The effect of disease associated point mutations on 5β-reductase (AKR1D1) enzyme function

Article abstract of DOI:10.1016/j.cbi.2010.12.020

The stereospecific 5β-reduction of Δ⁴-3-ketosterols is very difficult to achieve chemically and introduces a 90° bend between ring A and B of the planar steroid. In mammals, the reaction is catalyzed by steroid 5β-reductase, a member of the aldo-keto reductase (AKR) family. The human enzyme, AKR1D1, plays an essential role in bile-acid biosynthesis since the 5β-configuration is required for the emulsifying properties of bile. Deficient 5β-reductase activity can lead to cholestasis and neo-natal liver failure and is often lethal if it remains untreated. In five patients with 5β-reductase deficiency, sequencing revealed individual, non-synonymous point mutations in the AKR1D1 gene: L106F, P133R, G223E, P198L and R261C. However, mapping these mutations to the AKR1D1 crystal structure failed to reveal any obvious involvement in substrate or cofactor binding or catalytic mechanism, and it remained unclear whether these mutations could be causal for the observed disease. We analyzed the positions of the reported mutations and found that they reside in highly conserved portions of AKR1D1 and hypothesized that they would likely lead to changes in protein folding, and hence enzyme activity. Attempts to purify the mutant enzymes for further characterization by over-expression in Escherichia coli yielded sufficient amounts of only one mutant (P133R). This enzyme exhibited reduced K_m and k_cat values with the bile acid intermediate Δ⁴-cholesten-7α- ol-3-one as substrate reminiscent of uncompetitive inhibition. In addition, P133R displayed no change in cofactor affinity but was more thermolabile as judged by CD-spectroscopy. When all AKR1D1 mutants were expressed in HEK 293 cells, protein expression levels and enzyme activity were dramatically reduced. Furthermore, cycloheximide treatment revealed decreased stability of several of the mutants compared to wild type. Our data show, that all five mutations identified in patients with functional bile acid deficiency strongly affected AKR1D1 enzyme functionality and therefore may be causal for this disease.

Full text of DOI:10.1016/j.cbi.2010.12.020

Chemico-Biological Interactions 191 (2011) 250–254
Contents lists available at ScienceDirect
Chemico-Biological Interactions
journal homepage: www.elsevier.com/locate/chembioint
The effect of disease associated point mutations on 5␤-reductase (AKR1D1)
enzyme function
Rebekka Mindnicha, Jason E. Drurya, Trevor M. Penninga,b,∗
a University of Pennsylvania, School of Medicine, Department of Pharmacology, Philadelphia, PA 19104-6084, USA
b Center of Excellence in Environmental Toxicology, Department of Pharmacology, University of Pennsylvania, Philadelphia, PA 19104-6084, USA
a r t i c l e i n f o
a b s t r a c t
The stereospecific 5␤-reduction of ꢀ4-3-ketosterols is very difficult to achieve chemically and introduces
a 90◦ bend between ring A and B of the planar steroid. In mammals, the reaction is catalyzed by steroid
5␤-reductase, a member of the aldo-keto reductase (AKR) family. The human enzyme, AKR1D1, plays an
essential role in bile-acid biosynthesis since the 5␤-configuration is required for the emulsifying proper-
ties of bile. Deficient 5␤-reductase activity can lead to cholestasis and neo-natal liver failure and is often
lethal if it remains untreated. In five patients with 5␤-reductase deficiency, sequencing revealed indi-
vidual, non-synonymous point mutations in the AKR1D1 gene: L106F, P133R, G223E, P198L and R261C.
