and permitted to come under the influence of the enzyme, the opposite change would be brought about, glucose would be formed, and the more rapidly the glucose was carried away by the blood, or otherwise placed out of reach of the enzyme, the more rapidly would fresh glucose be formed. Knowing as we do that not only is glycogen formed from glucose in the liver, but glucose is also formed from glycogen, and that this latter is the work of an enzyme that can be demonstrated in an extract of the organ, it is specially tempting to imagine that this is an instance of reversible zymolysis ; but we have no direct evidence that it is so. In this synthesis of glycogen, which consists in the formation of a polymeric anhydride of glucose, it is difficult to see how anything can take the place of the glucose, unless it is by being first converted into glucose. If, therefore, we have reason to believe that other substances of any kind besides glucose can give rise in the body to the formation of glycogen, the pre- sumption is that they do so because they can be converted into glucose. We may leave, therefore, the question of the formation 1 A. Croft Hill,/. C 5. 73, 634, 1898. . ' ; 24 ASSIMILATION OF CARBOHYDRATES [lect. of glycogen from substances other than glucose, till we are led back to it again by the consideration of the substances from which glucose can be derived. The other carbohydrate occurring free and uncombined in the animal body is lactose. This is a disaccharide differing from -maltose in that it is compounded, not of two molecules of glucose, but of a glucose molecule and . a molecule of galactose. The difference between galactose and glucose is a stereochemical one, consisting in the different spatial configurations of certain parts of the molecule. The formulae by which these differences are represented are : OH H OH OH COH.C . C . C . C . CHgOH for Glucose H OH H H and OH H H OH COH.C . C . C . C . CHgOH for Galactose. H OH OH H If the carbon atoms in these formulae are denoted by the numbers i to 6, beginning with the carbon atom in the aldehyde group at the left-hand end, as given here, then the difference between galactose and glucose consists only in the transposition of the hydrogen and hydroxyl groups connected with the fourth carbon atom, Now if galactose as well as glucose is present in the food of animals during the secretion of milk, it is possible that the synthesis of lactose is merely another case of the condensation of sugars to "form anhydrides, very similar to the condensation of glucose to form glycogen. Galactose is fairly widely dis- tributed in nature. It occurs principally in the form of poly- saccharides, compounded of galactose and a pentose, either arabinose or xylose. Galacto-arabane is found in the seeds of beans, peas, vetches, and cresses, and the young plants of clover and lucerne : galacto-xylane is found in the cereals wheat and barley. It is possible, therefore, that a part of the galactose in the sugar of milk is derived without change from the galactose in the food of some animals, A cow yielding 2 to 3 II.] ORIGIN OF GALACTOSE 25 gallons, 20 to 30 lbs. of milk, turns out daily about i lb. of sugar, half of which is galactose. It is not known to what extent the galacto-pentanes are digested, and the galactose set free from them absorbed. But it is not necessary to suppose that any of the galactose in the sugar of milk is derived directly from this same sugar in the food. For it is known that carnivorous animals secrete undiminished quantities of lactose when kept on a pure meat diet,i so that the power of making galactose must be present at least in these animals : and it is not likely to be confined to them. It is most natural to suppose that this galactose is formed from glucose by the transposition of a hydrogen atom and hydroxl group, as indicated in the stereochemical structural formulae given above — such isomeric transformations among the sugars were the subject of the important studies of Lobry de Bruyn and van Ekenstein.^ They showed that glucose exposed to the action of bases in weak solutions gradually lost its rotatory power : for instance, if 20 g. of glucose in 500 c.c. of water containing 10 c.c. of normal potash was heated to 63° C. for 2| hours, the rotation sank from + 5° 30' to + 10' : and they proved that this change was due to the conversion of part of the glucose into fructose (Ia;vulose) and mannose, both of which they isolated and identified. Now the differences between these three sugars are confined to the first two carbon atoms in the formulae, written as above for glucose and galactose, thus : .OH .H COH.C . C — = Glucose. \h ^OH /H /H COH.C . C _ = Mannose. ^OH \0H /H CHjOH . CO . C — = Fructose or Laevulose. \0H ' Szubotin, V. A., p. 561, 1866. - Rfcueil de$ travaux chimiques des Pays Bos, xiv. and xvi., 1895 and 1897. 26 ASSIMILATION OF CARBOHYDRATES [lect. The Dutch chemists originally suggested that the inter- changeability of these three forms of sugar could be best explained by supposing that the aldehyde group took the hydrated form, and from this a hypothetical anhydride was formed which was a link between glucose and fructose, thus : . .OH .OH (i) HO . C — C — = Glucose hydrate. \h \h (2) HO . C — C — = hypothetical Anhydride \tt \tt of glucose hydrate. . /« (3) HO . C — CO — = Fructose, \h These changes being reversible, there would tend to be formed from fructose (3), in addition to the anhydride (2), another stereoisomeric anhydride, in which the oxygen atom linking the two carbon atoms took up position on the other side of the molecule, thus : ^H ^H (4) HO . C — C — = hypothetical Anhydride \q/ of mannose hydrate. (5) HO . C . C — = Mannose hydrate. \0H \0H And so fructose could be converted into either mannose or glucose, and glucose and mannose into each other through fructose. The same chemists, however, subsequently found that in addition to fructose, other ketone sugars are formed at the same time, and for these the theory they had given did not provide. But it does not follow that the theory is untenable .,.] ISOMERIC TRANSFORMATIONS 27 for' the changes in which fructose is concerned. There is reason for thinking that glucose also tends to assume the form in which the iirst and fourth carbon atoms are united to the same oxygen atom instead of the iirst and second, as supposed in the theory of Lobry de Bruyn (formula (2) supra). The methyl glucoside formed by heating glucose with methyl alcohol in the presence of hydrochloric acid has lost the power of reducing copper oxide, and therefore has no longer the aldehyde group, and occurs in two stereoisomeric forms, owing to the asymmetric properties acquired by the carbon atom of the aldehyde group. The formula ascribed to it by Fischer is: O (CH3.O or)H (H or) O . CHg It is clear that such a formula for glucose derivatives suggests that glucose and galactose may be convertible into each other by a reaction similar to the one proposed by Lobry de Bruyn, to explain the conversion of mannose into glucose. It is true that this chemist expressly investigated this question, and found that potash does not give rise to the formation of galactose from glucose. But in such points the action of different bases may well differ, just as the change from mannonic acid to gluconic acid is not effected by inorganic bases, but is by pyridine and quinoline. At any rate, in the animal body the isomeric transformation of glucose into galactose seems to occur, unless we prefer to think that the galactose is derived from substances less closely related to it ; and if it occurs, the problem as to how it is brought about has to be solved. And this is not the only case of isomeric transformation among the sugars which there is reason for believing to be brought about in animal chemistry. Laevulose certainly, and mannose almost certainly, as well as probably galactose itself, 28 ASSIMILATION OF CARBOHYDRATES [lect. have been shown to give rise to the formation of glycogen in the liver in the same way as glucose. It is also known that the glycogen so formed is the same glycogen that is formed from glucose, so that it is clear that the glycogen cannot be formed till the transformation of these sugars into glucose has taken place. In addition to the carbohydrates that occur free in the body, there are others that are found in combination with other groups of different kinds. The best known of these are the carbohydrate combinations in nucleic acids, in certain fatty acid compounds, and in certain proteids. In these cases, though it is true little is known as to the exact nature and constitution of these compounds, the nature of the sugar has been determined, and in each instance presents points of interest. Every addition to what is known of thfe chemical nature and properties of. nucleic acids must be of importance. Nucleic acids and protamines together make up almost the whole of the functionally essential part of the spermatozoa. Nucleic acids in combination with some form of proteid, in the nucleus of every cell, may fairly be looked to as the key to the mysteries of the chemistry of living matter. Not the least interesting of the many striking facts that have in recent years been elicited with regard to their nature and constitution, is that some form of carbohydrate is always to be recognised in their molecules. In all those of animal origin in which the nature of this carbohydrate has been determined, it has proved to be a pentose. Hammarsten was the first to suspect the presence of a five-carbon sugar in a nucleo-proteid ; ^ he was able to isolate, in the form of a syrup, a substance, obtained by boiling the nucleo-proteids of the pancreas with weak sulphuric acid, and this syrup gave reactions characteristic of pentoses. The osazone melted at about 1 50° C, the substance itself gave the phloroglucine reaction, and on distillation with hydrochloric acid, furfurol was found in the distillate. But since these reactions might have been accounted for by ' Hammarsten, ^.-5. Z. 19, 27, 1893. 11.] ORIGIN OF NUCLEIN PENTOSE 29 glycuronic acid, he was unable to say positively that the substance obtained was a pentose. Salkowski proved that it was, by analysing the osazone.^ It was then shown by Bang that the pentose was a component of the nucleic acid, and not of the proteid part of the molecule.^ And the exact nature of the sugar was determined by Neuberg,^ who showed that it was /-xylose, the sugar, that is, with the constitution represented by the formula — OH H OH COH . C C C CHoOH \ \ \ H OH H A pentose has been obtained, too, from the nucleic acids found in the liver, kidney, spleen, thymus, thyroid, submaxillary gland, brain, and muscle ; and in the case of that from the liver, it has been proved that it too is /-xylose.* In the nucleic acid from the lymph glands Bang was unable to find a pentose.^ Pentoses of all kinds appear to be but very imperfectly assimilated when taken with the food. Xylose itself under these conditions is excreted unchanged in the urine in amounts up to 55 per cent, of that taken; the xylose, therefore, in the nucleic acids is in all probability formed in the body. From the formulae for xylose and glucose it is clear that the structure of the two sugars is identical as far as the fourth carbon atom, and that the formula of xylose may be obtained by leaving out the fifth carbon atom in the formula for glucose. Xylose can be obtained from glucose : by oxidation of the terminal secondary alcohol group in glucose, glycuronic acid is formed ; and Salkowski and Neuberg showed that by the action of certain bacteria on glycuronic acid carbonic acid is split off and the sugar /-xylose formed.® Now it is almost certain that ' Salkowski, B. k. W., No. 17, p. 361, 1895 ; and, B.-S. Z. 27, p. 537, 1899. ^ Bang, H.-S. Z. 26, 145, 1898. ^ Neuberg, B. 35, 1467, 1902. * Wohlgemuth, H.-S. Z. 37, 475, 1903. ^ Bang, H. B. 4, 124, 1903. ^ Salkowski and Neuberg, H.-S. Z. 36, 261, 1902. 30 ASSIMILATION OF CARBOHYDRATES [lect. glycuronic acid is formed in the animal body from glucose. It is known that camphor is excreted in the urine, combined with glycuronic acid. P. Mayer found that a rabbit, when in addition to its ordinary food it was given 2 g. of camphor, excreted 2.18 g. of glycuronic acid ; if then kept without food for nine days, so as to reduce the stock of carbohydrate in its body, the same dose of camphor caused only 1.2 g. of glycuronic acid to appear in the urine : two days later the same animal, still with- out food, when given together with the same dose of camphor 10 g. of glucose, then excreted 2.06 g. of the acid.^ This points clearly to glucose being the source from which glycuronic acid is formed, in the rabbit at any rate. In small amounts, glycuronic acid is found constantly in normal urine. If, therefore, the reaction by which glycuronic acid gives up carbonic acid and becomes xylose is a general reaction in living organisms, and not a peculiarity of bacterial chemistry, it is possible to account for the xylose necessary for the synthesis of nucleic acids by the metabolism of glucose.^ The exact constitution of nucleic acids and the relationship of the pentose to all the different parts of the molecule, is not known. But the nucleic acids do not reduce; so that they are of the nature of glucosides, and the aldehyde group is effaced in combining, probably with an alcoholic hydroxyl in the glycerine phosphoric acid. This union is not dissolved by trypsine, but in the combined action of the enzymes of the .pancreatic cells, in autolysis of the pancreas, the pentose is set free, and can be separated, as the phenylosazone.^ The combinations in which carbohydrates are found associated with higher fatty acids are in part at any rate like the nucleic acids, in that their exact constitution has not yet been determined, although the nature of the sugar has been determined. Cerebrines, or cerebrosides as they have been called, to indicate the presence of a sugar in their composition, 1 P. Mayer, Z.f. k. M. 47, 68, 1902. 2 Cf. Salkowski and Neuberg, loc. cit. 3 Cf. Neuberg and Milchner, B. k. W., p. 1081, Oct. 10, 1904 ; cf. also, Bang, H.-S. Z. 31, 425, 1900 ; and, Burian, Ergeb. 3, 87, 1904 j and, B. 37, 696. 11.] COMPOUNDS OF FAT AND SUGAR 31 have been obtained in a crystalline form from the brain. The\ sugar that is set free on hydrolysis of these cerebrines has been shown to be galactose,^ while the rest of the molecule is probably composed of higher fatty acid radicals, or simple nitrogenous derivatives of these.^ In addition to the cerebrine compounds of galactose, higher fatty acids compounded with sugar were found in the substance jecorine, which Drechsel first obtained from the liver, and others after him in the blood and elsewhere. The nature of this too has not yet been exactly defined. \ Drechsel ' described it as a compound of sugar with lecithine and j some other group or groups, in which sulphur was present, and / characterised it by its intensely hygroscopic properties, its ' solubility in moist ether, and insolubility in alcohol. Combina- i tions of lecithine with sugar have been described as jecorine by Bing and others, who have not exactly followed Drechsel's \ methods of preparation.* . But beyond the fact that lecithine j may cause sugar to dissolve in ether, and sugar may cause i lecithine to dissolve in aqueous fluids, such as the blood, and 1 the fact that such combinations of sugar and lecithine are ) to be found in the yolk of eggs, and in the blood and certain organs, nothing definite is known of them. But jecorine and the cerebines alike serve to show the complexity of the relation- ships into which the simple sugars may enter in the animal body. Mayer, who has shown that glycuronic acid is normally present in the blood, is inclined to think that jecorine is a compound of glycuronic acid and lecithine; but this has not yet been demonstrated. With regard to the combinations into which carbohydrates enter with proteids, there is a great deal that is still not clear. But two facts at any rate stand out from much that is obscure : in the first place, that the number of proteids into the composi- tion of which recognisable derivatives of sugar enter are much more numerous than was generally believed a few years ago ; 1 Thierfelder, U.S. Z. 14, 209, 1890. 2 Bethe, S. A. 48, 73, 1902. ^ Drechsel, Ber. der Sachs. ges. der Wissensch., 1866, p. 44 ; and,y. Pr. Ch. 33. 425- * Bing, Sk. A. 9, 336, 1899 ; cf., too, Mayer, H.-S. Z. 32, 530, 1901. 32 ASSIMILATION OF CARBOHYDRATES [lect. and, secondly, that in a majority of cases at any rate the sugar has been found to be present in the form of glucosamine. It has long been known that a certain class of proteids when boiled with weak mineral acids give solutions that reduce Fehling's solution. The reduction is due to the presence of some substance closely related to sugar ; for from the solution a phenylosazone can be obtained. Proteids of this class were known, therefore, as the glyco-proteids, mucin being the most familiar of the class. From mucin and the mucoid proteids of egg-white and of ovarian tumours glucosamine has been definitely isolated. But in addition to these glyco-proteids many other proteids, not previously reckoned in this class, have been suspected of containing sugar radicals, and in some cases proved to contain glucosamine. So typical a "native albumin" as egg-albumin was the first to be shown to share this property with the glyco-proteids. The proteid of the yolk of eggs has also been found to contain glucosamine,^ and so too crystalline serum-albumin, from human as well as from horses' blood,^ and serum-globulin.' The amount of glucosamine that can be obtained from these different proteids is often very small, but in some cases is very considerable ; in the case of mucin and ovomucoid, more than 30 per cent, but in serum-globulin not more than 2 per cent, and still less in serum-albumin, while egg-albumin takes an intermediate position with 10, or perhaps nearer 15, per cent* Glucosamine has been shown by Fischer to be really an amido-glucose, so that the old name chitosamine is to be dropped. The constitution is expressed in the formula — /H /OH /OH COH . CH . NHj . C< . C< . C< . CH,OH \0H \h /OH K • \H 1 Neuberg, B. 34, 3963, 1901. ^ Langstein, B. B. i, 259, 1902. ^ Langstein, Ergeb. 3, 463, 1904. * Hofmeister, H.-S. Z. 24, 1 59 ; Langstein, Ergeb. i., 94, and iii., 463 ; and, H.-S. Z. 31, 491, 1900. 11.] SUGAR RADICALS IN PROTEIDS 33 the exact disposition of the groups connected with the second carbon atom being as yet undetermined.^ Evidence of the presence of sugar radicals in proteids, less convincing but easier to obtain than that given by the isolation and identification of glucosamine, or its oxidation product norisosaccharic acid,^ is afforded by certain colour reactions. The purple colour given by sugar with a-naphthol and sul- phuric acid, due to the formation of furfurol (Molisch's reaction), is also given by a large number of proteids, and the isolation of sugar derivatives from some of these has led to the inference that all proteids which give this reaction have a sugar group in their molecules. The test is extremely delicate ; it is given by as little as 0.5 mg. of cellulose, for instance, so that its very sensitiveness raises suspicions as to the legitimacy of such an inference. But even so there are some proteids that do not even by this test show signs of a carbohydrate group, notably casein and gelatine. Another reaction recommended for the testing for sugar groups in proteid molecules is the modified orcine test of Bial.^ The fluid to be tested is heated with two volumes of a solution of orcine in strong hydrochloric acid, as in the test for pentoses or glycuronic acid ; only a drop of ferric chloride solution is added. Dextrose and galactose, or disaccharides into which these sugars are compounded, give a bluish-green precipitate. A solution of this precipitate in amyl-alcohol gives a character- istic spectrum, in which the yellow and the edge of the green is blotted out. Lsevulose and glucosamine do not give the reaction. But certain proteids do ; egg-albumin and the proteid of the yolk of eggs, serum-albumin and serum-globulin, but not casein or pseudo-mucin. Since glucosamine does not react in this way, it is argued, there must be another carbo- hydrate group in addition to glucosamine in these proteids. And in some instances carbohydrate derivatives more complex than glucosamine have been obtained from proteids. Frankel ' Fischer and Leuchs, B. 36, 24, 1903. 2 Neuberg and Wolff, B. 34, 3840, 1901. 3 Bial, Z.f. k. M, 50, 417 ; M. /., 102, 1903. 34 ASSIMILATION OF CARBOHYDRATES [iect. described a compound carbohydrate containing nitrogen which he prepared from egg-albumin, and took for a diamino-disac- charide ; ^ and Langstein has obtained a polyhexosamine from serum-globulin in small quantities.^ Chondrosine was the first of all the compound carbohydrate derivatives to be obtained from any proteid, and Schmiedeberg ascribed to it the con- stitution of a compound of glucosamine with glycuronic acid, coupled by condensation of the aldehyde group of the latter with the amine group of the former. Orgler and Neuberg have prepared, also from cartilage, a substance of much higher mole- cular weight than corresponds to this formula, from which they failed to obtain either glucosamine or glycuronic acid. They found instead a tetra-oxyamido-caproic acid and some carbo- hydrate substance, the nature of which they did not determine.^ Besides these there are certain compounds of the nature, of albumoses, into the composition of which sugar enters, that have been described by Pick * and by Simon.^ It has been very commonly assumed that these sugar groups in proteid molecules are built up into the molecules as in gluco- sides. As we have seen, Neuberg and Milchner * think that this mode of union holds for the pentose in nucleic acid, but they argue for a different one for the glucosamine in proteids. The formation of a glucoside requires a hydroxyl or sulph-hydro group in an organic compound, in addition to the sugar. Of such groups in proteids, the only ones known are in tyrosine, cysteine, oxyproline, and certain other oxyamido acids. These oxy-acids are, however, present only in very small amounts: Cysteine does not occur as such, but as cystine, in which the — -SH group no longer exists ; and since the glyco-proteids give Millon's reaction, the phenol group in tyrosine cannot be used for glucoside formation. On these and other grounds, they 1 S. Frankel, M.f. Ch., 19, 819 ; M.J., p. 23, 1898. 2 L. Langstein, M. f. Ch. 24, 445 ; M. /., p. 30, 1903 ; and, Ergeb. iii., 460, 1904 ; cf., too, Leathes, 5. A. 43, 245, 1899. ^ Orgler and Neuberg, H.-S. Z. 37, 407, 1903. * Pick, H. B. ii., 481, 1902. ' Simon, S. A, 49, 457, 1903. * Neuberg and Milchner, B. k. W., p. ig8i, Oct. io, 1904 ; c/. sup. p. 30. II.] SUGAR SYNTHESIS IN GLYCOSURIA 35 think that it must be through the — NHj group that glucosamine is linked on to the proteid molecule. It is clear that there remains still very much to be learnt concerning the combinations into which glucosamine or other carbohydrate derivatives enter with proteids, and the manner in which the union is effected. If it is true that serum-albumin contains less than i per cent, of glucosamine, since glucosamine has the molecular weight 179, the molecular weight of the albumin comes out very high, over 1 8,000 ; and if the carbo- hydrate in the globulin of blood is, as Langstein thinks, a poly- hexosamine, and there is less than i per cent, of it,^ the molecular weight for globulin must come out considerably higher still, .something like 50,000. On the other hand, mucin containing 35 per cent, must either have a low molecular weight, or else present a very considerable number of groups into which the glucosamine can be welded. There are special reasons why it is important that we should get clearer conceptions as to the significance and prevalence of carbohydrate radicals in proteids. In the ordinary conditions of animal, and particularly of human life, starch and sugar enter so largely into the food on which that life is sustained, that it may seem unwarrantable to suppose that the carbohydrates of the body are derived from anything but the carbohydrates of the food. And yet the evidence is overwhelming that sugar can be formed in our bodies from substances which are not in the least related in the ordinary chemical sense to the sugars. It is well known that the stock of carbohydrate in the body is never large. Animals have a large amount of proteid in their blood and tissues, much of which can, if necessary, be made use of as a source of energy, both heat and work. The 5 or 6 litres of blood in a man's body contain about 18 to 20 per cent, of proteids, or about a kilogramme in all. The muscles contain about 20 per cent, of proteids ; and since they make up some 40 per cent, of the whole body-weight, their proteids amount to 8 per cent, of it, or nearly 6 kilo- grammes. So too with the fats, though they constitute a very * Langstein, M.f. Ck. 26, 531, 1905. 36 SYNTHESIS OF CARBOHYDRATES [i-ect. variable fraction of the whole weight of the body, nevertheless in normally nourished individuals they are hardly less abundant than the proteids. A moderately fat dog was found to have 26 per cent, of fat;^ and as much as 45 per cent, has been recorded in a dog. But with the carbohydrates the figures are very different from these. The sugar that can be detected in the blood in health at the most amounts to 0.2 per cent, or in all some 10 or 12 g. In the liver it is unusual to find more than 10 per cent, of glycogen, and more than 20 per cent, has not been ever observed ; in all, therefore, there may be about 200 to 400 g. at the outside in the liver. In the muscles 0.5 to I per cent, is probably not too low a figure at any rate, and that would give another 150 to 300 g. Exact determina- tions of the amount of glycogen in the human body have not been made, but it is seldom likely to be more than a kilo- gramme. In seven dogs, specially fed up in Pfliiger's laboratory in order to determine the maximum amount of glycogen that may occur in these animals, the average reckoned on the body- weight was about 2 per cent., but in one case nearly 4 per cent. These are certainly extreme figures. And yet it is on record that a diabetic patient on a strictly limited diet excreted no less than 11 50 g. of sugar daily.^ And clinical observers throughout the world are in agreement on this, that in the worst cases of diabetes, kept on the strictest diet, whatever it is that gives rise to the sugar that is excreted it cannot be only the carbohydrates of the body. Sugar must be synthe- sised under such circumstances, often on a very large scale. Experimental pathology confirms in this the conclusions of the clinical pathologists. Most of those who have studied the glycosuria set up by the administration of phlorrhizine have been led to the belief that the sugar is often excreted in such quantities that . it cannot be accounted for by the carbohydrate of the body. And in the glycosuria following removal of the pancreas in dogs it has been proved irrefutably that this is so. Liithje kept a dog weighing 18 kg. without food for five days, ' Mockel, Pfl. A. 108, 189, 1905. ^ Niccolini, B. Cbl. ii., p. 229, 19041 II.] PROTEIDS AS A SOURCE OF GLYCOGEN 37 and then excised its pancreas. During the next fourteen days the dog, still without food, excreted 228 g. of sugar ; for the teri subsequent days it had a diet consisting solely of proteid, a preparation of caseine, nutrose, free from carbohydrate of any kind, and during that time excreted 975 g. of sugar, After this, with no food at all for a week, it excreted 1 50 g. more. Altogether, therefore, it excreted 1350 g. of sugar without having any sugar provided, and as the experiment did not begin till the animal had been without food for five days, its body must have been but poorly stocked with glycogen to start with! This quantity of sugar, amounting to 7.5 per cent, of the animal's weight, is nearly twice as great as the highest amount of carbohydrate ever found under exceptional circum- stances in a dog, namely, 4 per cent, of the body-weight ; so that at least one-half of it must certainly have been made by synthetic processes in the body.^ Pfliiger excised the pancreas from a dog weighing at first 12 kg. In two months, 3097 g. of sugar were passed in the urine, whilst the dog's food contained no carbohydrate in any form, and practically nothing but pro- teid. Synthetic processes must in any case have produced as much as 2.5 kg. of sugar in this experiment.^ It is not only the sugar excreted in disease that may be the product of synthetic changes set up in other substances in the animal body. The glycogen of the liver and of the muscles is also most probably in part synthesised from proteid ; this was first taught by CI. Bernard, and subsequently by Kulz, Naunyn, V. Mering, and others, who took greater precautions to eliminate sources of error. It is true that the results of none of these experiments are as conclusive as those referred to above on the excretion of sugar. In the first place, the glycogen in an animal's liver cannot be determined more than once. In order, therefore, to decide whether there is more glycogen than there would have been if the treatment had been different, estimations have to be made in control animals. But animals treated and fed in exactly the same way, however similar they may be, are 1 Liithje, D. A. f. k. M. 79, 1904. 2 Pfliiger, Pfl. A. 108, 115, 1905. 38 SYNTHESIS OF CARBOHYDRATES [lect. found to have very widely different amounts of glycogen in their livers. So that the conclusions drawn from control animals, though they may be probably correct when a sufficient number of controls is taken, cannot be regarded even then as positively certain. Neither is it possible to be certain that any particular period of starvation will cause the glycogen com- pletely to disappear ; so that if, after a period of starvation followed by feeding with any particular food, it is found that glycogen is present in the liver, it is impossible to say that the glycogen has been formed from that food : it may be probable, but it cannot be certain. Pfliiger found in a dog that weighed 33 kg. nearly 50 g. of glycogen after starving for twenty-eight days, arid there was nearly 5 per cent, of glycogen in its liver. Such a case is exceptional, but it must be taken into account. But whether we are convinced by the experiments which Kiilz and others have advanced in proof of the formation of hepatic glycogen from proteid food, or whether with Pfliiger we insist on proof of this, from which there is no imaginable loophole of escape, and refuse absolutely to take account of probabilities, the fact is established that the synthesis of sugar can be carried out in the animal body from something which is not sugar. It has been established in diabetes and in the glycosuria following excision of the pancreas in dogs. It is therefore one of the cheniical reactions of animal metabolism, and it is improbable that this reaction is peculiar to the pathological conditions in which it has been possible to prove it. The material used for this synthesis must be derived either from proteid or from fat. Some of the results referred to above seem to point clearly to proteid as the source of this material. Liithje's dog, that after removal of its pancreas excreted 228 g. of sugar in the first fourteen days, during which it had no food, excreted 975 g. in the next ten days, when it had a proteid diet free from carbohydrate and fat ; that is, a daily average output of sugar eight times as great as before. In many cases of diabetes, it has been pointed out repeatedly, the output of sugar varies with the amount of proteid in the food. The experiments of Bernard, Kulz, Naunyn, and others, on the formation of glycogen II.] The ratio D:N 39 from proteid food may be inconclusive in the strict sense, but they establish some degree of probability that proteid can normally furnish material for carbohydrate synthesis. Min- kowski tried to identify the source of the sugar by seeing if there was any proportionality between the amount of sugar excreted and the amount of proteid broken down in the body, the nitrogen of which was cast off in the urine. From loo g. of proteid containing 17 g. of nitrogen, urea containing about 7 g. of carbon would be formed. The rest of the carbon in the proteid, not accounted for in the urea, would amount to about 45 g. Glucose contains 40 per cent, of carbon ; so that, if the whole of the 45 g. of carbon were converted into sugar, 112 g. of sugar could be formed from the 100 g. of proteid ; and if this were all excreted, the ratio between sugar excreted and nitrogen excreted rf, would be , or nearly 7. He found the ratio in dogs after N 17' -^ excision of the pancreas to be fairly constant at about 2.8.^ A constant value for this ratio has been found also by Lusk in animals excreting sugar under the influence of phlorrhizine, in dogs about 3.7, but in other animals, rabbits, goats, and cats, about the same as that found by Minkowski in his experiments. In diabetes, however, the ratio may have any value, up to even 1 1.8.^ Such a high value as this, if it is right to suppose that the nitrogen is excreted as rapidly as the sugar — an assumption that has been questioned^ — would point to the possibility of sugar being derived from the products of fat metabolism. Leaving, however, on one side for the present the question whether the synthesised sugar may be derived from substances formed from fat, and assuming that it may be derived from substances formed from proteid, it is at once obvious that what we have ' Minkowski, S.A. 31, 85, 1892. 2 Rumpf, B. k. W., No. 9, 1 899. In fifteen days, 1 1 70 g. of sugar and 98.8 g. nitrogen were excreted on a diet strictly confined to fat and proteid. Hartogh and Schumm, S.A. 45, 31, observed in a dog. treated with large doses of phlorrhizine, on a diet rich in fat but practically free from carbohydrate, the ratio -^ = 9 maintained for four days, and on one day as high as 13. 3 O. Loewi, S.A. 47, 68, 1901. ; . 40 SYNTHESIS OF CARBOHYDRATES [lect. learnt in recent years about the presence of sugar or simple derivatives of sugar in proteid molecules, raises a very important question. When glycogen or glucose is formed from proteids in the body, is this a fundamental synthesis of sugar from carbon and hydrogen in simpler combinations which are set free when proteid molecules go to pieces, or is it merely that ready-made sugar groups are picked out of these molecules as such, preserved from the general breakdown, and synthesised into glycogen or excreted as sugar ? If this latter conception of the change can be admitted, then the problem presented is a comparatively simple one. When once the carbohydrate is set free from the proteid, the formation of sugar or glycogen from proteid involves changes no more complicated than those by which one carbo- hydrate is in the body convertible into others. In order to decide this, we must know what the amount of carbohydrate in proteids is. The isolation of carbohydrates or their derivatives from the debris of demolished proteid molecules is not easy, and a quantitative isolation is not possible. It is possible to estimate the carbohydrate from the reducing power of the fluid obtained on hydrolysis of the proteid ; but we have no means of knowing that such estimations are reliable ; for all that reduces need not all be sugar, and some sugar may have been destroyed during the operation of hydrolysis. But, estimated in this way, the glucosamine in certain typical glyco- proteids, ovomucoid and the mucin in bronchitis sputa, amounts to about 35 per cent., in others from 25 to 30 per cent, of the compound. These glyco-proteids, however, are proteids which play but a small part in metabolism. Egg-albumin by this method is found to contain from 10 to 11 per cent, serum- globulin less than 2 per cent., serum-albumin less than 1 per cent. In some measure these values may be perhaps checked by deducting from the percentage elementary composition of these proteids the carbon and nitrogen of the glucosamine found in each, and determining the ratio of carbon to nitrogen in the remaining true proteid moiety. Thus, egg-albumin has 52.75 per cent. C. and 15.45 per cent. N., glucosamine 40 per cent. C. and 7.5 per cent. N. In 100 parts of this albumin, then. II.] GLYCOPROTEIDS AND GLYCOSURIA 41 the lo per cent, of glucosamine would account for 4 parts of carbon and 0.75 parts of N., and the 90 parts of proteid freed from glucosamine would contain 48.75 parts of carbon to 14.7 parts of N., so that the ratio ^^ = 3.31. Calculated in the same C way, the proteid part of ovomucoid gives the ratio 1^ = 3-5, and of serum-albumin and globulin, 3.25 and 3.32 respectively. The values given for the carbohydrate in these proteids seem, there- fore, to be relatively pretty nearly correct. At any rate the mucin group of proteids contain some 30 per cent, more carbohydrate than the albumin and globulin of the blood. It is said that the proteids of cell protoplasm do not give the a-naphthol reaction at all. That would imply that they were still poorer in carbo- hydrate than the blood proteids. But, on the other hand, Neuberg found that the human liver freed from glycogen as far as it is possible to free it by boiling with water, gives glucosamine, upon hydrolysis with hydrobromic acid, amounting to 3.6 per cent, of the solids, which is about 5 per cent, of the proteids. It is not possible to say how much too low these estimates are, if at all. But it is at any rate not fair to be so influenced by the fact that some proteids yield 35 per cent, as to argue that if others yield i or 2 per cent, this is too low a figure. It is to differences of this order that the elementary composition of the different proteids points. And there is no sure basis whatever for taking 10 per cent, as the average amount of carbohydrate in proteids generally throughout the body, as Pfliiger proposed. The proteids that give the high figures are those that are of least importance in general metabolism, and those that are most important and most abundant give the low figures. Even if we take Neuberg's estimate of the glucosamine in the liver, which amounts to 5 per cent, of the proteids in the liver, as a typical figure for tissue proteids, this amount of glucosamine would give a ratio 5^ = 0.3, or only about a tenth of the ratio which Minkowski found. So that it is not possible to account for the 42 SYNTHESIS OF CARBOHYDRATES [lect. sugar in his experiments, and others of the same nature, by supposing that it came from the ready-formed glucosamine groups of the proteids broken down. Besides, in so far as it is proved that proteids are the source from which the sugar in severe diabetes is derived, it is proved that the proteids which contain no glucosamine at all form sugar, no less than those that contain it. In a case of diabetes, Falta found that caseine caused an even more pronounced degree of glycosuria than egg-albumin or serum-globulin ; and similarly, Mohr's experiments with nutrose and meat gave a more marked effect than egg-white.^ It is even a question whether glucosamine should be con- sidered as the equivalent of sugar in the body at all. No direct evidence at any rate of the formation of glycogen from gluco- samine has been obtained. It has been administered to starving animals, as the hydrochloride and as the free base, but in neither case did it appear to have increased the amount of glycogen in the liver, since animals to which it was not given, but which were starved for the same period, had no less glycogen in their livers.^ But this method of experimenting does not give very sharp results, since, as Pfluger has pointed out, the amount of glycogen varies very much in different animals starved for the same length of time. And Lang has shown that glucosamine gives up ammonia, just as the amido acids do, in the presence of substances contained in the cells of many organs, including the liver and intestines.^ If the removal of ammonia by hydrolysis is the only change brought about in the glucosamine, then certainly glucose must be the product of the change. If, therefore, we are to look to proteids at all as the source from which the material for the formation of sugar in the body is derived, it can hardly be to any considerable extent the ready- made carbohydrate groups in certain of these proteids that . account for this sugar. The formation of sugar from proteids is 1 Cf. Ergeb. iii., p. 493 ; and, Mohr, Z.f. k. M., 1904. ^ Cf. Cathcart, U.S. Z. 39, 423, 1903, ' Lang, H. B. v., 340, 1904. II.] LEUCINE AS A SOURCE OF SUGAR 43 a problem that cannot be so easily disposed of; a synthesis of a really fundamental nature must be involved. Leucine was the first of the derivatives of proteids to be thought of as a possible source of sugar. It is the most abundant of the amido acids to be obtained from proteids. It has a chain of six carbon atoms, though this chain is not normal but branched ; and it is perhaps conceivable that after removal of the nitrogen, by oxidation of one of the hydrogen atoms attached to each of the carbon atoms, and by reduction of the carboxyl to the aldehyde group, and lastly, by conversion of the branched chain into a normal one, leucine might be converted into sugar without being first broken up. But that any one of this series of changes does take place, we have no evidence whatever. The difficulties are not removed by pointing, as has been done, to the formation of saccharinic acid with a branched chain from glucose with a straight normal chain, which is brought about by the action of lime-water. Saccharinic acid, Cxi3\ yix /H >C.OH.C< . C< . CHjOH COOH/ \OH \0H does not bring us very near to leucine, >CH . CH„ . C< . COOH CHg/ \nH2 And the tetra-oxyamido-caproic acid obtained by Orgler and Neuberg from cartilage,^ which has also been quoted in support of the conversion of leucine into sugar, is a substance the exact constitution of which is not known. If it were known, it is questionable whether, from what is at present known of its occurrence, it would be reasonable to infer that it was an inter- mediate stage in a reaction widely prevalent in the body by which leucine could be transformed into sugar. Leucine, in fact, suggests itself, before all other proteid cleavage products, as ^ Orgler and Neuberg, //.-S. Z. 37, 107, 1903. 44 SYNTHESIS OF CARBOHYDRATES [lect. a precursor of sugar' only because of its six carbon atoms. And the "hexone bases," the bases containing six carbon atoms, lysine, arginine, and histidine, are still less .likely to preserve their six carbon atoms in one molecule in any changes leading to the formation of sugar. Attempts to prove that leucine gives rise to sugar in the body have not been conclusive.' Rabbits have been kept without food for from four to six days, and then given 20 g. or more of leucine ; some hours later more glycogen has been found in the liver in one or two cases than in that of control animals that had had no leucine.^ Rabbits, fed with leucine after severe strychnine spasms, have been found to have no hepatic glycogen.^ On the other hand, Mohr found, in a case of severe diabetes kept on a constant diet, that the administra- tion of 20 g. of leucine increased the daily excretion of sugar to the extent of about 10 or 15 g.^ Of the other cleavage products of proteids, the one which has been regarded as likely to be specially concerned in the produc- tion of sugar is alanine. By the substitution of hydroxyl for the amido group in alanine, lactic acid would be formed. Lactic acid is known to be very readily formed from sugar under many different conditions. It is possible that the opposite change from lactic acid to sugar may occur. Neuberg and Langstein gave alanine to rabbits that had been kept without food for eleven days, and found lactic acid in the urine. In the liver they found glycogen amounting to from one to two grammes — more than might be expected after so long a period of starvation, but not more than may be accounted for without supposing a direct transformation of alanine into glucose;* Kraus estimated by Pfliiger's method the amount of glycogen in five cats that had been similarly fed for some time : he found the glycogen to be on an average 0.31 per cent, of the weight of the cats ; the maximum being 0.47, and the minimum 0.20 per cent. Five other cats also similarly dieted were then treated with ' Cohn, H.