Skip to main content
NCBI home page
Protein Science : A Publication of the Protein Society logoLink to Protein Science : A Publication of the Protein Society
. 2005 Jun;14(6):1485–1497. doi: 10.1110/ps.051353205

Comparison of crystal structures of human type 3 3α-hydroxysteroid dehydrogenase reveals an “induced-fit” mechanism and a conserved basic motif involved in the binding of androgen

Jean-François Couture 1,1, Karine Pereira De Jésus-Tran 1, Anne-Marie Roy 1, Line Cantin 1, Pierre-Luc Côté 1, Pierre Legrand 1,2, Van Luu-The 1, Fernand Labrie 1, Rock Breton 1
PMCID: PMC2253370  PMID: 15929998

Abstract

The aldo-keto reductase (AKR) human type 3 3α-hydroxysteroid dehydrogenase (h3α–HSD3, AKR1C2) plays a crucial role in the regulation of the intracellular concentrations of testosterone and 5α-dihydrotestosterone (5α-DHT), two steroids directly linked to the etiology and the progression of many prostate diseases and cancer. This enzyme also binds many structurally different molecules such as 4-hydroxynonenal, polycyclic aromatic hydrocarbons, and indanone. To understand the mechanism underlying the plasticity of its substrate-binding site, we solved the binary complex structure of h3α–HSD3-NADP(H) at 1.9 Å resolution. During the refinement process, we found acetate and citrate molecules deeply engulfed in the steroid-binding cavity. Superimposition of this structure with the h3α–HSD3-NADP(H)-testosterone/acetate ternary complex structure reveals that one of themobile loops forming the binding cavity operates a slight contraction movement against the citrate molecule while the side chains of many residues undergo numerous conformational changes, probably to create an optimal binding site for the citrate. These structural changes, which altogether cause a reduction of the substrate-binding cavity volume (from 776 Å3 in the presence of testosterone/acetate to 704 Å3 in the acetate/citratecomplex), are reminiscent of the “induced-fit” mechanism previously proposed for the aldose reductase, another member of the AKR superfamily. We also found that the replacement of residues Arg301 and Arg304, localized near the steroid-binding cavity, significantly affects the 3α–HSD activity of this enzyme toward 5α-DHT and completely abolishes its 17β–HSD activity on 4-dione. All these results have thus been used to reevaluate the binding mode of this enzyme for androgens.

Keywords: aldo-keto reductase, hydroxysteroid dehydrogenase, crystal structure, induced-fit mechanism

Hydroxysteroid dehydrogenases (HSDs), which form a subgroup in the aldo-keto reductase (AKR) superfamily (AKR1C), are well known to selectively catalyze the oxydo-reduction of hydroxyl/ketone groups found at specific positions on steroid molecules (Dufort et al. 1996, 1999, 2001; Zhang et al. 2000) (Fig. 1A ▶). Among the members of the AKR1C subgroup, human type 3 3α–HSD (h3α–HSD3; AKR1C2; EC 1.1.1.213) is able to exert its activity on structurally different steroids, such as androgens (C19-steroids) and progestins (C21-steroids). In addition to steroids (Dufort et al. 1996, 2001; Penning et al. 2000), this enzyme is also able to bind/transform numerous molecules with various shapes, such as acetate (Nahoum et al. 2001), prostaglandins (Lovering et al. 2004), polycyclic aromatic hydrocarbons (PAHs) (Palackal et al. 2001, 2002), 1-indanol (Matsuura et al. 1997), 9, 10-phenanthrenequinone (Penning et al. 1984), and 4-hydroxy-2- nonenal (Burczynski et al. 2001) (Fig. 1B ▶). Because all these structurally different molecules bind in the ligand-binding cavity, mainly via hydrophobic contacts, h3α– HSD3 must adapt the structure of its binding cavity in order to form an optimal binding site for each of them. This capacity to alter the orientation of numerous side chains surrounding the ligand-binding site, which was first described for the yeast hexokinase, is named the “induced-fit” mechanism (DelaFuente and Sols 1970; DelaFuente et al. 1970). The induced-fit mechanism is also known for members of the AKR superfamily and has been well studied, notably in the case of the aldose reductase, in a strategy for the design of inhibitors (Klebe et al. 2004; Sotriffer et al. 2004).

Figure 1.

Figure 1.

Open in a new tab

Schematic representation of enzymatic properties of HSDs member of the AKR superfamily. (A) Reactions catalyzed by h3α–HSD3 on 4-dione and DHT. (B) Overview of the molecules that bind and/or are transformed by HSDs member of the AKR superfamily (the figure has been created with SwissPDBViewer and rendered with Povray; Kaplan and Littlejohn 2001). (C) Orbital steering constraints proposed by Heredia et al. (2003). Here, the optimal distance from the catalytic Tyr residue, the angle of the hydride with the target ketone (in parentheses), and the distance of the C4 of the pyridine head of the cofactor are indicated (created with Molscript and POVScript; Kraulis 1991; Fenn et al. 2003).

Several crystallographic studies have been performed on mammalian 3α–HSDs. Rat 3α–HSD structure (r3α– HSD, AKR1C9) has been solved in ternary complex with testosterone (the product of the 17β-reduction of 4-dione) (Hoog et al. 1994; Bennett et al. 1996, 1997) (Fig. 1A ▶), whereas the structure of human 3α–HSD3 has been determined in ternary complex with testosterone/acetate (Nahoum et al. 2001) and ursodeoxycholate (Jin et al. 2001), a bile acid possessing a steroid-like architecture. These enzymes are folded as triose-phosphate isomerase barrels (α/β)8 (Hoog et al. 1994). They maintain the cofactor in an extended conformation (Bennett et al. 1996) and stabilize the steroids through numerous contacts, mainly with hydrophobic residues belonging to three large loops named A, B, and C. Biochemical studies performed on r3α–HSD demonstrated that the reaction catalyzed by this enzyme is triggered by the 4-Pro-R hydride transfer from the C-4 of the nicotinamide head of the cofactor to the carbon atom of the ketone group, thus reducing the steroid substrate. The reduction reaction is completed by a proton transfer from Tyr55 residue, which acts as a proton donor, to the oxygen atom through a “push–pull” mechanism (Schlegel et al. 1998). Recent studies have suggested that the orbital overlapping produced by an optimal orientation of the reacting atoms could play a major quantitative role in the catalytic power of these enzymes (Mesecar et al. 1997; Heredia et al. 2003). They determined that this optimal orientation is obtained when the carbon atom of the target ketone group is maintained at about 3.0 Å of the C-4 of the pyridine head of the cofactor and when a line joining these two atoms makes an angle of about 100° with the pyridine head (Fig. 1C ▶).

Considering these recent findings—the pronounced induced-fit adaptations shown for at least one member of the AKR enzyme family and the precise localization of the ketone group to be reduced inside the active site—and taking advantage of ternary complex crystal structures we have determined so far for many AKR1C enzymes (Nahoum et al. 2001; Couture et al. 2003, 2004), we decided to revisit all these structures, especially those for which the steroid substrate was found in the steroid-binding cavity but at a nonproductive position (Nahoum et al. 2001; Couture et al. 2003, 2004). We wished to investigate some pending issues: (1) the mechanism by which h3α–HSD3 binds/transforms numerous structurally different molecules, which implies an ability to discriminate various ligands from different steroid hormone classes; (2) the orientation adopted by a steroid in a productive complex; and (3) the geometry of the residues that surround the ligand in the productive complexes.