However, mapping these mutations to the AKR1D1 crystal structure failed to reveal any obvious involve-
ment in substrate or cofactor binding or catalytic mechanism, and it remained unclear whether these
mutations could be causal for the observed disease. We analyzed the positions of the reported mutations
and found that they reside in highly conserved portions of AKR1D1 and hypothesized that they would
likely lead to changes in protein folding, and hence enzyme activity. Attempts to purify the mutant
enzymes for further characterization by over-expression in Escherichia coli yielded sufficient amounts of
only one mutant (P133R). This enzyme exhibited reduced Km and kcat values with the bile acid intermedi-
ate ꢀ4-cholesten-7␣-ol-3-one as substrate reminiscent of uncompetitive inhibition. In addition, P133R
displayed no change in cofactor affinity but was more thermolabile as judged by CD-spectroscopy. When
all AKR1D1 mutants were expressed in HEK 293 cells, protein expression levels and enzyme activity
were dramatically reduced. Furthermore, cycloheximide treatment revealed decreased stability of sev-
eral of the mutants compared to wild type. Our data show, that all five mutations identified in patients
with functional bile acid deficiency strongly affected AKR1D1 enzyme functionality and therefore may
be causal for this disease.
Article history:
Available online 24 December 2010
Keywords:
5␤-reductase
AKR
Bile acid deficiency
© 2011 Elsevier Ireland Ltd. All rights reserved.
1. AKR1D1: a unique member of the AKR superfamily
Despite the observed low sequence conservation between dis-
tant members due to high evolutionary diversity, all AKR enzymes
share three important characteristics: (1) a conserved protein
fold that results in a characteristic (␣/␤)8-barrel structure; (2)
a conserved mode of cofactor and substrate binding; and (3)
a tetrad of amino acids most important in catalysis: D50, Y55,
K84 and H117 (based on the nomenclature of AKR1C9). In fact,
although substrate specificity varies significantly among AKRs,
including sugar aldehydes, ketosteroids, aflatoxin dialdehydes and
polycyclic aromatic trans-dihydrodiols, most AKR enzymes cat-
alyze the interconversion of aldehydes and ketones with primary
and secondary alcohols. However, AKR1D enzymes are a notable
exception. Instead of reducing an aldehyde or ketone group, they
catalyze the reduction of double bonds in ꢀ4-3-ketosterols in a
stereo-specific manner [3,4], and consequently, these enzymes are
also known as steroid 5␤-reductases. Considering the difference
between functionalities of these two enzymes, aldehyde/ketone
reduction and double bond reduction, it is remarkable that a
Human steroid 5␤-reductase (AKR1D1) is a member of the aldo-
keto reductase (AKR) superfamily, enzymes that exert important
metabolic functions in organisms of all phyla [1]. The superfam-
ily comprises over 150 annotated members that share 20–99%
sequence identity. The AKR nomenclature [2] distinguishes 16 fam-
ilies (AKR1-16) based on a minimum of 40% protein sequence
similarity. Family members with 60% or more sequence identity are
further grouped into subfamilies (e.g. AKR1A-E). Hence, AKR1D1
belongs to the AKR1 family and the AKR1D subfamily.
∗
Corresponding author at: Center of Excellence in Environmental Toxicology,
Department of Pharmacology, University of Pennsylvania, Philadelphia, PA 19104-
6084, USA.
E-mail addresses: penning@mail.med.upenn.edu, penning@upenn.edu
(T.M. Penning).
0009-2797/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.cbi.2010.12.020

R. Mindnich et al. / Chemico-Biological Interactions 191 (2011) 250–254
251
Fig. 1. Central role of 5␤-reduction in bile acid biosynthesis. AKR1D1 is a key enzyme in bile acid biosynthesis that catalyzes the 5␤-reduction of ꢀ4-cholesten-7␣-ol-3-one
and ꢀ4-cholestene-7␣,12␣-diol-3-one substrates that will ultimately result in the production of mature bile acids such as chenodeoxycholic and cholic acid. If 5␤-reduction
is inhibited through a functional defect in AKR1D1 or other factors, 5␤-reductase deficiency syndrome is observed. Physiological consequences include cholestasis and liver
damage and often result in death of the patient if untreated.
single amino acid substitution in the catalytic tetrad, from histi-
dine to glutamate, is sufficient to achieve this switch from one
function to another [5]. The recently determined crystal struc-
ture of AKR1D1 further elaborates the role of this residue where
it is proposed to act as a superacid and in concert with the
catalytic tyrosine facilitates hydride transfer to the ꢀ4-ene [6].
