Down-regulation of heme oxygenase-2 is associated with
the increased expression of heme oxygenase-1 in human
cell lines
Yuanying Ding
1
, Yong Z. Zhang
1
, Kazumichi Furuyama
1
, Kazuhiro Ogawa
2
*, Kazuhiko Igarashi
3
and Shigeki Shibahara
1
1 Department of Molecular Biology and Applied Physiology, Tohoku University School of Medicine, Sendai, Japan
2 Department of Molecular Pharmacology, Tohoku University School of Medicine, Sendai, Japan
3 Department of Biochemistry, Tohoku University Graduate School of Medicine, Sendai, Japan
Heme is an invaluable molecule that is essential for life
and is involved in many cellular processes that sense
or use oxygen. The intracellular concentration of heme
is maintained by the rate of its synthesis and degrada-
tion [1]. Many enzymes and their regulators are
responsible for heme synthesis [1,2]. On the other
hand, heme degradation is mediated by two structur-
ally related isozymes, HO-1 and HO-2, to generate
biliverdin IXa, carbon monoxide (CO), and ferrous
iron [3]. Biliverdin IXa is immediately reduced to bili-
rubin IXa. HO-1 has attracted particular attention,
because its expression is induced by its substrate,
heme, in animals [4,5] and in primary cultures of

HO-2 mRNA and protein are expressed in the eight human cancer cell
lines examined, and HO-1 expression is detectable in five of the cell lines,
including HeLa cervical cancer and HepG2 hepatoma. Down-regulation of
HO-2 expression with siRNA against HO-2 (siHO-2) caused induction of
HO-1 expression at both mRNA and protein levels in HeLa and HepG2
cells. In contrast, knockdown of HO-1 expression did not noticeably influ-
ence HO-2 expression. HO-2 knockdown prolonged the half-life of HO-1
mRNA twofold in HeLa cells. Transient transfection assays in HeLa cells
revealed that the 4.5-kb human HO-1 gene promoter was activated with
selective knockdown of HO-2 in a sequence-dependent manner. Moreover,
HO-2 knockdown caused heme accumulation in HeLa and HepG2 cells
only when exposed to exogenous hemin. HO-2 knockdown may mimic a
certain physiological change that is important in the maintenance of cellu-
lar heme homeostasis. These results suggest that HO-2 may down-regulate
the expression of HO-1, thereby directing the co-ordinated expression of
HO-1 and HO-2.
Abbreviations
GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HO, heme oxygenase; MARE, Maf recognition element; SA, succinylacetone; SnPP,
Sn-protoporphyrin.
FEBS Journal 273 (2006) 5333–5346 ª 2006 The Authors Journal compilation ª 2006 FEBS 5333
On the other hand, it has been reported that expres-
sion of HO-1 is decreased in several types of human
cell under various conditions, such as hypoxia [12,13]
and treatment with interferon-c [14,15] or desferriox-
amine, an iron chelator [12]. Likewise, the expression
of HO-2 is decreased in the placental tissues of abnor-
mal pregnancies [16,17] and in cultured human tropho-
blast cells [18]. We have recently shown that the
expression levels of HO-1 and HO-2 are decreased in
several human cell lines under hypoxia [19]. In mice,

down is associated with the induction of HO-1 expres-
sion in human cancer cell lines. HO-2 knockdown may
mimic a certain physiological change that is important
in the maintenance of cellular heme homeostasis. We
provide evidence that HO-2 may modulate the expres-
sion level of HO-1 by affecting HO-1 mRNA stability
and intracellular heme concentration.
Results
Expression profiles of HO-1 and HO-2 in various
human cell types
We initially analyzed the expression profiles of HO-1
and HO-2 in eight human cell lines by northern and
western blot analyses (Fig. 1A,B). Expression of HO-1
mRNA was detected in five of these cell lines, but
hardly at all in the other three (K562 erythroleukemia,
Jurkat T cell, and H146 small cell lung cancer), in
which HO-1 protein was also undetectable. In contrast,
HO-2 mRNA and protein were both expressed in all
eight cell lines (Fig. 1A,B). We also measured the heme
content of these cell lines: it was about twofold higher
in YN-1 and K562 erythroleukemia cells than in the
other cell types (Fig. 1C). Higher heme content may
reflect hemoglobin production in YN-1 and K562 cells
[33,34]. In fact, the population of hemoglobin-positive
cells was about 4.4% in YN-1 cells and 4.9% in K562
cells under basal conditions [19]. Otherwise, there was
A
B
C
Fig. 1. Expression profiles of HO-1 and HO-2 in various human cell

