639-1
IRVING I. GESCHWIND. CH0I-I HA0 LI
[CONTRIBUTION FROM
THE
AND
LIvIO BARNAFI
Vol. i 9
HORMOXE RESEARCHLABORATORY AVD THE DEPARTMENT OF BIOCHEMISTRY, UKIVERSITYOF CALIFORNIA, BERKELEY]
The Isolation, Characterization and Amino Acid Sequence of a Melanocyte-stimulating Hormone from Bovine Pituitary Glands BY IRVING I. GESCHKIND, CHOHHAOLI
AND
LIVIOB.ZRN.\FI
RECEIVED JUNE 20, 1957
A melanocyte-stimulating hormone has been isolated from bovine posterior lobe pow der by a procedure similar to that employed previously for the isolation of porcine p-MSH. The isoelectric point of the bovine hormone has been fouiitl t ( J be more basic than that of the porcine polypeptide. By means of chemical and enzymic degradationq, the structure of the bovine hormone has been shown t o be H.Asp.Ser.Gly.Pro.Tyr.Lys.Met.Glu.His,Phe.~4rg.Try,Gl~~.Ser.Pro.Pro,L~s,Asp.OH 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 T h e sequence of amino acids in the bovine hormone differs from that in the porcine in the substitution of a serine residue for a glutamic at position 2. A comparison is made between this “sery1-,3-MSH” and the bovine hISH preparation obtailleti b!Benfey and Purvis.’ Purification by Zone Electrophoresis on Starch.-PrelimThe purification and isolation of one or the other of two porcine melanocyte-stimulating hormones inary experiments indicated that the hISH activity in the crude concentrate had a greater cathodic mobility (MSH) have been described recently by a number abovine t pH 4.9 than did that in the porcine crude concentrate. of investigator^'-^ and the structure of one of these -ittempts were therefore made to perform the electrophoresis hormones, P-MSH, has been e l u ~ i d a t e das~ well. ~ ~ ~ , a~t a higher pH, and a pH 6.5 pyridine-acetic acid buffer was On the other hand, although considerable effort chosen. Although excellent resolution was obtained a t this H , the use of this particular buffer was discontinued hewas expended prior to 1950 on the purification of a pcause of certain technical difficulties. l m o n g these wcrc bovine MSH,’ interest in this hormone has been the increased trapping of bubbles in t h e starch during packsuperseded by the recent focus on the porcine hor- ing and the iricomplete sedimentation of the individu:il starch segments during centrifugation in the presence o f the mones, so that only a few reports in the past few buffer. Consequently, further purification by zone electroyears have been concerned with attempts to purify phoresis on starch was carried out mith the same buffer the bovine hormone. In part, the lack of interest and by the same procedure as that employed for the porcine in the latter hormone derives from the fact that hormone4-’*except that only 100 mg. was used per trough; beef glands are not as rich a source of -1ISH ac- a t higher loads the resolution was poor. -1typical clistribution pattern of material revealed by the reagent of Lowry, tivity as are pig glands; thus, various workers et al.,I3 is shown in the upper half of Fig. 1. The hulk of have reported that pig posterior lobes8 possess 5 the MSH activitJ-Ii was concentrated in segmeuts 22-31; timesg or 2 timeslo as much MSH activity as does the extracts of these segments mere pooled, filtered miti lyiiphilized. T h e recovered po\vder was routinely rerun uridcr an equal weight of beef posterior lobes. Never- identical conditions a t a load of 100 mg. per trough. The theless, concentration of bovine MSH activity has pattern obtained from such a rerun is shown in the 1i)xvc.r been achieved by Landgrebe and and half of Fig. 1. Purification by Countercurrent Distribution.-The purithe purification of a bovine MSH by means of countercurrent distribution has been reported by fied material (200-100 mg.) obtained from zone electrophoresis was submitted t o countercurrent distribution in the autoBenfey and Purvis.lb The present communication matic all-glass apparatus’s and in the same system that was describes the isolation and structure of a bovine used for porcine fi-MSH,i namely, %-hutanol-aqueous O.Sf)h MSH which differs markedly in amino acid com- trichloroacetic acid; the other details of the procedure wiverc position and distribution behavior from that re- similar to those used iti connection with the porcine horirioiie. h e volume of the lower phase was 5.0 ml., and t1i:it of tlic ported by Benfey and Purvis. A preliniitiary re- T upper phase, 5.4 xnl. port of these data has appeared recently.“ In a typical experiment, after a forerun of 16 transfers.
Isolation Procedure Crude MSH Concentrate.-This concentrate mas prepared in exactly the same manner as the concentrate from pig glands.4 (1) (a) B. G. Benfey and J. L. Purvis, THISJ O U R X A L , 77, 5167 (1955); (h) Biochem. J . , 62, 588 (1956). (2) A. B. Lerner and T . H. Lee, THIS JOL-RXAL, 77, 100ri (10>3). ( 3 ) J. 0. Porath, P . Roos, F. W Landgrehe and G . If.hlitchell, Biochinz. Biophrs. A c t a , 17, 598 (1955). (4) (a) I. I. Geschwind, C . H. Li and L. Barnafi, THISJ O U R N A L , 76, 4494 (195G); (b) I. I. Geschwind and C. H . L i , i b i d . , 79, 613 (1957). (5) J. I. Harris and P . Roos, Xotuve, 178, 90 (1950). (6) I. I. Geschwind, C. H . Li and L. Barnafi, THISJ O V R N A L , 7 9 , G?O 11957). (7) H . Waring and F. W.Landgrebe, in “ T h e Hormones,” x-01. 11, Ed. by G. Pincus and K. V . Thimann, Academic Press, I n c . , S e w 1950. York, N. T., (8) As used in this paper, “posterior lobe” denotes posterior plus intermediate lobes of t h e pituitary. (9) F. W .Landgrebe and C. Lf. LIitchell, Ozrarl. J . E n p e v . P h r r i o l . , 39, 11 (1951). (10) T. H . Lee and A . B. Lerner, J . B i d C h e w . , 221, 943 (1956). (11) I. I. Geschivind, C. H . T,i and T.. Barnafi, THISJ O U R N A L , 79, 1003 (1957).