Furthermore, strict conservation of this glutamate in all AKR1D
proteins and almost complete absence from other AKR subfami-
lies highlights the importance of this substitution for 5␤-reductase
function.
reduced structure occur in the bile acids of almost all mammals
and this configuration is essential for the proper emulsification of
dietary cholesterol, lipids, and essential nutrients and promotion
of their absorption and digestion [11,12].
Evidence that AKR1D1 is the only human 5␤-reductase and that
its most important physiological role is in bile acid biosynthesis
comes from patients with 5␤-reductase deficiency, a metabolic
syndrome first described by Setchell in 1988 [13]. The deficiency
syndrome develops in infants within the first few weeks after
birth and is characterized by reduced primary bile acid synthesis
and accumulation of ꢀ4-3-keto- and 5␣-reduced (allo-) bile acids
(Fig. 1). Furthermore, 5␤-reduced steroids are absent from serum
and urine. The effects of this rare metabolic disorder are severe
and manifest in cholestasis and neo-natal liver damage probably
caused by the accumulation of potentially cholestatic and hepa-
totoxic atypical bile acids. If untreated, 5␤-reductase deficiency
is often lethal. Loss of 5␤-reductase activity could either result
from a genetically determined enzyme defect or from secondary
effects such as changes in protein/mRNA stability due to liver dis-
ease [14] or enzyme modification through chemical modification
[15,16].
2. The physiological role of AKR1D1 in bile acid
biosynthesis
Steroid 5␤-reductase activity is most abundant in the liver and
led to the first isolation of the protein from rat liver [7]. More than
25 years later, the human gene, AKR1D1, was cloned based on a
previously released rat 5␤-reductase cDNA annotation [3,8]. Like its
rat homolog, AKR1D1 is most abundant in liver and recombinant
expression of the human protein in mammalian cells confirmed
5␤-reductase activity towards steroids and bile acid intermediates
[3,9,10].
In congruence with the enzyme’s expression pattern and
catalytic activity, AKR1D1 has been proposed to play a phys-
iological role in steroid hormone metabolism as well as bile
acid biosynthesis. In the latter, AKR1D1 catalyzes the central
5␤-reduction step and reduces bile acid precursors such as ꢀ4-
cholesten-7␣-ol-3-one and ꢀ4-cholestene-7␣,12␣-diol-3-one to
5␤-cholestan-7␣-ol-3-one and 5␤-cholestane-7␣,12␣-diol-3-one,
respectively. The stereo-specific reduction of the sterol precursors
is of physiological importance. While 5␣-reduction leaves the over-
all structure in its planar shape, 5␤-reduction introduces a 90◦
angle between the A and B ring. Sterols with this characteristic 5␤-
3. Disease associated AKR1D1 mutations affect conserved
protein sites
In an effort to link putative genetic defects in AKR1D1 to the
5␤-reductase deficiency syndrome, the genotype of AKR1D1 had
been determined in several patients. Sequencing identified mis-
sense mutations leading to truncated, non-functional proteins in
two patients [17,18], and five individual point mutations lead-
ing to non-synonymous amino acid substitutions in the AKR1D1
protein (L106F, P133R, P198L, G223E, and R261C) in three other
patients [17–19]. Analysis of the mutations in the AKR1D1 crys-

252
R. Mindnich et al. / Chemico-Biological Interactions 191 (2011) 250–254
Fig. 2. Disease-associated mutations affect regions highly conserved in the AKR1 family. Multiple sequence alignment and calculation of conservation per site of AKR1
sequences annotated at the AKR website was performed with ClustalW and relative conservation visualized with JalView. Amino acid position in AKR1D1 is indicated on the
x-axis; degree of conservation on the y-axis. Residues mutated in patients with 5␤-reductase deficiency syndrome are indicated in black; the residues forming the catalytic
tetrad are marked in light grey. Secondary structure elements of AKR1D1 are indicated above the plot. Notably, residues changed in the disease are highly conserved or are
located in a highly conserved region of the protein.
tal structure however, did not reveal involvement of the affected
residues in catalysis, or cofactor or substrate binding [6]. There-
fore, it remained unclear whether the detected mutations could
indeed cause the 5␤-reductase deficiency phenotype observed in
these patients.
