<<

2 9 4 F. M. VERONESE, E. BOCCU’. C. A. BENASSI, AND E. SCOFFONE

Preferential Hydrolysis of Kynurenine Peptides F. M, V e r o n e s e , E. Boccu’, C. A. B e n a s s i, and E. S c o f f o n e * Institute of Pharmaceutical Chemistry (Centro Naz. di Chimica del Farmaco e dei Prodotti Bio- logicamente Attivi del C.N.R.), and * Institute of Organic Chemistry (Centro Naz. di Chimica delle Macromolecole del C.N.R. — Sez. VIII), University of Padova (Italy) (Z. Naturforschg. 21 b, 294—300 [1969] ; eingegangen am 7. August 1968)

The preferential cleavage of kynurenine peptides bonds was studied through reduction to y-(o- aminophenyl)-homoserine derivatives followed by mild acidic as well as mild alkaline hydrolysis. The reduction of keto group of kynurenine to hydroxy one was achieved by controlled potential electrolysis at —1.05 Volts. In order to understand the mechanism of hydrolysis some model com­ pounds related to kynurenine peptides were also synthesized and studied. Hydrolysis of y-(o-amino- phenyl)-homoserine peptides in acidic media is related only to the assistance of y-hydroxy function, whereas in alkaline media also aromatic amino group is involved.

The conversion of to kynurenine was These studies, already accomplished with kynu­ extensively investigated in our laboratories through renine alone, have now been extended to some kynu­ performic acid oxydation 1 and, with better results, renine peptides in 5% and 50% acetic acid. in peptides and proteins by ozonization 2’ 3 and dye As reported in Table I the examined kynurenine sensitized photooxidation 4 w ith the aim to reach a derivatives present similar half-wave potentials of preferential chemical cleavage at the tryptophan polarographic reduction as that done by kynurenine residue. The preparation of kynurenine peptides itself. Some variations among the compounds could was firstly achieved from the corresponding trypto­ depend on the different molecular structure. phan peptides 5 and modified lysozyme w as obtained and studied 6. Compounds Volts vs. SCE In this paper we present results about the pre­ CH3COOH 5% CH3COOH 50% * ferential hydrolysis of kynurenine peptides through Kynurenine - 0.960 - 0.952 reduction by controlled potential electrolysis of the HC1 • H-Kyn-Glv-OEt - 0.992 — — keto group to the hydroxy followed by mild acidic HC1 • H-Ala-Kyn-OMe - 0.996 HC1 • H-Phe-Kyn-OMe - 0.990 — or basic hydrolysis. HC1 • H-Ala-Kyn- In addition some kynetic studies on the cleavage Glv-OEt - 0.992 — Z-Kyn-Gly-OEt — - 1.008 of selectively reduced kynurenine peptides and of Z-Leu-Kyn-OMe — — 1.006 related derivatives are reported *. Z-Ala-Kvn-Gly-OEt — - 1.000 Selective reduction of the keto group of kynure­ Z-Phe-Kyn-OMe — - 1.001 nine (1) to hydroxy group of y-(o-aminophenyl) - Table I. Polarographic Reduction: Half-Wave Potentials of homoserine (2) was previously studied polarogra- Kynurenine and Kynurenine Peptides in 5% and 50% Acetic phically by one of us 8 and was found to occur over Acid a. a For experimental conditins, see text. * 50% acetic acid was employed owing to the insolubility of protected kynu­ a large interval of pH with the consumption of two renine peptides in 5% acetic acid. electrons: 0II NH, I H \ / OH NH, I Controlled potential electrolysis may be better ac­ CH c - c h 2- c h complished at —1.05 Volts vs. SCE, the reduction COOH 2 e " C 00H of keto group being controlled by the lowering of 1 2 current flow and also checked by recording polaro-

