Adaptive changes in the expression of nuclear and mitochondrial
encoded subunits of cytochrome
c
oxidase and the catalytic
activity during hypoxia
C. Vijayasarathy
1,
*
,†
, Shirish Damle
1,
*, Subbuswamy K. Prabu
1,
*, Cynthia M. Otto
2
and Narayan G. Avadhani
1
1
Department of Animal Biology and
2
Department of Clinical Studies, School of Veterinary Medicine, University of Pennsylvania,
Philadelphia, PA, USA
The effects of physiologically relevant hypoxia on the
catalytic activity of cytochrome c oxidase (CytOX), mito-
chondrial gene expression, and both nuclear and mito-
chondrial encoded CytOX mRNA levels were investigated
in murine monocyte macrophages, mouse C2C12 skeletal
myocytes and rat adrenal pheochromocytoma PC12 cells.
Our results suggest a coordinated down regulation of mito-
chondrial genome-coded CytOX I and II and nuclear
genome-coded CytOX IV and Vb mRNAs during hypoxia.

ferrocytochrome c to molecular oxygen, constitute the
catalytic site of the enzyme complex. The three catalytic
subunits, CytOX I, II and III are coded by the mitochon-
drial DNA and are synthesized within mitochondria.
Heme a, heme a
3
and Cu
b
are ligated to subunit I, while
Cu
a
is ligated to subunit II which is also the binding site for
cytochrome c [4,5]. The remaining 10 subunits of the
mammalian enzyme, namely, IV, Vb, VIa, VIb, VIc, VIIa,
VIIb and VIII are encoded by the nuclear genome,
synthesized in the cytosol and imported into mitochondria
[1–3]. Some of the nuclear-encoded subunits in the mam-
mals are regulated developmentally and occur as tissue
specific isoforms [6,7]. Although the nuclear encoded
subunits, such as CytOX VIa and VIb, have been shown
to enhance the catalytic efficiency of the enzyme [8,9], the
precise role of many nuclear-encoded subunits in the
mammalian enzyme complex remains unknown.
Oxygen as a substrate and heme as a prosthetic group, are
closely interlinked in the function of the enzyme complex.
Studies in yeast have shown that both oxygen and heme act
as physiological modulators and regulate the expression of
the enzyme complex [10]. In the yeast CytOX complex, the
nuclear encoded subunit V is expressed as two distinct
isoforms, Va and Vb, that are regulated by heme and O

enzyme following heme reduction. These studies suggested
that heme might regulate not only the mammalian CytOX
gene expression but also the catalytic activity of the
enzyme by affecting its stability or composition. Although
succinylacetone and CoCl
2
are known inhibitors of heme
biosynthesis, these agents also elicit nonspecific and toxic
effects in animals. To ascertain that the effects of these
agents, on the catalytic activity and subunit composition
of CytOX were related to their hypoxia-specific effects, we
have extended our investigation to physiological hypoxia
in cultured cells. We focused our attention on the
expression of mitochondrial genome encoded catalytic
subunits I and II, and the nuclear genome encoded
subunits, IV, Vb and VIIa. The mammalian CytOX
subunit IV is a homolog of the yeast subunit V, with the
latterexpressedinanO
2
and heme regulated manner.
In this study therefore, we investigated how O
2
modulates
the expression of mRNAs for CytOX subunits and also the
catalytic activity of the enzyme complex. Our results show
that the levels of some of the select mitochondrial and
nuclear genome encoded CytOX mRNAs are uniformly
down regulated during hypoxia. Results also show changes
in the composition and activity of the enzyme complex,
which is accompanied by alterations in cellular ATP and

)1
). All the cells were grown to 80–90%
confluence in a controlled humidified environment (21%
O
2
,5%CO
2
, remainder N
2
at 37 °C).
Exposure of cells to hypoxia
The normal range of tissue oxygen tension (in nonpul-
monary tissues) measured under in vivo conditions ranges
from 5 to 71 Torr, with most tissues maintaining a pO
2
of
40 Torr or less. Simulation of realistic in vivo hypoxia
requires that O
2
tension is maintained at less than 5 Torr
[13]. We have used modular incubator chambers (Billups-
Rothernberg, CA, USA) for creation of a nonfluctuating
hypoxic environment. The chambers were maintained at
37 °C in a humidified incubator. Cells grown in tissue
culture dishes were introduced into the chambers that
were directly connected to certified premixed compressed
gas cylinders. The modular chambers were purged with a
constant flow of premixed gas that was certified to contain
1 Torr (hypoxic) or 141 Torr of oxygen (normoxic), all
with 5% CO