biosynthetic pathway. SnPP is a competitive inhibitor
of HO activity [35]. The heme content of all three cell
lines was significantly decreased after treatment with
SA, but increased after treatment with SnPP
(Fig. 2A,B). Thus, an appropriate balance between
heme synthesis and heme degradation is responsible
for maintenance of cellular heme contents. More
importantly, these results indicate that measurement
of heme content is useful for evaluation of heme
dynamics in cultured cells.
Regulatory role of free heme in expression
of HO-1
To explore the role of free heme in HO-1 and HO-2
expression, we treated HeLa and HepG2 cells with
SA and determined the expression levels of HO-1 and
HO-2 (Fig. 3). HO-1 mRNA expression was signifi-
cantly reduced in HeLa and HepG2 cells after treat-
ment with SA for 6 h, whereas HO-2 mRNA
expression was not noticeably changed by SA treat-
ment (Fig. 3A,C). Western blot analysis revealed that
treatment with SA reduced the expression of HO-1
protein in HeLa and HepG2 cells, but did not change
the HO-2 protein concentration (Fig. 3B,D). These
results suggest that a certain threshold concentration
of free heme may determine the basal expression levels
of HO-1.
Effects of HO-1 or HO-2 short interfering RNA
(siRNA) on the expression of HO-1
To explore the functional significance of HO-1 and
HO-2, we selectively reduced the expression of HO-1

contents are shown as ng ⁄ 10
6
cells. The data are mean ± SEM
from three independent experiments. *P < 0.05, **P<0.01.
Y. Ding et al. Heme oxygenase-2 down-regulates heme oxygenase-1
FEBS Journal 273 (2006) 5333–5346 ª 2006 The Authors Journal compilation ª 2006 FEBS 5335
We then confirmed the effects of HO-2 knockdown
using another HO-2 siRNA with a different sequence
(HO-2 siRNA1) [32] and a scrambled HO-2 siRNA
(siHO-2-R) in HeLa and HepG2 cells (Fig. 5). HO-2
siRNA1 efficiently decreased the expression of HO-2
mRNA and protein and induced the expression of
HO-1 mRNA and protein. In contrast, the scrambled
HO-2 siRNA did not affect the expression of HO-1
and HO-2 mRNAs and proteins in HeLa and HepG2
cells (Fig. 5). Thus, the induction of HO-1 is due to
the selective repression of HO-2 expression achieved
with HO-2 siRNA.
Knockdown of HO-2 expression causes
time-dependent induction of HO-1 mRNA
expression
We then performed a time course study to confirm the
effects of HO-2 siRNA on the expression of HO-1
mRNA in HeLa cells (Fig. 6A). HO-2 siRNA efficiently
reduced the expression of HO-2 mRNA at 6 h, which
was further decreased at 12 h. In contrast, expression of
HO-1 mRNA was time-dependently increased, reaching
a maximum at 24 h (Fig. 6A,B). HO-2 siRNA had not
noticeably changed the concentrations of GAPDH
A