1 C. A. Benassi, F . M. Veronese. A. De Antoni, and P. 8 G . G a l i a z z o , G . J o r i , and E. S c o f f o n e , Biochem. bio­ P a je t t a , Gazz. chim. ital. 97, 3 [1967]. physic. Res. Commun. 31. 158 [1968]. 2 E. S c o f f o n e , A. P r e v i e r o , C. A. B e n a s s i , and P . P a j e t t a , * A preliminary account of this wTork was presented at the Peptides, Proc. VIth European Sympos. Athens, 1963 VIII Peptides Symposium, Noordwijk, September 1966 7. (Pergamon Press, Oxford 1966), p. 183 — 188. 7 E. S c o f f o n e , A. F o n t a n a . F . M a r c h i o r i , and C. A. 3 A. P r e v i e r o and E. B o r d i g n o n , Gazz. chim. ital. 94, 630 B e n a s s i , Peptides. Proc. V I I I th European Sympos. Noord­ [1964], wijk (1967) [North Holland Publishing Company, Amster­ 4 C. A. B e n a s s i , E. S c o f f o n e , G . G a l i a z z o , and G . J o r i , dam] , p. 189. Photochem. Photobiol. 6, 857 [1967]. 8 L. G r i g g i o . E. V i a n e l l o , and C. A. B e n a s s i , Gazz. chim. 5 F . M. V e r o n e s e , A. F o n t a n a , E. B o c c u ’, and C. A. ital. 93, 1412 [1965], B e n a s s i , Gazz. chim. ital. 97, 321 [1967], PREFERENTIAL HYDROLYSIS OF KYNURENINE PEPTIDES 2 9 5 grams without changing the experimental condi­ tions 9. Reduction is accompanied by disappearance of the absorption maximum at 360 m/<, whereas a new maximum is noted at 280 m/< (see Fig. 1 ) with a contemporaneous decrease of fluorescence under UV light.

J______I______I______L -0.4 -0.6 -0.8 -1.0 -1.2 VOtt-VS SCE Fig. 2. Pattern of polarographic reduction of kynurenine in 5% acetic acid.

For these reasons the electrochemical reduction is preferable, especially when small amounts of sub­ stance are available. In order to study the mechanism of cleavage some protected kynurenine peptides were prepared d and the carbobenzoxy group was removed by hydrogeno- Wavelength [m fi] — ► lysis. Their melting points, rotatory power, chro­ Fig. 1. Absorption spectra of kynurenine (1) ( ------) and matographic behaviour and analytical data are re­ y- (o-aminophenyl) -homoserine (2) ( ------) in phosphate ported in Table II. buffer 0.05 M, pH 6. All these kynurenine peptides were reduced to the Furthermore, reduced compounds [ 7- (o-amino- corresponding /-(o-aminophenyl)-homoserine deri­ phenyl)-homoserine derivatives] show on paper and vatives and then hydrolized in mild acide (0.2 N TLC chromatograms a bright yellow reaction with HC1) as well as in mild alkaline conditions (0.5 M p-dimethylaminobenzaldehyde reagent. NaHCOg) in a sealed evacuated tube at 100°. Controlled potential electrolysis appears a very The amino acids released from reduced derivati­ selective method of reduction since no other amino ves at various times of hydrolysis were determined acid is affected with the exception of cystine which is by the analyzer. Fig. 3 a shows the rate quantitatively converted to 10. Fig. 2 shows of liberation of the N-terminal or C-terminal amino acid from the 7-(o-aminophenyl)-homoserine pepti­ a pattern of polarographic reduction of kynurenine des during acidic hydrolysis. The results obtained in 5% acetic acid. with y- (o-aminophenyl) -homoseryl- ethyl Kynurenine and its peptides may alternatively be ester (3), phenylalanyl-/-(o-aminophenyl)-homo­ reduced to y (o-aminophenyl) -homoserine with methyl ester (4), and alanyl- 7- (o-aminophe­ NaBH4 11 but this method is less accurate than the nyl) -homoseryl-glycine ethyl ester (5) are here tabu­ electrochemical one, since over reduction may result, lated. Comparable data have been also obtained with and pure products are obtained only after crystalli­ other model compounds as alanyl-/-(o-aminophe­ zation. nyl)-homoserine methyl ester, carbobenzoxy-y- (o-

9 L. M e i T e s , in: Polarographic Technique (Interscience Pu­ 10 G. M a r k u s , Federat. Proc. 19, 340 [1960]. blishers, Inc., New York, N. Y. 1955). 11 A. P r e v i e r o , M . A. C o l e t t i P r e v i e r o , and P . J o l l e s , Biochim. biophysica Acta [Amsterdam] 124, 400 [1966]. 6 9 2

