This reaction is made possible by the dual nucleotide specificity of the dehydrogenat-ing enzyme.

The placental hydroxysteroid dehydrogenase which oxidizes estradiol 17/3 is not alone in this class of enzymes in having dual nucleotide specificity. A soluble 3 a-hydroxysteroid dehydrogenase isolated from liver reacts with both DPN and TPN.10 Recent experiments29 have shown that liver microsomes of various species contain a firmly bound 3a-hydroxysteroid dehydrogenase. These particles also interconvert hydrocortisone and cortisone by the action of a specific 11β-hy-droxysteroid dehydrogenase. Both microsomal enzymes react at comparable rates with DPN and with TPN.

Detailed studies of the specificities and substrate affinities of highly purified, adaptive hydroxysteroid dehydrogenases of bacteria have revealed that these enzymes bind certain steroids tenaciously.9,30 The magnitudes of the Michaelis constants correspond with the concentrations at which many steroids exert their physiological actions. Moreover, the binding of steroids by these enzymes is very sensitive to even minor structural alterations of the steroid molecule. A number of striking similarities have been shown to exist between the molecular features necessary for physiological activity and those which favor efficient binding of the steroid to the enzyme surfaces.9,30 The observation that these enzymes exhibit both the high affinity and the specificity for steroids demanded by hormonal activity led to the suggestion that hydroxysteroid dehydrogenases are intimately concerned with the hormonal action of steroids.9,30 The importance of the state of oxidation of steroids had already been pointed out by Huggins and his co-workers31 from considerations of the relationship of molecular structure to the growth-promoting activity of these hormones.

The present experiments strongly suggest that steroids can participate in transhydrogenation between pyridine nucleotides. On the assumption that the enzyme in placenta which catalyzes the reduction of TPN and DPN by estradiol 17β is the same as that which promotes the steroid-activated transfer of hydrogen from TPNH to DPN, it is possible to regard the pyridine nucleotides as substrates for the transhydrogenation, and the steroid as a coenzyme for the latter reaction. In this way it becomes possible to assign a functional role to the mammalian hydroxysteroid dehydrogenases which possess dual nucleotide specificity. In the presence of truly minute amounts of steroids, they act as mediators of hydrogen transfer between pyridine nucleotides.

It would appear that the thermodynamic properties of steroid hormones are well suited for the promotion of hydrogen transfer in such systems. Equilibrium constants for the interconversion of hydroxy- and ketosteroids have recently become available from measurements made with highly purified bacterial hydroxysteroid dehydrogenases.30 The free-energy changes involved in these reactions are such as to favor the existence of significant amounts of both oxidized and reduced forms of steroid at physiological hydrogen ion concentrations. Thus, at pH 7 and at 298° K., the equilibrium ratio of 4-androstene-3,17-dione to testosterone is in the vicinity of 0.4, assuming that the oxidized and reduced pyridine nucleotides are present in equimolar amounts. The calculated change in free energy for the interconversion of these two steroids is thus Δ F° = 5.2 Kcal/mole. For several 3-hydroxysteroids, the equilibrium ratio of ketone to alcohol varies from 0.01 to 0.07 (Δ F° = 6.0-7.2 Kcal/mole) under the same conditions. The oxidation of a steroid hydroxyl group to a ketone with a pyridine nucleotide as a hydrogen acceptor involves the stoichiometric liberation of hydrogen ion; hence the equilibrium is strictly dependent upon pH. It follows that the ability of steroids to effect transhydrogenation by their reversible oxido-reduction would vary in a predictable manner with the hydrogen ion concentration. For example, at pH 9.0 and in the presence of the appropriate hydroxysteroid dehydrogenase and equimolar concentration of oxidized and reduced pyridine nucleotide, 17β hydroxysteroids would exist in the ketone form to an extent of 97 per cent. Thus, at pH 9.0, one would not expect the steroid-hydroxy-steroid dehydrogenase system to function as an efficient mediator of transhydrogenation. Similar considerations would hold if the hydrogen ion concentration was too far removed from neutrality on the acid side. If the present interpretation of our findings is correct, these deductions are in agreement with the finding of Gordon and Villee' that the activation of the reduction of DPN in crude extracts of placenta by estradiol 17/3 is maximal at pH 7.3.

It may be emphasized that the oxidation of the 17β-hydroxyl group of steroids gives, at neutrality, an equilibrium ratio of alcohol to ketone closer to 1 than that of any other steroid oxido-reduction which has been studied.'30 It is, perhaps, not altogether fortuitous that the 17β-hydroxyl function is of paramount importance for the biological activity of estrogenic and androgenic hormones.

The few known mammalian pyridine nucleotide-linked enzymes which exhibit dual nucleotide specificity with non-steroidal substrates, e.g., glutamic dehydrogenase" and glucose dehydrogenase," may not be expected to function as efficiently as hydroxysteroid dehydrogenases to effect a transfer of hydrogen from TPNH to DPN, or vice versa, at pH 7. The equilibria which they catalyze are unfavorable, and the affinities for their substrates are too low in this respect.