HO-2 mRNA was normalized with respect to the intensity of the 18S rRNA signal. The ratio of each normalized value to the value in
untreated cells is shown as the relative expression level of HO-1 or HO-2 mRNA (**P<0.01). (B and D) Western blot analysis of HO-1 and
HO-2 proteins. The intensity representing HO-1 or HO-2 protein was normalized with respect to the intensity of the b-actin signal. The ratio
of each normalized value to the control value in siRNA-untreated cells (control) is shown as the relative expression level of HO-1 or HO-2
protein (*P < 0.05, **P<0.01). The data are mean ± SEM from three independent experiments.
Y. Ding et al. Heme oxygenase-2 down-regulates heme oxygenase-1
FEBS Journal 273 (2006) 5333–5346 ª 2006 The Authors Journal compilation ª 2006 FEBS 5337
knockdown may evoke a certain metabolic change,
which in turn induces HO-1 mRNA expression.
HO-2 knockdown increases stability of HO-1
mRNA
Consequently, we analyzed the stability of HO-1
mRNA in HeLa cells treated with HO-2 siRNA. In
this series of experiments, HeLa cells were left untrans-
fected or transfected with the indicated siRNA, and
cultured for 12 h before addition of actinomycin D.
The half-life of HO-1 mRNA was about 3 h in
untransfected and siGAPDH-treated HeLa cells
(Fig. 7A,B), which is in good agreement with the half-
life of HO-1 mRNA determined in HeLa cells [37].
Interestingly, the half-life of HO-1 mRNA was pro-
longed to about 7 h in HeLa cells transfected with
siHO-2. Thus, the induction of HO-1 mRNA with
HO-2 knockdown may be in part due to the increased
stability of HO-1 mRNA.
Effects of HO-2 knockdown on HO-1 and HO-2
promoter activities
To assess the biochemical consequences of HO-2
knockdown, we analyzed whether HO-2 siRNA affects
the promoter activity of the human HO-1 gene, using

Heme oxygenase-2 down-regulates heme oxygenase-1 Y. Ding et al.
5338 FEBS Journal 273 (2006) 5333–5346 ª 2006 The Authors Journal compilation ª 2006 FEBS
HO-2 knockdown may transactivate the promoter of
phHOLUC45 through a heme-independent mechanism.
Moreover, HO-2 knockdown may cause a metabolic
change similar to that evoked by cadmium [36] or
sodium nitroprusside [38], each of which activated the
expression of a reporter gene under the regulation of
the 4.5-kb HO-1 gene promoter. It should be noted,
however, that heme activates the HO-1 gene promoter
[40], but the relevant cis-acting element is not present
in the 4.5-kb promoter region. It is therefore conceiv-
able that HO-2 knockdown may induce HO-1 expres-
sion through not only the heme-independent
mechanism but also the heme-dependent mechanism.
Heme accumulation caused by knockdown of
HO-1 or HO-2 expression
To explore the biological implication of the knock-
down experiments and to evaluate the relative contri-
bution of HO-1 and HO-2 to the total amount of
heme degradation, we measured the heme content of
HeLa cells and HepG2 cells after transfection with
HO-1 or HO-2 siRNA. There were no significant chan-
ges in heme content in HeLa cells (Fig. 9A) and
HepG2 cells (Fig. 9B), which were transfected with
each HO siRNA. Thus, heme content may be main-
tained by a compensatory mechanism, or the changes
in heme content may be below the detectable limit of
the assay method used. Accordingly, we treated HeLa
and HepG2 cells for 12 h with 1 lm hemin, a subopti-

resulted in the accumulation of heme in the hemin-
treated HepG2 cells. Taken together with the siHO-2-
mediated induction of HO-1 expression, these results
suggest that HO-2 rather than HO-1 may play the pre-
dominant role in heme degradation in cultured human
cells.
Discussion
We have shown that the selective knockdown of HO-2
expression with each of two different siRNAs is consis-
tently associated with increased expression of HO-1
mRNA and protein. Moreover, we provide evidence
that at least three mechanisms may account for the
siHO-2-mediated induction of HO-1 expression:
increased stability of HO-1 mRNA and heme-depend-
ent and heme-independent transcriptional regulation of
the HO-1 gene. These metabolic consequences of HO-2
knockdown suggest a regulatory role for HO-2 in the
co-ordinated expression of HO-1 and HO-2. We there-
fore propose that HO-2 may modulate the expression
of HO-1.
Incidentally, the three cell lines with low HO-1
expression, K562, Jurkat, and H146, were maintained
in suspension culture (Fig. 1), although HO-1 is
expressed in two other cell lines, YN-1 and KG1, that
were also maintained in suspension culture. These
results suggest that the cellular microenvironment,
such as cell attachment, may influence the expression
of HO-1. The dominant expression of HO-2 protein, in
comparison with the low expression of HO-1 protein,
in H146 small cell lung cancer cells is of particular