Yield Solvent of Mp [°C] lift Formula Calcd [%] Found [% Compounds [%] crystallization [not eor] [ I U— .M EOEE E OC’ C .BNSI AD E.SCOFFONE AND C.BENASSI, A. BOCCU’. E. VERONESE. M.F. Ö > i 0 1 q i l IO 0 5 d i Ö X Tf l> o iO CO f r s Ö iO 1 tc iq ^ — 5 0 0 -ri 5 0 o c 0 0 5 0 O »-H 5 0 _ 10 ai c i CM ( a I \ C- CH, H -C o H -C ■C c seryl-glycine ethyl ester (3). Phe Phe (3). ester ethyl seryl-glycine y- hnllnly oaiohnl hmsrn mty etr (4). ester methyl -homoserine (o-aminophenyl) phenylalanyl-y- ia aio cd of acid amino minal ester. methyl serine ty etr ( ester ethyl cdc laae f h Ntria ad -emnl amino C-terminal and N-terminal the of of acid cleavage acidic eesdfo glycyl- from released — + ( * Gly car- and ester - ethyl -phenylalanyl bobenzoxy inophenyl)-homoseryl-glycine am Gly esd rm lnl^pey-ooey-lcn ehl se (6). ester ethyl alanyl-^-phenyl-homoseryl-glycine from leased i. . ) id cdc laae f h Ntria ad C-ter- and N-terminal the of cleavage acidic Mild a) 3. Fig. homoseryl-glycine ethyl ester (6) and in addition addition in and (6) ester ethyl homoseryl-glycine one. inal C-term diffe­ the a than in but rapidly more bonds, leased peptide both labilizes group phenyl-v-hydroxy-butyryl-glycine ethyl ester (8 ), in in ), (8 ester ethyl phenyl-v-hydroxy-butyryl-glycine studied. Thus we have prepared alanyl-;;-phenyl- alanyl-;;-phenyl- prepared have we Thus studied. re­ is acid ino am N-terminal the fact In extent. rent hnl hmsrn ppie, oe oe cm­ com model some the of lacking peptides, pounds -homoserine phenyl) glycyl- hc nt ny rmai aio ru i missing, is group amino atic arom only not which involved in the preferential cleavage of /-(o-am ino- ino- /-(o-am of cleavage preferential the in involved (H— * Glycine ethyl ester, during acidic or basic hydrolysis is par­ is hydrolysis basic or acidic during ester, ethyl Glycine * - .OH (-mnpey)-ooey-lcn ehl se () b Mild b) (5). ester ethyl -homoseryl-glycine (o-aminophenyl) OH h c h sm f lcn ad lcn ehl se dtrie by determined ester ethyl glycine and glycine of sum the authomatic amino acid analyzer. acid amino authomatic mn ai i te iei eprmns a cluae as calculated was experiments kinetic the released the of in amount acid the thus amino glycine, to converted tially The obtained results show that the }'-hydroxy }'-hydroxy the that show results obtained The In order to ascertain if only the hydroxy group is is group hydroxy the only if ascertain to order In (-■•— (G- * G) 2------h n -cooc2 5 2h c o o c 2- h c - n h - o c h c o c I 7 7 -phenyl-y-hydroxy-propylamide (7) and and (7) -phenyl-y-hydroxy-propylamide ) eesd rm -hnlyhdoy butyryl-glycine y-phenyl-y-hydroxy from released b) - L - -phenyl-homoserine derivatives: Gly Gly derivatives: -phenyl-homoserine c o n h ---- 8 - - 8 Gly* ). h c h c h c J I ) eesd from released h) 3 2- - and Ala Ala and h n 7 c o o c y- pey-’hdoypoyaie 7. Gly* (7). -phenyl-}’-hydroxy-propylamide 2 (-mnpey)-ooeie peptides: -homoserine (o-aminophenyl) (O 2 h ortho ----- 5 (O y G) ----

(-miohnl o o­ - hom inophenyl) -(o-am ^ ^ am ino group have been been have group ino am O) and Ala Ala and (A y- c (o-aminophenyl) -homo- (o-aminophenyl) ------eesd rm alanyl- from released H -nh2 h n 2- h c - c o - h n OH ch 2- eesd from released A ch 0~~O) ~ ~ (0 2 (A' ----- A) re­ y-