A, 5 lgÆmL
)1
aprotinin, and 1 m
M
phenylmethanesulfonyl
fluoride]) and lysed by three cycles of freezing and thawing
in liquid nitrogen. The lysates were centrifuged at 4 °Cfor
30 min in an Eppendorf centrifuge tube to remove debris
and unlysed cells. The supernatant was collected and stored
at )80 °C till further use. Protein concentrations were
determined using Lowry’s method [14].
Mitochondria were isolated form intact cells by differen-
tial centrifugation as described earlier [17]. The cell pellets
were suspended in H medium (70 m
M
sucrose, 220 m
M
mannitol, 2.5 m
M
Hepes, pH 7.4 and 2 m
M
EDTA) and
ruptured by homogenization through a Dounce homo-
genizer. Submitochondrial particles (SMP) were prepared
according to the method of Pederson et al. [18], and washed
three times with mitochondrial isolation buffer. All steps of
subcellular fractionation and isolation of SMP were carried
out at 4 °C.
Northern blot analysis
RNA was isolated from cells grown under hypoxic and

pH 7.4, 60 m
M
KCl, 10 m
M
MgCl
2
,5m
M
2-mercapto-
ethanol, 10 m
M
KH
2
PO
4
(pH 7.4), 0.14
M
sucrose, 2 m
M
ATP, 1 m
M
each of GTP and CTP, 5 m
M
pyruvate, 5 m
M
creatine phosphate, 0.2 mgÆmL
)1
creatine phosphokinase
and 100 l
M

and heme content
CytOX was assayed in membrane fragments (SMP) by the
method of Smith [20], wherein the rate of oxidation of
ferrocytochrome cwas measured by following the decrease in
absorbency of its a band at 550 nm. The reaction medium
contained 50 m
M
PO
4
(pH 7.0), 1% sodium cholate, 80 l
M
ferrocytochrome c,1m
M
EDTA and 1–2 lg of protein in a
total volume of 1 mL. Reaction rates were measured using
Cary-1E spectrophotometer (Varian Instruments Walnut
Creek, CA, USA). First order rate constants were calculated
from mean values of four measurements. The heme aa
3
content was calculated from the difference spectra (dithio-
nate/ascorbate reduced minus ferricyanide oxidized) of
mitochondria or SMP solubilized in 2% lauryl maltoside
using an absorption coefficient of 24 m
M
)1
Æcm
)1
at 605–
630 nm [21].
Electrophoresis of proteins and immunoblot analysis

Bis-Tris, pH 7.0) and
analyzed by BN/PAGE. Marker proteins such as b-amy-
lase, 200 kDa; apo-ferritin, 443 kDa and thyroglobulin,
669 kDa (Sigma Chemical Company) were included as
standards. Electrophoresis was carried out initially at 100 V
until the protein samples were within the stacking gel, and
then at a constant current of 18 mA (500 V) for 5–6 h. The
proteins were transblotted onto a poly(vinylidene difluoride)
membrane and probed with subunit-specific monoclonal
or polyclonal antibodies and appropriate horseradish
peroxidase (HRP)-conjugated secondary antibody. The
immunoblot was developed using the Super Signal ULTRA
chemiluminescent substrate kit (Pierce Chemical Co),
imaged and quantitated in a Bio-Rad Fluor-S imaging sys-
tem. Subunit-specific monoclonal antibodies for CytOX I,
IV and Vb proteins were obtained from Molecular Probes
(Eugene, OR, USA) and the specificity of each antibody was
tested by immunoblot analysis of purified CytOX complex.
Polyclonal antibody to rat liver F
1
ATPase was a kind gift
from P. L. Pederson (Johns Hopkins University, Baltimore,
MD, USA).
Measurement of cellular ATP levels
Cellular ATP levels were measured using a somatic cell ATP
assay kit (Sigma Chemical Co, St Louis, MO, USA), which
is based on the assay of ATP driven luciferin luciferase assay
system. Cells were lysed with ATP releasing agent as per
manufacturer’s protocol and ATP levels were measured in a
TD-20/20 Luminometer (Turner Designs, Sunnyvale, CA,