mice.
Unlike the partial lethality of HO-1
– ⁄ –
[27], HO-2
– ⁄ –
mice are able to survive for at least a year under basal
conditions [28]. These mice can probably compensate
for the loss of HO-2 by increasing the expression of
HO-1, which is supported by the following observa-
tions. First, HO-2
– ⁄ –
mice show no noticeable changes
or only a marginal decrease in arterial carboxyhemo-
globin, a marker of overall heme degradation [27,28].
Secondly, no differences in heme concentration were
A
B
Fig. 7. HO-2 knockdown increases the stability of HO-1 mRNA. (A)
Northern blot analysis. HeLa cells, which were left untransfected
(Control) or transfected with the indicated siRNA, were cultured for
12 h, and then treated with actinomycin D (AMD) (1 lgÆmL
)1
) for
the indicated time (h). Each lane contains 15 lg total RNA. The lane
labeled 0 h contained RNA prepared from cells harvested just
before the addition of AMD (0 h). (B) Relative expression levels of
HO-1 mRNA. The intensity representing HO-1 mRNA was normal-
ized with respect to the intensity of 18S rRNA. The intensity repre-
senting HO-1 mRNA at the time of addition of AMD (0 h) under
each condition was considered to be 1. One representative of two

ency is characterized by hypobilirubinemia, persistent
hemolytic anemia, and iron deposits in the liver [43].
HO-2 knockdown increased the transient expression
of phHOLUC45 through the enhancer region of the
human HO-1 gene, located between )4.5 kb and
)4 kb. The increased expression of phHOLUC45 sug-
gests that the cellular microenvironment generated by
HO-2 knockdown may mimic the metabolic change
evoked by cadmium [36] or sodium nitroprusside [38],
each of which activates the expression of a reporter
gene under the regulation of the 4.5-kb HO-1 gene
promoter. This region contains the cadmium-respon-
sive element [36] and the MARE [44,45], but lacks
the element required for full activation by hemin
[9,36,40]. It is the MARE site that is bound by Nrf2,
a transcription activator, or Bach1, a transcription
repressor, each of which functions as a heterodimer
with a member of the Maf family [46]. Bach1 is a
heme-responsive repressor, and its repression activity
is lost when Bach1 is bound by heme, which in turn
leads to transcriptional activation of the HO-1 gene
A
B
Fig. 8. Knockdown of HO-2 expression increases the HO-1 promoter activity. HeLa cells were transiently transfected with each reporter con-
struct (3 lg), then incubated for 24 h, and either re-transfected with each siRNA (A) or treated with 50 l
M hemin (B), as described in Experi-
mental procedures. After 24 h of incubation, luciferase activity was measured. The test promoters analyzed are shown on the left. Relative
luciferase activity is shown as the ratio to the normalized luciferase activity obtained with pGL3Basic; the normalized luciferase activity used
was that in cells treated with siGAPDH in (A) and that in cells untreated with hemin in (B). The data are mean ± SEM from three independ-
ent experiments. **P < 0.01.

from Sigma Chemical (St Louis, MO, USA). SnPP was
from Porphyrin Products (Logan, UT, USA).
Cell cultures
Human cell lines used were HeLa cervical carcinoma cells,
HepG2 hepatoma cells, K562 and YN-1 erythroleukemia
cells, Jurkat T-lymphocyte cells, KG1 myeloid cells, H146
small cell lung cancer cells, and HMV-II melanoma cells.
H146 small cell lung cancer cells were obtained from
ATCC (HTB-173) and cultured in RPMI-1640 medium.
HMV-II melanoma cells were obtained from Riken Cell
Bank and cultured in nutrient mixture Ham’s F12 med-
ium. HeLa and HepG2 cells were maintained in Dul-
becco’s modified Eagle’s medium (Sigma). YN-1 cells were
maintained in Iscove’s modified Dulbecco’s medium (Sigma),
and K562, KG1 and Jurkat cells were maintained in
RPMI-1640 medium (Sigma). Each medium contained
10% heat-inactivated fetal bovine serum, penicillin G
(100 UÆmL
)1
), and streptomycin sulfate (100 lgÆmL
)1
).
Cells were incubated at 37 °C under 5% CO
2
⁄ 95% room
air, unless otherwise specified. HepG2, HeLa and YN-1
cells were treated with 50 lm SnPP or 5 mm SA for up to
48 h. SnPP was freshly prepared and added immediately
to the culture medium. The culture dishes were placed in
the incubator.