Here we report the crystal structure of h3α–HSD3 in binary complex with NADP(H), a complex in which a citrate and an acetate molecule, added during the crystallization step, were found deeply bound inside the steroid-binding cavity. Comparison with the testosterone/acetate ternary complex structure reveals that loop A undergoes a slight contraction toward the citrate. Moreover, numerous residues found in the cavity, namely Val54, Val128, Ile129, Trp227, Leu306, and Leu308, alter the orientation of their side chains, narrowing the cavity. All these structural changes are reminiscent of the induced-fit mechanism previously proposed for the aldose reductase (Sotriffer et al. 2004). We also found that, in order to respect the orbital steering constraint, the optimal position of the ketone of a steroid during the reduction reaction is somewhat similar to that of the acetate molecule found in the active site. Manual positioning of 4-dione and DHT in an ideal productive ternary complex revealed that the nucleus of the steroid must be oriented toward the C-terminal domain (loop C) of the protein. To verify this, we proceeded to a leucine-scanning mutagenesis study on many residues of loop C, and found that the replacement of residue Arg301 or Arg304, located 10 Å from the catalytic Tyr55, significantly affects the activity of this enzyme. These results suggest that residues involved in the formation of a mature binding cavity in h3α–HSD3 differ according to the nature of the steroid substrate, the position of the ketone group on the steroid nucleus, and the activity exerted by this enzyme.

Results and Discussion

Overall structure of the h3α–HSD3 binary complex

The crystal used to determine the binary complex structure belonged to the R32 space group and contained two molecules per asymmetric unit (monomers A and B). The crystal structure of the h3α–HSD3/NADPH binary complex was refined to a crystallographic R-factor of 17.4% (Rfree=19.9%) at 1.9 Å of resolution. No major differences were found when both monomers were compared (root-mean-square deviation, calculated for 327 Cα atoms is 0.36 Å). The model obtained was of good quality since more than 99.7% of the dihedral angles of the main chain were found in the allowed region of the Ramachandran plot (Table 1). Residue Ser221 was found in the disallowed region due to a hydrogen bond between its nitrogen and the pyrophosphate group of NADPH. The protein was folded as a triose phosphate isomerase barrel motif [(α/β)8-barrel], in which eight β strands forming the cylindrical core of the barrel are surrounded by eight helices. The first 17 residues are folded as a two-strands anti-parallel beta-sheet that closes the bottom of the barrel. Finally, three large loops, named A (Phe118 to Asp143), B (Ser217 to Asp238), and C (Leu299 to Tyr323) are situated in a manner that creates a funnel-shaped pocket at the carboxy end of the β-strands, the top of the barrel (Fig. 2A ▶).

Table 1.

Data collection and refinement statistics (data of the last shell of resolution)

Data collection h3α-HSD3-NADP(H)
Unit cell dimensions (Å) a = 143.12; b = 143.12; c = 204.31
α = γ = 90°; β = 120°
Space group R32
No. of monomers per a/u 2
Resolution (Å) (last resolution shell) 1.90 (1.90;2.00)
Total no. of reflections 277,363 (33,473)
Unique reflections 59,186 (7945)
Completeness (%) 93.4 (88.9)
I/σI 17.8 (3.4)
Rmergea 0.063 (0.466)
Redundancy 4.7
Refinement
Reflections used (Rfree set) 56,844 (3025)
Rcrystb 0.174
Rfree 0.199
R.m.s. deviation from ideal bond length (Å)/angles(°) 0.006/1.4
No. of nonhydrogen atoms
Protein (A/B) 5183
    NADP 94
    Sulfate 10
    β-mercaptoethanol 4
    Ethylene glycol 8
    Citrate 13
    Acetate 12
    Water 477
Average B-factors (Å2)
    Protein 28.6
    NADP 33.0
    Acetate 36.7
    Citrate 43.1
    Water 39.9
PROCHECK
Allowed regions 91.7
Generously allowed regions 8.0
Disallowed regions 0.3
Open in a new tab

aRmerge = ∑∑i|I(h) − I(h)i|/∑∑iI(h), where I(h) is the mean intensity after rejections.

bRcryst = ∑|Fo − Fc|/∑|Fo, where Fo and Fc are the observed and calculated structure-factor amplitudes for the reflections with Miller indices h=(h, k, l). The free R-factor is calculated for a “test” set of reflections that was not included in atomic refinement (5%).

Figure 2.

Figure 2.

Figure 2.

Open in a new tab

Structure of h3α–HSD3 in binary complex with NADP(H) crystallized in the presence of citrate and acetate salts. (A) Overall view of a superposition of the h3α–HSD3 complex structures with testosterone (RCSB PDB entry, 1J96; black) and with citrate and acetate (monomer A) (RCSB PDB entry, 1XJB; gray). (B) Close-up stereo view of the citrate binding cavity of h3α–HSD3. The 2(Fo−Fc) electron density map for citrate (dark gray) and acetate (light gray) are respectively contoured at 0.9 σ and 1.0 σ level. (C) Schematic representation of interactions made by citrate and acetate (thick lines) within the binding cavity. Only residues that interact with ligands are shown. The dashed lines and the number in parentheses represent the hydrogen bonds and the H-bond distances, respectively. (D) Stereo view of the superposition of the h3α–HSD3 complex structures with citrate and acetate (monomer A; white) and with testosterone (gray). Only residues affecting the size of the steroid-binding cavity are depicted for both complexes.

Cofactor-binding site

During model building, a clearly visible Fo−Fc electron-density map corresponding to the cofactor was found. The NADP(H) is maintained in an extended conformation with the adenine part lying on the surface of the protein. The pyridine head of the cofactor is positioned in the center of the barrel forming the bottom of the steroid-binding cavity. This mode of binding of the cofactor seems highly conserved since it is also observed in an impressive number of AKR members for which the three-dimensional structure has been determined (Rondeau et al. 1992; Bennett et al. 1996; Khurana et al. 1998; Kavanagh et al. 2002; Kozma et al. 2002). Interestingly, in spite of our many attempts, it was impossible to crystallize the apo form of h3α–HSD3. Because no cofactor was added to the protein preparation during the purification and crystallization steps, we suppose that the NADP(H) found in all our crystal structures came from the Escherichia coli cells used to overproduce the h3α–HSD3 protein. The fact that the endogenous bacterial NADP(H) was kept by h3α–HSD3 through all the purification steps is probably an indication of the high affinity of this enzyme for its cofactor.

