Recent Studies on Steroid Isomerases From Pseudomonas Testosteroni and Pseudomonas Putida

Temperature effects on the catalytic activity of the D38E mutant of 3-oxo-Delta5-steroid isomerase: favorable enthalpies and entropies of activation relative to the nonenzymatic reaction catalyzed by acetate ion

3-oxo-Delta5-steroid isomerase (ketosteroid isomerase, KSI) catalyzes the isomerization of 5-androstene-3,17-dione (1) to 4-androstene-3,17-dione (3) via a dienolate intermediate (2-). KSI catalyzes this conversion about 13 orders of magnitude faster than the corresponding reaction catalyzed by acetate ion, a difference in activation energy (DeltaG) of approximately 18 kcal/mol. To evaluate whether the decrease in DeltaG by KSI is due to enthalpic or entropic effects, the activation parameters for the isomerization of 1 catalyzed by the D38E mutant of KSI were determined. A linear Arrhenius plot of kcat/KM versus 1/T gives the activation enthalpy (DeltaH = 5.9 kcal/mol) and activation entropy (TDeltaS = -2.6 kcal/mol). Relative to catalysis by acetate, D38E reduces DeltaH by approximately 10 kcal/mol and increases TDeltaS by approximately 5 kcal/mol. The activation parameters for the microscopic rate constants for D38E catalysis were also determined and compared to those for the acetate ion-catalyzed reaction. Enthalpic stabilization of 2- and favorable entropic effects in both chemical transition states by D38E result in an overall energetically more favorable enzymatic reaction relative to that catalyzed by acetate ion.

Nature of the intermediate in the 3-oxo-.DELTA.5-steroid isomerase reaction

The role of Tyr-14 of 3-oxo-delta 5-steroid isomerase (KSI) was probed by analysis of the spectra of 3-amino-1,3,5(10)-estratrien-17 beta-ol (4) and equilenin (5) bound to the active site of KSI. The ultraviolet spectrum of 4 bound to KSI is identical to that for 4 in neutral solution. This observation indicates that Tyr-14 does not protonate the amine group of 4 at the active site. By analogy, it is argued that the 3-oxo group of steroid substrates for KSI is not protonated during the reaction. In contrast, the fluorescence excitation spectra of 5 bound to KSI show characteristics of an ionized phenol, even at pH values as low as 3.8. It is concluded that the pKa of equilenin is perturbed from its value in solution of 9 to less than or equal to 3.5 at the active site of KSI. Similarly, the pKa of the intermediate dienol in the KSI reaction should be lowered to less than or equal to 4.5 when it is bound to KSI. Thus, the function of Tyr-14 as an electrophilic catalyst is likely the stabilization of the anion of the dienol by hydrogen bonding rather than by proton transfer.

Mechanistic insights from the three-dimensional structure of 3-oxo-Delta(5)-steroid isomerase

3-Oxo-Delta(5)-steroid isomerase (KSI) catalyzes the isomerization of beta,gamma-unsaturated 3-oxosteroids to their conjugated isomers through the formation of an intermediate dienolate. The three-dimensional structure of the enzyme from Pseudomonas testosteroni was solved by multidimensional heteronuclear magnetic resonance spectroscopy. This protein, a 28-kDa symmetric dimer, exhibits a three-dimensional fold with the two independently folded monomers packed together via extensive hydrophobic and electrostatic interactions. The previously identified catalytically important residues Tyr-14 (general acid) and Asp-38 (general base) are located near the bottom of a deep hydrophobic cavity and are positioned in a manner consistent with previous mechanistic hypotheses. The structure also revealed the presence of an unexpected acid group (Asp-99) located in the active site adjacent to Tyr-14. Mutagenesis and kinetic studies show that Asp-99 has an anomalously high pK(a) (>9), which allows it to contribute to catalysis by donating a hydrogen bond to the intermediate and to the transition states. In support of this hypothesis, effects on the kinetic parameters of the mutations Y14F and D99A are additive in the Y14F/D99A mutant.

