3540
COMMUNICATIOXS TO THE EDITOR
A
A
dL
Fig. 2.-Electrophoretic patterns of Armour bovine serum albumin in various buffer solvents a t PH 4.0: A, 0.002 ionic strength NaAc-HAC 0.198 ionic strength 0.195 XaC1; C, 0.20 SaXcNaC1; B, 0.005 Sa.ic-H.lc H.lc. Protein migrated toward the cathode. The product of field strength and time is approximately the same in all these experiments.
+
+
aries, ;.e., those corresponding to more positively charged components, a t the expense of slower moving ones. That this represents an effect of increasing Ac- andlor HAC rather than decreasing CI- is indicated by the observation that the patterns obtained in 0.01 ionic strength NaAc-HAC, PH 4.7, are essentially the same as those obtained 0.015 NaC1 (the chief differin 0.01 NaAc-HAC ence being that the more rapidly moving boundary in the rising pattern was not resolved into two peaks at the lower ionic strength) .z While binding of acetate has been either demonstrated or postulated in a number of protein systems, this is the first demonstration that such binding has a major effect on protein structure. (Such binding does not, however, change the sedimentation constant.) Recognition of this effect increases our understanding of the structure of BSX. The non-enantiographic nature of the electrophoretic patterns of this protein has been interpreted by Longsworth and Jacobsen3 to indicate continual readjustment of an equilibrium across the moving boundaries. But other workers assume that BSA is a simple mixture of proteins.? A solution of BSA may well contain molecules which
+
12) T h e possibilities that the complex patterns arise from electrical anomalies or a- i i result of poor buffering action, h a r e also been eliminated.
( 4 ) I< .\ A l l , c r l y , i h i t i , 5 3 , 1 1 4 (l9401.
VOl. 7s
differ ~hemically.~However, we conclude from our experiments that BSA is indeed an equilibrium mixture of electrophoretic components ; that any chemically different molecules participate in this equilibrium ; and that the equilibrium composition depends upon the composition of the supporting medium, particularly the acetate concentration.6 Two independent observations support these conclusions, first, if a t the end of the usual movingboundary electrophoresis experiment of duration t, the polarity of the electrodes is reversed and the experiment continued for another length of time f, then the final boundaries which have now returned to their initial positions are not single symmetrical ones as expected of a simple mixture of proteins but are complex showing as many as three distinct, although poorly resolved peaks. This suggests that an equilibrium was being continuously readjusted as the boundaries returned to their initial positions. Experiments of this type were carried out at several ionic strengths and concentrations of KaAc-HAC. The second piece of evidence is afforded by small differences in the ultraviolet absorption spectra of BS-4 in 0.1 NaAqc-Hiic and in 0.002 hTa.4c-HAc f 0.098 NaCl. These differences are confirmed by a differential spectrum whigh shows two main bands a t 2350-2-100 and 2900A.. and indicate that binding of Ac- and.'or HAC by BSA produces structural changes in the protein, perhaps due to making or breaking of hydrogen bonds. The effect described above has also been observed with ovalbumin and y-globulin and accounts for the complex electrophoretic patterns observed' for these proteins in certain buffers. Experiments are in progress to elucidate the nature of the effect of various buffer constituents (such as homologous series of aliphatic acids) on protein structure. ( 5 ) H . A. Sober. F. J. Gutter, 11.h i . Wyckoff and E . A . Peterscrn, TXISJ O U R N A L , 78, 756 (19.56), (6) The electrophoretic patterns of BSA are also dependent on whether the buffer cation is N a + or I; +, pH, and to some extent ionic
strength ( 7 ) .4.Saifer and H . Corey, J . Biui. Chem., 217, 23 (19Z.5): Abstracts of Papers. 138th Meeting American Chemical Society, hlinneapolis, h l i n n . , Sept. 11-16, 1935, p. 40-C.