Steroid-binding site

During the refinement process, two unexpected positive peaks appeared inside the steroid-binding cavity of monomer A. Citrate and acetate molecules, added to the mother liquor during the crystallization step, were fitted into the Fo−Fc difference Fourier map and refined with the rest of the model (Fig. 2B ▶). These two molecules were surrounded by residues Tyr24, Ala25, Val54, Tyr55, Trp86, His117, Ile129, Asn167, Tyr216, His222, Glu224, Trp227, Leu306, Leu308, Ile310, and Phe311 (Fig. 2C ▶). Together, these residues form a narrow cavity where the acetate and the citrate are tightly bound. At the bottom of the cavity, the acetate molecule occupies the catalytic binding site of the h3α–HSD3 enzyme with the C=O of its carboxyl group mimicking the ketone group of a steroid substrate. Indeed, one of the oxygen atoms of this carboxylate group is involved in a hydrogen bond network that includes the OH group of Tyr55 (2.7 Å), the Nɛ atom of His117 (2.7 Å), in addition to the Nζ of Lys84 (located at 3.1 Å from the tyrosine hydroxyl group), while the carbon atom is located at 3.3 Å from the C-4 of the nicotinamide head of the cofactor, the hydride donor atom. The presence of an acetate molecule well stabilized inside the catalytic binding site of an enzyme of the AKR superfamily has been reported before (Nahoum et al. 2001; Lovering et al. 2004), but this is the first time it is seen orientated so as to interact with the catalytic residues (Tyr55 and His117) and, at the same time, participate in the ligand stabilization, here the citrate molecule. Indeed, one of the oxygen atoms of the acetate molecule carboxylate group is equidistant (2.5 Å) from two oxygen atoms (O2δ and O7δ) of the citrate molecule with which it could share a hydrogen atom (Fig. 2C ▶). In addition, two other oxygen atoms of the citrate molecule are involved in hydrogen bonds with residues of the h3α–HSD3 enzyme: O2δ with the side chain (Nɛ2) of residue His222 (2.6 Å) and O5δ with the hydroxyl group of Tyr24 (2.7 Å) and with the nitrogen atom of the indole ring of Trp227 (2.8 Å).

The numerous potential hydrogen bonds that stabilize the acetate and citrate in the substrate-binding cavity of h3α–HSD3 led us to think that these two molecules could probably interfere in the binding of the steroid, which normally forms only one hydrogen bond with the enzyme, and thus could be potential inhibitors of h3α–HSD3. We thus tested the capacity of the acetate and citrate to inhibit the 3α-reduction of DHT and found that none of them, alone or together, could measurably inhibit the h3α–HSD3 activity (data not shown). This result suggests that weaker and less specific forces such as van der Waals bonds or hydrophobic interactions, which are well known to mediate the binding of steroids and other structurally different ligands by h3α–HSD3, are probably much more important than anticipated, provided of course that the plasticity of the substrate pocket is sufficient to insure the steric complementarity between the enzyme and its ligands.

Citrate bound in a steroid-binding cavity suggests a ligand induced-fit mechanism

The side chains of residues Tyr55 and His117 form an anionic binding site that could bind molecules bearing a negative charge. This, combined with the fact that the substrate-binding site is funnel shaped and largely open on the protein surface, probably explains why human 3α– HSD3 can bind and transform numerous molecules that vary in length and shape. Indeed, it has been shown that this enzyme can bind acetate (Nahoum et al. 2001), citrate (this paper), indanol (Matsuura et al. 1997), steroids (Dufort et al. 1996; Penning et al. 2000), and PAHs (Burczynski et al. 1998). To understand the mechanism underlying this plasticity we have superposed and compared our model (h3α–HSD3/NADP(H)/Citrate/Acetate) with another h3α–HSD3 complex structure (h3α–HSD3/NADP(H)/Testosterone/Acetate). We first observed that a short segment of loop A, between residues Pro124 and Phe139, makes a movement against the citrate molecule that slightly closes the top of the binding cavity (Fig. 2A ▶). Loop A is one of the three loops involved in the formation of the testosterone-binding cavity in rat and human 3α–HSD enzymes. It has been previously suggested for HSDs of the AKR superfamily that this loop plays a key role in the recognition, binding, and transformation of their substrates since a chimera rat 3α–HSD having the loop A found in human 20α–HSD sequence acquires 17β- and 20α–HSD activities on testosterone and progesterone, respectively (Ma and Penning 1999). Among conformational changes observed in this segment, the backbone of residues Val128 and Ile129 moves 0.8 Å toward the citrate, contributing to a reduction of the cavity opening. These two residues are fairly important since they directly interact with and participate in the binding of testosterone and 20α-OHProg by h3α–HSD3 and h20α–HSD, respectively (Nahoum et al. 2001; Couture et al. 2003). In addition, previous studies on h20α–HSD enzymes revealed that the side chain of another residue of loop A, Glu127, rotates upon binding of the steroid and slightly closes the top of the cavity, preserving the hydrophobic environment of the steroid-binding site (Couture et al. 2003). All these observations support the importance of the role played by loop A in the formation of the ligand-binding cavity and in the plasticity of this site. In comparison with loop A, the main chains of loops B and C do not undergo a significant structural shift (Fig. 2A ▶) but close inspection of the super-imposed binding site regions reveals that some residues that belong to these loops (Trp227, Leu306, and Leu308) or which delineate the cavity (Val54) make important conformational changes and become closer to the citrate molecule (Fig. 2D ▶).

Compared with the model of h3α–HSD3 in ternary complex with testosterone/acetate, the side chain of residue Val54 in the present model undergoes a rotation of 180° around its Cβ, placing its Cγ atom toward the binding cavity (Fig. 2D ▶), probably to stabilize the citrate molecule. Interestingly, the importance of this residue for the selectivity of HSDs of the AKR superfamily has been previously demonstrated by site-directed mutagenesis experiments (Matsuura et al. 1997; Couture et al. 2003, 2004). For example, Matsuura et al. (1997) demonstrated that the replacement of residue Val54 by a leucine (residue found in the h20α–HSD sequence) confers to the mutated h3α–HSD3 almost the same catalytic properties as h20α–HSD. Now, our structural results tend to demonstrate that residue Val54, through its capacity to adopt different conformations, is also important for the plasticity of the ligand-binding cavity of h3α–HSD3.

We also observed that the indole ring of residue Trp227 moves toward the citrate molecule, where it is stabilized through a hydrogen bond between the O5δ and the nitrogen atom of its indole ring (Fig. 2D ▶). In the ternary complex structure of h3α–HSD3 with testosterone, this residue makes a π-stacking interaction with cycles B and C of testosterone and greatly contributes to the formation of a mature steroid-binding cavity (Nahoum et al. 2001). In r3α–HSD, the replacement of Trp227 has been shown to affect the capacity of this enzyme to bind and transform testosterone and progesterone (Heredia et al. 2004) and, more importantly, its replacement can affect the activity on decalone and 9,10-phenanthrenequinone (Fig. 1B ▶). Combined with the h3α–HSD3/NADP(H)/Acetate/Citrate ternary complex structure study, these results suggest that upon the binding of a non-steroid-shaped ligand, this residue is of great importance for the optimal formation of a mature cavity and for the plasticity of this enzyme.

Finally, the side chains of residues Leu306 and Leu308 also rotate 80° around their Cγ toward the citrate molecule (Fig. 2D ▶). This “leucine-flip” thus seems to play a role in the plasticity of this cavity. Similarly, a recent study on aldose reductase has pointed out the Leu300, a residue found in loop C (also called specificity pocket), as an important factor in the modulation of the shape of this cavity. It appears that the conformation adopted by the side chain of this residue with the backbone shift observed for residues Ala299 and Leu300 are directly related to the shape of the bound molecule (Sotriffer et al. 2004). Our finding indicating that the conformational change undergone by Leu306 and Leu308 reduced the size of the binding cavity of h3α–HSD3 upon binding of the citrate is thus in total agreement with the model proposed for the aldose reductase, even if the contraction movement observed here does not affect the backbone position of these two residues.