4. Disease associated AKR1D1 mutations affect protein
integrity and enzyme activity
Further evidence of diminished structural integrity comes from
our efforts to purify and enzymatically characterize the mutant
AKR1D1 isoforms, and from monitoring expression levels and
stability of these mutants when expressed in human cells. Recombi-
nant protein expression in Escherichia coli yielded abundant protein
in lysed cells [24]. However, subsequent efforts to isolate the
enzymes either failed or yielded extremely low amounts of purified
protein in comparison to the wild type enzyme (Table 1). Only the
P133R mutant was obtained in sufficient purity and amounts that
allowed further characterization. Determination of melt curves by
CD spectroscopy revealed decreased thermal stability of the P133R
mutant in the absence of cofactor. Melt curves were recorded fol-
lowing the absorbance at 222 nm while raising the temperature
from 14 to 94 ◦C in increments of 2 ◦C. The Tm, the temperature
where half of the protein was unfolded according to measurement
of molar elipticity, was 41 ◦C for the mutant compared to 48 ◦C for
the wild type protein. However, the presence of cofactor eliminated
the observed difference [24]. Following expression of recombinant
AKR1D1 in HEK 293 cells, levels of mutant proteins were signif-
icantly decreased compared to wild type (Fig. 3). For two of the
mutations, L106F and R261C, protein expression levels were less
than 1% of that of the wild type enzyme. In addition, cycloheximide
treatment revealed decreased stability of mutant protein as expres-
sion levels declined compared to wild type AKR1D1 over a 24 h time
course.
Non-synonymous mutations not involved in enzyme function
may still affect protein folding and stability if they are found within
highly conserved residues or domains of a protein family. In this
case, conservation reflects less the enzyme properties associated
with catalysis but rather establishment of the common and charac-
teristic protein fold of a family. These cues are the basis for instance
in sequence based algorithms that identify conserved domains (e.g.
CD search at NCBI) [20,21] or in homology modeling (examples
for AKR1D1 include [22,23]). To test this hypothesis we examined
sequence conservation within the AKR1 family (Fig. 2). We find that
allthe AKR1D1mutations occureitherin conserved residuesorcon-
served domains. Of these, two mutations, L106F and R261C, reside
in ␣-helices 3 and 8, respectively, and could affect protein stabil-
ity. The remaining mutations are located in loop or turn regions of
the protein. These mutations severely alter the physico-chemical
properties of the residue at the affected site. In two cases, the small
but rigid proline is replaced by either a large positively charged
amino acid (P133R) or a similar small hydrophobic, but flexible
side chain (P198L). In G223E, the ␣-hydrogen is replaced by a car-
boxylic acid side chain. It is notable, that two mutations, P133R
and G223E, reside in loops A and B, respectively, which convey
substrate recognition and binding properties. These findings sup-
port the notion that the observed mutations could indeed result
in changes in protein structure and stability and ultimately lead to
deficient 5␤-reductase activity.
However, the mutations observed in patients with 5␤-reductase
deficiency not only affected expression levels and stability but also
Table 1
Purification of recombinant AKR1D1 mutant protein.
Ni-sepharose fraction of
Volume (mL)
Total protein (mg)
Total activity
Specific activity
Purification
factor
Yield (%)
(nmol min−1
)
(nmol min−1 mg−1
)
WT
4.0
14.5
3.9
4.1
5.0
5.7
3.3
0.4
0.5
1.5
462
0.2
ND
ND
36
81
0.05
ND
ND
24.4
9.9
1.25
–
–
9.47
56
7
0
0
16
L106Fa
P198L
R261C
P133R
a
By SDS-PAGE analysis, sample is only 10% pure.

R. Mindnich et al. / Chemico-Biological Interactions 191 (2011) 250–254
253
Table 2
Effect of the P133R mutation on Km and kcat of 5␤-reduction.