PREFERENTIAL HYDROLYSIS OF KYNURENINE PEPTIDES 2 9 7

but also the a-carboxy and a-amino functions are the behaviour of alanyl-^-phenyl-homoseryl-glycine respectively absent. ethyl ester (6), glycyl-y-phenyl- 7-hydroxy-propyl- These derivatives were subjected to mild acidic am ide (7) and j'-phenyl-y-hydroxy-butyryl-glycine hydrolysis and the rate of cleavage of peptide bonds, ethyl ester (8). determined on the basis of released amino acids, are reported in Fig. 3 b. A comparison of the results reported in Figs. 3 a and 3 b shows that there is no appreciable difference in the rate of acidic hydrolysis of both peptide bonds of the different compounds. These results point out that the aromatic amino group is not involved in the acidic hydrolysis of re­ 300 0 duced kynurenine peptides. Moreover neither a-carb- tim e [m in ] oxy, nor a-amino group have any influence on the Fig. 4. a) Mild basic cleavage of the N-terminal and C-termi- hydrolysis of the near peptide bond. nal amino acid of y - (o-aminophenyl)-homoserine peptides: Gly* (H------h) released from y- (o-aminophenyl) -homoseryl- These data suggest the following pathway for the glycine ethyl ester (3). Phe (A ----- A) released from phenyl- acidic hydrolysis of C-peptide bond of }'-(o-amino- alanyl-}'- (o-aminophenyl) -homoserine methyl ester (4). Gly* phenyl) -homoserine and ;’-phenyl-homoserine pep­ ( □ ——□ ) and Ala ( O ----- O ) released from alanyl-y-(o- aminopheyl) -homoseryl-glycine ethyl ester (5). b) Mild basic tides: cleavage of N-terminal and C-terminal amino acid of y-phe- nyl-homoserine derivatives: Gly (A ----- A) released from -CH A__ CH n © ,C 2 CH —NH3 2^CH-NH, 2v-c h - n h , glycyl-y-phenyl-y-hydroxy-propyl-amide (7). Gly* (H ---- — + ) I 1 3 I 3 released from y-phenyl-yhydroxy-butyryl-glycine ethyl ester OH i = o C-OH - - i = 0 rsjHF? NHR H® (8). Gly* ( □ ----- □ ) and (Ala ( O ------O ) released from alanyl- 7-phenyl-homoseryl-glycine ethyl ester ( 6). ♦ H,N-R It is noteworthy that a) no hydrolysis of the N- This mechanism is analogous to that demonstrated terminal amino acid bond takes place when the aro­ by B r u ic e and M a r q u a r d t 12 in the acidic hydro­ matic amino group of y-(o-amino-phenyl)-homoserine lysis of y-hydroxy-butyramide. is lacking, and b) the rate of hydrolysis of the C-ter­ On the other hand, the cleavage of the TV-peptide minal amino acid bond is decreased. bond of y- (o-aminophenyl) -homoserine peptides These results prove that the aromatic amino might be explained on the basis of a “N — 0 acil group of reduced kynurenine is involved in the al­ migration” as postulated by Z a h n and ZÜRN 13 for kaline hydrolysis. In our opinion this group does the acidic hydrolysis of hydroxylysyl peptides and not act, however, attacking the peptide bonds since also by S a n g e r and Tuppy 14 in the hydrolysis of it would occur through a very unlike eight or seven- serine and peptides. member ring.

CH, CHo \ c h 7 \ 'CH, V/CH2 .c ^ 2^ ch - c o o h C^ 2 CH-COOH "CH-COOH I 1 C/HOH H ^NH ------0 ^ „ .NH ®NH, Z' H< H © p'C-R / C\ 1 u -nh2 0 HO R 9 10 r1 = _ N H 2 - H .

Considering now the effect of basic medium, also To explain the role of the aromatic amino group in this case both peptides bonds of y- (o-aminophe­ in the hydrolysis it is tentatively proposed the for­ nyl) -homoserine are splitted, in agreement with the mation of a hydrogen bond with the y-hydroxy results of P r e v i e r o et al. n . group, with the result to increase the basicity of this Fig. 4 a shows the rate of cleavage of y- (o-amino­ last. phenyl) -homoserine peptides, while Fig. 4 b reports In order to control the validity of this hypothesis 12 T. C. B ruice and F. H. Marquardt , J. Amer. chem. Soc. the I.R. spectra of a-methyl-o-aminophenyl-carbinol 1,365 [1962]. (9) and a-methyl-phenyl-carbinol (10) were com ­ 13 H. Zahn and L. Zürn, Liebigs Ann. Chem. 613, 76 [1959]. 14 F. Sanger and H. Tuppy, Biochem. J. 49, 463 [1951]. pared. 2 9 8 F. M. VERONESE, E. BOCCU’, C. A. BENASSI, AND E. SCOFFONE