To ascertain that the size of the h3α–HSD3 substrate cavity really changes upon binding of its structurally different substrates, and thus confirm our first visual finding, we decided to precisely measure the volume of this surface groove in the presence of the acetate/citrate (monomer A) and compare it with that of the same enzyme in complex with testosterone/acetate (Nahoum et al. 2001) (RCSB PDB entry, 1J96) and ursodeoxycholate (Jin et al. 2001) (RCSB PDB entry, 1IHI). We thus found that the substrate cavity volume was directly related to the size of the bound ligand. Indeed, the cavity volume for the acetate/citrate (Fig. 3A ▶) was 704 Å3 compared to 776 Å3 for testosterone/acetate (Fig. 3B ▶), in agreement with the size of these molecules (236 Å3 vs. 333 Å3). Interestingly, the cavity volume in the ursodeoxycholate complex (734 Å3) is smaller than the testosterone cavity, although ursodeoxycholate (362 Å3) is slightly bigger (Fig. 3C ▶). This apparent contradiction could be easily explained by the difference in the binding mode adopted by these two steroidal substrates. The ursodeoxycholate is bound in such a manner that its long and thin 17β-(1-methyl-3- carboxypropyl) arm is positioned close to the catalytic site of the enzyme, pushing almost half of the steroidal nucleus (cycle A and a large part of cycle B) out of the cavity (Fig. 3D ▶). In this position, the ursodeoxycholate is in fact a “smaller” ligand than testosterone (248 Å3 vs. 277 Å3), which is almost totally inside the cavity.

Figure 3.

Figure 3.

Open in a new tab

Variation in the surface of the steroid-binding cavity of h3α–HSD3 in complex with structurally different ligands. Surfaces of the steroid-binding cavity of h3α–HSD3 in complex with (A) acetate/citrate, (B) testosterone/acetate (Nahoum et al. 2001), and (C) ursodeoxycholate (Jin et al. 2001) have been calculated with the program VOIDOO (Kleywegt 1994). (D) Superposition of various ligands (testosterone, blue, and ursodeoxycholate, green) in complex with the h3α–HSD3 enzyme to compare their mode of binding. Acetate and citrate molecules, which are profoundly engulfed inside the steroid-binding cavity, are not visible from this angle. Figures have been rendered with Pov-Ray (Kaplan and Littlejohn 2001).

Together, all these conformational changes observed for many residues of loops A, B, and C, which contribute to the modification of the overall shape of the substrate-binding cavity in the presence of structurally different substrates (Figs. 2 ▶,3 ▶), are reminiscent of the induced-fit mechanism previously proposed for the aldose reductase (Sotriffer et al. 2004). Our results strongly suggest that this mechanism can be applied to h3α–HSD3 and probably to many other members of the AKR1C enzymes subfamily.

What is the real productive ternary complex?

Residues found in loops A, B, and C are of great importance in the plasticity of the steroid-binding cavity of h3α–HSD3, a characteristic that allows this enzyme to bind molecules with various shapes and structures. In addition, these loops must optimally orient the incoming molecules to promote their transformation. Unfortunately, each HSD member of the AKR superfamily for which the ternary complex structure has been solved shows its steroid substrate in a nonproductive position, i.e., in a position (or in an orientation) that does not permit the oxido-reduction reaction to take place (Bennett et al. 1997; Nahoum et al. 2001; Couture et al. 2004). According to recent reports, HSDs of the AKR superfamily would transform steroids through an orbital steering mechanism where the optimal distance and angle between the carbon atom of the substrate ketone group and C4 of the nicotinamide would be around 3.0 Å and 100°, respectively (Heredia et al. 2003). Close inspection of the h3α–HSD3 structure with the citrate and acetate molecules bound in the substrate cavity (monomer A) reveals that this optimal position corresponds exactly to the carbon atom of the carboxylate group of the acetate molecule.

To better understand how h3α–HSD3 can bind androgen in a productive complex, we manually modeled steroid molecules (4-dione and DHT) in a position that respects the orbital steering constraints. When these two structurally similar steroids were modeled with their reactive group in an optimal position (the C3 for the DHT and C17 for the 4-dione), their steroid nuclei were found in different positions (Fig. 4 ▶). For DHT, the nucleus of the steroid was roughly directed toward loop C and clashed with the indole ring of Trp227, a residue belonging to loop B and that is highly conserved among AKR members. Interestingly, we had previously found that Trp227, with many other residues of loop B, can undergo significant conformational changes to promote the binding of the steroid in the crystal structure of rb20α–HSD in ternary complex with testosterone (Couture et al. 2004). In fact, we have found that Trp227 with some other residues of loop B that point toward the steroid-binding cavity in the rb20α–HSD binary complex, probably thus maintaining the hydrophobic character of the steroid cavity, are at the surface of the protein and exposed to the solvent upon binding of testosterone. This relaxation movement results in an enlargement of the steroid-binding cavity and in the mobilization of some residues belonging to the C-terminal part of the protein (loop C) to create the mature steroid-binding site (Couture et al. 2004). Crystal structure determination of r3α–HSD in ternary complex with testosterone also reveals that the indole ring of Trp227 can rotate by 180° around its Cγ, a movement that drastically changes the shape of the cavity and allows DHT to bind in a productive complex.

Figure 4.

Figure 4.

Open in a new tab

Stereo view of the hypothetic orientation of two different androgens for their optimal reduction by h3α–HSD3. The target ketone group of 4-dione (light gray) and DHT (dark gray) were manually superimposed on the acetate molecule. Only residues in proximity of the steroid are depicted. Figure was created with POVScript (Fenn et al. 2003) and rendered with Pov-Ray (Kaplan and Littlejohn 2001).

On the other hand, when 4-dione was superimposed on the acetate molecule with its C17-ketone group in a position that allowed its 17β-reduction, the body of the steroid occupied a space delineated by the side chain of residues Leu306 and Leu308 with its C3-ketone group lying down near the side chain of residue Arg304. When modeled in a productive ternary complex, DHT and 4-dione do not occupy the same position in the binding cavity but both steroids interact with residues of the C-terminal end of the protein (Fig. 4 ▶). These hypothetical orientations are in total agreement with previous site-directed mutagenesis studies that have demonstrated that the C-terminal domain of HSDs of the AKR superfamily protein is responsible for the selectivity of these enzymes. For example, replacement of Val306 by a phenylalanine residue in rb20α–HSD greatly impairs the activity of this enzyme toward DHT and 4-dione (Couture et al. 2004). Likewise, the exchange of loop C sequence between h3α–HSD1 and h20α–HSD gives to the chimerical enzyme 20α–HSD3α–HSD1 (h20α–HSD with 39 C-terminal residues of h3α–HSD type 1) new properties and transforms it into a potent reductase enzyme for C19-steroids (Matsuura et al. 1998). Together, all these observations strongly support the hypothesis that the C-terminal end of these enzymes is of paramount importance in the discrimination of the incoming steroid substrates.

A C-terminal basic motif, a concerted way to bind androgen?

We thus found that, in order to respect the orbital steering constraints, the steroid should be maintained approximately parallel to the pyridine head of the cofactor with the body of the steroid pointing toward the C-terminal part of the protein. To evaluate the role of residues of this domain in steroid discrimination and binding, we undertook a systematic leucine scanning mutagenesis study. We found that the replacement of two residues, Arg301 and Arg304, leads to a drastic effect on the capacity of h3α– HSD3 to transform androgen. Indeed, when the kinetic constants for the reduction of DHT were compared with the wild-type h3α–HSD3 (Table 2), we found that Km for the 3α-reduction of DHT was barely affected, whereas kcat was 60- and 227-fold lower for the Arg301Leu and Arg304Leu mutants, respectively. For the 17β-reduction of 4-dione, we found that the replacement of Arg301 affects the Km and the kcat values by 3.8- and 7.3-fold, respectively. More strikingly, we found that the 17β-activity on 4-dione was totally abolished for the Arg304Leu mutant enzyme. No activity was detectable even by using radio-activity, the most sensitive technique to detect the transformation of a steroid (Luu-The et al. 1995).