assaying isocitrate dehydrogenase [26], hexokinase [27] and
phosphofructokinase [28] enzyme activities by published
methods.
Ó FEBS 2003 Cytochrome c oxidase subunit expression in hypoxia (Eur. J. Biochem. 270) 873
Results
As a measure of the hypoxic effect on CytOX gene
expression, we measured the steady state CytOX mRNA
levels. The Northern blot in Fig. 1 shows the effect of
hypoxia on the levels of mitochondrial genome- and nuclear
genome-coded mRNAs for CytOX subunits. There were no
detectable changes in mRNA levels for CytOX subunits in
macrophages (Fig. 1A), PC12 and C2C12 cells (results not
shown) after 3 h of exposure to hypoxia. As CytOX has a
low K
m
for oxygen, it is conceivable that an adaptive
response to hypoxia might only be seen after a long period
of exposure to low O
2
. Changes in mRNA levels became
apparent only after 6 h of exposure to hypoxia. The
mitochondrial genome coded CytOX I and II mRNAs were
reduced by 60–70% in macrophages (Fig. 1A), PC12 cells
(Fig. 1B) and differentiated C2C12 myotubes (Fig. 1C),
after hypoxic exposures ranging from 6 to 10 h. The time
point at which a 40–60% reduction in mRNA levels
occurred varied between different cell types. Macrophages
and C2C12 myotubes showed a 50% reduction in mRNA
levels after 6 h of exposure to hypoxia, while PC12 cells
showed a similar reduction after 10 h of exposure to

genes coding for CytOX are coordinately down regulated
during hypoxia. Although not shown, the reversibility of
mRNA levels and also other biochemical parameters tested
in this study were limited to hypoxic exposure up to a
threshold limit. This threshold limit ranged from 10 to 16 h
for different cells. Our results point to differential sensitivity
of cell types to hypoxia, though the effect on CytOX gene
expression was similar in all the cell types studied. Based on
Fig. 1. Effects of hypoxia on the steady state levels of nuclear and
mitochondrial coded CytOX mRNAs. Northern blot analysis was car-
ried out with total RNA (25 lg each) from macrophages (A, E), PC12
cells (B,E), myotubes (C) and myocytes (D) exposed to normoxia
(141 Torrr O
2
)orhypoxia(1TorrO
2
) for 10 h. Hybridization with
32
P-labeled probes was carried out as described in the Materials and
Methods section. The stripped blots were rehybridized with a [
32
P]18S
rDNA probe to determine the RNA loading. The blots were scanned
in a Bio-Rad GS-525 Molecular Imager. The values were normalized
to the 18S rRNA level in each lane.
874 C. Vijayasarathy et al.(Eur. J. Biochem. 270) Ó FEBS 2003
this observation we restricted our subsequent investigations
to macrophages and PC12 cells.
In order to understand the basis for reduced CytOX I
and II mRNA levels, we studied the rate of transcrip-

levels were restored to near normal levels following exposure
of cells to normoxic conditions. It is likely that the reduced
mtTFA level is a factor in reduced mitochondrial mRNA
levels during hypoxia.
Fig. 2. Hypoxia mediated inhibition of mitochondrial transcription. In organelle RNA synthesis with isolated mitochondrial particles from cells
grown under control (141 Torrr O
2
) and hypoxia (1 Torr O
2
) for 10 h was carried out using [a-
32
P]UTP as described in Materials and methods. Rate
of
32
P incorporation in the RNA fraction was determined as described before [21]. RNA isolated from in vitro incubated mitochondria (5 lgeach)
was hybridized to CytOX subunit I and II encoding DNA from the mouse mitochondrial genome by slot blot hybridization. The blot was
quantified in a Bio-Rad GS-525 Molecular Imager. Rates of
32
P incorporation by mitochondria from these cells (A), transcription rates as in PC12
cells (B) and macrophages (C) were shown. (D) shows the levels of mtTFA in PC12 cells and macrophages grown under normoxia and hypoxia by
immunoblot analysis. Immunoblot analysis was carried out as described in Materials and methods using 30 lg mitochondrial protein in each case.
Ó FEBS 2003 Cytochrome c oxidase subunit expression in hypoxia (Eur. J. Biochem. 270) 875
Oxygen is essential for the biosynthesis of heme and
hence heme is considered, an oxygen sensor [31]. Addition-
ally, our previous study [11] indicated that heme not only
regulates the catalytic activity of the CytOX complex but
may also affect its stability. Based on these observations, we
determined heme aa
3
levels as well as the catalytic activity of