Wild type
P133R mutant
kcat (min−1
)
Km (␮M)
kcat (min−1
)
Km (␮M)
Testosterone
Cortisone
4-Cholesten-7␣-ol-3-one
7.1 1.7
9.9 0.1
1.7 0.1
2.7 1.2
15.1 0.3
0.8 0.2
2.7 0.3
0.6 0.02
0.2a
12.7 2.2
1.3 0.1
0.08a
Activity measurements were determined fluorimetrically.
a
Estimated.
enzymatic activity. A study of the purified P133R mutant revealed
drastic changes in kinetic constants with different steroid and
bile acid intermediate substrates (Table 2). With ꢀ4-cholesten-
7␣-ol-3-one as substrate the reaction proceeded with a highly
reduced Km and kcat and kinetics were reminiscent of uncom-
petitive inhibition. Hence, the P133R mutation converts AKR1D1
from a low affinity, high capacity enzyme, to one with high affin-
ity and low capacity. The mutant enzymes also showed severely
decreased enzymatic activity when compared to wild type follow-
ing expression in human HEK293 cells. The AKR1D1 mutants P198L
and G223E did not show measurable 5␤-reductase activity after
60 h following addition of substrate to the culture medium. The
mutants L106F, P133R and R261C exhibited 5␤-reductase activ-
ity that was only a fraction of that of wild type AKR1D1 (Fig. 4).
Quite notably, L106F and R261C showed detectable enzyme activ-
ity despite extremely low expression levels indicating that these
mutations did not necessarily affect enzyme activity and that low
protein expression accounted for their reduced activity in the HEK-
293 cells.
Fig. 4. AKR1D1 disease mutants had reduced or no enzyme activity when expressed
in HEK-293 cells. Wild type and mutant AKR1D1 were expressed in HEK 293 cells
and 5␤-reductase activity was measured radiometrically by following the conver-
sion of tritiated testosterone to 5␤-DHT. Product verification and quantification was
carried out by thin layer chromatography. Initial velocities were estimated from
linear phase of product formation. In case of the mutants P198L and G223E no sub-
strate conversion was detected within 60 h following addition of testosterone to the
medium.
5. Conclusions
In the AKR family, AKR1D1 is the only enzyme that does not
catalyze the interconversion of aldehyde and ketone groups with
primary and secondary alcohols but instead catalyzes the reduction
of double bonds in ꢀ4-3-ketosteroids and sterols.
In contrast to the closely related AKR1C subfamily, only one
functional AKR1D1 gene is present in all vertebrates except fish.
Furthermore, compared to other AKR1 family members the num-
ber of SNPs and other genomic variations is low in AKR1D1 [25]
and sequence conservation between distantly related 5␤-reductase
members is significantly higher than between 5␣-reductase
homologs [26]. This agrees well with the central physiological func-
tion of AKR1D1 in bile acid biosynthesis, an important metabolic
pathway in all vertebrates. Defects in bile acid metabolism are
rare (<0.1% of population) but can have severe consequences, as
seen in patients with 5␤-reductase deficiency. We showed that five
mutations detected in these patients had either reduced protein
expression and stability or enzyme activity, or both. As a result,
net levels of physiologically active 5␤-reductase would be drasti-
cally reduced. Although the disease-associated mutations did not
seem to directly affect substrate binding or catalysis, they signifi-
cantlyalteredphysico-chemicalproperties ofconserved residuesor
Fig. 3. AKR1D1 disease mutants showed reduced protein expression and stabil-
ity versus wild type enzyme. Expression of wild type and mutant AKR1D1 in HEK
293 cells was observed by Western Blotting with a specific anti-AKR1D1 antibody.
Expression levels were determined with densitometry and analysis with the Uniscan
and imageJ software packages. Expression levels of the L106F and R261C mutants
were so low that they could only be detected by a highly sensitive detection method
(SuperSignal West Femto Kit (Thermo Scientific)). Time indicates time points after
cycloheximide treatment.