Table 2.

Steady-state kinetic parameters for steroid reduction catalyzed by homogeneous recombinant h3α-HSD3s

Activity measured (steroid used) Enzyme Km (μM) kcat (min−1) kcat/Km (min−1μM−1)
3α-activity (DHT) WTa 5.7 ± 0.8 1338.51 234.82
Arg301Leu 5.9 ± 0.7 22.05 4.33
Arg304Leu 12.0 ± 1.6 5.88 0.49
17β-activity (4-dione) WT 10.5 ± 1.8 19.88 1.88
Arg301Leu 38.8 ± 1.4 2.68 0.07
Arg304Leu Ndb Nd —
Open in a new tab

a (WT) Wild type.

b (Nd) Nondetectable activity.

It has been shown previously that some residues found in the C-terminal domain of AKR superfamily members are involved in the binding of the cofactor (Rees-Milton et al. 1998). To verify that our mutant enzymes did not compromise the binding of NADP(H), fluorescence titration experiments were performed. Compared to wild-type h3α–HSD3, the Kd values for NADPH were almost identical for the Arg301Leu and Arg304Leu mutant enzymes, demonstrating that the replacement of this basic motif does not impair the binding of the cofactor (Table 3). We also verified that the formation of a mature binding cavity was not affected by the replacement of the Arg301 and Arg304 residues. We thus performed enzymatic experiments on 1-indanone, a molecule that possesses two cycles and a single ketone to be reduced (Fig. 1B ▶) (Matsuura et al. 1997). The Arg301Leu and Arg304Leu mutants showed nearly the same catalytic activity toward 1- indanone as the wild-type enzyme (Table 4). These results strongly suggest that Arg304 is essential in the formation of a mature binding cavity specific for 4-dione and/or for the 17β-reduction activity of h3α–HSD3, whereas its role in the 3α-reduction of DHT is less important. More importantly, they also demonstrate that the residues involved in the formation of a mature cavity for 4-dione are not necessarily the same as those having the same role for a structurally similar C19-steroid, DHT.

Table 3.

Effect of mutations on the binding of NADPH by homogeneous recombinant h3α-HSD3s

Enzyme Kd (μM)
Wild type 0.88 ± 0.23
Arg301Leu 0.84 ± 0.25
Arg304Leu 1.74 ± 0.73
Open in a new tab

Table 4.

Transformation of 1-indanone by homogeneous recombinant h3α-HSD3s

Enzyme Specific activity (μmol/min per μg)
Wild type 1.59 ± 0.02
Arg301Leu 1.22 ± 0.04
Arg304Leu 1.46 ± 0.01
Open in a new tab

Surprisingly, Arg301 and Arg304 residues are found at the surface of the protein and are roughly located 10 Å from the catalytic Tyr55, a distance corresponding approximately to the length of a steroid. As mentioned above, numerous reports highlight the importance of the carboxy-terminal end (loop C) of AKR enzymes in the specificity for their ligand. We have already reported that the replacement of residue Arg304 in h20α–HSD severely affected the Km value of this enzyme for the 20α-reduction of progesterone (Couture et al. 2003). Moreover, site-directed mutagenesis studies on human aldehyde reductase, another representative member of the AKR superfamily, have demonstrated that Arg311 is likely involved in the binding of aldehydes containing a carboxyl group and being longer than 7–8 Å (Barski et al. 1996). A superposition of the C-terminal loop sequence for HSDs of the AKR superfamily, the less conserved region among the AKR1C subfamily, reveals that Arg301 and Arg304 are conserved in 10 HSDs out of 12 (Fig. 5 ▶). Except for the AKR1C11 sequence in which a serine residue is found at position 301, all the residues at positions 301 and 304 are arginines, lysines, or histidines, three residues possessing similar positive charges at neutral pH. Nahoum et al. (2003) have reported that the structural elements that compose the architecture of the estrogen-binding site (C18-steroid) in the estrogen-specific proteins should include a proton acceptor/donor residue interacting with the steroid O3 and a proton donor interacting with O17. Human 3α–HSD3 possesses these structural characteristics since a basic motif, the Arg301 and Arg304 residues acting as the proton donors, is at one end of the steroid molecule, whereas the catalytic Tyr55, the proton acceptor/donor residue, is at the other end.

Figure 5.

Figure 5.

Open in a new tab

Sequence alignment of the C-terminal end of 12 HSDs member of the AKR superfamily. The alignment of the C-terminal sequences (residues 291 to 323 or 322 for AKR1C9) was done with CLUSTALW. Residues 301 and 304 in the human 3α–HSD type 3 sequence are boxed while corresponding residues in other AKR members are indicated in bold characters. Abbreviated names for HSDs are AKR1C1, human 20α–HSD; AKR1C2, human 3α–HSD type 3; AKR1C3, human 17β–HSD type 5; AKR1C4, human 3α–HSD type 1; AKR1C5, rabbit 20α–HSD; AKR1C6, mouse 17β–HSD; AKR1C7, bovine prostaglandin F synthase 1; AKR1C8, rat 20α–HSD; AKR1C9, rat 3α–HSD; AKR1C10, frog rho-crystallin; AKR1C11, prostaglandin F synthase 2; and AKR1C18, mouse 20α–HSD.

Combined with the orbital steering constraints proposed by Heredia et al. (2003), our observations on the h3α–HSD3 enzyme suggest that, for triggering of the reaction, 4-dione is likely maintained parallel to the pyridine head of the cofactor in order to optimally orient its C17-ketone group, with respect to the 4-proR hydride of the cofactor. In this position, the nucleus of the steroid lies down toward the C loop of the enzyme where the guanido group of Arg304 interacts with the steroid, either directly by forming an ionic bond with the C3-ketone group or indirectly by the formation of a positively charged environment essential for perfect positioning of the steroid substrate. It thus appears that, together, these structural characteristics are essential for the formation of the mature steroid-binding cavity and are the basis of the specificity of this enzyme for a steroid, a subgroup of steroids as well as other molecules.

These observations also have important physiological relevance since, in the prostate tissues, h3α–HSD3 is able to inactivate the most potent androgen, DHT, into a weak androgen, 3α-diol, while this enzyme is also involved in the formation of testosterone, a potent androgen, through its 17β-reduction activity on 4-dione. If the residues involved in the binding of DHT and 4-dione differ, it is tempting to speculate that it would be feasible to synthesize an inhibitor specific for the 17β-reduction of 4-dione that does not affect, or very weakly affects, the 3α-inactivation of DHT. In this regard, it will be essential to take into account that some residues of the h3α–HSD3 binding cavity involved in the binding and maintaining of the steroid substrates can undergo significant conformational changes, in part dictated by the shape of the steroid itself. Similar to the “ligand induced-fit mechanism” reported for the aldose reductase, this binding site adaptation is of major importance in the efficiency of inhibitors targeted against this enzyme.

Materials and methods

Material

NADPH, 1-indanone, and other chemical products were purchased from Sigma-Aldrich Canada. Unlabeled steroids and [14C]-labeled steroids were purchased from Steraloids and PerkinElmer, respectively.

