6. Purification

Langer and Engel (74) have described a fiftyfold purification of placental 17β-hydroxysteroid dehydrogenase. A somewhat different procedure, which results in about hundredfold purification and involves stabilization with estradiol-17β, has been reported from our laboratory (112). Accurate and comparable transhydrogenase measurements could not be obtained with assay systems involving TPNH, as such, or TPNH-generating enzymes. This led to the suggestion that the transhydrogenase activity be measured by differential spectrophotometry from DPNH as donor to the acetyl-pyridine analog of DPN as acceptor. This measurement in turn is complicated by the existence in crude systems of enzymes (probably flavoproteins) which catalyze hydrogen transfer from reduced pyridine nucleotides to analogs of higher oxidoreduction potential (133). The measurements were corrected for this "nonspecific" transhydrogenase component by including an additional reaction vessel to which TPN was added. Evidence was presented that TPN was without influence on the steroid-independent activity but completely suppressed the estradiol-17β-dependent transhydrogenation from DPNH to the acetylpyridine analog of DPN.

Using DPN as acceptor for dehydrogenase assays, and the transfer of hydrogen from DPNH to the acetylpyridine analog of DPN for the assay of transhydrogenase, it was possible to follow the enzyme purification and measure the two activities at each step. The ratios of the activities varied by as much as a factor of 2 to 4 during the hundredfold purification. This could not be considered as evidence for separation of the enzymes because of the uncertainty of the assay procedures. It has now been asserted by Hagerman and Villee (39, 40) that there are in placenta two 17β-hydroxysteroid dehydrogenases, specific for DPN and for TPN respectively, in addition to the transhydrogenase. We have repeated our purification, and have measured dehydrogenase activities with six nucleotides, at pH 7.4 and 9.0. These measurements were carried out at different stages of purification when the specific activities ranged from 7.3 to 112 dehydrogenase units per milligram protein. Assuming the rate of reduction of DPN at pH 7.4 as equal to 100 for each enzyme preparation, the rates of reaction of the other nucleotides varied from 10 to 500. However, the relative rates of reduction of the nucleotides were remarkably constant. The transhydrogenase activity of the same preparations was also examined in three systems at pH 8.4. Assuming that the rate of the transfer of hydrogen from DPNH to the acetylpyridine analog of DPN was 100, the relative rates of the transfers from DPNH to acetylpyridine analog of TPN and from TPNH to the acetylpyridine analog of TPN were similar in preparations of differing purity (Table IV).

We have stored many preparations of purified placental 17β-hydroxysteroid dehydrogenase for periods up to one year, during which time they lost 75% or more of their activity. The relative reactivity with the six nucleotides has remained identical.

7. Summary

The evidence that there is a single 17β-hydroxysteroid dehydrogenase in placenta, catalyzing estrogen-dependent transhydrogenations, may be summarized as follows: (a) The dehydrogenase activities with several nucleotides and the transhydrogenase activities parallel each other in each preparation throughout more than hundredfold purification of the enzyme.

Table IV. Dehydrogenase And Transhydrogenase Activities Of Placental 17β-Hydroxysteroid Dehydrogenase With Various Nucleotides

* Dehydrogenase activities were measured at 25° C in systems of 3.0 ml. final volume containing: 300 umoles Tris buffer of pH 7.4 or 9.0, 80 ug. estradiol-17β in 0.04 ml. acetone, 25 mg. bovine serum albumin, 1 mg. of each nucleotide, and sufficient enzyme to give a convenient rate. Blank cuvettes contained all ingredients except the steroid. Rates were measured at the maximum absorption wavelengths of the reduced nucleotide and are corrected for differences in extinction coefficient (see Table I). The rate with DPN at pH 7.4 was taken as 100 for each enzyme preparation.