Heme content of cultured cells
Heme content (expressed as ng ⁄ 10
6
cells) was determined as
described previously [19]. Cell pellet was dissolved in
0.5 mL 2 m oxalic acid by shaking vigorously and immedi-
ately heated for 30 min at 100 °C. Mixtures that had not
been heated were used as a blank for each measurement of
endogenous porphyrins. After cooling down, fluorescence
was measured in a RF-5300PC spectrofluorometer (Shim-
adzu Corp., Kyoto, Japan). Under the conditions used, the
lowest limit of detection is about 1 ng heme ⁄ assay. Cells
(1 · 10
6
) were used to determine heme content in all assays.
In some experiments, HeLa cells and HepG2 cells were cul-
tured for 12 h after the transfection with siHO1, siHO2, or
siGAPDH, then treated with 1 lm hemin for 12 h, and har-
vested for heme measurement.
SiRNA and expression plasmids of HO-1 and
HO-2
A specific siRNA against HO-1, siHO-1, which was repor-
ted by Miralem et al. [53], was used. HO-2 siRNA (target
base 248–272), named siHO-2, was designed and syn-
thesized by iGENE Therapeutics (Tsukubu, Japan), and
scrambled HO-2 siRNA was used as a negative control:
HO-2 siRNA: sense, 5¢-AGGACUUCUUGAAAGGCAA
CAUUAAAG-3¢, antisense, 3¢-UAUCCUGAAGAACUU
UCCGUUGUAAUU-5¢; scrambled HO-2 siRNA: sense,
5¢-UAUAAGAGUCAGUACACAUCAUGGAAG-3¢,anti-

were: forward, 5¢-
AAGCTTCATGTCAGCGGAAGTG
GAAAC-3¢; reverse, 5¢-CTGCAGTCACATGTAGTACC
AGGCCAA-3¢. The sequence underlined is an artificial
HindIII site. A full-length HO-2 cDNA fragment was iso-
lated from pCR-hHO-2-1 with EcoR1 and subcloned into
the pMACS4-IRES vector (Miltenyi Biotec Inc., Bergisch-
Gladbach, Germany), generating the HO-2 expression
vector, pMACS-hHO-2.
Effects of HO-2 siRNA on the stability of HO-1
mRNA
To study the effects of HO-2 knockdown on the stability of
HO-1 mRNA, HeLa cells were left untrasnfected or trans-
fected with siHO-2, and incubated for 12 h, followed by the
addition of actinomycin D (1 lgÆmL
)1
) [37]. The cells were
further incubated for up to 7 h, and then harvested at each
time point for RNA preparation.
Luciferase assays
The luciferase reporter constructs used were the human
HO-1 gene constructs, phHOLUC45 and phHOLUC40
[14,36,55,56], and the human HO-2 gene constructs,
phHO2()1494) and phHO2()663) [19]. The test plasmids,
pRBGP2 and pRBGP4, contain three copies of the MARE
and mutated MARE, respectively, in each promoter region
linked to the luciferase gene [39]. HeLa cells were plated
1 day before transfection and grown to 50–70% confluence
in 24-well plates.
For siRNA experiments, HeLa cells at 50% confluence