-S. Z. 28, 210, 1899. 2 Simon, H.-S. Z. 35, 320, 1902. 3 Mohr, Z.f. k. M., 1904. * Neuberg and Langstein, D. R. A., supp., p. 514, 1903. M.] ALANINE AND ASPARTIC ACID 45 phlorrhizine, and kept without food for from five to eight days, when they were, killed. The sugar that had been excreted in this time together with the glycogen still remaining in the bodies of the cats amounted to 0.67 per cent, of their weight. The maxima were 1.23 and 0.77 per cent, given by two animals to which 5 g. of alanine had been given daily : the minimum was 0.33 per cent, in an animal that had had i g. of leucine daily.^ Embden and Salomon experimented on dogs with glycosuria, following removal of the pancreas. In one case 34 g. of alanine were given, and 14 g. of sugar above the amount excreted with constant diet were found in the urine of the next two days. Another dog kept without food excreted daily for four days on an average 3.5 g. of sugar, and then in the two days following a dose of 20 g. of alanine excreted 25 g.^ They also found that sodium lactate increased very considerably the output of sugar in animals with pancreatic diabetes.^ The amount of alanine that can be obtained from proteids is, it is true, but small ; but it is remarkable that compounds of alanine are numerous ; tyrosine, phenyl-alanine, trypto- phane, cystine, serine, histidine, are all compounds of alanine, and it is possible that from some, at any rate, of these alanine may be available for the synthesis of sugar. Another amido acid for which there is the same sort of evidence, that it can undergo changes ultimately leading to the formation of sugar, is aspartic acid. Embden and Salomon found that in their dogs a dose of 20 g. of aspartic acid increased the sugar excretion by six or seven grammes. A similar result was observed by Knopf in phlorrhizine glycosuria.* And twenty years ago R5hmann noted that rabbits fed on carbohydrates and asparagine always had more hepatic glycogen than others fed on carbohydrates alone, up to even eleven times the amount.^ It would be possible to connect the conversion of asparagine or ^ F. Kraus, D. M. W. 14, 1903 ; and, B. k. W. i, 4, 1904. ^ Embden and Salomon, H. B. v., 508, 1904. ^ H. B. vi., 63, 1904. * Knopf, 5. A. 49, 135, 1903 ; cf., too, Nebelthau, M. M. W. 917, 1902. « Rohmann, Cbl.f. k. M. 35, 2, 1884. 46 SYNTHESIS OF CARBOHYDHATES [lect. aspartic acid into sugar with that of alanine, if it were shown that aspartic acid give up COg in the body, thus : COOH.CHo.c/ — COOH ^C0o + CH3.C< — COOH Such a change is not without parallel ; and if it occurs, then in this case too the formation of sugar from proteid cleavage products may be conceived of as occurring through lactic acid as an intermediate stage. If these results be taken to point to lactic acid as a probable stage in the synthesis of sugar in animals, then it is important in this connection to take note of the striking result obtained by Liithje with glycerine. Glycerine is a substance which has certain relationships with lactic acid in biological chemistry. Glyceric aldehyde, cA - c/" - cf \OH \0H \H is isomeric with lactic acid, and from glyceric aldehyde it is known that sugar can be synthesised. Liithje fed a dog, that had had its pancreas removed, with large quantities of glycerine mixed with serum. In fourteen days the dog took 2890 c.c. of glycerine, and during this period 1408 g. of sugar were passed in the urine. The dog weighed to begin with 15.2 kg., so that if there were in its body the maximum amount of glycogen ever found in a dog, the equivalent of 41 g. of sugar per kg., then 623 g. of sugar could have been formed from the glycogen of the animal's body. This leaves 785 g. to be accounted for from the glycerine or from proteids or fats. The nitrogen excreted in the fourteen days amounted to 210 g. If we suppose that for each gramme of nitrogen excreted, 3 g. of sugar were formed from the proteid after the removal of this amount of nitrogen ; if we take, that is, the ratio ^^ as = 3, which is a high value for this ratio, then 630 g. of sugar may have come from the proteid broken down. That leaves 155 g. to be II.] GLYCERINE: LACTIC ACID: GLYCOCOLL 47 accounted for from the glycerine or from fats. It is not easy to escape from the conviction that glycerine in this case was converted into sugar.^ If, therefore, glycerine and those amido acids which may give rise to lactic acid in the body may serve for the synthesis of sugar, it certainly seems as if the synthesis of sugar probably is effected by condensation of chains each containing three carbon atoms, as in the well-known synthesis of sugar from the oxidation products of glycerine, glyceric aldehyde and dioxy-acetone. But Embden and Salomon obtained as marked results with glycocoU as those quoted above, which they obtained with lactic acid, alanine, and aspartic acid. If these results are also to be taken as evidence for the direct synthesis of sugar from the substance administered, then we have no hypothesis so near at hand to account for a synthesis such as this implies. It is difficult to imagine the formation of lactic acid or any three carbon chain from glycocoll. P. Mayer has injected glycollic aldehyde, into rabbits, and found about lo per cent, of the theoretical amount of glucose in the urine, and this he is inclined to suppose was formed by condensation of the aldehyde directly to sugar.^ Even if the sugar was formed from the substance injected, it is not proved that the aldehyde condensed without undergoing any other change. We are equally at liberty to suppose that a sugar synthesis occurs in the body, not only from this substance, but from glycocoll and from any fatty acid, whether derived from amido acids or from fats, by condensation of formic aldehyde groups. Till we know how the simple fatty acids such as acetic acid are dealt with in the body, we cannot say whether it may not be possible that formic aldehyde groups should be snatched out of the burning of these substances and ^ Liithje, D. A.f. k. M. 80, 98, 1904. 2 p. Mayer, H,-S, Z. 38, 148, 1903. 48 SYNTHESIS OF CARBOHYDRATES [lect. used for the synthesis of sugar by a condensation similar to that which Bayer supposed to be brought about in plants.i Strict proof, it must be remembered, has not been obtained that the sugar that appears in the body, and cannot be derived from the carbohydrates of the body, has necessarily been derived from material supplied by proteids. It may be that the proteids, or proteid cleavage products, which often appear to give rise to an increased formation of sugar, do so in some indirect way. Pfliiger, who has subjected all the evidence for the formation of sugar from proteids to what will at times appear an unnecessarily stringent criticism, interprets all the cases in which sugar appears to be derived from proteids, by supposing that proteids and proteid cleavage products stimulate the liver to increased activity in converting fats into carbo- hydrates. This latter change is no less difficult to understand than the change from proteid cleavage products, so that our difficulties are not removed by this supposition. Whether it be fats or proteids, or both, that supply the material for sugar synthesis, we are for the present brought up to a stop by the fact that we cannot follow the reactions by which either fats or proteids are oxidised. We know something about the hydrolytic processes in which these substances are involved, but in order to be able to advance further we need to know something of the later stages in their breakdown. At some point, clearly, some substance or substances arise which by a sifJe^r-eaction can condense and give rise to the synthesis of carbohydrate groups. However tempting it may be to speculate as to the nature of this substance, whether it be formic, glycollic, or glyceric aldehyde, or any other possible derivative of lactic acid, the fact remains for the present that we do not know what the nature of the substance is. If, as we shall see there is reason for believing to be the case, the amido acids lose their amido 1 groups, they become simple fatty acids or non-nitrogenous derivatives of these, the oxy-acids, and, according to one view ^ Dakin has shown that formic aldehyde is produced when glycocoU is oxidised with hydrogen peroxide and ferrous s\i\^h.a.ts.— Journal of Biological Chemistry, i., 171, 1906. II.] THE REAL STARTlNG-POtNT OP THE SYNTHESIS 49 at any rate of the breakdown of the higher fatty acids, these substances too will be formed from fats. So that it may well be that some one and the same simple non-nitrogenous com- pound may arise at some point in the course of the reactions by which both fats and proteids alike are broken down ; and that this substance is capable, under the conditions obtaining in the body, of undergoing a synthetic condensation to sugar. LECTURE III CARBOHYDRATE CATAEOLISM The final products of the metabolism of non-nitrogenous substances in the body are carbonic acid and water. Physiology has been much concerned with the mode of origin of the principal nitrogenous substances, that are also final products in metabolism ; but in the matter of the origin of carbonic acid and water, we have still to be content with the simile of the combustion furnace : the fats and carbohydrates are completely burnt up, and leave the body as carbonic acid and water. There is something, however, known about the decomposition of sugar in vitro, and much of this has almost certainly important bearings on the physiology of carbohydrates and the course of the reactions by which in the body they are finally converted into these end-products. If the view of biological chemistry which is being generally adopted is correct, that the reactions in living organisms are not different in kind from those that can be observed in the same material where no life is present, but that the differences depend upon the acceleration of certain phases in these reactions, which is effected by catalytic agents, then whatever can be made out as to the natural lines of cleavage in the molecules of sugars and fats must be of significance in the physiology of these substances. Catalysis, which is essentially merely a change in the velocity of reactions, may, it is true, by hurrying these reactions past the points in their course at which secondary reactions are prone to occur, acquire a directive 60 LECT. III.] ACTION OF ALKALIES ON SUGARS 51 control, but though it may thus alter the yield, it must work along the lines which determine also the general course of change in vitro. There are many reasons why for us special interest should attach to the familiar reaction by which sugar is converted into lactic acid. When sugar solutions are treated with alkalies at a temperature of 60-70° they rapidly undergo changes which result, except in the case of cane sugar, in a discoloration of the fluid, familiar under the name of Moore's test. What the yellow or brown colour is due to is not known, but the most abundant derivative from the decomposed sugar, under certain conditions at any rate, is lactic acid. That this acid is formed from sugar by alkalies, was shown by Hoppe-Seyler and Schiitzenberger,^ and in 188 1-2 Nencki made a special study of the change and its conditions.^ A solution containing 10 per cent, of glucose and 20 per cent, of caustic alkali after twenty-four hours at blood heat contains only traces of sugar, but lactic acid amount- ing to 50 per cent, of the sugar taken. The conversion takes place in the complete absence of air, and is to be observed not only in the case of glucose, but also with lactose, maltose, laevulose, galactose, the pentoses, arabinose, and xylose,^ but not cane sugar — with those sugars, that is, that contain the unaltered aldehyde or ketone group, and for that reason give Moore's reaction. Ammonia and the alkaline carbonates do not produce the change, but neurine and organic ammonium bases on the other hand do. Duclaux has since then shown that even at the room temperature, in the presence of sunlight, baryta water forms lactic acid from sugar to the amount of 60 per cent. ; and this has been confirmed by Buchner, who also found that even in the dark, lactic acid was formed from a 5 per cent, solution of glucose in 5 per cent, caustic potash, amounting in eleven months to 15 per cent, at the ordinary temperature.* * Hoppe-Seyler, B. 4, 396, 1871. ^ Nencki,/. Pr. Ch. 24, 498, and 26, i, 1881-2. ^ Katsuyama, A 35, 669, 1902. •• Duclaux, A. P. I. 7, 751, and 10, 168 ; Buchner and Meisenheimer, B. 38, 620, 1905 ; B. 27, 417, I9°4- 52 CARBOHYDRATE CATABOLISM [lect. Even more familiar than this action of inorganic and organic bases, is the resolution of sugar into lactic acid by the action of micro-organisms. Countless varieties of bacilli, including B. typhosus and B. coli ; cocci, vibrios, including the cholera vibrio ; and sarcina, are known to bring about this change. The bacilli that cause milk to curdle are capable of forming as much as 80 per cent, of lactic acid from glucose, so that in this case, even if none of the sugar fails to yield one molecule of the acid, 60 per cent, of the sugar must undergo changes, summed up in the equation — CeHiaOg = 2 CgHgOg In some of the other fermentations of sugar by micro- organisms lactic acid is believed to be formed at one stage of the fermentation. Pasteur showed that a bacillus that formed butyric acid from sugar formed it also if lactic acid were substituted for the sugar. This has not been found to hold by any means for all the micro-organisms that cause butyric fermentation, but it is commonly and fairly assumed that those which can convert calcium lactate into butyrate, when they convert sugar into butyric acid, do so by first converting it into lactic acid. Buchner has recently expressed the view that also in the alcoholic fermentation of sugar by yeast the first stage is the conversion of the sugar into lactic acid. His reasons for this belief are that lactic acid is frequently to be found in the expressed juice of yeast cells, and in that case may disappear during fermentation ; or on the other hand, it may be absent in the juice originally, and then subsequently appear during fermentation. This, he suggests, points to the existence of two reactions in alcoholic fermentation ; the first, the conversion of sugar into lactic acid, and the second, the formation of alcohol and carbonic acid from lactic acid ; the two reactions he ascribes to two distinct enzymes, the name "zymase" being given to the enzyme concerned in the first stage, and " lactacidase " to that concerned in the second. Normally, both are present and almost equally active in the yeast cell, so that the lactic acid is III.] LACTIC ACID IN FERMENTATIONS 53 used up at nearly the same rate as that at which it is formed.^ Harden and Young have shown that the expressed juice of yeast when filtered through gelatine under pressure, is separated into two parts, neither of which alone has the power of ferment- ing sugar, while the mixture of the filtrate with the residue, that cannot pass through the filter, has this power like the original juice. But this separation is not the separation of Buchner's two enzymes. The substance in the filtrate which is necessary in the fermentation is not altered by boiling, and the exact rela- tionship of the parts played by the two constituents of the zymase has not yet been determined ; but provisionally it can be com- pared to that found by Magnus between the ferment and co- ferment, which together produce the lipolytic action of liver cells.^ Some support for Buchner's conception of the process in alcoholic fermentation may be derived from the fact that Maz6 has shown that Eurotiopsis Gayoni produces alcohol from lactic acid ; and perhaps, too, from the fact that Thomas finds that formic acid is produced by yeast, especially when certain substances such as urea or certain salts of ammonia are added to the media on which it is grown.^ Whatever the result of the further investigations on this point may be, there is no doubt that the formation of lactic acid from sugar would make the chemistry of both alcoholic and butyric fermentation more intelligible. For lactic acid is an a-oxy-fatty acid, and it appears to be a general reaction for such compounds to split into formic acid and the aldehyde of the next lower member of the series of acids. Le Sueur has shown this to hold for all the fatty acids from stearic to lauric* And in the case of lactic acid, it is well known that in a number of different conditions the tendency to break up in this way is manifested. CH,.C< — COOH >-CH,.C< + H.COOH • Buchner and Meisenheimer, B. 37, 417 ; and B. 38, 620, 1904-5. - Harden and Young,//. Phys. 32, i. ; Magnus, B.-S. Z. 42, 149, 1904. ' Maz^, A. P. I. 16, 446, 1902 ; P. Thomas, C. R. 136, 1015, 1903. * Le Sueur,/. C. S., Dec. 1905. 54 CARBOHYDRATE CATABOLISM [lect. When heated to a high temperature, about 440° C, the change occurs spontaneously ; in the presence of dilute sulphuric acid, it occurs at 1 30° C. Electrolysis of lactic acid gives the same products. And the same reaction clearly underlies the fact observed by Duclaux, that whereas the weaker bases, baryta or lime-water, acting on sugar in sunlight, form lactic acid, the stronger base, potash, under the same conditions forms alcohol and carbonic acid. For the formic acid, on breaking up into hydrogen and carbonic acid, would furnish the hydrogen for the reduction of the aldehyde to alcohol.^ If lactic acid be treated with sulphuric acid and manganese dioxide or peroxide of lead, aldehyde and carbonic acid are liberated ; the two atoms of hydrogen that would otherwise have been set free from the formic acid being oxidised to water as soon as formed. If, therefore, lactic acid is a precursor of alcohol and carbonic acid in alcoholic fermentation, the changes undergone by sugar in this process are far more intelligible than when crudely summed up in the equation : CeHiA = 2C2HeO + 2C02 And similarly the equation, by which the final results of butyric fermentation is represented, C^HiA = C,H302 + 2H2 + 2C02 also becomes intelligible if we suppose that two molecules of lactic acid are first formed from the sugar, and that these break up into two molecules of acetic aldehyde and two of formic acid. For one of the most familiar of the changes to which aldehyde is liable is the condensation of two molecules to form aldol— .0 /OH .0 2QYi.ff = CH3C/ — CH^C^ H ^H ^H ' Duclaux, A. P. I. 7, 751, and 10, 168. ,n.] LACTIC ACID ^ ACETIC ALDEHYDE 55 And this aldol, or j8-oxybutyric aldehyde, by reacting with two molecules of water, thus — CH3.C< H OH -OH CH, C // O ^H + H \ H OH would give CH3 . CH2 . CH2 . COOH + 2 H2O, or butyric acid. Butyric acid fermentation, therefore, like alcoholic fermentation of sugar, may reasonably be expected to depend on the tendency of sugar to split up into lactic acid ; so, too, with acetic acid fermentation, in which case it may be noted that the acetic acid produced always contains some acetic aldehyde. How the formation of lactic acid from sugar is to be ex- plained is not yet quite clear. From the fact that galactose ferments less readily with yeast than glucose, fructose, or mannose, in all three of which the atoms round the two middle carbon atoms are disposed thus — H OH — C — C — OH H while in galactose the arrangement about these two carbon atoms is H H — C — C — OH OH it is possible that the former arrangement is particularly favour- able to a reaction with water, thus — HO "^OH — G H OH H 56 CARBOHYDRATE CATABOLISM [lect. resulting in the formation of two molecules of glyceric aldehyde, CH„OH.C< — Cr \0H \h or its hydrate. Glyceric aldehyde has been shown by Nef to be converted into lactic acid, under conditions in which glucose and fructose also yield lactic acid, with caustic soda at a slightly raised temperature. And the formation of glyceric aldehyde from glucose would be, if reversed, the familiar reaction by which a-acrose (z-fructose) is formed from the oxidation products of glycerine : CH,OH.C^ . c/ + CH,OH . CO . CH,OH -■OH ^O .H /H ^OH = CHpH.c/ . c/ —c( — CO.CHoOH \0H \0H \H If this is the way in which the chain of six carbon atoms splits up into two chains of three, the transformation, 6f glyceric aldehyde to lactic acid has still to be accounted; for. The change is precisely the same as that from aldol to butyric acid referred to above, and might be explained iu the same way. '-^ Nef has, however, proved that in the conversipn of sugar into lactic acid by the action of caustic alkali, py'hivic aldehyde, CHg.CO.Cf^ is an intermediate product. The probability of the occurrence of pyruvic aldehyde in the synthesis of imidazol compounds in t;he bodj^ will be referred to in a subsec^uent lecture. Bu Zillessen, H.-S. Z. 15, 392, 1891. ^ Marcuse, Pfl. A. 39, 425, 1886. 3 Astaschevsky, H.-S. Z. 4, 397, 1880; Warren, Pfl. A. 24, 391, 1881 ; and, Monari, M. J. 303, 1889. 64 CARBOHYDRATE CATABOLISM [lect. reaction were carried one step further than lactic acid, there would still be no evolution of energy. The physiological significance of the appearance of lactic acid in muscle is not that which is sometimes assigned to it. If, therefore, the formation of lactic acid from sugar is a constant normal factor in the metabolism of muscle, it is merely preliminary to subsequent oxidation processes in which the energy of the sugar serves for purposes of work or heat pro- duction in this tissue. We should not look on lactic acid as a waste product excreted by the muscle as useless and done with, or even as necessarily the product of a miscarriage of metabolism, as Hoppe-Seyler seemed to imply. It may be that the appearance of lactic acid in the blood and urine after exertion is due to the escape of but a very small part of the whole amount of this substance that has been produced in the course of the ordinary metabolism of the muscles, and that it has escaped merely because the other subsequent stages of sugar catabolism have lagged behind this earlier one ; a sign perhaps of overwork, or of inadequate oxygen supply, as Hoppe- Seyler himself supposed. Such a conception of the position of lactic acid in the chemistry of muscle may help to make intelligible the discrepant results, too, which have been obtained with excised muscles, in which excitation has sometimes caused a diminution, sometimes an increase of lactic acid in the tissue. Such results are exactly comparable with the observations of Buchner on the production of lactic acid in yeast juice : in some samples he found the lactic acid would increase, in others it would diminish. If lactic acid is not a final product of the metabolism of muscle tissue — neither the raw material nor the finished article— it ceases to be strange that the amount of material in a half-worked condition found in stock at any moment should differ in different cases. In the series of reactions by which sugar is broken down, if the first phases to fail are those that come earliest, then there must be a diminu- tion of lactic acid ; if, on the other hand, the later stages come to a stop, while the earlier ones are still actively carried on, then there will be an increase. in.] GLYCOLYSIS iN THE BLOOD 65 There is, therefore, reason for connecting the lactic acid found in the organs we have been considering with the sugar or glycogen observed to disappear from them. The disappear- ance of sugar from body fluids or tissues has been the subject of a number of important studies, in which even the temporary appearance of lactic acid in the place of the sugar has not been definitely determined. The observations from which all these investigations upon the phenomena of glycolysis started were made by Claude Bernard. He found that the sugar in the blood diminished very rapidly in the first hours after the blood was withdrawn from the vessels. In the first five hours 6o per cent, of the original amount of sugar might disappear, and at the end of twenty-four hours none might be left. Pavy confirmed Bernard's results, and Lupine more recently developed out of them and his own experiments on this point a theory of the disorder of metabolism in diabetes.^ He found that the glyco- lytic properties of the blood resided in the leucocytes, since the blood corpuscles were more active than the serum, and the lymph than the blood. But the glycolytic agent could be extracted with normal saline, so he regarded it as an enzyme. This enzyme he thought at first was formed in the pancreas, for three reasons. First of all, the glycolytic power of the blood was diminished by removal of the pancreas. (This has not been found to be the case by others who have repeated the experiments.) Then it was increased by tying the duct of the pancreas or cutting its nerves. And thirdly, the blood of the pancreatic veins was more active than that of the splenic vein. (But this also has been disputed by others.) These ex- periments, and the attempt to derive from them a complete theory of the causation of diabetic glycosuria, have led to a large number of investigations upon the glycolytic action of the blood and tissues, and the relations of the pancreas to this action. Claude Bernard himself thought that the sugar was con- verted into lactic acid. Seegen could not, however, detect any increase in the amount of this substance present in shed blood. ' Lepine, M.J., 1891 to i8g6. E 66 CARBOHYDRATE CATABOLISM [lect. So that if lactic acid was formed it must be formed only as an intermediate product, and must itself undergo further change. If this further change should be of the same nature as that which Buchner believes to be set up in the lactic acid produced by yeast, then it should be possible to demonstrate the forma- tion of carbonic acid and also of alcohol. Oppenheimer accord- ingly tried to determine whether alcohol was formed, but could get no evidence of it except traces of some substance that gave the iodoform reaction, but was not acetone. At the same time, he failed to detect any evidence of the formation of lactic acid. Very similar results were obtained by Herzog, not only with blood but also with the pancreas. Blumenthal found that the juice expressed by hydraulic pressure from a variety of organs developed carbonic acid when mixed with sugar solutions ; but he could not find any alcohol, till Feinschmidt in his laboratory subsequently isolated a small quantity of alcohol produced by the action of the fluid expressed from the liver on sugar solu- tions in the presence of antiseptics. In some cases a precipitate obtained from the expressed juice by means of alcohol and ether was dried, and the solutions of this precipitate were found to give somewhat better results than the juice itself. In the meantime, Stoklasa and others working with him found that various organs, especially the pancreas, but also the liver and muscles, contained a substance, precipitated with alcohol, and ether from the expressed cell juice, which produced in an atmosphere of hydrogen a vigorous fermentation when added to sugar solutions. Feinschmidt in part confirmed Stoklasa's results, but others have either completely failed to do so, or shown that the fermentation which they obtained was due to bacteria.! The glycolytic action of the fluid expressed from muscles is said by Cohnheim to be elicited only under the influence of something present in the pancreas which can be extracted from it by boiling water and is soluble in alcohol. Neither the muscle juice nor the alcoholic extract of pancreas has any • Maze, A. P. /., i8, 378 ; Portier, ii. 633, 1904 ; Cohnheim, I/.-S. Z. 39, 348, 1903 ; Arnheim and Rosenbaum, H.-S. Z. 40, 233, 1903. in.] THE PANCREAS AND GLYCOLYSIS 67 glycolytic action by itself, but the two together, provided there is not an excess of the latter, produce a powerful action. The nature of the substances formed from the sugar in his experi- ments he has not determined. Results in some respects similar to Cohnheim's have been obtained by Rachel Hirsch.i But here, too, there are considerable difficulties in definitely adopting the scheme of correlated interactions of different parts of the body as laid down by Cohnheim. For the difficulty of eliminating bacterial growth and preserving at the same time the conditions necessary for the development of the reactions described, has in the hands of others so far proved insuperable.^ Till these conditions are more clearly defined, and it becomes possible to control them more completely than at present is the case, we cannot be certain that such experiments tell us anything about reactions due to the cell- substances themselves, and consequently those which are carried out in the living body. Glycolysis, whether in the blood or in the tissues, is a subject therefore on which it is possible for the present to make but few positive statements : neither the conditions under which it occurs, nor the nature of the change involved, has been determined in such a way as to give final and conclusive results. Occasion arose in the course of the last lecture to refer to one way in which some of the sugar in the body is believed to be broken down. We then saw that glycuronic acid, which differs from glucose only in having a carboxyl group substituted for the terminal primary alcohol group of the sugar molecule, had been proved by P. Mayer to be formed from glucose in the rabbit. This substance occurs in traces in normal human urine, about half a decigramme daily, combined for the most part with 1 R. Hirsch, H. B. 4, 535, 1903. 2 Embden and Glaus, H. B. 6, p. 214, 1905. Harden and Young have repeated Cohnheim's experiments, with the one modification that the muscle was ground with sand, instead of being frozen and cut in shavings with Kossel's apparatus, as done by Cohnheim. But they found no trace of the action described. {Private communication^ 68 CARBOHYDRATE CATABOLISM [lect. phenol : ^ it is also present in bile,^ in the blood,^ and in faeces. But the administration of camphor or chloral causes it to appear, combined with these substances, in the urine in considerable quantities. This fact may be taken to indicate that it is formed as an intermediate product in metabolism, possibly in still larger quantities, and that it does not come into evidence unless it happens to be caught and fixed by combination with such a substance as phenol, because it is otherwise completely oxidised. Mayer found in rabbits no indication that doses of glycuronic acid up to about 5 g. were not as completely oxidised as glucose itself When larger amounts were given, however, the acid was found in the urine, partly free and partly combined with the phenol that would otherwise have been present combined with sulphuric acid. There was no increase in the phenol, but more of it was paired with glycuronic acid and less with sulphuric acid.* The fact that normally only a small fraction of the phenol pairs with glycuronic acid may be taken to indicate that glycuronic acid does not occur in the body normally in such large quantities as in these experiments, even as an intermediate product of metabolism, and therefore that the whole of the sugar oxidised in the body does not go through the changes which in the first instance give rise to the formation of this acid. But the absorption of glycuronic acid from the intestine, it must be remembered, is not the same as the formation of glycuronic acid in the cells in the course of oxidation of sugar. Stress cannot always be laid on arguments from the results of administering intermediate products of metabolism, but it is remarkable that doses of from 15 to 25 g. of glycuronic acid proved fatal to rabbits, and in such animals considerable quantities of oxalic acid were found in the liver, 60 to 80 mg. instead of mere traces. All that we can say, therefore, is that it does not appear probable that the normal course of sugar metabolism necessarily and always ' Mayer and Neuberg, H.-S. Z. 29, 267, igoo. 2 Bial, H.-S. Z. 45, 263, 1905. 3 ivlayer, H.-S. Z. 32, 518, 1901. * Mayer, Z.f. k. M. 47, 68, 1902. HI.] GLYCURONIC ACID AND XYLOSE 69 involves the formation of glycuronic acid, though clearly some of the sugar is oxidised in this way. It has, however, even been contended that this substance is not derived from sugar at all. Loewi concludes, from his experiments on the administration of camphor to dogs in phlorrhizine diabetes, that, at any rate under these conditions, it is not. In one case a dose of lo g. of camphor, which would and apparently did combine with as much glycuronic acid as could be formed from 12 g. of sugar, instead of diminishing the amount of sugar excreted to this extent, was accompanied actually by an increase of the sugar excreted, amounting to about 20 g., or nearly 30 per cent, of the average amount excreted on the previous days. In the three other experiments, in each of which two doses of camphor were given, on four occasions the sugar excretion was diminished, on two unchanged.! The results, therefore, varied somewhat widely, and it is questionable whether they furnish proof that in the dog under these conditions the glycuronic acid is not formed from carbohydrates. If it is possible for sugar to be synthesised in the body from material derived from either proteids or fats, then clearly glycuronic acid may be remotely connected with either of these substances, though directly formed from sugar. The probability that the xylose found in nucleic acids may be the next product after glycuronic acid in the series starting from glucose along this line of catabolic change, has also been referred to in the last lecture. But beyond this possibility we cannot trace the catabolism of sugar in this direction at all. In connection with the origin of xylose we have to take into account the phenomena of pentosuria. This remarkable anomaly of metabolism — it does not amount to a disease, since it is not accompanied necessarily by any disturbance of health — was discovered ^ before it was known that pentoses entered into the composition of any substances in the body. When this came to be known, there was not unnaturally a tendency to associate the pentose in pentosuria with the pentose in the ' O. Loewi, S. A. 47, 56, 1902. ^ Salkowski and Jastrowitz, C5/. Jlfed. Wtss., Nos. 19 and 35, 1892. 70 CARBOHYDRATE CATABOLISM [lect. nucleic acids. But with the determination of the exact nature of these pentoses, the identification of active /-xylose in the nucleic acids from the pancreas and liver/ and of inactive arabinose in the urine,^ this tendency was found to lead to difficulties. The sugar in the nucleic acids is represented by the formula — OH H OH COH.C . C . C . CH^OH I r I H OH H The sugar in the urine in pentosuria is the inactive mixture of H OH OH COH.C . C . C . CHgOH (ti-Arabinose) I I I OH H H and OH H H I I I COH.C . C . C . CH,OH (/-Arabinose), III H OH OH The second of these (/-arabinose) is related to /-xylose as galactose is to glucose. If, therefore, it were the optically active /-arabinose alone that were found in the urine in these cases, we might explain its presence either by a stereoisomeric transformation of /-xylose similar to that which has been suggested to account for the formation of galactose from glucose, or by deriving it from active galactose, as xylose has been derived from glucose. But apart from the fact that neither by feeding dogs on large quantities of pancreas could pentosuria be induced, nor in the subjects of pentosuria could any relationship be observed between the pentose excreted and the nature or amount of the ' Neuberg, B. 35, 1467, 1902 ; and, Wohlgemuth, H.-S. Z. 37, 475, 1903. 2 Neuberg, B. 33, 2243, 1900. HI.] PENTOSURIA 71 carbohydrates taken as food/ it would be contrary to all that is known concerning optically active compounds to suppose that inactive arabinose should be formed directly from either active xylose or active galactose. But if the terminal carbon atoms of galactose undergo any change by which the groups into which they enter become indistinguishable ; if, for instance, both are oxidised to carboxyl, giving rise to mucic acid, or if, by reduction of the aldehyde group both become primary alcohol groups, giving dulcite, COOHf OH H H OH ^ COOH or J C . C . C . C I or CHjOHi H OH OH H J CH^OH the substances formed are, owing to the symmetrical disposition of the asymmetric carbon atoms in the middle of the molecule, optically inactive. If, then, a derivative of galactose fulfilling this condition by subsequent changes gave rise to arabinose, as Neuberg suggests,^ both modifications must occur equally readily and the arabinose be inactive. For instance, if carbonic acid were removed from one end of mucic acid, it must be equally easily removed from either end, and the resulting arabonic acid must be inactive, and on reduction give inactive arabinose. The occurrence of inactive arabinose as a product of animal metabolism is so remarkable a phenomenon, that it is important to be alive to its significance and its bearings on the metabolism of carbohydrates in general. 1 Bial and Blumenthal, D. M. IV., No. 22, 1901. ^ Neuberg, Ergeb., iii., 429. LECTURE IV THE ASSIMILATION AND SYNTHESIS OF FAT The assimilation of carbohydrates involves in all cases some change in the nature of the carbohydrates. Starch and sugars are converted into hexoses ; the hexoses undergo certain trans- formations, and are built up into glycogen. But it may possibly be questioned whether carbohydrates are synthesised under ordinary circumstances from material which is not carbohydrate in nature to start with. The synthesis of sugar can certainly be effected in animals, but we do not yet certainly know that such a synthesis commonly takes place in normal metabolism, nor do we know certainly what is the exact nature of the chemical combinations made use of in this synthesis, when it occurs. _In the metabolism of the fats the situation is different. There is no doubt — it was proved fifty years ago — that fats are built up from material of quite a different nature, in animals that are in every respect normal. On the other hand, all the evidence points to fats being taken up and assimilated without transformation. The glycerine and fatty acids are, it is true, temporarily dissociated in digestion for the purpose of absorp- tion ; but they are recombined almost immediately without any other change, and they are found stored up, when not required for immediate use, in the identical form in which they were taken in the food. Thus it is familiar that the nature of the fat in an animal's body largely depends on the nature of the fat in its food. In certain vegetable oils fatty acids occur as glycer- ides which are not commonly found in animal fats. Colza oil 78 LECT. IV.] FAT IS ASSIMILATED WITHOUT CHANGE 73 or rape-seed oil, for instance, contains in large quantities the glyceride of an acid of the oleic series, erucic acid, CjaH^jOg ; and linseed oil is composed largely of linolein, linoleic acid being a derivative of stearic acid, with unsaturated linkages at two points in the chain, and having the formula C^gHggOa. If either of these oils is made use of as food, or if fats even in which the unsaturated carbon atoms have been saturated with the halogens iodine or chlorine are taken, then these unusual glycer- ides are laid down in the connective tissues without change, just as if it were perfectly normal for them to be there. . Even in man it is known that, if erucic acid be given by the mouth, this acid, as glyceride, but otherwise unchanged, is found in the chyle. A dog's fat is usually so compounded that it can absorb, by reason of the amount of olein that it contains, from 5 1 to 56 per cent, of its weight of iodine. But if a dog, after a period of starvation in which most of its own fat is consumed, is fattened on fat mutton, the fat now found in its connective tissues absorbs only about 42 per cent, of iodine. It is almost as poor in olein as mutton fat itself, which absorbs only about 36 per cent, of iodine. One form of fat appears therefore to serve as well as another, if only it is absorbed. So that the assimilation of fat is a subject on which there is not much to be said. The character of the fat in the body is determined primarily by the nature of the fat in the food ; in part, too, doubtless by the nature of the fat synthesised from substances other than fat within the body, and also in part by the ease with which the different kinds of fat are absorbed. Fats are absorbed with difficulty if their melting point is higher than the body temperature. Olein is in all animals, but especially in cold-blooded animals such as fish, much more easily absorbed than fats with a high melting- point, such as paJmitin and stearin. But whatever fat is absorbed is ipso facto assimilated. No transformation is neces- sary, for a,ll the fatty acids are equally adapted to the reactions by which their chemical energy is subsequently liberated in the body. In the synthetic metabolism of fats, therefore, we need not stop over problems of assimilation. It is the actual building up 74 ASSIMILATION AND SYNTHESIS OF FAT [lect. of fat from substances of a different kind, that furnishes us with all our problems in this division of our subject. The fats are glycerine esters of higher fatty acids. It is necessary therefore to give what account we can of the origin of the fatty acids, of the origin of the glycerine, and of the process by which these are combined. The last of these three subjects, to begin with the one that presents least difficulty, has not failed to attract its full share of the attention of physiologists. The union of glycerine and higher fatty acids can be effected in the laboratory by heating the reagents together to high temperatures. The reverse process, saponification of the glycerides, can be brought about by the use of alkalies, especially in the presence of alcohol, or by steam at 300° C. In the body both saponification and the reverse operation are familiar, and apparently both are effected by one and the same means. The reaction, a reversible one, is the work of an enzyme or enzymes which are found widely distributed in the body ; in the pancreas and its secretion, in the s_tomach and its secretion, in the liver, the m amm a, the i ntestine , the connective tissues, and perhaps in the blood. A lipase has been shown to be present in all these tissues, which saponifies not only fats, but other esters like ethyl butyrate. Aqueous extracts of these organs treated with this ester in the presence of toluene develop an acid reaction, while controls to which no ester is added do so to a comparatively slight degree, and other controls which are boiled remain unchanged: treated with butyric acid and ethyl alcohol, they bring about the synthesis of the ester, to judge at least by the smell.