BB
ELSEVIER
Biochimica et Biophysica Acta 1271 (1995) 29-33
Biochi
f ic~a A~ta
etBiophysica
Mitochondrial dysfunction during anoxia and acute cell injury
Dean P. J o n e s *
Department of Biochemisto', Emory Unicersity, Atlanta, GA 30322, USA
Abstract
Mitochondrial function is closely linked to the maintenance of mitochondrial integrity. During short-term anoxia, ion-transport systems
in the inner membrane are inhibited to protect against loss of the promotive force and associated osmotic imbalance that can cause
irreversible loss of mitochondrial integrity and function. In two models of chemically induced mitochondrial failure, a prostaglandin B i
derivative, di-calciphor, protected against mitochondrial failure and prevented cell death. Characteristics were similar to those observed in
mitochondria during short-term anoxia. Thus, the results indicate that di-calciphor may represent a new type of mitochondrial protectant
that inhibits ion transport and thus slows the loss of osmotic stability and delays mitochondrial dysfunction under traumatic and
toxicologic conditions.
Keywords: Cell injury; Mitochondrial dysfunction; Anoxia; Di-calciphor
1. Introduction
Since the general acceptance of chemiosmotic coupling
as the central mechanism of oxidative phosphorylation
about 20 years ago [ 1,2], investigations have shifted from a
primary focus on mitochondrial function to a focus on
mitochondrial dysfunction. The present report provides a
summary of our studies which address the interface of
function and dysfunction, namely, conditions in which
oxygen deficiency shifts mitochondrial function to dysfunction.
Oxygen deficiency is commonly associated with many
pathologic processes. As a result, the consequences of
impaired mitochondrial function and the mechanisms of
mitochondrial protection during oxygen deficiency are of
great importance both to the understanding of cellular
homeostasis and to developing improved means to protect
against mitochondrial failure. This is most relevant to
individuals with mitochondrial genetic defects or genomic
deletions associated with aging because they have compromised mitochondria that are more susceptible to failure.
The current presentation provides a summary of studies
which show that oxygen deficiency substantially increases
cellular susceptibility to oxidative injury and describes
* Corresponding author. E-mail: [email protected]. Fax: + 1 ( 4 0 4 )
7273231.
0925-4439/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved
SSDI 0 9 2 5 - 4 4 3 9 ( 9 5 ) 0 0 0 0 6 - 2
mechanisms that exist in mammalian cells to protect against
mitochondrial failure during short-term anoxia.
2. Results and discussion
2.1. Oxygen deficiency increases susceptibiliO' to acute
cell inju~ .from oxidants
In pioneering studies more than a decade ago, Granger
and associates [3] showed that antioxidants could protect
against tissue injury during post-ischemic reoxygenation.
In the intervening period, it has become apparent that
reactive oxygen species are commonly generated during
this period and can result in cell injury [4]. In addition,
several studies have indicated that the mitochondria are a
primary target of oxidative injury. For instance, in rat
hepatocytes, treatment with 3-hydroxy-4-pentenoate results
in selective depletion of mitochondrial glutathione [5] and
greatly potentiates toxicity due to t-butylhydroperoxide
(Fig. 1). Thus, we addressed whether conditions that limit
mitochondrial function, namely, states of oxygen deficiency, affect the susceptibility of cells to oxidative injury.
We have examined this question with three different
experimental models. In the first, we addressed whether
cells exposed acutely to oxygen deficiency were more
vulnerable to oxidative injury than normoxic cells [6]. The
results showed that substantially more cell killing occurred
in hepatocytes incubated under steady-state hypoxic condi-
D.P. Jones/Biochimica et Biophysica Acta 1271 (1995) 29-33
30
100
able to oxidative injury. Thus, conditions which could be
tolerated under normoxia can become pathologic following
transient interruption of blood flow.
