Effect of coenzymes and thyroid hormones on the dual activities
of
Xenopus
cytosolic thyroid-hormone-binding protein (xCTBP)
with aldehyde dehydrogenase activity
Kiyoshi Yamauchi and Jun–ichiro Nakajima
Department of Biology and Geoscience, Faculty of Science, Shizuoka University, Shizuoka, Japan
A cytosolic thyroid-hormone-binding protein (xCTBP),
predominantly responsible for the major binding activity of
T
3
in the cytosol of Xenopus liver, has been shown to be
identical to aldehyde dehydrogenase class 1 (ALDH1)
[Yamauchi, K., Nakajima, J., Hayashi, H., Horiuchi, R. &
Tata, J.R. (1999) J. Biol. Chem. 274, 8460–8469]. Within this
paper we surveyed which signaling, and other, compounds
affect the thyroid hormone binding activity and aldehyde
dehydrogenase activity of recombinant Xenopus ALDH1
(xCTBP/xALDH1) while examining the relationship
between these two activities. NAD
+
and NADH (each
200 l
M
),andtwosteroids(20l
M
), inhibit significantly the
T
3
-binding activity, while NADH and NADPH (each
200 l

using cultured Xenopus cells demonstrates
xCTBP/xALDH1 bound T
3
within living cells. These results
raise the possibility that an unknown factor(s) besides
NAD
+
and NADH may modulate the thyroid-hormone-
binding activity of xCTBP/xALDH1. In comparison, thy-
roid hormone, at its physiological concentration, would
poorly modulate the enzyme activity of xCTBP/xALDH1.
Keywords: cytosolic thyroid-hormone-binding protein;
aldehyde dehydrogenase; retinoic acid synthesis; Xenopus
laevis.
Hydrophobic molecules that signal via nuclear receptors,
such as thyroid and steroid hormones, retinoic acid and
vitamin D
3
, predominantly exist within plasma and within
intracellular compartments bound to specific proteins. The
kinetics and the nature of the cellular responses to these
signaling molecules are determined by these specific binding
proteins. This has been well documented for cytosolic
retinoic acid and retinol binding proteins where it has been
suggested that these binding proteins may act, not only as
buffers or reservoirs of intracellular retinoids to maintain
significant levels of free retinoids, but also as modulators
transporting retinoids to their target sites, the retinoid
responsive genes within the nucleus and the metabolic
enzymes within the cytoplasm [1–3]. Although similar

kidney [11], although xCTBP/xALDH1 mRNA was found
predominantly in the kidney and the intestine rather than in
the liver [10]. The restricted tissue-distribution of xCTBP/
xALDH1, particularly at the metamorphosing stages, raises
the possibility that xCTBP/xALDH1 could modulate the
actions of T
3
in a tissue-dependent manner. By controlling
the intracellular concentrations of free T
3
, xCTBP/
xALDH1 might play a critical role in regulating T
3
access
to its target sites within the nucleus and the cytoplasm [12].
There have been several reports demonstrating interac-
tions between mammalian ALDH1 and bioactive
Correspondence to K. Yamauchi, Department of Biology and
Geoscience, Faculty of Science, Shizuoka University, 836 Oya,
Shizuoka 422-8529, Japan.
Fax: + 81 54 2380986, Tel.: + 81 54 2384777,
E-mail:
Abbreviations: CTBP, cytosolic thyroid-hormone-binding protein;
xCTBP, Xenopus CTBP; ALDH1, aldehyde dehydrogenase class 1;
xALDH1, Xenopus ALDH1; T
3
,3,3¢,5-triiodo-
L
-thyronine; T
4

