LIBERATION OF TYROSINE DTJRING TRYPTIC PROTEOLYSI8. 145 XVIII.-The Liberation of Ty-osine during Tryptic Pro teol ysis. By ADRIAN JOHN BROWN and EDMUND THEODORE MILLAR. WHEN studying the various methods suggested for the purpose of measuring the activity of proteolytic enzymes, it occurred to us t h a t a method of directly estimating tyrosine by bromination recently described by James H. Millar (Trccns. Guinmss Reseamh Laboratory, 1903, 1, Part I) might fiimish a means of determining the course of proteolytic change in those cases i n which tyrosine is liberated during the breaking down of the protein molecule. An investigation in this direction was therefore commenced and the r e s u h so far obtained are described i n this paper. The paper may be summarised as follows : (1) J.H. Millar’s method of estimating tyrosine by bromination is applicable to the estimation of tyrosine in tho presence of proteins and their earlier cleavage products due to enzyme action, if suitable control{experiments are employed. VOL. LXXXIX. L146 BROWN AND MILLAR: THE LIBERATION OF (2) Tyrosine is nat a late product of tryptia proteolysis, as is usually supposed ; on the contrary, the tyrosine nuoleus of a protein is attacked and the whole of the tyrosine liberated during the first stage of tryptio digestion, (3) The resistance of the protein tyrosine nuoleue to peptic hydrolysis is confirmed. (4) Attention is called to the similarity of Emil Fiscber and E. Abderhalden’s recent observations on the actions of tryptic and peptic enzymes on polypeptides containing a tyrosine nucleus (Zeit.physiol, Chsrn., 1905,46, 52) to the authors’ observations on the aotions of the same enzymes on proteins containing a tyrosine nucleus. reliable means of differentiating enzymes of a peptic from those of a tryptic nature, and may assist in throwing some light on the confused state of knowledge with regard to the existence of a tyrosine nucleus in the dieerent albs- moses resulting from peptic and tryptic proteolysis. ( 5 ) The authors’ investigations appear to indicate EX P ERIM E N TA L. J. H. Millar’s method of directly estimating tyrosine (Zoc. cit.) i s based on its reaction with free bromine, by which a bromine compound of tyrosine is formed. Tyrosine is dissolved in hydrochloric acid to which potassium bromide is added.The solution is then titrated with a N/5 sodium bromate solution. The liberated bromine resulting from the interachion of t h e sodinm bromate a.nd bromide in acid solution is rapidly absorbed by the tyrosine present, and the end of the reaction determined by employing starch and potassium iodide as an indicator for free bromine. J. H. Millar’s experiments with pure tyrosine show that the reaction results in the formation of dibromotyrosine aacording to the following equation : C6H,(HO)*CH2*CH(NH,)*C02H + 4Br = C,H2Br2(HO)*CH2*CH(NH2)*C02H + 2HBr. The method of estimating tyrosine is shown by J. H. Millar to be applicable not only to the accurate estimation of the pure substance, but also to the substance when it exists in intermixture with ammonium salts and amides and amino-acids such as asparagine, aspartic acid, leucine, and phenylalanine, which result from complete acid proteolysis.J. EL Millar’s work does not, however, show whether his method is applicable to the determination of tyrosine in the presence of proteins or their primary cleavage products, such as albumoses or peptones ; itTYROSINE DURING TRYPTIC PROTEOLYSIS. 147 was necessary, therefore, for us to investigate this point as preliminary to an attempt to employ the method for the estimation of tyrosine when present among the products of enzyme proteolysis. Preliminary experiments with solutions of egg-albumin, edestin, and gelatin indi- cated that they possessed to some extent the property of absorbing bromine under the conditions employed by J.H. Millar to estimate tyrosine. Following on this observation, a solution of edestin was prepared and divided into two equal vclumes. One part was titrated direct with N/5 bromate solution and its power of absorbing bromine noted. A known amount of tyrosine was dissolved in the second volume and it was also titrated with broniate solution. It was then found, after correcting the result of the second titration for tho amount of bromine absorbed by the edestin alone, indicated by the first titration, that an accurate measure was obtained of the amount of tyrosine introduced. Similar results were also obtained when gelatin