We studied the effects of chronic in vivo oxygen deficiency on sensitivity to oxidative injury in a third model
with rat hepatocytes [ 10]. Rats were maintained for 8 to 10
days under conditions of normoxia or a moderately severe
hypoxia (10.5% 02 under normobaric conditions), hepatocytes were isolated, and experiments were performed in
vitro. Cells from animals that were previously hypoxic
were comparable in viability and several other characteristics but were considerably more sensitive to toxicity
induced by t-butylhydroperoxide (Fig. 2). Thus, with three
different models of oxygen deficiency, cells are more
susceptible to injury from oxidants.
Because mitochondria are a sensitive target of injury
due to oxidants and are functionally impaired by oxygen
deficiency, these results suggest that hypoxic episodes may
be particularly important contributors to mitochondrial
dysfunction. Mitochondrial genomic damage associated
with age-related diseases in humans is thought to occur by
oxidative mechanisms [11]. Our results show that impaired
circulatory and respiratory supply of 0 2 to tissue may
substantially exacerbate the oxidative damage and therefore be an important contributing factor to age-related
disease.
80
60
.c~
40
20
0
0
I
I
I
20
40
60
Time (rain)
Fig. 1. Effect of depleting mitochondrial GSH on the susceptibility to
oxidative injury. Rat hepatocytes were incubated with 3-hydroxy-4pentenoate to selectively deplete mitochondrial GSH and then treated
with a concentration of t-butylhydroperoxide (0.3 mM) that did not kill
control cells. Measurements of Trypan blue exclusion showed that the
oxidant did not kill the control cells ( © ) but killed the cells in which
mitochondrial GSH was depleted ( 0 ) . 3-Hydroxy-4-pentenoate alone
was not toxic and selective depletion of cytoplasmic GSH did not
potentiate toxicity (data from [5]).
tions (Fig. 2). Studies of the mechanistic basis of this
effect showed that NADPH supply for reduction of peroxides by the GSH-dependent pathways was limited [7], as
was the rate of GSH synthesis [8].
In a second model, we examined whether hepatocytes
previously exposed to anoxia were more vulnerable to
oxidative injury than cells that had not been exposed to
anoxia [9]. The results showed that a prior exposure to
30-60 min of anoxia dramatically increased the sensitivity
of cells to a constant oxidative challenge (Fig. 2). This
finding is particularly important because it demonstrates
that even without an enhanced generation of oxidants
during post-ischemic reoxygenation, cells are more vulner-
2.2. Protection against mitochondrial dysfunction during
short-term anoxia
The period of time that mammalian cells can tolerate
anoxia is highly variable [12], depending upon cell type
and conditions such as functional load and nutritional
status. Of particular interest, cell death is substantially
delayed from the time required for redox changes in the
mitochondrial electron transport chain and bioenergetic
changes as reflected in the cellular ATP pool (Fig. 3).
100
~
so
80 ~
60
8o
~6o
40
40
0
20
N 20
20
I
0
60
121
Time (min)
0
-60
~
0
~
60
Time (min)
0
120
0 20 40 60 80
Time (min)
Fig. 2. Oxygen deficiency increases the susceptibility of cells to oxidative injury. In panel A, rat hepatocytes were incubated under a physiological oxygen
concentration (20/zM; triangles) or anoxic conditions (closed circles) with 0.6 mM t-butylhydroperoxide. The results showed that anoxic cells were more
susceptible to injury as measured by trypan blue exclusion. Untreated anoxic cells were not killed under these conditions (open circles). Data from [6]. In
panel B, cells were incubated for 1 h under aerobic (open symbols) or anaerobic (closed symbols) conditions and then reoxygenated. Following treatment
with 0.3 m M t-butylhydroperoxide (circles), extensive cell death occurred in the cells that were previously anoxic while little toxicity was observed in the
cells that were continuously aerobic (triangles). Data from [9]. In panel C, toxicity of 0.3 mM t-butylhydroperoxide (circles) in cells from normoxic rats
(open symbols) was compared to that in cells from animals that had been maintained for 8 - 1 0 days in an environment with 10.5% oxygen (closed
symbols). The results show that in vivo hypoxia increases the sensitivity of cells to oxidative injury. Data from [10].