254
R. Mindnich et al. / Chemico-Biological Interactions 191 (2011) 250–254
areas within the protein thereby disrupting stability and enzymatic
activity of AKR1D1.
[10] A. Charbonneau, V.L. The, Genomic organization of a human 5beta-reductase
and its pseudogene and substrate selectivity of the expressed enzyme, Biochim.
Biophys. Acta 1517 (2) (2001) 228–235.
[11] D.W. Russell, K.D. Setchell, Bile acid biosynthesis, Biochemistry 31 (20) (1992)
4737–4749.
[12] D.W. Russell, The enzymes, regulation, and genetics of bile acid synthesis, Annu.
Rev. Biochem. 72 (2003) 137–174.
[13] K.D. Setchell, F.J. Suchy, M.B. Welsh, L. Zimmer-Nechemias, J. Heubi, W.F. Bal-
istreri, Delta 4-3-oxosteroid 5 beta-reductase deficiency described in identical
twins with neonatal hepatitis. A new inborn error in bile acid synthesis, J. Clin.
Invest. 82 (6) (1988) 2148–2157.
Conflict of interest statement
The authors declare that there are no conflicts of interest.
Acknowledgements
[14] P.T. Clayton, Delta 4-3-oxosteroid 5 beta-reductase deficiency and neonatal
hemochromatosis, J. Pediatr. 125 (5 Pt. 1) (1994) 845–846.
We thank Dr. Robert Brown for his advice on using cyclo-
heximide in measuring protein stability in cell culture. R.M. was
supported by USPHS NIH Grant R13-AA019612 to present this
work at the 15th International Meeting on Enzymology and Molec-
ular Biology of Carbonyl Metabolism in Lexington, KY USA. This
work was supported by grants NIH 5-R01-DK047015-13 and P30-
ES013508 awarded to T.M.P.
[15] Y. Zhu, M.J. Fillenwarth, D. Crabb, L. Lumeng, R.C. Lin, Identification of the 37-
kD rat liver protein that forms an acetaldehyde adduct in vivo as delta 4-3-
ketosteroid 5 beta-reductase, Hepatology 23 (1) (1996) 115–122.
[16] R.C. Lin, M.J. Fillenwarth, X. Du, Cytotoxic effect of 7alpha-hydroxy-4-
cholesten-3-one on HepG2 cells: hypothetical role of acetaldehyde-modified
delta4-3-ketosteroid-5beta-reductase (the 37-kD-liver protein) in the patho-
genesis of alcoholic liver injury in the rat, Hepatology 27 (1) (1998) 100–107.
[17] I. Ueki, A. Kimura, H.L. Chen, T. Yorifuji, J. Mori, S. Itoh, K. Maruyama, T.
Ishige, H. Takei, H. Nittono, T. Kurosawa, M. Kage, T. Matsuishi, SRD5B1
gene analysis needed for the accurate diagnosis of primary 3-oxo-Delta4-
steroid 5beta-reductase deficiency, J. Gastroenterol. Hepatol. 24 (5) (2009)
776–785.
References
[1] J.E. Drury, D. Hyndman, J. Yin, T.M. Penning, The aldo-keto reductase superfam-
ily homepage: 2006 update, in: H. Weiner, E. Maser, R. Lindahl, B. Plapp (Eds.),
Enzymology and Molecular Biology of Carbonyl Metabolism, Purdue University
Press, West Lafayette, Indiana, 2007.
[2] J.M. Jez, T.G. Flynn, T.M. Penning, A new nomenclature for the aldo-keto reduc-
tase superfamily, Biochem. Pharmacol. 54 (6) (1997) 639–647.
[3] K.H. Kondo, M.H. Kai, Y. Setoguchi, G. Eggertsen, P. Sjoblom, T. Setoguchi, K.I.
Okuda, I. Bjorkhem, Cloning and expression of cDNA of human delta 4-3-
oxosteroid 5 beta-reductase and substrate specificity of the expressed enzyme,
Eur. J. Biochem. 219 (1–2) (1994) 357–363.