3
moieties are associated with the
mitochondrial genome encoded CytOX subunit I [4,5].
The effects of hypoxic inhibition of transcription on the
subunit contents of the complex were assessed by BN/
PAGE, which allows the separation of large oligomeric
complexes based predominantly on size. Equal amounts of
SMP protein from control and hypoxia exposed macro-
phages were resolved on BN/PAGE and transferred to
PVDF membrane. The enzyme resolved as two major
complexes, which comigrated with Apo-ferritin and
b-amylase. Based on the rates of migration, the slow
migrating complex may be a dimmer and the faster
migrating complex migrating with an apparent molecular
mass of 200 kDa may be the monomeric form. It is
interesting that the levels of mitochondrial encoded
CytOX I and nuclear encoded CytOX IV and Vb in both
complexes were reduced, although we observed a more
pronounced reduction in the putative dimeric form, which
is thought to be the more active form (Fig. 3A). It is also
seen that the level of ATPase complex as determined by
immunoblotting with antibody to the F
1
ATPase did not
change under these conditions (Fig. 3B). Quantification of
the blots shows that the levels of CytOX subunits I, IV
and Vb were reduced by 50–75% in the two complexes
combined as compared to cells grown under normoxia (see
Fig. 3C). A nearly 50% reduction in enzyme activity
(Table 1) under hypoxia seems to accompany a change in

Macrophages PC12 Cells
Normoxia Hypoxia Normoxia Hypoxia
CytOX activity (nmolÆmin
)1
Æmg
)1
SMP) 2429 1160 2633 2166
Heme aa3 content (nmolÆmg
)1
SMP
)1
) 0.225 0.103 0.179 0.112
TN (nmolÆnmol heme aa
À1
3
Æs
)1
) 180 187 245 314
Fig. 3. Hypoxia induced changes in CytOX complex. SMP (30 lg
protein) from standard (141 Torr O
2
)orhypoxia(1TorrO
2
)exposed
cells were solubilized by treatment with 1% digitonin and analyzed by
BN/PAGE on 6–13% acrylamide gels as described in Materials and
methods. Proteins were transblotted to poly(vinylidene difluoride)
membrane and probed with subunit specific monoclonal antibodies
and HRP-conjugated anti-(mouse IgG) Ig (A). Stripped blots were
also probed with antibody to F

Studies in yeast have demonstrated that oxygen acts as a
molecular switch and alters the expression of the two
nuclear genome coded isoforms of CytOX V [10,32]. The
regulation of genes CytOX5a and CytOX5b, coding for the
two isofoms Va and Vb, parallels that of genes CYC1 and
CYC7, which encode iso-1 and iso-2 of yeast cytochrome c,
respectively. CytOX 5a and CYC1 are coexpressed under
aerobic conditions (O
2
>0.5l
M
), whereas CytOX 5b and
CYC7 are co-expressed under hypoxic (O
2
<0.5l
M
)and
heme deficient conditions [11]. The coexpression of specific
subunit V and cytochrome c isoforms indicates that these
isoform pairs act synergistically to regulate electron transfer
rates in enzyme function. These variant subunit isoforms
have been shown to affect the turnover rate (TN) of the
holoenzyme markedly by altering the rates of intramole-
cular electron transfer between heme a and the binuclear
reaction center. Thus the yeast, CytOX V, functions as an
oxygen/heme sensor [32].
This investigation was undertaken, to determine the effect
of hypoxia on (a) CytOX gene expression and (b) CytOX
activity. The objective was to determine if oxygen/heme
dependent regulation of mammalian CytOX genes is similar