370 mg. of material containing approsimately 90 tng. of starch from the electrophoretic run TWS :tclded to the combined lower phases of the first two tubes. The mixture WIS stirred and centrifuged. The supernatant fluid was collected with a pipet, and the precipitate (insoluble starch) w a s washed once with the combined upper phases of the first two tubes. The upper and lolyer phases were put back inti] the machine which \vas then shaken tii:~iiuall~ t i l permit equilibration of the phases in the first twri tubes. .iftcr the phases had separated, the remainder of the run x i s cxricti out automatically. .After 240 transfers, the optical density of the contents of every fourth tube (lower p l ~ a s efroiri tubes 0 t o 160; upper phase from tubes 161 to 239) wis determined a t 277 nip. Over 95yoo f the material w a s I)rcsetit (12) T h e run v a s carried out a t 1 O . I n a previous puhlicatir~n.“’ because of a typographical error, this temperature \\‘a> erroneouily reported as ? O c . . I,. Farr and R. J. Randall. (13) 0. H . Lowry, E.J. Rosebrough. A J . B i d . C h o n . , 193, P G 5 (1051). (11) T h e melanocyte-stimulntiny activity was determined by- t h e method described by E;.Shizume, A. B. Lerner and T. B. Fitzpatrick ( E , i d o o i , i o l o g y , 64, 5 . 3 (1931)); the unit is the came :IS thxt descrilxrl by these in\.estigators. (15) L. C. Craig. W .Hausmann, E. H . Ahrens and 73. J. IIarfenist, Anal. Chciiz., 23, 1276 (1931).
Dee. 20, 1957
I S O L A T I O N OF A
0mot
6 395
h
o,70:m
'; h
MELANOCYTE-STIMULATING HORMONE
42
46
50
&I
5 0600 C
E 0500 Q
0 0.300
I", IO
0
r 14
' 18
1
1
22
26
1
1
I
130 3 4 J 38
I
42
Segment Number (cm.).
Fig. 1.-Upper, pattern obtained from zone electrophoresis of crude bovine MSH concentrate on starch; 60 cm. trough, 0.1 ill pyridine-acetic acid buffer a t p H 4.9, 230 volts, 48 hours. T h e activity is concentrated in segments 22-31. Lower, electrophoretic pattern obtained with material that had previously been purified by zone electrophoresis on starch. Conditions the same as above. in a major peak with a K value of approximately 0.6 and a trailing shoulder with a K value of about 0.4. T h e remaining optical density was present as a high, irregular baseline extending from tubes 162 t o 222. The contents of these latter tubes were then removed, and were replaced by fresh upper and lower phases. T h e machine was then set for cycling and was allowed t o proceed automatically t o a total of 702 transfers. T h e optical densities of the lower phases of every fourth tube were then determined, and, once the main peaks were located, the contents of every other tube making up the peaks also mere analyzed. The distribution obtained is shown in the upper part of Fig. 2. There are present one main peak ( K = 0.53) and two smaller peaks ( K = 0.37 and 0.30). More than 9070 of the biological activity was located in the main peak. T h e contents of tubes 226-250 (pool I ) and of tubes 251278 (pool 11)were separately pooled, concentrated in a rotary evaporator, extracted with diethyl ether and subsequently lyophilized. T h e distribution apparatus was then filled with fresh upper and lower phases; pool I (87 mg.) was added t o the lower phase of the first tube of the upper train and pool I1 (98 mg.) was added t o the lower phase of the first tube (tube 120) of the lower train. The machine was set for recycling and allowed t o proceed automatically for 251 transfers. Analysis of lower phases in alternate tubes revealed the distribution shown in the lower half of Fig. 2. Both peaks have practically identical partition coefficients (0.52, 0.53) and both show some trailing, which is more evident in the larger peak (pool 11). -4 small but definite peak with a K value of approximately 0.3 is also seen. T h e material representing each of the two main peaks was individually pooled, concentrated and lyophilized; 69 mg. of pool I and i 2 mg. of pool I1 were obtained (13.1Y0 3; trichloroacetate) . From 1 kg. of acetone-dried powder of bovine posterior pituitaries, approximately 150 mg. of MSH was isolated. The average activity of such preparations is 0.2 X lo7 u./mg.14 Characterization Amino Acid Analysis .-Quantitative amino acid analyses by the method of Levy16have been performed on 3 different samples of MSH obtained following countercurrent distribution. In all cases, 2 t o 3-mg. samples were hydrolyzed (16) A. I,. Levy, Zaluue, 174, 126 (1954).
Tube Number. Fig. 2.-Upper, countercurrent distribution (702 transfers) of material obtained from zone electrophoresis on starch. System, 2-butanol-0.5yc aqueous trichloroacetic acid. Lower, countercurrent distribution (251 transfers) of the two halves of the main peak obtained above (upper, Fig. 2). The material found in tubes 226-250 after the first distribution was placed in tube 0 and that found in tubes 251-278 was placed in tube 120. System same as above. a t 100" in 5.7 N HC1 for 24 hours. The hydrolyzates were dinitrophenylated according to the method of Sanger17 in 6370 ethanol containing 2yo S a H C 0 3 and 470 fluorodinitrobenzene ( F D S B ) . T h e reaction mixture was shaken for 3 hours a t room temperature on a simple shaker designed by L. Barnafi.'* The dinitrophenylated amino acids were then submitted to two-dimensional chromatography on paper, in the "toluene" and 1.6 -2.1phosphate systems described by Levy16; each sample was run in triplicate or quadruplicate. T h e yellow spots were excised and eluted, and the optical densities were determined. I n this manner the molar ratios for all the amino acids, with the exception of tryptophan, were obtained. Tryptophan19 and amide ammoniazo were estimated separately. I n order to determine the optical configuration of the amino acids, a fourth sample was hydrolyzed and submitted to microbiological assay.*I T h e results of the F D S B analyses on three different countercurrent preparations of bovine hISH and of the single microbiological assay (Table I) suggest the following empirical formula for the bovine MSH
Asp~Glu1Ser~G1~~Pro~~~et1PhelTyrlLys?His~~~rg~Try~ All the optically active amino acids are of the L-configuration (see Discussion for configuration of tryptophan). (17) F. Sanger, Biochem. J., 39, 507 (1945). (18) This shaker has as 'its basic component a vacuum-operated automobile windshield-wiper motor. (19) T. W. Goodn-in and R. 0 . Morton, Biochenz. J.,40, 628 (19.16). (20) A. L. Levy, I. Geschwind and C. H. Li, J . B i d . Chem., 213, 187 (1955). (21) T h e microbiological assay was carried o u t by Shankman Laboratories, Los Angeles. Only organisms which respond solely t o t h e L-form were used. No determination of glycine was made, and in the case of glutamic acid only a few determinations were made,
6 396
IRVING
I. GESCHWIND, CHOH HAOLI
T h e minimum molecular weight calculated from this formula is 2135; t h e theoretical nitrogen content is 17.0%. Determination of Isoelectric Point.-The electrophoretic mobilities of bovine MSH on starch were determined as previously d e ~ c r i b e d . ~4 s shown in Fig. 3, the isoelectric point of the hormone was found to be p H 7.0.