T
3
-binding activity of this protein, whereas NADH,
NADPH and iodothyronines inhibit the ALDH activity.
Detailed studies revealed that NAD
+
and T
3
each act as a
noncompetitive inhibitor on the T
3
-binding and enzyme
activities of the protein, respectively.
MATERIALS AND METHODS
Materials
T
3
,D-T
3
,
L
-thyroxine (T
4
), 3,3¢,5-triiodo-
L
-thyroacetic acid
(Triac), all-trans-retinal, all-trans-retinoic acid, androster-
one, cortisone, 11-deoxycorticosterone, dehydroisoandros-
terone, 17-b estradiol, progesterone and testosterone were
purchased from Sigma. NADP

Escherichia coli
E. coli BL21 bearing an expression vector containing
xALDH1-I (pET15b/xALDH1-I) cDNA [10] was grown
and expression of the recombinant proteins was induced by
0.2 m
M
isopropyl thio-b-
D
-galactoside. Purification of the
recombinant proteins was performed as described previously
[10]. In brief, bacteria were collected by centrifugation at
1200 g for 30 min at4 °C. After resuspending in 0.3
M
NaCl,
50 m
M
Tris/HCl, pH 8.0, 10 m
M
imidazole, 1 mgÆmL
)1
lysozyme, 1 m
M
benzamidine hydrochloride, 1 m
M
phenyl-
methanesulfonyl fluoride and 50 m
M
2-mercaptoethanol,
the cells were disrupted by sonication (UR200P type, Tomy,
Japan) for 10 s repeated three times. The extract was

I]T
3
bound to proteins was separated
from free [
125
I]T
3
by the Dowex method [9] and radioac-
tivity levels were measured in a c-counter (Auto Well
Gamma System ARC-2000, Aloka, Japan). The amount of
[
125
I]T
3
bound nonspecifically was obtained by measuring
the radioactivity level within the samples incubated with
5 l
M
unlabeled T
3
. The nonspecific binding value was
subtracted from the amount of total bound [
125
I]T
3
to give
the values of specifically bound [
125
I]T
3

absence of fetal bovine serum for 0.5–1.0 h at 24 °C. The
cytosol, contained within a 0.5-mL Eppendorf tube, and
the Xenopus cells, spread on a 35-mm plastic Petri dish,
were placed on a UV crosslinker (CL-1000, Funakoshi
Co., Japan), and exposed to UV light (254 nm, 40 W) for
3min at 0°C. The resultant cytosolic proteins, and
Xenopus cells, detached from the Petri dish with 0.05%
trypsin, were mixed separately with an equal volume of
2 · SDS-sample buffer, followed by boiling for 5 min. The
proteins were resolved by SDS/PAGE. The affinity-labeled
proteins were detected by autoradiography, exposed to
X-ray XAR5 film (Kodak) on an intensifying screen at
)85 °C for 1–3 weeks.
Aldehyde dehydrogenase activity
Photometric assays were performed in triplicate in 400 lL
of 50 m
M
Tris/HCl, pH 8.0, 3.3 m
M
pyrazole, 100 m
M
KCl, 1 m
M
dithiothreitol, 0.33 m
M
NAD
+
and 30 l
M
retinal, unless otherwise stated [10]. The amount of retinoic

3
binding
to recombinant xCTBP/xALDH1 was examined in the
presence of each compound listed in Table 1. Of three
iodothyronines and Triac, T
3
was the most potent compet-
itor of [
125
I]T
3
binding. The resulting affinity order of
T
3
‡ D-T
3
>T
4
> Triac, agreed with the order of their
relative binding affinity to xCTBP in the Xenopus cytosol
from adult and metamorphosing tadpole liver [9,11]. At
pH 7.5, 50% inhibition of [
125
I]T
3
binding to xCTBP/
xALDH1 was achieved with T
3
and D-T
3

M
, inhibited [
125
I]T
3
binding by more than 50%
while retinal, at a concentration of 12 l
M
, activated [
125
I]T
3
binding by 36%, although no significant difference was
obtained. The other compounds exhibited little effect on T
3
binding (Table 1). The effect of NAD
+
is shown to be dose-
dependent (Fig. 1B). The concentration of NAD
+
neces-
sary to inhibit 50% of [
125
I]T
3
binding to xCTBP/xALDH1
(IC
50
) was 40 l
M