[4] A. Okuda, K. Okuda, Purification and characterization of delta 4-3-ketosteroid
5 beta-reductase, J. Biol. Chem. 259 (12) (1984) 7519–7524.
[18] H.A. Lemonde, E.J. Custard, J. Bouquet, M. Duran, H. Overmars, P.J. Scambler,
P.T. Clayton, Mutations in SRD5B1 (AKR1D1), the gene encoding delta(4)-3-
oxosteroid 5beta-reductase, in hepatitis and liver failure in infancy, Gut 52
(10) (2003) 1494–1499.
[19] E. Gonzales, D. Cresteil, C. Baussan, A. Dabadie, M.F. Gerhardt, E. Jacquemin,
SRD5B1 (AKR1D1) gene analysis in delta(4)-3-oxosteroid 5beta-reductase defi-
ciency: evidence for primary genetic defect, J. Hepatol. 40 (4) (2004) 716–718.
[20] A. Marchler-Bauer, A.R. Panchenko, N. Ariel, S.H. Bryant, Comparison of
sequence and structure alignments for protein domains, Proteins 48 (3) (2002)
439–446.
[21] A. Marchler-Bauer, S.H. Bryant, CD-search: protein domain annotations on the
fly, Nucleic Acids Res. 32 (Web Server issue) (2004) W327–W331.
[22] J.M. Jez, M.J. Bennett, B.P. Schlegel, M. Lewis, T.M. Penning, Comparative
anatomy of the aldo-keto reductase superfamily, Biochem. J. 326 (Pt. 3) (1997)
625–636.
[5] J.M. Jez, T.M. Penning, Engineering steroid 5 beta-reductase activity into rat
liver 3 alpha-hydroxysteroid dehydrogenase, Biochemistry 37 (27) (1998)
9695–9703.
[6] L. Di Costanzo, J.E. Drury, T.M. Penning, D.W. Christianson, Crystal structure of
human liver Delta4-3-ketosteroid 5beta-reductase (AKR1D1) and implications
for substrate binding and catalysis, J. Biol. Chem. 283 (24) (2008) 16830–16839.
[7] O. Berseus, Conversion of cholesterol to bile acids in rat: purification
and properties of a delta-4-3-ketosteroid 5-beta-reductase and a 3-alpha-
hydroxysteroid dehydrogenase, Eur. J. Biochem. 2 (4) (1967) 493–502.
[8] Y. Onishi, M. Noshiro, T. Shimosato, K. Okuda, Molecular cloning and sequence
analysis of cDNA encoding delta 4-3-ketosteroid 5 beta-reductase of rat liver,
FEBS Lett. 283 (2) (1991) 215–218.
[23] W.H. Lee, P. Lukacik, K. Guo, E. Ugochukwu, K.L. Kavanagh, B. Marsden, U.
Oppermann, Structure–activity relationships of human AKR-type oxidore-
ductases involved in bile acid synthesis: AKR1D1 and AKR1C4, Mol. Cell.
Endocrinol. 301 (1–2) (2009) 199–204.
[24] J.E. Drury, R. Mindnich, T.M. Penning, Characterization of disease-related
5beta-reductase (AKR1D1) mutations reveals their potential to cause bile acid
deficiency, J. Biol. Chem. 285 (32) (2010) 24529–24537.
[25] R.D. Mindnich, T.M. Penning, Aldo-keto reductase (AKR) superfamily: genomics
and annotation, Hum. Genomics 3 (4) (2009) 362–370.
[9] R.B. Iyer, J.M. Binstock, I.S. Schwartz, G.G. Gordon, B.I. Weinstein, A.L.
Southren, Human hepatic cortisol reductase activities: enzymatic properties
and substrate specificities of cytosolic cortisol delta 4-5 beta-reductase and
dihydrocortisol-3 alpha-oxidoreductase(s), Steroids 55 (11) (1990) 495–500.
[26] V.S. Langlois, D. Zhang, G.M. Cooke, V.L. Trudeau, Evolution of steroid-5alpha-
reductases and comparison of their function with 5beta-reductase, Gen. Comp.
Endocrinol. 166 (3) (2010) 489–497.