IR investigations of a 0.001 M solution in carbon Methods. — UV spectra were obtained with a Beck­ tetrachloride give evidence for an intramolecular man Model DB spectrophotometer connected to a Sargent recorder. Absorptions at single wavelengths hydrogen bond between amino and hydroxy group were determined with a Hilger Uvispek Model H. 700 as appears from observed v OH-frequencies (3600 spectrophotometer. The IR spectra were obtained with and 3680 cm-1 respectively). a Perkin-Elmer Model 337 spectrophotometer. The pH The intramolecular partecipation of amino and was determined with a Beckman expanded scale pH- hydroxy group in alkaline hydrolysis of ;’-(o-amino- meter, Model 76. The rotatory powers were measured with a Perkin-Elmer Model 141 polarimeter; concen­ phenyl) -homoserine peptide bonds may than be de­ trations are given in grams per 100 ml of solvent. The picted as follows: melting points were determined on a Tottoli apparatus ^coo© „C 0 0 © (Biichi) and are uncorrected. Hv CH?—CH. K ^CH-C ning and sometimes during the course of the reduction. As shown in Fig. 4 this participation of the ortho Reduction of kynurenine peptides was considered ended amino group is a prerequisit for the cleavage of the at the disappearance of the characteristic reduction N-terminal bond of y- (o-aminophenyl) -homoserine wave of its keto group (E i/ 2 ^ — 1.0 Volts, see Table I ) . Amino Acid Analyzer was a Technicon apparatus peptides, whereas in the cleavage of the C-peptide (Technicon Instruments Company Ltd, Chertsey, Eng­ bond it probably acts also increasing the reaction land) . rate. Thin layer chromatography (Si02) was performed In this regard we remember that in alkaline media using the following solvent mixtures: n-butyl : not only the C-terminal bond of y-phenyl-homo- water: glacial acetic acid (3:1:1) (Rfi) ; ethyl ace­ tate: : wrater: glacial acetic acid (60:20:14:6) serine peptide is cleaved (see above) but also amide (Rf2). The compounds were in turn detected by nin- bond of y-hydroxy-butyramide (see B r u ic e and hydrin spray, hypochlorite test20 and reaction with M a r q u a r d t 12) . p-dimethylaminobenzaldehyde. The participation of aromatic amino group of Kynurenine peptides. — The V-benzyloxycarbonvl group of protected kynurenine peptides, prepared ac­ y- (o-aminophenyl) -homoserine by hydrogen bond cording our procedure 5, was removed by hydrogenoly- coordination seems to play an assistance to the cata­ sis in the presence of palladium on charcoal as a cata­ lysis similar to that observed in some enzymes lo. lyst in methanol or ethanol. The catalyst was removed and the solvent evaporated in vacuo. The residual oil, Experimental after addition of an equimolecular amount of dry HC1 in ethanol, was cristallized from various solvents (see Materials. — /2-Benzoyl propionic acid 16, ^-benzoyl Table II). ethylamine17 and a-amino-/?-benzoyl-propionic acid 18 ß-Benzoyl-propionyl-p-nitrophenyl ester. — To a were prepared according to literature. This last com­ solution of 3.56 g (20 mmoles) of /?-benzoyl propionic pound was also prepared from acetophenone ac­ acid, 2.78 g (20 mmoles) of p-nitrophenol in 100 ml of cording to a new procedure below reported. a-Methyl anhydrous , cooled to 0°, 4.12 g (20 mmoles) o-aminophenyl carbinol and a-methyl phenyl carbinol of dicyclohexylcarbodiimide were added. After 30 mi­ were prepared respectively from o-nitroacetophenon and nutes at 0C and 24 hrs at room temperature, the pre­ acetophenon 19. The substances.for these synthesis were cipitate of dicyclohexylurea was filtered off and washed Fluka products (Fluka AG, Basle, Switzerland). All with chloroform. After washings with 5% NaHC03 and other reagents, when not otherwise specified, were water, the solution was evaporated to dryness and the Merck reagent grade products (Merck AG, Darmstadt, residue crystallized from methanol-ethyl ether. The Germany). yield was 4.78 g (80,'o), mp 124°.

15 D. F in d l a y , D. G. H e r r ie s , A . P. M a t h ia s , B . R . R a b i n , 18 M. M. F r a s e r and R. A. R a p h a e l , J. chem. Soc. [London] an d C. Ross, Nature [London] 190.781 [1961]. 1930. 2245. 16 L. F . S o m e r v il l e a n d C. F . A l l e n , Org. Syntheses 2, 81 19 U. N a g a i , T. S h i s h i d o , R . C h i b a , and H. M i t s u h a s h i , N e w Y o rk . Tetrahedron [London] 21,1701 [1965], 17 W. J. H a l e an d E. C. B r i t t o n , J. Amer. chem. Soc. 41, 841 20 S. C. P o r n and J. T. D u t c h e r , Analvtic. Chem. 38. 836 [1919]. [1956], PREFERENTIAL HYDROLYSIS OF KYNURENINE PEPTIDES 2 9 9