^ Hanriot found that lipase causes the combination of butyric acid and glycerine, and obtained monobutyrin in this way ; since it can also saponify esters in which the alcohol and acid may be of widely different characters, its action is perhaps a general one, consisting in the acceleration of the approach to the equilibrium point in the system, acid + alcohol ^ >" ester + water. The significance of this action in physiology appears to be this : when fat is to be ^ Kastle and Loewenhart, Am. Ch. Jl. 24, 491, 1900 ; and, Loewenhart, Am. Jl. Phys. vi., 331, 1902. IV.] ACTION OF LIPASE 75 transferred across certain cell membranes, it is saponified ; when it is to come to rest, the ester is formed. Under certain conditions this reaction proceeds in the one direction ; under others, in the other. Thus, in digestion saponification of fats precedes their entry into the epithelium ; in starvation, it enables the connective tissue fat to leave the fat cells, and so reach the circulation, to be distributed to the organs that require chemical energy with which to do their work. Exactly how and why saponification aids the passage of fats through cell membranes is not known. The fatty acids are no more soluble in water than the fats themselves, and their alkaline salts have been shown to be poisonous when injected into the circulation — 0.15 g. of sodium oleate per kg. or less of the palmitate or stearate is fatal for rabbits.i They cause a great fall of blood pressure, and then diastolic arrest of the heart and loss of coagulability of the blood. But the free fatty acids though insoluble in water are soluble in certain fluids found in the body ; in the bile, for instance, and still more in the fluid found in the intestine ; ^ and it is probable that it is as fatty acid dissolved in some unknown agent in the body fluids that the fats are transferred through cell membranes. Neither is it easy to form a conception of the conditions which determine the direction in which this reversible reaction is to take place : how it is, for instance, that during inanition the stored fat is so easily put into circulation, and in ordinary conditions of nutrition is left at rest ? If it is the enzyme in the connective tissues that brings about the synthesis and deposition of the glycerides, how it is that this synthesis continues to be carried out even when the cells are already loaded up with fat ? If it is right to suppose that this widely distributed enzyme, lipase, serves the purpose of setting in motion changes in the distribution of fat in the body, it is clear that its function is important. But the change it effects is isothermic. One ^ Munck, D. R. A., supp., p. 117, 1890. 2 Moore and Rockwood, //. of Phys. 21, 58, 1897. 76 ASSIMILATION AND SYNTHESIS OF FAT [lect. gramme-molecule of ethyl butyrate gives, on combustion, 851.3 Cal. I g. mol. of Ethyl alcohol gives . 325.7 Cal. I g. mol. of Butyric acid gives . 524.4 Cal. 850.1 Similarly, 3 g. mols. of Stearic acid on combustion give . 801 7 Cal- I g. mol. of Glycerine gives .... 396 Cal. 8413 and I g. mol. of stearin (i g. = 9.43 Cal.) would give 8393 Cal. The differences fall within the limits of experimental error. In whichever direction, therefore, a lipolytic enzyme acts, it neither adds to nor deducts from the fund of energy on which the organism has to draw. In the complex of chemical changes that constitute the life of an organism, there are many which, though of service, are not intimately associated with the phenomena of its life. It may be of service in digestion that the free hydrochloric acid of the gastric juice should be neutralised in the duodenum ; but such a reaction is not one of peculiarly physiological moment in the sense in which the liberation of the chemical energy of sugars or fats is, when with that energy the heart, for instance, is enabled to do its work. The saponification of fat, or the reverse process, is not one of the changes in which a vital transformation of energy is brought about. The fatty acids are at one moment combined with glycerine, at the next they are free : it makes no difference how many times this to-and-fro movement is executed, the groups concerned are in either form equally charged with what they are to contribute to the life of the organism. But the synthesis of the high fatty acids from the food-stuffs other than fat is a chemical change so peculiarly characteristic of living matter that its importance is of a different order. That carbohydrates can be converted into fat at all, we know only from the physiological study of animal metabolism. It is a change for which living organisms have the monopoly. The IV.] LAWES AND GILBERT'S EXPERIMENTS 77 earliest method of studying metabolism, what is sometimes called the balance-sheet method, which consists in the analysis of the food on the one hand and the excretory output from the body on the other, as well as sometimes the whole body of the animals employed, and then balancing the elementary items, has rendered no greater service to physiology than this. LaWes and Gilbert's experiments on the fattening of farm stock, carried out on the experimental farm at Rothamsted, nearly fifty years ago, were the first of a large number of similar experi- ments, which all prove what was first suggested by Liebig, that large quantities of fat are built up in animals from the carbo- hydrates of the food. Now these fats contain about 96 per cent. I o f their weig ht of the higher fatty acidsJong^normal_^cliains, I that is, of 16 or 18 carbon _atmns,_sa±urated_oralmost saturated with hydrogen, and, except for the 2 oxygen atoms of the acid- carboxyl group, with hydrogen alone ; while in the sugars, the chains of 6 carbon atoms are oxidised all along the line. The change is remarkable however it is to be looked at. Supposing for a moment that 3 sugar molecules could join up end to end in a normal chain, no less than 16 out of the 18 oxygen atoms must be dislodged for it to become stearic acid, and those that remain be rearranged. And the more we call to our aid of what synthetic chemistry has to teach, the more remarkable the change appears ; so that for the chemist, the only solution of the problem has often appeared to be to regard the whole matter as a physiological fiction. At any rate, the fifty years since it was first proved, which have been more eventful and fruitful in the history of chemistry than any similar period has perhaps ever been in the history of any other science, have hardly produced a serious attempt to solve the problem. The problem gains in interest, too, if we take into account the fact that fats and oils are most probably synthesised from carbohydrates wherever they are synthesised in Nature. Whatever the solution may prove to be, there are certain facts, for the niost part very familiar, which can hardly fail to be of significance as clues to its elucidation. Of the fatty acids that occur in Nature, those that contain more than 5 carbon atoms 78 ASSIMILATION AND SYNTHESIS OF FAT [lect. are, with hardly an exception, the members of the series which contain an even number of carbon atoms.^ Of these even number acids the i6 and i8 carbon acids are much the most abundant, but, nevertheless, all the other intermediate members are found, in traces at any rate. The connective-tissue fat, it is true, does not contain these lower acids, but that is fat which is out of the current of metabolism passively waiting for its time to come. But in milk, besides comparatively large quantities of the glyceride of butyric acid, caproin, caprylin, caprin, laurin, and myristin are all found, in traces at least, the glycerides, that is, of the acids with 6, 8, lo, I2, and 14 carbon atoms. And in all cases these acids are the normal acids, with straight un- branching chains. Then in addition to these normal saturated acids certain unsaturated acids occur, principally oleic acid, but also, in plants at any rate, others, some of which are unsaturated in more than one linkage of the chain. And even in animals the twice unsaturated acid, linoleic acid, C^^H^fi^i i^ said ^ to account for 25 per cent, of the fatty acids in the lecithine of the hen's egg, and also is believed to be present in lard and other animal fats : for instance, the fat of the hare and the horse, in the latter of which it amounts to about 10 per cent, of all the fatty acids. In plants, too, besides unsaturated acids, at any rate one oxy-acid, ricinoleic, or hydroxyoleic acid, occurs abundantly in castor oil ; and other oxy-acids are said to occur in the wool fat of animals. It is possible that these facts should all be associated together, and that there is some genetic relationship between these several forms in which fatty acids occur ; that they indicate, in fact, steps in the process by which all alike are formed — a process which tends to be completed, and result in the evolution of the highest members of the group, those which actually occur in the largest quantities. If there is no such relationship, then each fatty acid will present a problem of its own, and our diffi- 1 Belief in the occurrence in nature of the acid C1VH34O2, margaric acid, has been of late resuscitated from time to time, but apparently on insuffi- cient and mistaken evidence. — v. Holde, B. 38, 1247, 1905. 2 Henriques and Hansen, Si. A. 14, 390, 1903. IV.] THE GENESIS OF HIGHER FATTY ACIDS 79 culties will be the greater. Thus the occurrence of palmitic acid side by side with stearic makes it improbable that the 1 8 carbon atoms of stearic acid are joined up in groups of 6 at a time. And the occurrence of all the even number acid makes it probable that the chains are built up by the addition of 2 carbon atoms at a time. If this conception can be combined with one which will account for the appearance on occasions of oxy-acids and of unsaturated linkages in the chain, then the problem of the building up of fatty acids will begin to present a comparatively simple aspect, such as it is probable that the natural process has. Since all fatty acids alike are equally available for catabolic processes, it is probable that the reactions on the down-grade of fat metabolism are not specific for each fatty acid ; that palmitic acid, for instance, because it contains i6 atoms of carbon, does not break down differently from stearic acid with i8. The reactions in the catabolism of fat are such as to apply equally to them all. The reactions by which they are generated are similarly likely to be the same for all. Nearly thirty years ago, in a series of papers on the changes effected by micro-organisms (Gahrungs-processe) Hoppe-Seyler drew attention to the similarity between many of these changes and those brought about by caustic alkalies.^ And in particular he devoted especial study to the fate of lactic acid under these influences. He found that lactic acid heated with caustic alkalies to a temperature of 220° C. began to give off hydrogen : as the temperature was slowly raised, the reaction became at first more energetic, then gradually slowed down, and finally stopped at about 300° C. The fused mass, dissolved in water and acidified with sulphuric acid, gave up, besides carbonic acid, upon distillation considerable quantities of fatty acids, principally acetic and butyric, but also caproic acid. In addition to these volatile acids, however, he always found small quantities of solid insoluble acids which melted in hot water to form oily drops : their alkaline salts were soluble in water or alcohol, and the aqueous solutions of these salts when concentrated and cooled 1 Hoppe-Seyler, I/.-S. Z. 3, 351, 1879. 80 ASSlMiLATiON AND SYNf HESIS OF P'AT [lect. set to a jelly, as a soap solution does. The melting point of the acids he found to vary in different experiments, but it was always lower than that of palmitic acid, and their molecular weight also varied, and was lower than that of palmitic acid. He concluded that these insoluble acids were mixtures of higher even number members of the acetic series formed from the lactic acid in order, in diminishing quantities as the chains became longer, and by the same reaction as that by which the butyric and caproic acid were formed. Pasteur, and also Fitz, had previously described the forma- tion of butyric and caproic acid from lactic acid by bacteria, and Hoppe-Seyler looks to these reactions as the simplest form of the change of such great importance in the physiology of \ plants and animals alike by which carbohydrates are converted into fats. On repeating these experiments of Hoppe-Seyler's it is easy to confirm the formation of volatile fatty acids, and also to obtain small quantities of a substance or substances which are not volatile, are insoluble in water, and in some respects behave like higher fatty acids. But the dark oily drops have a tarry appearance, and if separated from the strong salt solution in which they are first formed, dissolved in alkali, and reprecipitated, they tend no longer to float, but to stick to the sides and bottom of the vessel. The products of this reaction have been recently reinvestigated by Raper at the Lister Institiite, with the result that besides formic and acetic acids together v^ith butyric, and probably caproic, also isobutyric acid, are formed : the insoluble acids, described as higher fatty acids by Hoppe- Seyler, are of low molecular weight, are unsaturated compounds, since they decolorise bromine water and reduce permanganate in the cold, and they are heavier than water. Their exact nature was not determined, but they are clearly not the higher fatty acids, such as palmitic acid, for which Hoppe-Seyler took them.i Hoppe-Seyler's experimental support for his conception of the manner in which the long fatty acid chains are built up, goes 1 Raper,//. of Phys. 32, 216, 1905. IV.] HOPPE-SEYLER'S EXPERIMENTS 81 no further, therefore, than caproic acid at any rate. But the formation of even butyric and caproic acid from lactic acid must involve a synthesis, and just such a synthesis as that which occurs in Nature, leading to a series of acids each containing two carbon atoms more than the preceding one. The reaction, it is true, as carried out by him, cannot be traced beyond the second step, but if it were continued it must lead to the other even number members of the series. It is possible that, though mis- taken in supposing that he had obtained molecules of the order of magnitude of palmitic and stearic acid, he was really on the right lines, and that the reaction failed to reach the stage which he thought it reached, only because it was diverted by the inter- vention of side reactions : a selective catalytic action might be able to direct and keep it on the straight course by accelerating just those phases of the reaction which are necessary, and by rapidly carrying it past the points where side reactions are otherwise apt to lead it astray. It is certainly a remarkable coincidence (if only a coincidence), that not only in this reaction with alkalies caproic acid should appear as well as butyric, but that in the butyric fermentation of sugar by micro-organisms, caproic acid should also be found among the products so regularly that commercial caproic acid is obtained from crude fermentation butyric acid. It seems that lactic acid readily gives off some compound containing 2 carbon atoms that tends to condense with itself, once to form butyric and twice to form caproic acid. Exactly how the groups of 2 carbon atoms are successively added on, even if the reaction goes no further than caproic acid, we can of course only speci^ate. Hoppe-Seyler's suggestion as it stands {loc. cit., p. 357) hardly commends itself, as it supposes a tendency in ethyl alcohol to condense with acetic acid. But since lactic acid, as is well known, is very prone under the most varied treatment to split into aldehyde and formic acid,^ and aldehyde condenses with itself very readily, there is, for the present at any rate, more to be said for such a scheme as the following, which is based on a suggestion ■ Cf. supra, p. 53. F 82 ASSIMILATION AND SYNTHESIS. OF TAT [lect. made in the first instance by Nencki and developed by Magnus Levy.^ /H /H (I) 2CH3.C< - COOH = 2CH3.C/ + 2H.COOH \0H ^O Lactic acid splitting into Aldehyde and Formic acid. /H CH„ . 0. Aldehyde condensing to Aldol. \, o (3) + H- CHo.C /H HO H CH, -^ C = CH,.CH„.CH,.C00H + 2H„Q \ OH H ^O — OH . Aldol, with 2 molecules of Water, giving Butyric acid and Water. For the higher members of the series similar reactions may account. Aldehydes with 6 carbon atoms have been obtained from acetic aldehyde by Riban and Kekul^ : these are in the one case a twice unsaturated aldehyde, and in the other an aldehyde unsaturated at one linkage and retaining the aldol or oxy-form in the other. That aldol and similarly constituted bodies readily pass into unsaturated compounds by the loss of water, is familiar ; aldol itself, for instance, is very apt to become crotonic aldehyde — /H CHo.CH = CH.C< %o Dialdanie, a condensation product of 2 aldol molecules, was first described by Wurtz, who showed that it contained 8 carbon ^toms. He prepared the acid by oxidation, but did not deter- 1 M. Levy, Z>. i?. 4., p. 365, 1902. IV.] ALDOL CONDENSATION AND FATTY ACIDS 83 mine whether it contained a branched or a normal chain, nor whether the water that went out left an unsaturated compound or one that was constructed like a lactone, though he assumed that the product was unsaturated, and therefore had the formula CHg.CcC . CH,.CH = CH.C< . CH„.C^ \0H ' \OH ' ^H Raper at the Lister Institute has also prepared a condensa- tion product of aldol. When examined as soon after isolation as possible by the cryoscopic method, this had approximately the molecular weight of a dialdol. The molecular weight found was 183; that of CgHjgO^ is 176. But it was an unstable sub- stance tending to polymerise, as after three weeks the molecular weight had risen to 278. On analysis, it showed the composi- tion of the substance CgH^^Og, mixed with a small quantity of CgHijOj. The iodine absorptions agreed rather with the lactone formula than with that of an unsaturated compound, though it took up some iodine. In order to determine whether the compound contained a branched or a normal chain, it was first converted into the acid (molecular weight from the barium salt 168.8; that of CgHi^O^ being 174), and the acid was reduced with hydriodic acid in the way in which Fischer pre- pared heptylic acid from mannose carboxylic acid. The acid was distilled in steam, and from the sodium salt the free acid was obtained, but on distillation it underwent decomposition ; this fact also probably pointing to the lactone formula, as this, like the lactone of y-oxycaproic acid, is not reduced by hydriodic acid. His work on this subject has had to be interrupted, but will, he hopes, be resumed before long, and in the meanwhile he has allowed me to refer to it. The aldol condensation is a general reaction of aldehydes, and it is not impossible that conditions may occur under which, by means of it, the fatty acids should be derived in series from acetic aldehyde. It is true that condensation products of aldehyde formed under ordinary laboratory conditions contain mixtures of many most intractable bodies, and that each 84 ASSIMILATION AND SYNTHESIS OF FAT [lect. aldehyde as it is produced will condense in this way, not only with itself, but with every other aldehyde present in the system ; it is also true that in the condensation of such an aldehyde as butyric aldehyde with acetic aldehyde in vitro leads, not to a product with a normal chain, but to one in which condensation has taken place at the a-carbon atom, thus : .H .H CHg.C — H so that at present it seems impossible to reproduce the con- ditions under which the series of even number fatty acids can be formed by aldol condensation from acetic aldehyde as the starting point. But it is just some such highly reactive body as acetic aldehyde, condensing under the guiding influence of some catalytic agent, that would best fit the data that we have for forming a conception of the manner in which fats are produced in living organisms from carbo- hydrates. Magnus Levy has shown that during aseptic autolysis of the liver the strongly acid reaction that develops is due to the formation of lactic acid, acetic acid, and butyric acid. His analyses go to show that the source of these acids is probably mainly carbohydrate, and he regards the reactions involved as typical of those by which the higher fatty acids are formed from carbohydrates.^ The particular acids that occur are sug- gestively in favour of the scheme given above, by which the lactic acid in giving rise to acetic aldehyde would lead to the formation of acetic acid by oxidation, and of butyric acid, by, in the first place, aldol condensation, and then by changes the net result of which is to transpose the /3-hydroxyl and the aldehyde hydrogen. Supposing that the synthesis of the higher fatty acids from ' Magnus Levy, H. B. 2, 261, 1902. IV.] NENCKI'S HYPOTHESIS 85 sugar is effected on these lines, then the energy equations would run as follows : — 1 g. inol. Glucose, \ / 2 g. mols. Aldehyde + 2 g. moIs.Formic acid, 677.2 Cal. /, I 2x275.5+2x61.7 = 674.4031, 2 g. mols. Aldehyde, I I i g- mol. Aldol, 1 I „ ^" .™° ". , or, tracing the same change on as far as palmitic acid : 4 g. mols. Glucose,\ /ig.mol. Palmitic acid + Sg.mols.Formicacid, 2708 Cal. I \ 2362 Cal. +494 Cal. = 2856. In the first stage of the synthesis, the reaction leading to butyric acid, the net result would be, supposing the formic acid to be oxidised, that some 160 cal., or nearly 25 per cent, of the whole energy would be rendered available for other purposes. In the later stages leading to palmitic acid, some of the energy derived from the oxidation of the formic acid would be required for effecting the synthesis, and only about 12.5 per cent, of the original amount contained in the sugar would be set free. What the organ or organs in the body are that effect this I change of sugar into fat, we have not the means of deciding. ' In Magnus Levy's experiments butyric acid was shown to be formed in the liver, and probably, as he himself thinks, from the lactic acid. Those livers in which he found that much butyric and acetic acid had been formed contained but little lactic acid, and vice versd. And the appearance of butyric acid was accompanied by the evolution of hydrogen and carbonic acid, just as it is in butyric fermentation, whether of sugar or of lactic acid.^ The conversion of lactic acid into butyric acid, a 3 carbon chain into one of 4 carbon atoms, must imply a condensation. And Magnus Levy's experiments suggest, as he points out, that under the conditions obtaining in the ' The hydrogen and carbonic acid would be provided for in the scheme suggested above by the breaking up of formic acid into these gases, a change that occurs in vitro at 1 60° C, or under the catalytic influence of certain metals, such as rhodium, at lower temperatures. 86 ASSIMILATION AND SYNTHESIS OF FAT [lect. body the liver may be able to complete the sefies of condensa- tions, which it appears to begin under the conditions of these experiments, and so carry out the synthesis of the higher fatty acids also from lactic acid. There are general reasons, too, for thinking the liver to be specially concerned in this process. In addition to its situation on the path of absorption, there is the fact that the change in question, however it be brought about, is one that involves energetic reduction, and therefore is more likely to be feasible in an organ where the larger part of the blood-supply is venous than in others where this is not the case. Siegert, however, found that there was no change in the amount of fat contained in pieces of liver that were allowed to undergo aseptic autolysis.^ Under other conditions, the total amount of higher fatty acids contained in a pulp of liver cells after incubation in a current of air has sometimes been found to be greater than before, and still greater when glycogen was added.^ A similar change has been observed by Hahn in blood, and in this case the addition of sugar made the difference more marked.^ Twenty years ago, a second source from which fat was believed to be derived in the body figured more prominently in physiological writings than the carbohydrates. It was thought that proteids were converted into fat. This was rhainly due to Voit and Pettenkofer's interpretation of their experiments, in which dogs were fed on large quantities of lean meat. It was found that all the nitrogen was eliminated, but not all the carbon. Since the amount of carbon retained in the body was larger than could be accounted for as glycogeri, they inferred that it was' retained and stored in the form of fat. A dog, for instance, was given 2.5 kg. of lean meat, which con- tained 85 g. of nitrogen, and according to their reckoning 42 g. of carbon niore than the amount excreted. This reckoning was based on the assumption that for every gramme of nitrogen 1 Siegert, H. B. i, 114, 1902. ■ 2 Hildesheim and Leathes, //. of Phys. 31, 1904. 3 Hahn, M. M. W., April 19, 1904. IV.] VOIT'S VIEW OF fAt FORMATION sf in lean meat there were ^,68 g. of carbon, whereas, as was pointed out by Pfliiger in 1891, according to the analyses of Rubner, this figure should be 3.28, For every gramme of nitrogen, therefore, they calculated 0.4 g. of carbon too much, giving an error, with 85 g. of nitrogen, amounting to 34 g. of carbon at one stroke. But even the lower figure is misleading. It is not right to assume that the whole of the carbon is in . C the form of proteid. What the ratio of ^ in the proteid of meat really is, we do not know. For myosin, according to Kiihne and Chittenden's analyses, the value of this ratio is 3.14. But myosin is by no means the only proteid in meat • it forms, according to Danilewski, less than half of the whole proteid of meat, and there are proteids in which this ratio has a still lower ratio. Besides, however lean it is, meat isl composed of cells, and we cannot know from inspection how much proteid it contains, nor how much fat or glycogen — not accurately enough to be able to say precisely how much proteid carbon is or is not retained in an experiment. Lean meat contains both fat and glycogen in larger quantities than the methods of ex- traction in use at the time of Voit's experiments were capable of revealing. In 2.5 kg. of lean meat it is not likely. that' there was less than 10 g. of fat ; the muscles of cats and rabbits after very careful cleaning contain i per cent. ; and according to the recent analyses of Pfliiger the amount of glycogen is nbt likely to have been any less. It is clear, therefore, that such experiments do not furnish a secure foundation for the belief that fats are made from proteids. The balance-sheet method cannot be finely enough adjusted to decide this question. The second corner-stone in the building up of the doctrine that fat could be derived from proteids was supplied by pathologyj Virchow, as is well known, distinguished two forms of fatty change in the cells of the principal organs of the body : a fattyi infiltration, in which fat imported from other parts was deposited' in the cells ; and a fatty degeneration, in which the fat was formed in the cells from their own protoplasm. The need for strict proof of the chemical possibility of a transformation of t8 ASSIMILATION AND SYNTHESIS OF FAT [lect. I proteid into fat was in this case hardly felt, as the niicroscopical ' appearances familiar to pathologists seemed to point so clearly to it. Moreover, the same change was believed to take place / when milk is formed in the mammary gland, or sebum in the I glands of the skin, when the brain undergoes softening, or nerve tissue in other parts degenerates, and when living \ leucocytes are converted into dead pus corpuscles : the fats of a caseating gland, of the corpus luteum, or of an encysted extra-uterine gestation were all supposed to have a similar origin, and were derived from the proteids of the dying cells, This pathological dogma that protoplasm when failing in its functions tended to turn into fat, though so generally accepted, did not pass unchallenged. In 1883 Lebedeff devised an ingenious experiment to determine the origin pf the fat in the liver after phosphorus poisoning. He had previously shown that dogs, which had been reduced by a month's starvation, and after that fed on a diet containing large quantities of linseed oil or mutton fat, fattened and acquired a fat that in the former case was fluid at 0° C, and in the latter was still solid at 50° C. ; in both cases different from normal dog's fat, but similar to I the fat of the food. If, now, such a dog were poisoned with phosphorus and died with " fatty degeneration " of the liver, upon the supposition that the fat of the diseased liver would be formed in the cells of the liver out of their own proteids, it should be similar to that in the liver of any other dog that had died of phosphorus poisoning. There is no reason why fat made from the proteids of the liver should be affected by the fact that an abnormal fat happened to be present in the subcutaneous tissue. But if the fat in the liver were imported from other parts as ready-made fat, then it should be similar to the fat stored in these parts. In a dog fattened on linseed f oil and poisoned with phosphorus, Lebedeff showed that the ^ fats obtained from the liver consisted largely of linolein. He saponified the fat, prepared the lead salts, separated those soluble in ether from those that were insoluble, and found that the former gave liquid fatty acids corresponding to 67 per cent., the latter solid acids to 23 per cent, of the fat. On treating 'V.] LEBEDEFF'S EXPERIMENT 89 the liquid acids with nitrous acid, he obtained an amount of elaidic acid corresponding to only one-fifth of these acids. The rest he argued must be linoleic acid, which does not solidify with nitrous acid.^ Lebedeffs experiment has been repeated more recently by Rosenfeld, and also by Leick and Winkler in Krehl's laboratory. The estimation of linoleic acid in liquid fatty acids by the elaidin test is unreliable, and this was a weak point in Lebedeffs argument ; for even olive oil, nearly thre^-quarters of which consists of olein, may under certain conditions yield no elaidin with nitrous acid. In the later modifications of this experiment the nature of the fat in the "degenerated" organs has been determined by the iodine absorption. Fats containing un- saturated acids like oleic or linoleic acid absorb iodine, and the amount of unsaturated acid can be determined by the percentage amount of iodine absorbed, which is called the iodine value of the fat. Thus Leick and Winkler found the following iodine values for fat from the situations indicated : ^ guboutsneous Pat. Myooar(}ial Fat. Normal Dog Dog poisoned with Phosphorus Normal Sheep Sheep poisoned with Phosphorus , 56.1 58.6 38.« 36.9 70.7 8a.9 58.2 64.1 A dog was then starved to emaciation : it was then fed on a full diet of meat with large quantities of mutton fat, and finally phosphorus was given. The subcutaneous fat after death had the iodine value 43.3, and the fat from the degenerated heart muscle 64.6. This last figure is almost identical with that found in the sheep when poisoned with phosphorus, and very different > LebedeflF, Pfl. A. 31, 11, 1883. ^ Leick and Winkler, .S". A. 48, 163, 1902. so ASSIMlLAtlOisr AND SYNTHESIS OF FAT [lkct. from that which is normal for the dog under these circumstances. Rosenfeld obtained similar results with phlorrhizine, which under certain conditions also causes "fatty degeneration" of the liver. He pointed out, too, that the liver in this case, though it may contain from 25 to even 75 per cent, of fat, does not contain more than a gramme or two less proteid than it should, and that the accumulation of fat can very rapidly be disposed of and the disturbance completely cured. This could hardly be the case if the protoplasm of the liver cells had degenerated, and this great mass of fat was formed from the dell proteids. Neither is it likely that the fat is derived from the proteids of other organs; for though there is an increased output of nitrogen, and therefore proteids are brokerl down under the ' action of the drug, the carbon excretion is also increased, owing to the severe glycosuria, and there is no retention of carbon.^ A further important point, first noted by Lebedeff iri a case of phosphorus poisoning in man, has been confirmed experimentally in animals by Rosenfeld. The patient whose case is referred to by Lebedeff was in a state of extreme emaciation when the poison took effect, and at the post-mortem examination he observed that there was no fatty change in the liver. Rosenfeld showed in a series of dogs that the amount of fatty change in the liver after phosphorus poisoning varied directly with the amount of fat in the body, and in emaciated animals there might be none appreciable. A similar relation was established also in fowls. '-''■. These experiments and observations prove that Lebedeff was almost unquestionably right, and that in these most typical forms of fatty degeneration in Virchow's sense the fat is not produced by degradation of the proteids in the cells, but imported into them from the storage places . of the body. Many attempts have been made to determine this point in another way, by ascertaining whether the total amount of fat in, the body is increased in phosphorus poisonirig. This can only be done by comparing poisoned animals with normal 1 Rosenfeld, Z.f. k. M. 36, 232, 1898 ; cf., too, M.J., p. 53, 1897 ; and, Ergeb. ii., p. 50 seq. IV.] THE FAT IN DISEASED ORGANS 91 animals, a method which can only' give uncertain results. In some cases the answer has been given in the afifirmative, but not in the most convincing. Athahasiu, for instance, in Pfliiger's laboratory took 1 24 pairs of frogs, the frogs in each pair being of equal size ; one frog out of each pair was given phosphorus, the other serving as a control. No appreciable difference in the ratio of fat to body-weight could be detected on comparing the poisoned frogs with the normal frogs.^ Kraus and Sommer, working with mice, found that phosphorus actually diminished the total amount of fat in the body, although by altering its distribution the amount of it found in the liver was increased.^ This observation is in accord with the account of the change caused by phosphorus which was given by Lebedeff and by Rosenfeld. It remains, therefore, necessary to try and harmonise this conclusion with all the other instances cited by Virchow of a similar change, by which proteids were believed to be converted into fats. As Rosenfeld points out, Virchow was led astray by microscopic appearances. Virchow speaks of the brain tissue^ in the condition known as yellow softening, entering into a state exactly similar to that of a secreting mammary gland, drops of fat being secreted in the protoplasm in each case. " When milk is manufactured in the brain instead of in the mamma, the process is a form of softening of the brain." Because no fat drops are to be seen in normal brain tissue, it does not follow that the drops when they appear are formed from proteid. But Hoppe-Seyler ^ and Walther proved that degenerated nerve fibres, though densely loaded with fat drops, really contain less fat than the corresponding normal nerves from the other side of the body, in which no drops of fat are visible at all. Similarly, Rosenfeld compared the amount of fat in a patch of softening in the occipital lobe on one side with the amount in the corresponding part of the brain on the other normal side : 18 g. of softened brain tissue, containing 3.06 g. ^ Athanasiu, Pfl. A. 74, 511, 1899. ^ Kraus and Sommer, H. B. 2, 86, 1902. » Hoppe-Seyler, V. A. 8, 127 ; Walther, V. A. 20, 426. 92 ASSIMILATION AND SYNTHESIS OF FAT [lect. solid matter, gave 36.3 per cent, of extract (6.17 per cent, of the fresh tissue) ; 16 g. of healthy brain substance with 3.26 g. of solids containing 43.3 per cent, of extract (8.8 per cent, of the fresh tissue). And even the fat in the secretion of the mammary gland, we have now the best reasons for believing, is not transformed proteid, but simply fat brought to the gland by the blood from the fat stores of the body. For, just as the fat on which an animal has been fed can be traced into the subcutaneous tissue, so it can too be traced into the milk. Cows fed on maize oilcake give butter the melting point of which is so low that it is unsaleable. If animals during or before lactation are fed on fats, the unsaturated acids of which have been saturated by iodine, these same fats are found in the milk, retaining their iodine.^ Rosenfeld fattened a bitch after a period of starvation on a diet rich in mutton fat, and then, with the beginning of lactation, fed it only on lean meat. The milk contained fat the iodine value of which was practically that of mutton fat.^ And what is true of the mammary gland is equally true of the sebaceous glands. The secretion of these glands, which Virchow regarded as the product of a fatty degeneration of protoplasm, has been proved by Plato and by Rijhmann to contain the fats which are present in the food. Sesame oil gives a characteristic colour reaction with hydro- chloric acid and furfurol. In geese that have been fed on this oil, the secretion of the sebaceous glands about the tail also gives this reaction, and when tested by determinations of iodine value, melting point, and molecular weight, the nature of the fatty acids present in the secretion is found to vary according to the fat in the food.^ But it is not only in nerve tissue and the glands which secrete fat that microscopic appearances have led to miscon- ceptions. Normal human cardiac muscle contains no fat that is visible under the microscope; but according to Rosenfeld's analyses 15 per cent, of the solid matter in it is fat. The 1 Wintemitz, H.-S, Z. 24, 425, 1898. 2 Rosenfeld, Allgem. Med. Central. Zeitung, No. 60, 1897. ^ Rohmann, H. B., v., 1 10, 1904. IV.] FATTY COMPOUNDS NOT STAINING AS FAT 93 kidney — in which fat cannot, either, be detected microscopically — contains still more than the heart, about i8 per cent, by the same method ; and a fatty kidney contains no more than a normal one. Dudgeon has made preparations of the hearts of guinea-pigs that were killed by injection of diphtheria toxin. Unlike the normal heart muscle, these preparations show very large quantities of fat when stained with scharlach. The difference is most striking, and suggests that the tissue must contain a very great deal more fat than the normal muscle, which, in fact, shows with scharlach no sign of containing fat at all. Dr Dudgeon, however, allowed me to determine the amount of fat in some of these hearts. Twelve guinea-pigs were taken, six of which received fatal doses of diphtheria toxin and died within forty-eight hours. The other six, of approximately the same size, were then killed. The hearts of the guinea-pigs that had been killed by the toxin stained intensely with scharlach, the others not at all. The results of the analysis were as follows : — Weight of Fiesh- muscle Substance. Weight, Dry and Powdered. Ether "l Extract, | Bosen- >■ fold's method, j f Per cent, of ~ 1 Sub- V stance. Extract Saponified. Insoluble ^ Patty Acids >■ from f Extract. J [ Per 1 cent, of (. stance. Hearts of Six normal Guinea- pigs Six poisoned Guinea-pigs . 6.542 10.945 1. 406 1 1.9997 0.2451 0.3954 17-43 19-77 0.1344 0.2440 9-55 12,20 These figures show that the hearts in which the toxin had set up changes contained more fats than the others ; according to the last column, about 25 per cent. more. But the dried normal hearts contained 17.43 P^i" cent. ; and yet no trace of this could be detected microscopically. In this case it is clear that the toxin causes, in addition to an accumulation of fat which is not very great, changes in the condition in which the 94 ASSIMILATION AND SYNTHESIS OF FAT [lect. fat is present in the muscle substance. That the accumulation of fat was not sufficient to account for the microscopical appear- ances is practically certain — reckoned on the fresh tissue, the increase in higher fatty acids was from 2.05 per cent, in the normal hearts to 2.23 per cent, in those on which the toxin had acted. It is difficult to believe that the intense scarlet staining of the latter was due to a substance that formed only about 0.2 per cent, of the tissue. The probability is, that as in the fatty kidney, in which Rosenfeld could detect no abnormal amount of fat, and as in the degenerated nerve tissue, the manner in which the fatty acids are combined is altered by the poison. In the normal heart, kidney, or nerve, the fatty acids are not present as simple glycerides : they are combined in such a way that they do not react histologically as fat. Under the influence of diphtheria toxin these combined fats or fatty acids are set free from the heart substance as they are in degenerating nerves, and then they answer to the histological tests for fat. If we extend the term protoplasm to include these unknown compounds of fat, such as the myelin of nerve tissue, and if we do not regard protoplasm as a synonym for proteids, then we may retain the term fatty degeneration of protoplasm for the pathological changes described by Virchow : only it must always be quite clear that the fat is fat that has been unmasked, when it is not fat that has been imported, and that it certainly is not fat that has been made in the cells by a chemical change from ordinary proteid substances. Bainbridge has described the same histological change in the liver after ligature of the hepatic artery. The normal liver tissue of cats stains very slightly with scharlach. But after ligature of the artery within forty-eight hours there are marked changes in the microscopic appearance of the tissue when treated with this reagent, the peripheral parts of the lobule particularly sfairiing intensely. The analyses, however, show that these abnormal livers frequently contain less fat than the normal ones which hardly give any histological reactions for fat. \ The evidence, therefore, for the formation of fat from the IV.] THE SOURCE AND ORIGIN OF GLYCERINE 95 proteids of the body has melted away, no less than that for the / transformation of the proteids in food into fat. But to admit ' this, is not the same as to say- that such a transformation of material cannot occur. The removal of nitrogen froih proteid substances, which was .hypothesised by the school of Voit in their account of fat formation, has become an established fact in metabolism, owing to the work of Lang and others ; and though we cannot say that the carbon compounds left after the amido acids have lost their nitrogen are retained and converted into fat, we cannot say that this can never be the case. If alanine were set free from all the compounds in which it occurs in proteids and by the loss of its NHj group were converted in lactic acid, this is the same substance that we have felt compelled to regard provisionally as the connecting link between carbo- hydrates and fats. And at any rate we do not know enough about the subsequent fate of denitrified products of proteid hydrolysis to justify a direct negation of the possibility of the formation of fat from some of them, or from substances derived from some of them. All we can say is, that what has sometimes passed as proof of the change has not stood the tests to which it has been submitted. Lastly, with regard to the origin of glycerine, there can be j no doubt that this is a substance that can be formed in the ordinary course of metabolism. Munk's experiments on the absorption of free fatty acids and their appearance as glycerides in the chyle show this. The fact, too, that fats are synthesised from carbohydrates implies that the glycerine as well as the / acids are so synthesised, since glycerine is not commonly intro-* duced into the body except in the form of neutral fats. The experiments of Luthje^ have made it practically certain that ' glycerine can be converted_jnto sugar in the body, and there is, | therefore, good ground for regarding sugar as a source from which the glycerine in the body, when not derived from the fats of the food, takes its origin. If glyceric aldehyde is the first derivative from sugar in the production of lactic acid, the change from the aldehyde to the alcohol is one that is not improbable. ' Liithje, D. A.f. k. M. 80, loi, 1904 ; cf. supra, p. 46. 96 ASSIMILATION AND SYNTHESIS OF FAT [lect. iv. J. Schmid, in three experiments on dogs treated with phlorrhizine, found that fatty acids added to a meat diet diminished the excretion of sugar.