and egg- albumin mere employed in tbe place of edestin. Our preliminary experiments therefore showed that it was possible to determine tyrosine in the presence of proteins if control experiments were made in order to correct for the bromine absorbed by the proteins.A series of experiments were then made in which edestin* was digested with pancreatic extract and the products of change examined by the bromine method as follows : A 1 per cent. solution of edestin was prepared by dissolving 2 grams of the dry substance in 200 C.C. of a 0.5 per cent. sodium carbon- ate solution, 50 c.c of this solution being placed in each of four flasks. To each of three of these flasks, 5 C.C. of active pancreatic extract (Benger) were added, and to the fourth flask, employed as a control, 5 C.C. of pancreatic extracb were added which had been previously heated to looo to render it inactive. All the flasks were placed in a water-bath kept at 32'.After 24 hours, the control and the contents of one of the flasks containing active pancreatic extract were titrated with bromate solution after the addition of 20 C.C. of SO per cent. hydrochloric acid and 10 C.C. of a 20 per cent. solu- tion of sodium bromide; and after 72 and 144 hours respectively the contents of the second and third flasks containing active pancreatic extract were titrated in a similar manner. The results obtained are given in the following table : * Edestin was employed in this and many of the followiiig experiments, contains a tyrosine nucleus and can also be readily prepared in a comparatively state, as it pure148 BROWN AND MILLAR: THE LIBERATION OF N/5 - Bromate solution after Time of NiB-Bromate deducting digestion.solution used. control. Active digestion ... 24 hours 0.80 C.C. 0.38 C.C. Control ............ 24 ,, 0'42 ? 7 7 7 72 7 , 0.80 ), 0.38 ,, > I 0.80 ,, 0-38 ,, ,, 144 9 , - Calculated per cent. of tyrosine formed from edestin during proteolysis. 4'06 4-06 4 '06 - It appeared from the above experiments, if the method of estimating tyrosine adopted was reliable, that 4.06 per cent. of tyrosine resulted from the tryptic digestion of edestin during the periods of 24, 72, and 144 hours, Oh the supposition that proteolysis had pro- ceeded far enough during 24 hours-the shortest period employed-to liberate the whole of the tyrosine from its containing nucleus in the edestin molecule, the results appeared quite reasonable, but on other grounds they were open to question.I n the first place, it was questionable whether the small volumes of bromate solution consumed in the above experiments measured the tyrosine present with any approach to accuracy. It was found, however, on experimenting with known amounts of tyrosine, com- parable with those measured in the preceding experiments, that very accurate results were obtained considering the small volumes of bromtlte solution employed. A second more difEcult objection to meet questioned the accuracy of the correction obtained from the control experiment. The control indicated the amount of bromine absorbed by the pancreatic extract and the undigested protein, and was subtracted as a correction from the total amount of bromine absorbed by an intermixture of digested products and pancreatic extract in order to arrive at the amount of bromine absorbed by the tyrosine liberated.It was open to doubt whether the correction remained constant under these conditions. When solutions of edestin and pancreatic extract of similar concentration to those employed in the foregoing experiments were titrated separately, it was found that each absorbed bromine t o some extent.* For instance, when 0.5 gram of edestin and 2.5 C.C. of pancreatic extract were titrated separately with bromate solution, bromine equal to 0.3 C.C. of bromate solution was absorbed by the edestin and to 0.1 C.C. by the pancreatic extract. It was possible to digest pancreatic extract alone as a control in order to ascertain whether its original power of absorbing bromine underwent any change during digestion, and it was found on doing so that no alteration took place.The accuracy of the control experiment PO far as it concerned the pancreatic extract was therefore established. But with * Tyrosine has been fount1 in all pancreatic extracts examined.TYKOSINE DURING TRYPTIC PROTEOLYSIS. 