Site-directed mutagenesis

Arg301Leu and Arg304Leu mutations were created using Quick- Change Site-Directed Mutagenesis Kit (Stratagene) with these forward primers: Arg301Leu, AAAGCCATAGATGGCC TAAACCTAAATGTGCGATATTTGACCCTTGATAT and Arg304Leu, CATAGATGGCCTAAACAGAAATGTGCTAT ATTTGACCCTTGATATTTTTGATGGCC. To confirm the presence of the desired mutation and to ascertain that no other mutations had occurred, the complete coding region of the mutated cDNA was sequenced with a T7 sequencing kit (Amersham Bioscience).

Overexpression and purification of wild-type and mutant h3α–HSD3

Briefly, the cDNA encoding h3α–HSD3 was subcloned in a pGEX vector (Amersham Bioscience), overproduced in E. coli BL21(DE3)pLysS as a fusion protein with glutathione-S transferase (GST) and purified by a combination of affinity, anionic, and size-exclusion chromatographies, yielding about 5 mg of a highly purified enzyme preparation per 100 mL of cell culture. Mutated h3α–HSD3s were overproduced and purified to homogeneity as described for the wild-type enzyme.

Determination of kinetic parameters

Reduction reactions on [14C]-steroids were performed as previously described, with minor modifications (Luu-The et al. 1995; Couture et al. 2002). Kinetic parameters (Km and kcat/Km) were calculated using the EnzFitter program (Biosoft). Because some HSDs of the AKR superfamily have been shown to lose their activity when they are purified to homogeneity (Dufort et al. 1999), a kinetic characterization was performed with DHT and 4-dione, two natural ligands of this enzyme, to verify that the purification steps did not affect the enzyme activity. Within the experimental error, Km values obtained for the 3α-reduction of the DHT (5.7 ± 0.8 μM) and for the 17β-reduction of 4-dione (10.5 ± 1.8 μM) (Table 2) were very similar to those published earlier (1.4 ± 0.66 μM and 9.73 ± 1.8 μM [Penning et al. 2000; Dufort et al. 2001]), indicating that our h3α–HSD3 preparation was suitable for subsequent experiments.

Reduction of 1-indanone was measured at room temperature in a 50-μL system containing 100 μM 1-indanone in 2% (v/v) ethanol, 200 μM NADPH, and 100 mM KH2PO4–K2HPO4 (pH 7.5). Reactions were started when 10 μg of homogenous recombinant type 3 h3α–HSDs were added to the system in which 1-indanone transformation was monitored with a Beckman DU-7400 spectrophotometer by measuring the rate of change of absorbance of NADPH (ɛ = 6270 cm−1 M−1) in a 1-cm light path.

Fluorescence

Measurements of the fluorescence of wild-type, Arg301Leu, and Arg304Leu enzymes were conducted on an SLM 8000 fluorometer. Scans of the fluorescence emission spectra were performed at 25°C in a 200-μL microcell containing 2 μM of the purified enzyme, 50 mM Na2HPO4/NaH2PO4 buffer (pH 7.5), 20% glycerol, 1 mM EDTA, and variable concentrations of NADPH (0–4000 nM). Tryptophan fluorescence was detected using 290 nm as the excitation wavelength, while the emission was scanned from 300 to 600 nm (at 120 nm/min). Emission and excitation bandpasses were both set to 4 nm.

Crystallization

A homogenous preparation of h3α–HSD3 enzyme freshly purified was used for crystallization experiments using the hanging-drop vapor diffusion method at 4°C. Crystals were obtained by mixing equal volumes of h3α–HSD3 concentrated to 14 mg/mL and of mother liquor (0.1 M NaCitrate at pH 6.5, 0.1 M NH4Ac, and 24%–30% PEG4000). Crystal used for X-ray diffraction experiments was directly soaked into the cryopreserving solution (20% of ethylene glycol added to the mother liquor) and flash-cooled at 100 K in a nitrogen gas stream.

Structure determination and refinement

X-ray diffraction oscillation images of h3α–HSD3/NADPH binary complex were recorded on an R-AXIS IIc image-plate detector mounted on a Rigaku RU-200 rotating-anode generator equipped with focusing mirrors (Rigaku MSC). Data were integrated and scaled using the December 2000 version of the XDS program (Kabsch 1993) and the crystal structure was directly determined by performing a rigid body refinement using the Refmac program (Murshudov et al. 1997). The h3α–HSD3 structure (PDB code 1J96; Nahoum et al. 2001) was used as a model and molecular replacement solution was calculated with two molecules per asymmetric unit. Crystallographic statistics are shown in Table 1. The refinement procedure was performed using CNS (Brunger et al. 1998) and Refmac (Murshudov et al. 1997) programs. The initial model issued from rigid body refinement was submitted to a cycle of simulated annealing at 3000 K followed by energy minimization cycles. After electronic density maps calculations and manual rebuilding using O (Jones et al. 1991), NADPH was added to the model. Atomic models were then refined by simple energy minimization followed by isotropic B-factors refinement (restrained, individual B-factor refinement) and corrected by manual rebuilding. Missing parts of the model, sulfate ion, β-mercaptoethanol, ethylene glycol, citric acid, acetate, and water molecules were progressively added during the refinement procedure. Finally, the quality of the model was verified with PROCHECK (Table 1) (Laskowski et al. 1993).

The substrate cavity volumes of h3α–HSD3 in the presence of its different ligands were measured with the program VOIDOO (Kleywegt 1994) using program default parameters. The volume values obtained this way were finally corrected using the equation given by Chakravarty et al. (2002).

Accession numbers

Atomic coordinates of this binary complex have been deposited in the Protein Data Bank with the accession code 1XJB.

Acknowledgments

This work was supported by the Medical Research Council (MRC) of Canada and Endorecherche Inc. Jean-François Couture and Pierre-Luc Côté are the recipients, respectively, of a doctoral scholarship provided by the Laval University Foundation, Hydro-Québec and a master’s degree scholarship provided by the Natural Sciences and Engineering Research Council of Canada. The authors thank Sylvie Méthot for careful reading of the manuscript.

Abbreviations

  • h3α–HSD3, human 3α-hydroxysteroid dehydrogenase type 3

  • h20α–HSD, human 20α-hydroxysteroid dehydrogenase

  • h3α– HSD1, human 3α-hydroxysteroid dehydrogenase type 1

  • 17β–HSD, 17β-hydroxysteroid dehydrogenase

  • NADP(H), reduce form of nicotinamide adenine dinucleotide phosphate

  • GST, glutathione sulfotransferase

  • PDB, Protein Data Bank

  • AKR, aldo-keto reductase

  • progesterone, 5-pregnen- 3,20-dione

  • 4-dione, 4-androsten-3,20-dione

  • PAH, polycyclic aromatic hydrocarbon

  • EDTA, ethylene diamine tetraacetic acid.

Article and publication are at http://www.proteinscience.org/cgi/doi/10.1110/ps.051353205.