^ Since the animals had been treated with phlorrhizine for some days they probably contained but little glycogen. The fatty acids, therefore, in these cases, combined with glycerine which was derived from material which was not carbohydrate, but which would otherwise have been converted into sugar. No other experimental evidence on this point is at present available. ' J. Schmid, 5. A. 53, 429, 1905. LECTURE V THE CATABOLISM OF FAT The assimilation and physiological synthesis of fat have long occupied much of the attention of physiologists, as the familiar discussions on the formation of fat from carbohydrates and from proteids, and on the problems of obesity and pathological fatty change bear witness. The subsequent fate of fat, whether assimilated or made in the body, presented apparently fewer points of interest. It was sometimes taught that fat was principally of value as a source of heat, and that in the arctic regions a lining of fat was more effectual than furs and blankets in preventing the loss of body-heat. But in what organs the evolution of heat from fat took place, where the long fatty acid chains were taken to pieces and oxidised ; how such changes were brought about, and what functions, if any, were served by them in addition to that of supplying the body with heat — these were questions on which curiosity received but little encouragement. But it is a fact that is now generally appreciated, that the fat molecules are at least as important physiologically as the carbohydrate. In the experiments carried out on the pro- fessional starving man, Cetti, it was found that nitrogenous metabolism calculated per unit of body-weight remained fairly constant between the fifth and tenth days of starvation, ranging between 0.15 and 0.20 g. daily :^ on an average, therefore, a little over i g. of proteid per kg. daily. To convert this amount ' Zuntz and Lehmann, B. k. W., No. 24, 1887. 97 ^ 98 THE CATABOLISM OF FAT [lect. of proteid into urea, carbonic acid, sulphuric acid, and water, nearly 2 g. of oxygen would be required, having the volume 22 -^=1.4 litre, which amount, in twenty-four hours, is at the rate of nearly i c.c. per minute. But Cetti's oxygen-consumption was at the rate of nearly 5 c.c. per kg. minute ; so that four- fifths of the oxygen absorbed was required for the oxidation of non-nitrogenous substances. A gramme of oxygen yields very nearly the same number of calories whether it is used for the oxidation of proteid, carbohydrate or fat ; so that four-fifths of the starving man's energy was derived from non-nitrogenous material, much the greater part of which must have been fat during the greater part of the time in which these observations were made. Cetti was not well stocked with fat when he began this fast, and his nitrogen excretion was higher than that recorded in other cases of inanition. In dogs, Rubner and Voit reckon that during starvation only from 10 to 16 per cent, of the total energy is derived from proteids, the rest, up to 90 per cent., that is, being derived from fat. It is clear, therefore, that, at any rate in starvation, arrange- ments exist by which the greater part of the energy needed for life can be obtained from the oxidation of fats. How much of the total chemical energy transformed in metabolism is con- verted directly and in the first instance into heat, and how much primarily, if not finally, into other forms of energy, is not known ; but it is difficult to suppose that none of the work done by the heart, the respiratory muscles, kidneys, or other glands is performed by means of the oxidation of fat. Zuntz reckons that the functions of the circulation and respiration by them- selves account for from 10 to 20 per cent, of the total energy expenditure in starvation.^ It is remarkable, too, that the organs which have the most unceasingly continuous work to do 1 at. V. Noorden, Pathologische Stoffwechsel, p. 97. The work done by the heart is reckoned to be equal to about 27,000 kg. m. Supposing the efficiency of heart-muscle as a working machine to be 33 per cent., then 27,000 kg. m. work X 3 = 200 Cal.=nearly 50 g. proteid — nearly the whole proteid metabolism of the body in Cetti. v.] THE USE OF FAT IN THE BODY 99 are richly stocked with fat In the human heart muscle, 1 5 per cent, of the solids are soluble in ether, and more than half of the ether extract is composed of fat: this would be enough by itself to supply the heart with fuel for six or seven hours' work. In the kidney, according to Rosenfeld, there is still more fat than in the heart. Of the voluntary muscles the diaphragm, and in the rabbit the deeply placed red muscles, such as the soleus, pectineus, crureus and semi-tendinosus, which are believed to be especially concerned with the more protracted forms of muscular activity, differ from the paler, more super- ficial muscles in containing considerably more fat,' though less than the heart. Attempts to demonstrate a diminution in the amount of fat in the muscles during activity have been made ; both sciatic nerves have been divided, and one of them tetanised for some hours, and the fat in the corresponding muscles on the two sides compared. But no evidence for the immediate con- sumption of fat in muscular activity has been obtained in this way. In one experiment of this kind, I found that both in the gastrocnemius and the tibiales on the tetanised side, there was rather more fat, reckoned in percentage of the dry sub- stance, than on the side that had been kept at rest. Zuntz and Bogdanoff found that the effect of stimulating muscles was to diminish in them the amount of those com- binations of fatty acids which cannot be extracted directly with ether.^ It is doubtful whether the utilisation of fat in muscular activity can be proved by stimulating the muscles with the circu- lation normally maintained through them. But in other ways it has been established beyond doubt that the muscles can and do make use of fat as a source of energy. To begin with, Zuntz and his pupils have shown that muscular activity does not alter the respiratory quotient unless the work is severe enough to interfere with the oxygen-supply to the muscles. This may hold when the work done is sufficient to increase the oxygen consumption more than threefold. This great increase in the ' Leathes, Jl. of Phys., xxxi., p. ii., 1904. 2 Zuntz and BogdanoflF, D. R. A. 13, 1897 ; Pfl. A. 65, 81. 100 THE CATABOLISM OF FAT [- rate of oxidation in the body is unaccompanied by any increase, or at least any material increase, in the nitrogen output : so that the energy must be supplied by non-nitrogenous material. If this increased metabolism involved only carbohydrates, the respiratory quotient must be raised : since it is not, fat as well as carbohydrate must be made use of to supply the muscles with what they require for their work. Even in fasting animals there is only very little increase in the excretion of nitrogen when they are made to work. A dog on the sixth and seventh days of starvation was made to do work in a treadmill equivalent to climbing to the height of 1400 metres. The nitrogen excretion rose from 6 g. to 6.6. The energy for this work must have been derived in the main from fat. Zuntz, in fact, finds that fat can be used for muscular work no less economically than either proteids or carbohydrates : he determined the oxygen-consumption and respiratory quotient in a man resting and working on three different diets — one principally fat, one principally carbohydrate, and the other principally proteid — and found, as the results in the following table show, that slightly less oxygen and energy was required to do work on the fat diet than on the others. Diet principally. Besting. Worlcing. M. Kg. otWorlt done. PerM.Kg. of Worlj. c.c. Oxygen used per min. Besp. Quotient. c.c. Oxygen used per min. Besp. Quotient. c.c. Oxygen used. Oal. Fat . Carbohydrate . Proteid . 319 277 306 0.72 0.9 0.8 1029 1029 II27 0.72 0.9 0.8 354 346 345 2.01 2.17 2.38 9-39 10.41 11.35 Similarly, in dogs, even after a period of starvation followed by phlorrhizine glycosuria, when the carbohydrate stores must have run very low, work is done on the fat of the body as ■•] WORK DONE IN THE MUSCLES WITH FAT 101 economically as when proteid food or carbohydrates are abundant. In the following table this is shown. Diet, eic. Besp. Quotient. Oxygen used per M. Kg. Heat-equivalent of this amount of Oxygen. I , Proteid diet 0.78 0.57 C.C. 2.58 Cal. 2. Proteid diet, but no food given on day of experiment 0.72 0.53 „ 2.43 „ 3. As in 2, only Sugar was given before and during work .... 0.83 0.54 .. 2.58 „ 4. Little Proteid, much Starcli . 0.88 0.55 „ 2.6 „ 5. Starved : Phlorrhizine glycosuria 0.71 0-59 ,- 2-71 ,> In the fifth of these experiments certainly the work must have been done with energy derived from fat, and from the con- sumption of oxygen and the calculated expenditure of energy it is clear that it was done as economically as in the other experiments. The fats, then, can be used as a source of energy by muscles as well a;s proteids and carbohydrates, and the yield of work for a given amount of chemical energy in the form of fat is as good as in the case of either of the other kinds of material.^ Of the physiological importance of the catabolic changes in 1 It has been shown that the carbonic acid output of fowls' eggs from the ninth day of incubation on is weight for weight equal to that of the fully grown bird, that the respiratory quotient is about 0.7, and that of the 5 or 6 g. of ether extract that can be obtained from a fresh egg about one-half is used up before the chick is hatched. The heat-loss from the egg in an incubator must be considerably less than in the bird at large, although evaporation causes some loss of heat that must be replaced. The oxidation of fat, therefore, must be needed for other transformations of energy than heat-production ; the work done in the contractions of the heart is all converted back into heat, so that it is difficult with the data which are available to account for the loss of chemical energy during incubation. Bohr and Hasselbalch, M.J., p. 522, 1899 ; Liebermann, M. J., p. 234, 1888. 102 THE CATABOLISM OF FAT [lect. fat there can be no question, whatever the use made of these changes in the economy of the organism may be. But when we attempt to form a conception of the way in which these changes are brought about, we are confronted with the greatest difficulties. The hydrolysis of fats, resulting in the separation of the glycerine from the fatty acids, it is true we know something about : but this hardly touches the fringe of the subject. Until we know how the fatty acids are oxidised, we do not know the essential part of the process. Our ignorance is fundamental. And it stretches into all departments of metabolism. If we do not know how simple carbon chains are oxidised, we can form no conception of the way in which those transformations of chemical energy which constitute life are brought about in the metabolism of carbohydrates or proteids any more than in that of the fats themselves. In the case of the carbohydrates, we have already seen that it is at this point at any rate that we are brought up to a stop. For even supposing that it is right to fancy that we can trace sugar metabolism through lactic acid to aldehyde and formic acid, the energy of these products is the same as that of the sugar from which they were derived ; and assuming that the aldehyde becomes acetic acid, this would still leave 540 calories out of the original 677, or about 80 per cent, of the energy of the sugar, still latent in the acetic and formic acids ; and the whole kernel of carbohydrate metabolism would lie in the oxidation of the simple fatty acids. And the same holds for the metabolism of proteids, in so far as that metabolism is concerned in the evolution of energy from chemical combina- tions. For there is good ground for thinking that the cleavage products of proteids to a great extent lose their nitrogen com- paratively early in the course of their catabolism, and it is a fact that the removal of the nitrogen hardly affects the energy of the molecules ; consequently the crucial stage in proteid catabolism consists in the oxidation of simple fatty acids. Physiological chemistry teaches us to look upon life as the transformation of chemical energy into energy of a different nature, in conformity to the requirements of the organism. But it fails us just where its work would begin to have significance unless it can help us to v.] THE OXIDATION OF FATTY ACIDS 103 follow the processes by which the simple fatty acids are made to render up their great stores of energy. In the oxidation of fatty acids there are reasons for thinking that the problems are essentially the same whether we have stearic or butyric acid to deal with ; that is to say, that the long chains are step by step reduced in length by a process which is simply repeated at each step. It is at any rate true that the intermediate members of the fatty acid series present chemical combinations in which the energy is just as available for the purposes of the body as those commonly taken in fatty food. Palmitin is no less valuable as a food than stearin, and in Salkowski's laboratory Ludwig Meyer proved that lauric acid, CijH^^Oj, and myristic acid, Ci^HjgOg, keep down proteid metabolism just like the higher members of the fatty acid series, and, allowing for their lower potential energy and indifferent absorption, are effective foodstuffs.i From stearic acid, therefore, down to lauric acid the fatty acids are interchangeable. Of the lower fatty acids, it is known that butyric, valerianic, and caproic acids, when given with food, are not excreted in the urine; whereas, of the acetic acid contained in 25 g. of sodium acetate given by the mouth, about 10 per cent., and of the formic acid in 25 g. of sodium formate, about 26 per cent, can be recovered from the urine.^ When caproic acid is injected into the veins it is oxidised ; acetic and formic acid under these con- ditions are not.' It is also known that, while an economy of proteid effected by lactic acid can be detected, this is not the case with acetic acid : acetic acid is not completely oxidised, and, acting as a diuretic, actually increases the amount of nitrogen excreted.* But if this acid, when absorbed into the blood from the intestine, or injected directly into a vein, is not all of it avail- able as a source of energy in the body, it does not follow that when generated in the organs in the course of their metabolism it is not capable under these conditions of supplying the cells 1 L. Meyer, I/.-S. Z. 40, 550, 1904. ^ Schotten, H.-S. Z. 7, 375, 1883. ^ Scheremetjewski, B. der Sachs. Ges. der Wiss., p. 154, 1868. •• Weiske and Flechsig, M. J., p. 408, 1889. 104 THE CATABOLISM OF FAT [lect. with all its energy. In so far at any rate as the acids from stearic to lauric are all equally available as foods, it is fair to assume that the reactions by which stearic acid is made to yield up its energy are equally applicable to palmitic, myristic, and lauric acids. The fact that it is only the even number members of the series of fatty acids above caproic acid that occur in Nature, and the fact that all of these are found in milk, may possibly indicate that the demolition of the higher fatty acids is effected by the removal of two carbon atoms at a time ; just as when oleic acid is fused with potash, acetic acid is split off and palmitic acid formed. Otherwise, we may bear in mind that Le Sueur has shown that the fatty acids may be derived in succession from each other by changes in which the chain of carbon atoms are shortened by one atom at a time. The a-oxy-fatty acids, oxy- stearic no less than lactic acid, when heated to 260 to 280° C. split up into the next lower aldehyde and formic acid, or carbon monoxide and water. By the oxidation of the aldehyde the acid is obtained, which can then be converted into the a-oxy-acid and the process repeated. The reaction is the same at each step, and can begin with lauric acid or any other acid in the series equally well as with stearic acid. In that case, oleic acid may figure as the first derivative from stearic acid, and by taking up water give rise to the oxy-acid. Only it must be borne in mind that the unsaturated linkage in oleic acid is not in the a-/3 position, as was formerly supposed, from the fact that on fusion with potash it yields palmitic and acetic acids, and not in the ^-y position, as has also been thought, but is known to be exactly in the middle — ■■ CgHiy . CH : CH . CyHi, . COOH. For by more than one reaction it can be made to yield on oxidation normal nonylic acid (pelargonic acid) and azelaic acid, COOH . CyHj^ . COOH.^ But that the unsaturated linkage can be transferred under certain conditions from the centre of ' Baruch, B. 27, 173, 1894 ; and, Le Sueur,/. C. S., p. 1708, 1904. v.] OXIDATION OF THE ^-CARBON ATOM 105 the molecule to the a-^ position, seems to be suggested by the result of potash fusion. The balance between these two alternatives — the removal of two carbon atoms or of one at each step — may perhaps be affected by some very interesting results recently published by Knoop on the fate of aromatic derivatives of the lower fatty acids in the body.i When phenyl-acetic acid is given to a dog, it combines like benzoic acid with glycocoll, and is excreted as phenaceturic acid — /^ch„.co.n/ ^CHj . COOH \/ the side chain is not oxidised. Phenyl-propionic acid is oxidised in the side chain, two carbon atoms are removed, and it is excreted as hippuric acid. Phenyl-butyric acid is also oxidised in the side chain, but appears in the urine, not as hippuric but as phenaceturic acid ; here too, therefore, two carbon atoms are removed, but not the third. Phenyl-valerianic acid, however, gives hippuric acid ; four carbon atoms are in this case struck off. In each instance, therefore, the oxidation appears to be effected at the j8-carbon atom ; in the case of phenyl-valerianic acid this occurs twice, and when there is no |8-carbon atom no oxidation takes place. We cannot, of course, say that the oxidation of these aromatic derivatives of the fatty acids is necessarily worked out on the same lines as that of the fatty acids themselves. But the influence of the aromatic group must be diminished the longer the side chain is, and the fate of phenyl-valerianic acid is certainly very suggestive. If we may argue from these most interesting results, then the oxidation of stearic acid must lead to palmitic, of palmitic acid to myristic, and so on down the series ; and the reason why we can find traces of each of the even number members in this series, but none of the others, is at once apparent. The higher fatty acids are extraordinarily stable compounds. ' Knoop, H. B., vi., 150, 1905. 106 THE CATABOLISM OF FAT [lect. It is said that palmitic and stearic acids have been obtained from tombs in Egypt, where they must have been for several thousand years. Micro-organisms do not live in pure fats or oils, though funguses, especially Penicillium glaucum, grow and thrive on moist material which is sunk in oil. Apparently they make use of the oil to some extent as food, hydrolysing it, probably removing glycerine, and leaving some at any rate of the free fatty acid to crystallise out.i The changes that fats and oils undergo when they become rancid are not the work of micro-organisms. Dry, sterilised fats become rancid on exposure to air and sunlight.^ These changes, which are more prone to occur the more unsaturated acids such as oleic acid are present in the fat, consist partly in the formation of the volatile lower fatty acids, and partly in other changes the nature of which is not understood. Those fats or oils that contain the doubly or trebly unsaturated acids linoleic and linolenic acid, undergo changes which are referred to as " drying," and which result in the formation of a varnish. In this case there is a gain in weight from absorption of oxygen. But the products of the change and the nature of the reactions by which they are pro- duced have been but little studied, and certainly very little is known of them. The process of " blowing " oils, in which at a raised temperature air is blown through the oil, results in the introduction of hydrogen and hydroxyl groups where there are unsaturated linkages : for the products combine with more acetic anhydride than the original oil, combine with less iodine, and a part of the acids acquire a higher molecular weight. At the same time some of the chains must be broken, as volatile acids appear and the saponification value is raised. Changes in some respects similar to these have been observed by V. Fiirth to occur during germination in sunilower and castor- oil seeds. Part of the oil in the seeds was used up, and part of what was left was hydrolysed. But the saponification value (the amount of potash in milligrammes which is necessary for neutralising the total fatty acids free and combined in i g. of ^ Duclaux, Microbiologie, iv., 691. 2 Duclaux, A. P. /., 1887 ; Mjoen, Ch. CM., ii., 526, 1897. v.] UNSATURATED ACIDS AND OXY-ACIDS 107 the oil) was raised, and the mean molecular weight of the acids therefore diminished ; some of the acids of high molecular weight had broken down into acids of lower molecular weight. And in this process the number of unsaturated linkages and of hydroxyl groups had both diminished, indicating that these are the points in the chains where the cleavage occurs.^ Changes of a similar nature appear to occur in the fat in the body. In different parts of the body the fats certainly are different in their properties. The fat obtained from the heart muscle absorbs, for instance, considerably more iodine than that in the connective tissue. In the sheep_ 58 per cent, as compared with 38 per cent, and in the dog 70 per cent, as compared with 56 per cent. The fat obtained from the liver, according to Rosenfeld, gives still higher figures. It seems, therefore, that though the fat is deposited in the connective tissues unchanged, changes subsequently take place in it, with the result that it contains more of the unsaturated acids before it is used in the organs in which it is broken up. The unsaturated linkages become more numerous, presumably because it is at these points that the chains of carbon atoms are to break. If we could catch the process at a more advanced stage, we should find, as in the plant, that some of the unsaturated linkages had disappeared, and the mean molecular weight of the fatty acids had diminished. To sum up, therefore, on the very scanty evidence that we at present possess, we may expect to find that the fatty acids undergo oxidation step by step, each time at the /3-carbon atom ; that an unsaturated linkage is the first move towards this oxidation, and probably the formation of a saturated oxy- acid the second ; the first of these preparatory changes takes place either in the organs where the oxidation is carried out, or before it reaches them ; but after it leaves the storage places, possibly in the liver. These are problems in animal metabolism of a fundamental nature : for the present, we can hardly be more than aware of their existence. But the dim outlines miis^t become sharply ' V. Fiirth, H. B., iv., 430, 1903. 108 THE CATABOLISM OF FAT [lect. defined before we can hope to control or even to follow the course of the chemical changes that constitute either health or disease ; and the first step to this end is to be aware of them. In the meanwhile, however, quite unexpected light has been thrown on one phase of the catabolism of fat by the physiological and pathological study of the condition known as acetonuria. It has long been known that in diabetes a group of substances occur in the urine which are genetically related to each other : ;S-oxybutyric acid by oxidation gives rise to aceto- acetic acid, and this by losing carbonic acid to acetone — CHg . C< . CH2 . COOH — ^ CH3 . CO . CH2 . COOH \0H ^CH„.CO.CH, + CO, The first of these changes is peculiar to the body ; the second, which occurs probably in the bladder, is one that can hardly be prevented, unless the ester of aceto-acetic acid and not the free acid or its salts be obtained. When first these substances were found in urine there was a tendency to associate them with arrested or dis- ordered sugar catabolism, for it was in the urine of diabetic patients that they were first recognised. A little laterv'Kowever, it was noticed that they were excreted also under conditions in which no disturbance of sugar metabolism existed — in fevers, in starvation, even in healthy people on a meat diet; and the common factor in all these conditions, diabetes included, which was then believed to be the cause of acetonuria, was an increase in the rate at which proteids were being broken down. This theory also was soon found to be unsatisfactory. There is no strict parallelism between the amount of acetone bodies excreted and the rate of proteid destruction. During the first days of starvation the nitrogen excretion diminishes, the acetonuria increases. In a case of diabetes, M. Levy recorded the excre- tion in the course of three days of 342 g. of /3-oxybutyric acid, while the nitrogen excreted corresponded only to 262 g. of proteid.i , Similarly, in a case with acetonuria, Satta found that ^ Magnys Levy, i", A. 42, 149, 1899 ; and, S. A. 45, 389, 1901. v.] ACETONURIA 109 the variations in the ratio between the amounts of sulphuric acid and acetone in the urine ranged from 29 to 0.33, or 9000 per cent. ; so that, taking the sulphuric acid as a measure of the proteid catabolism, no kind of correspondence could be traced between the production of acetone and the oxidation of proteid.^ And diabetic acetonuria has been observed during nitrogenous equilibrium, and even when nitrogen was being retained in the body and no sugar excreted in the urine.^ Rosenfield, who first described the acetonuria that is constantly induced in healthy people by restricting the diet so as to exclude also carbohydrates, found that the excretion of acetone was stopped as soon as the sugar or starch was again included in the dietary.^ Hirschfeld found that it was less marked even if the diet consisted only of proteids, but was more liberally taken ; and similarly,* Waldvogel showed that if in a period of starvation a day was interpolated on which 100 g. of proteid was taken, the amount of acetone excreted fell considerably.^ And then, finally, a number of observers, first among whom was Geelmuyden, described an intensification of acetonuria as the result of adding to the amount of fat in the food." Thus, in an experiment on a healthy person the dietary consisted of 250 g. of butter, 200 g. of oil, and a little wine. This was taken, and nothing else, for five days. Intense acidosis came on ; diacetic acid, j8-oxybutyric acid, and acetone were found in the urine in amounts such as occur only in the severest cases of diabetes. On the last day these acids caused so much of the nitrogen excreted in the urine to appear as ammonia, that when the total nitrogen excreted amounted to 5.7 g., only 2.7 g. was in the form of urea, and as ' much as 2. i g. in the form of ammonia.'^ ' Satta, H. B. 6, 22, 1904. ^ Weintraud, 5. A. 34, 169, 1894. ' Rosenfeld, M.J., p. 467, 1885. * Hirschfeld, M. J., p. 565, 1895 ; cf. Rosenfeld, Cbl.f. 1. M., 1895. * Waldvogel, M. /., p. 833, 1899. ^ Geelmuyden, H.-S. Z. 23, 431, 1897 ; and, 26, 381, 1898 ; and, 41, 128, 1904. ' Landergren, Sk. A. 14, 126, 1903 ; cf. Gerhardt and Schlesinger, S. A. 42, 105, 1899. 110 THE CATABOLISM OF FAT [lect. In practically all forms of acetonuria, increasing the amount of fat in the food intensifies the condition : in that due to the exclusion of carbohydrates from the food, to diabetes in man, and to phlorrhizine glycosuria in dogs and rabbits kept other- wise without food. Under certain conditions, then, excess of fat in the food causes acetonuria; under other conditions, in which a common feature is malnutrition, in which, therefore, exces- sive calls have to be made on the fat stored in the body, the excretion of this group of substances related to acetone has been shown to be due to the metabolism of fats, since it has nothing to do with that of either carbohydrates or proteids. Here, then, we have evidence as to a part of the course which, under certain conditions at any rate, the catabolism of fat may take. Is this likely to be the course taken under normal conditions ? Traces of acetone, it is true, are constantly found in the breath and in the urine of man. But both men and animals when given j8-oxybutyric acid, or even aceto-acetic acid, normally show no signs of acetonuria ; so that we may assume that these substances in the normal organism are as completely oxidised as the fats themselves. Acetone, on the other hand, is very incompletely oxidised, the greater part being excreted by the lungs and kidneys unchanged. These facts are compatible with the hypothesis that /8-oxybutyric is a product of the normal metabolism of fat ; that, in fact, butyric acid undergoes ;8-oxida- tion, and then gives rise to aceto-acetic acid, only the aceto- acetic acid does not normally lead to the formation of acetone to any extent, but is converted presumably into acetic acid. On this interpretation of the facts the disorder in acetonuria would consist in the failure of the organism to carry out the final steps in the progressive oxidation of the fatty acids, a disorder due to the overtaxing apparently of the power of carrying out the reaction repeated at each step in the oxidation of fatty acids. The aceto-acetic acid cannot be dealt with, and either it or the acetone into which it so readily decomposes, or the substance immediately preceding it in the series of deriva- v.] ACETONURIA 111 tives from the higher fatty acids, jS-oxybutyric acid, is excreted in the urine. If we provisionally adopt this interpretation, then all the fatty acids in undergoing repeated j8-oxidation must give rise to butyric acid ; each molecule of stearic or palmitic acid to one molecule of butyric acid. And it may be noted that the experimenters, who have found that acetonuria is aggravated by the administration of fats or fatty acids, have found that it is fats such as butter, in which the mean molecular weight of the acids present is comparatively low, that produce the most marked effect; and of the fatty acids, it is the lower members of the series, and especially butyric acid, that are most active.^ If stearic acid with the molecular weight 284 acts in this way because it undergoes a process of gradual oxidative erosion until it finally becomes butyric acid, it is clear that butyric acid with the molecular weight 88 should 284 be more than three times as active, since -——, or more than 00 three times as much stearic acid, must be given to produce the same effect. The data at present available are not, of course, capable of determining exact numerical ratios of this kind. And though it may be convenient to combine and co-ordinate the facts into some relationship to each other and to the general problems of the metabolism of fat, such an attempt can only be made provisionally and for convenience, until the subject has been more investigated. And there are many striking facts connected with this subject which it is difficult to com- prehend in any simple conception of it. In the dog, for instance, simple starvation does not bring on acetonuria, as it does in man. And yet dogs treated with phlorrhizine, pro- vided that they are not kept in nitrogenous equilibrium, exhibit the phenomenon in a marked degree.^ Then, again, it is said that oleic or erucic acid exerts a much less marked influence than the saturated fatty acids.^ But, to summarise our knowledge on this point in human metabolism, we may 1 Geelmuyden, B.-S. Z. 26, 385, 1898 ; Schwarz, M. J., p. 976, 1903. 2 Baer, S. A. 51, 271, 1904. ^ Schwarz, loc. cit. 112 THE CATABOLISM OF FAT [lect. say that acetonuria is brought on by those conditions in which the fat of the body or of the food has to be taken into use to an unusual extent, whether because the body has lost the power of utilising sugar, as in diabetes, or because there is no carbohydrate available. In addition to the substances from which acetone is derived, sugar, according to a view that has been coming more and more into prominence in recent years, may also be derived from fat in the course of metabolism. It was observed by Voit, and repeatedly since by others, that the amount of oxygen taken up by animals, and even man, could not all be accounted for in the final products of combustion ; the respiratory quotient when oxidation is carried out to completion cannot sink below 0.7. If fats which contain about 1 1 per cent, of oxygen could in the course of oxidation give rise to substances out of which sugars containing more than 50 per cent, of oxygen could be formed, and these sugars stored in the body, then clearly respiratory quotients lower than 0.7 must occur. They have been observed ; figures even lower than 0.3 have been recorded in hybernating animals. Under these conditions the weight of an animal taking no food may, paradoxically, rise steadily for some time. This of course does not prove that sugar is formed from fat, though it certainly shows that substances richer in oxygen are being formed from substances that are poorer, and that the combustible material in the body when it is oxidised does not suddenly and in one explosive reaction collapse into carbonic acid, water and the other final products of vital oxidation. The changes occur in stages, and though usually the conditions permit of each stage following its pre- decessor without interruption, such interruption may in certain circumstances occur at some intermediate stage. It is believed that the corresponding phenomenon, which was observed by Wiesner in plants the seeds of which contain oil, and cause during germination over mercury a diminution in the volume of the gas in the vessel in which they are contained, is correctly explained as being due to the formation of starch from oil.^ 1 Cf. Seegen, Pfl. A. 39, 140, 1886. v.] LOW VALUES OF R.Q. : HIGH VALUE OF D : N 113 But it is only in recent years that evidence has been collected for the view that sugar is formed in the animal body from fat. Some have maintained, especially Rumpf, that the sugar excreted in diabetes is in part derived from the fat of the body. Working under him, Hartogh and Schumm have shown that if dogs are made to work hard without food for some days, and then fed liberally on fat for a week or more, then again made to work and finally treated with phlorrhizine, the glycosuria that is brought on may become very intense. One animal excreted as much as 145 g. of sugar in twenty-four hours. It is difficult to believe that after the drastic preparatory treatment sufficient glycogen was left in the tissues to produce such quantities of sugar as they observed, and the amount of sugar was greater than could conceivably have been derived from the proteid broken down in the same period. If every carbon atom in the proteid, except those that are excreted as urea, were used for the manufacture of sugar, the sugar pro- duced could not weigh more than seven times the amount of nitrogen excreted. In these experiments the sugar excreted in one period of five days amounted to as much as nine times the weight of the nitrogen found in the urine.^ No one can suppose that even 7 is an actually possible value for the ratio of the sugar derived from proteid to the nitrogen of that proteid. The value found by Minkowski for this ratio, ^, in dogs after removal of the pancreas, was 2.85. On the con- stancy of this ratio in his experiments he bases his view that the sugar is derived in these animals from the proteids. But in human diabetes, and also in dogs after . removal of the pancreas, values even higher than that found by Hartogh and Schumm have been observed — 11, and even up to 12.2.^ In such cases it is physically no less than chemically impossible . that proteids can have supplied the material for the formation of the whole of the sugar. ' Hartogh and Schumm, S. A. 45, 17, igoi. 2 Rumpf, B. k. W., No. 9, 1899; Luthje, D. A. f. k. M. 80, 98, 1904. H 114 THE CATABOLISM OF FAT [lf.ct. Pflugeri has recently given strong expression to his belief that, when sugar is produced in the body which cannot be accounted for by the carbohydrates of the tissues or the food, it must be derived from the fats and not from the proteids. In an earlier paper he had argued against all the evidence that has been adduced for the belief that proteids give rise to sugar in the course of metabolism ; showed that there were many flaws in this evidence, and that it did not irrefutably prove what it was designed to prove ; and finally concluded, not that the conversion of proteid into sugar was not established, but that it never occurs. Accepting this conclusion, then, his new experi- ments certainly prove that sugar is formed from the fats. " He removed the pancreas in dogs by Sandmeyer's operation,^ that is to say, excised the greater part, and severed the connection of the remainder with the intestine, a procedure which enables the animals to live for much longer than they do if the gland is entirely extirpated, and causes sooner or later a glycosuria which, at first moderate, subsequently becomes intense. The addition of raw ox-pancreas to the food of such animals improves the absorption of both fats and proteids, and at the same time greatly increases the glycosuria. Pfliiger's dogs were fed for several months after the operation on a preparation of casein (nutrose) and boiled cod, the flesh of which contains, in the winter months when his experiments were carried out, no glycogen and only traces of fat. One of the dogs on this diet excreted in the course of two months more than 3 kg. of sugar. The largest amount of glycogen ever found in a dog amounted, when reckoned as sugar, to 4.1 per cent, of the animal's body-weight. Supposing that in this experiment the dog, when it was put on the nutrose and boiled cod diet, and when it weighed 10.3 kg., had this maximum amount of glycogen in its tissues, even then 422 g. of sugar is the most that can possibly be accounted for by the carbohydrates of its tissues. More than 2| kilogrammes of sugar must have ^ Pfliiger, Das Glykogen, 2nd edit., Bonn, 1905 ; and, Pfl. A. 108, 115, ■1905 2 Z.f B. 31, 12, 1894. v.] PFLOGER derives sugar from fat 115 been produced from material other than carbohydrate : either the proteids of the food or the animal's own fat. Either of these could have supplied sufficient carbon ; for the proteid food was abundant, and was taken well ; and the fat in a dog may amount to 46 per cent, of its weight, and on the fourth day of starvation in pancreatic diabetes, Pfluger has found as much as 26 per cent, of fat in a dog. If all the carbon in fat were available for sugar formation, then 100 g. of fat could yield 192 g. of sugar; but according to Pfliiger's conception of the reaction by which fat yields sugar in the body, 130 g. would be formed from 100 g. of fat; the 3 kg. of sugar, therefore, would require 2.3 kg. of fat, which is not more than the dog weighing 10.3 kg. may have had. The dog certainly was reduced to a skeleton by the end of the experiment, and after death had no visible fat in its connective tissues. During the greater part of the experiment the dog was in nitrogenous equilibrium, gaining about 25 g. of nitrogen in forty- four days. But there was no constant ratio between the sugar and the nitrogen excreted. For thirty-four days the average value of this ratio was about 2.27. But it was much lower than this at first, and fell off again towards the end, several days before the power of assimilating proteid food showed signs of failing. Pfluger believes that the sugar was derived from the fat of the body. And the seat of this transformation he believes was the liver. For although the animal was as wasted as if it had died of starvation, and had lost 40 per cent, in weight, its liver weighed nevertheless as much as it probably weighed at the beginning of the experiment. Like the heart and the brain in ordinary starvation, he argues it had been active up to the end; and this activity consisted in the conversion of fat into sugar, a form of activity of which the liver is capable, he thinks, at all times. The commonly observed fact, that both in men and animals glycosuria may be increased by a more abundant proteid diet, he explains as due to the stimulating action exerted by proteids on all the functions of the liver, including this one by which fat is converted into sugar. Besides proteids, other 116 THE CATABOLISM OF FAT [lect substances influence the activity of this function in the same way — amido acids, ammonium carbonate, adrenaline, carbon monoxide, phlorrhizine, etc. In this experiment of Pfluger's one fact stands out clear and incontestable, that sugar is synthesised in the body, and, under the conditions observed, on a very extensive scale ; the material used in this synthesis can have been derived only from the fat of the body or the proteid of the food. But if it has not been strictly proved, as Pfliiger maintains, that proteids ever supply what is necessary for this synthesis, neither does this experiment positively prove that fat does so. It proves it only on the assumption that proteid cannot serve as a source of the necessary material, and that has not been proved. It is no easier to conceive a solution of the problem presented by the transformation of fat into sugar than of the change in the opposite direction, and it cannot be more difficult. The suggestion made by Pfliiger is that the carbon atoms in stearic acid are oxidised one at a time, and in order beginning at the terminal methyl group, by the intervention of ammonia. When the sixth carbon atom is reached the chain breaks, giving the aldehyde character to this atom of carbon, and removing the seventh as carbonic acid. This process repeated leaves the four terminal carbon atoms in the form of butyric acid, as is required to account for the /3-oxybutyric acid and its derivatives, which are known to arise in the metabolism of fat. Nothing analogous to this can be quoted, though the final result is of course the one needed for the theory. If the fatty acids furnish material for sugar synthesis there would be less difficulty in imagining some simple carbon compound being formed in the oxidation of the fatty acids, and condensing to give rise to sugar molecules some simple aldehyde such as those that are known to be capable of such condensation — formic, for instance, or possibly glycollic or even glyceric aldehyde. Before leaving the review of the problems presented by the destructive metabolism of fat, there are certain disorders to the v.] FAILURE OF FAT CATABOLISM 117 study of which we may look as likely to prove as important in this connection as the study of diabetes, in its bearing on the destructive metabolism of sugar. In what is still known as fatty degeneration we have now learnt to see not a degeneration in the sense originally intended, not, that is to say, a degrada- tion of the proteids of the cells resulting in their conversion into fat. We have learnt that the excess of fat in the degener- ated heart, as in the degenerated liver, is fat that has been imported from the storage places for fat in the connective tissues. The current view of the essence of the disorder in such cases seems to be that an excessive importation of fat has taken place ; that it is not, therefore, the degenerated organs themselves that are primarily at fault, but the tissues which should hold fat in reserve, or the blood that transports it, which have conspired together, as it were, to dump fat in otherwise healthy organs, and so paralyse the normal traffic of those parts. And yet it is hard, when we find a heart or liver the seat of fatty degeneration, to give up the idea that the disease has struck at these tissues themselves in a vital function, and it is not easy to assent to a view that looks upon them merely as the victims of obesity. Neither is it necessary. It is more compatible with our general conceptions of physiological pro- cesses and their disorders to suppose that since fat is the most valuable source of energy that we have in our bodies, it should be made use of in those organs whose needs are greatest and most insistent. We have seen that there is much evidence that fat is an available source of energy for the heart and other muscles ; the more constant the activity of a muscular organ, the larger the amount of fat we find in it. It is not unreason- able, then, to suppose that it is normally provided that there should be a regular system of supply of fat from the connective tissues by the blood-stream to such organs. But in order to balance this supply there must be a certain activity in the chemical processes by which the fat is made to yield up its energy in those organs. If the chemical reactions by which this result is brought about — the catabolic changes which liberate energy by the demolition of fatty acids — fail, while the supply 