Anal. Calcd. for C16H13N105: C, 64.21; H, 4.38; In order to confirm the complete reduction of the N, 4.68. Found: C, 64.17; H, 4.33; N, 4.64. keto group, a sample was hydrolyzed in HC1 6 N at 110 ß-Benzoyl-propionyl-glycine ethyl ester. — To a so­ for 20 hrs and then chromatographed on Whatman n. 1 lution of 2.94 g (21 mmoles) of glycine ethyl ester paper sheet (solvent mixture n-butyl alcohol-water-ace­ hydrochloride with 21 mmoles of triethylamine in 100 acid, 40:50:3). Ninhydrin reaction revealed only ml of anhydrous chloroform, 5.98 g (20 mmoles) of two spots: glycine, Rf = 0.3 and y-phenyl-y-hvdroxy-pro- yS-benzoyl-propionyl-p-nitrophenyl ester were added, pyl-amine, Rf 0.67, whereas /J-amino-propiophenone (Rf and the mixture was stirred for 48 hrs. After removing 0.53) did not appear on the chromatogram. the precipitate of the triethylammonium salt and alka­ Glycyl-y-phenyl-y-hydroxy-propylamide hydrochloride. line and acidic washings, the solvent was removed — A solution of carbobenzoxy-glycyl-}'-phenl-y-hydroxy- in vacuo and the residue crystallized from methanol- propylamide (1.7 g) in 60 ml of methanol and few ethyl ether. The yield was 3.94 g (35%), mp 78°. drops of acetic acid, was hydrogenated in the presence Anal. Calcd for C14H1704N: C, 63.86; H, 6.51: of 0.3 g of 10% palladium on chorcoal for 6 hrs. The N, 5.32. Found: C, 63.90; H, 6.37; N, 5,26. filtrate was evaporated in vacuo after removal of cata­ y-Phenyl-y-hydroxybutyryl-glycine ethyl ester. — To lyst, and the residual oil added of equimolecular a solution of 1.32 g (5 mmoles) of /?-benzoyl-propionvl- amount of dry HC1, crystallized from ethanol-ethyl glycine ethyl ester and 1.5 g oxalic acid in 10 ml of ether. The yield was 4g (83%), mp 100°; single nin­ 50% methanol, 1.85 g (50 mmoles) of NaBH4 were hydrin positive spot (Rfi 0.75, /?f2 0.50). added slowly under vigorous stirring. When a negative Anal. Calcd. for C11H170 2N2C1: C, 54.10; H, 6.97; acetone-test (2,4-dinitropheyl hydrazine) was observed N, 11.48. Found: C, 54.00; H, 6.93; N, 11.40. (about 30 minutes), few drops of conc. HC1 were added to destroy the excess of NaBH4 . The reaction mixture, a-Amino-ß-benzoyl-propionic acid. — It was pre­ diluted with water and freed from methanol under re­ pared according to F r a s e r and R a p h a e l from benzoyl- duced pressure, was extracted with ethyl acetate. After acrylic acid 18. This product was utilized for the suc­ washing with water, the ethyl acetate solution was eva­ cessive synthesis, in addition, in order to control the porated to dryness and the residue crystallized from retention volume of its reduction product, a-amino-/i- -ligroine. The yield was 1.19 g (90%), mp 113°. benzoyl-propionic acid was also reduced with NaBH4 . Anal. Calcd. for C14H190 4N: C, 63.36; H, 7.22; The obtained compound y-phenyl-homoserine, which is N, 5.28. Found: C, 63.41; H, 7.30; N, 5.33. the substance formed during hydrolysis of y-phenyl- Carbobenzoxy-glycyl-ß-benzoyl-ethylamide. — To a homoseryl peptides, emerged from the column in two solution of 1.85 g (10 mmoles) of /?-aminopropiophe- distinct peaks. none hydrochloride and 10 mmoles of triethylamine in In fact the possibility to obtain, from benzoyl-acrylic 60 ml anhydrous chloroform, 3.33 g (11 mmoles) of acid and ammonia, a mixture of a-amino-/?-benzoyl-pro- carbobenzoxy-glycine-p-nitrophenyl ester were added. pionic and /?-amino-/?-benzoyl-propionic acids was al­ After 48 hrs at room temperature the solution was ready discussed by several authors 18, 21, 22. On the aim washed with 1 N HC1 and 5% NaHC03, dried over to overcome such a doubt we devised a new unambigous anhydrous Na2S04 , concentrated to small volume and route of synthesis for a-amino-/?-benzoyl-propionic acid. cooled at 0°. The white crystalline product was col­ A solution of 60 ml of anhydrous methanol, 0.6 g of lected by filtration and dried in vacuo. The yield was Na, 5.7 g (26.2 mmoles) of diethyl acetaminomalonate 3.06 g (90%), mp 116 — 117°, single ninhydrin posi­ and 4.9 g (24.6 mmoles) of phenacylbromide was tive spot (Rfx 0.90; Rf2 0.75). heated at reflux for 6 hrs. After removal of the solvent Anal. Calcd. for C, 67.04; H, 5.93; in vacuo, the residual oil was extracted with chloro­ N, 8.23. Found: C, 67.10; H, 5.95; N, 8.13. form; the solution was concentrated to dryness and the residue hydrolyzed with a mixture of 50 ml of conc. Carbobenzoxy-glycyl-y-phenyl-y-hydroxy-propylamide. HC1 and 20 ml of glacial acetic acid for 6 hrs at 90“. — To a solution of 1.70 g (9 mmoles) of carbobenz- The reaction mixture, diluted with 300 ml of water, oxy-glycyl-/3-benzoyl-ethylamide and 1.5 g of oxalic was extracted for 24 hrs in a liquid-liquid apparatus acid in 10 ml of 50% methanol, 1.85 g (50 mmoles) of with ethyl ether. The ethereal extract left a residue NaBH4 were added slowly under vigorous stirring. (3 g) that dissolved in 100 ml of water and brought to After 30 minutes (negative acetone test) a few drops of pH 7 with NaOH solution, was applied to a 3 x 30 cm conc. HC1 were added, the mixture was diluted with column of strong cationic resin Amberlite IR 120(H®). water and cooled. The collected precipitate was washed The first 500 ml water washings were discharged and with water and dried. The yield was 1.30 g (76%) mp the column was then eluted with 2 / of 0.2 N HC1 which 133°. yielded 1.2 g of a crystalline compound identified as Anal. Calcd. for C19H*>.,04N2: C, 66.64; H, 6.49; glycine hydrochloride by its chromatographic be­ N, 8.18. Found: C, 66.57; H, 6.44; N, 8.06. haviour and analytical data. The column was finally