-binding activity was examined by incubating the purified xALDH1 with 0.1 n
M
[
125
I]T
3
for 30 min at 0 °C, as described in Materials and methods. Nonspecific binding was determined from the samples incubated in the presence
of 5 l
M
unlabeled T
3
and subtracted from the total binding. The activity of the retinoic acid formation was examined by incubating the purified
xALDH1 with 0.33 m
M
NAD
+
and 30 l
M
retinal for 1–2 min at 24 °C [10]. Data are mean ± SEM from at least triplicate determina-
tions.*P < 0.05; **P < 0.01; ***P < 0.001.
Effector Concentration T
3
-binding activity ALDH activity
Control 100 ± 6 100 ± 2
Retinoic acid 12 l
M
99.3 ± 5.4 133 ± 4**
NAD
+
200 l

18.6 ± 2.2***
1 l
M
35.7 ± 0.6***
L
-Thyroxine 0.32 l
M
60.4 ± 4.4**
1 l
M
36.3 ± 0.4***
L
-3,3¢,5-Triiodothyroacetic acid 0.32 l
M
95.5 ± 4.1
1 l
M
39.5 ± 1.3***
Control 100 ± 2 100 ± 5
Testosterone 20 l
M
83.7 ± 5.8 104 ± 3
Androsterone 20 l
M
92.3 ± 3.0 96.4 ± 3.1
Dehydroisoandrosterone 20 l
M
87.2 ± 1.9** 101 ± 1
Progesterone 20 l
M

d
values between the
NAD
+
-treated and untreated samples, 66 ± 11 n
M
(n ¼ 3) vs. 53 ± 5 n
M
(n ¼ 5), respectively, as shown in
Fig. 2. This result indicated that the inhibitory mode of
NAD
+
was noncompetitive. Progesterone, at 2 l
M
,
appeared to affect both the K
d
(75 ± 2 n
M
, n ¼ 3)
and MBC (310 ± 28 pmolÆmg
)1
protein, n ¼ 3) values,
although no significant differences were obtained for these
values when compared with the K
d
and MBC values for the
untreated samples.
Characterization of ALDH activity of recombinant
xCTBP/xALDH1

3
in the presence of various concentrations of
unlabeled T
3
with (open symbols) or without (d) the effector: 200 l
M
NAD
+
(s), 2 l
M
progesterone (h), for 30 min at 0 °C. Nonspecific
binding was subtracted from total binding. Each value is the mean of
triplicate determinations. This experiment was repeated at least three
times.
Fig.3. EffectofT
3
on retinoic acid synthesis from retinal, catalyzed by
xCTBP/xALDH1. ALDH activity was measured as the rate of retinoic
acid synthesis. The reaction was performed at 24 °Cwith5lgof
xCTBP/xALDH1 in the presence of various concentrations of T
3
.The
inset illustrates the Hill plot, log[v
c
/v
i
)1] vs. the logarithm of T
3
molar
concentration, the slope of which yields the Hill coefficient. v

Triac (n) was added, whereas, in (B), progesterone (s)orNAD
+
(d)
was added. Nonspecific binding was subtracted from total binding to
give values for specific binding. Each value is the mean ± SEM of
triplicate determinations.
2260 K. Yamauchi and J. Nakajima (Eur. J. Biochem. 269) Ó FEBS 2002
To determine how thyroid hormones interact with
xCTBP/xALDH1, resulting in the decrease in the formation
of retinoic acid from retinal, kinetics of the inhibition of
xCTBP/xALDH1 by T
3
was examined by variation of
NAD
+
concentration within the reaction mixture. The K
m
value, 9 l
M
, was independent of the concentration of T
3
,
but the V
max
value decreased from 0.18 to 0.08 lmolÆ
min
)1
Æmg
)1
with increasing concentrations of T

