These Enzymes Are 3#, 17,3-Hydroxysteroid Dehy

INHIBITION BY 2'-DEOXYADENOSINE OF ENZYME INDUCTION IN PSEUDOMONAS TESTOSTERONI BY ARNOLD D. WELCH

INSTITUT FUR THERAPEUTISCHE BIOCHEMIE, UNIVERSITAT FRANKFURT, AND THE DEPARTMENT OF PHARMACOLOGY, YALE UNIVERSITY SCHOOL OF MEDICINE Communicated by Carl F. Cori, September 29, 1965 Among various systems in which induction of new enzyme synthesis can be initiated, that developed by Talalay,'-3 in which the metabolism of a variety of steroids by Pseudomonas testosieroni is induced by either testosterone or certain other steroidal compounds, has been particularly rewarding. Recent studies of this system by Wacker et al.4' 5 have suggested that a complex may be formed between the inducer, in this case testosterone, and repressors, possibly proteins,4 thus leading to derepression and the formation of the appropriate ribonucleic acids and specific enzyme proteins. Although testosterone is metabolized rapidly and completely by Pseudomonas testosteroni, the inducible enzymes so far extensively studied catalyze only early steps in catabolism; these enzymes are 3#, 17,3-hydroxysteroid dehydrogenase, 3a-hydroxysteroid dehydrogenase, and A5-3-ketosteroid isomerase.1-5 Certain cytotoxic substances, e.g., 5-fluorouracil, mitomycin C, puromycin, p- fluorophenylalanine, and ethionine, appear to affect the induced formation of enzymes more than processes involved in the reproduction of the microorganism;5 however, no reports of the effects of either natural or unnatural nucleosides on the induction process have appeared. This paper is concerned with studies of the inhibition of induction of certain enzyme activities, particularly z5-3-ketosteroid isomerase, in Pseudomonas testosteroni, caused by the addition to the medium of 2'- deoxyadenosine, an effect much more marked than that observed with other nucleosides or bases tested, including 3'-deoxyadenosine and adenosine. Materials and Methods.-The procedures used differed but little from those employed by Marcus and Talalay2 I and by Wacker et al.4 I Pseudonwnas testosteroni (ATCC 11996; the majority of the experiments were carried out with a culture kindly provided by Prof. P. Talalay) was grown in a medium of pH 7 with a composition similar to that described by Wacker et al.,5 except that sodium lactate (5 gm per liter) was employed instead of glucose, and casamino acids and yeast extract (both obtained from the Difco Laboratories) were present at the level of 5 gm of each per liter. The organisms used as an inoculum were taken from an agar slant and grown overnight at 300 in a liquid medium (without steroid or amino acids); of this culture, 4.3 ml were added to each 100 ml of growth medium at the beginning of each experiment. Incubation of 5.0 ml of the inoculated medium was carried out for 7 hr in conical fermentation flasks (125 ml) at 30° in a gyrorotatory shaker (New Brunswick Scientific Co.) operated at 170 4 20 oscillations per min. In all cases, the inducer, M-androstene-3,17-dione, 1.25 mg dissolved in dimethylsulfoxide (DMS0), 0.05 ml, was added 2 hr after the initiation of incubation. In concentrations up to at least 5%, DMSO had no detectable effect on either the growth of the organism or the induction of enzyme activities. Adenine sulfate (Nutritional Biochemicals Corp. #4829), adenosine (Pabst #1408), and various lots of 2'-deoxyadenosine or its monohydrate (the most satisfactory material was that of Calbio- chem, #590893) were added to the flasks either as solids or as solutions in DMS0, of which not more than 0.1 ml was necessary per 5 ml of culture medium. 3'-Deoxyadenosine (generously provided by Dr. H. Klenow), di-sodium 2'-deoxyadenylate (5') (Calbiochem #30475), and the other 2'-deoxyribonucleosides tested (obtained from various sources) were added at appropriate times to the culture flasks as solid substances. After removal of the flasks from the incubator, the turbidity of dilutions of the cultures was 1359 Downloaded by guest on September 27, 2021 1360 BIOCHEMISTRY: A. D. WELCH PROC. N. A. S.