21 J. B o u g a u l t , Ann. Chim. analyt. IS, 491 [1908], 22 J. B o u g a u l t and P. C h a b r i e r , C . R. hebd. Seances Acad. Sei. 226, 1378 [1948]. 3 0 0 PREFERENTIAL HYDROLYSIS OF KYNURENINE PEPTIDES washed with 3 / of 1 N HC1 releasing, after concentra­ removal, the filtrate was added of equimolecular amount tion, 1.5 g of a-amino-/5-benzoyl-propionic acid hydrochlo­ of dry HC1 in ethanol and concentrated to dryness. The ride chromatographically and analytically identical to residue was crystallized from ethanol-ethyl ether. The the substance prepared from benzoyl acrylic acid. It yield was 1.70 g, mp 110°, single ninhydrin positive was then reduced with NaBH4 to y-phenyl-homoserine spot (Rh 0.73; R,2 0.46). which also gave two distinct peaks when chromato­ Anal. Calcd. for C14H18N204 • HCl: C, 53.40; H. graphed in the amino acid analyzer. This fact suggested 5.77; N, 8.90. Found: C, 53.60; H, 6.01; N, 8.85. that the two peaks had to be related to the existence of the two pairs of enantiomers. Carbobenzoxy-alanyl-a-amino-ß-benzoyl-propionyl-gly- Carbobenzoxy-a-amino-ß-benzoyl-propionic acid. — A cine ethyl ester. — To a solution of 2.08 g (6 mmoles) solution of 1.93 g (10 mmoles) of a-amino-/?-benzoyl- of a-amino-/?-benzoyl-propionyl glycine ethyl ester hy­ propionic acid in 200 ml of 1 M NaHC03 , cooled at drochloride in 10 ml of pyridine, 1.72 g (5 mmoles) of O , was added under constant stirring of 2.04 g (12 carbobenzoxy-alanyl-p-nitrophenyl ester were added mmoles) of benzyloxycarbonyl chloride, the pH of the under stirring. After 48 hrs at room temperature the solution being continuously held at pH 9 by addition solution was concentrated to dryness in vacuo. The of N NaOH. After 3 hrs the solution was extracted with ethyl acetate solution of the residue was washed with ethyl ether, cooled at 0° and acidified to Congo red 1 N HC1, 5% NaHCO;? solution and water. After drying with 5 N HC1. The precipitate, formed on standing, over Na2S04 the solution was concentrated in vacuo was collected, washed with water and dried. The yield and added of petroleum ether. After standing for was 2.84 g (87%), mp 128°; ninhydrin negative spot several hours at 0°, a crystalline product separated out. (Rfl 0.65; R,2 0.35). The yield was 2 g (83%), mp 135°. Anal. Calcd. for C18H17N05: C, 66.05; H, 5.24; Anal. Calcd. for C9-,Ho9N30 7: C, 62.10; H, 6.05; N, 4.28. Found: C, 66.15; H, 5.25; N, 4.15. N, 8.69. Found: C, 62.40; H, 6.02; N, 8.83.