References

  1. Barski, O.A., Gabbay, K.H., and Bohren, K.M. 1996. The C-terminal loop of aldehyde reductase determines the substrate and inhibitor specificity. Biochemistry 35 14276–14280. [DOI] [PubMed] [Google Scholar]
  2. Bennett, M.J., Schlegel, B.P., Jez, J.M., Penning, T.M., and Lewis, M. 1996. Structure of 3 α-hydroxysteroid/dihydrodiol dehydrogenase complexed with NADP+. Biochemistry 35 10702–10711. [DOI] [PubMed] [Google Scholar]
  3. Bennett, M.J., Albert, R.H., Jez, J.M., Ma, H., Penning, T.M., and Lewis, M. 1997. Steroid recognition and regulation of hormone action: Crystal structure of testosterone and NADP+ bound to 3 α-hydroxysteroid/dihydrodiol dehydrogenase. Structure 5 799–812. [DOI] [PubMed] [Google Scholar]
  4. Brunger, A.T., Adams, P.D., Clore, G.M., Delano, W.L., Gros, P., Grosse-Kunstleve, R.W., Jiang, J.-S., Kuszewski, J., Nilges, M., Pannu, N.S., et al. 1998. Crystallography & NMR system: A new software system for macromolecular structure determination. Acta Crystallogr. D Biol. Crystallogr. 54 905–921. [DOI] [PubMed] [Google Scholar]
  5. Burczynski, M.E., Harvey, R.G., and Penning, T.M. 1998. Expression and characterization of four recombinant human dihydrodiol dehydrogenase isoforms: Oxidation of trans-7, 8-dihydroxy-7,8-dihydrobenzo[a]pyrene to the activated o-quinone metabolite benzo[a]pyrene-7,8-dione. Biochemistry 37 6781–6790. [DOI] [PubMed] [Google Scholar]
  6. Burczynski, M.E., Sridhar, G.R., Palackal, N.T., and Penning, T.M. 2001. The reactive oxygen species- and Michael acceptor-inducible human aldo-keto reductase AKR1C1 reduces the α,β-unsaturated aldehyde 4-hydroxy- 2-nonenal to 1,4-dihydroxy-2-nonene. J. Biol. Chem. 276 2890–2897. [DOI] [PubMed] [Google Scholar]
  7. Chakravarty, S., Bhinge, A., and Varadarajan, R. 2002. A procedure for detection and quantitation of cavity volumes proteins. Application to measure the strength of the hydrophobic driving force in protein folding. J. Biol. Chem. 277 31345–31353. [DOI] [PubMed] [Google Scholar]
  8. Couture, J.F., Cantin, L., Legrand, P., Luu-The, V., Labrie, F., and Breton, R. 2002. Expression, crystallization and preliminary X-ray analysis of human and rabbit 20α-hydroxysteroid dehydrogenases in complex with NADP(H) and various steroid substrates. Acta Crystallogr. D Biol. Crystallogr. 58 135–139. [DOI] [PubMed] [Google Scholar]
  9. Couture, J.F., Legrand, P., Cantin, L., Luu-The, V., Labrie, F., and Breton, R. 2003. Human 20α-hydroxysteroid dehydrogenase: Crystallographic and site-directed mutagenesis studies lead to the identification of an alternative binding site for C21-steroids. J. Mol. Biol. 331 593–604. [DOI] [PubMed] [Google Scholar]
  10. Couture, J.F., Legrand, P., Cantin, L., Labrie, F., Luu-The, V., and Breton, R. 2004. Loop relaxation, a mechanism that explains the reduced specificity of rabbit 20α-hydroxysteroid dehydrogenase, a member of the aldo-keto reductase superfamily. J. Mol. Biol. 339 89–102. [DOI] [PubMed] [Google Scholar]
  11. DelaFuente, G. and Sols, A. 1970. The kinetics of yeast hexokinase in the light of the induced fit involved in the binding of its sugar substrate. Eur. J. Biochem. 16 234–239. [DOI] [PubMed] [Google Scholar]
  12. DelaFuente, G., Lagunas, R., and Sols, A. 1970. Induced fit in yeast hexokinase. Eur. J. Biochem. 16 226–233. [DOI] [PubMed] [Google Scholar]
  13. Dufort, I., Soucy, P., Labrie, F., and Luu-The, V. 1996. Molecular cloning of human type 3 3 α-hydroxysteroid dehydrogenase that differs from 20 α-hydroxysteroid dehydrogenase by seven amino acids. Biochem. Biophys. Res. Commun. 228 474–479. [DOI] [PubMed] [Google Scholar]
  14. Dufort, I., Rheault, P., Huang, X.F., Soucy, P., and Luu-The, V. 1999. Characteristics of a highly labile human type 5 17β-hydroxysteroid dehydrogenase. Endocrinology 140 568–574. [DOI] [PubMed] [Google Scholar]
  15. Dufort, I., Labrie, F., and Luu-The, V. 2001. Human types 1 and 3 3 α-hydroxysteroid dehydrogenases: Differential lability and tissue distribution. J. Clin. Endocrinol. Metab. 86 841–846. [DOI] [PubMed] [Google Scholar]
  16. Fenn, T.D., Ringe, D., and Petsko, G.A. 2003. POVScript+: A program for model and data visualization using persistence of vision ray-tracing. J. Appl. Crystallogr. 36 944–947. [Google Scholar]
  17. Heredia, V.V., Kruger, R.G., and Penning, T.M. 2003. Steroid-binding site residues dictate optimal substrate positioning in rat 3α-hydroxysteroid dehydrogenase (3α–HSD or AKR1C9). Chem. Biol. Interact. 143–144 393–400. [DOI] [PubMed] [Google Scholar]
  18. Heredia, V.V., Cooper, W.C., Kruger, R.G., Jin, Y., and Penning, T.M. 2004. Alanine scanning mutagenesis of the testosterone binding site of rat 3α-hydroxysteroid dehydrogenase demonstrates contact residues influence the rate-determining step. Biochemistry 43 5832–5841. [DOI] [PubMed] [Google Scholar]
  19. Hoog, S.S., Pawlowski, J.E., Alzari, P.M., Penning, T.M., and Lewis, M. 1994. Three-dimensional structure of rat liver 3 α-hydroxysteroid/dihydrodiol dehydrogenase: A member of the aldo-keto reductase superfamily. Proc. Natl. Acad. Sci. 91 2517–2521. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Jin, Y., Stayrook, S.E., Albert, R.H., Palackal, N.T., Penning, T.M., and Lewis, M. 2001. Crystal structure of human type III 3α-hydroxysteroid dehydrogenase/bile acid binding protein complexed with NADP(+) and ursodeoxycholate. Biochemistry 40 10161–10168. [DOI] [PubMed] [Google Scholar]
  21. Jones, T.A., Zou, J.Y., Cowan, S.W., and Kjeldgaard, M. 1991. Improved methods for the building of protein models in electron density maps and the location of errors in these models. Acta Crystallogr. A 47 110–119. [DOI] [PubMed] [Google Scholar]
  22. Kabsch, W. 1993. Automatic processing of rotation diffraction data from crystals of initially unknown symmetry and cell constants. J. Appl. Crystallogr. 26 795–800. [Google Scholar]
  23. Kaplan, W. and Littlejohn, T.G. 2001. Swiss-PDB Viewer (Deep View). Brief Bioinform. 2 195–197. [DOI] [PubMed] [Google Scholar]
  24. Kavanagh, K.L., Klimacek, M., Nidetzky, B., and Wilson, D.K. 2002. The structure of apo and holo forms of xylose reductase, a dimeric aldo-keto reductase from Candida tenuis. Biochemistry 41 8785–8795. [DOI] [PubMed] [Google Scholar]