measured either in a Lange5 or Klett colorimeter (66 filter), or with a Zeiss spectrophotometer (660 mM). The medium and cells were transferred to centrifuge tubes of either stainless steel (20 ml) or glass (Corex no. 8441; 15 ml) and these were centrifuged for 10 min at -5° and -25,000 X g. After discarding the supernatant medium, the cells were resuspended in sodium- potassium phosphate buffer (5 ml; pH 7; 0.1 M) with the aid of a Vortex Jr. mixer. Subsequent to centrifugation (as before) and decantation of the supernatant fluid, 10 ml of acetone at -50 to -70° were added to the centrifuge tubes; after standing at room temperature for a few minutes, the frozen material was broken up with a spatula and stirred; the tubes were centrifuged (as before), the acetone was discarded, and the air-dried residues were incubated with phosphate buffer (5.0 ml; pH 7; 0.1 M) for 1 hr at 300 in the same centrifuge tubes (using silicone stoppers) placed horizontally in the gyrorotatory shaker. A final centrifugation for 10 min gave a clear supernatant fluid suitable for assay of enzyme activities. The assay for A5-3-ketosteroid isomerase activity is direct and involves measurement of the rate of increasing absorption at 248 my as A4-androstene-3,17-dione is formed from the corresponding As-compound;6 the latter, prepared by the method of Schmidt-Thom6 and Butenandt,7 was generously provided by either Dr. J. Drews, Institut fur Therapeutische Biochemie, Uni- versity of Frankfurt, or Prof. P. Talalay, The Johns Hopkins University. 3a-Hydroxysteroid dehydrogenase activity was measured in a sodium phosphate buffer (0.1 M) at pH 9, with other minor modifications of the method of Marcus and Talalay;2 3 the oxidation of androsterone was coupled to the conversion of NAD + to NADH, as measured at 340 mjg. Units of enzyme activity were calculated in the manner described by Wacker et al.,5 and corrections were made for minor depressions in the extent of bacterial growth, as indicated by measurements of turbidity. If major inhibitions of growth occurred, for example, with 7-hr turbidity values less than 60% of maximal, the results were not regarded as interpretable. Conversions of 2'-deoxyadenosine 5'-phosphate to the nucleoside were carried out by Dr. M. Chu and Dr. G. A. Fischer, using purified intestinal alkaline phosphatase (Worthington), in 0.013 M glycine buffer (pH 8.8) containing Mg2+ (5.5 X 10-3M). In terms of release of inorganic phosphate and ultraviolet absorption, conversion was 82-90% complete; in the tests of the products, which purposely were not isolated, correction for incomplete conversions was made, the pH values were adjusted to 7.1, and the materials were added to the culture media as vacuum-dried residues. Results.-In this report only results obtained with A4-androstene-3,17-dione as an inducer will be described; as reported by Wacker et al. , however, testosterone was not more active than this inducer of enzyme activity in Pseudomonas testosteroni. Although the findings presented are derived from the study of only one prototype- induced enzyme, A5-3-ketosteroid isomerase, subsequently referred to as isomerase, analogous results were obtained with 3a-hydroxysteroid dehydrogenase with respect to both induction and the inhibition of induction by 2'-deoxyadenosine (2'-AdR). In Figure 1 are presented the results of an experiment on the effect of additions of 2'-AdR, final concentration of 8 mM, at various times during the induction phase, on the development of isomerase activity in the bacterial cells. The dashed lines in the figure show the inhibitions caused by 2'-AdR of the subsequent development of enzyme activity. These results are typical of a number of experiments carried out under identical conditions; in none of these was the addition of 2'-AdR inhibitory to the further development of bacterial turbidity. At lower levels of 2'-AdR, inhibitions of the appearance of isomerase activity usually were less marked; thus, at 6 mM, added at 4 hr, the levels of enzyme activity attained at 7 hr, expressed as per- centages of maximal values, were 9, 46, and 48, respectively, while at 3 mM some of the analogous enzyme activity levels were 33, 56, 82, and 85 per cent of the maximal values. Of particular importance to the conclusion that the effects of 2'-AdR on enzyme induction cannot be attributed to an impurity in various samples of this nucleoside Downloaded by guest on September 27, 2021 VOL. 54, 1965 BIOCHEMISTRY: A. D. WELCH 1361