118 THE CATABOLISM OF FAT [lect. of fat by the blood-stream is maintained, then the result must be an accumulation of fat — of imported fat — in the cells of the tissues that normally use it up as fast as it is imported, but now no longer can. Phosphorus, or arsenic, or diphtheria toxin, for instance, may act as negative catalysing agents and retard the rate at which a reaction normal to the cells takes place, and so by attacking a vital function poison the cells by robbing them of their power of using what may in some conditions at any rate be their principal source of energy. The application of such a conception as this to the liver, it is true, is less obvious. The liver is the great riddle of the body ; but we do not look to the liver as to an organ in which transformations of energy are to be most conspicuous — chemical transformations certainly, but changes involving in the main but little liberation of energy. It is not likely that the most valuable source of energy in the body should be extensively consumed in the liver. If, therefore, fat is brought to the liver in large quantities, as undoubtedly appears to have been the case in livers that are the seat of fatty degeneration, it is brought there to undergo changes of a different kind from those that are its fate in a working organ like the heart. It is possible, of course, that these changes involve reactions which, as well as the quite different reactions in the heart, are paralysed by the same poisons. But it is also possible, and perhaps easier, to look on the accumulation of fat in the liver as a kind of fat congestion. If we suppose that the fat from the connective tissues has to undergo some preparatory treatment in the liver before it is adapted to the requirements of the organs in which it is actually oxidised and consumed, then as those organs no longer use the fat brought to them by the blood, the blood will have an in- creasing difficulty in getting rid of the fat that it carries from the liver, and be less able to remove the altered fat from the liver; although, so long as the liver is able to take up the unaltered fat brought by the blood from the connective tissues, the blood may still be able to carry the unaltered connective tissue fat. There are many conditions which are accompanied v.] FATTY DISEASE OF THE LIVER 119 apparently by an increase in the fat in the liver which tempt one to look on this fatty change as a sign that the stored fat is being called up at a greater rate than usual. Rosenfeld has especially studied the fatty liver that he finds in dogs that have been starved for a week and then treated for two days with phlorrhizine. The animals thus prepared are prevented by their glycosuria from making use of what is left them, after their week of inanition, of their stock of carbohydrate, and the fat has to be made use of Pancreatic diabetes is accompanied in the same way by a fatty liver, and Pfliiger's dog fed for months on pure proteid food, having used up almost all the stored fat of its body still had a normal amount of fat in its liver, 1.7 per cent, of the fresh organ, or 11.2 percent, of the dry substance : its last non-nitrogenous reserves had been called up, and were being prepared in the liver for the use of the tissues. This, too, is how Pfliiger regarded it ; only he goes further, and considers the preparation of the fat to consist in its conversion into sugar. In the discussion of the nature of fatty degeneration and its bearings on the physiological problems of the meta- bolism of fat it must not be forgotten that there are certain phenomena included under this general term which require a different interpretation from those we have considered so far, an interpretation which is perhaps not very different from that which was in Virchow's mind when he introduced the term fatty degeneration. The nerve sheaths do not contain free fats, though they contain complex combinations of higher fatty acids and glycerine with other substances. They do not react with the histological reagents in common use for staining fats, not even with osmic acid, if by the use of fixing agents they are prevented from undergoing the changes which they other- wise slowly undergo in the presence of osmic acid. When the nerves degenerate after division or in disease, free fats appear in them, and can be at once demonstrated by any of the re- agents for fats. In the myelin, therefore, fat exists in the form of compounds, of the nature of which we know very little but that 120 THE CATABOLISM OF FAT [lect. they do not react like ordinary fat. In the heart and the kidney, too, the considerable quantities of fat that can be shown to be contained in the cells are normally combined in such a way as not to give the histological reactions that characterise free fats. It is well known, too, that this fat is but very imperfectly removed from the dried and powdered tissue of these organs by extraction with solvents for fats. In or der to extract theJ at completely, eith er digestion with "ge|)sm and hydrochloric asid, or boiling wTtn alcohol, or solution and saponification Jn^stxa»g " caustic "alKaiiS inQli^HO^ By "ffGeration of the fatty adds by means of a mineral acid, must "pre^aS'^'ffie extraction. In the heart, however, which has undergone fatty degenera- tion, as the result for instance of poisoning by diphtheria toxin, the tissue gives the histological reactions of fat, just as the degenerated nerve does. And though in the heart there may be some increase in the amount of fat present, and therefore there may be a question whether the histo- logically demonstrable fatty change is not all due to imported fat which is in excess, in the degenerated nerve, and, according to Rosenfeld, also in the degenerated kidney there is no abnormal amount of fat, and may be even less than is normal ; as was found in the case of the " fatty " liver too, that results from ligature of the hepatic artery, to which reference was made in the last lecture. In such cases the essence of the disorder is clearly not an accumulation of imported fat, and must consist in a change in the mode of combination of the fat. This change we may class with the autolytic changes by which proteids are broken up in cells, and, as was pointed out in the last lecture, may refer to as degeneration of protoplasm, so long as it is clear that by proto- plasm is not meant merely the proteids of the cells, but other complex components of living matter as well, into the composi- tion of which, for all that we know, proteids may not enter in any way at all. Such considerations seem to point to the idea that the catabolic changes in fats are brought about, not in the free simple glycerides of higher fatty acids, but in complex com- v.] FATTY DEGENERATION 121 binations of these glycerides with other substances such as we know exist in the medullary substance of nerves, though even there we know as yet only very little concerning their exact nature and constitution, or their relations to the rest of the substance which we refer to as protoplasm. LECTURE VI THE ASSIMILATION AND SYNTHESIS OF PROTEIDS Since the first days of modern physiology, when the group of substances known as proteids were first recognised, down to comparatively recent times, the well-known fact that proteids are a necessary part of the food of all animals was interpreted to mean that no synthesis of proteids occurred in the animal body. It was common to draw this distinction between plants and animals : plants, with hardly an exception, live without proteid food, and must therefore make their own proteids for themselves ; animals find their proteids ready-made in their food, and therefore need not and do not make them. This implied that all the proteids in existence are the product of vegetable life. Many of the names given to vegetable proteids show how this implied conception, though not always expressed, was always present : vegetable " fibrin " and vegetable " casein " occur as well as vegetable " albumins " and " globulins," names given on account of the similarity of certain superficial physical properties of these proteids to those of the homonymous sub- stances of animal origin, and with no regard whatever to chemical constitution. But as the differences between different kinds of proteids came to be more clearly appreciated, it was obvious that no proteid found in any animal was the same as any proteid found in any plant"; and we may now probably go further, and say that no proteid found in any species of animal is identical with any proteid found in any other species. Even the haemoglobin found in different species of animals 122 LECT. VI.] SPECIFICITY OF ANIMAL PROTEIDS 123 presents differences in its properties which point to differences in constitution. The crystal form, though very generally similar, is different in some species, and the solubility even when the crystal form is the same, as in man, the horse, and the dog, may still be very different. The solubility of haemoglobin from the dog is given by Hoppe-Seyler as 2 parts in lOO of water at 5°C. Dogs' blood contains roughly from 12 to 15 per cent, and when laked, spontaneously yields crystals of haemoglobin in the cold. Human blood, containing about the same amount, treated in this way, does not crystallise. The solubility of human haemoglobin has not been exactly determined, but is certainly greater than that of the haemoglobin of dogs' blood. The analyses of hemoglobin obtained from different species also show that the substance is not one and the same in all animals. The differences are probably confined to the proteid part of the molecule, haematine being always the same sub- stance, so far as is known. Two atoms of sulphur are found for one of iron in the haemoglobin of the horse, ox, and pig, but three in that of the dog. The other proteids of the blood also present differences in their properties in different species. The serum-albumins can be crystallised with very different degrees of readiness ; the specific rotatory power of corresponding proteids prepared in the same way from the blood of different species is notably different. There is ground, too, for thinking that the specific immunities and specific reactions of the blood of different species are properties residing in the blood-proteids, and depending upon their chemical constitution. The chemical differences that have been brought to light in the composition of the pro- tamines obtained from the sperma of different species of fish illustrate the sort of differences that may be expected to exist on a larger scale in the larger molecules of the true proteids, such as the serum-albumins or globulins found in different species. But however that may prove to be, enough is already positively known to show that the proteids taken as food cannot find a place in the economy of the animal body till they have been, as it were, melted down and recast. 124 ASSIMILATION AND SYNTHESIS OF PEOTEIDS [lect. At the same time, our notions about the meaning of the changes undergone by proteids in digestion have been pro- foundly modified. It was formerly supposed that proteids being indiffusible colloids could not be absorbed, until they had been converted into diffusible peptones, and that this was the special purpose for which digestion was necessary. But, in the first place, it has been shown that solutions of even such proteids as serum- and egg-albumin are absorbed from a loop of intestine which has been washed till no trace of pepsine or trypsine can be detected in the washings. These albumins are among the most typically indiffusible members of the class of proteids, and yet they may be absorbed, if introduced into the bowel in solution, although they are not digested. In one of the experiments as much as 90 per cent, was absorbed, and the average extent of absorption was 22 per cent.i And, secondly, it must be remembered that almost all the ordinary forms of proteid in human food are insoluble, coagulated in cooking, or in the case of the cereal proteids, insoluble even if not baked or boiled. The apparent exception, casein, is rendered insoluble in the stomach. The digestive action of the gastric and pan- creatic excretions is, therefore, before all things necessary in order to convert these insoluble proteids into something soluble, and so capable of passing into the mucous membrane. This is, then, the first point in which our interpretation of the physiology of proteid digestion has been modified. The second is, that peptones and albumoses are not the only sub- stances formed from proteids in digestion. That this is true of the digestive action of the pancreatic ferment, trypsine, has been known for nearly forty years. But Kuhne, who first showed that leucine and tyrosine are formed when this ferment is at work,^ underestimated the extent of the demolition of proteids effected by it. He believed that one-half of the proteid mole- cule resisted its action entirely. It is now known that this is not so ; that trypsine can do with proteids nearly all that boiling mineral acids can ; that all the known products of acid proteo- 1 Friedlander, Z.f. B. 33, 1896. 2 Kiihne, V. A. 39, 1867. VI.] DIGESTIVE CHANGES IN PROTEIDS 125 lysis are with very few exceptions ^ liberated also by trypsine ; and that, finally, nothing may be left, where trypsine has been at work, that can be called in any sense of the word proteid at all.^ But more than this, we have learnt that the action of pepsine does not, either, stop with the formation of peptones. It has been found, for instance, that when crystalline serum-albumin or casein are digested with pepsine and hydrochloric acid, not only are substances containing nitrogen formed which do not give the biuret reaction, but after so short a time as two hours, more than so per cent, of the nitrogen originally contained in the proteid may appear in some form in which it is not precipitated by phosphotungstic acid — in a form, that is, which is certainly not proteid, not even peptone — and this 50 per cent, can include only the mono-amido acids, the basic products presumably set free at the same time being included, together with the albumoses and peptones, in the precipitate.^ A considerable number of the common products of extreme proteid hydrolysis have been isolated from the fluids in which pepsine has been acting — leucine, amido-valerianic acid, aspartic and glutamic acids, cystine, tyrosine, and lysine ; and besides these, certain peculiar substances formed from these, in part by the loss of COg, as in the case of pentamethylene diamine from lysine, tetramethylene diamine from ornithine, and oxyphenyl ethylamine from tyrosine, and lastly leucinimide, an anhydride of leucine.* It is clear, therefore, that pepsine also can destroy to a great extent the proteid character of the substance on which it acts. But in addition to what has been recently learnt with regard to the action of trypsine and pepsine, the discovery of the ' Cf. infra, p. 132. ^ Kutscher, H.-S. Z. 25, 195 ; 26, no ; and, 28, 88, 1898 and 1899. * Zunz, H.-S. Z. 28, 146, 1899 ; Pfaundler, H.-S. Z. 30, 90, 1900. * Lavroff, H.-S. Z. 33, 312, 1901 ; Langstein, H. B. i, 507, and 2, 229, 1901 and 1902 ; Salaskin, H.-S. Z. 32, 592, 1901. With regard to the forma- tion of diamines from diamido acids, I have failed to confirm Lavroffs statement. The mucous membrane of dogs' stomachs autolysed with HCl for two days and filtered, was treated with crystalline pure lysine bichloride. After two months, by means of phenyl isocyanate, no cadaverine could be detected. 126 ASSIMILATION AND SYNTHESIS OF PROTEIDS [lect. ferment, erepsine, has also contributed to the necessity for a revision of our views on the meaning of proteid digestion. This ferment does not act on all proteids alike : it acts upon albumoses and peptones, and not upon other proteids, with the possible exception of casein and fibrin.^ And when it acts it gives rise to substances which do not give the biuret reaction, amongst which tyrosine, leucine, lysine, arginine, and histidine have been identified. It is found in the succus entericus ; so that solutions of Witte's peptone introduced into a washed loop of intestine are in an hour or two transformed into solutions in which a far larger proportion of the nitrogen escapes precipitation with tannic acid than was the case before. But according to Cohn- heim the amount of the enzyme found in the mucous membrane itself is much greater than that in the succus entericus, and it is mainly therefore on the albumoses and peptones that have been absorbed into the intestinal cells that he supposes it to act.^ Erepsine, it is true, is a ferment, on the nature and even existence of which there is some difference of opinion. Cohnheim demon- strated the presence of this ferment in the intestine by washing the mucous membrane free of pepsine and trypsine. But Embden and Knoop deny that this is possible. They found that pieces of intestine obtained from normal animals, however much they were washed, did undergo digestive changes when kept at body temperature for from one to three hours; the nitrogenous substances that could not be removed by heat- coagulation increased. But this they ascribed not to the action of intracellular ferments, such as erepsine, but to the presence of trypsine in the recesses of the mucous surface, from which no amount of washing can remove it. For if, a week before testing the intestine in this way, the pancreatic ducts were ligatured so that no trypsine could reach the gut, then these changes could not be detected, nor could any alteration be made out in the intensity of the biuret reaction due to albumoses or peptones.^ 1 Lambert, C. R. S. B. 55, 418, 1903 ; Embden and Knoop, H. B. 3, 127 ; Kutscher and Seemann, H.-S. Z. 35, 433, 1902. ^ Cohnheim, H.-S. Z. 35, 136, and 36, 13, 1902. ' Embden and Knoop, H. B. 3, 120, 1902. VI.] THE FINAL PRODUCTS OF DIGESTION 127 It may be questioned whether this necessarily proves that the changes to be observed in the normal intestine were due to adherent trypsine : for after this operation the intestine was abnormal, not only in containing no trypsine, but also in other ways, and the cells may not have contained the enzymes which are present in cells from a normal intestine. Kutscher and Seemann isolated a loop of gut by means of a Thiry-Vella fistula, and found that the secretion obtained from the loop thirty days after the operation had the power of digest- ing deutero-albumose as well as boiled fibrin. But they calcu- late, by methods which are not altogether unexceptionable, that not more than s per cent, of all the proteid digested by the animal could possibly be accounted for by the action of this intestinal enzyme, and conclude that erepsine plays an insignifi- cant part in digestion in the lumen of the bowel ; and that it does not act within the cells, they argue from the fact that during digestion they could isolate no leucine or other crystal- line derivative of proteid, either from the portal blood or from the mucous membrane. In their account of proteid digestion, however, what erepsine loses in significance trypsine gains, so that the net result is the same, and they, like Cohnheim, think that the hydrolysis of proteids in digestion is complete ; the substances actually absorbed are not the proteids themselves, but the simple crystalline derivatives, amido acids of all kinds.i What is known, therefore, of the action of each of these three ferments, pepsine, trypsine, and erepsine, all points to the probability that the demolition of proteid molecules in digestion is far more complete than used to be supposed. It has often been argued that, although trypsine is capable of splitting off amido acids from proteids, it does not actually do so in normal digestion : the grounds on which this belief is generally based are principally two. In the first place, it is said, if the amido acids are set free by trypsine in the bowel, it should be possible to find them there when digestion is in progress. And Schmidt Miilheim searched repeatedly for leucine and tyrosine in the intestine of dogs during digestion, and found only traces or ' Kutscher and Seemann, H.-S. Z. 35, 432, 1902, 128 ASSIMILATION AND SYNTHESIS OF PROTEIDS [lect. none.^ MacFadyen, Nencki, and Sieber also could find none in the discharge from an intestinal fistula in a patient who had had a strangulated hernia involving the ileo-colic valve.^ But Kiihne,^ on the other hand, obtained about 3 decigrammes of tyrosine and the same amount of leucine from the duodenum of a dog into which he had introduced, four hours previously, 20 grammes of fibrin. And even earlier than that, Kolliker and Miiller found leucine and tyrosine in the small intestine during digestion, much more in the upper parts than the lower, and never any in the large intestine. They thought they came from the pancreatic juice, in which they had shown that these substances were present. And recently Kutscher and Seemann found not only leucine and tyrosine constantly, but also lysine and arginine. It is true that the quantities which are obtained of these substances are usually small ; but, then, if they are formed they are formed for absorption, and not in order to accumulate in the intestine, so that the amount formed cannot be judged from the amount left unabsorbed in the bowel at any particular moment. The other principal ground on which this belief is based is a teleological one. If proteids are necessary for the nutrition of the body, it cannot be supposed that any large quantity of this indispensable material will be destroyed in digestion. Kiihne himself regarded that part of the proteid which was broken down into leucine and tyrosine as so much wasted proteid. But it is not clear why we should look upon leucine and tyrosine as wasted material. They are both of them when absorbed as completely oxidised as the proteids them- selves, and, weight for weight, as sources of energy they are each of them worth more than proteids. A gramme of leucine yields on combustion 6.5 Cal., a gramme of tyrosine 5.9. Even glycocoll, of all the carbon compounds which are split off from proteids the one with the smallest molecule, still contains chemical energy equal to 3 Cal. per gramme. The energy con- tained in a given weight of a substance does not depend on ^ Schmidt Miilheim, D. R. A. p. 39, 1879. ^ Nencki, MacFadyen, and Sieber, 5. A. 28, 311, 1891. ^ Kiihne, V. A. 39, 155, 1867. VI.] THE ENERGY EXCHANGE IN PROTEOLYSIS 129 the number of. atoms aggregated into the molecules of the substance. The other hydrolytic changes effected in digestion leave the sum of chemical energy practically unaffected. The energy-value in Calories of a gramme-molecule of maltose, cane sugar, or lactose, for instance, is in each case just over 1350 Cal. : that of the two gramme-molecules of dextrose formed from one of maltose is 1347.4: that of the gramme-molecule each of glucose and fructose from cane sugar is 1 349.6, and of galactose and glucose from lactose is 1343.6. In the two first cases the difference comes to about 3 Cal., in the last to nearly 8 Cal. Even this difference is hardly appreciable. In the hydrolysis of ethyl butyrate the exchange is 850.1 Cal. in the alcohol and acid for 851.3 Cal. in the ester from which they are derived. In the case of the proteids and the substances derived from them by hydrolysis, a direct determination was in part made by Rubner for O. Loewi.^ A pancreatic digest, freed by filtra- tion from the leucine and tyrosine that had crystallised out, was dried and the residue powdered : i g. of the powder was found to yield 4.6 Cal., which, as Loewi points out, is only 10 per cent, less than the amount given by i g. of dry powdered meat; and since considerable quantities of leucine and tyrosine with high heat equivalents had been removed, this fact must account for a great part of this difference. But besides this, the figure given for meat is for a mixture of chemical substances, in which some fat at any rate must have been included, whereas the filtered acid pancreatic digest must have lost in addition to most of its leucine and tyrosine all its derivatives of higher fatty acids. The margin left, therefore, for a difference between the energy-value of the proteids and that of the substances derived from them in this experiment seems to be very small. No other more direct determinations of the energy exchange in the hydrolysis of proteids appear to have been made. But everything points to its being inconsiderable. It is a mistake, therefore, to argue that if the value of proteid food as a source of energy is not to be sacrificed, proteid digestion cannot be carried beyond the peptone stage. There is no ground for ^ Loewi, 5. A. 48, 328, 1902. I 130 ASSIMILATION AND SYNTHESIS OF PROTEIDS [lect. thinking that changes that are merely hydrolytic can materially affect the sum of energy. But, more than this, we have actual experimental proof that the final products of tryptic digestion, amongst which no detectable trace of anything that cain be called proteid is left, may nevertheless serve in the place of proteid in the food of an animal. The first experiments establishing this were those of Otto Loewi. Dogs were fed on the solution of pancreas obtained by allowing the glands to undergo complete autolysis in the presence of chloroform. This solution was filtered and a quantity of crystalline products removed, and the filtrate, which gave no trace of a biuret reaction, was administered, together with starch and fat, to make up the requisite sum total of Calories in the animal's food. The most successful experiment was one in which the dog took in the course of eleven days 66.8 g, of nitrogen in this form, and on each day the amount of nitrogen excreted was less than that ingested, so that altogether 9.8 g. were retained in the eleven days, or a daily average of 0.9 g., while the animal's weight went up from 1 1.9 kg. to 12.9 kg.^ More recently still, Henriques and Hansen have carried out similar experiments on rats, and obtained similar results. Animals were kept in nitrogenous equilibrium for twelve to fourteen days while being given no proteid, but in place of this the substances contained in a digest of ox-pancreas and dog's intestine, from which, under the influence of the enzymes present, everything that could give the biuret reaction had disappeared. But in addition to this, two further points of no little interest were determined. If the digested fluid was precipitated with phosphotungstic acid and filtered, the nitrogenous substances in the filtrate were still sufficient to keep the animals in nitrogenous equilibrium. In this case the basic substances, and also the polypeptides according to what is at present known, must have been absent from the food. In other experiments the digest was treated repeatedly with 96 per cent, alcohol at 50° C, and so separated into two portions, one soluble, the other insoluble in alcohol; and on ' Q. Loewi, 5. A. 48, 303, 1902. VI.] NITROGENOUS EQUILIBRIUM WITHOUT PROTEIDS 131 feeding the rats with these it was found that the substances soluble in alcohol were efficient substitutes for proteid, but not those that were insoluble in alcohol. But though the substances formed by the hydrolytic action of the enzymes of the pancreas and intestine were shown under these three different sets of conditions to contain all that is necessary for maintaining nitrogenous equilibrium, if the substances formed from casein by the hydrolytic action of mineral acids were substituted for these, then the animals lost nitrogen, and lived no longer than others that were fed on food containing no nitrogen.^ This remarkable result has been obtained by Abderhalden and Rona as well, in both mice and a dog.^ There is after all something apparently that the enzymes leave intact but that acids destroy, which, whatever its nature may be, is certainly not in any sense proteid, but is nevertheless indispensable in the synthesis of proteid. The presence or absence of this just makes all the difference in the food-value of the mixture of nitrogenous compounds given in these experiments. What is the nature of these essential substances which escape hydrolysis by trypsine, but are destroyed by acids? To a very great extent the products of hydrolysis by enzymes and acids are the same, and there may have been sometimes a tendency to assume that they are throughout identical. But this is improbable. The acids hydrolyse indiscriminately almost all that can be hydrolysed — carbohydrates, for instance, as well as proteids ; whereas the enzymes are, as is known, restricted within very narrow limits in their action. To take the familiar instances, the enzymes that hydrolyse the polysaccharides leave the disaccharides unaltered. We are beginning to learn some- thing of the hydrolytic powers of trypsine and its limitations from the experiments of Emil Fischer, who has tested the action of this enzyme on definite synthetic compounds of amido acids. Many of these are hydrolysed, the racemic ones asymmetrically, while others escape unchanged. The factors * Henriques and Hansen, H.-S. Z. 43, 417, 1905. ^ Abderhalden and Rona, H.-S. Z. 42, 530, and 44, 200, 1904-5. Cf. Henderson and Dean, M. J. 862, 1903 ; and, Lesser, Z.f. B. 45, 497. 132 ASSIMILATION AND SYNTHESIS OF PROTEIDS [lect. determining these different results have already been in part indicated by Fischer.^ One compound that resists the action of trypsine, but is broken up by acids, has been successfully traced by Fischer and Abderhalden among the products of the pancreatic digestion of casein and other proteids. It was observed that neither pyrrholidine carboxylic acid nor phenyl-alanine could be detected in the fluid obtained by the prolonged digestion of casein with trypsine, unless the ester method involving the exposure of the substances to the action of hydrochloric acid was employed. And the reason for this was found to be, that these particular cleavage products were not set free at all by trypsine, but were left combined with other amido acids in a polypeptide; this polypeptide could be obtained by precipitation with phospho- tungstic acid, was not a peptone, since it did not give the biuret reaction, but when hydrolysed with mineral acids was decom- posed and then yielded as much phenyl-alanine and pyrrholidine carboxylic acid as the casein itself when so treated.^ These results, no less than the experiments of Henriques and Hansen, show how much we have to learn about the changes that proteids undergo in digestion. Even what we have learnt shows more than anything else how little we know. For, according to the results obtained by Henriques and Hansen, it cannot be the polypeptides that must remain unaltered in order that the products of hydrolysis should serve in the place of proteids. For these polypeptides are precipitated by phospho- tungstic acid, and the experiments expressly show that what is precipitated by this reagent is not essential. But whatever the nature of the compounds may be which account for the physiological difference between the products of tryptic digestion and the products of acid hydrolysis, it can hardly be that the substances spared by trypsine are the only parts of the proteid that are required for proteid synthesis in the body. The great bulk of the substances set free in the ' E. Fischer and Bergell, B. 36, 2592, 1903 ; and, B. 37, 3103, 1904 ; E. Fischer and Abderhalden, H.-S. Z. 46, 52, 1905. 2 E. Fischer and Abderhalden, H. S. 39, 81 ; and 40, 215, 1903. VI.] THE CONDITIONS OF PROTEID SYNTHESIS 133 hydrolysis of proteids by enzymes and by acids are the same, and these substances enter into the composition of the proteids synthesised in the body in similar proportions to those in which they occur in the proteids of the food. For the present, all we can say is that there appears to be some kind of linkage between certain groups in the proteid molecules which is not uncoupled by the enzymes in the body, and that when it is uncoupled, as in acid hydrolysis, then it is impossible for it to be coupled up again in the body. This combination, which the cells can neither take to pieces nor put together again, must be present, in order that the other component parts of the proteid molecule may gather about them and group themselves round them when the synthesis of proteids is to occur. These con- siderations appear to suggest that the synthetic processes here involved may be the work of the same agent as the hydrolytic, the limitations in its hydrolytic power determining the limita- tions of its synthetic activity, as in reversible zymolysis. It is well known that those who have thought that the digestive disintegration of proteids did not go beyond the formation of albumoses and peptones have found great difficulty in understanding how it is that these substances are not to be found in the blood. It has, in fact, been long known that they are in no small degree toxic, intravenous injections affecting the coagulability of the blood, and causing a considerable fall in blood-pressure. In young or feeble animals the injection of o. I g. per kilo may be fatal.^ And when the amount injected does not cause death, the albumose is removed from the blood and is excreted unchanged in the urine. Spiro and Pick, it is true, purified albumoses in such a way as to do away with their toxic properties, but even so they were excreted in the urine.^ Neither can we suppose that under normal conditions they are taken up from the portal blood by the liver, and there altered in such a way as"^to abolish their toxicity and at the same time render them available for nutrition. For Neumeister showed that they were not to be found in the portal any more than ' Schmidt Mulheim, D. R. A., p. 30, 1879 ; Fano, D. R. A., p. 277, 1881. ° Spiro and Pick, H.-S. Z. 31, 235, 1900. 134 ASSIMILATION AND SYNTHESIS OF PROTEIDS [lect. the systemic blood, and if injected into the portal vein they appear in the urine, in spite of having to pass first through the liver. Nor can they be found in the lymph or the tissues any more than the blood.^ There are two well-known hypotheses that have been advanced to remove this difficulty. Hofmeister suggested that the leucocytes, which do contain albumose, might act as carriers, take the absorbed albumose up and convey it to the tissues. The phenomenon of post-prandial leucocytosis was quoted as supporting such a conception. But Heidenhain found that it was quantitatively impossible to attribute to the leucocytes such a fundamentally important function in nutrition.^ And post- prandial leucocytosis, though very commonly observed, is not constant, as it should be, if it were necessary for proteid-absorp- tion. Besides the hypothesis seems to require that it should be a leucocytosis due to an increase of lymphocytes, which is not observed to be the case. The explanation, therefore, which has most generally found acceptance is, that the albumoses before they leave the intestinal epithelium are resynthesised into proteids such as are found in the blood, albumin or globulin. This was the explanation given of Salvioli's often-quoted experi- ments done in Ludwig's laboratory. He sent blood through the vessels of excised loops of intestine into which a solution of Witte's peptone had been placed, and after three or four hours found that peptone had disappeared from the bowel, but was not to be found in the blood.^ Intestinal perfusion experi- ments which have been carried out at the Lister Institute with the apparatus devised by C. J. Martin — by a method, that is, which is certainly an improvement on the simple rough method employed by Salvioli — have shown that this is almost un- questionably not the right interpretation of his results. It is true that the intestine under such treatment may exhibit move- ments for some hours, but whatever these movements indicate ' Neumeister, Z.f. B. 27, 315, 1890. 2 Heidenhain, Pfl. A. 43, suppl. i, 1888 ; Shore, //. of Phys. xi., 528, 1890. ' Salvioli, D. R. A., suppl. 95, 1880. vi.j THE INTESTINE AND PROTeID SYNTHESIS 13o as to the condition of the muscular coat, they do not necessarily indicate that the mucous membrane is living and normal. The mucous membrane desquamates, no absorption takes place, and at the end of the experiment the intestine is distended with a thick grumous fluid, generally deeply stained with blood. In this last respect Salvioli's experience seems to have been similar. The fluid recovered from the intestine is found on examination, after removal of coagulable proteids, to contain no less nitrogen, sometimes more, than was introduced in the peptone solution. Peptone has disappeared, it is true ; for much more of this nitrogen is in the form of compounds that are not precipitated by tannic acid than was the case in the fluid put into the bowel. The peptone has disappeared owing to the action of the ferments liberated from the mucous membrane, not because it has been absorbed. And the peptone is not found in the blood, for the simple reason that no peptone has been absorbed.^ Hofmeister tried to obtain experimental proof for this same hypothesis in another way. The mucous membrane of the stomach and intestines was removed from animals killed during digestion, and thoroughly washed. Parts of this were boiled at once, or heated merely to 60° C. for a short time, other parts were kept at the body temperature for two or three hours. Those portions which had been heated at once were found to contain more albumose and peptone than those the temperature of which had not been raised. The disappearance of albumose and peptone was interpreted as due to the synthesis of coagul- able proteid from it.^ Neumeister looked for proof of the same hypothesis by another method of experimentation. He floated washed mucous membrane taken from the intestine in a solution of peptone containing blood that was kept aerated, and found that peptone disappeared from the solution and was not to be recovered from the mucous membrane.^ But these results and those obtained by Hofmeister are capable of another inter- ^ Cathcart and Leathes, //. ofPhys. 33, 462, igo6. ^ Hofmeister, H.-S. Z. 6, 69, 1882 ; and, 5. A. 19, 20 and 22. ' Neumeister, Z. f. B. 27, 315, 1890. 136 ASSIMILATION AND SYNTHESIS OF PROTEIDS [lect. pretation, if enzymes, present either in the cells or on the surface of the mucous membrane, were at work upon the albumoses, converting them into substances of a simpler nature that would give no biuret reaction. And Embden and Knoop found that this was so under the conditions under which those results were obtained.^ The hypothesis, therefore, that proteids are synthesised in the intestinal epithelium from albumose absorbed during digestion, and that these synthesised proteids are then passed on to the blood, and circulate in the blood for the nutrition of the body, remains a hypothesis. It may appear to be inevitable ; but no direct proof has yet been advanced. If, however, we believe that the changes undergone by proteids in digestion do not stop with the formation of albumoses or peptones, then a third explanation of the absence of these substances from the blood may be offered. If the proteolytic ferments in the intestinal canal form still simpler substances than these, and if albumoses that are absorbed into the epithelium are there acted on by erepsine, that which reaches the blood in proteid absorption may no longer be proteid at all. It may be amido acids and similar final products of hydrolysis that are circulated and supply the body with what is necessary for the restitution processes of metabolism and for growth. But if so, it should be possible to detect these substances in the blood ; and the problem that now confronts us is to determine whether this can be done, more especially during the absorption of proteids. But, first of all, we may do well to consider how much nitrogen in such forms we can expect to find, even supposing that the whole of the proteid in the food reaches the blood in the form of simple final products of hydrolysis, and as such is conveyed to the tissues where it is required. A man takes lOO g. of proteid in the day, we will suppose, containing about i6 g. of nitrogen. Absorption begins soon after the first meal, and lasts with little if any interruption till four or five hours after the last meal. The average rate of absorption, therefore, ^ Embden and Knoop, H. B. 3, 128, 1902. VI.] THE SEARCH FOR AMIDO ACIDS IN THE BLOOD 137 of the i6 g. of nitrogen cannot be much more than about i g. an hour. The circulation time of the intestine is short, almost as short as that of the lungs, and is probably much less than a minute, perhaps not more than half a minute ; and, therefore, if there is one-tenth of the blood, or about half a litre, in the intestine during absorption of food — and there is probably more than this^half a litre of blood passes through the intestine every minute ; that is to say, 30 litres in an hour at the lowest computation, or 60 litres by what is probably a fairer mode of reckoning.. Thirty to sixty litres are available, therefore, for the removal of i g. of nitrogen corresponding to 6.25 g. of proteid.^ This amount of nitrogen by the hypothesis is dis- tributed over a large number of substances formed by hydrolysis from the proteid, and is not present all of it in one form. Leucine is probably the most abundant of any of these substances, and if we take 20 per cent, for the amount of leucine obtained from the proteid, the amount of leucine in the 30 to 60 litres of blood would be 1.25 g., or in i litre 20 to 40 mg. To detect this amount of leucine in blood, which contains some 20 per cent, of proteids, may be possible, but it cannot be easy. Tyrosine, of which proteids generally contain less than S per cent., would be present in less than a quarter of this amount ; and the other cleavage products, also obtainable from proteids in small amounts, one may well abandon all hope of isolating in a recognisable form, since they are certainly less easy to identify than these. Kutscher and Seemann, however, tried to isolate crystalline cleavage products of proteids from the portal blood of dogs during proteid absorption, and they failed in every case. Their procedure was to give dogs 500 g. of meat, that is about 100 g. of proteid, at about two o'clock in the afternoon, and again another 100 g. about ten at night, and the next morning at ten to bleed from the portal vein. ' Cybulski (5. A. 27, P- 39, 1895) measured the rate of flow through the portal vein of a dog weighing 9.5 ko., and found it to be on an average 9090 c.c. per hour, or about 150 c.c. per minute. In a man of eight times the weight of this dog, at the same rate the flow would be 1200 c.c. per minute, or 72 litres per hour. 138 ASSIMILATION AND SYNTHESIS OF PROTEIDS [lecT. After coagulating the proteids by heat, the filtrate was evaporated, to allow leucine or other recognisable substances to crystallise out ; no crystals of this nature were in any case detected. On one occasion, in a dog that six hours previously had had a large meal of meat, they connected the portal vein with the vena cava by means of an oiled tube, and then tied the aorta below the origin of the superior mesenteric branch, and also the subclavian and carotid arteries close to their origin, and the renal vessels on both sides, so that the blood then circulated through the intestines, heart, and lungs, while the liver and nearly the whole of the rest of the body was excluded from the circulation. With the help of artificial respiration, the heart was kept beating for about an hour, but at the end of that time no amido acids could be isolated from the blood. The conclusion to which they came was that these substances, which they believe to be the principal products of proteid digestion, are built up in some way into proteid in the intestine, just as Ludwig and Salvioli, Hofmeister and Neumeister had supposed that the albumoses and peptones are. In support of this conclu- sion they found that a substance could be extracted from the mucous membrane, which is not precipitated by phosphotungstic acid, but gives, after boiling with sulphuric acid, leucine crystals on evaporation. This they regarded as a synthetic product formed from the absorbed amido acids in the mucous membrane.