Satoshi Endo et al.
Archives of Biochemistry and Biophysics 700 (2021) 108755
described. The purified enzymes showed single bands on SDS-PAGE
analysis (Supplementary Fig. S1).
3. Results and discussion
3.1. Coenzyme and substrate specificity
2.3. Assay of enzyme activity
The amino acid sequence deduced from the isolated cDNA for R1C19
(Fig. 1) was identical to that predicted from its gene (akr1c19), except
that the residue at position 265 is Ile instead of Asn in the predicted
sequence. R1C19 showed the highest ASI (92%) to mouse AKR1C19 [14]
among members of the AKR superfamily, whereas 20HSDL shared low
ASI (61–69%) with mouse and other rat AKR1Cs listed in Table 1.
Crystallographic and site-directed mutagenesis studies of AKR1Cs have
shown that the residues at positions 270 and 276 are critical for coen-
zyme specificity: Those are Lys270 and Arg276 for the preference for
NADP(H), and are Gln270 and Glu276 for the NAD(H) specificity [2,20].
Gln270 and Glu276 are conserved not only in NAD+-preferring mouse
AKR1C19, but also in R1C19 and 20HSDL. Since R1C19 and 20HSDL
are, thus, expected to exhibit coenzyme specificity for NAD(H), we pu-
rified the two recombinant 39-kDa monomeric enzymes and tested their
NAD+-linked dehydrogenase activities towards hydroxysteroids and
xenobiotic alcohols that are substrates of other rat AKR1Cs [13,15–20]
and mouse AKR1C19 [14].
Reductase and dehydrogenase activities were assayed by measuring
the rate of change in NAD(P)H absorbance (at 340 nm) and its fluores-
cence emission (at 455 nm with an excitation wavelength of 340 nm),
respectively, in a 1-mL reaction mixture containing 0.1 M potassium
phosphate (pH 7.0), 0.1 mM NADH or 1 mM NAD(P)+, substrate and
enzyme [13]. The substrate concentrations were 0.1–200 μM (for PGs),
10–100 μM (for steroids), 1–25 mM (for monosaccharides) and 0.1–2
mM (for other carbonyls and alcohols). One unit (U) of enzyme activity
was defined as the enzyme amount that catalyzes the oxidation or for-
mation of 1 μ
mol of NAD(P)H per min at 25 ◦C. The Km and Vmax values
were determined over a range of five substrate concentrations by fitting
Michaelis-Menten equation and/or Lineweaver-Burk plot analysis of the
initial velocities using the program Hyper 32 (University of Liverpool,
UK). The enzyme concentrations used in the kinetic studies were 2–10
μ
μ
g/mL (for R1C19, AKR1C16, AKR1C17 and AKR1C19) and 0.8–4
g/mL (for 20HSDL and AKR1C24). kcat value was calculated from the
V
max value based on molecular weights of His-tagged R1C19 (38,959),
R1C19 oxidized three 17β-hydroxysteroids and β-ionol, in addition
to (S)-1-indanol, 3β-hydroxyhexobarbital and cis-benzene dihydrodiol
that are reported as the substrates of mouse AKR1C19 [14] (Table 2).
His-tagged 20HSDL (39,047), and AKR1C19 (37,046). The IC50 values
for inhibitors were determined in the NAD+-linked oxidation of 2
μM
PGD2 (for 20HSDL, 0.8
μ
g/mL) and 25
μM β-ionol (for R1C19, 2
μg/mL).
Among the substrates, 5β-androstane-3α,17β-diol and β-ionol were the
The kinetic constants are expressed as the means ± SE of at least three
most efficient substrates, showing high kcat/Km values. By contrast, most
of the R1C19 substrates were not oxidized by 20HSDL, which instead
exhibited dehydrogenase activities towards trans-benzene dihydrodiol
and (1S,2S)-trans-1,2-cyclohexanediol. Neither R1C19 nor 20HSDL
oxidized other xenobiotic alcohols and hydroxysteroids (listed in
Table 2), most of which are substrates of other rat AKR1Cs [13,15–20].
determinations.
2.4. Product identification
The reaction was conducted in a 2-mL reaction mixture, containing 1
mM NAD+, PG, enzyme (25–50
μ
g) and 0.1 M potassium phosphate, pH
It should be noted that 20HSDL does not accept 20α-hydroxysteroids as
7.4. The substrate concentrations and incubation times were 25
μ
M and
its substrates, contrary to the current protein name encoded in the
15 min, respectively, for the oxidation of PGD2 and 13,14-dihydro-15-
keto-PGF2α, and were 50 M and 30 min, respectively, for that of other
RGD1564865 gene.