was added to the reaction mixture. The
Hill coefficient did not change significantly in incubations
with and without 5 l
M
T
3
, 2.3 ± 0.1 vs. 2.2 ± 0.1 (Fig. 5,
inset). These results indicated that T
3
acts as a noncompet-
itive inhibitor against both NAD
+
and retinal upon the
enzyme activity of xCTBP/xALDH1.
T
3
binding to xCTBP/xALDH1 in intact
Xenopus
cells
The present studies on the dual activities of xCTBP/
xALDH1 have indicated that NAD
+
is required at
concentrations of 10
-5
)10
-4
M
for expression of ALDH
activity, whereas  10

with values reported previously [27,28]. On the other hand,
Xenopus liver had a low NAD content, 201 ± 23 lmolÆ(kg
fresh weight)
)1
(n ¼ 6), less than one third of that in rat liver
(Table 2). There were no significant differences in NAD
contents among various Xenopus tissues. Next, T
3
-binding
activity of xCTBP/xALDH1 was directly examined by
photoaffinity-labeling using intact Xenopus cells. Analyses
of the cytosol obtained from the cell lines (KR and XL58)
and the adult liver revealed the presence of single labeled
59-kDa xCTBP (lanes 1–3 in Fig. 6). Photoaffinity-labeling
of [
125
I]T
3
using intact KR and XL58 cells revealed, via
autoradiography, a labeled protein band of the same size
(lanes 4 and 5 in Fig. 6), demonstrating that xCTBP/
xALDH1 is capable of binding T
3
within the Xenopus cells.
Fig. 4. Kinetics of the inhibition of xCTBP/xALDH1 by T
3
when
NAD
+
concentration was varied within the reaction mixture. The

minations. SEMs, which were less than the size of symbols, are not
shown. This experiment was repeated six times, each with similar
results.
Table 2. Contents of NAD in rat liver and various Xenopus tissues. Data
are expressed as the mean ± SEM (number of samples). NAD content
is the sum of the oxidizaed and reduced forms.
Species/tissue NAD (lmolÆ kg wet weight
)1
)
Rat
Liver 756 ± 49 (3)
Xenopus
Liver 201 ± 23 (6)
Kidney 234 ± 83 (5)
Stomach 232 ± 7 (3)
Intestine 291 ± 69 (3)
Ovary 294 ± 94 (4)
Heart 177 ± 29 (3)
Skeletal muscle 199 ± 37 (3)
Ó FEBS 2002 Dual activities of xCTBP/xALDH (Eur. J. Biochem. 269) 2261
DISCUSSION
The present work was undertaken with the aim of deter-
mining which signaling molecules, and other molecules,
affected the T
3
-binding and ALDH activities of xCTBP/
xALDH1. We have obtained evidence that the
[
125
I]T

hydrocortisone [15–17] and benzo[a]pyrene [18,19]. As the
binding of the first three compounds to xALDH1 was also
witnessed in the present study (Table 1), the ability of
ALDH1 to bind the compounds appears to have occurred
at an early step during vertebrate evolution.
Detailed studies revealed that NAD
+
noncompetitively
inhibited the T
3
-binding activity of xCTBP/ALDH1
whereas T
3
inhibited the ALDH activity in a noncompet-
itive fashion against both NAD
+
and retinal. These results
suggested the formation of a ternary complex consisting of
xCTBP/xALDH1, NAD
+
and T
3
. For human mitochon-
drial and cytoplasmic ALDHs, T
3
and Triac were compet-
itive inhibitors against NAD
+
and uncompetitive inhibitors
against propionaldehyde [31]. These distinct inhibitory