Electrophilic assistance by Asp-99 of 3-oxo-Delta 5-steroid isomerase

3-Oxo-Delta 5-steroid isomerase (Delta 5-3-ketosteroid isomerase, KSI; EC 5.3.3.1) catalyzes the conversion of a variety of beta, gamma-unsaturated 3-oxosteroids to their corresponding alpha, beta-unsaturated isomers at rates that approach the diffusion limit for specific substrates. The reaction proceeds through a dienolate intermediate, with two amino acid residues (Asp-38 and Tyr-14) known to be involved in catalysis. When the complete three-dimensional structure of KSI was determined recently by NMR methods, an additional polar residue (Asp-99) was found in the active site and this group was shown to be important for catalytic activity. In this work, we examine the properties of several mutant KSIs to determine the nature of catalysis by Asp-99 of KSI. The electrophoretic mobilities of wild-type (WT) KSI and several mutants (D99A, D99N, D38N, and D38N/D99A) on native gels were determined at pH values ranging from 6.0 to 8.5. The results demonstrate that the pKa of Asp-99 is >8.5 in wild-type KSI. The pH-rate profiles for the D99A, D99N, and D38H/D99A mutants of KSI were also determined. For all three mutants, kcat and kcat/KM do not decrease at high pH, in contrast to those for WT and D38H, which lose activity above pH 9 and 8, respectively. Mutation of Asp-99 to Asn decreases kcat for the substrate 5-androstene-3,17-dione by 27-fold and kcat/Km by 23-fold, substantially less than the loss of activity (3000-fold in kcat and 2200-fold in kcat/Km) observed when Asp-99 is mutated to Ala, consistent with a hydrogen bonding role for Asp-99. Taken together, these results provide evidence that Asp-99 participates in catalysis in its protonated form, with a pKa of >9 in WT and approximately 8.5 in the D38H mutant. Asp-99 likely donates a hydrogen bond to O-3 of the steroid, helping to stabilize the transition state(s) of the KSI-catalyzed reaction.

Solution structure of 3-oxo-delta5-steroid isomerase

The three-dimensional structure of the enzyme 3-oxo-delta5-steroid isomerase (E.C. 5.3.3.1), a 28-kilodalton symmetrical dimer, was solved by multidimensional heteronuclear magnetic resonance spectroscopy. The two independently folded monomers pack together by means of extensive hydrophobic and electrostatic interactions. Each monomer comprises three alpha helices and a six-strand mixed beta-pleated sheet arranged to form a deep hydrophobic cavity. Catalytically important residues Tyr14 (general acid) and Asp38 (general base) are located near the bottom of the cavity and positioned as expected from mechanistic hypotheses. An unexpected acid group (Asp99) is also located in the active site adjacent to Tyr14, and kinetic and binding studies of the Asp99 to Ala mutant demonstrate that Asp99 contributes to catalysis by stabilizing the intermediate.

Mechanism of proton transfer in the isomerization of 5-androstene-3, 17-dione by 3-oxo-delta 5-steroid isomerase and its D38E mutant

The stereochemistry of proton transfer in the isomerization of [4 beta-2H]-5-androstene-3,17- dione (1d) to 4-androstene-3,17-dione (3) catalyzed by 3-oxo-delta 5-steroid isomerase (KSI) has been reinvestigated. In H2O, approximately 65% of the label is retained in the product (3); of this, one-third is at C-4 and two-thirds at C-6 beta. When the same reaction is catalyzed by the D38E mutant of KSI, ca. 60% of the label is retained in the product, but almost all of it is at C-4. These reactions run in deuterium oxide result in 13% incorporation of a second deuterium with the wild type (WT) enzyme and 75% incorporation with the D38E mutant. When unlabeled 1 is isomerized in D2O, there is little incorporation of deuterium with WT (ca. 5 at. %) but substantial incorporation with D38E (130 at. %). These results are consistent with competitive abstraction of both the C-4 alpha and C-4 beta protons, as proposed by Viger et al. [(1981) J. Am. Chem. Soc. 103, 4151], and demonstrate that the KSI reaction is not completely stereospecific. A mechanism is proposed to account for these observations.

Crystallization and preliminary x-ray crystallographic studies of ketosteroid isomerase from Pseudomonas putida biotype B

The delta(5)-3-ketosteroid isomerase from Pseudomonas putida biotype B has been crystallized. The crystals belong to the space group P2(1)2(1)2(1) with unit cell dimensions of a = 36.48 angstrom, b = 74.30 angstrom, c = 96.02 angstrom, and contain one homodimer per asymmetric unit. Native diffraction data to 2.19 Angstrom resolution have been obtained from one crystal at room temperature indicating that the crystals are quite suitable for structure determination by multiple isomorphous replacement.