149 regard to the part of the correction applying to the bromine absorbed by edestin previous to digestion, it still remained open to doubt whether it could be employed with accuracy after digestion had taker1 place and the edestin molecule had been broken down to a greater or less extent. There appeared to be no way of obtaining an answer to this question by means of experiments with the tryptic digestion products of edestin, but experiments with the tryptic digestion products of gelatin pointed to the conclusion that no change takes place.Gelatin, unlike a typical albumin or globulin, does not contain a tyrosine nucleus ; it appeared, therefore, that if an examination of its digestion products by the bromine method were made, the complicating presence of tyrosine would be avoided and some light might be thrown on the constancy of the control referred to above. A series of digestion experiments with a 1 per cent. solution of gelatin and pancreatic extract were made under similar conditions to those with edestin (p. 148) with the following result : 50 C.C. of gelatin solution and 2.5 C.C. pancreatic Time of ex tract. digestion. 1. Active digestion ... 48 hours 2- , Y 7 , 100 ,, 3. 7 9 > > 148 Y , 4.Control ... ... ... ... - Calculated HIS-Bromate percentage of solution used tyrosine formed N15-Bromate after deduction during solution used. of control. proteolysis. 0.35 C . C . none. none. 0-35 ,, 9 , ,$ 0.35 ), ,? , 9 0.35 ,, - - It will be seen from the foregoing results that the amounts of bromine absorbed by the different digestions of the gelatin do not vary from the amount originally absorbed by the gelatin prior to diges- tion, which shows that for gelatin at least the cleavage products of its molecule absorb the same amount of bromine as the original molecule prior to hydrolysis. It seemed probable, therefore, that the same conditions might obtain with edestin. More convincing evidence of this was, however, obtained by an examination of the products of a peptic digestion of edestin itself.There was good reason to anticipate from the results obtained by previous investigators that it would be found that the tyrosine nucleus of a protein such as edestin was not attacked during peptic digestion, and consequently that free tyrosine would not be present among the products of peptic proteolysis. On this assumption, the following experiment on the digestion of edestin by peptase was made: 3 grams of edestin were dissolved in 300 C.C. of a solution containing 0.27 per cent. hydrochloric acid, and 50 C.C. of liquor pepticus (Benger) were then added. Immediately after intermixture, 50 C.C. of the solution were vithdrawn and titrated with N/5 bromate solution in150 BROWN AND MlLLAR: THE LIBERATION OF order t o ascertain the amount of bromine absorbed by the original mixture of edestin and liquor pepticus prior to digestion.The rest of the solution was kept in a water-bath a t 32”, and during digestion portions were withdrawn and titrnted with bromate solution a t succes- sive intervals of time. The results obtained are given below : N/5-Bromate solution employed to titrate 50 C.C. of edestin solution. Control, prior t o conmencement of digestion ... 0.50 C.C. After 24 hours’ digestion .............................. 0.55 ,, ,) .............................. 0.50 ,, ) ) .............................. 0’60 ’, 8 ’ .............................. 0.55 ,, ) ) ............................. 0-55 ,, 9 , 48 9 , 9 , 72 , t 9 , 96 > Y >, 192 7 3 During the course of the prolonged peptic digestion of edestin in the above experiment, it will be noticed that-within errors of experi- ment-the original edestin and its digestion products absorbed equal amounts of bromine--a result markedly different from that which was obtained when digesting edestin with pancreatic extract.The experi- ment therefore confirmed the impression that tyrosine is not liberated during peptic digestion, and further strengthened the view that our method of employing a control for the bromine absorbed by proteins in digestion experiments was reliable. Before proceeding to make use of J. H. Millar’s bromine method for a further investigation of the conditions governing the liberation of tyrosine during tryptic proteolysis, it seemed desirable, however, to inquire as to the existence of another possible source of error.It i n known that tryptophane (scatoleaminoacetic acid) is very generally found among the products of proteolysis, and that it readily forms derivatives with free bromine. The presence of tryptophane was, moreover, recognised by us among the tryptic digestion products of edestin. It seemed possible, therefore, that