"00.51
\ 9.0pH
0.5 i
\
f
a
-2 2.5
% I I
W
3.0
Fig. 3.-Electrophoretic mobility of bovine MSH on starch as a function of PH. The mobilities have been corrected for electroosmosis. Degradation Procedures N-Terminal Amino Acid Sequence.-The K-terminal residue of the MSH molecule was determined on 3 mg. of h l S H by the F D S B " procedure; the sole ether-soluble D S P amino acid found following acid hydrolysis mas identified by the chromatographic procedure of LevyI6 as D S P aspartic acid; a yield of 0.6 p M per 3 mg. of the dinitrophenylated MSH was obtained. This represents arecoveryof 0.6 mole/ mole, assuming t h a t 5 D K P residues (reaction with one aS H z , two E-SHZ,one 0-Tyr and one Im-His) have been introduced into the dinitrophenylated hormone, thereby increasing the molecular weight of the hormone from approximately 2100 to 3000. Since there have been reports by other workers t h a t AT-terminal aspartic or asparagine residues may be lost occasionally in the course of dinitrophenylation,22pz3this possibility was investigated. Examination of t h e ether extracts of t h e acidified dinitrophenylation reaction mixture revealed the presence of approximately 0.05 mole/mole of D S P aspartic acid. This value, though extremely small, was observed in two different experiments. The aqueous solution remaining after ether extraction of t h e hydrolyzate of the D S P peptide was chromatographed on paper in 1 M Sa2HP04, a system which permits the separation of CY- and E - D N P lysines.24 -4 recovery of 1.7 p i l l of e-DNP lysine was obtained. S o other water-soluble, colored, D S P amino acid was evident. The sequence of amino acids at the N-terminus was determined by means of the paper-strip modification46 of t h e phenyl isothiocyanate procedurez6; 5 mg. of the hormone was employed for each experiment. The phenylthiohydantoins ( P T H ) obtained at each step were identified following chromatography of a n aliquot on p a p e r in the heptane-pyridine aliquot was hydrolyzed with system of S j o q ~ i s t . ~ Another ' (22) (23) (24) (25) (26) (27)
E. 0. P. Thompson, Biochim. Biophys. Acta, 10,633 (1953). F. Weygand and R. Junk, Z. physiol. Chem., 300, 27 (1955). J. Cummins, personal communication. H. Fraenkel-Conrat, THISJ O U R N A L , 76, 3006 (1954). P. Edman, Acta Chem. S c a d . , 4, 283 (1960). J. Sjoquist, i b i d . , 7, 447 (1953).
AND
LIVIOBARNAFI
VOl. 79
TABLE I MOLARRATIOS OF AMIXOACIDSIN BOVISE P-XSH Amino acid
CCI"
Glutamic acid 1.1 Aspartic acid 1.8 Serined 2.0 Glycine 2.0 Proline 3.0 Methionine 1.2 Phenylalanine 1 .0 Tyrosine 0.9 Lysine 1.8 Histidine 1.0 Arginine 0.9 Tryptophan 1.2 Amide KH3
RISH preparations CCIIQ CCIIIb CCIV':
0.9 2.2 1.9 2.0 3.0 1.0 1.1 1.1 1.9 1.1 1.0 1.0 0.0
0.88 2.08 2.05 1.98 3.13 1.00 0.90 1.00 1.95 1.00 1.13
0.0
0.82 2.10 1.98
2.90 0 97 1.11 0.93 2.10 1.02 0.89
Residues to the nearest integer
1 I )
2 2 3 1 1 1 2 1 1 1
..
Total 18 Analysis by F D S B methodi6; average of three chromntograms. Analysis by F D K B method; average of fuur chromatograms. e Analysis by microbiological assay .zl d Corrected for 10% destruction. constant boiling HC1 at 150" for 16 hoursz8t o regenerate the parent amino acids which were identified after chromatography in the system n-butanol-acetic a c i d - ~ a t e r . ~ g The degradation procedure was employed successfully for seven steps, revealing in order the following amino acids: Asp, Ser, Gly, Pro, Tyr, Lys and M e t . In t h e second step, along with the P T H of serine, a very definite spot of the P T H of glycine was also evident on the chromatograms. This latter spot reached its maximum at the third step, was still evident a t the fourth, and was no longer visible by the fifth step. Digestion with Leucine Aminopeptidase (LAP) .30-Digestion with LXP was employed as a n enzymic approach for determining the sequence of amino acids a t the K-terminus. One ml. of the enzyme solution (0.6 mg. enzyme) treated with diisopropyl f l u ~ r o p h o s p h a t ewas ~ ~ added to 0.6-0.7 p W 2 of the hormone in 1 ml. of solution (0.02 31 tris buffer, PH 8.1; 0.002 M hlgc+) t o give a n enzpme:substrate molar ratio of 1:300-1:350. The resulting solution was incubated a t 40". A4t designated intervals up to 24 hours, aliquots of the reaction mixture were removed and allowed to react with FDhTB. T h e free D K P amino acids were then extracted from the acidified reaction mixture into ether and the amounts of the D S P amino acids were quantitatively determined by the method of Levy. The rate of release of amino acids by L.4P is plotted in Fig. 43 and, for comparison, the rate of release of amino acids from porcine 8-MSH is presented in Fig. 4b. Even after 24 hours, only 0.43 mole per mole of the S-terminal aspartic acid of bovine MSH had been released (these results are typical of three different digestions carried out with different batches of bovine MSH). The only other amino acid found was serine, the amount of which also increased steadily during the 2 1 hours of digestion. A comparison of the results of these experiments on bovine hfSH with those obtained with pMSH will be reserved for the Discussion. C-Terminal Amino Acid Sequence.-The sole approach applied to the investigation of the sequence of amino acids at t h e C-terminus was enzymic, For these investigations carboxypeptidase (Worthington) was treated with a 50-fold molar excess of D F P , and 0.6 mg. of the treated enzyme was incubated at 40" with 1 pilf of the hormone (enzyme:sub(28) A. L. Levy, Biochim. Biophys. A c l a , 1 6 , 589 (1954). (29) S. M. Partridge, Biochcm. J . , 42, 238 (1948). (30) D. H. Spackman, E. L. Smith and D. R l . Brown, J . Bid. Chem., 212, 255 (1955). Some of t h e enzyme used was a gift from D r . I,. K. Ramachandran. (31) R. I,. Hill and E. L. Smith, unpublished manuscript (personal communication). ( 3 2 ) T h e molecular weight of MSH assumed in this paper is t h e minimal value obtained from t h e amino acid analyses-2135. At all times weights of materials were corrected t o t h e true peptide weight (based on N Content).