D.P. Jones / Biochimica et Biophysica Acta 1271 (1995) 29-33
Several mechanisms exist in cells to protect against cell
death during anoxia, such as enhanced ATP production by
glycolysis, inhibition of phospholipases by low pH, and
inhibition of ion transporters and channels in the plasma
membrane. Our research has provided evidence for an
additional mechanism that functions during short-term
anoxia to protect against mitochondrial swelling and lysis
[13-15]. This mechanism allows preservation of the structural and functional integrity of mitochondria in the absence of energy supply from electron transport and extends
the time of O 2 deficiency that can be tolerated without
irreversible injury.
The proposed mechanism is based upon data acquired
from analyses of rat hepatocytes under anoxic conditions
and interpreted in terms of a coordinated multisite regulation model [16] for oxidative phosphorylation (Fig. 4). In
this model, 5 principal control sites are recognized. These
include the control of flow of electrons into the mitochondrial chain that is regulated by NAD+-linked dehydrogenases and the flow of electrons to O 2 at cytochrome
oxidase. Coordinated regulation at these two sites provides
control over the energy available for mitochondrial functions dependent upon the mitochondrial membrane potential (A~0) and pH gradient (ApH). Utilization of A~O and
ApH for ATP supply to the cytoplasm occurs by three
systems, the phosphate transporter (H 2 P O / / O H antiport
system), the ATP synthase and the adenine nucleotide
transporter (ANT). Regulation at these sites allows control
of energy utilization for ATP synthesis independently of
other mitochondrial energy utilization, such as for Ca 2+
cycling [17]. Coordinated regulation is necessary because
phosphate uptake in the absence of ATP synthesis or
export would rapidly result in mitochondrial swelling and
lysis. During anoxia, this type of regulation allows for
inhibition of utilization of A~0 and ApH for ATP production. This preserves both A~ and ApH so that osmotic
100'
m
.'_
,~,
80'
L)
60
t~
.<
40 i
~-
20
0
i
0
60
i
120
180
Time (mln)
I
240
Fig. 3. Time-course of effects of anoxia on pyridine nucleotide redox
state (O), ATP (C)) and cell viability (z~) in rat hepatocytes. Data from
[19] and [24].
Regulation of
generation of
Ap by control
of electron
flow
31
+c,l+oxl+,
NADlinked
DH's
~I~
NADPH
Ap
/
Regulation of
use of Ap for [
cytosolic ATP
Sc;77, ,on
+
:
cycling,
otherenergyneeds
"N~
ATe ANT fl
synthase
supply, C a+2
PiT
transport
Cytosolic ATP supply
Fig. 4. Coordinated multisite control model for mitochondrial function.
Generation of the proton-motive force (Ap) is controlled by coordinate
regulation of electron flow into the mitochondrial electron transport chain
at the NAD+-linked dehydrogenases and out of the chain at cytochrome
oxidase. Utilization of Ap for ATP supply to the cytoplasm is controlled
by coordinate regulation of the ATP synthase, adenine nucleotide translocator (ANT) and the phosphate transporter (PIT). This allows ATP supply
to the cytoplasm to be relatively independent of the other uses of Ap in
the mitochondria. This coordinated multisite regulation also provides a
mechanism to preserve osmotic balance and prevent swelling in the
mitochondria under anoxic conditions because simultaneous inhibition of
the ATP synthase, ANT and phosphate transporter substantially decreases
the rate of dissipation of 3p.
stability can be maintained for an extended time even
though cytoplasmic ATP drops precipitously [13].
2.3. Effect of KCN on mitochondrial function and cell
viabili O,
Isolated mitochondria exposed to anoxia have a very
weak protective response, suggesting that the mechanisms
protecting against mitochondrial failure may involve a
diffusible species. Such a species could be generated by
subunits of cytochrome oxidase that have unknown functions [18]. Indeed, cytochrome oxidase would provide an
ideal sensor for oxygen deficiency because it normally
exists with cytochrome a 3 in the ferric form which becomes reduced to the ferrous form during hypoxia. A shift
in redox state (from < I% to > 99% ferrous) with an
associated conformational change could provide a very
sensitive means to signal O 2 deficiency.