b Transhydrogenase activities were measured at 25° C. in systems of 3.0 ml. final volume containing: 300 umoles Tris buffer of pH 8.2, 4 u.g. estradiol-17p in 0.01 ml. acetone, a convenient quantity of each enzyme and the following mixtures of nucleotides: 0.4 umole DPNH plus 2.3 umoles acetylpyridine-DPN; or, 0.4 µmole DPNH and 0.5 umole acetylpyridine-TPN; or, 0.18 umole TPNH and 0.5 umole acetylpyridine-TPN. All measurements were carried out at 385 mu.. An additional blank was included which contained no estradiol-17β, and 5 µmoles of TPN. The rates are corrected for the rate of nonspecific hydrogen transfer in the presence of TPN (0-30% of total rate). The rate of hydrogen transfer from DPNH to acetylpyridine-DPN was taken as 100 for each enzyme preparation.

(b) The steroid specificity of both reactions is identical, (c) The pyridine nucleotide specificity of the two reactions is also identical, (d) The inactivation by heat and the stability to storage of both activities remain parallel to each other, (e) During the course of transhydrogenation there is an interconversion of estradiol-17β and estrone, (f) 17β-Hydroxysteroid dehydrogenase has the same stereospecificity for pyridine nucleotides as the transhydrogenase reaction (side II).

In addition to these direct experiments with the purified placental enzyme, certain other findings strengthen the contention of identity of hydroxysteroid dehydrogenase and transhydrogenase. It has been demonstrated that partially purified 3a-hydroxysteroid dehydrogenase from rat liver can promote hydrogen transfer between pyridine nucleotides with appropriate steroids (55). In addition, we have shown an active, steroid-dependent hydrogen transfer between DPNH and certain analogs of DPN by highly purified bacterial hydroxysteroid dehydrogenases (111).

IV. Hydrogen Transport By Other Hydroxysteroid Dehydrogenases. 1. Bacterial Hydroxysteroid Dehydrogenases

Although the bacterial hydroxysteroid dehydrogenases require DPN as hydrogen acceptor, and are inert toward TPN, they also react with certain of the analogs of DPN. We have found (111) that purified bacterial 3a-and (3- and 17-)β-hydroxysteroid dehydrogenases can promote efficient hydrogen transfer between DPNH and suitable analogs of DPN in the presence of catalytic quantities of appropriate steroids. These reactions occur by virtue of the alternate reduction of ketosteroids by the donor nucleotide, and oxidation of the resultant steroid alcohol by the acceptor nucleotides. These studies provide useful models for the steroid-dependent transhydrogenations catalyzed by mammalian hydroxysteroid dehydrogenases which react with both DPN and TPN.

a. Bacterial 3a-Hydroxysteroid Dehydrogenase, i. Reactivity oj 3a-kydroxysteroid dehydrogenase with DPN analogs. The highly purified enzyme (83) reduces DPN, the pyridine aldehyde, the acetylpyridine and the thionicotinamide analogs of DPN, in the presence of substrate amounts of androsterone (111). The rates of reduction of the nucleotides were measured as a function of their concentration in pyrophosphate buffer at pH 9.5. In the presence of saturating concentrations of androsterone, the dependence of velocity on nucleotide concentration obeyed the kinetic predictions of Michaelis and Menten. The Michaelis constants (KM) and maximum velocities (VM) were calculated from Lineweaver-Burk plots and are shown in Table V. At pH 9.5, the thionicotinamide analog of DPN was bound more firmly and reduced more rapidly than the corresponding acetyl-pyridine and pyridine aldehyde analogs.

Table V. Reduction Of Dpn And Analogs Of Dpn Catalyzed By Bacterial 3a-Hydroxysteroid

Dehydrogenase*

* Velocities were measured at 25° C. in systems of 3.0 ml. final volume containing: 100 umoles sodium pyrophosphate buffer of pH 9.5; 30 ug. androsterone in 0.03 ml. methanol; 0.25 to 5.0 umoles of each nucleotide; and 8.6 ug. of purified 3a-hydroxy-steroid dehydrogenase (specific activity 83,700 units per milligram protein). Controls contained all ingredients except the steroid. Measurements were carried cut at wavelengths of maximum absorption of the reduced nucleotides and are corrected for differences in extinction coefficient (Table I).