tion of cellular iron and heme: endless findings and
growing evidence of tight regulation. Tohoku J Exp
Med 205, 297–318.
2 Furuyama K & Sassa S (2000) Interaction between suc-
cinyl CoA synthetase and the heme-biosynthetic enzyme
ALAS-E is disrupted in sideroblastic anemia. J Clin
Invest 105, 757–764.
3 Yoshida T & Kikuchi G (1978) Purification and proper-
ties of heme oxygenase from pig spleen microsomes.
J Biol Chem 253, 4230–4236.
4 Tenhunen R, Marver HS & Schmid R (1970) The enzy-
matic catabolism of hemoglobin stimulation of microso-
mal heme oxygenase by hemin. J Lab Clin Med 75,
410–421.
5 Pimstone NR, Engel P, Tenhunen R, S eitz PT, Marver HS
& Schmid R (1971) Inducible heme oxygenase in the
kidney: a model for the homeostatic control of hemo-
globin catabolism. J Clin Invest 50, 2042–2050.
6 Pimstone NR, Tenhunen R, Seitz PT, Marver HS &
Schmid R (1971) The enzymatic degradation of hemo-
globin to bile pigments by macrophages. J Exp Med
133, 1264–1281.
7 Shibahara S, Yoshida T & Kikuchi G (1978) Induction
of heme oxygenase by hemin in cultured pig alveolar
macrophages. Arch Biochem Biophys 188, 243–250.
8 Shibahara S, Yoshida T & Kikuchi G (1979) Mechan-
ism of increase of heme oxygenase activity induced by
hemin in cultured pig alveolar macrophages. Arch Bio-
chem Biophys 197, 607–617.
9 Okinaga S, Takahashi K, Takeda K, Yoshizawa M,

oxygenase-1 by nitric oxide donors. J Neurochem 67,
482–489.
16 Zenclussen AC, Lim E, Knoeller S, Knackstedt M, Her-
twig K, Hagen E, Klapp BF & Arck PC (2003) Heme
oxygenases in pregnancy II: HO-2 is downregulated in
human pathologic pregnancies. Am J Reprod Immunol
50, 66–76.
17 Lash GE, McLaughlin BE, MacDonald-Goodfellow SK,
Smith GN, Brien JF, Marks GS, Nakatsu K & Graham
CH (2003) Relationship between tissue damage and heme
oxygenase expression in chorionic villi of term human
placenta. Am J Physiol Heart Circ Physiol 284,
H160–H167.
18 Appleton SD, Mark SGS, Nakatsu K, Brien JF,
Smith GN, Graham CH & Lash GE (2003) Effects of
hypoxia on heme oxygenase expression in human
chorionic villi explants and immortalized trophoblast
cells. Am J Physiol Heart Circ Physiol 284, H853–H858.
19 Zhang Y, Furuyama K, Kaneko K, Ding Y, Ogawa K,
Yoshizawa M, Kawamura M, Takeda K, Yoshida T &
Shibahara S (2006) Hypoxia reduces the expression of
heme oxygenase-2 in various types of human cell lines.
FEBS J 273, 3136–3147.
20 Zenclussen AC, Joachim R, Hagen E, Peiser C,
Klapp BF & Arck PC (2002) Heme oxygenase is down-
regulated in stress-triggered and interleukin-12-mediated
murine abortion. Scan J Immunol 55, 560–569.
21 Han F, Takeda K, Yokoyama S, Ueda H, Shinozawa Y,
Furuyama K & Shibahara S (2005) Dynamic changes in
expression of heme oxygenases in mouse heart and liver