To test whether inhibition of mitochondrial function
under conditions where cytochrome a 3 was trapped in the
ferric form would interfere with mitochondrial protection.
we examined the effects of KCN on mitochondrial functions and cell viability [19]. The results showed that cytosolic ATP fell with the same kinetics as seen for anoxia
(Fig. 5). In contrast, control of mitochondrial ion transport
and osmotic regulation was lost in the presence of cyanide,
and cyanide-treated cells died much more rapidly than did
anoxic cells ([19]; Fig. 5). Thus, in the presence of KCN,
mitochondria fail and cells die much more rapidly than
they do when they are exposed to anoxia.
32
D.P. Jones / Biochimica et Biophysica Acta 1271 (1995) 29-33
~,100
g
20
.~ 80
._~
ao
"~ 10
j 4o
[ 20
0
10
20
30
~
0
0
i
i
i
1
2
3
Time (h)
Time (rain)
Fig. 5. Comparison of the effects of KCN and anoxia on ATP depletion
(left panel) and viability (right panel) in hepatocytes. Time-course of
effects of 0.5 mM KCN (O) and anoxia (O) on ATP was measured by
HPLC and effects on viability were measured by Trypan blue exclusion.
Data from [19].
2.4. Protection against mitochondrial .failure by di-calciphor
Using this model for mitochondrial failure induced by
KCN, we examined the effects of a prostaglandin derivative (di-calciphor) that had previously been reported to
protect mitochondria [20]. Addition of di-calciphor to
KCN-treated hepatocytes did not alter the kinetics or extent of ATP loss but protected against cell death ([21]; Fig.
6). Di-calciphor protected against mitochondrial loss of
A~b, loss of ApH, loading of phosphate and swelling (Fig.
6). Thus, the results indicate that di-calciphor may protect
against cell death by protecting against mitochondrial failure. The characteristics of protection by di-calciphor appear to be identical to the endogenous mechanisms that
protect during short-term anoxia, indicating that a more
detailed understanding of this mechanism may provide the
basis for developing therapeutic means to protect against
mitochondrial dysfunction and disease.
However, because KCN is a relatively non-specific
poison and can affect systems other than mitochondria, we
felt that it was necessary to determine whether di-calciphor
could protect against mitochondrial failure induced by a
very specific mitochondrial poison. Consequently, we examined the effects of di-calciphor on cell death induced by
antimycin A [22]. As found with KCN-treated hepatocytes,
di-calciphor protected against cell death but did not alter
the loss of ATP. Mitochondrial A~b and ApH were preserved by di-calciphor, and phosphate uptake and swelling
were prevented. Thus, di-calciphor appears to protect cells
specifically by preventing mitochondrial failure.
In ongoing studies, we have recognized that di-calciphor exists as multiple positional and stereoisomers, only
some of which are functional in providing protection
against mitochondrial failure [23]. Because of this heterogeneity, future studies will be needed to establish the
precise stereochemistry of the active isomers as a prelude
to defining the specific biomolecules through which dicalciphor exerts its remarkable biologic activity.
,~ 30
100
90
~0 2o
80
"~ 70
..~
.2
O
~ 10
e~
60
50
~150
i
0
i
i
1 2
Time (h)
i
3
"< 0
C DC KCN KCN
+DC
C DC KCN KCN
+DC
~
100
Q
50
o
o~
C DC KCN KCN
C DC KCNKCN
C DCKCN KCN
+DC
+DC
+DC
Fig. 6. Effects of di-calciphor on the KCN-induced changes in hepatocyte viability, ATP loss, pH gradient, mitochondrial membrane potential,
mitochondrial phosphate content (nmol/106 cells) and mitochondrial volume (pA/106 cells): (O) cells with 0.5 mM KCN; (O) cells with 0.5 mM KCN
and 10 /~M di-calciphor; and (A) cells with di-calciphor alone. Other panels were from studies done at 30 rain with l0 /~M di-calciphor and 0.5 mM
KCN. Data from [21].