Dec. 20, 1957
ISOLATION OF A MELANOCYTE-STIMULATING HORMONE:
Glycine
Time. Hours,
Fig. 4.-Liberation of amino acids from bovine MSH (left) and porcine 0-MSH (right) by leucine aminopeptidase.
was found; from C h 3 B , diDNP-lysine; 'Ch-4B, DNPglycine and e D N P lysine; and from Ch-5B, DSP-aspartic acid. T h e yield of DNP-glycine from Ch-4B was approximately 25y0 of the theoretical value.38 The N-terminal amino acid sequences of peptides Ch-3B and Ch-4B were investigated by allowing one-quarter of the eluate to react with DFPtreated leucine aminopeptidase," with the tris buffer employed above. \Vith peptide Ch-3B and 37.5 pg. of the enzyme (E:S = 1:l X l o 4 ) ,the peptide was completely hy-
6397
I
Gh-SB
strate molar ratio 1: 5 5 ) in 1 ml. of 1% NaHC03. At designated intervals aliquots were removed, were subjected to dinitrophenylation, and were otherwise treated in the same manner as the aliquots from the leucine aminopeptidase experiment reported above. After 24 hours only one free amino acid, aspartic acid, was found to be present, in a yield of 0.2-0.25 mole per mole. Chymotryptic Digestion.-Approximately 6 p M of the hormone in aqueous solution (2 ml.) was incubated with 0.75 mg. of chymotrypsin (Armour). The solution was adjusted with S H 4 0 H t o p H 8.3, a t which p H the solutio; was still slightly turbid. Digestion was carried out a t 40 for 18 hours, a t the end of which time the solution was clear. T h e products of digestion were separated by zone electrophoresis on paper in a r-collidine-acetic acid-water buffer33 of p H 7.0. The details of this procedure were exactly the same as those reported for the enzymic digests of @-MSH.6 At the termination of the electrophoretic run, guide-strips were cut from the dried paper and sprayed with ninhydrin. T h e pattern obtained was similar to that seen with P-MSH,B with the exception that in the digestion products from the bovine hormone the only band that migrated toward the anode (Ch-5B) had a lesser mobility than did the analogous band derived from the porcine hormone (Fig. 5 ) . The mobilities of the three major bands moving toward the cathode were apparently the same in the digests of the bovine and porcine hormones. Spray reagents specific for the various amino acids revealed in which bands arginine, 3 4 tyrosine, 35 histidine36and tryptophan3' were located; in each case the results were the same as those observed previously with the digest of P-MSH. Each band was demarcated on the paper, with the guidestrips serving as reference, and was subsequently eluted with 1 ,V NH,OH as previously described.6 Amino Acid Analyses of the Peptides Obtained from the Chymotryptic Digest.-From each eluate one-tenth of the solution was evaporated to dryness and hydrolyzed in 5.7 iV HCl for 24 hours a t 110". The hydrolyzate was then taken for duplicate quantitative amino acid analysis by the method of Levy.lG T h e results are presented in Table 11. T h e empirical formulas for each of the peptides, as derived from the amino acid analyses, are as follows: Ch-PB (-4rg,Try); Ch-3B (Lys,Met,Glu,His,Phe); Ch-4B (Gly, Ser ,Proz,Lys,Asp); and Ch-5B (Asp,Ser ,Gly ,Pro,Tyr) . N-Terminal Amino Acids and Amino Acid Sequences of the Peptides Obtained from the Chymotryptic Digest.-From each eluate two-tenths of the solution was allowed to react with FDA-B (2%) for 2 hours. T h e DSP-peptides were extracted from the acidified reaction mixture into ether or ethyl acetate, taken to dryness, and hydrolyzed in 5.7 N HCl for 16 hours a t 110". Each peptide possessed only one N-terminal amino acid. From Ch-2B, DNP-arginine
intervals after the addition of the enzyme, it was possible to determine . ~~~~~~. (Table 111) h a ,t lv~ t_ sine was the S-terminal amino acid, and that, following its release, methionine I was released b i the enzymic action. Subsequently glutamic acid, histidine and phenylalanine all appeared almost simultaneously. In contrast to its effect on Ch3B, the enzyme released only one amino acid, glycine, from Ch-4B even after reaction for 24 hours a t 40" a t an enzyme t o substrate molar ratio of 1:625. .After 2 hours of digestion all of the glycine had appeared, and no other amino acid appeared subsequently. C-Terminal Amino Acid Sequences of the Peptides Obtained from the Chymotryptic Digest.-From each eluate two-tenths of the solution was allowed to react Fig. 5.-Resolution of a for 2.1 hours a t room tempera- chymotryptic digest of boture with 0.6 mg. of DFP- virle MSH by electrotreated carboxypeptidase. The solution was then allowed Phoresis on paper. System, to react with F D K B ( 3 7 , ) for r-collidine-acetic acid-water 3 hours, and the D X P amino at PH 7 . 0 ; 11 yolts/cm, for 6 acids formed were identified bv the Levr Drocedure.16 The hours' results arepiesented in Table-IV. I t will be seen that, aside from what is observed with the dipeptide Ch-2B, only from peptide Ch-3B was more than one amino acid obtained; the results indicate that in this particular peptide the C-terminal sequence is . . .His.Phe. Stepwise Degradation of Peptide Ch-4B.-In order to determine the sequence of amino acids in Ch-4B, the final quarter of the eluate containing Ch-4B was taken for stepwise degradation by a m o d i f i c a t i ~ nof~ ~the phenyl isothiocyanate reaction. The reaction cycle was repeated for four steps. The phenylthiohydantoins formed were identified by chromatography of an aliquot on paper in the heptanepyridine system, and further identified by regeneration of the parent amino acid from another aliquot (vide supra). In this manner the sequence gly.ser.pro.pro was formulated. Finally, the aqueous solution remaining after extraction of the fourth phenylthiohydantoin was taken to dryness and then allowed to react with F D S B . After hydrolysis of the
(33) I. M. Lockhart and E. P. Abraham, Biochem. J., 88, 633 (1951). (34) J . P. Jepson and I. Smith, h'atrrrc, 172, 1100 (1953). (35) R. Acher and C. Cracker, Biochim. Biophys. Actn, 9, 704 (1952). (36) F. Sanger and H. T u p p y , Biochem. J . , 49, 463 (1951). (37) I. Smith, Nature, 171, 43 (1953).