the accuracy of the estimation of tyrosine by means of bromine might be influenced by the presence of tryptophane, although the results obtained in the experiments with edestin previously described appeared to render this unlikely. The investigations of 8. Vines indicate that tryptophane is liberated in gradually increasing quantities during tryptic proteolysis ; however, the amounts of bromine absorbed in our experiments with edestin (p.148) remained constant during digestion for very varying intervals of time, a result not likely t o be obtained if tryptophane takes part in absorbing the bromine. I n order, however, to settle this point definitely, tryptophane was prepared by Hopkins and Coles’ method (Sourn. Physiol., 1901, 27, 418), and subjected to the test of direct experiment. During protein digestion, tyrosine and tryptophane are said to beTYROSINE DURING TRYPTIC PROTEOLYSIS. 151 liberated in the approximate proportions of 4 : 1.5. In order, there- fore, to obtain conditions in some degree parallel with the digestion experiments with edestin (p l48), tyrosine and tryptophane in the proportion of 4 : l5-that is, 0.5 gram of tyrosine and 0.018 gram of tryptophane-were dissolved in 50 C.C.of water and titrated with bromate solution : bromine absorbed 2 0.9 C.C. N/5 bromate solution. A solution of 0.5 tyrosine alone in 50 C.C. of water was also titrated : bromine absorbed = 0.9 C.C. N/5 bromate solution. The experiments, therefore, indicated that no bropine was absorbed by the tryptophane employed. A second series of experiments in which more tryptophane was employed than in the foregoing series led to a similar conclusion. I n a further experiment in which 0.03'7 gram of tryptophane alone mas titrated with bromate solution, no bromine was absorbed.* As the evidence obtained from the experiments described above appeared t o show that the bromine method, when employed with a control, was capable of estimating tyrosine when present among the products of tryptic proteolysis, further experiments on the tryptic digestion of edestin were made.The first series of experiments with edestin (p. 148) indicated that the maximum amount of tyrosine was liberated within the first twenty- four hours of tryptic digestion, and this appeared to show that the tyroaine nucleus of the protein was attacked a t a much earlier stage of digestion than is usually supposed. Five hundred C.C. of a 1 per cent. solution of edestin, rendered alkaline by the addition of 2.5 grams of sodium carbonate, were digested with 25 C.C. of pancreatic extract (Benger) in the presence of a little toluene at 32'. A control experiment in which the pancreatic extract was rendered inactive by heat was also prepared, The following results were obtained : Percentage of tyrosine ht/5=Bromate calculated after solution used.deducting control. Active digestion after 30 minutes ... 0'8Q C.C. 2 *2 7 7 7 , )) 1 hour ...... 0'95 ) ) 8.8 7 7 Y 2 ,, If hours ... 0'90 ), 3.3 1 9 ,? > > 2 J 7 ..* Q'96 ,, 3.8 7 ) $ 3 7 9 20 7 1 * . * 0.95 ,, 3.8 - C011trol .................................... 0.80 ), The above experiments show that the tyrosine nucleus of the edestin was attacked during a very early stage of proteolysis. Although the digestion was carried on under conditions which did not favour very i+ The non-absorption of bromine by tryptophane under the conditions of J. H. Millar's method of estimating tyrosine is probably due to the hydrogen chloride which is present.152 BROWN AND MILLAR: THE LIBERATION OF rapid hydrolysis, within thirty minutes more than half the tyrosine was liberated, and within one hour the whole of it was set free.Other experiments with edeetio, which it is not considered necessary to describe in detail, also indicated that the whole of the contained tyrosine is liberated within a remarkably short period of time after digestion commences. Egg-albumin subjected to tryptic digestion under similar conditions to edestin also appeared to yield the whole of its tyrosine within three hours. From the above experiments with edestin and egg-albumin, it appeared, therefore, that the tyrosine nucleus of these proteins was one of the first constituent parts of their molecule to be attacked and hydrolysed during pancreatic digestion.As it appeared desirable to confirm the presence of tyrosine among the first products of tryptic digestion by some means other than the bromine one hitherto employed, 100 C.C. of a solution containing 1 gram of edestin