uc 100 ;TURBIDITY 0pt 100 = t 80 (A) A480 X FIG. 1.-The rapid rise in A5-3-ketoisomerase 5 60 < activity of cells of after m_ 60 x 60 5 hr of incubation, andPseudomona3 hr subsequenttestodteronito addition 0 l °0 at arrow (I) of A4-androstene-3,17-dione, is shown by 0 = l at the solid line (- X -). The effect of addition of - 40 E-40I- X ISOMERASE -> 2'-deoxyadenosine (final concentration: 8 milli- zO ACTIVITY no molar) at 4, 5, and 6 hr, respectively, after the initiation of incubation, on the levels of enzyme ., 0 0I-_ activity after incubation for 7 hr, is shown by the 0 * (A) dashed lines (X --- X). Turbidity values (-O-) >a 20 i II 20 in excess of 80% of maximum had been reached n ,x X control cultures (without added 0 A deoxyribonucleo- lo (A) - XloU side) by the time significant enzyme activity ap- M(I) J X- _ peared in the cells; additions of the nucleoside had o no effect on the development of bacterial turbidity. 0 2 4 5 6 7 DURATION OF INCUBATION ( HOURS) were experiments with the di-sodium salt of 2'-deoxyadenosine 5'-phosphate (2'- dAMP(5')), a compound that exerted no inhibitory effect on either the development of bacterial turbidity or enzyme induction, even at levels as high as 12 mM. After treatment of the nucleotide with purified alkaline phosphatase, however, the products were as inhibitory to the appearance of isomerase activity as was 2'-AdR of commercial origin. Some of the results obtained with the nucleotide and the products of phosphatase cleavage are presented in Table 1. TABLE 1 EFFECT ON LEVELS OF A5-3-KETOSTEROID ISOMERASE OF ADDITION OF SOME COMPOUNDS RELATED TO 2'-DEOXYADENOSINE Percentage of Percentage of Millimolarity maximal maximal Compound added* of compound isomeraset turbidityt 2'-dAMP(5')Na2§ 6 107 91 it 12 144 85 It 6 (treated with phosphatase) 10 85 Phosphatase control 75 104 3'-Deoxyadenosine 6 92 83 It 8 87 100 Adenine 6 67 85 it 9 48 84 Adenosine 6 118 80 (a 9 100 80 2'-Deoxyinosine 6 100 100 * All additions were made at 4 hr after initiation of incubation (2 hr after addition of inducer). t Isomerase (A'-3-ketosteroid isomerase activity) determined in cells after incubation for a total of 7 hr. t Turbidities determined after incubation for 7 hr. § Di-sodium salt of 2'-deoxyadenosine 5'-phosphate. Table 1 also shows the inactivity of deaminated 2'-AdR, i.e., 2'-deoxyinosine, and further evidence for the structural specificity of 2'-AdR is afforded by the lack of activity of 3'-deoxyadenosine. The corresponding ribonucleoside, adenosine, also exerted no significant activity, while the base, adenine, although moderately inhibitory to the induction of enzyme activity, was insufficiently so to account for the results with 2'-AdR, as is demonstrated by either Figure 1 or the results with phosphatase-treated 2'-dAMP(5') (Table 1). A number of experiments have been conducted with the 2'-deoxyribonucleosides Downloaded by guest on September 27, 2021 1362 BIOCHEMISTRY: A. D. WELCH PROC. N. A. S.

of guanine, cytosine, and thymine, and although the results have been somewhat variable, ranging from slight inhibition to possible stimulation of isomerase levels, the findings do not require tabulation to permit the conclusion that these derivatives of DNA do not have effects comparable to that of 2'-AdR. 2'-AdR had no effect on the activities of either isomerase or 3a-hydroxysteroid dehydrogenase, once these activities had appeared in the bacterial cells. Thus, the addition of 2'-AdR to the cells at the end of a 7-hr period of growth, either with or without an additional hour of incubation, or to extracts of the cells, did not diminish the amount of enzyme activity already present. Discussion.-The results presented indicate clearly that the induction by A4- androstene-3,17-dione of A5-3-ketosteroid isomerase activity in Pseudomonas testosteroni, grown in a salts-lactate-amino acid-yeast extract medium, can be promptly and markedly suppressed by the addition to the medium, preferably 2 or 3 hr after exposure to the inducer, of 2'-AdR in concentrations of 3-8 milli- molar. Other naturally occurring 2'-deoxyribonucleosides tested had much less activity and the specificity of the structural requirements is indicated also by the inability of 3'-deoxyadenosine (cordycepin), adenosine, and 2'-deoxyadenosine 5'-phosphate to arrest the induction process. The results suggest that the 5'- phosphate ester of 2'-AdR, added to the environment of this microorganism, is not converted to the nucleoside, but, more importantly, exposure of the nucleotide to alkaline phosphatase, which converts it to 2'-AdR, resulted in the appearance of the expected inhibitory activity; this finding supports the conclusion that the inhibitory effects of commercially obtained samples of 2'-AdR are attributable to the nucleoside per se and not to an impurity. Similarly, the activity of 2'-AdR cannot be accounted for by the release of adenine, since the free purine was significantly less active as an inhibitor of the induction process than was the deoxyribonucleoside. Although information concerning the fate of 2'-AdR, when incubated with Pseu- domonas testosteroni, is not yet available, studies of its penetration into the organism, as well as of its catabolism and anabolism are planned. The high concentrations of 2'-AdR required to inhibit enzyme induction suggest either that considerable break- down of the deoxyribonucleoside may occur or transport into the bacterial cells may proceed with difficulty, or both. It has been found (although data concerning this enzyme are not presented here) that the induction of 3a-hydroxysteroid dehydrogenase activity is affected in the same manner by 2'-AdR as is the induction of isomerase activity. It would be expected that 303, 173-hydroxysteroid dehydrogenase activity would be affected in the same way, but data concerning this enzyme are not yet available. It is conceivable that, if the mechanism of induction by certain steroids involves a strong binding between the inducer and a repressor,4 with resultant derepression, 2'-AdR or, more likely, a metabolically formed derivative of it, may cause a disso- ciation of such a complex, with restoration of repression. Such a mechanism might involve a modification of the tertiary structure of the repressor when in combination with the steroid. A related possibility for consideration is the formation of an actual complex between the steriod inducer and deoxyadenosine. Scott and Engel8 and Munck et al.9 have studied the complexing of a variety of steroids, including testosterone and A4-androstene-3,17-dione, with adenine and adenosine, as well as phosphate esters of the latter, but no data are available, as yet, on either 2'- or 3 - Downloaded by guest on September 27, 2021 VOL. 54, 1965 BIOCHEMISTRY: A. D. WELCH 1363