and K
i
values. It may be possible
that xCTBP/ALDH1 forms different conformations when
bound to NAD
+
and/or T
3
, This possibility is considered
due to the presence of positive cooperativity upon ALDH
activity (the Hill coefficient, h ¼ 2.2) when the concentra-
tion of retinal was varied (Fig. 5) and the presence of
positive cooperativity upon the inhibition of ALDH activity
(h ¼ 2.4) when the concentration of T
3
was varied (Fig. 3).
T
3
may be a selective, allosteric inhibitor of the xALDH1
enzyme. Such an allosteric conformational change was
proposed for human alcohol dehydrogenase when bound to
testosterone, where testosterone acts as a noncompetitive
inhibitor with respect to ethanol and NAD
+
[33]. Alter-
natively, it is possible that thyroid hormone alters the
equilibrium between the tetramer and dimer conformations
or between the dimer and monomer conformations of
xCTBP/ALDH1, as found in glutamate dehydrogenase,
where T

reached the same result by eluting human ALDHs bound to
AMP-affinity column with T
3
or Triac. These results
support the possibility of a dehydrogenase-specific binding
site for thyroid hormone as the coenzyme-binding domains
within dehydrogenases have a relatively conserved ternary
structure [42] when compared to their catalytic domains.
However, K
i
values for thyroid hormone binding to all
dehydrogenases, including those calculated for xCTBP/
xALDH1, were in the 10
-7
)10
-4
M
range. These are high
concentrations, even if the local distribution or accumula-
tion of intracellular thyroid hormones was considered.
The present studies demonstrate that xCTBP/xALDH1
can bind T
3
in intact cells (Fig. 6). However, the NAD
content corresponding to 0.2 m
M
concentration would
restrict T
3
-binding activity of xCTBP/xALDH1 within the

ACKNOWLEDGEMENTS
We would like to thank Mr Takashi Honda for the preparation of
recombinant xCTBP/xALDH1. We also wish to thank Drs S. Iwamuro
and R. J. Denver for providing the Xenopus cell lines. This work was
supported by Grant-in-Aid for Scientific Research (B) from the Japan
Society for the promotion of Science (no. 13559001).
REFERENCES
1. Chanbon, P. (1995) The molecular and genetic dissection of
the retinoid signaling pathway. Recent. Prog. Horm. Res. 50,
317–332.
2. Napoli, J.L., Posch, K.C., Fiorella, P.D., Boerman, M.H.E.M.,
Salerno, G.J. & Burns, R.D. (1993) Roles of cellular retinol-
binding protein and cellular retinoic acid-binding protein in the
metabolic channeling of retinoids. In Retinoids. Progress in
Research and Clinical Applications. (Livrea, M.A. & Packer, L.,
eds), pp. 29–48. Marcel Dekker, New York.
3. Napoli, J.L. (1996) Retinoic acid biosynthesis and metabolism.
FASEB J. 10, 993–1001.
4. Kato, H., Fukuda. T., Parkinson, C., Mcphie, P. & Cheng, S.Y.
(1989) Cytosolic thyroid hormone-binding protein is a monomer
of pyruvate kinase. Proc. Natl Acad. Sci. USA 86, 7861–7865.
5. Yoshizato, K., Kistler, A. & Frieden, E. (1975) Metal ion
dependence of the binding of triiodothyronine by cytosol proteins
of bullfrog tadpole tissues. J. Biol. Chem. 250, 8337–8343.
6. Hashizume, K., Miyamoto, T., Ichikawa, K., Yamauchi, K.,
Kobayashi, M., Sakurai, A., Ohtsuka, H., Nishii, Y. & Yamada,
T. (1989) Purification and characterization of NADPH-dependent
cytosolic 3,5,3¢-triiodo-thyronine binding protein in rat kidney.
J. Biol. Chem. 264, 4857–4863.
7. Kobayashi, M., Hashizume, K., Suzuki, S., Ichikawa, K. &