and 0.5 gram of sodium carbonate were digested with 5 C.C. of pancreatic extract for one hour at 32' and a t once precipitated with trichloroacetic acid. The dense white precipitate of protein matter was filtered off and the filtrate evaporated to a small volume. After standing, crystals of tyrosine of characteristic appear- ance were obtained. Following on this, the crystals, together with the mother liquor, were titrated with sodium bromate solution, with the result that 4.3 per cent. of tyrosine, calculated on the original edestin employed, was found.I n a control experiment carried on under similar conditions to the above, but in which the pancreatic extract was rendered inactive by heat, no crystals of tyrosine were obtained. In a second experiment,25 grams of edestin dissolved in 2500 C.C. of 0.5 per cent. sodium carbonate solution were digested with 125 C.C. of pancreatic extract for forty-five minutes a t 32' and the digestion products precipitated with phosphotungstic acid in presence of dilute sulphuric acid. The precipitate was filtered off and the filtrate treated with barium hydroxide. The solution was again filtered to remove barium sulphate and the filtrate concentrated by evaporation to a small volume. On standing, tyrosine crystallised out freely. The tyrosine after separation was redissolved in dilute hydrochloric acid and again recrystallised from the solution after the addition of ammonium hydroxide.By this means, tyrosine was obtained in apparently a pure state. The above experiments therefore confirmed our original impression that the tyrosine nucleus of proteins is attacked and hydrolysed during a very early stage of tryptio digestion. An attempt was then made to gain some knowledge regarding the extent of protein degradation accompanying the breaking down of theTYROSJNE DURING TRYPTIC PROTEOLYSIS. 153 tyrosine nucleus and the liberation of tyrosine. A method employed by Weiss (Compt. rend. Tvccv. Laboratoh de Carhberg, 1903, 5, 11, 133) when investigating the proteolytic enzymes of germinating barley was used for this purpose.This investigator showed if a solution of a protein such as edestin is precipitated by tannic acid in the presence of sodium acetate, that almost the whole of the protein is thrown out of solution, and that the filtrate Contains a mere trace of nitrogen. If, on the other hand, the protein in solution is subjected to the action of a proteolytic enzyme previous to precipitation by tannic acid, varying amounts of the original protein nitrogen are found in solution, and the amounts found provide to some extent a measure of the proteolytic change which has taken place. As the primary cleavage products of proteolysis, such as albumoses and peptones, are precipitated by tannic acid, it appears from Weiss’s work that the soluble nitrogen found after proteolysis and precipitation with tannic acid is the nitrogen of arnino-acids and other substances of simpler constitution than albumoses or peptones.A solution containing approximately 2 per cent. of edestin and 0.5 per cent. of sodium carbonate was prepared : (1) A nitrogen determination by Kjeldahl’s method on 50 C.C. of the solution indicated that it contained 0.15 18 gram of nitrogen. (2) Fifty C.C. of the same solutim were digested with 5 c c . of pancreatic extract until a control digestion of the same volume of the extract showed by the bromine method of titration that the maximum quantity of tyrosine was liberated. The time taken for digestion t o this point was forty-five minutes. Digestion was then stopped by pre- cipibating the solution with tannic acid according to Weiss’s method.After filtration, i t was found that a volume of the filtrate equal to 50 C.C. of the original solution contained 0.0548 gram of nitrogen. (3) A control experiment, which is required when employing Weiss’s method, was made by digesting 50 C.C. of the original solution of edestin with 5 C.C. of pancreatic extract, the activity of which WAS previously destroyed by heat, and treating the solution in an exactly similar manner to Expt. 2. The nitrogen in 100 C.C. of the filtrate, equal to 50 C.C. of the original solution of edestin, was 0-0416 gram. This amount represents the nitrogen of the pancreatic extract employed which has not been precipitated by tannic acid, and also that of a very Rma31 amount oE edestin not precipitated by tannic acid.The total amount of nitrogen in the control experiment, 0.0416 gram, must therefore be subtracted