D.P. Jones / Biochimiea et Biophysica Aeta 1271 (1995) 29-33
3. Summary
In summary, our studies have shown that mitochondria
contain an endogenous protective mechanism that protects
against mitochondrial failure during short-term anoxia. The
prostaglandin B~ derivative, di-calciphor, appears to activate this protective mechanism and delay irreversible
swelling and lysis by inhibiting ATP synthase and other
kinetically active ion transport systems. The mechanism
thus preserves mitochondrial integrity and allows recovery
of mitochondrial function even following extensive decline
in cellular energetics as reflected in the cytoplasmic ATP
pool.
References
[1] Boyer, P.D., Chance, B., Ernster, L., Mitchell, P. Racker, E. and
Slater, E.C. (1977) Annu. Rev. Biochem. 46, 955-1026.
[2] Mitchell, P. (1979) Science 206, 1148-1159.
[3] Granger, D.N., Rutili, G. and McCord, J. (1981) Gastroenterology
81, 22-29.
[4] Rao, P.S., Cohen, M.V. and Mueller, H.S. (1983) J. Mol. Cell.
Cardiol. 15. 713-716.
[5] Shah, X., Jones, D.P., Hashmi, M. and Anders, M.W. (1993) Chem.
Res. Toxicol. 6, 75-81.
[6] Tribble. D.L., Jones, D.P. and Edmondson, D.E. (1988) Mol. Pharmacol. 34, 413-420.
[7] Tribble, D.L. and Jones, D.P. (1990) Biochem. Pharmacol. 39,
729-737.
33
[8] Shan, X., Aw, T.Y, Shapira, R. and Jones. D.P. (1989) Toxicol.
Appl. Pharmacol. 101,261-270.
[9] Kowalski, D.P., Aw, T.Y., Park, Y. and Jones, D.P. (1992) Free
Radicals Biol. Med. 12, 205-212.
[10] Shah, X., Aw, T,Y., Smith, E.R., Ingleman-Sundberg, M., Mannervik, B., Iyanagi, T. and Jones, D.P. (1992) Biochem. Pharmacol.
43, 2421-2426.
[I 1] Wallace, D.C. (1992) Annu. Rev. Biochem. 61, 1175-1212.
[12] Tenney, S.M. and Lamb, T.W. (1965) in Handbook of Physiology,
Section 3: Respiration, Vol. 2 (Fenn, W.O. and Rahn, H., eds.), pp.
979-1010, American Physiological Society, Washington, DC.
[13] Andersson, B.S.. Aw, T.Y. and Jones. D.P. (1987) Am. J. Physiol.
252, C349-C355.
[14] Aw, T.Y., Andersson. B.S. and Jones D.P. (1987) Am. J. Physiol.
252. C356-C361.
[15] Aw, T.Y., Andersson, B.S. and Jones. D.P. (1987) Am. J. Physiol.
252, C362-C368.
[16] Jones, D.P., Shan. X. and Park, Y. (1992) Annu. Rev. Nutr. 12,
327-343.
[17] Lee, C.P., Sciamanna, M. and Peterson, RL. (1993) Methods Toxicol. 2.41-50.
[18] Anthony, G., Reimann, A, and Kadenbach, B. (1993) Proc. Natl.
Acad. Sci. USA 90, 1652-1656.
[19] Aw, T.Y. and Jones, D.P. (1989) Am. J. Physiol., 257, C435-C441.
[20] Devlin, T.M., Krupinski-Olsen, R., Uribe, S. and Nelson, G.E.
(1986) Biochem. Biophys. Res. Commun. 137, 215-221.
[21] Park, Y., Devlin, T.M. and Jones, D.P. (1992) Am. J. Physiol. 263,
C405-C411.
[22] Park, Y., Devlin, T.M. and Jones~ D.P, (1994) Tox. Appl. Pharmacol. 126, 33-38.
[23] Jones, D.P. (1995) in Cell Biology of Trauma, (Lemasters, J.J. and
Oliver, C.N., Eds,), CRC Press, Boca Raton, FL, in press.
[24] Jones, D.P. and Kennedy, F.G. (1982) Am. J. Physiol. 243, C247
C253.