(vitamin A1 aldehyde) to vitamin A acid by mammalian steroid-
sensitive aldehyde dehydrogenase. Biochim. Biophys. Acta 64,
430–437.
15. Pereira, F., Rosenmann, E., Nylen, E., Kaufman, M., Pinsky, L.
& Wrogemann, K. (1991) The 56 kDa androgen binding protein is
an aldehyde dehydrogenase. Biochem. Biophys. Res. Commun.
175, 831–838.
16. Wrogemann,K.,Pereira,F.,Belsham,D.,Kaufman,M.,Pinsky,
L. & Rosenmann, E. (1988) An abundant 56 kDa protein with low
affinity androgen binding: another member of the steroid/thyroid
receptor family? Biochem. Biophys. Res. Commun. 155, 907–913.
17. Pereira, F., Belsham, D., Duerksen, K., Rosenmann, E., Kauf-
man, M., Pinsky, L. & Wrogemann, K. (1990) The 56-kDa
androgen-binding protein in human genital skin fibroblasts: its
relation to the human androgen receptor. Mol. Cell. Endocrinol.
68, 195–204.
18. Lesca, P., Peryt, B., Soues, S., Maurel, P. & Gravedi, J.P. (1993)
Detection and characterization of a novel hepatic 8 S binding
protein for benzo[a]pyrene distinct from the Ah receptor. Arch.
Biochem. Biophys. 303, 114–124.
19. Lesca, P., Pineau, T., Galtier, P., Peryt, B. & Derancourt, J. (1998)
The 8S benzo(a)pyrene-binding protein is an aldehyde dehy-
drogenase regulated by the Ah receptor. Biochem. Biophys. Res.
Commun. 242, 26–31.
20. Banfi, P., Lanzi, C., Falvella, S., Gariboldi, M., Gambetta, R.A. &
Dragani, T.A. (1994) The daunorubicin-binding protein of M
r
54,000 is an aldehyde dehydrogenase and is down-regulated in
mouse liver tumors and in tumor cell lines. Mol. Pharmacol. 46,
896–900.

Zool. 284, 207–216.
31. Zhou, J. & Weiner, H. (1997) Binding of thyroxine analogs to
human liver aldehyde dehydrogenases. Eur. J. Biochem. 245,
123–128.
32. Kitson, T.M. (1982) The activation of aldehyde dehydrogenase by
diethylstilboestrol and 2,2¢-dithiodipyridine. Biochem. J. 207,
81–89.
Ó FEBS 2002 Dual activities of xCTBP/xALDH (Eur. J. Biochem. 269) 2263
33. Ma
˚
rdh, G., Falchuk, K.H., Auld, D.S. & Vallee, B.L. (1986)
Testosterone allosterically regulates ethanol oxidation by homo-
and heterodimeric c-subunit-containing isozymes of human alco-
hol dehydrogenase. Proc. Natl Acad. Sci. USA 83, 2836–2840.
34. Wolff, J. (1962) The effect of thyroxine on isolated dehydro-
genases. II. Sedimentation changes in glutamic dehydrogenase.
J. Biol. Chem. 237, 230–235.
35. Wolff, J. & Wolff, E.C. (1957) The effect of thyroxine on isolated
dehydrogenases. Biochim. Biophys. Acta 26, 387–396.
36. Wolff, J. (1962) The effect of thyroxine on isolated dehydrogen-
ases. III. The site of action of thyroxine on glutamic dehydroge-
nase, the function of adenine and guanine nucleotides, and the
relation of kinetic to sedimentation changes. J. Biol. Chem. 237,
236–242.
37. Maggio, E.T. & Ullman, E.F. (1978) Inhibition of malate dehy-
drogenase by thyroxine and structurally related compounds.
Biochim. Biophys. Acta 522, 284–290.
38. McCarthy, K., Lovenberg, W. & Sjoerdsma, A. (1968) The
mechanism of inhibition of horse liver alcohol dehydrogenase
by thyroxine and related compounds. J. Biol. Chem. 243,