(38) No values for the recoveries of theother D N P K - t e r m i n a l a m i n o acids are given because the amounts of D N P peptides available were too small to weigh. T h e recovery of DNP-glycine was obtained b y comparing it with the recovery of e-DNP-lysine, also obtained from the same peptide. (39) H. Fraenkel-Conrat and J. I. Harris, THISJOURNAL, 76, 6058 (1954).
Origin Gh-4B
~~
~~
Gh-3B
Ch-2B
L Ca
thode
.
6398
IRVING I. GESCIIWIND, CHOHHao LI
AND
LIVIOBXRNAFI
Vol. 79
porcine crude concentrate. Thus, it may be noticed (Fig. 1) that the material purified by a single electrophoretic run possesses a greater cathodic mobility than does the crude material (inigraPeptide, +I4 Amino acid Ch-PB Ch-3B Ch-lB Ch-5B tion of 17 and 23 cm. for the crude and purified Glutamic acid 0.241 ( 1 . 0 4 1 ~ ~ materials, respectively). Futhermore, the pattern Aspartic acid 0 2IX (O,!J7) 0 2 i 3 ( 1 . 0 2 ) for the purified material appears considerably Serine ,274 (1.00) ,234 (0 o i i ) (note in the original electrophoretic Glycine .xu(1 07) . ~ 0 2 ~ i . o sharpened ~) Proline ,;lj(j (2.06) ,218 ( 0 . 0 3 ) pattern the broad cut from which the purified Methionine 0 21hj (1 1.)) material was obtained). Both effects can be exPhenylalanine 231; ( 1 n 2 j plained on the basis of an interaction between the Tyrosine ,251 (0.91) hormone and some other component(s) present in Lysine 202 ( 0 . S X ) , 2 > c I0.ljl) Histidine 210 ( n . ! i i j the crude material, although i t is possible that Arginine 0,349 some complete and irreversible alteration in the Tryptophan 0.05s hormone occurs in the course of the first isolation * -1nalysis on :in aliquot equal t o 1 / 2 0 the total amount of from starch. peptide; average of 2 such analpes. Values in parentheses are the molar ratios for each uniiiio acid in the peptide. The amino acid and N-terminal group analyses of material which has been twice submitted to zonc TABLE I11 electrophoresis, and of the product obtained from THE RELE.4SE O F AMINOACIDS FROM PEPTIDE C H - ~ BBY the countercurrent procedure, are identical. Iri THE ACTIOS OF LEUCISEAMIXOPEPTIDASE' this respect the bovine material purified by zone Timeb electrophoresis differs from the porcine hormone at Amino acid 0 3 10 15 20 GO rhl PPI1 rhf fin1 rR.1 /A131 the comparable stage of the isolation procedure. Lysine 0.005 0.038 0.053 0.060 0.062 0.064 In the latter case, a number of contaminating Methiuniiie , . , ,023 ,042 ,050 ,054 ,064 amino acids not present in the final product are Glutamic present in the electrophoretic p r ~ d u c t . ~Despite acid , . , ,00i . 026 ,041 ,047 . OB4 the apparent purity of the bovine hormone a t this Histidine , , , ,010 ,029 ,040 ,043 ,064 point in the isolation scheme, countercurrent disPhenyltribution has resulted in the resolution of three conialniiitie , , . ,008 . 0 2 i ,042 ,046 ,064 ponents of differing partition coefficients. HowReaction carried out a t room temperature; Enzyme: ever, the quantitative amino acid analysis of each substrate molar ratio - 1: 1 X IO4. Minutes after addi- of these components is identical. The entire pattion of enzyme. tern is highly reminiscent of that observed with the porcine hormone, with the exception that the K TABLE IV C-TERXISALAMINOACIDS RELEASEDBY CARBOXYPEP?.I-values for the bovine components are generally, and reproducibly, l,iyOlower than those obtained DASE FROM THE PEPTIDES OBTAIXED FROM CHYMOTRYPTIC for the components of porcine origin. Because of DIGESTOF BOVINEp-MSH the lack of sufficient quantities of purified porcine Peptide Amino acids and bovine hormones and of the large number of Ch-2B Tryptophan (0.875 p X ) o Ch-3B Phenylalanine (OS71 p J J ) ; histidine (0.126 $11) transfers that would be required to separate coniponents with K values of 0.61(porcine) ancl 0.53 Ch-4B Aspartic acid (0.432 p X ) (bovine) in this system, no attempt has been made Ch-5B Tyrosine (0.884 ptdP) to see whether the K values obtained for the bulk a DXP-arginine was obtained from the water-soluble of the hormone from the two species are indeed fraction. really different and whether the analogous comD S P product and extraction of an>- ether-soluble compoponents from the different species may be resolved. nents, paper chromatography in 1 IIP Xa2HP0124 revealed It is of some interest a t this point to compare the presence of a yellow spot with the same Ri as ol-DXPthe isolation procedure described above with that lysine (both a-DSP- and e-DSP-lysine were run as controls on the same sheet). From the ether-soluble fraction only used by Benfey and Purvislb for their bovine NSH dinitrophenol was obtained in any yield. preparation. The starting material used by these investigators was obtained by the methods deDiscussion The procedure employed for the isolation of veloped by Stehle,-'Oin which dilute acetic acid exbovine p-AfSH is patterned after the procedure tracts are treated, in order, with ethanol, methanol developed for ,&LISH of porcine origin,?from which and ethyl acetate, and are then further fractionated it differs only in minor details. Once again, the by precipitation with picric acid. U7hen such most important single step in the isolation proce- material was submitted to countercurrent distridure appears to be the selective adsorption of the bution in the same system that we have employed, hormone on oxycellulose. As already has been namely, 2-butanol-aqueous 0.5c/; trichloroacetic shown, this one procedure permits approximately acid, a K d u e of T was obtained, in contrast with a -Ofold purification of bovine melanocyte-stimu- the value found in this paper of 0.33 for the box inc hormone. The system actually used by 13e1ifcy lating activity. and Purvis for the purification of their bovine prepThe material eluted from oxycellulose has been submitted to zone electrophoresis on starch. In arcition was ?-butanol-aqueous 0. li,a trichlorothe course of this procedure, certain reproducible acetic acid, in which the melancrcyte-stiinulating effects have been observed with the bovine hor( 4 0 ) i n , K 1, Stchle R L L L a u a d & d , 3 , 108 (1(414), (ti) i i i z t J mone which have not been previously seen with a I ' i r u Y i t L U ' o l 8 , 435 (19x3) TABLE I1