μ
Intriguingly, both R1C19 and 20HSDL oxidized seven 9
α
-hydroxy-
PGs. The reaction was stopped by adding 1 M Na2SO4 (0.5 mL) and
diethyl ether/methanol/citric acid (30:4:1, v/v; 2 mL), and the substrate
and product were extracted into the ether. After the ether phase was
dried under vacuum, the residue was dissolved in methanol and
analyzed by thin-layer chromatography using silica-gel plates in a sol-
vent system of ethyl acetate/methylene chloride/acetone/methanol/
acetic acid (90:30:10:8:1, v/v) [26]. The PGs and authentic samples (13,
14-dihydro-15-keto-PGF2α, 13,14-dihydro-15-keto-PGE2, PGD2, PGK2,
PGs, of which 13,14-dihydro-15-keto-PGF2 was the most efficient
α
substrate showing the highest kcat/Km value, followed by 15-keto-PGF2
and PGD2 (Table 2 and Supplementary Fig. S3). Such activity has noαt
been reported for other rodent AKR1Cs as shown in Table 1. R1C19 and
20HSDL showed a similar reactivity towards changes in the two hy-
drocarbon chains of PG: 13,14-Dihydro- and 15-keto-derivatives were
oxidized more efficiently than PGF2α, whereas the hydrogenation of the
C5=C6 double bond on the opposite hydrocarbon chain (13,14-dihydro-
15-keto-PGF1 and PGD1) markedly decreased the catalytic efficiency
compared toαthe corresponding PGs with this double bond (13,14-
PGF2α, PGE2, 9α,11β-PGF2,11β-PGE2) were visualized with iodine vapor.
2.5. Determination of Kd value for coenzyme
dihydro-15-keto-PGF2 and PGD2, respectively). In addition to the
α
structure of the hydrocarbon chains, the substituent on the five-
membered ring may affect the catalytic efficiency of the enzymes,
because PGD2 with 11-keto group was more efficiently oxidized than
The Kd values for NAD(P)+ to R1C19 and 20HSDL were determined
by measuring intrinsic tryptophan fluorescence (at 334 nm with an
excitation wavelength of 290 nm) of 1.0
μ
M enzyme in 0.1 M potassium
PGF2 and 9α,11β-PGF2 with 11α- or 11β-hydroxy group. No activity
α
phosphate, pH 7.4, at 25 ◦C, as described previously [27]. The Kd value
is expressed as the mean ± SE of three determinations.
was observed towards PGs without 9α-hydroxy group (PGF2β, PGE2, 11β-
PGE2, 13,14-dihydro-15-keto-PGE2, PGI2, PGA2, PGA1, PGB2 and PGB1).
The substrate specificity indicates that the two enzymes catalyze the
2.6. RT-PCR
oxidation of the 9α-hydroxy-PGs into their 9-keto metabolites. Indeed,
the oxidized products of 13,14-dihydro-15-keto-PGF2α, PGF2 and
α
Total RNAs of the female Wistar rat tissues and a male rat testis were
prepared, and subjected to RT-PCR as described above. PCR was per-
formed using the Quick Taq HS DyeMix and primers for R1C19, 20HSDL
and β-actin (as an internal control), which are shown in Table S1
(Supplementary data). The PCR products were separated by agarose gel
electrophoresis, and revealed with ethidium bromide.
9α,11β-PGF2 by 20HSDL were identified with 13,14-dihydro-15-keto-
PGE2, PGE2 and 11β-PGE2, respectively, on thin-layer chromatography
analysis (Fig. 2). In the oxidation of PGD2, the substrate amount was
decreased compared to the no-incubation blank, but a very small
amount of the reaction product with a Rf similar to authentic PGK2 was
formed. This might result from the instability of the assumed 9-keto
product, PGK2, because PGK2 could not be detected in the analysis of
the extract of a reaction mixture containing PGK2, instead of PGD2.
For R1C19, the kcat/Km values for 13,14-dihydro-15-keto-PGF2α and
15-keto-PGF2α were higher than those for the xenobiotic and steroidal
substrates (β-ionol and 5β-androstane-3α,17β-diol, respectively),
3
Satoshi Endo