CONTRIBUTION S o . 43 FROM THE ROBERT A . PHELPS DEPARTMENT O F BIOPHYSICS J O H NR. CASX FLORESCE R . SABIN LABORATORIES UNIVERSITY OF COLORADO MEDICAL CENTER DESVER,COLORADO RECEIVED APRIL2, 1956 SYNTHESIS AND ANABOLIC ACTIVITY OF 4-SUBSTITUTED TESTOSTERONE ANALOGS
SLY. \Ve wish to mhke a preliminary announcement of the synthesis and anabolic activity of 4-substituted testosterone analogs. Treatment of testosterone with alkaline hydrogen peroxide gave a mixture of the (I- and a-epoxides, which could be separated by crystallization : 4~,5-oxidoetiocholane-l7(I-ol-3-one, m.p. 156-l57Oj ( c r \ l ~ f145' and 3a,3-oxidoandrostane-l7(I-ol-3one, 1ti.p. 14i--13S', ( a ) -~3 3 O [Calcd. for C I ~ H ~ S (II C , 7-1.96; H , 9.27. Found: C . 74.86;H , 9.3.5 and C 7-5.16; H. !).?SI;acetates 111 11. 155-lsj7', ( d i ) +1:30° (CHCI,) ant1 m.p. 172- IT:%', ( N ) D
July 20, 1956
COMMUNICATIONS TO THE EDITOR
3541
Most of the described compounds showed re-62' (CHC13) [Calcd. for C21H~004c, 72.80; H, 8.73. Found C, 72.74; H, 8.88 and C, 72.86; markable anabolic activity and low androgenic H, 8.741. The assignment of configuration was effect. The following anabolic androgenic ratios, based on rotatory power, taking as reference 4P,5- determined according to Hershberger, Shipley and oxido coprostane-3-one, whose structure has been Meyer,4a t 500 y daily doses, were obtained: demonstrated. Testosterone propionate 0.28 4-Chlorotestosterone acetate .88 Rearrangement of the p- or a-epoxide with boron 4-Hydroxytestosterone acetate .61 trifluoride in benzene yielded A4-androstene-4,17p4-Fluorotestosterone acetate .35 diol-3-one, m.p. 222 -223', strong ferric chloride .72 19-Nortestosterone cyclopentylpropionate test, while acetic acid containing 2% of H2S04 gave 4-Chloro-19-nortestosterone cyclopentylpropionate 1 .82 4-Chloro-19-nortestosterone acetate 1.60 A4-androstene-4,17-P-diol-3-one 17-acetate (4-hydroxytestosterone 17-acetate) m.p. 194- 196'., We are indebted t o Dr. Sala and G. Baldratti of ( a ) +~8 3 O , Xmax 277 mp, E = 12,100 [Calcd. for our Department of Pharmacology for the data on CalHaoO4:C, 72.80; H, 8.73; Found: C, 72.40; anabolic and androgenic potency and to Dr. F. H, 8.801. Canal for the microanalyses. The latter compound formed with O-phenyleneL. G. Hershberger, E. G. Shipley and R. K. Meyer, Proc. SOC. diamine a quinoxaline derivative and with acetic E x(4) P . B i d . M e d . , 83, 176 (1953). anhydride in pyridine a 4,17-diacetate, m.p. LABORATORI RICERCHE FARMITALIA 170-172', ( a ) +l05', ~ Xmax 246 mp, E = 15,500 SEZIONE CHIMICA B. CAMERINO [Calcd. for C23H3205:C, 71.10; H, 8.30. Found: VIA DEI GRACCHI B. PATELLI 35 MILASO,ITALY A. VERCELLONE C, 71.03; H, 8.251. RECEIVED MAY29, 1956 The 4~,5-oxidoetiocholane-17~-ol-3-one in acetone containing dil. rearranged to 2ahydroxytestosterone, m.p. 160-162', ( a ) ~ 166' A NEW PANCREATIC CARBOXYPEPTIDASE (reported2: m.p. 161-162', (LY)D +120'), while Sir: acetate gave the 4a,5-oxidoandrostane-l7~-ol-3-one A hitherto unreported carboxypeptidase has the normal product of epoxide fission androstanebeen found in the euglobulin fraction obtained from 4/3,50,17P-triol-3-one 17-acetate, m.p. 213-215' [Calcd. for C21H3205: C, 69.20; H, 8.85; Found: bovine pancreas after autolysis. It manifests a rapid activity in releasing lysine and arginine from C, 69.07; H, 9.181. Treatment of the 4/3,5-oxidoetiocholane-l7j% the carboxyl