All types of naturally occurring steroid hormones undergo a variety of oxido-reductions in animal tissues, and the enzymes catalyzing these transformations are widely distributed in nature.4 A number of these enzymes possess the properties requisite for them to act as transhydrogenating systems between pyridine nucleotides and, possibly, between other coenzymes engaged in hydrogen transport. From the limited evidence at hand, it appears that these enzymes differ considerably in their affinities and specificities for steroids and in their intracellular and tissue distribution. Thus some of the enzymes which permit the oxidation or reduction of oxygen substituents on steroids may have significance not only as mechanisms for the inactivation of hormones4 but also as catalysts for the interconversion of the oxidized and reduced forms of hydrogen-transporting coenzymes. For this reason, further studies on the properties and distribution of the family of enzymes which catalyze the oxido-reduction of steroids in animal tissues are of the utmost importance. The fact that certain steroid hormones exert a profound action in more than one physiological domain may well relate to the lack of rigid specificities of various hydroxysteroid dehydrogenases.

Summary

Soluble enzyme preparations of human placenta promote the transfer of hydrogen from TPNH to DPN in the presence of minute amounts of certain steroid hormones. The same enzyme preparations catalyze the oxidation of these steroids by both forms of pyridine nucleotide. Evidence is presented for the identity of a placental hydroxysteroid dehydrogenase with the transhydrogenase activity.

It is proposed that the metabolic function of hydroxysteroid dehydrogenases with dual pyridine nucleotide specificity is to act as pyridine nucleotide transhydrogen-ases. Properties of some hydroxysteroid dehydrogenases which favor transhydrogenation are: (1) their high affinity and specificity for particular steroids; (2) their ability to react with both DPN and TPN; and (3) the suitable equilibria between steroid alcohols and ketones which obtain at physiological hydrogen ion concentrations.

*Supported by grants from the American Cancer Society.

† Scholar in Cancer Research of the American Cancer Society.

1 C. A. Villee, J. Biol. Chem., 215, 171, 1955. 2E. E. Gordon and C. A. Villee, J. Biol. Chem., 216, 215, 1955. 3 C. A. Villee and E. E. Gordon, J. Biol. Chem., 216, 203, 1955. 4 P. Talalay, Physiol. Revs., 37, 362, 1957.

5O. Warburg and W. Christian, Biochem. Z., 310, 384, 1941.

6G. Siebert, J. Dubuc, R. C. Warner, and G. W. E. Plaut, J. Biol. Chem., 226, 965, 1957. 7B. Hurlock and P. Talalay, J. Biol. Chem., 227, 37, 1957.

8 L. Langer and L. L. Engel, Federation Proc., 15, 296, 1956.

9 P. I. Marcus and P. Talalay, Proc. Roy. Soc. London, B, 144, 116, 1955. 10 G. M. Tomkins, J. Biol. Chem., 218, 437, 1956.

11 R. G. Langdon, J. Bioi. Chem., 226, 615, 1957.

12 F. Lynen and K. Decker, Ergeb. Physiol., 49, 327, 1957.

13 F. M. Huennekens, Y. Hatefi, and L. D. Kay, J. Biol. Chem., 224, 435, 1957.

14 J. G. Flaks, L. Warren, and J. M. Buchanan, J. Biol. Chem., 228, 215, 1957.

15 Y. Hatefi, M. J. Osborn, L. D. Kay, and F. M. Huennekens, J. Biol. Chem., 227,637,1957.

16. C. Mueller, Cancer Research, 17, 490, 1957. 17 T. Mann, The Biochemistry of Semen (London: Methuen, 1954). 18 H. G. Hers, Biochim. et Biophys. Acta, 22, 203, 1956.

19 H. G. Williams-Ashman, J. Banks, and S. K. Wolfeon, Jr., Arch. Biochem. and Biophys. (in press).

20 H. G. Williams-Ashman, Abstracts of Communications, Third International Congress of Biochemistry, Brussels, 1956 (New York: Academic Press, 1956), p. 64. 21 H. G. Williams-Ashman and S. Liao (in preparation). 22 S. J. Nyden and H. G. Williams-Ashman, Am. J. Physiol., 172,588,1953. 23 G. Glock and P. McLean, Biochem. J., 61, 388, 1955. 24 F. L. Hoch and F. Lipmann, these Proceedings, 40, 909, 1954. 25 N. O. Kaplan, S. P. Colowick, and E. F. Neufeld, J. Biol. Chem., 195, 107, 1952. 26 N. O. Kaplan, S. P. Colowick, and E. F. Neufeld, J. Biol. Chem., 205, 1, 1953. 27 G. F. Humphrey, Biochem. J., 65, 546, 1957.

28 N. O. Kaplan, M. N. Schwartz, M. E. Freeh, and M. M. Ciotti, these Proceedings, 42, 481, 1956.

29 B. Hurlock and P. Talalay (in preparation). 30 P. Talalay, Record Chem. Progr., 18, 31, 1957.

31 C. Huggins, E. V. Jensen, and A. S. Cleveland, /. Exptl. Med., 100, 225, 1954. 32 J. A. Olson and C. B. Anfinsen, Biol. Chem., 202, 841, 1953. 33 H. J. Strecker and S. Korkes, /. Biol. Chem., 196, 769, 1952.