deoxyadenosine. In view of the findings described in the present paper, which demonstrate the relative inactivity of adenine, adenosine, and 3'-deoxyadenosine, as compared to 2'-AdR, and the complete inactivity of 2'-dAMP(5') (although this compound may not penetrate the cells), it would seem unlikely that an explanation of the rather striking effect of 2'-AdR is to be accounted for in this manner; however, an investigation of the capacity of 2'-AdR to complex steroids must be made before this possibility can be excluded. These studies must be extended to other systems, including mammalian cells, before any conclusions can be drawn concerning the general significance, if any, of the phenomenon described in Pseudomonas testosterani. Summary.-Studies of the induction by A4-androstene-3,17-dione of A5-3-ketosteroid isomerase activity in Pseudomonas testosteroni, grown in a salts-lactate-amino acid-yeast extract medium, have shown that 2'-deoxyadenosine, added to the medium at various periods during the induction process, promptly and markedly inhibits the appearance of this enzyme activity within the bacterial cells. Rela- tively inactive are 3'-deoxyadenosine (cordycepin), adenine, adenosine, 2'-deoxyinosine, and other naturally occurring 2'-deoxyribonucleosides. 2'-Deoxyadenosine 5'-phosphate is without activity until converted enzymatically to 2'-deoxyadenosine. Similar results have been obtained with a second induced enzyme, 3a- hydroxysteroid dehydrogenase. The activity of either enzyme, once formed, is not affected by 2'-deoxyadenosine or its metabolically produced derivatives. Some possible mechanisms of action of 2'-deoxyadenosine as an inhibitor of the induced synthesis of the enzymes are discussed.

The author expresses his thanks to Professor A. Wacker, and to other members of the staff of the Institut fur Therapeutische Biochemie, for the hospitality extended to him while on sabbatical leave at the University of Frankfurt, a stay made possible by generous financial assistance from The Commonwealth Fund. It is a pleasure for him to acknowledge the suggestions and interest of Professors P. Talalay, E. S. Canellakis, and C. F. Cori, and the assistance of Miss Heidi Hum- pert. The costs of these investigations were supported by grants from the U.S. Public Health Ser- vice (CY-2817), provided by the National Cancer Institute, and the Upjohn Company. 1 Talalay, P., Physiol. Rev., 37, 362 (1957). 2 Marcus, P. I., and P. Talalay, J. Biol. Chem., 218, 661 (1956). 3 Talalay, P., and P. I. Marcus, J. Biol. Chem., 218, 675 (1956). 4 Wacker, A., J. Drews, W. B. Pratt, and P. Chandra, Angew. Chem., 77, 172 (1965); English edition, 4, 155 (1965). 6 Wacker, A., J. Drews, W. B. Pratt, H. Laurent, and K. Petzoldt, Z. Naturforsch., 20, 547 (1965). 6 Kawahara, F. S., S.-F. Wang, and P. Talalay, J. Biol. Chem., 237, 1500 (1962). 7 Schmidt-Thome, J., and A. Butenandt, Ber., 69, 882 (1963). 8 Scott, J. F., and L. L. Engel, Biochim. Biophys. Acta, 23, 665 (1957). 9Munck, A., J. F. Scott, and L. L. Engel, Biochim. Biophys. Acta, 26, 297 (1957). Downloaded by guest on September 27, 2021