Calvin DC, Demas GE, Schramm LP, Tonegawa S,
Nelson RJ, Snyder SH, et al. (1998) Ejaculatory
abnormalities in mice with targeted disruption of the
gene for heme oxygenase-2. Nat Med 4, 84–87.
30 Dennery PA, Spitz DR, Yang G, Tatarov A, Lee CS,
Shegog ML & Poss KD (1998) Oxygen toxicity and iron
accumulation in the lungs of mice lacking heme oxyge-
nase-2. J Clin Invest 101, 1001–1011.
31 Adachi T, Ishikawa K, Hida W, Matsumoto H,
Masuda T, Date F, Ogawa K, Takeda K, Furuyama K,
Zhang Y, et al. (2004) Hypoxemia and blunted hypoxic
ventilatory responses in mice lacking heme oxygenase-2.
Biochem Biophys Res Commun 320, 514–522.
32 Williams SE, Wootton P, Mason HS, Bould J, Iles DE,
Riccardi D, Peers C & Kemp PJ (2004) Hemoxygenase-
2 is an oxygen sensor for a calcium-sensitive potassium
channel. Science 306, 2093–2097.
33 Endo K, Harigae H, Nagai T, Fujie H, Meguro K,
Watanabe N, Furuyama K, Kameoka J, Okuda M &
Hayashi N (1993) Two chronic myelogenous leukaemia
cell lines which represent different stages of erythroid
differentiation. Br J Haematol 85, 653–662.
34 Rutherford TR, Clegg JB & Weatherall DJ (1979) K562
human leukaemic cells synthesise embryonic haemoglo-
bin in response to haemin. Nature 280, 164–165.
35 Sassa S (1997) Inhibition of carbon monoxide produc-
tion by tin-protoporphyrin. J Pediatr Gastroenterol Nutr
6, 167–171.
36 Takeda K, Ishizawa S, Sato M, Yoshida T & Shibahara S
(1994) Identification of a cis-acting element that is

S & Snyder SH (1997) Targeted gene deletion of heme
oxygenase 2 reveals neural role for carbon monoxide.
Proc Natl Acad Sci USA 94, 14848–14853.
43 Yachie A, Niida Y, Wada T, Igarashi N, Kaneda H,
Toma T, Ohta K, Kasahara Y & Koizumi S (1999) Oxi-
dative stress causes enhanced endothelial cell injury in
human heme oxygenase-1 deficiency. J Clin Invest 103,
129–135.
44 Oyake T, Itoh K, Motohashi H, Hayashi N, Hoshino H,
Nishizawa M, Yamamoto M & Igarashi K (1996) Bach
proteins belong to a novel family of BTB-basic leucine
zipper transcription factors that interact with MafK and
regulate transcription through the NF-E2 site. Mol Cell
Biol 16, 6083–6095.
45 Igarashi K & Sun J (2006) The heme-Bach1 pathway
in the regulation of oxidative stress response and ery-
throid differentiation. Antioxid Redox Signal 8, 107–
118.
46 Ogawa K, Sun J, Taketani S, Nakajima O, Nishitani C,
Sassa S, Hayashi N, Yamamoto M, Shibahara S,
Fujita H, et al. (2001) Heme mediates derepression of
Maf recognition element through direct binding to tran-
scription repressor Bach1. EMBO J 20, 2835–2843.
47 Sun J, Hoshino H, Takaku K, Nakajima O, Muto A,
Suzuki H, Tashiro S, Takahashi S, Shibahara S, Alam J,
et al. (2002) Hemoprotein Bach1 regulates enhancer
availability of heme oxygenase-1 gene. EMBO J 21,
5216–5224.
48 Alam J, Stewart D, Touchard C, Boinapally S, Choi AM
& Cook JL (1999) Nrf2, a Cap’n’Collar transcription

54 Shibahara S, Yoshizawa M, Suzuki H, Takeda K,
Meguro K & Endo K (1993) Functional analysis of
cDNAs for two types of human heme oxygenase and
evidence for their separate regulation. J Biochem
(Tokyo) 113, 214–218.
55 Shibahara S, Sato M, Muller RM & Yoshida T (1989)
Structural organization of the human heme oxygenase
gene and the function of its promoter. Eur J Biochem
179, 557–563.
56 Takahashi S, Takahashi Y , Ito K, Nagano T, Shibahara S
& Miura T (1999) Positive and negative regulation of
the human heme oxygenase-1 gene expression in cul-
tured cells. Biochim Biophys Acta 1447, 231–235.
Heme oxygenase-2 down-regulates heme oxygenase-1 Y. Ding et al.
5346 FEBS Journal 273 (2006) 5333–5346 ª 2006 The Authors Journal compilation ª 2006 FEBS