^ The results of their experiments, though they are compatible with their conclusion, do not prove it ; for they are equally compatible with other hypotheses different from theirs ; for instance, that the absorbed amido acids undergo other changes in the process of absorption not necessarily of a synthetic nature, or that they are removed from the blood by the tissue cells as quickly as they enter it. Leaving on one side for the present any other changes that there may be ground for believing the amido acids may undergo in the intestine or liver ,2 and supposing, as is after all probable, that the products of digestive hydrolysis of proteids that enter 1 Kutscher and Seemann, I/.-S. Z. 34, 528, 1902. ^ Cf. infra. VI.] AMIDO ACIDS FOUND IN THE BLOOD 139 the blood unchanged are taken up by the tissue cells at a rate not appreciably different from the rate at which they are absorbed by the blood from the intestine, then, as we saw from the comparison of the blood flow with the rate of absorption, the amount of any one cleavage product present in the blood at any moment may be too small for the isolation of crystals of this substance to be practicable. But it might still be possible to detect by some test for amido acids in general an increase in the sum total of these bodies in the blood during proteid absorption. The most delicate reagent for amido acids is the chloride of naphthalene sulphonic acid, which reacts with the amido group, just as benzoyl chloride does when together with glycocoll it forms hippuric acid, the products that are formed being insoluble and crystallising well.i By the use of this reagent the presence of glycocoll has been detected in the urine of rabbits poisoned with phosphorus, and glycocoll is, owing to its solubilities, exceedingly difficult to trace in fluids that contain only traces of it. v. Bergmann used this acid chloride to test for the presence of amido acids in the blood from a case of acute yellow atrophy of the liver. After coagulating by heat the proteids in 270 c.c. of blood the filtrate gave a precipitate with the sulphochloride amounting to more than 2 g. ; but actual isolation of the component substances, which were in part crystalline, proved to be impossible. So, too, with the blood of dogs killed during digestion, precipitates were obtained, though the identification of any single component was not practicable.^ It is clear that if it were possible to prevent the removal by the tissues of any cleavage products of proteids that are absorbed into the blood, our chance of finding them in the blood would be increased. Cohnheim has for this purpose made experiments on the intestines of Octopus vulgaris and Eledone moschata, in which animals the intestine floats as it were in the blood. He introduced a solution of peptone into the excised and ligatured intestines, and floated them in blood, which he kept oxygenated ; ' Fischer and Bergell, B. 35, 3779, 1902 ; Abderhalden and Bergell, H.-S.Z.sg, II, 1903. 2 V. Bergmann, ff. B. 6, 40, 1904. 140 ASSIMILATION AND SYNTHESIS OF PROTEIDS [lect. at the end of twenty hours he found that the blood contained lysine, arginine, leucine, and tyrosine, whereas the blood during digestion in the living animal contained none of these sub- stances that could be detected.^ In some experiments carried out at the Lister Institute on this point the attempt was made in the first instance to obtain the absorbed products in a greater concentration by allowing the blood to circulate only through the intestine, using Martin's perfusion apparatus ; but, as mentioned above, it has proved so far impossible to keep the mucous membrane in the excised intestine of dogs in such a condition that it should absorb any nitrogenous substances at all. The appearances suggest that the mucous membrane perishes through auto-intoxication. If only it should be possible to overcome this difficulty, it should be much easier to detect an increase of amido acids in the blood in this way than in the entire animal. But even in the entire animal, with a solution of peptone, of albumose, or of the final products of tryptic digestion introduced into the bowel between ligatures, when absorption takes place, as is the rule, there is quite constantly on examining the blood an increase in the nitrogen, in the form of compounds which are not precipitated by tannic acid. The nature of these compounds of nitrogen added to the blood during absorption has not been determined, but they cannot be proteid, not even albumose. It has been shown that the very small increase in the ammonia in the blood is by far not sufficient to account for them. But urea, on the other hand, appears to account for one half, while the rest almost certainly must be in the form of amido acids and similar sub- stances.^ The whole question of the absorption of nitrogenous foods, and as to the nature of the material used in the synthesis of proteids in the animal body, is, however, still further complicated by the fact that it has recently been shown in more than one laboratory that, occasionally at any rate, albumoses may be present in the blood. Embden and Knoop found this to be the ' O. Cohnheim, H.-S. Z. 35, 407, 1902. ^ Cathcart and Leathes, //. of Phys. 33, 462, 1906. VI.] ALBUMOSE IN THE BLOOD 141 case repeatedly ; but since in blood from animals killed during absorption of proteids they sometimes failed to find albumose, whereas in blood from starving animals they sometimes had positive results, they are unwilling to draw any conclusions as to the significance of their results.^ So, too, v. Bergmann and Langstein found that after removing coagulable proteid from the blood plasma of dogs killed, during digestion there was still a considerable portion of the total nitrogen of the plasma to be found in the filtrate, in one case as much as 14.7 per cent., and some of this was in one instance shown to be in the form of primary albumose. Human blood also has occasionally been found to contain albumose.^ It may be that none of these recorded observations of the presence of albumose in the blood have any direct bearing on the question with which we are concerned here. But they throw some doubt, especially those in the last paper referred to, on the conclusiveness of the results which have been hitherto the foundation of our conceptions of proteid absorption. The situation may be summed up in a few words. Till recently, it was believed that proteids were absorbed mainly if not entirely as albumoses and peptones ; that these substances were converted by a synthetic change carried out in the intestinal mucous membrane into the coagulable proteids found in the blood ; that these blood proteids supplied the needs of the body, and were the material used for all tissue repair. We have gradually learnt that the first of these articles of belief requires considerable modification, we have to recognise that the second remains purely hypothetical, and that, therefore, the third is little if anything more than a preconception. The problem as it presents itself to us now, is rather this : is the synthesis of proteid, which is so important a factor in the metabolism of all growing and living animals, a function only of the intestinal epithelium ? Direct evidence is not as yet forth- coming, and we must be content with some working hypothesis put together from general physiological and biological considera- ' Embden and Knoop, H. B. 3, 120, 1902. ''■ Bergmann and Langstein, H. B. 6, 37, 1904. 142 ASSIMILATION AND SYNTHESIS OF PROTEIDS [lect. tions. But the result of attempting to form such a hypothesis at the present time will be somewhat different from what it was a few years ago. We know that in the seeds of plants proteids are often stored in considerable quantity as food stuff for the growing seedling. During germination these proteids are hydro- lysed, and circulated in the sap in the form of the familiar cleavage products. Schultze and Winterstein have isolated from the seedlings of various species of plants a long list of these cleavage products: leucine, iso-leucine, amido-valerianic acid, alanine, glutamic acid, aspartic acid, phenyl-alanine, tyrosine, pyr- rholidine carboxylic acid, cystine, tryptophane, lysine, arginine, liistidine. All these substances can be formed during digestion in animals. In the plant it is in these forms that the nitro- genous material is supplied to the cells during the period of most active growth, and from these unquestionably the proteids are synthesised. In animals, till recently we have believed that the intestine synthesised from these or more complex substances the serum-albumin and globulin found in the blood, and that it was with these highly organised coagulable proteids that the cells of the body were actually nourished. No account has been commonly taken of the fact that these proteids of the blood must be taken to pieces and again put together, rearranged on a different plan, if they are to serve for the making of proteids and nucleo-proteids in the cells of the muscles and other organs in which the destructive changes of life are felt. The proteids circulating in the blood are a currency which is not legal tender. And no account has been commonly taken of the familiar fact that when no food is obtainable, certain organs maintain for themselves a normal composition at the expense of the substance of other organs. When the spleen, liver, or the muscles of the limbs dissolve away in starvation, the heart feeds on what they supply. Are the proteids of these organs converted into serum- albumin and globulin, or are they melted down by autolytic processes into the same cleavage products as are formed in the digestion of food, and in this form thrown into the circulating blood, which is thus in a position to supply the heart and diaphragm with just what they are accustomed to receive in the vi.J AS IT APPEARS TO US NOW 143 blood from the digestive organs ? In the equally familiar and often quoted marvel of animal metabolism, the salmon in fresh water, when the flesh of its own muscles takes the place of food and supplies the fish with what is necessary for the development of the sexual glands, the synthesis of proteids, of a kind pro- foundly different from those that supply the material for this synthesis, must be effected out of fragments of the original molecules rearranged and put together in new combinations, by processes in which the intestine can hardly be supposed to play a part. And what applies to the salmon or the starving animal applies also to patients in acute fevers, for instance, or in hysterical or other conditions in which insufficient food or none at all is taken. The hypothesis to which such considerations as these are leading us is that the synthesis of proteids is a function of every cell in the body, each one for itself, and that the material out of which all proteids in the body are made is not proteid in any form, but the fragments derived from proteids by hydrolysis, probably the amido acids, which in different combina- tions and different proportions are found in all proteids, and into which they are all resolved by the processes, autolytic or digestive, which can be carried out in every cell in the body. LECTURE VII PROTEID CATABOLISM The catabolism of proteids in the body results in the discharge of most of the nitrogen, commonly not far short of 90 per cent., in the form of urea. From 100 g. of proteid about 30 g. of urea are formed, containing only 6 g. of carbon out of the 52 or more contained in the proteid. So that nearly 90 per cent, of this carbon is disposed of in other ways, and finally leaves the body altogether dissociated from the nitrogen, as carbonic acid. In the study of proteid catabolism we have to account, therefore, for the formation of urea and of any other constant nitrogenous excretory products, but also for the disposal of the greater part of the carbon of the proteids, and at the same time to form some conception as to how far these two sides of proteid catabolism are associated together ; whether, that is, the carbon and nitrogen are separated early or late in the series of changes which end with the production of little but urea, carbonic acid, and water. It is customary in discussing proteid catabolism to work back from these end-products, or more generally only those that contain nitrogen, and, by trying to give an account of the manner in which they come into being, to aim at getting in this way a general view of the destructive changes to which proteids are submitted in the body. It is possible that this general view will be more easily attained by reversing this process, starting from the proteids themselves and working down to the end- products, not omitting the carbonic acid. As soon as sufficient 144 LECT. VII.] INTRACELLULAR PROTEOLYSIS 145 data have been collected, this must certainly be the order of procedure, and even now it may be worth while seeing how far it is possible to follow it. The catabolism of proteids takes place within the cells of the various tissues. And one of the most outstanding facts among those that have come to light in recent years is, that the cells of almost all the organs of the body have been shown to contain enzymes that hydrolyse proteids — have within them, that is, the means of taking to pieces those complex condensation products of simple and compound amido acids which are known as proteids, almost, if not quite as completely as strong boiling mineral acids. Salkowski,i about fifteen years ago, was the first to show that, like the unicellular yeast plants, the liver and muscles of dogs, kept at body-temperature in the presence of antiseptics, underwent changes, to which he gave the name of auto-digestion, and that in those changes leucine, tyrosine, and other soluble nitrogenous substances were produced at the expense of the proteids. Jacoby^ extended this study, and introduced the word autolysis, which has been generally adopted as the name for such changes. He showed that the excised liver gradually undergoes liquefaction, and in the process, besides leucine, tyrosine, glycocoU, and tryptophane, ammonia and other basic substances are formed ; and also that the autolysis is far more rapid if aseptic precautions render the use of antiseptics unnecessary. The enzyme which effected this could be precipitated with ammonium sulphate, and a solution of the precipitate retained its power of setting up these changes. In the spleen, thymus, lymphatic glands, kidneys, and heart, Hedin ^ found that similar autolytic processes could be traced ; and the enzymes separated and distinguished from each other ; while the study of the substances formed showed that the disintegration of proteids is as complete as 1 Salkowski, D. R. A., 1890; Z.f. k. M., i8go; Schwiening, V. A. 136, 1894; and, Biondi, V. A. 144, 1896. ^ Jacoby, H.-S. Z. 30, 149, 1900 ; cf. Conradi, H. B. i, 144, 1902. ' Hedin and Rowland, H.-S. Z. 32, 341 and 531, 1901 ; and, Hedin, //. ofPhys. 30, 155, 1903. K 146 PROTEID CATABOLISM [lect. in the action of the ferments secreted by the digestive glands.i These autolytic processes are, it is true, frequently referred to as if they were merely death changes, and as if they corre- sponded to nothing that goes on during the life of the cells. And the discussion that was raised and so long kept up over Claude Bernard's discovery of the ferment in the liver that hydrolyses glycogen, has threatened to come up again over the proteolytic enzymes of the cells. The exact part played by them during life in the cells does not lend itself to investigation, perhaps, but it is certainly easier to believe that they do operate in the metabolism of the living cell, controlled and checked by the conditions that balance during life the tendencies to chemical reaction one against the other, than to suppose, as is otherwise implied, that every cell is furnished, as it were, with the means of putting an end to itself, to which recourse is to be had under no circumstances till the worst comes, and life is no longer possible. For, that they are not merely escaped trypsine, as Neumeister suggested, was shown by the fact that extirpation of the pancreas does not cause them to disappear.^ They have been shown to be concerned in the removal of pneumonic exudation, since, while normal lung and lung in the stage of red hepatisation is not autolysed, in the grey hepatised lung autolysis results in the formation of lysine, histidine, leucine, and tyrosine.^ The changes in the liver in phosphorus poisoning are due to autolytic activity getting out of control, for Jacoby found that the autolytic changes are much more rapid in the excised liver of animals poisoned with phosphorus than in that of normal animals ; in the former case the tissue is reduced to a fluid mass in twelve hours.* Autolysis may be supposed to account for the liquefaction of pus, the softening and breaking down of new growths, and the absorption of 1 Leathes,//. of Phys. 28, 360, 1902 ; Dakin, ib. 30, 84, 1903 ; Cathcart, ib. 32, 299, 1905. 2 Matthes, 5. A. 51, 442, 1904. ^ Simon, D. A.f. k. M. 70, 604, 1901. * Jacoby, H.-S. Z. 30, 176, 1900. VII.] AUTOLYSIS IN METABOLISM 147 infarcts and thrombosed parts. The absorption of gummata under treatment with potassium iodide has been ascribed to the promotion of autolysis by this drug.^ In all these cases, however, even though it be granted that the operation of the cellular enzymes within the body is the cause of the change, it may be said that the cells are dead, and it is because they are dead that the autolytic process is manifested ; the fact that the process is in these conditions found at work within the body does not prove that they play a part in the metabolism of the normal cells. That is clearly true ; but normal metabolism is a struggle between life and death, in which life just manages to get the upper hand, the catabolic processes of death being more than compensated for by the restitutive processes of life. We have to choose between two explanations : either, when cells die, enzymes are called into existence to act as licensed scavengers; or, the death of the cells manifests itself in a disturbance of the normal balance between competing changes, and in the failure of those conditions that with a normal circulation, or in the absence of poisons such as those intro- duced on the administration of phosphorus, hold in check and regulate these catabolic processes which we call autolytic. And if we choose the latter, then we are at the same time able to understand how the proteids are broken down in normal metabolism, and how it is that the muscles, liver, spleen, and other organs are in starvation able to supply the heart and respiratory muscles with all the nitrogenous material they require. There is no difficulty in suggesting provisional ex- planations of how it is that the action of these cellular enzymes during normal life is not conspicuous, as it is in the pathological conditions referred to above : how it is that we do not find the products of their action normally in the liver, for instance, as we do after poisoning with phosphorus ; it may be that these products undergo further changes on the spot, or that they are removed by the circulation to be disposed of elsewhere : how it is, again, that the organs do not normally liquefy and disappear under their action ; it may be that this is held in check by ' Oswald, B. Cbl., iii., 367, 1905. 148 PROTEID CATABOLISM [lect. anti-catalytic agents — Hedin showed that the serum of the ox contains such a substance for the cellular enzyme of the spleen and other organs ^ — or it may be that it is compensated for by the due adjustment of restitution processes, the conditions obtain- ing for a local reversion of the process. However this may be, even before the cellular proteolytic enzymes came to light, the catabolism of proteids was very commonly conceived of as being in the first instance simply hydrolytic. It has been obvious for many years that such a conception was compatible with known facts : leucine, tyrosine, aspartic acid, and glycocoll, when taken with food are not excreted ; their nitrogen is found in the urine as urea, and the carbon, so far as is known, undergoes the same fate as that of the proteids themselves : no unaltered glycocoll, alanine, leucine, or phenyl-alanine is found in the urine, even with the delicate reagent for such substances, naphthalene sulphonyl chloride, after administering 3 to 8 grammes of them to rabbits.^ That a hydrolytic resolution of proteids does form part of normal catabolic processes, is indicated by the fact that some of the products of this change are found in the bile and urine, fixed as it were by condensation with other compounds, and so escaping further destructive changes, in much the same way as glycuronic acid does. Glycocoll is thus fixed by condensation with both cholalic and benzoic acids ; diamido-valerianic acid or ornithine, which is formed by the hydrolysis of arginine, is in birds similarly combined with benzoic acid ; and cystine, after oxidation of the mercaptan group and the loss of carbonic acid, is, it is also known, the source from which the taurine in taurocholic acid is derived.^ It is difficult to account for the occurrence of these substances in the organism otherwise than by supposing them to be formed by the hydrolysis of proteid, and if so, it is difficult to suppose that they are the only parts of the proteid molecules thus split off, or that hydrolysis in the cells is in its broad outlines different from hydrolysis effected > Hedin, //. of Phys. 30, 195, 1903. 2 Abderhalden and Bergell, H.-S. Z. 39, 9, 1903. ^ V. Bergmann, H. B. 4, 132, 1903. VII.] AMIDO ACIDS FOUND IN METABOLISM 149 by the ferments secreted in digestion, or by those that act in autolysis outside the body. The isolation of the products of this action from the organs of the body after death, except in cases in which the risk of post-mortem autolytic changes giving rise to them was actually foreseen and avoided, is not a secure foundation for an argument on this point. And more striking than the fact that arginine has been isolated from the spleen, cystine from the liver, leucine from most of the organs of the body, and so forth, is the fact that from the normal liver and intestine, examined with due precautions against this risk, it has not been possible to isolate those even that are most easily isolated, such as leucine and tyrosine. And the liver and intestine we know contain, at any rate after removal from the body, the necessary enzymes for their formation. But if we cannot base our belief that the first step in proteid catabolism in the cells is hydrolysis upon the presence of the resulting substances in the tissues, their absence need not cause us to abandon it. We are apt to form an exaggerated notion of the rate at which proteid catabolism is carried on in the body. Even supposing that the urea in the urine is a measure of proteid catabolism in the tissues (which it is probably wrong to do), a man of 70 kg. in nitrogenous equilibrium on 120 g. of proteid daily, uses up 5 g. of proteid on an average in an hour in his whole body. The amount of proteid in his body, disregarding his blood, must be considerably more than 5 kg., for the muscles alone form 40 per cent, of his weight, and since they contain about 20 per cent, of proteid, they alone account for more than this. The average rate at which proteid is used up is therefore con- siderably less than 0.1 per cent, per hour: this means that 100 g. of tissue, containing 20 per cent, of proteid, will use up on an average less than 20 mg. of proteid in an hour, and produce in that time less than 4 mg. of, say, leucine, forming one-fifth of the weight of the proteid. So that, in order for the products of proteid hydrolysis to occur in any organ in sufficient quantity to be isolated in an ordinary experiment, it would be necessary to suppose either a rate of formation very considerably greater 150 PROTEID CATABOLISM [lbct. than the average rate, or else stagnation of the other processes by which these products are further acted upon or removed. We are also apt to form our notion of the stability of substances that occur in the body from their behaviour when isolated and examined outside the body. Polysaccharides, fats, and proteids alike, when isolated, may require energetic treat- ment in order to separate them into their component parts. But in living organisms the work of hydrolysis and condensation is so simply effected, and with so little if any loss of energy, that we ought to accustom ourselves to quite different notions of their stability in physiology. To take the most familiar instances : sugar is built up in the liver into glycogen at one moment, to be reduced, as we believe, to its former condition as sugar again the next, and again may become glycogen a moment later on reaching the cells in other parts. The fatty acids are divorced from glycerine to enter the intestinal epithelium, and reunited as they leave it to enter the lymph, and it seems probable that they are again separated on reaching the blood, and remain separated till they settle down in the connective tissues. Even there, too, we find that agencies exist for bringing about another separation, which probably takes place before the next stage in the life history of these groups is entered upon. It is only putting our physiological concep- tions of the proteids in the body on a footing with those that we have to form of the fats and carbohydrates, if we regard them as labile aggregations of amido acids which ih the body have far less cohesion than their behaviour when isolated suggests. There is therefore much that points to the hypothesis, which is in fact commonly adopted, that proteid catabolism begins by a resolution of the proteids into their simple coni- ponent parts, such as is effected also in digestion. If, as is probable, this is a change which, like the corresponding change in fats and carbohydrates, is practically isothermic, then this change is merely a preamble to those in which the proteids are made use of as a source of energy ; and proteid catabolism, in so far as this is concerned with heat-production or work VII.] DENITRIFICATION OF AMIDO ACIDS 151 done in muscles or elsewhere, is the catabolism of nothing more complex than the amido acids, simple and compound, thus set free. The accounts of the course of proteid metabolism, starting from the proteids themselves, which it has hitherto been possible to give, have at this point had to cut matters short and jump to the end-products. But it is now possible that a further step may be taken. It has recently been shown that in many of the organs and tissues a reaction takes place, the work pre- sumably of an enzyme, which shows itself by the liberation of ammonia from amido acids.^ Lang in Hofmeister's laboratory treated the pulp obtained from various organs very thoroughly with toluene, by shaking mechanically for some time, and then estimated, after keeping them for some hours at the body- temperature in an incubator, the amount of ammonia present in samples to which various amido acids had been added, and in others to which no addition had been made. More ammonia was found when leucine or glycocoll, tyrosine or' cystine, was added, in almost all of a large number of experiments. The most marked action was with the intestine, liver, or pancreas treated with glycocoll or leucine. But it was clear that the toluene, as is unfortunately so commonly the case with anti- septics, exerted a very unfavourable influence on the reaction ; since liver pulp, prepared as far as possible aseptically, produced more ammonia from glycocoll in an hour and a half than was produced by the liver in the presence of toluene in any other experiment after several days. Sufficient growth of the few imported bacteria can hardly have taken place in this short time to have produced any detectable ammonia. These results are an extension of results arrived at pre- viously by Jacoby,^ who found that, in the fluid obtained from the liver after grinding it up with sand, the ammonia increased, and that this ammonia was derived from substances which like the amido acids do not give up their nitrogen when boiled with acids. O. Loewi had also shown that the amido group in • Lang, H. B., v., 321, 1904. 2 Jacoby, J^.-S. Z. 30, 149, 1900. 152 PROTEID CATABOLISM [lect. glycocoU was changed, in the presence of a pulp of liver cells, into some substance soluble in alcohol, not urea, which like urea gives up ammonia when treated with fixed alkalies, the firm union of nitrogen to carbon as it exists in the amido acid having been dissolved. It seems, therefore, that the power of removing ammonia from amido acids may be a general property of the cells of many tissues in the body ; just as we know that bacteria remove ammonia from tyrosine in putrefaction, or from tryptophane, and from urea. And, at any rate, this property is marked in the intestinal mucous membrane and liver. This explains the fact observed by Nencki ^ and his fellow-workers, that the amount of ammonia in the portal blood is during digestion greater than that in the systemic blood, as much as four times the amount being sometimes found : ammonia is split off from amido acids in the intestine, conveyed to the liver, and there removed from the blood and converted into urea. The stomach, intestines, liver, and pancreas all contain, accord- ing to the same observers, considerably more ammonia than the muscles, brain, kidneys, or spleen. Lang's^ experiments, it may be, require exterision and confirmation under varying conditions, though they leave no doubt about the main fact. According to them, however, while some of the amido acids, unlike leucine and glycocoll, are apparently hardly acted on at all, phenyl-alanine, for instance, and in some cases cystine and tyrosine, amides such as acetamide, but in a much more marked degree the amides of amido acids, asparagine, and glutamine, give up ammonia as well. Now it is known that a certain amount of the nitrogen in proteids is given off as ammonia in hydrolysis both by enzymes and mineral acids, and this nitrogen is, therefore, commonly referred to as the " amide " nitrogen of the proteid molecule. This amide nitrogen may, therefore, be fairly assumed to contribute to the nitrogen removed from proteids in the liver as well as other organs, whether the removal be effected in the 1 Nencki and Zaleski, 5. A. 36, 385 ; and, A. St P., 1895. Salaskin and Zaleski, H.-S. Z. ii,i) ; cf. Folin, H.-S. Z. 37, 174, 1902. * Lang, H. B. 5, 340, 1^04. vii.] LIBERATION OF AMMONIA 153 course of proteid hydrolysis or in this special denitrifying process. The amide nitrogen constitutes in the case of casein as much as 1 3 per cent, of the whole nitrogen ; in other proteids, serum- and egg-albumin and the primary albumoses obtained from fibrin, generally about 7 or 8 per cent. ; while in gelatine it is less than 2 per cent.i The denitrification of amides is, however, not quite a universal reaction, to judge from the fact that oxamic acid is not converted into oxalic acid at all, but appears to be converted directly into urea when administered to rabbits. This is, however, rather an exceptional form of amide.^ Whether the diamido acids contribute to this ammonia formation, we do not know : but from the following fact it seems probable. Just as Neuberg and Langstein found that after administration of alanine to rabbits in considerable doses, small quantities of the denitrified product, lactic acid, escaped oxida- tion and appeared in the urine,^ so P. Mayer showed that diamido-propionic acid injected subcutaneously into rabbits gave rise to the excretion of a small quantity of the corre- sponding dioxy-acid, glyceric acid ; the small yield not necessarily meaning, of course, that no more was formed, since such a substance as glyceric acid could hardly be expected to run the gauntlet in the body and appear to any great extent in the urine unaltered.* Lysine, therefore, and the ornithine set free by arginase in the body,^ may very probably share the fate of the mono-amido acids and lose their nitrogen before they are otherwise acted on in the cells. Now the amido acids are nitrogenous substances in which the combination of carbon and nitrogen is particularly firm, as is obvious from the stability of these compounds under the action of the proteolytic enzymes and of boiling mineral acids. The removal of the nitrogen from amido acids by the cells of the body is a remarkable reaction, and one that has hitherto ' Hausmann, H.-S. Z. 27, 95, and 29, 136, 1899 and 1900. 2 Schwarz, S. A. 41, 60, 1898. ' Neuberg and Langstein, M. J., p. 603, 1903. * P. Mayer, H.-S. Z. 42, 59, 1904. " Kossel and Dakin, H.-S. Z. 41, 321, 1904. 154 PROTEID CATABOLISM [lect. not been generally taken into account in considering the course and nature of proteid catabolism. We do not know to what extent it takes place in either the assimilation of proteid food or cell metabolism. But in future account must be taken of it, and a little reflection must show that it may have important bearings on some of the most prominent problems of nitrogenous metabolism. In the first place it may be as well to take note of the fact that the removal of the nitrogen does not very materially affect the energy value of the acids. The heat equivalents of some of the amido acids and the corresponding fatty and oxy-acids are given in the following table : — Gal. per 1 g. Cal. per 1 gm.-mol. Difference per cent. Leucine fi.52 854.9 I , Caproic acid . 7.16 830.2 J ' Leucic acid . ... ... - Alanine 4-36 389 I =, Propionic acid 4-95 366.9 J '■' Lactic acid . (?) 3.7 (?) 338 1 331 15 GlycocoU 3-13 ^35 I .0. Acetic acid . 3-49 209 GlycoUic acid 2.10 160 22 1 This has not been directly determined apparently ; but, as was pointed out in Lecture IIL, p. 63, has probably about this value. The figures given in the last column show the difference of the energy value of the fatty acid as compared with that of the corresponding amido acid, reckoned in percentage, of the latter. Even in the case of the smallest molecule, glycocoll, it amounts only to about 10 per cent. We do not know what the non- VII.] NOT TO BE MEASURED BY UREA EXCRETION 155 nitrogenous compounds left after removal of the nitrogen are : it is perhaps more likely that they are the oxy-acids than the unoxidised fatty acids, although it is oxyphenyl propionic acid, not oxyphenyl lactic acid, that is produced by bacteria from tyrosine ; and in that case the energy set free will be a larger proportion of the whole. But, even so, if proteids in assimila- tion are first hydrolysed, and then to any considerable extent the amido acids denitrified, the resulting products would hardly have given up more than about lo to 15 per cent, of the energy in the proteid molecule from which they were derived. In other words, the nitrogen, or a great part of it, may be removed from the proteid, converted into urea, and expelled with the urine before the oxidation of the rest of the molecule has been started upon ; and the fact that we can trace in the urine excreted in a given time, all or the greater part of the nitrogen of the proteid taken at a meal, tells us nothing what- ever about the fate of that part of the proteid which contains, it may be, as much as 80 or 90 per cent, of the total energy of the proteid food. Proteid metabolism in so far as it is concerned with the evolution of energy, proteid metabolism in its exo- thermic stages, may be almost entirely non-nitrogenous meta- bolism. Our habit of looking upon the appearance of so much urea in the excreta as a sign that so much proteid has been " used up," of calculating the total energy corresponding to that amount of proteid, and saying that that amount has been derived from the combustion of proteid, may be based entirely upon a misconception. We shall have in future to bear this reaction in mind: it has been ^proved in the case of the cells which are the first cells in the body with which the absorbed nitrogenous substances come into relations ; it is a reaction which leaves the greater part of the energy of the proteid molecule untouched, but, nevertheless, puts the greater part of the nitrogen in the way for being thrown off as urea in the urine. It is well known that in dogs, as in man, after a proteid meal, the urea excretion rises, and in the first six or seven hours this rise may account for fully half the nitrogen of the 156 PROTEID CATABOLISM [lect. meal, and in a few hours more for almost the whole of it ; ^ and since in dogs, after a large meal of proteid, absorption from the intestine is known to be extended often over fourteen hours or more, the whole of the nitrogen may be traced into the urine within an hour or two of its absorption. It is true that the carbonic acid excretion is simultaneously somewhat increased, but this is to be explained almost entirely, in the first place, by the increased muscular activity of the digestive tract which, it is known, influences the carbonic acid output just as the activity of the voluntary muscles does ; and in the second place, by the increased activity of the digestive and excretory organs : but there is no strict parallelism between the excretion of urea and carbonic acid. The two main end-products of proteid meta- bolism, urea and carbonic acid, are, to a great extent, produced independently of each other, and the reactions which result in the discharge of the nitrogen are not those in which energy is set free, work done, and carbonic acid produced. This rise in the rate of excretion of urea after a proteid meal, which we have learned to regard, with both Voit and Pfluger, as a sign that the cells of the body prefer to use proteids for all their requirements, if only they can get them, and that it is only when the proteids are used up that the cells fall back upon the non-nitrogenous stores, we must now be prepared to learn is a sign of nothing of the kind, but rather a sign that the body has no need for all this nitrogen, and that it must be got rid of before the really valuable part of the proteid molecules is admitted into the general circulation. One of the most familiar facts with regard to nitrogenous metabolism is, that the amount of nitrogen excreted in starvation is much less than the minimum amount that must be given in the form of proteid in order to maintain nitrogenous equilibrium. This fact appears in a new light, if we suppose that a long succession of generations in the past, which have lived from choice or necessity on a diet rich in proteids, have handed down to us, as our inheritance, a constitution in which arrangements exist for the removal of nitrogen from a considerable part of ^ Falck, M.J. 247, 1876. VII.] NITROGEN CANNOT BE STORED 157 this proteid. The fact that the amount of proteid taken is readjusted to suit the actual needs of the body, though it makes these arrangements unnecessary, will not necessarily remove them. The denitrifying enzyme, which has been trained to keep guard over the entrances by which nitrogenous substances are admitted into the body, will continue to levy its toll of nitrogen, even when the amount of proteid presented to it is no more than the tissues which it serves actually require. Again, it is one of the cardinal laws of proteid metabolism that the store of nitrogenous substances in the body is not increased by, or not in proportion to, an increase in the nitrogen intake. There is a classical experiment of Voit's which is everywhere referred to, in which a dog, after being kept in nitrogenous equilibrium on. a daily ration of 500 g. of meat, was given three times that amount on seven days in succession. On the very first day the nitrogen excreted corresponded to 1222 g. of meat, on the third to 1390 g., and on the seventh nitrogenous equilibrium was re-established. In seven days 7 kg. of meat had been given above and beyond what was sufficient for the animal's normal existence ; yet only about 1 1 g. of the nitrogen contained in this meat had been retained in the body, and when the daily allowance was reduced to 1000 g. of meat, or double the amount that was formerly sufficient, most of what had been gained was lost in the next five days. Siv^n, in his experiments on himself,^ came across the same phenomenon. With an exceedingly low nitrogen intake, maintained for some weeks, he lost about 38 g. of nitrogen. Then, during a week, with 12 g. daily, he gained only 14.5 g. in all, and in the succeeding six days, with a daily intake of 22.6 g., a further 6.5 g., making a gain altogether of only 21 g. in thirteen days. He concluded, as has been commonly done in such circumstances, that it is a tendency of the body, even with an abundant supply of nitrogen, and after a period of nitrogen loss, to strive towards nitrogen equilibrium ; and commenting on this, lays it to the credit of the " circulating proteid " theory of Voit. This phenomenon has always appeared ' Siven, Sk. A. 11, 308, igoi. 158 PROTEID CATABOLISM [lect. a most remarkable one, and if we may explain it in the light of what we have recently learnt of the reactions carried out in the cells of the intestine and liver, it may be a relief to many who have felt the strain of adhesion to the current explanations. We riiay then see that much of the proteid does not get past the liver as nitrogenous matter at all ; all excess of proteid above what is indispensable for repair is stripped of its nitogen and can only reach the tissues in combinations in which nitrogen does not occur ; the nitrogen that is not wanted is removed as ammonia, converted into urea, and expelled from the system by the kidneys. And when nitrogen is wanted to repair loss, even then, because of this tendency of the substances derived from proteid food to lose their nitrogen before admission to the body, only a small portion of them can run through past the liver and become available for this repair. The question whether the urea is formed from " circulating " or " tissue- proteid " does not come in at all, and we can escape, it may be, from the dilemma of having to take for our own, one or other of those two historical watchwords, each of which has been so stoutly defended in the literature of physiology. The ideas underlying this famous controversy are doubtless fundamental, and must be extended to include, or, rather perhaps confined to, the non-nitrogenous components of cells. What is the relation of cell-protoplasm and of nucleo-plasm to the non- nitrogenous material in the oxidation of which the vital transformations of chemical energy are made manifest? Does it become a part of that protoplasm, or is it acted on by mere contact with it ? It may be that in some such form as this we may look to this question reappearing in the future. But it is of little use trying to formulate it now, since for the present we hardly have sufficient data to enable us to ask ourselves intelligent questions on such subjects.^ 1 O. Folin has recently called attention to the weakness of the principal experiment, which has very commonly been regarded as giving the supporters of the organ proteid theory the best of the argument : Schondorff s well-nourished dog, before the perfusion experiment, was turning out nitrogen at the rate of 38 g. a day, or nearly 1.5 g. an hour : the blood VII.] THE NITROGEN MINIMUM 159 But whatever significance we read into the work that has come from Hofmeister's laboratory, the fact of course remains that animals die without proteid, or, at any rate, without the products of proteid hydrolysis in their food. And therefore it is hardly to be supposed that the nitrogen of the whole of these substances is removed before they can get past the liver, so that for replacing tissue the combinations have to be re- synthesised. If that were so, it should be possible to maintain animal life on ammonia and non-nitrogenous foods. The utmost that the data warrant is that from a part of them the nitrogen is so removed, it may be a considerable part, and sufficient to account for the phenomena that we have just been discussing. But some of it must remain. What is suggested before all things by these results, as it appears, is that it is no longer possible to take the rate of urea formation as necessarily a measure of proteid metabolism, whether by that we mean the rate at which cell protoplasm decays or merely the rate at which proteid is used as a source of energy in the body. It is not a measure of the true proteid catabolism, the decay of living nitrogenous matter, because a great part of it is formed from nitrogen that has never been beyond the liver, and it is not a measure of the energy derived from proteid, because it is largely derived from proteid by reactions which leave the energy value of the molecules from which it is derived but little altered. Further, it must occur to us that if a considerable part of the proteid is only allowed to reach the tissues of the body generally in a non-nitrogenous form, it should be possible to replace this part by non-nitrogenous food-stuffs from the outset ; and that the low nitrogen dietaries that in some races are found to be compatible with great physical activity may be perfused for four and a half hours through the hind-limbs and liver collected 25 mg. of nitrogen in the form of urea. In the entire living animal during that time it would have conveyed about 6 g. of nitrogen as urea to the kidney, or considerably more than two hundred times as much. As Folin points out, 25 mg. of nitrogen is not sufficient foundation for so weighty a superstructure. 