transcription rates (Fig. 1 and Table 2). This is supported
by the observations of Schumacker et al. who have noted an
inhibition of cellular respiration and suppression of ATP
utilization during hypoxia [37–40]. The mammalian mito-
chondrial RNA polymerase requires a high concentration
of ATP (0.5–1 m
M
) for maximal activity. Narasimhan
and Attardi [41] showed that a high concentration of
5¢-adenylylimidodiphosphate was able to stimulate the
Table 2. Effect of hypoxia on some biochemical parameters related to mitochondrial function. ATP levels in total cell extracts were measured using the
somatic cell ATP assay kit, which is based on the assay of ATP driven luciferin luciferase assay system. ATP levels were measured in a TD-20/20
luminometer. Respiration coupled ATP synthesis by isolated mitochondria was measured by incubating mitochondria in a medium supplemented
with ADP and succinate as described in the Materials and methods section. Hexokinase and phosphofructokinase activities were measured in the
cytosolic fractions. Isocitrate dehydrogenase and NADH:ubiqinone oxidoreductase (complex I) activities were measured in isolated mitochondria
by standard methods as indicated in the Materials and methods section. Values are given as means ± SD calculated from four estimates.
Macrophages PC12 Cells
Normoxia Hypoxia Normoxia Hypoxia
Total cellular ATP (nmolÆmg protein
)1
) 10.3 ± 1.132 7.04 ± 0.774 22 ± 1.986 11 ± 1.431
Respiration coupled ATP synthesis in isolated
mitochondria (nmolÆmg protein
)1
)
36 ± 3.24 28 ± 5.7 84 ± 9.52 37 ± 5.12
Enzyme activity
Isocitrate dehydrogenase (lmolÆmin
)1
Æmg protein

Similar to that reported for chemical hypoxia with
CoCl
2
and succinyl acetone [12], physiologically relevant
hypoxia in cultured cells also leads to rapid depletion of
heme aa
3
pools (Table 1). The observed heme depletion is
closely associated with lowered CytOX subunits I, IV and
Vb content and altered enzyme activity. Notably the
reduced subunit level is more apparent in the slow
migrating putative dimeric form of the enzyme (Fig. 3),
suggesting that reduced heme and altered subunit levels
may interfere with the formation of the more active
dimeric complex. As heme is involved in reactions that
transfer electrons from cytochrome c to molecular oxygen,
its depletion reflects alterations in the catalytic efficiency
of the enzyme complex as assessed by the TN for
cytochrome c oxidation or oxygen utilization (Table 1).
Even under hypoxia induced heme depletion and reduced
enzyme content the TN of the CytOX complex for
cytochrome c oxidation essentially remained unaltered in
macrophages, while the TN was slightly enhanced in
PC12 cells (Table 1). This is in sharp contrast to twofold
to fourfold higher TN rates for cytochrome c oxidation,
which we reported for the enzyme from heme-depleted
tissues in CoCl
2
treated animals [12]. Although the
mechanism of action of CoCl

decrease in TN of the enzyme might be a late event in
intact cells exposed to hypoxia. It is more likely that these
differences in TN represent the different stages of adaptive
response to hypoxia.
Based on the results of our study and those of
Schumacker, it is reasonable to conclude that the effects
of hypoxia on CytOX gene expression and its activity are
secondary to suppression of respiration during hypoxia. In
the absence of differential regulation of a specific nuclear
gene coding for the subunits of CytOX, changes in the
microenvironment of the cell may induce alterations in the
catalytic efficiency of mammalian enzyme. Selective loss of
CytOX subunits I/II and IV might be important factors in
altered catalytic activity [12]. Alternatively, subunit phos-
phorylation as suggested in studies by Kadenbach [43,44]
and others, including our own [12] might be involved in
altered catalytic function of the enzyme complex. In
summary, altered respiration and oxygen-regulated alter-
ations in ATP and heme pools might have a direct effect on
the activity of the complex.
Acknowledgements
We are thankful to members of the Avadhani lab for useful discussions
and comments during the course of this work. We also thank Dr David
Clayton for providing antibody to mouse mtTFA. This research was
supported in part by National Institute of Health (USA) grant GM-
49683.
References
1. Taanman, J.W. (1997) Human cytochrome c oxidase: structure,
function, and deficiency. J. Bioenerg. Biomemb. 29, 151–163.
2. Capaldi, R.A. (1990) Structure and function of cytochrome c