Cloning, nucleotide sequence, and overexpression of the gene coding for delta 5-3-ketosteroid isomerase from Pseudomonas putida biotype B

The structural gene coding for the delta 5-3-ketosteroid isomerase (KSI) of Pseudomonas putida biotype B has been cloned, and its entire nucleotide sequence has been determined by a dideoxynucleotide chain termination method. A 2.1-kb DNA fragment containing the ksi gene was cloned from a P. putida biotype B genomic library in lambda gt11. The open reading frame of ksi encodes 393 nucleotides, and the amino acid sequence deduced from the nucleotide sequence agrees with the directly determined amino acid sequence (K. Linden and W. F. Benisek, J. Biol. Chem. 261:6454-6460, 1986). A putative purine-rich ribosome binding site was found 8 bp upstream of the ATG start codon. Escherichia coli BL21(DE3) transformed with the pKK-KSI plasmid containing the ksi gene expressed a high level of isomerase activity when induced by isopropyl-beta-D-thiogalactopyranoside. KSI was purified to homogeneity by a simple and rapid procedure utilizing fractional precipitation and an affinity column of deoxycholate-ethylenediamine-agarose as a major chromatographic step. The molecular weight of KSI was 14,535 (calculated, 14,536) as determined by electrospray mass spectrometry. The purified KSI showed a specific activity (39,807 mumol min-1 mg-1) and a Km (60 microM) which are close to those of KSI originally obtained from P. putida biotype B.

Energetics of 3-oxo-delta 5-steroid isomerase: source of the catalytic power of the enzyme

Knowledge of the partitioning of the putative dienol intermediate (2) by steroid isomerase (KSI) (Hawkinson et al. 1991), in conjunction with various steady-state kinetic parameters, allows elucidation of the detailed free energy profile for the KSI-catalyzed conversion of 5-androstene-3,17-dione (1) to 4-androstene-3,17-dione (3). This free energy profile shows four kinetically significant energy barriers (substrate binding, the two chemical steps, and dissociation of product) that must be traversed upon conversion of 1 to 3. Thus, no single step of the catalytic cycle is cleanly rate-limiting. The source of the catalytic power of KSI is discussed via comparison of the free energy profile for the KSI-catalyzed isomerization with those for the acetate-catalyzed isomerization and the aqueous reaction at pH 7. Similarities between the energetics of the KSI-catalyzed and triosephosphate isomerase catalyzed reactions are also noted.

Crystal structure of cholesterol oxidase from Brevibacterium sterolicum refined at 1.8 A resolution

Cholesterol oxidase (3 beta-hydroxysteroid oxidase, EC 1.1.3.6) is an FAD-dependent enzyme that carries out the oxidation and isomerization of steroids with a trans A : B ring junction. The crystal structure of the enzyme from Brevibacterium sterolicum has been determined using the method of isomorphous replacement and refined to 1.8 A resolution. The refined model includes 492 amino acid residues, the FAD prosthetic group and 453 solvent molecules. The crystallographic R-factor is 15.3% for all reflections between 10.0 A and 1.8 A resolution. The structure is made up of two domains: an FAD-binding domain and a steroid-binding domain. The FAD-binding domain consists of three non-continuous segments of sequence, including both the N terminus and the C terminus, and is made up of a six-stranded beta-sheet sandwiched between a four-stranded beta-sheet and three alpha-helices. The overall topology of this domain is very similar to other FAD-binding proteins. The steroid-binding domain consists of two non-continuous segments of sequence and contains a six-stranded antiparallel beta-sheet forming the "roof" of the active-site cavity. This large beta-sheet structure and the connections between the strands are topologically similar to the substrate-binding domain of the FAD-binding protein para-hydroxybenzoate hydroxylase. The active site lies at the interface of the two domains, in a large cavity filled with a well-ordered lattice of 13 solvent molecules. The flavin ring system of FAD lies on the "floor" of the cavity with N-5 of the ring system exposed. The ring system is twisted from a planar conformation by an angle of approximately 17 degrees, allowing hydrogen-bond interactions between the protein and the pyrimidine ring of FAD. The amino acid residues that line the active site are predominantly hydrophobic along the side of the cavity nearest the benzene ring of the flavin ring system, and are more hydrophilic on the opposite side near the pyrimidine ring. The cavity is buried inside the protein molecule, but three hydrophobic loops at the surface of the molecule show relatively high temperature factors, suggesting a flexible region that may form a possible path by which the substrate could enter the cavity. The active-site cavity contains one charged residue, Glu361, for which the side-chain electron density suggests a high degree of mobility for the side-chain. This residue is appropriately positioned to act as the proton acceptor in the proposed mechanism for the isomerization step.