160 PROTEID CATABOLISM [lect. explained by taking this into account. In point of fact, this is just the conclusion that certain workers on proteid metabolism in laboratories both in Europe and America have arrived at, quite apart from these considerations. Siv^ni found in experiments on himself that he could lead his ordinary life and maintain nitrogenous equilibrium without losing weight, when taking only between 4 and 5 g. of nitrogen daily, provided, of course, that the total Calories supplied in his food amounted to a little over 40 per kg. Landergren,^ in four experiments on a man of 70 kg., found that if carbohydrates were given in sufficient quantity to provide from 38 to 45 Calories per kg., and no nitrogen, except about I g. daily, which was inseparable from the forms of carbo- hydrate food employed, then the nitrogen excretion sank on the fourth day to from 3 to 4 g. ; about a fifth, that is, of the amount ordinarily excreted. Chittenden finds that he himself and his collaborators, and also men undergoing military training, and students actively engaged in athletic competitions, maintain their weight and nitrogenous equilibrium on a diet containing only from 30 to 50 per cent, of the amount of proteid which has been commonly regarded as normal for man ; and in some cases his observations extend continuously over several months.^ It may be, therefore, that the conventional dietary of 100 g. of proteid, or more, should be very considerably cut down, and that that part of this allowance of proteid which is merely denitrified and used by the body in non-nitrogenous forms should be habitually replaced by carbohydrates, and in part by fat. But it does not necessarily follow that this is so, or that it is unphysiological for man to take more than the minimum amount of nitrogen necessary for equilibrium, any more than it is unphysiological to take any food which yields more than the minimum amount of faecal refuse. The experiments on the nitrogen minimum, establishing an amount that holds for a few ^ Siven, Sk. A. 10, 91, and 11, 308, 1900-1. 2 Landergren, Sk. A. 14, 112, 1902. ^ CbatitaAea, Physiological Economy in Nutrition, Heinemann, 1905. vii.] is THE MINIMUM NORMAL? 161 days or weeks, do not necessarily prove that it holds for pro- longed or for lifelong use. The food of an infant in the second half of the first year is commonly and normally about two pints of milk ; even taking this to contain only 1.5 per cent, of proteids, that gives about 17 g. daily, or 2 g. per kg., and this estimate, which is certainly not high, is more than the conventional adult diet provides, and from five to ten times as much as the minimum. It is, indeed, a well-known fact that the rate of urea excretion in infancy is higher in proportion to the body-weight than at any other period of life. If ten times the minimum rate is the normal diet provided by Nature, then, even after making full allowance for the necessities of growth, the minimum can hardly be normal for the adult, nor the amount ordinarily taken a very great deviation from the prescriptions of Nature. Again, if the removal of nitrogen from certain amido acids is established as a fundamental physiological reaction, it is not proved that it applies to all the compounds of nitrogen formed from proteids : it is said, for instance, that tyrosine and phenyl-alanine, unlike most of the amido acids, when injected into the blood, do not increase the urea excretion,^ and in Lang's experiments tyrosine reacted feebly, if at all, and phenyl-alanine not at all : there are nitrogen compounds in proteids which are of a different nature altogether ; histidine, the indol group in tryptophane, and perhaps pyrrholidine may be included with these; and one of these, histidine, there are reasons for thinking, contains a nitrogenous group which is related to one of those present in nucleic acids. We know nothing of the fate of the nitrogen in these compounds, but it may be that one or other of them is required in larger amounts for cell repair, and that it is only the ordinary amido acids that are not required in the amount commonly taken. There still remains that side of proteid catabolism which represents the decay of protoplasm. This, we may suppose, is the source of the nitrogenous excreta when all superfluous proteid is excluded from the food — when all the proteid that is merely denitrified in the process of assimilation is dispensed with. In the experiments of Siven and Landergren this was ^ Stoltz, H. B., v., 25, 1903. L 162 PROTEID CATABOLISM [lect. shown to yield about one-fifth of the amount of nitrogen commonly taken as normal. It becomes a point of great interest to know in what form this nitrogen is excreted. The denitrification processes result, we must suppose, in the discharge of nitrogen, principally in the form of urea, with possibly some ammonia, and even some uric acid ; since, according to Wiener,^ uric acid can be synthesised in the mammalian organism, as in birds, from urea and an organic acid free from nitrogen contain- ing a chain of three carbons. If these excretory substances, resulting from the denitrification process, are subtracted from the urine, then the distribution of the nitrogen over the several different nitrogenous substances of the urine may be much altered. For the breakdown of cell protoplasm, including the nucleoplasm, may give rise to these same substances in other proportions, or even to other substances altogether. If, therefore, on a diet containing the minimum amount of proteid, or better ;Still, a diet free from proteid, but of a sufficient energy equivalent, it is found that the distribution of nitrogen is altered, this might fairly be regarded as a corroboration of the hypothesis to which we have been led. As a matter of fact, Folin ^ has found just such a difference, and on this ground has been led to take a general view of proteid metabolism which agrees almost exactly with the one to which the work of Lang and others in Hofmeister's laboratory has led us. The principal points which he has determined are, that under these conditions the nitrogen in the form of urea sinks from 87 per cent, of the total nitrogen to about 60 per cent., and the inorganic sulphates at the same time from 90 to 60 per cent, of the total sulphur; the kreatinine and neutral sulphur in absolute amount remain almost unchanged, in percentage of the total nitrogen or sulphur are therefore greatly increased, the kreatinine nitrogen rising from 3.6 per cent, to 17.4 per cent., and the neutral sulphur from S per cent, to 26 per cent. The other forms of nitrogen which he determined, ammonia and uric acid, as well as those forms which he did not determine directly, but estimated together by ^ Cf. infra, page 187. ^ Folin, Am. Jl. Phys. 13, 66, and 117, 1905. VII.] AFTER DISCOUNTING DE NITRIFICATION 163 subtracting from the total nitrogen the sum of the various forms of nitrogen directly determined — all these three, as well as the ethereal sulphates, were diminished in absolute amount, but in percentage of the total nitrogen or sulphur were increased. The diets he employed for his« comparison were, on the one hand, an egg and milk diet, containing about 19 g. of nitrogen, but no kreatine and no purine ; and on the other, a starch and cream diet, containing only about i g. of nitrogen. These results are of great interest and importance, and fully justify, as I believe, the complete reconsideration of the position of physiology as to the nature and course of proteid metabolism which he has advocated. Taken in connection with the account we have been led to above, they indicate that the catabolism of the tissues accounts for all the kreatinine excreted on a diet free from meat, and all the neutral sulphur, a larger proportion of the uric acid, ammonia and ether sulphates, and a smaller proportion of the urea and inorganic sulphates than the denitrification processes ; while the fact that the reduction of the proteid supply, so as to exclude those changes which the proteid food is liable to in the digestive organs and liver, is accompanied by a diminution of only those nitrogenous constituents of the urine which we know may be formed from ammonia, is some confirmation of the significance ascribed above to the denitrification processes. The idea that a nitrogen-free residue might result at some stage of proteid metabolism before the final one, is by no means new. Minkowski, in calculating the utmost conceivable amount of sugar that could be formed from proteid, claimed the whole of the carbon that was not wanted for urea ; Seegen, in his theory of the origin of the sugar in the liver from peptone, and many others, have more or less dimly tried to realise some such stage in the metabolism of proteids, as Lang's experiments and those that preceded them have revealed. The problem of the synthesis of sugar from proteid has taken a new aspect, and one step forward has been taken which may lead to others, and finally make clear how the change may be effected. As we have already seen in an earlier lecture, several experiments have 164 PROTEID CATABOLISM [lect. vii. already been published which were designed to test the possibility of a sugar synthesis from amido acids or the oxy- acids which are formed from them. But much more remains to be done before the matter becomes clear. It seems, indeed, that, in order to be able to follow the processes of animal metabolism, what we require to know more than anything else is the way in which the lower fatty acids and their simple derivatives are oxidised. If we are right in believing that it is the oxidation of the substances obtained after removal of the amido groups from amido acids that give proteids their value as sources of energy, then the reactions which do the work and supply heat in the body are concerned with similar compounds, whether in the first instance it be proteids, carbohydrates, or fats from which they are derived. The greatest step forward will be taken when we come to know how the cells dispose of simple organic compounds, such as lactic or acetic acid, what stages the reactions pass through, and, therefore, what side reactions can occur ; whether substances may be formed, such as formic or glyceric aldehyde, from which sugar, or acetic aldehyde, from which, according to Nencki's suggestion, higher fatty acids may be put together. Then we shall be able at last to make intelligible the problem of the formation of both sugar and fat from proteids, which is never likely to be the case so long as these supremely complex substances themselves are all that we deal with in our experi- ments and discussions on metabolism. LECTURE VIII THE METABOLISM OF CYCLIC FORMATIONS The greater part of proteid molecules is composed of the amido and diamido fatty acids, with which, together with the unknown amide combinations, we were principally concerned in the last lecture. But in addition to these there are a number of combinations of nitrogen with carbon found in proteids, or associated with proteids in the body, which have not yet been touched upon. And some of these are present in substances which have furnished physiology with not the least difficult of its problems in metabolism. We may begin with the guanidine group, imido urea, NH„ / C = NH NH, which is found in arginine combined with amido-valerianic acid and also in kreatine. We know this about arginine, that in many of the organs — the liver, kidney, intestine, thymus, and lymph glands, but notably not in the spleen, pancreatic secretion, or bile — an enzyme is found which attacks this group, removing urea and leaving the diamido-valerianic acid, ornithine.! Besides this enzyme acting on free arginine, an ' Kossel and Dakin, //.-S. Z. 41, 321, and 42, 181, 1904 165 166 THE METABOLISM OF CYCLIC FORMATIONS [lect. enzyme exists in the intestine that has the power of removing urea from the arginine in one at any rate of the protamines, clupeine, without completely disintegrating it, the guanidine group being accessible to it even if the arginine is not detached ; so that after it has acted, the clupeone, as the altered clupeine is called, gives, when hydrolysed with acids, ornithine, which was not the case before. It is remarkable that kreatine, in which the same guanidine group is also attached to an acid, is not acted on by the arginase : whether this is due to the methyl group attached to the nitrogen atom at which the cleavage has to take place, or whether it is because in acid media kreatine tends to pass into the anhydride form, kreatinine, is not known. But we have in this fact an explanation of the behaviour of these substances in the body : how it is that while boiling with baryta removes urea from both kreatine and arginine, in the body arginine is completely destroyed, while the kreatine that is taken with food in meat passes through the body unattacked, and appears in the urine as kreatinine. Kreatinine is formed in the urine even when the food contains no kreatine. This endogenous kreatinine has been shown by Folin to be a remarkably constant quantity in each individual. In different people the amount is found to vary considerably, but for each person the amount excreted remains practically the same, whether much proteid is taken or little, provided that the food contains no kreatine. There is no other nitrogenous component of the excreta of which this is known to be true ; so that, whatever may be true of the other forms in which nitrogen is discharged from the body, the kreatinine excreted on a kreatine-free diet is a measure of the metabolic decay of tissue. On a low, proteid diet kreatinine may account for nearly 20 per cent, of the whole nitrogen excreted. There are many reasons for believing that it is in the muscles that it is formed, principally at any rate.i ^ Gregor, H.-S. Z. 31, 98, 1900 ; Liebig, Ann. 1848 ; Monari, M.J. 296,. 1889 ; Folin, Am. Jl. Phys. 14, 72, 1905 ; Burian, H.-S. Z. 43, 545, 1905. VIII.] ENDOGENOUS KREATININE 167 Muscular activity is followed by an increased output of kreatinine : more kreatine is found in working muscles than in those that have been at rest : the amount of kreatinine excreted is increased in fevers.^ The individual variations in the excretion of this substance seem to be determined by the mass of muscle.2 How the kreatine in the muscles comes into existence we do not as yet know. If we attach fundamental importance to the guanidine feature of the molecule, we shall be especially interested by the fact that Kutscher and Otori^ have found guanidine among the substances formed in autolysis of the pancreas, in traces also among the products of hydrolysis of. pseudo-mucin, and also that it is formed when certain proteids are oxidised by permanganate ; in this case in greater quantity than could be accounted for by supposing it to be derived from the arginine in those proteids. And that may lead to the suspicion that guanidine groups other than that of arginine are present in proteids generally, the muscle proteids more than the rest, and that the kreatine formed in muscles is a fragment of the proteid molecule, which in the process of separation has itself undergone but little change. Or, on the other hand, it is possible that the guanidine is an acquired characteristic of the compound ; that kreatinine is the more fundamental form with the ring structure, shown in its formula, CHo.N CH„ I I NH . C CO \/ NH and that this ring structure gives rise to the guanidine con- figuration, incidentally, as it were, when its ring is opened. This is a point to which we may return later. 1 Moritz, D. A. f. k. M. 46; Hofmann, V. A. 48, 1869; Salkowski, V. A. 88, 391. According to van Hoogenhuyze and Verploegh, H.-S. Z. 46, 415, 1905, it is only in starvation that muscular exertions increase the output of kreatinine appreciably. ^ Folin, Am.Jl. Pkys., xiii., 85, 1905. ' Kutscher and Otori, H.-S. Z. 42, 453, and 43, 98, 1904. 168 THE METABOLISM OF CYCLIC FORMATIONS [i.ect. In any case, it is important to remember that the kreatinine in the urine is probably a measure of physiological decay of muscular tissue. It is a substance particularly rich in nitrogen, which forms 37 per cent, of its weight, and so may account for a large proportion of the nitrogen waste of the muscles. Urea, at any rate, is formed in them to a very small extent in com- parison with the body generally : to what extent uric acid is also produced in the metabolism of muscle, we may leave for consideration later. One further point with regard to kreatine may be noted here : the presence of a methyl group attached to nitrogen, constituting it a methylamine derivative, is remarkable. Choline and its derivatives, and adrenaline, are the only other well- known instances of this combination in the body. But the methyl mercaptan found by Nencki^ in normal urine in traces, and in larger amounts after taking asparagus, and the excretion of tellurium ^ as methyl tellurium after admini- stration of tellurous acid, and the methylatibn of xanthine to heteroxanthine ^ (y-methyl xanthine) in the dog, are curiosities in this respect related to the methylamine derivatives. The other nitrogenous combinations found in, or associated with proteids, are all of them heterocyclic ring formations ; rings, that is, in which some of the atoms are not carbon atoms. The simplest of these rings is that found in pyrrhol — HC CH II II HC CH \/ NH Pyrrhol is formed when proteids are heated with baryta to 1 50° C. But the first simple derivative of pyrrhol to be iso- lated and identified among substances formed from proteids 1 Nencki, S. A. 28, 206, 1891. 2 Hofmeister, S. A. 33, 198, 1894. ' J^euberg, Salowski's Festschrift, 1904. viii.] PROLINE, ARGININE, AND GLUTAMIC ACID 169 under physiological conditions was pyrrholidine carboxylic acid — HnC CH, I I H„C CH . COOH \Z NH — proline, as it is called for short, which Emil Fischer i first found among the products of hydrolysis of casein and other proteids, in amounts varying from about 5 per cent, in gelatine to 3 per cent, in casein, and 1.5 per cent, in egg-albumin and haemo- globin. The question was at once raised by Fischer himself, whether this new compound was not formed by a secondary reaction in the course of the esterification process by which it was obtained ; — either arginine or diamido-valerianic acid could conceivably have given rise to it, but neither of these substances, treated by themselves in the same way as that in which the mixed products of proteid hydrolysis had been treated, yielded anything that behaved like proline.^ He has obtained it from casein by boiling with alkalies instead of acids ;3 but though it was not liberated by tryptic digestion, a polypeptide which resisted tryptic action for seven months, a substance precipitated by phos- photungstic acid but not showing the biuret reaction, was found, which, when boiled with acids, gave up all the proline as well as the phenyl-alanine that could be obtained from casein by direct hydrolysis.* Peptic digestion, on the other hand, especially if followed by tryptic, hydrolyses the greater part of this com- pound in which the proline is contained, and under these circumstances the proline can be isolated without exposing the mixture in which it is found to the action of hot acids at any stage of the process.^ Fischer concludes, therefore, that the proline is as much a primary product of proteid hydrolysis as leucine itself. It is even possible that it may prove to be the parent substance from which other constituents of the proteid molecule are descended, for another derivative, an oxyproline,* ' E. Fischer, JI.-S. Z. 33, 152, 1901. ^ lb., p. 170. 3 H.-SZ. 35, 227, 1902. * n.-S. Z. 39, 81, 1903. 5 H.-S. Z. 40, 216, 1903. <= B.-S. Z. 39, 155. 170 THE METABOLISM OF CYCLIC FORMATIONS [lect. has been found by Fischer in casein and gelatine. The exact position of the hydroxyl group has not been determined, but if it should prove to be in the S position, CH2 CHj I I HO . CH CH . COOH \/ NH then it may be evidence for the intramolecular oxidation of the pyrrhol ring to glutamic acid. From glutamic acid a pyrrholidone carboxylic acid ^ has been obtained, CH„ CH, I I CO CH . COOH \/ NH and Fischer came across this latter substance among those formed in the hydrolysis of horn, but preferred to regard it as a secondary product derived from glutamic acid. The possibility of a close relationship between proline and glutamic acid, so long as it remains a possibility, adds to the physiological interest of this pyrrhol derivative. Another substance, unquestionably of first-rate importance, in which the pyrrhol ring occurs is hsematine. Pyrrhol is formed when haematine is subjected to dry distillation,^ and also, as Nencki showed, when it is reduced with tin and hydro- chloric acid and distilled with alkali, or when it is fused with potash.^ No less important in vegetable physiology than haematine in the physiology of animals is chlorophyll, and the phylloporphyrine obtained from chlorophyll by Schunck and Marchlewski, which they proved to differ in its elementary composition from hsematoporphyrine only in containing two atoms of oxygen less, and spectroscopically to be almost indistinguishable from it, like h^matoporphyrine also gives ^ M.f. Ch. 3, 228. 2 Hoppe-Seyler, Med. chem. Untersuchungen, p. 533. 2 Nencki and Sieber, 5. A. 418, 1884. vtii.] H^MATINE 171 pyrrhol on distillation with zinc dust.* Nencki and Zaleski^ then obtained a further link connecting these two pigments, by preparing -from haemine or haematoporphyrine on reduction with phosphonium iodide two substances: mesqporphyrine, with the formula Ci8H;i8N202i coming, therefore, between TiaematD- porphyrine, CigHigNaOs, and phylloporphyrine, CigHj^gNjO ; « and besides mesoporphyrine, haemopyrrhol, which was also obtained by Marchlewski and Nencki from phylloporphyrine.* This haemopyrrhol, then, with the formula CgHjgN, is a common derivative of both chlorophyll and hasmatine, and from which- ever source it is obtained, it is oxidised on exposure to the air to urobiline. For haemopyrrhol Nencki and Zaleski suggested two possible formulae, of which one, that of a methyl propyl pyrrhol, II II HC CH \/ NH has been shown to be probably correct. Kiister^ oxidised haematine or haematoporphyrine with bichromate in glacial acetic acid, and obtained a substance, a haematinic acid with the formula CgHgNO^, which on dry distillation also gave pyrrhol, and when treated with alkali gave up ammonia, and was converted into a second haematinic acid, which differed \/ from the other by having exchanged an NH group for an atom of oxygen. The relation of these two acids was clearly to be expressed by the formulae RC ===CR' RC = CR' CO CO and CO CO NH O ^ Schunck and Marchlewski, Ann. 278, 284, 288, and 290, 1896. ^ Nencki and Zaleski, B. 34, 997, 1901. ' These formulas are not final ; cf. Zaleski, ff.-S. 37, 73, 1902. * Marchlewski and Nencki, B. 34, 1687, 1901. ^ Kuster, I/.-S. Z. 28, i, 1899 ; and, 44, 391, 1905. 172 THE METABOLISM OF CYCLIC FORMATIONS [lect. the imide and anhydride respectively of the same tribasic acid. The value of R and R' in these formulae was determined by the fact that the anhydride on further oxidation gave succinic acid, COOH . CHg. CHg. COOH, and also could be converted by the loss of CO2 into a substance identical with methyl ethyl maleic anhydride — CHq , C =^^= C • CijHk I I CO CO \/ o The value of R, therefore, is CHg, and that of R' is CHj-CHj. COOH ; and the constitution of haemopyrrhol is almost proved to be CHo . C C . CoHb ,11 II H.C C.H \/ NH — methyl propyl pyrrhol. Its synthesis by reduction of methyl propyl maleic imide has indeed been already attempted, though the results are not con- clusive as yet.-' It is, therefore, beginning to be possible to look forward to the synthesis of the blood pigments, and to the opening up of entirely new prospects over the field of biological chemistry. The pyrrhol formation occurs in another important combina- tion. Indol, with the formula, CH ^\ HC C CH I II II = C^H^N HC C CH CH NH may be regarded as compounded of benzene and pyrrhol grafted on to each other so as to share two carbon atoms in common. All the known derivatives of indol occurring in the body can be traced to one and the same origin, one of the ' Marchlewski and Buraczewski, J/.-S. Z. 43, 410, 1905. vin.] TRYPTOPHANE: ITS CONSTITUTION 173 compound amide acids that enters into the composition of proteids, with but few exceptions such as gelatine, and is set free by the hydrolytic action both of acids and of trypsine, from the latter of which properties it received its name of tryptophane long before it was identified. The constitution of this substance has not been precisely determined. Hopkins and Cole.i who first isolated and described it, found that the substances produced by bacteria in the intestine, known as scatol acetic and carboxylic acids, and scatol itself, Were formed by bacteria from tryptophane, which they therefore regarded as scatol amido-acetic acid. But it has since been shown ^ that the substance which had been regarded as scatol carboxylic acid is really indol acetic acid. For it is identical with the product of the action of zinc chloride upon the phenyl hydrazone of aldehydo- propionic acid, %N . N : CH . CHg . CH2 . COOH, condensation taking place thus — CgH^ C . CHj . COOH NH CH The acid, therefore, known as scatol acetic must really be indol propionic, and tryptophane be indol amido-propionic acid. But Ellinger made another interesting contribution to the chemistry and physiology of tryptophane by discovering that when administered to dogs it was converted into kynurenic acid, which he found in the urine.' Now this substance was known to have the constitution represented in the formula OH /\/\ COOH N ^ Hopkins and Cole,//, of Phys. 27, 418, and 29,451, 1903. 2 Ellinger, B. 37, 1801, 1904. 3 Ellinger, H.-S. Z. 43, 325, 1904. 174 THE METABOLISM OF CYCLIC FORMATIONS [lect The conversion of an indol amido-propionic acid into a body of this constitution is difficult to imagine. Ellinger attempted to solve the difficulty by ascribing to tryptophane the constitu- tion a-indol /3-amido-propionic acid, and represented the change graphically thus : CH, NH. yCH . COOH y \COOH C _^ I I OH CH NH But it is a question whether this is the only or the most probable way in which the change could occur, and it is, therefore, not necessary to regard tryptophane as a /3-amido acid. The amido acids that occur in Nature are almost without exception the a- acids, and it is questionable whether there is any justification for the supposition that in tryptophane we have an exception to this rule. The only instance of a /3-amido acid in the chemistry of living organisms is the iso-cystine described by Neuberg and Mayer.i This is a cystine which they found in the urinary calculi, differing in several respects from that which is obtained from proteids and from that found in the urine of cystinuric patients, so far as we know at present. This iso-cystine they think is derived from ^S-amido a-thiolactic acid instead of the a-amido ^S-thiolactic acid which is ordinary cysteine : from CH2.NH2 HS.CH2 HS . CH and not from CH . NH„ I I COOH COOH Morner,^ also, from a study of the cystine obtained from proteids, came to the conclusion that these two kinds of cysteine entered into its composition. For, when he heated it under ' Neuberg and Mayer, U.S. Z. 44, 472, 1905 ; cf. H.-S. Z. 43, 338, 1904. ^ Morner, H.-S. Z., p. 363, 1904. VIII.] ISO-CYSTINE 175 pressure with hydrochloric acid, besides ammonia and sulphu- retted hydrogen, he found that both a-amido-propionic acid and a-thiolactic acid were formed. Now both Neuberg and Friedmann ^ have made it clear that in cystine the sulphur and the nitrogen are united to different carbon atoms, and not the same one as Baumann supposed. In the molecule, therefore, from which a-thiolactic acid was formed, the amido group, he argues, must have been in the ^ position. The occurrence side by side of these two cysteines would be most easily intelligible on the supposition, proposed by Neuberg,^ that they were both formed from the same substance, COOH I HS.CH I NHXH I COOH, by the removal of CO^ from the two different ends of the molecule : just as from the corresponding oxy-aspartic acid, which has actually been obtained by Skraup* from casein, it may be expected that in addition to serine, /3-oxy a-amido- propionic acid, the iso-serine with the hydroxyl and amido groups transposed, which has been synthesised by Fischer and Leuchs,* should be formed. But the iso-cystine found by Neuberg and Mayer is at any rate unusual, even in cystine stones, and the actual sample they prepared was not free from tyrosine, or at any rate gave Millon's reaction, as Fischer^ showed ; and Fischer suggests that this impurity may possibly account for some of the peculiarities of the cystine, which were the only ground for supposing it to contain the amido group in the j8 position. We do not, therefore, positively know of any (8-amido acid ' Neuberg, B. 35, 3161, 1901 ; Friedmann, ff. B. 3, i, and 184, 1902. = Neuberg, B.-S. Z. 43, 340. ^ Skraup, H.-S. Z, 42, 274, 1904. * Fischer and Leuchs, B. 35, 3790, 1902. ° Fischer and Suzuki, H.-S. Z. 45, 410, 1905. 176 THE METABOLISM Of CYCLIC FORMATIONS [leci'. in Nature, and are perhaps justified in hesitating to accept a constitutional formula for tryptophane which has this anomaly.^ But, however we may have to account for the formation of kynurenic acid from tryptophane, the possibility of the deriva- tion of a quinoline compound from this group, which is present in all typical proteids, is one of great interest, and, as Ellinger suggests, it is possible that the alkaloids with a quinoline or pyridine nucleus that occur in plants may be derived from tryptophane by similar reactions in vegetable metabolism. The part played by tryptophane in animal metabolism is obscure. It is well known that in the intestine it is set free by the pancreatic juice, and that bacteria in the intestine decom- pose it in so far as it is not absorbed unchanged. In this decomposition, however, it is the side chain alone that is attacked ; denitrification gives rise to indol propionic acid, oxidation of this to indol acetic acid, and this losing COg becomes scatol, or by oxidation nearer the chain, and further loss of CO, indol itself ^ Of these products of bacterial change, indol acetic acid may be found in the urine,^ and indol and scatol oxidised to the corresponding phenol and combined with glycuronic or sulphuric acid are constantly found in small amount in human urine. But the principal part of these changes is the work of bacteria, and not the metabolism of the body itself Tryptophane is contained in the proteids which break down in metabolism, and is set free, not only by trypsine, but by the cellular enzymes. Does this endogenous tryptophane and the tryptophane that is absorbed unchanged from the intestine undergo the same changes in the cells as the bacteria bring about in the intestine ? If so, does the indol ring under these conditions escape into the urine unbroken, and contribute to the urinary indican? This is a question which has been 1 More recently Ellinger has synthesised the acid formerly known as scatol acetic acid, and from the synthesis it is clear, not only that this acid is indol propionic acid, but that the indol is attached to the /3-carbon atom, and not as in Ellinger's suggested formula for tryptophane to the o atom. — B. 38, 2884, 1905. '^ Salkowski, H.-S. Z. 9, 27, 1885. ^ jj^d. VIII.] ORIGIN OF URINARY INDICAN 177 warmly disputed. It was taught till recently that all the indican in the urine was the result of bacterial decomposition. Jaffe i showed how obstruction in the small intestine in dogs caused a great increase in the indican excreted ; obstruction of the large intestine, a much smaller increase, if any — clinical experience in man agrees with this. Ellinger and Prutz^ have shown that the obstruction caused by anti-peristalsis, induced by dividing the small intestine in two places a few inches apart and joining the intervening length of intestine on again the wrong way round, caused the indican excretion to be increased enormously — twenty and thirtyfold. No such pronounced alteration as this is known to be set up in any way in which the operation of intestinal decomposition can be excluded. Recently, however, Blumenthal ^ has maintained that some of the urinary indican is of a different origin. Rabbits, under ordinary conditions, excrete, he says, no indican, but in starvation, when they are living on their tissues, indican is found in their urine. Diabetic puncture in these animals also brings on indicanuria, and in man he traces certain cases of marked indicanuria to nervous disturbances. The increased proteid destruction set up by phlorrhizine in rabbits also is accompanied by indicanuria.* But the indican excreted by starving rabbits is differently explained by Ellinger.^ These animals, when food is withheld, begin almost on the first day to consume their own dejecta, and if by effectual muzzling they are prevented from doing this, the indicanuria is not observed. The experiments with phlorrhizine, repeated by Scholtz,^ gave negative results. So that the view that some of the indican in the urine is formed in tissue metabolism is not well supported by evidence. As a matter of fact, subcutaneous injection of tryptophane in rabbits and dogs does not increase the indican in the urine,^ while ' Jafr^ V. A. 70, p. JT, 1870. ^ Ellinger and Prutz, H.-S. Z. 38, 400, 1903. ^ Blumenthal, M. J., p. 817, 1902. * Lewin, H. B. i, 493, 1902. ' Ellinger, H.-S. Z. 39, 52, 1903. ° Scholtz, H.-S. Z. 38, 513, 1903. ' Ellinger and Gentzen ,H. B., 4, 174, 1903 ; Rosenfeld, H. B., v., 83, 1903. M 178 THE METABOLISM OF CYCLIC FORMATIONS [lect. subcutaneous injections of indol does.' So that it is not likely that tryptophane is dealt with by the cells in the body in the same way as it is by the bacteria in the intestine, or, in other words, it is improbable that the breakdown of tryptophane in metabolism leads to the liberation of indol. In some animals, at any rate rabbits and dogs, it is partially changed into kynurenic acid,^ but the bulk of it apparently is oxidised, and the indol ring itself is not spared. Closely related to pyrrhol is imidazol, which is also a ring of five atoms, but of these five, two are nitrogen : HC NH HC— N The corresponding 6-atom ring, containing one additional carboij atom, is pyrimidine, N = CH HC CH • ■ ■ II II ' ■ N CH in which it is customary to number the several atoms, beginning from the nitrogen at the top of the above molecular diagram, as No. I, working down the left-hand column and then up the right-hand column to the carbon No. 6 at the top on the right. These two rings are most familiar to us in the form of the compound ring that results from their fusion, and is purine. Just as indol may be regarded as benzene grafted on to pyrrhol, purine is pyrimidine grafted upon imidazol. Since both imidazol and pyrimidine rings occur in substances of physiological importance, they may be dealt with individually before we pass to the more familiar purine. Imidazol has recently been proved to enter into the forma- ' Grosser, H.-S. Z. 44, 320, 1905. 2 PUinirer, H.-S. Z. 43, 325, 1904. viii ] IMIDAZOL DERIVATIVES 179 tion of histidine. This base, first described by Kossel,i who obtained it by hydrolysing the protamine sturine, was shown by Hedin^ to be present in a large number of proteids. Its empirical formula, CgHgNgOj, showed that it could not be built on a saturated open chain of carbon atoms, as the amount of hydrogen was too small. Certain facts with regard to its con- stitution came to light from time to time,^ but a correct inference as to the arrangement of the atoms in the molecule was first made by Pauly,* who pronounced for a compound of imidazol and alanine. This was proved to be correct by Knoop and Windaus, who synthesised imidazol propionic acid, and showed that it was the same substance as that formed from histidine by substituting hydroxyl for the amido group and then reducing the product.^ Histidine, therefore, like tyrosine, phenyl-alanine, cysteine, tryptophane, and serine, is a compound of amido-propionic acid. But histidine is not the only substance familiar in physiology in which the imidazol ring occurs. Kreatinine is also built up upon this plan, as is seen from its formula — CH,- N/ >c = NH CO NH' and it is possible that this may furnish a clue as to the mode of its formation in muscle. And the interest attaching to such imidazol derivatives has been greatly increased by the remark- able fact observed by Knoop and Windaus,' that out of glucose, in the presence of ammonia and zinc oxide, methyl imidazol is formed in large quantities under the influence of light. This they suppose to take place through the condensation of pyruvic ^ Kossel, II.-S. Z. 22, 176, 1896. 2 Hedin, H.-S. Z. 22, 191, 1896. ^ Herzog, H.-S. Z. 37, 248 ; Frankel, M. J., p. 152, 1903. * Pauly, H.-S. Z. 42, 508. ^ Knoop and Windaus, H. B. 7, 144, 1905. ^ Knoop and Windaus, B. 38, 5 ; H. B., vi., 392, 1905. 180 THE METABOLISM OF CYCLIC FORMATIONS [lect. aldehyde with two molecules of ammonia and one of formal- dehyde — CH„ C=0 + NH, .H CH„ -O c = o NH 't" + NH, \h \cH HC N Pyruvic aldehyde has, as we have seen, been regarded, as one of the probable intermediate products in the formation of lactic acid from sugar. They point out that the methyl imidazol formation occurs in purine, as is best seen from the formula for xanthine — NH— CO CO C— NH. \ II >CH NH— C — N -^ which is an ureide of imidazol carboxylic acid. The simple derivatives of pyrimidine, Ni=flCH H . Cj gCH II II N3 *CH that have been obtained from the animal body have all of them been separated from the products of hydrolysis of nucleic acids. If nucleic acids are heated under pressure with acids, or boiled for twelve hours or more with 30 to 40 per cent, sulphuric acid, one or more of the following substances are found to have been split off — uracil, thymine, or cytosine. The nucleic acid of herrings' sperma gives all three of them. Thymine has not been obtained from nucleic acids of vegetable origin ; nucleic acids of animal origin give one or other, generally two, of these bodies. Uracil has been synthesised by Fischer and Roeder ^ ' Fischer and Roeder, B. 34, 3751, 1901. VIII.] PYRIMIDINE DERIVATIVES 181 from acrylic acid and urea by removing two hydrogen atoms from the dihydro uracil iirst formed when these two substances condense, and thymine in the same way was obtained by them from urea and methyl acrylic acid. Their constitution is therefore represented thus : HN— CO H . N C=0 II II OC CH 0=C C.CH, I II I II HN— CH H . N— CH Uracil = 2.6-Oxy pyrinidine. Thymine = 5-Methyl uracil. Cytosine has also been synthesised by Wheeler and Johnson,^ and its constitution therefore finally determined. It is 2-oxy 6-amido pyrimidine, N=C.NH„ I I 0=C C . H I II HN C . H The question whether these pyrimidine derivatives are actually as such contained in the nucleic acids from which they have been obtained, or whether they are derived from the purine bases of the nucleic acids, has been discussed but not conclusively settled. The opinion of those who have worked most on these bodies seems to be, generally, that the pyrimidine nucleus is present in the nucleic acids in the simple form as well as in the form of purine.^ And in the case of thymine, at any rate, the methyl group does not correspond to anything in any of the purine bases. But much more energetic treatment is necessary to split them off than is usually necessary in hydrolysis, and though the purine bases, uncontaminated with other substances when treated in this way, are not decomposed so as to yield pyrimidine derivatives, it is possible that they ' Wheeler and Johnson, Am. Ch. Jl. 29, p. 492, 1903. '^ Osborne and Harris, H.-S. Z. 36, 109 ; Kossel and Stendel, H.-S. Z. 38, 49- 182 THE METABOLISM OF CYCLIC FORMATIONS [lect. might be so decomposed in the presence of the carbohydrate and other constituents of the nucleic acids.^ Purine may, as we have seen, be considered as a compound ring in which imidazol is grafted on pyrimidine. And it may be as well to bear this conception of the structure of purine in mind. It is true that nothing that is at present known points to this relationship having any physiological significance. But, then, at present we do not know how the purine bases come into being in the body, as they certainly may. Young animals, during suckling at any rate, take up from the milk something which is not purine, from which they make these bases, and the increase in the amount of bases in their bodies is about proportional to the increase in weight.^ A synthesis of purine must also take place in the developing chick, since eggs contain no purine. It has been suggested by Knoop and Windaus, though as yet there is no direct evidence for the suggestion, that the methyl imidazol which can be formed from sugar and ammonia in sunlight could, after oxidation of the methyl group, condense with urea and so form xanthine. The possibility of a simple synthesis of a ring so closely related to purine from just the sort of material that would be available, is at any rate interesting. The question of the origin of purine in the body introduces the whole of what is still one of the most difificult and vexed questions in metabolism — the origin and meaning of uric acid in the body. It was diiificult, in the days before the chemical constitution and relationships of uric acid had been cleared up, and before the distribution and physiological importance of the xanthine bases in the nuclei of all cells had been appreciated. But even now that these advances have been made, the subject remains a difficult one, and at the same time has gained in significance. It divides itself now into a number of chapters. Uric acid is formed in the body from related substances which are present free or combined in the food ; this is known as the exogenous uric acid. It is supposed to be formed from similar substances, constituents of cells in the body, set free when the 1 Cf. Burian, Ergeb., vol. iii., p. 99. '^ Burian and Schur, U.S. Z. 23, 55, 1897. VIII.] OXIDATION OF PURINE BASES 183 cells perish, or when, as a result possibly of their activity, they give up certain components of their nuclei. And this endo- genous uric acid, like the exogenous uric acid, is formed by the oxidation of less highly oxidised derivatives of purine. But besides this; in birds and reptiles certainly, the purine ring is put together from substances constitutionally not related to it, all forms of nitrogenous combinations contributing nitrogen for the elaboration of this highly complex excretory substance, and there are grounds for thinking that, to some extent at any rate, this synthesis takes place in mammalian animals and man. This question, then, has to be considered, whether uric acid is ever made from material that has not already the purine stamp upon it ; and then there still remains the question the converse of this, whether whatever has the shape of purine is necessarily indestructible, and though it may be oxidised to uric acid, must then be excreted as uric acid, being incapable of undergoing changes by which it would cease to be purine at all ; whether variations in the amount of uric acid excreted in man, for instance, may not depend on variations in the activity of normal processes in which the purine ring is split, and uric acid or its immediate predecessors destroyed. Under these four headings may be grouped most of what has recently been done to advance our knowledge of this side of the nitrogenous metabolism of the body. To begin with the exogenous uric acid, it is now universally acknowledged that uric acid is formed from the purine bases, free and combined, present in the food. But this has not been long established. It was some years after Miescher and Kossel had indicated the significance of the wide distribution of xanthine bases and their presence in all nuclei, that v. Mach and Minkowski ^ showed that in birds, after the removal of the liver, hypoxanthine given by the mouth, unlike most nitrogenous compounds, was still converted into uric acid. The power of synthesising uric acid was to a great extent lost, but the oxidation of hypoxanthine to uric acid was a different matter, and was not affected by removal of the liver. Attempts to 1 V. Mach, S. A. 24, 398, 1888. 