AMINOACID ANALYSES"OF THE PEPTIDES OBTAINED FROM CHYMOTRYPTIC DIGESTOF BOVINE8-MSH
Q
Dec. 20, 1957
ISOLATION O F A MELANOCYTE-STIMUL.4TING
activity had a K value of 0.74-0.85.1b The preparation of Benfey and Purvis also varies considerably from the bovine preparation reported herein with respect to its amino acid composition. For example, i t contains large quantities of alanine, leucine and valine (3, 3 and 5 residues, respectively, per mole consisting of a total of 47 residues). In addition, cystine, isoleucine and threonine are present. None of these six amino acids is present in the bovine ,B-MSH, whose analysis is given in Table I . There are, as well, great differences in the quantities of those amino acids that are common to both preparations. In the light of the marked differences in K values and amino acid composition between the two, these preparations would appear to be unrelated. Although there is apparently no connection between the bovine preparation of Benfey and Purvis and the porcine a-MSH of Lerner and Lee2S1Othe isolation of two unrelated melanocyte-stimulating hormones from bovine posterior lobes* is reminiscent of the discovery of a- and ,BMSH in porcine posterior lobes. The amino acid analyses recorded in Table I indicate that the composition of the bovine hormone differs from that of P-MSH in that the former contains one more residue of serine and one less residue of glutamic acid than does the porcine hormone. The substitution of a neutral residue for an acidic one is sufficient to explain the more basic isoelectric point of the bovine preparation. There is illustrated in Fig. 6 a comparison of the mobilities of the bovine hormone and porcine P-&ISH in a pyridineacetic acid buffer, @H4.9. To one trough, 1 mg. each of the bovine and porcine hormones was added a t the origin; to a parallel trough 1 mg. of the bovine preparation alone was added. The separation achieved, effected during a 24 hour run a t 5 v./cm., unequivocally illustrates the greater cathodic mobility of the bovine preparation a t this pH. Knowledge of the amino acid composition also permits the calculation of the isoionic point of the hormone, p H 7 . 2 , which compares favorably with the experimentally determined isoelectric point, 7.0. The isoionic point calculated from the data of Benfey and Purvis is approximately 6.8, if a complete absence of amide groups is assumed. Waring and Landgrebe’ reported that in crude bovine posterior lobe extracts the melanocyte-stimulating activity had an isoelectric point of approximately 4.1. Ketterer and Kirk41 have stated that the “melanophore-expanding activity of ox posterior lobe powder is complex,” but that the “center of gravity for all. . .activity” was in the region of 4.S-4.9. I n the light of the findings reported above, which demonstrated decreased cathodic mobilities in crude LUSH preparations, one may suspect that interactions of the type suggested above may satisfactorily explain the observations of IYaring and Landgrebe and Ketterer and Kirk. At each pH utilized for the determination of the isoelectric point, as well as in pyridine-acetic acid buffers of pH 4.9 and 6.4, the material purified by the countercurrent procedure appeared to be homogeneous. The determination of the mobility a t pH 8.4 was performed in a Verona1 buffer of 0.1 (41) B Ketterer a n d R L Klrk, J Eitdocvinol, 11, 19 (1954).
HORMONE
P
09001
$ 1
E
6399
/
0750
Porcine
n hI
Segment Number (cm.), Fig. 6.-Patterns obtained from zone electrophoresis on starch of a purified bovine MSH ( A ) and of a mixture of bovine MSH and porcine P-MSH (0). Origin for both runs is designated by the arrow; 40 cm. trough, 0.1 M pyridine-acetic acid buffer a t PH 4.9, 200 volts, 24 hours.
ionic strength, and in this buffer the material also appeared to be homogeneous. However, in a borate buffer of similar 9 H (in the trough) and ionic strength, the spread of the material is considerable. This specific effect of borate buffer has been previously observed a t times in this Laboratory by I. D. Raacke, and apparently is due to some interaction between material, buffer and starch. The material obtained by the recorded isolation procedure was considered sufficiently pure to permit an investigation of the structure of the hormone. In the light of the known structure of porcine 6-MSH and the concept that the heptapeptide core5S6common to ,B-MSH and the corticotropins would probably remain inviolate, it was assumed that a simple substitution of serine for glutamic acid occurred a t the second residue in the ,B-MSH structure to produce the bovine hormone. Preliminary support for this idea was forthcoming from the N- and C-terminal group analyses. Further support was obtained from the degradation of the N-terminal sequence with phenyl isothiocyanate which revealed a serine residue next to the Nterminal aspartic acid. However, a t this point a fair amount of glycine was usually evident also. Apparently under the acidic conditions needed for the cyclization of the terminal aspartic phenylthiocarbamate, the particularly labile seryl-glycine bond is partially broken. No such effect was observed with the comparable glutamyl-glycine bond of the porcine P-ICISH. In all, the phenyl isothiocyanate method was employed for the elucidation of a sequence of seven consecutive amino acids. At each step the quantities of phenylthiohydantoins formed were determined spectrophotometrially,^^ but, as reported previously,6 the amount of identifiable phenylthiohydantoins evident on the chromatograms was much less than that indicated by spectrophotometry. Recently, Elliott and Peart42 reported similar findings in the course of (4’2) U. F. Elliott and W. S. Peart, Biocheiiz. J . , 65, ‘2.28 (1957).