Detection of a transient enzyme-steroid complex during active-site-directed irreversible inhibition of 3-oxo-delta 5-steroid isomerase

The reaction of the active-site-directed irreversible inhibitor (17S)-spiro[estra-1,3,5(10),6,8-pentaene-17,2'-oxiran]-3-ol (5 beta) with 3-oxo-delta 5-steroid isomerase has been monitored by repetitive scanning ultraviolet spectroscopy of a solution of 5 beta plus isomerase against a blank containing only 5 beta. Upon initial mixing of 5 beta with the isomerase an absorbance maximum at ca. 250 nm appears. With time, this peak decreases and is replaced with a new peak near 280 nm. These results directly demonstrate the existence of a transient enzyme-steroid intermediate in the inactivation reaction. The ultraviolet spectrum suggests that the steroid in the transient complex resembles the ionized phenol, while the phenolic group in the irreversibly bound complex is un-ionized. These spectral studies support our previous proposal that there are two enzyme-steroid complexes that are related by a 180 degree rotation about an axis perpendicular to the plane of the steroid nucleus. This hypothesis offers an explanation for the reaction of 17 beta-oxiranes with the same residue (Asp-38) that is thought to be involved in the catalytic mechanism. Two new oxiranes, (17S)-spiro[estra-1,3,5(10)-triene-17,2'-oxiran]-3 beta-ol (6 beta) and (17S)-spiro[5 alpha-androstane-17,2'-oxiran]-3-one (8 beta), were also found to be potent active-site-directed irreversible inhibitors of the isomerase (k3/KI = 31 M-1 s-1 and 340 M-1 s-1, respectively). The relationship of these results to the nature of the active site of the isomerase is discussed.

Photoaffinity modification of delta 5-3-ketosteroid isomerase by light-activatable steroid ketones covalently coupled to agarose beads

In order to identify the minor site(s) of photoattachment of unsaturated steroid ketones to delta 5-3-ketosteroid isomerase from Pseudomonas testosteroni, we have developed a solid-state photoaffinity labeling technique. Two solid-state reagents, O-carboxymethylagarose-ethylenediamine-succinyl-17 beta-O-19-nortestosterone and O-carboxymethylagarose-ethylenediamine-succinyl-17 beta-O-4,6-androstadien-3-one, have been synthesized. Under anaerobic conditions, isomerase bound to these resins is photoinactivated by UV light (lambda greater than 290 nm) whereas isomerase bound to O-carboxymethylagarose-ethylenediamine-deoxycholate or isomerase in the presence of O-carboxymethylagarose-ethylenediamine-acetate is almost completely stable to irradiation under the same conditions. Photoinactivation under anaerobic condition promoted by the resin-bound steroid ketones results from a reaction at the active site since the competitive inhibitor, sodium cholate, which does not absorb light above 290 nm, provides protection toward photoinactivation. Preliminary analysis of isomerase that has been photolyzed in the presence of O-carboxymethylagarose-ethylenediamine-succinyl-17 beta-O-4,6-androstadiene-3-one has established that the enzyme is converted to at least two different forms. One form binds more tightly to the resin than does the native enzyme. This form can be eluted by a sodium dodecyl sulfate containing buffer. The second form is not eluted by this buffer but can be released from the resin by cleavage of the ester bond linking the steroid to the derivatized agarose. We presume that the latter form is covalently coupled to the resin-linked steroid. In the presence of oxygen, additional nonspecific inactivation reactions occur, but these can be suppressed by the singlet oxygen trap, L-histidine. The application of solid-state photoaffinity reagents to some areas of receptor isolation and characterization is discussed.