  25. Khurana, S., Powers, D.B., Anderson, S., and Blaber, M. 1998. Crystal structure of 2,5-diketo-D-gluconic acid reductase A complexed with NADPH at 2.1-A resolution. Proc. Natl. Acad. Sci. 95 6768–6773. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Klebe, G., Kramer, O., and Sotriffer, C. 2004. Strategies for the design of inhibitors of aldose reductase, an enzyme showing pronounced induced-fit adaptations. Cell. Mol. Life Sci. 61 783–793. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kleywegt, G.J. 1994. Detection, delineation, measurement and display of cavities in macromolecular structures. Acta Crystallogr. D Biol. Crystallogr. 50 178–185. [DOI] [PubMed] [Google Scholar]
  28. Kozma, E., Brown, E., Ellis, E.M., and Lapthorn, A.J. 2002. The crystal structure of rat liver AKR7A1. A dimeric member of the aldo-keto reductase superfamily. J. Biol. Chem. 277 16285–16293. [DOI] [PubMed] [Google Scholar]
  29. Kraulis, P.J. 1991. MOLSCRIPT: A program to produce both detailed and schematic plots of protein structures. J. Appl. Crystallogr. 24 946–950. [Google Scholar]
  30. Laskowski, R.A., MacArthur, M.W., Moss, D.S., and Thornton, J.M. 1993. PROCHECK: A program to check the stereochemical quality of protein structures. J. Appl. Crystallogr. 26 283–291. [Google Scholar]
  31. Lovering, A.L., Ride, J.P., Bunce, C.M., Desmond, J.C., Cummings, S.M., and White, S.A. 2004. Crystal structures of prostaglandin D(2) 11-ketoreductase (AKR1C3) in complex with the nonsteroidal anti-inflammatory drugs flufenamic acid and indomethacin. Cancer Res. 64 1802–1810. [DOI] [PubMed] [Google Scholar]
  32. Luu-The, V., Zhang, Y., Poirier, D., and Labrie, F. 1995. Characteristics of humantypes1, 2 and 317β-hydroxysteroid dehydrogenase activities: Oxidation/reduction and inhibition. J. Steroid Biochem. Mol. Biol. 55 581–587. [DOI] [PubMed] [Google Scholar]
  33. Ma, H. and Penning, T.M. 1999. Conversion of mammalian 3α-hydroxysteroid dehydrogenase to 20α-hydroxysteroid dehydrogenase using loop chimeras: Changing specificity from androgens to progestins. Proc. Natl. Acad. Sci. 96 11161–11166. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Matsuura, K., Deyashiki, Y., Sato, K., Ishida, N., Miwa, G., and Hara, A. 1997. Identification of amino acid residues responsible for differences in substrate specificity and inhibitor sensitivity between two human liver dihydrodiol dehydrogenase isoenzymes by site-directed mutagenesis. Biochem. J. 323 (Pt. 1) 61–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Matsuura, K., Hara, A., Deyashiki, Y., Iwasa, H., Kume, T., Ishikura, S., Shiraishi, H., and Katagiri, Y. 1998. Roles of the C-terminal domains of human dihydrodiol dehydrogenase isoforms in the binding of substrates and modulators: Probing with chimaeric enzymes. Biochem. J. 336 (Pt 2) 429–436. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Mesecar, A.D., Stoddard, B.L., and Koshland Jr., D.E. 1997. Orbital steering in the catalytic power of enzymes: Small structural changes with large catalytic consequences. Science 277 202–206. [DOI] [PubMed] [Google Scholar]
  37. Murshudov, G.N., Vagin, A.A., and Dodson, E.J. 1997. Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr. D 53 240–255. [DOI] [PubMed] [Google Scholar]
  38. Nahoum, V., Gangloff, A., Legrand, P., Zhu, D.W., Cantin, L., Zhorov, B.S., Luu-The, V., Labrie, F., Breton, R., and Lin, S.X. 2001. Structure of the human 3α-hydroxysteroid dehydrogenase type 3 in complex with testosterone and NADP at 1.25-A resolution. J. Biol. Chem. 276 42091–42098. [DOI] [PubMed] [Google Scholar]
  39. Nahoum, V., Gangloff, A., Shi, R., and Lin, S.X. 2003. How estrogen-specific proteins discriminate estrogens from androgens: A common steroid binding site architecture. FASEB J. 17 1334–1336. [DOI] [PubMed] [Google Scholar]
  40. Palackal, N.T., Burczynski, M.E., Harvey, R.G., and Penning, T.M. 2001. The ubiquitous aldehyde reductase (AKR1A1) oxidizes proximate carcinogen trans-dihydrodiols to o-quinones: Potential role in polycyclic aromatic hydrocarbon activation. Biochemistry 40 10901–10910. [DOI] [PubMed] [Google Scholar]
  41. Palackal, N.T., Lee, S.H., Harvey, R.G., Blair, I.A., and Penning, T.M. 2002. Activation of polycyclic aromatic hydrocarbon trans-dihydrodiol proximate carcinogens by human aldo-keto reductase (AKR1C) enzymes and their functional overexpression in human lung carcinoma (A549) cells. J. Biol. Chem. 277 24799–24808. [DOI] [PubMed] [Google Scholar]
  42. Penning, T.M., Mukharji, I., Barrows, S., and Talalay, P. 1984. Purification and properties of a 3 α-hydroxysteroid dehydrogenase of rat liver cytosol and its inhibition by anti-inflammatory drugs. Biochem. J. 222 601–611. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Penning, T.M., Burczynski, M.E., Jez, J.M., Hung, C.F., Lin, H.K., Ma, H., Moore, M., Palackal, N., and Ratnam, K. 2000. Human 3α-hydroxysteroid dehydrogenase isoforms (AKR1C1-AKR1C4) of the aldo-keto reductase superfamily: Functional plasticity and tissue distribution reveals roles in the inactivation and formation of male and female sex hormones. Biochem. J. 351 67–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Rees-Milton, K.J., Jia, Z., Green, N.C., Bhatia, M., El-Kabbani, O., and Flynn, T.G. 1998. Aldehyde reductase: The role of C-terminal residues in defining substrate and cofactor specificities. Arch. Biochem. Biophys. 355 137–144. [DOI] [PubMed] [Google Scholar]
  45. Rondeau, J.M., Tete-Favier, F., Podjarny, A., Reymann, J.M., Barth, P., Biellmann, J.F., and Moras, D. 1992. Novel NADPH-binding domain revealed by the crystal structure of aldose reductase. Nature 355 469–472. [DOI] [PubMed] [Google Scholar]
  46. Schlegel, B.P., Jez, J.M., and Penning, T.M. 1998. Mutagenesis of 3 α-hydroxysteroid dehydrogenase reveals a “push-pull” mechanism for proton transfer in aldo-keto reductases. Biochemistry 37 3538–3548. [DOI] [PubMed] [Google Scholar]
  47. Sotriffer, C.A., Kramer, O., and Klebe, G. 2004. Probing flexibility and “induced-fit” phenomena in aldose reductase by comparative crystal structure analysis and molecular dynamics simulations. Proteins 56 52–66. [DOI] [PubMed] [Google Scholar]
  48. Zhang, Y., Dufort, I., Rheault, P., and Luu-The, V. 2000. Characterization of a human 20α-hydroxysteroid dehydrogenase. J. Mol. Endocrinol. 25 221–228. [DOI] [PubMed] [Google Scholar]

Articles from Protein Science : A Publication of the Protein Society are provided here courtesy of The Protein Society

RESOURCES