IRVING I. GESCHWIND, CHOHHAOLI
6400
AND
LIVIOBARNAFI
VOl. 79
the degradation of hypertensin. These findings in terms of mobility, N- and C-terminal groups, and may be explained in terms of E d m a n ’ ~recent ~~ amino acid analysis, with peptides obtained from formulation of the mechanism of the phenyl iso- the chymotryptic digestion of porcine P-l\ISH.6 thiocyanate degradation. On the other hand, the only peptide which miFor confirmation of the N-terminal sequence, grated toward the anode during the electrophoretic leucine aminopeptidase30 was employed. Al- run (Ch-5B) did so a t a lesser mobility than that though the enzyme promoted the release of aspar- observed with peptide Ch-5 from the porcine hortic acid and serine, only 0.43 mole of the N-ter- mone hydrolysate. Furthermore, though the Nminal aspartic acid was released from the bovine and C-terminal groups of Ch-5B were identical hormone in 24 hours under the conditions chosen with those found for the comparable porcine pep(Fig. 4a). Under the same conditions, 1.0 mole of tide, its amino acid composition differed in one aspartic acid was released from porcine P-hlSH more serine and one less glutamic acid, thus ex(Fig. 4b). Thus the rate of release of identical N- plaining the decreased mobility of the bovine pepterminal amino acids from two different substrates tide. was markedly modified by the nature, in each inWith respect to the analyses of the peptides obstance, of the adjacent amino acid. It may be fur- tained by chymotryptic digestion, reference should ther noted that in both cases little or none of the be made to a number of findings. First, from the third amino acid, glycine, appeared. Although amino acid analyses it may be calculated that apglycine is the amino acid most slowly released by proximately 5 pmoles of each peptide was obtained. the failure of the enzyme to release glycine This value represents a recovery of greater than from both hormones is believed to be due to the 80% of theory, without the application of any corglycyl-proline linkage which is involved. I t has rection for losses encountered through guide strips, been our general experience that amino acids in incomplete elution, etc. Second, the amount of peptide linkage with the imino group of proline are C-terminal aspartic acid released by carboxynot released. Thus, it will also be recalled that peptidase from Ch-4B is greater than that obtained LAP hydrolyzed the release of only the terminal from the intact hormone (Table IV). A similar glycine of peptide Ch-4B. The initial sequence finding was made for the comparable porcine pepfound for this peptide was Gly4er.Pro. . . . Ilere, tide and hormone. Third, in the course of the actoo, a bond involving the imino group of proline tion of leucine aminopeptidase on Ch-3B, after was not split.45 On the other hand, Elliott and lysine and methionine are released, the following Peart42have found that LA?? (enzyme to substrate three amino acids, ;.e., glutamic acid, histidine and ratio unrecorded) promoted the release of proline, phenylalanine, are released almost simultaneously and the amino acids which precede it, from the hy- (see Table 111). This may be explained as due to pertensin molecule. Consequently, we have re- the rate-limiting release of glutamic acid, followed ported our experiences with this enzyme without by rapid hydrolysis of the bond between histidine and phenylalanine. This effect is the same as that attempting to generalize our findings. Since submission of the intact hormone to step- observed by White48during the hydrolysis by LAP wise chemical and enzymic degradative procedures of Corticotropin A, in which the same sequence permitted the identification of only seven amino . . .Glu.His.Phe. . . occurs. Finally, the release of acids a t the N-terminus along with the C-terminal tryptophan from Ch-2B by carboxypeptidase aspartic acid, proteolysis with chymotrypsin was supplements the data from microbiological assay employed to cleave the hormone into the four com- presented in Table I by establishing the L-configuraponent peptides which, on the basis of the aro- tion for this amino acid; thus, it has been found matic amino acid content of the hormone, are that all the amino acids in this hormone are of this theoretically possible. Of the four peptides re- configuration. The large number of amino acids solved by paper electrophoresis, three were identical released by LAP and by carboxypeptidase from the intact molecule and from the peptides obtained (43) P. E d m a n , Acla C h e m . Scand., 10, 761 (1950). from the chymotryptic digest is consistent with (44) E. L. Smith and D. H. Spackman, J . B i d . Chewt., 213, 271 this finding. (1955). (45) Another example t h a t may be cited is the action of L4P The only peptide whose structure was not readily (unpublished observations) on Tr-2,3, the C-terminal peptide derived determined either from the analytical data or from from a tryptic hydrolysate of a-corticotropin.46 T h e initial sequence the results of the action of LAP was Ch-4B. As of this peptide, as derived by t h e phenyl isothiocyanate procedure, is Val.Tyr.Pro.. . T h e enzyme releases valine from this peptide mentioned above, LAP promoted the release o f very rapidly, a n d even a t enzyme to subtrate ratios of 1 :30 for 24 hours only the N-terminal glycine, and so, in order to no tyrosine is released, despite the fact t h a t this amino acid is a very determine the sequence of amino acids in this pepfavorable substrate for t h e enzyme.44 Once valine and tyrosine are tide, the phenyl isothiocyanate procedure was einremoved from this peptide by the phenyl isothiocyanate procedure, L A P releases proline readily since L A P may split off proline with a ployed. After the first four amino acids had bemi f v e e imino group, A final example derived from work in this Laboraremoved, the remaining dipeptide (c-phenylthiotory has been obtained with lactogenic hormone. T h e initial sequence carbamyl-lysyl-aspartate) was dinitrophenylated 47 Employing t h e performic acidof this hormone is T1ir.Pro.Val to give a-DNP-ephenylthiocarbamyl-lysyl-aspuroxidized protein as a substrate, J. Cummins (personal communication) has found t h a t L A P does not promote t h e release of a n y amino acid. tate, which upon hydrolysis gave rise to a-DNPOne m a y also refer t o t h e findings of Smith and Spackmanaa t h a t t h e lysine and aspartic acid. dipeptide glycyl-proline is not attacked b y LAP, although prolylIVith the elucidation of the structure of peptide glycine is split a t a rate 2.77, of leucyl-leucine. Ch-4B, the sequence of amino acids in the hormone (40) C. H. I,i, I. I. Geschwind, R . D. Cole, I. D. Raacke, J. I. Harris and J. S . Dixon, S o l u r e , 176, 687 (1955). could be formulated (Table V). It may be seen ,46
(47) R. D. Cole, I. I. Geschwind and C. H. Li, J. B i d . C h e m . , 244, 399 (1957).