regulated isoforms in cytochrome c oxidase. J. Exp. Biol. 201,
1163–1175.
12. Vijayasarathy, C., Damle, S., Lenka, N. & Avadhani, N.G. (1999)
Tissue variant effects of heme inhibitors on the mouse cytochrome
c oxidase gene expression and catalytic activity of the enzyme
complex. Eur. J. Biochem. 266, 191–200.
13. Otto, C.M. & Baumgardner, J.E. (2001) Effect of culture PO2 on
macrophage (RAW 264.7) nitric oxide production. Am.J.Physiol.
280, C280–C287.
878 C. Vijayasarathy et al.(Eur. J. Biochem. 270) Ó FEBS 2003
14. Lowry, O.H., Rosebrough, N.J., Farr, A.L. & Randall, R.J.
(1951) Protein measurement with the Folin Phenol reagent. J. Biol.
Chem. 193, 265–275.
15. Chomczynski, P. & Sacchi, N. (1987) Single-step method of RNA
isolation by acid guanidinium thiocyanate-phenol-chloroform
extraction. Anal. Biochem. 162, 156–159.
16. Wellauer, P.K. & Dawid, I.B. (1979) Isolation and sequence
organization of human ribosomal DNA. J. Mol. Biol. 128, 289–303.
17. Bhat, N.K., Niranjan, B.G. & Avadhani, N.G. (1982) Qualitative
and comparative nature of mitochondrial translation products in
mammalian cells. Biochemistry 21, 2452–2460.
18. Pedersen, P.L., Greenawalt, J.W., Reynafarje, B., Hullihen, J.,
Decker, G.L., Soper, J.W. & Bustamente, E. (1978) Preparation
and characterization of mitochondria and submitochondrial par-
ticles of rat liver and liver-derived tissues. Methods Cell. Biol. 20,
411–481.
19. Kantharaj, G.R., Bhat, K.S. & Avadhani, N.G. (1983) Mode
of transcription and maturation of ribosomal ribonucleic acid
in vitro in mitochondria from Ehrlich ascites cells. Biochemistry
22, 3151–3156.

16086.
30. Larsson, N.G., Wang, J., Wilhelmsson, H., Oldfors, A., Rustin,
P., Lewandoski, M., Barsh, G.S. & Clayton, D.A. (1994)
Mitochondrial transcription factor A is necessary for mtDNA
maintenance and embryogenesis in mice. Nat. Genet. 18, 231–236.
31. Kwast, K.E., Burke, P.V. & Poyton, R.O. (1998) Oxygen sensing
and the transcriptional regulation of oxygen-responsive genes in
yeast. J. Exp. Biol. 201, 1177–1195.
32. Poyton, R.O. (1999) Models for oxygen sensing in yeast: impli-
cations for oxygen-regulated gene expression in higher eucaryotes.
Resp. Physiol. 115, 119–133.
33. Chinenov, Y., Schmidt, T., Yang, X.Y. & Martin, M.E. (1998)
Identification of redox-sensitive cysteines in GA-binding protein-
alpha that regulate DNA binding and heterodimerization. J. Biol.
Chem. 273, 6203–6209.
34. Kwong, M., Kan, Y.W. & Chan, J.Y. (1999) The CNC basic
leucine zipper factor, Nrf1, is essential for cell survival in response
to oxidative stress-inducing agents. Role for Nrf1 in gamma-gcs
(1) and gss expression in mouse fibroblasts. J. Biol. Chem. 274,
37491–37498.
35. Marin-Garcia, J., Ananthakrishnan, R., Agrawal, N. & Gold-
enthal, M.J. (1994) Mitochondrial gene expression during bovine
cardiac growth and development. J. Mol. Cell Cardiol. 26, 1029–
1036.
36. Wiesner, R.J., Aschenbrenner, V., Ruegg, J.C. & Zak, R. (1994)
Coordination of nuclear and mitochondrial gene expression dur-
ing the development of cardiac hypertrophy in rats. Am.J.Phy-
siol. 267, C229–C235.
37. Chandel, N., Budinger, G.R., Kemp, R.A. & Schumacker, P.T.
(1995) Inhibition of cytochrome-c oxidase activity during pro-