Carbobenzoxy-a-amino-ß-benzoyl-propionyl-p-nitrophe- Carbobenzoxy-alanyl-y-phenyl-homoseryl-glycine ethyl nyl ester. — To a solution of 3.27 g (10 mmoles) of ester. — To a solution of 2.42 g (5 mmoles) of carbo­ carbobenzoxy-a-amino-/?-benzoyl propionic acid, 1.5 g benzoxy - alanyl - a - amino - ß -benzovl-propionyl - glycine (11 mmoles) of p-nitrophenol in 60 ml of anhydrous ethyl ester and 1.5 g of oxalic acid in 20 ml of 50% chloroform cooled at 0°, 2.06 g (10 mmoles) of di- methanol, 1.85 g (50 mmoles) of NaBH4 were slowly cyclohexylcarbodiimide were added under stirring. added under stirring. At the end of the reaction, After 48 hrs at room temperature the solution was checked with the disappearance of reactivity toward washed with 5% NaHC03 , dried over Na2S04, con­ 2,4-dinitrophenyl-hydrazine, few drops of conc. HC1 centrated in vacuo. The residue, crystallized from ethyl were added to destroy the excess of NaBH4 . The reac­ acetate-ethyl ether gave yellow crystals (4.12 g) melting tion mixture was diluted with water, freed from MeOH at 224°, and giving a single spot, hypochlorite starch by evaporation under reduced pressure and extracted positive, Rjx 0.9. with ethyl acetate. After washing with water, solution Anal. Calcd. for C.>4H.,0NoO7: C, 64.28; H. 4.50; was evaporated to dryness and the residue, crystallized N, 6.25. Found: C, 64.00: H, 4.47; N. 6.28. from ethyl acetate-ethyl ether. The yield was 1.37 g, mp 148°, Carbobenzoxy- a- amino -ß- benzoyl - propionyl - glycine ethyl ester. — To a solution of 1.7 g (12 mmoles) of Anal. Calcd. for C2-,H31N30 7: C, 61.84; H, 6.44: glycine ethyl ester hydrochloride and 12 mmoles of tri- N, 8.66. Found: C, 61.76; H, 6.38; N, 8.60. ethvlamine in 60 ml of anhydrous chloroform, 4.48 g Alanyl-y-phenyl-homoseryl-gly cine ethyl ester. — A (10 mmoles) of a-amino-/3’-benzoyl propionic acid-p- solution of carbobenzoxy-alanyl-r-phenyl-homoseryl- nitrophenvl ester were added. After 24 hrs at room glycine ethyl ester (1 g) in 100 ml of methanol, was temperature the solution, washed with 1 N HC1 and 5% hydrogenated in the presence of 10% palladium on NaHC03 , was dried over Na2S04 and concentrated in charcoal. After 12 hrs the catalyst was removed and vacuo. The solid residue (3.2 g; 78%), crystallized the solution was added of equimolecular amount of twice from ethyl acetate, melted at 153°. dry HC1 in methanol and concentrated to dryness. The Anal. Calcd. for CooHo4N>06: C, 64.06; H, 5.87; residue was crystallized from methanol-ethvl ether. The N, 6.79. Found: C, 63?93H.“5.94; N, 6.78. yield was 0.59 g, mp 138°; single ninhydrin positive spot {Rfi 0.68; Rf2 0.40). The same product (70% yield, mp 154°) was also obtained with mixed anhydride method (ethyl chloro- Anal. Calcd. for C17H25N305-HCl: C, 52.71; H, carbonate) from carbobenzoxy-a-amino-/?-benzoyl-pro- 6.72; N, 10.85. Found: C’ 52.68; H, 6.58; N, 10.79. pionic acid and glycine ethyl ester hydrochloride. a-Amino-ß-benzoyl-propionyl-glycine ethyl ester hy­ The authors wish to thank Prof. F. P u l i d o r i (Che­ drochloride. — A solution of carbobenzoxy-a-amino-/3- mical Institute of Ferrara) for the polarographic benzoyl propionyl-glycine ethyl ester (3 g) in 350 ml of measurements, Dr. A . P ietrogrande for the elemen­ acetic acid was hydrogenated in the presence of 0.3 g tal analysis and Miss S. A g o s t i n e t t i for technical as­ of 10% palladium on charcoal for 6 hrs. After catalyst sistance.