(48) W. F. White, THISJOURNAL, 77, 4691 (1955).
Dec. 20, 1957
SOME
7-METHYLPURINES AS
POTENTIAL PURINE
ANTAGONISTS
6401
TABLE V THESEQUENCE Method
Substrate
OF
AMINOACIDS I N BOVINE8-MSH (“SERYL-O-MSH”)
Product
PDNB H“ LAP H H PITC Chymotrypsin H Ch-5Bb H Ch-3Bb Chymotrypsin LAP Ch-3B Chymotrypsin H Ch-2Bb Ch-4Bb H Chymotrypsin LAP Ch-4B PITC Ch-4B Carboxypeptidase H Complete sequence:
Sequence
Asp. Asp.Ser.
Asp.Ser.Gly.Pro.Tyr.Lys.Met Asp(Ser,Gly,Pro)Tyr Lys(Met,Glu)His.Phe Lys.Met( Glu,His,Phe) Arg.Try Gly(Ser,Pro,Pro,Lys)Asp Gly.
Gly.Ser.Pro.Pro(Lys,Asp) ASP
Asp.Ser.Gly.Pro.Tyr.Lys.Met.Glu.His.Phe.Arg.Try.Gly.Ser.Pro.Pro.Lys.Asp
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 H = hormone. The N-terminal amino acid was obtained by the F D N B procedure, and the C-terminal amino acid or sequence by the carboxypeptidase method. a
that of all the peptides found in the chymotryptic digest, only the dipeptide Arg.Try (Ch-2B) must have its position assigned by elimination, for the contiguity of Ch-5B and Ch-3B is assured by the results of the phenyl isothiocyanate procedure when applied to the intact hormone, and the position of Ch-4B must be C-terminal since i t alone possesses a C-terminal aspartic acid. As suspected, the sequence of amino acids in the bovine hormone differs from that found for porcine pMSH516only a t the second residue. I t is therefore suggested that porcine and bovine /3-MSH be henceforth called “glutamyl-p-MSH” and “seryl-@-MSH,” respectively. The interchange of a neutral amino acid for an acidic one has not been previously encountered among biologically active peptides derived from different species, where experience to date has demonstrated interchanges of the neutralneutral or basic-basic types. Finally, in a series of 8 separate comparative
assaysI4 the porcine hormone has been found to be approximately 2.5 times as active as the bovine hormone. Thus, the interchange of amino acids a t the position of the second residue has resulted in a modification of the hormonal activity. Seryl-pMSH is comparable to the porcine hormone, however, in that upon being heated in 0.1 N NaOH for 15 minutes, “protection”7 occurs with little or no “potentiation,”7 in confirmation of the findings of Landgrebe and Mitchells with an oxycelluloseeluate of bovine posterior lobe powder. Acknowledgments.-The authors wish to acknowledge the able technical assistance of Miss Winifred Lilly. This work has been supported in part by grants from the National Institutes of Health of the United States Public Health Service (Grant No. G-2907) and the Albert and Mary Lasker Foundation. BERKELEY 4, CALIFORNIA
[CONTRIBUTION FROM THE DEPARTMENT OF CHEMISTRY, NEW MEXICOHIGHLANDS UNIVERSITY]
Potential Purine Antagonists. VIII. The Preparation of Some 7-Methylpurinesl BY RAJ NANDAN PRASAD2 AND ROLAND K. ROBINS3 RECEIVED MAY13, 1957
A new method of synthesis of 7-methyladenine (111) and 7-methylhypoxanthine (V) has been accomplished from 4amino-1-methyl-5-imidazolecarbonitrile(11) and 4-amino-1-methyl-5-imidazolecarboxamide(IV), respectively. The chlorination of V yielded 6-chloro-7-methylpurine (VI), Various new 7-methyl-6-substituted purines have been prepared from VI. 2,6-Dichloro-7-methylpurine (XIV) has been prepared from 7-methylxanthine (XIII). Several new 2,6-disubstituted-7-methylpurine derivatives have been prepared from XIV.
I n the course of a general program for the synthesis of potential purine antagonists as antitumor agents, it was discovered that 6-chloro-9methylpurine4 showed definite activity. It there(1) (a) This investigation was supported in part b y research grant C-2845 from t h e National Cancer Institute of t h e National Institutes of Health, Public Health Service. (b) Presented in part before the Division of Medicinal Chemistry a t the 131st Meeting of t h e American Chemical Society, April, 1957, a t Miami, Florida. (2) On leave-of-absence from Chemistry Department, B. N. College, P a t n a University, India. (3) Department of Chemistry, Arizona S t a t e College, Tempe, Arizona. (4) R. K. Robins a n d H . H. Lin, TIXISJOURNAL, 79, 490 (1957).
fore seemed desirable to prepare G-chloro-7methylpurine (VI) as a candidate anti-tumor agent. Since preliminary methylation studies of 6-chloropurine gave a difficultly separable mixture of the 7-methyl and 9-methyl isomers, the possible preparation of 6-chloro-7-methylpurine (VI) from 7methylhypoxanthine (V) was investigated. Earlier methods of preparation of 7-methylhypoxanthine6 proved unsatisfactory for large scale preparation. This prompted investigation of a new route t o the preparation of 7-methylhypoxanthine (V) . (5) (a) E. Fischer, B e y . , SO, 2400 (1897); (b) 81, 104 (1898).