from the total nitrogen found in Expt. 2 in order t o ascertain the amount of nitrogen which bas been rendered soluble during the digestion of the original edestin. The amount found was 0.01 32. The total amount of nitrogen in tbe edeotin present in 60 C.C. of the154 BROWN AND MILLAR: THE LIBERATIOX OF original solution was 0.1518 gram, therefore only 8.7 per cent. of this nitrogen was present in such form as to remain in solution unprecipi- tated by tannic acid after proteolytic digesticn had proceeded sufficiently to liberate the whole of the tyrosine. I n a second experi- ment with edestin, it was found that 9.1 per cent. of its contained nitrogen was rendered soluble.About one-third of the amount of soluble nitrogen found in the above experiments can be accounted for as being present in the tyrosine liberated ; the condition in which the remaining two-thirds exists is a t present unknown. The results of the above experiments should be regarded as only roughly indicating the maximum amoiin t of decomposition of edestin into substances not precipitated by tannic acid during the liberation of tyrosine ; it appears very probable that further investigation will show that the amount is less. However this may be, the experiments confirmed our previous conolusion that liberation of tyrosine takes place during the first stage of the tryptic hydrolysis of edestin. E. Fischer and E. Abderhalden have recently shown (Zoc.cit.), when synthetically prepared polypeptides containing a tyrosine nucleus are submitted to tryptia digestion, that they are hydrolysed and tyrosine is liberated. On the contrary, the same polypeptides are shown to resist the action of peptic digestion, and consequently no tyrosine is liberated. It appears interesting to compare these results with our observations on the actions of peptic and tryptic enzymes on proteins containing a tyrosine nucleus. Although the protein molecule is of far greater complexity than that of the polypeptides referred to, the behaviour of the two enzymes with regard to it has the appearance of being the same. This suggests that the tyrosine nucleus of both protein and polypeptide constitutes a point of attack for the tryptic enzyme.But, on the other hand, E. Fischer and E. Abderhalden (Zoc. cit.) show that some peptides which do not contain a tyrosine nucleus, such as alanyl- glycin, are hydrolysed by the tryptic enzyme, whilst others of some- what similar constitution, such as glycyl-alanin, are not decomposed. I n those cases in which hydrolysis takes place, i t cannot here be associated with a tyrosine nucleus, and it appears desirable t o bear this in mind when considering the mode of action of tryptase on the protein molecule. The liberation of tyrosine may be merely a secondary effect accompanying the cleavage of the molecule at some other point than the tyrosine nucleus. A reliable means of differentiating enzymes of a peptic from those of a tryptic nature is required, for attempted classification from the behaviour of these enzymes in acid or alkaline solution has proved iwu$cient, particularly with regard to vegetable proteolytic enzymes,TYROSINE DURING TRYPTIC PROTEOLYSIS. 155 It appears that the rapid arid complete liberation of tyrosine during tryptic digestion may furnish a satisfactory means of differentiating enzymes of a tryptic from those of a peptic nature. At present, the state of knowledge with regard to the existence of a tyrosine nucleus in the different albumoses resulting from tryptic and peptic proteolysis is in a somewhat confused state, and contradictory statements are frequently met with regarding t h i s question. From the results of our experiments, presumably no albumose resulting from tryptic digestion contains a tyrosine nucleus, since the whole of the tyrosine appears to be liberated in the free state in the earliest stage of digestion. On the other hand, one or more of the albumoses or other of the earlier cleavage products of peptic digestion should contain the whole of the protein tyrosine. A t present we have not experimented with a1 bumoses formed during tryptic digestion, but some preliminary experiments with a1 bumoses formed during peptic digestion indicate the presence of a, tyrosine nucleus in some and not in others. We have some reason t o believe that tyrosine is liberated from edestin during a very early stage of acid prot,eolysis as well as during tryptic pro t eol ysis. SCHOOL OF BREWING, UNIVERSITY OF BIRMINGEAM.