184 THE METABOLISM OF CYCLIC FORMATIONS [lect. prove the direct conversion of xanthine bases into uric acid in mammalian animals and man, however, were for some time unsuccessful. Horbaczewski in 1891 found that food containing nucleine increased the uric acid excretion in rabbits and men ; but he explains this by supposing that it was the Jeucocytosis set up by the nucleine and not the nucleine itself, that gave rise to the uric acid ; others subsequently confirmed his result again and again, but showed that his explanation would not hold.^ But Minkowski proved that even in dogs, though much less than in men, feeding with nucleic acid from the testes of the salmon caused a rise in the excretion of uric acid, and that the xanthine bases themselves too, especially hypoxanthine, had the same effect,^ The old idea that uric acid was formed in the catabolism of proteids in general, but was for the most part normally converted into urea, and that excess of uric acid could be accounted for by failure of this final stage of nitrogenous metabolism, had to be abandoned ; for proteids such as those in eggs, which contain no purine, do not increase the uric acid excretion.^ The mistake had been fostered by the fact that meat contains purine bases, though but little nucleine : 100 g. of meat has on an average 60 mg. of purine bases in it, only one quarter of which is combined.* Incidentally it should be noted that the uric acid excretion was found to be less than the theoretical yield for a given amount of purine, in man as well as in the dog; but in this animal the loss of purine in the body was much greater than in man. But the formation of uric acid from nucleic acid and purine contained in the food, does not account for all the uric acid in the urine, since it is well known that the urine contains uric acid when no purine of any kind is allowed to enter the body. The source that suggests itself for this endogenous 1 Horbaczewski, M. f. Ch. 12, 221 ; Weintraud, B. k. W., 1895 5 Hess and SchmoU, S. A. yj, 243, 1896. 2 Minkowski, 5. ^.41, 375, 1898. 3 Hess arid SchmoU, loc. cit. * Burian and Schur, Pfl. A.%o, 241, 1900. VIII.] ENDOGENOUS URIC ACID 185 purine is, naturally enough, the nucleic acid of the cells in the body. The chemical changes would be the same in the one case as in the other. But it is an assumption that is not necessarily justifiable, to suppose that nuclear substance, wherever it may be in the body, when breaking down, breaks down in this particular way merely because the reactions involved are not altogether foreign to animal organisms. And methods have not yet been discovered by which the secrets of nuclear metabolism can be revealed. Such evidence as we have, however, is strongly in favour of this hypothesis, which is, in fact, very generally adopted. We know that in autolysis nucleic acids give up purine bases.^ We know that while the bases actually present in nucleic acids and set free by hydrolysis with acids are the amino bases adenine and guanine (6-amino purine and 2-amino 6-oxypurine respectively),^ the bases that are found in the solution of the organs that results from their autolysis are much rather the oxypurines corresponding to these, hypoxanthine and xanthine. The amino purines tend to disappear and give place to the oxypurines during autolysis,^ and it often happens that no trace of either adenine or guanine can be detected in the final products. Also if these bases are added to an extract of the organs, under these circumstances too they are oxidised. In some cases the reaction is confined to adenine, and guanine is unaffected, in others both alike are thus denitrified and oxidised.* These autolytic reactions, which are ascribed to enzymes known as adenase and guanase, bring us nearer to uric acid. But on referring to the molecular diagrams of these substances it will be obvious that a second reaction, different from this, is neces- sary in order to produce uric acid. The group in adenine or ' Salkowski, Z. f. k. M., 1890; cf, Salomon, M. /., p. 106, 1881. 2 Schmiedeberg, 5. A. 43, 57, 1900 ; Kossel and Neumann, B. 27, 2215, 1894 ; Osborne and Harris, H.-S. Z. 36, 102, 1902 ; Levene, H.-S. Z. 32, 545, 1901 ; and 37, 404, 1902. Cf. Burian, Ergeb. iii., p. 86. 3 Jones, H.-S. Z. 42, 343, 1904. * Jones, H.-S. Z. 45, 89, 1905. 186 THE METABOLISM OF CYCLIC FORMATIONS [lect. guanine which is affected by the change we have so far con- sidered is C = NH, and this is converted into C = 0. But to I I convert xanthine into uric acid or hypoxanthine into xanthine the change is a different one, and involves the oxidation of the group — N — NH \ \ C. H to the group 0=0. The formation of uric acid from purine bases resulting from this change was first observed by Horbaczewski, in his well- known experiments with the spleen, which marked an epoch in the physiology of uric acid.i Spleen pulp was heated with eight to ten volumes of water at 50° C. for some hours till decomposition began. It was then precipitated with basic lead acetate and filtered ; and the filtrate mixed with an equal volume of blood was kept at 38° C. with a current of air passing through it. After a few hours adenine and guanine could not be found, and uric acid had been formed. He thought that the initial bacterial decomposition was necessary, but that at a certain stage it must be checked by the use of lead acetate. But it was proved that in that he was mistaken. Spitzer 2 found that extracts of the spleen and liver treated with thymol chloroform or sodium fluoride converted the xanthine and hypoxanthine which he added to them, when supplied with a current of air at 50° C, into uric acid. And Schittenhelm ^ was able to precipitate with two volumes of saturated ammonium sulphate solution something contained in the extract of the liver, spleen, or lungs, a solution of which in a current of air converted guanine in some cases quantitatively into uric acid. If the air-stream were omitted, xanthine was formed, and not 1 Horbaczewski, M.f. Ch. 12, 221, 1891. 2 Spitzer, Pfl. A. 76, 192, 1899. ^ Schittenhelm, H.-S. Z. 42, 253, and 43, 228, 1904. Cf., too, Burian, H.-S. Z. 43, 497, 1904. viii.J SYNTHESIS OF URIC ACID IN THE BODY 187 uric acid ; in other words the reaction involving denitrification took place, but not the second reaction by which the CH group II is oxidised. In this way we have learnt to look upon the formation of uric acid from nuclear purine as a probable reaction in nuclear metabolism. But it is not proved in the same way that the formation of uric acid from food purine is proved. And in order to account for the uric acid that is excreted when no purine is taken with the food, we are not compelled to take this probability, however great it may be, for proof. For there is another way in which it is possible that such endogenous uric acid may arise, and that is by synthesis from material that is not already in the purine form. It has long been known that birds and reptiles excrete 70 per cent, of their nitrogen as uric acid. This amount cannot all have been purine to start with, and in fact all kinds of nitrogenous compounds in these animals contribute to the output of uric acid. Minkowski's experiments on the removal of the liver in geese made it probable that in this synthesis ammonia is first converted into urea, and the urea instead of being ejected at once is further elaborated by condensation with lactic acid to uric acid. This conception was confirmed by Salaskin and Kovalevski,^ who on perfusion of the excised livers of birds with blood containing ammonium lactate found that uric acid was formed. And Wiener 2 proved it in another way. Since urea administered to birds is excreted as uric acid, whatever else is necessary for the synthesis of uric acid from urea must be present in the organism in larger amount than is required for the ordinary output of uric acid. But if the dose of urea be progressively increased, a point is reached at which the animal begins to excrete the urea unchanged, at which, it may be assumed, the supply of the other factor necessary for the synthesis is exhausted. If at this point lactic acid be injected in addition to the urea, then the urea is no longer excreted unchanged, but uric ' Salaskin and Kovalevski, H.-S. Z. 33, 210, 1901. ^ Wiener, H. B., ii., 42, 1902. 188 THE METABOLISM OF CYCLIC FORMATIONS [i.ect. acid begins to be formed again. Still more effectual than lactic acid in this respect are certain other organic acids, also containing three carbon atoms in chain, notably malonic acid, COOH.CH2.COOH, and tartronic acid, COOH.CHOH.COOH. These are substances which might well be formed from lactic acid by oxidation ; and the reason why they give better yields of uric acid than does lactic acid itself may well be that lactic acid does not all of it undergo oxidation into these substances, and that part of it, reacting in other ways, is diverted into other channels, and so rendered unavailable for the synthesis. In his experiments, malonic, tartronic, and mesoxalic acids gave quantitative yields: glycerine, lactic and pyruvic acids about 40 per cent. ; while the four carbon acids, butyric, succinic, and malic, gave none. With the clearer insight into the mode of origin of uric acid in birds thus obtained, the question arose whether the reactions involved were peculiar to those animals whose nitrogen excretion is mainly in this form. Wiener tested some of these acids on human beings, and found a definite, though small, increase in the uric acid excretion after giving malonic, lactic or dialuric acid (the monoureide of tartronic acid, /NH . CO. CO. >CHOH) ; \nh.co/ and finally he found that an extract of ox-liver, containing 0.2 per cent, sodium fluoride, after being shaken with tartronic acid for four hours, contained more uric acid than a portion similarly treated but with no addition. These experiments led him to the view that a synthesis of uric acid from urea and oxidation products of lactic acid occurred in the liver of mammalian animals as in that of birds, though, of course, to a very much smaller extent. Wiener was led to making these experiments by some curious results noted by Hopkins and Hope.i They found that when they took sweetbread which had been digested artificially ' Hopkins and Hope,//, of Phys. 23, 271, 1898. VIII.] SYNTHESIS OF HYPOXANTHINE 189 with pepsine, and then freed by filtration from nucleine,, a marked increase in the uric acid excreted was constantly observed in spite of the removal of the nucleine. But in these experiments it was not quite clear that there were not sufficient purine bases free in the filtered solution of digested thymus to account for the increased uric acid. And Wiener's experiments, though sugges- tive, are not universally accepted as establishing the synthesis which he supposes. Burian is also one of those who are not satisfied that the nuclear metabolism of the body should be made to account for the whole of the endogenous uric acid.^ But he has been led to a conception of the synthesis of uric acid which is entirely different from that propounded by Wiener. Endogenous uric acid takes its origin, according to him, principally in the muscles. Here in the first instance hypoxanthine is formed by synthesis, and this is then subsequently oxidised to uric acid by the ferment effecting this change which the muscles are known to contain. This view is based on the fact determined by him, that muscular activity is followed in the first hour by increased excretion of purine bases other than uric acid, and then for an hour or two more by increased excretion of uric acid.^ And in confirmation of this he finds that excised muscles of dogs, per- fused with blood, give up to the blood during activity, at first more bases and afterwards more uric acid. He thinks, then, that the muscles are the principal source of endogenous uric acid, that in producing it they build up the purine ring, and that hypoxan- thine is a product of muscular metabolism, just as kreatine is a substance which there is also reason for thinking is produced in quantities that vary with muscular activity. How the hypoxan- thine and kreatine are put together is not yet known, but it is of interest in this connection to remember that kreatinine is essen- tially an imidazol derivative, that the imidazol formation occurs in purine derivatives, and that methyl imidazol can be obtained in vitro directly from sugar and ammonia by the action of sun- light. There is not a particle of evidence that there is any ' Burian, H.-S. Z. 43, 533, 1905. 2 Cf., too, van Hoogenhuyze and Verploeghj H.-S. Z, 46, 442, 1905. 190 THE METABOLISM OF CYCLIC FORMATIONS [lect. genetic relationship of this kind between these substances ; but there is not a particle of evidence for any theory to account for their origin in the muscles, either independently or in association. Then, lastly, in addition to the three modes in which it has been supposed that uric acid may arise in the body, two of which are by oxidation of ready-made purine, and the third by some synthetic process, we have to remember that uric acid itself is not indestructible. Nearly sixty years ago, Wohler and Frerichs ^ showed that uric acid given to dogs was converted into urea, and this has been confirmed again and again since then. Spiegelberg found that he could in these animals recover, as uric acid in the urine, only about 5 per cent, of the uric acid administered by the mouth if the dogs were full-grown, but in puppies as much as 50 per cent. Burian and Schur^ determined what they call the integrative factor for uric acid ; that is to say, the number of milligrams of uric acid that must reach the blood in order that I milligram should be excreted in the urine : in dogs this integrative factor is 22, agreeing closely with Spiegelberg's figure, in rabbits 6, in man 2. This factor is clearly a measure of the destruction of uric acid in the organism of the species ; and they argued that they could determine the amount of uric acid that had been thrown into the circulation, whether exo- genous or endogenous, by multiplying the amount excreted by the factor for the species. Whether such calculations are con- vincing or not, they showed that the destruction of uric acid is carried out largely in the liver. For excision of the kidneys caused no accumulation of uric acid in the blood so long as it was allowed to circulate through the liver ; but if the liver was cut out of the circulation, in addition to the kidneys being removed, then accumulation of uric acid did take place. And there seems to be no doubt whatever that besides the liver other organs are also able to destroy uric acid. Wiener^ found that ' Wohler and Frerichs, Ann. 65, 1848. ^ Burian and Schur, Pfl. A. 80, 241, 1900 ; and 87, 239, 1901. 3 Wiener, S. A. 42, 375, 1899. Cf., too, Jacoby, V. A. 157, 235, 1899; Lang, /r. B., v., 330, 1904. viii.] DESTRUCTION OF URIC ACID IN THE BODY 191 extracts of the liver of dogs and pigs, and of the kidney, though not the liver, of oxen, shaken for four hours at body-temperature with uric acid, contained at the end of that time a much smaller amount of uric acid than portions that had been boiled, but otherwise treated in the same way. In some cases, out of 140 mg. of uric acid added to each sample only 3 or 4 mg. could be recovered, while in the boiled sample nearly the whole was recovered. Schittenhelm ^ has been able by means of uranyl acetate to separate the substance in the kidney to which this action is due, and found that a solution of it, supplied with air at 38° C. in the presence of chloroform, destroyed from 80 to 100 per cent, of the uric acid that he added, whereas if the solution were boiled this was not the case. But what it is that is formed from the uric acid which dis- appears in the body, or in these experiments in vitro, has not been determined. The decomposition of uric acid under almost every possible variety of conditions was studied, in the early attempts to determine its constitution. And many different substances were obtained from it. When heated with hydro- chloric acid to 170° C. it breaks up into glycocoll, carbonic acid, and ammonia. It was accordingly surmised that glycocoll might be formed when uric acid was destroyed in the body. Wiener ^ tested this conjecture by saturating rabbits with benzoic acid introduced into the stomach, giving quantities larger than the animal's own stock of glycocoll could combine with to form hippuric acid. If uric acid were then injected subcutaneously, more hippuric acid was formed, and at the same time the amount of benzoic acid that could be given without killing the animal was increased. He also found that when an extract of ox- kidney, obtained by shaking the pulped organ with normal saline containing 0.2 per cent, of sodium fluoride for some time and then straining, was digested with uric acid, not only did the uric acid disappear, as we have seen above, but glycocoll was found in larger quantities than was the case if no uric acid was added. The amount of glycocoll actually found was enough to ' Schittenhelm, H.-S. Z. 45, 160, 1905. 2 Wiener, S. A. 40, 313, 1898. 192 THE METABOLISM OF CYCLIC FORMATIONS [lect. account for about half the amount of uric acid that had disappeared, on the supposition that a molecule of uric acid gives rise to one of glycocoll. It is possible that the glycocoll is not formed directly from the uric acid ; some intermediate product may intervene in a reaction taking place in two stages, and the amount of glycocoll found in this experiment may have been low, for the reason that some of the uric acid had been carried through the first stage but not the second. But if it is proved that uric acid in the kidney gives rise to glycocoll, it does not follow that uric acid is disposed of in the same way in other organs of the body in which it is known to be destroyed. There are other destructive decompositions of uric acid that have been well known from the early days of the chemistry of this substarice. With permanganate of potash in the cold, uric acid is oxidised to allantoine — NH . CO NH„ /I / CO C— NH. —->■ CO .CO.NH. \ II >co \ >CO NH . C— NH/ NH.CH.NH/ Now allantoine is found in the urine of calves and eows,^ in that of new-born children,^ and not infrequently in that of dogs and even adult human beings.' On autolysis of the liver allantoine is formed, and in hydrazine poisoning, in which the liver undergoes severe disintegrative change, this organ is found to contain much allantoine.* Salkowski administered 8 g. of uric acid to a dog, and found that the urine contained about 1.5 g. of allantoine ; and he notes that dogs fed on meat frequently excrete allantoine, but show in this respect individual peculiari- ties, the allantoine excreted in some animals being replaced by ^ Salkowski, //.-S. Z. 42, 220, 1904. 2 Prout, Med.-CMr. Tram. 8, 526, 1818. ^ Cf. Minkowski, S. A. 41, 397, 1898. « Pohl, 5. A. 48, 367, 1903. vin.] ALLANTOINE AND OXALIC ACID 193 uric acid in others,^ Minkowski found that hypoxanthine administered to dogs appeared as allantoine in the urine, while in man it was excreted as uric acid. After feeding on thymus, too, dogs excreted considerable amounts of allantoine, while in men under these conditions none was excreted. All this suggests that allantoine is formed when uric acid is destroyed in the body, and that allantoine is not commonly found in the urine of either dogs or men, because it is itself further acted on and destroyed. Given by the mouth, it is true, it is not broken up in the dog ; as much as 90 per cent, has been recovered unchanged in the urine.^ But it may well be that, when it arises from uric acid in the cells, it is attacked on the spot, and not allowed to reach the blood or the kidneys. In man, on the other hand, even when taken by the mouth, from 50 to 70 per cent, of it disappears, and only the smaller part is found in the urine. Then, again, from allantoine in vitro oxidising agents give rise to the formation of oxalic acid and urea. Even potash splits oiif oxalic acid from this body.^ So, too, the oxidation of uric acid itself with nitric acid finally results in the production of oxalic acid. And oxalic acid occurs constantly in small quantity in the urine, occasionally in larger amount as a result of disordered metabolism, in diabetes, or in chronic gastric disturbances.* Now this oxalic acid is very often in part derived from ingested oxalic acid, for it has been proved that this substance, not only when absorbed from the stomach, but also when small quantities are injected sub- cutaneously, is excreted, in the latter case quantitatively, in the urine.^ But in addition to the varying amounts that are introduced into the system with vegetable foods, this acid is 1 Salkowski, B, 7, 719, 1876 ; and 11, 501, 1878 ; cf., too, H.-S, Z. 35, 493, 1903. 2 Poduschka, S.A. 44, 64, 1900; cf., too, Luzzatto, H.-S. Z. 38, 537, 1903. 3 Ad. Glaus, B. 7, 226, 1874. * Lewin, H. B. i, 490 ; Baldwin, M.J. 715, 1900. ^ Pierallini, M. J. 714, 1900 ; Pohl, S. A. 37, 413, 1896 ; Faust, .S", A. 44, 235, 1900. N 194 THE METABOLISM OF CYCLIC FORMATIONS [lect. a product of the metabolism of the body, and the fact that whereas casein does not increase the output, meat arid still more thymus does, led Salkowski to suspect that it was derived from hucleine.^ This may mean that it is formed in the destruction of uric acid, and be one of the substances that we are searching for. It has been, as a matter of fact, found that allantoine in rabbits gives rise to a very definite degree of oxaluria ; but, on the other hand, neither in dogs nor in rabbits is there any more oxalic acid excreted as a result of adding uric acid to the food.^ It is not possible, therefore, to say for certain what becomes of the uric acid that is disposed of in the body before it can reach the urine. Glycocoll, allantoine, and oxalic acid may each of them, in certain cases at any rate, account for some of it. But there is no doubt that uric acid is not necessarily a final product that undergoes no further change. The purine ring is not indestructible, and this fact has to be taken into account in pathology as well as in physiology. And if there is doubt about the last stages in the metabolic processes leading to the destruction of purine, the earlier stages are intelligible, and. probably throw light on the fate of ring structures in general in the body. The forms of purine which actually occur in the nucleic acids are the amino derivatives. By the gradual in- sinuation of oxygen atoms wherever possible, leading to the formation of uric acid, the ring is so weakened that unless excreted at this stage it breaks up. Till we know more about the origin of kreatine in the muscles, it may be fanciful to draw a parallel in the case of this substance. But it may be that in kreatinine we have the last stage before the disappearance of an imidazol ring. In this connection we are led finally to what is known of the fate of the aromatic ring in tyrosine and phenyl-alanine. Our conceptions on this point have been made clearer by the study of the disorder known as alcaptonuria. ^ Salkowski, B. k. W. 20, 1900 ; Liithje, M. J. 584, 1898; Lommel, i^/./. 336, 1899. ''■ Luzzatto, H.-S. Z, 37, 228 ; and, 38, 542. vFii.] ALCAPTONURIA . 195 In this interesting anonmaly of metabolism the urine contains considerable amounts of homogentisic acid ; that is to say, 1.4 dioxybenzene acetic acid,i HO OH CHg . COOH Baumann, who first discovered the presence of this abnormal substance in the urine, showed that the administration of tyrosine to persons who had alcaptonuria caused an increase in the amount of homogentisic acid in the urine. But he thought that the homogentisic acid was formed by some peculiar variety of bacteria in the intestine. It has since then been shown that the homogentisic acid bears a constant proportion to the nitrogen excreted, whatever the amount of proteid taken may be, and even when proteid is almost excluded from the diet. Since the nitrogen excreted under such circumstances comes from the interrikl metabolism of the body, it is difficult to ascribe a different source to the homogentisic acid, and Baumann's idea that it is not a product of metabolism at all has been abandoned. 2 From the study of the ratio of the homogentisic acid to nitrogen excreted another point was made out, that this ratio was higher than could be accounted for, even on the supposition that all the tyrosine set free in proteid metabolism underwent this change.^ It was then found that phenyl-alanine behaved like tyrosine in persons with this affection, and was almost entirely converted into homogentisic acid.* This makes ' Wolkow and Baumann, U.S. Z. 15, 228, 1891. ^ Falta, B. CM. 3, 173, 1904 ; cf. Embden, U.S. Z. 18, 304. ' Langstein andE. Meyer, D. A.f. k. M. 78 ; and, D. R. A., p. 383, 1903. * Falta and Langstein, H.-S. Z. 37, 513, 1903. — It is well known that many fungi contain a " tyrosinase " which oxidises tyrosine to a dark brown or black substance. It is said that homogentisic acid is formed. Raper and I tried to see if these fungi oxidised phenyl-alanine similarly ; we collected four varieties of Russula that contain tyrosinase, and found that while the juice oxidised and blackened tyrosine, it had no effect apparently upon phenyl-alanine, 196 THE METABOLISM OF CYCLIC FORMATIONS [lect. it somewhat easier to understand the conversion of tyrosine in which the hydroxy! group is in the para position to the side chain, into a substance in which two hydroxyl groups are in the 1-4 positions, the one ortho and the other meta to the side chain. If the tyrosine be reduced to phenyl-alanine it would still be converted into the same product in these persons. Further, it has been shown that phenyl-lactic acid and phenyl- pyruvic acid also give rise to homogentisic acid, while phenyl- acetic acid does not,^ so that it has been suggested that the sequence of events in the metabolism of tyrosine under these conditions is to be represented by the following formulae : ^ — OH /\ CH,.C -> H COOH NH„ Tyrosine. CH .C< \0H -^ COOH CH2 . C< . COOH \nh, Phenyl-alanine. /\ _^H0 /\ CH 2-C O.COOH CH OH 2 . COOH Phenyl-lactic acid. Phenyl-pyruvic aciJ. Homogentisic acid. Now the physiological importance of all this is derived from 1 Falta, M. M. W., p. 1846, 1903 ; Embden, U.S. Z. 18, 316. " The occasional occurrence of uroleucic acid. HO /\ / OH CH,.C. •H OH , COOH in alcaptonuric urine would imply that the oxidation of the ring would occur at an earlier stage than indicated in this scheme. VIII.] TYROSINE IN METABOLISM 197 the fact that in normal men the aromatic ring in gentisic acid H0/\ !joH COOH taken by the mouth disappears almost entirely, just as in tyrosine and phenyl-alanine it does ; but in alcaptonuric patients it is, like the homologous acid, excreted quantitatively unchanged.^ It is argued, therefore, that normally these phenyl derivatives, tyrosine and phenyl-alanine, are oxidised to homogentisic acid just as they are in alcaptonuria, but at this point the difference between the normal and the abnormal metabolism shows itself; normally the introduction of the two hydroxyl groups into the benzene ring is the signal for the destruction of the ring. But in the alcaptonuric individual the ring for some reason will not split : this is the reaction that fails.^ It will be remembered that tyrosine in the intestine is to a varying extent converted by bacteria in oxyphenyl-propionic acid and the corresponding acetic acid compound, both of which may appear in the urine or be converted into cresol or phenol itself And these phenols are excreted as esters of sulphuric or glycuronic acid, the ring having escaped without any fundamental change. Now it has been found that in alcaptonuria phenyl-propionic and phenyl-acetic acid are not oxidised to homogentisic acid.^ It seems, therefore, that the bacteria denitrify tyrosine to oxyphenyl-lactic acid, but then reduce this to phenyl-propionic acid, and thereby at once render the ordinary metabolic changes by which tyrosine, including the aromatic ring, is completely oxidised in the body no longer ' Neubauer and Falta, H. S. 42, 84, 1904. ^ It is interesting to note that a disturbance apparently identical with alcaptonuria can be induced in plants, metabolism being arrested so that homogentisic acid accumulates in the roots. Cf. Czapek, Ber. der deutschen Botan. Ges., vols. 20 and 21, 1902 and 1903. 3 Embden, H. 5. 18, 316, 1893. 198 THE METABOLISM OF CYCLIC FORMATIONS [lect.viii. possible. The reduction of the lactic acid group to propionic acid does not prevent the oxidation of the side chain, though this probably is done by the bacteria too, and not in the body ; but it does prevent the ring from being broken up. It appears that the normal course of metabolism is for the lactic acid chain to be oxidised to an acetic acid chain by way of pyruvic acid, /" — CH, . C< . COOH -> — CH„ . CO . COOH -^ —CH. . GOGH \0H and that if this is the course of the reaction the simultaneous oxidation of the ring to hydroquinone is affected ; while if the lactic acid is first reduced to propionic acid, this oxidation in the ring, which in normal conditions but not in alcaptonuria leads to its destruction, is no longer possible. This is in itself curious and interesting. But it is still more so if, as appears probable, it applies exactly to tryptophane too. We know, at any rate, that bacteria in denitrifying tryptophane produce indol-propionic acid, not indol-lactic acid ; that the indol- propionic acid is further converted into indol-acetic acid, scatol, and indol itself, but that such indol groups appear in the form of indoxyl and scatoxyl esters in the urine with the ring intact. We know at the same time that tryptophane, when not appropriated by bacteria, in the normal course of metabolism disappears entirely. But we do not know exactly how the reactions leading to the destruction of the indol ring differ from those carried out by bacteria in the intestine. But the indol ring must be attacked before it is stripped bare of its side chain, just as the phenyl group must in tyrosine, for the naked indol ring introduced into the body is not attacked. INDEX Absorption of proteids, rate of, 137 Acetic acid in liver, 58 metabolism, 103 Acetic aldehyde in synthesis of fat, 81 from lactic acid, 53, 81 Acetone in normal expired air, no Acetonuria, 108 Adamkiewicz, reaction of, 5 Adenase, 185 Adrenaline, 168 Adsorption and catalysis, 18 Alanine, 7 as source of sugar, 44 compounds of, 45, 179 Albumins, glucosamine in, 32, 40 Albumoses in blood, 133, 140 Alcaptonuria, 195 seq. Alcohol from sugar in muscles, 66 Alcoholic fermentation, 14, 53 Aldehydes, formation of, from oxy acids, S3 Aldol condensation, 21, 55, 82 Alkalies, action of, on sugar, 25, 51 Alkaloids related to tryptophane, 176 AUantoine from uric acid, 192 from hypoxan thine, 193 Amide nitrogen in proteids, 152 Amido acids as source of sugar, 43 seq. formed in digestion, 127 seq. heat equivalents of, 128, 154 and nitrogenous equilibrium, 130 199 Amido acids in blood, 137 seg. in tissues, 149 in metabolism, 148, 151 seq. Ammonia in acetonuria, 109 in blood during digestion, 140, 152 Ammonium lactate to uric acid in per- fusion of liver, 187 Anti-catalysis in disease, 118 in health, 148 and antiseptics, 151 Arabinose, 70 Arginine, 11, 165 Arginine as source of proline, 169 Arsenious acid and lactic acid, 59 and fatty liver, 118 Aspartic acid as source of sugar, 45 in peptic digestion, 125 Azelaic acid, 104 Bayer, synthesis of carbohydrates in plants, 9, 20 Bial's reaction, 33 Biuret reaction, 3 Blood, proteids of, 141, 142 rate of flow of, in intestine, 1 37 Blown oils, 106 Buchner, zymase, 12, 52 Butter in acetonuria, 161 Butyric fermentation, 53 seg. Butyric acid in liver, 58 in metabolism, 103 200 INDEX Capronic acid from lactic acid, 80 in butyric fermentation, 81 in metabolism, 103 Carbohydrates, synthesis in plants, 9,20 amount of, in animal body, 36 digestion of, isothermic, 129 effect of, in acetonuria, 108 Cartilage, 34, 43 Casein in glycosuria, 42 amide nitrogen in, 153 Catalysis, Ludwig, 13 Ostwald's definition, 16 negative in disease, 118 Cerebrines, 30 Cetti, metabolism in starvation, 97 Choline, 168 Chondrosine, 34 Clupeine, action of arginase on, 166 Co-ferments, 53 Cohnheim on glycolysis in muscles, 66 Cysteine, conversion into tausine, 1 1 Cystine in peptic digestion, 125 from taurine, 11, 148 Cytosine, 180 D N . 39> 113. 115 Denitrification of amido acids, 151 seg. of purine bases, 185 Diabetes, acetonuria in, 108 sugar from fat in, 113 Lupine's theories, 65 Diabetic puncture and indicanuria, 177 Dialdane, 82 Diamido acids, fate of, in the body, IS3 Diamines in peptic digestion, 125 Digestion of proteids, 124 seg. Digestive changes, isothermic, 129 Diphtheria toxin and fatty heart, 118, 120 Drechsel, lysine, 10 Drying oils, 106 Eggs, metabolism during incubation, lOI synthesis of purine in, 182 Enzymes and vital action, 12 intracellular, 13, 57, 145 physiological classification, 13 Erepsine, 126 seg. Erucic acid in acetonuria, 1 1 1 Faraday, catalysis, 19 Fat, amount of, in animal body, 35 in different kinds of muscles, 99 use of, in muscular activity, 100 extraction of, 120 conversion into sugar, 112 seg., 119 derived from proteids, 86, 164 from carbohydrates, 77, 85 composition of, in milk, 78 source of energy in starvation, 98 effect on acetonuria, 109 iodine value of, in heart, 107 compounds of, with sugar, 31 Fatty acids, in metabolism, 103, 105 stability of, 106 oxidation of, its importance, 164 Fatty degeneration, 87 seg., 117 seg. Fermentation, 14, 52 seg. Fever, acetonuria in, 108 proteid metabolism in, 143 kreatinine excretion in, 167 Folin, metabolism on diet free from proteids, 162 Formic acid in metabolism, 103, 105 produced by yeast, 53 from lactic acid, 81 Formic aldehyde and sugar synthesis, in plants, 10, 20, 48 in animals, 47, 164 from sugar, iSo- Fructose converted into glucose, 25 INDEX 201 Galactose, origin of, 25 in fermentation, 55 in cerebrines, 31 relation to arabinose, 71 Globulin, glucosamine in, 32, 40 Glucosamine, constitution, 32 estimation of, in proteids, 32, 40 as source of glucose and glycogen, 42 Glucose, conversion into Isevulose and mannose, 25. conversion into imidazol, 179 Glucoside nature of proteids, 34 Glutamic acid in peptic digestion, 125 relation to proline, 170 Glyceric acid from diamido propionic, 153 Glyceric aldehyde as precursor of lactic acid, 56 Glycerine as source of sugar, 46 origin of, 95 Glycoalbumoses, 34 GlycocoU as source of sugar, 47 in urine after phosphorus, 139 from uric acid, 191 Glycogen, amount of, in animal body, 36, 114 in starvation, 38 derived from glucose, 22 from other sugars, 27 from proteids, 37 from glucosamine, 42 from alanine, leucine, 44 in rigor mortis, 59 in muscular activity, 5i, 62 Glycolic aldehyde, 47 Glycolysis in blood, 65 Glyco-proteids, 32, 40 Glycosuria, amount of sugar excreted in, 36 phlorrhizine, 36, 39 pancreatic, 37-39> 45j 46 V. Diabetes. Glycuronic acid converted into xylose, 29 derived from glucose, 30 from proteids, 69 in urine, 30, 67 Glyoxyiic acid reaction, 5 Guanase, 185 Guahidine, 165 derived from proteids, 167 H^MATINE, 123, 170 Hsematinic acids, 171 Haemoglobin, 122 Heat, effect of, on reaction velocity, 16 Heart, fat in, 99, 120 iodine, value of, 107 Heteroxanthine from xanthine, 168 Hexone bases as source of sugar, 44 Hippuric acid, 10 from phenyl propionic acid, etc., 105 from uric acid, 191 Histidine, constitution, 179 Homogentisic acid, 195 seq. Hoppe-Seyler, fatty acids from lactic acid, 79 Hybernation, respiratory quotient in, 112 Hydrazine poisoning, allantoine in, 192 Hydrogen set free in autolysis of liver, 59 Hydrolytic changes in metabolism, 150 Hypoxanthine, synthesis in muscles, 189 Imidazol, 178, 179, 182 Indicanuria, 176 seg. Indol, 172 seq. Intestine^ perfusion of, 134 synthesis of proteid from peptone in, 135, 141 202 INDEX Intestine, fatty acids dissolved in, 75 Intracellular enzymes, 13, 57 Iodine value of fat, 73, 89 from different sources, 107 I odo fatty acids, 73 Isobutyric acid from lactic acid, 80 Iso-cystine, 174 Isomeric transformations of sugar, 25 Iso-serine, 175 Jecorine, 31 Kidney, fat in, 99, 120 of ox, uric acid destruction in, 191 Kreatine not attacked by arginase, 166 origin of, 194 Kreatinine excretion on different diets, 162 endogenons, 166 and imidazol, 179 Kiihne on proteid digestion, 124 Kynurenic acid, 173 Lactacidase, 52 Lactic acid from alanine, 44, 61, 153 from sugars in fermentation, Ji seq. in the blood, 65 from proteids, 60 as source of sugar, 45 economises proteid, 103 in muscle, 59, 61, 63 in liver, 58 in urine of birds after removal of liver, 60 in fatigue, 57 in glycolysis in blood, 65 in synthesis of uric acid, 187 heat equivalent of, 63, 154 Lasvulose converted into glucose, 25 Lang, denitrification, 151 seq. Laurie acid in metabolism, 103 LebedefTs experiment, 88 Lecithine in jecorine, 31 Le Sueur, decomposition of oxy-acids^ 104 Leucine as source of sugar, 43 in peptic digestion, 125 heat equivalent of, 128 Leucolytosis, post-prandial, 134 and uric acid, 184 Linoleic acid, 73, 78, 106 Lipolysisj 74 co-ferment in, 53 isothermic, 76, 129 Liver, acute yellow, atrophy of, 139 autolysis of, 84 fat, synthesis in, 85 fattydiseasein, 118 change in phlorrhizine glycosuria, 119 in pancreatic glycosuria, 115, 119 fat from, iodine value, 107 uric acid, synthesis in, 188 destruction in, 191 allantoine formed in, 192 removal of, in birds, 183 oxidation of purine bases in, 186 Lobry de Bruyn, 25 Ludwig on catalysis, 13 Lungs, oxidation of purine bases in, 186 Liithje, pancreatic glycosuria, 37, 46 Malonic acid in synthesis of uric acid, 188 Mannose, conversion into glucose, 25 Mannonic acid, action of pyrridine on, 27 Margaric acid, 78 Meat, purine bases in, 184 Mesoporphyrine, 170 Mesoxalic acid, 188 INDEX 203 Methylamine derivatives, i68 Methyl mercaptan, i68 Milk, fats in, 78 synthesis of purine bases from, 182 Minkowski, — , 39 N Molisch's reaction, 33 Moore's test, 5 1 Mucin, 32 Muscles, lactic acid in, 59, 61 seg. fat in different varieties of, 99 glycolysis in, 66 Muscular activity, fat used in, 100 kreatinine excretion in, 167 hypoxanthine formed in, i8g Myristic acid in metabolism, 103 Naphthalene sulphonyl chloride, 139, 148 Nitrogen, minimum intake of, 160 cannot be stored, 157 excretion after proteid meal, 155 not a measure of proteid com- bustion, 155 on a diet free from proteids, 162 Nitrogenous equilibrium on diet free from proteids, 130 Norisosaccharic acid, 33 Nucleic acid, pentose in, 28 glucoside nature, 30 pyrimidine derivatives in, 180 in autolysis, 185 Oils, " drying," "blown," and rancidity of, io5 Oleic acid, constitution, 104 in acetonuria, 1 1 1 Ostwald on catalysis, 16 • Oxalic acid, 68 origin of, 193 Oxamic acid, fate of, in body, 153 Oxidation of fatty acids, 102, 105, no Oxy-acids, decomposition of, 104 Oxy-aspartic acid, 175 Oxygen consumption in starvation, 98 on various diets, 100 in chick, 10 1 Pancreas, autolysis of nucleic acids in, 30 removal of, effect on erepsine, 126 extract of, causing glycolysis in muscles, 66 V. Glycosuria Pelargonic acid, 104 Pentoses in nucleic acids, 27 lactic acid from, 41 Pentosuria, 69 Peptic digestion, amido acids formed in, 125 action on polypeptides, 132 PflUger, pancreatic glycosuria, 37 on origin of sugar from fat and proteid, 1 14 seg. criticism of Voit on fat formation, 87 Phenyl acetic acid series, fate of, in body, 105 Phenyl alanine, 7 not formed in tryptic digestion, 132 effect on urea output, 161 liberated by pepsine, 169 effect on alcaptonuria, 195 Phlorrhizine, acetonuria in dogs, in glycosuria, fat in, 113 indicanuria, 177 Phosphorus poisoning and fatty heart, 89, 118 and amido acids in urine, 139 autolysis of liver in, 146 Phosphotungstic acid, 6 Phylloporphyrine, 170 Pneumonic exudation, absorption of, 146 204 INDEX Polypeptides, 5 unattacked by trypsine, 131, 169 Proline, 169 not formed in tryptic digestion, 132 Protamines, 123, 166 Proteid, colour tests, 3 amount of, in animal body, 35 as source of sugar, 37 seq., 115 seq. fat, 86, 164 lactic acid, 60 acetone bodies, 106 synthesis of, in animals, 123, 133 seq. digestion of, 124 seq. specificity of proteids in blood, 123 heat equivalents of cleavage pro- ducts, 129 rate of catabolism in tissues, 149 metabolism on proteid diet, 156 " circulating," and " tissue " proteid controversy, 158 minimum intake, 160 Purine, 178 seq. as source of pyrimidine bases, i8i synthesis, 182, 187 Pyridine derivatives related to .trypto- phane, 176 Pyrimidine, 178, 180 Pyrrhol and derivatives, 168 Pyrrholidine carboxylic acid. See Proline Pyruvic aldehyde, precursor of lactic acid, 56 in formation of imidazol, 179 QuiNOLINE derivatives from trypto- phane, 176 Rancidity, 106 Respiratory quotient in hybernation, 112 Reversible zymolysis, 23, 74 Ricinoleic acid, 78 Rigor mortis, lactic acid in, 59 Saccharinic acid from glucose, 43 Salmon, metabolism of, in fresh water, 143 Salvioli's perfusion experiments, 134 Sandmeyer's operation, 1 14 Saponification value of fats, 106 Scatol, scatol carboxylic, and scatol acetic acids, 173 Seeds, germinating, oil in, 106, 112 proteolysis in, 142 Serine, 7 Soaps, toxicity of, 75 Spleen, oxidation of oxypurine bases in, 186 Starvation, metabolism in, 97 acetonuria in, 108 causes no acetonuria in dogs, 1 1 1 autolysis in, 142, 156 Stomach synthesis of proteid from peptone in, 135 Succinic acid from hasmatine, 172 Succus entericus, erepsine in, 126 Sugar decomposition in muscles, 66 effect of alkalies on, 25, 51 fermentation of, 52 seq. converted into fat, 112 seq. derived from fat, 112 proteid, 37 seq., 115 seq. amount of, in blood, 36 output in diabetes, 36 Sulphur, excretion of, on different diets, 162 Sweetbreads, effect on endogenous uric acid, 189 Tartronic acid in synthesis of uric acid, 188 Taurine, 11, 148 Tetra-oxyamido-caproic acid, 34, 43 Thiolactic acid, 175 Thymine, 180 INDEX 205 Tissues, nitrogen metabolism of, i66 Tryptic digestion, 125 peptides resisting, 131 seq. Tryptophane, 6, 173, 177, 198 Tyrosine, heat equivalent of, 128 in peptic digestion, 125 effect on urea output, 161 and homogentisic acid, 195 Uracil, 180 Uranyl acetate, 191 Urea synthesis, 10 in acetonuria, 109 in blood during digestion, 140 excretion not a measure of proteid combustion, 155 after proteid meal, 1 56-7 on diet free from proteid, 162 in uric acid synthesis, 187 Uric acid, 182 seq. synthesis, 162, 187 seq^ exogenous, 183 endogenous, 184 destruction in body, 190 Valerianic acid in metabolism, 103 Virchow on fatty degeneration, 87 Voit, formation of fat from proteid, 86 Work and oxygen consumption, 100 See Muscular activity. Xanthine, methylation of, in dogs, 168 relation to methyl imidazol, 180, 182 Xanthine oxidase, 186 Xylose constitution, 29 from glycuronic acid, 29 in nucleic acids, 69 Yeast, fermentation by enzymes in, 12,52 produces formic acid, 53 Zymase, 12, 52 Zuntz, energy expenditure on respira- tion and circulation, 98 use of fat in muscles, 100 Printed by- Oliver and Boyd Edinburgh