terminal position of the synthetic subol-3-one acetate with the requisite hydrogen halide strates, a-N-benzoylglycyl-L-lysine (BGL) and in acetic acid formed the 4-bromotestosterone ace- benzoylglycyl-L-arginine (BGA), which is not tate, m.p. 196-197', Xmax 261 mp, E = 11,600 ascribable to the known carboxypeptidase (CP). [Calcd. for C21H2903Br: C, 61.61; H, 7.14. The new enzyme is tentatively designated basic Found: C, 61.66; H, 7.181, the 4-chlorotestosterone carboxypeptidase (BCP) because of its apparent acetate (3), m.p. 228-230°, (a)D + 1 1 8 O , Xmax 255 requirement for a basic carboxyl terminal amino Like C P i t exists in fresh frozen pancreas as mp, E = 13,300 [Calcd. for C21H2803C1: C, 69.11; aacid. proenzyme which is liberated in active form upon H, 8.01; Found: C, 69.37; H, 8.221 and the 4fluorotestosterone acetate, m.p. 178-180', Amax 241 treatment of pancreas extract with trypsin. In Table I are shown the proteolytic coefficients mp, E = 11,670 [Calcd. for C21H2903F;C, 72.38; of some of the fractions obtained during a typical H , 8.38. Found: C, 72.50; H, 8.611. The reaction of the 4a,5-oxidoandrostane-l7~TABLE I ol-3-one acetate with hydrogen chloride in acetic PROTEOLYTIC COEFFICIENTS~ OF CARBOXYPEPTIDASE PREacid produced the 4P-chloroandrostane-5a,17PPARATIVE FRACTIOXS WITH DIFFERENT SUESTRATES diol-3-one-17-acetate, m.p. 202-204' [Calcd. for Substrateb CGPC BGLd BGAO CZIH~IO~CI: C, 65.86; H, 8.15. Found: C, 65.68; 1.12 0.85 0.64 H, 8.321 with on prolonged heating eliminated Crude pancreas extract water to give 4-chlorotestosterone acetate. 3 26 2.55 1.90 Crude euglobulin precipitate In the same way the a-epoxide acetate on reac- Barium hydroxide extract 3.62 3.00 2 . 8 5 tion with hydrogen bromide in acetic acid yielded Once crystallized carboxypeptidase 18.00 2.85 2.10 Three times crystallized carboxypepdirectly 4-bromotestosterone acetate. tidase 19.40 The p- and a- epoxides reverted t o testosterone on treatment with aqueous hydrogen iodide in Six times crystallized carboxypepti0 dase 19.10 chloroform or by heating with potassium iodide in acetic acid. The corresponding derivatives were a First order proteolytic coefficient defined as the first obtained also from 116-hydroxytestosterone, 17a- order velocity constant, Kt = log (a/a - x ) , per mg. proN/ml. The extent of hydrolysis was measured on 0.2methyltestosterone, 19-nortestosterone, progester- tein ml. samples by the alcoholic KOH titration procedure of one, 1lp-hydroxyprogesterone, desoxycorticoster- Grassman and Heyde, Z . physiol. Chem., 183, 32 (1929). Substrates employed a t 0.02d M concentration with 0.025 one, cortisone and other steroids and will be deM "tris" buffer PH 7.65 a t 25 . c Hofmann and Bergmann, scribed in forthcoming papers. J . BioZ. Chem., 134,225 (1940). Hofmann and Bergmann,
+
'
(1) P1. A . Plattner, H. Heusser and A. B. Rulkarni. Helo. Chim. Acto, 31, 1822 (1948). (2) F. Sondheimer, St. Kaufmann, J. Romo, H. Martinez and G. Rosenkranz, THISJOURNAL, 76, 4712 (1953).
(3) 4-Chlorotestosterone acetate was prepared also by direct chlorination of testosterone acetate.
ibid., 134, 225 (1940); Folk, Arch. Biochem. Biophys., in
press. Folk, unpublished data. f Five t o 10% hydrolysis in 24 hours at a substrate-enzyme molar ratio of 300 to 1. 0 Very little hydrolysis, only detectable by paper chromatographv after 24 hours a t a substrate-enzyme moldr ratio of 30 tu 1.
