Diversity in Mitochondrial Function Explains Differences in
Vascular Oxygen Sensing
Evangelos D. Michelakis, Vaclav Hampl, Ali Nsair, XiCheng Wu, Gwyneth Harry, Al Haromy,
Rachita Gurtu, Stephen L. Archer
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Abstract—Renal arteries (RAs) dilate in response to hypoxia, whereas the pulmonary arteries (PAs) constrict. In the PA,
O2 tension is detected by an unidentified redox sensor, which controls K⫹ channel function and thus smooth muscle cell
(SMC) membrane potential and cytosolic calcium. Mitochondria are important regulators of cellular redox status and
are candidate vascular O2 sensors. Mitochondria-derived activated oxygen species (AOS), like H2O2, can diffuse to the
cytoplasm and cause vasodilatation by activating sarcolemmal K⫹ channels. We hypothesize that mitochondrial
diversity between vascular beds explains the opposing responses to hypoxia in PAs versus RAs. The effects of hypoxia
and proximal electron transport chain (pETC) inhibitors (rotenone and antimycin A) were compared in rat isolated
arteries, vascular SMCs, and perfused organs. Hypoxia and pETC inhibitors decrease production of AOS and outward
K⫹ current and constrict PAs while increasing AOS production and outward K⫹ current and dilating RAs. At baseline,
lung mitochondria have lower respiratory rates and higher rates of AOS and H2O2 production. Similarly, production of
AOS and H2O2 is greater in PA versus RA rings. SMC mitochondrial membrane potential is more depolarized in PAs
versus RAs. These differences relate in part to the lower expression of proximal ETC components and greater expression
of mitochondrial manganese superoxide dismutase in PAs versus RAs. Differential regulation of a tonically produced,
mitochondria-derived, vasodilating factor, possibly H2O2, can explain the opposing effects of hypoxia on the PAs versus
RAs. We conclude that the PA and RA have different mitochondria. (Circ Res. 2002;90:●●●-●●●.)
Key Words: rotenone 䡲 K⫹ channels 䡲 redox 䡲 pulmonary circulation 䡲 oxygen sensor
ing hydrogen peroxide (H2O2), are important K⫹ channel
openers.3 K⫹ channels contribute to regulation of vascular
tone through their control of SMC membrane potential (Em).
Closing of K⫹ channels depolarizes Em, which causes an
increase in the open probability of the voltage-gated L-type
Ca2⫹ channels, Ca2⫹ influx, and vasoconstriction. In contrast,
K ⫹ channel openers cause hyperpolarization and
vasodilatation.4
The major source of AOS and peroxides in SMCs are
cytochrome-based oxidases, such as nicotinamide adenine
dinucleotide phosphate (NADPH) and NADH, and the electron transport chain (ETC) in mitochondria.2 Both vascular
oxidases and the ETC produce AOS in proportion to PO2, and
thus have been proposed as candidates for vascular oxygen
sensors.5,6 Despite the low KM for O2 of the mitochondrial
cytochromes, lung mitochondria make AOS in direct proportion to PO2 over the physiological range.7
Most O2-sensitive systems consist of a sensor that produces
a mediator in response to changes in PO2, which in turn alters
the function of an effector. O2- and redox-sensitive K⫹
channels have been shown to be the effectors in O2 sensing in
several tissues including the PA, the ductus arteriosus, the
T
he normoxic pulmonary circulation is vasodilated and
accommodates the entire cardiac output at much lower
pressures than the systemic circulation. During hypoxia, the
pulmonary arteries (PAs) constrict (hypoxic pulmonary vasoconstriction, HPV), whereas systemic arteries, such as renal
arteries (RAs), dilate. The mechanism of this opposing
control of tone between the two vascular beds is unknown.
Although the response of each bed to hypoxia is significantly
modulated by the endothelium, the mechanism for the opposing responses to hypoxia appear to lie within the vascular
smooth muscle cells (SMCs). Hypoxia increases intracellular
Ca2⫹ ([Ca2⫹]i) and contracts PASMCs; in contrast, isolated
SMCs from systemic arteries display decreased [Ca2⫹]i and
relax in response to hypoxia.1
Both the control of tone and the response of O2-sensitive
tissues to hypoxia involve redox-sensitive mechanisms (for
review, see Wolin2). Activated oxygen species (AOS) are
now recognized as important mediators in vascular cellular
signaling. Several kinases and sarcolemmal potassium channels (K⫹ channels) are redox-sensitive and modulated by
AOS. Cysteine-rich K⫹ channels are inhibited when reduced
and activated when oxidized, and oxidants and AOS, includ-
Original received October 30, 2000; resubmission received December 18, 2001; revised resubmission received May 23, 2002; accepted May 23, 2002.
From the Department of Medicine (Cardiology) and the Vascular Biology Group (E.D.M., A.N., X.W., G.H., A.L., R.G., S.L.A.), University of Alberta,
Edmonton, Alberta, Canada; and Charles University, Physiology (V.H.), Prague, Czech Republic.
Correspondence to Evangelos D. Michelakis, MD, Cardiology, University of Alberta, 2C2.36 Walter Mackenzie Health Sciences Centre, Edmonton,
AB, Canada T6G 2B7. E-mail [email protected]
© 2002 American Heart Association, Inc.
Circulation Research is available at http://www.circresaha.org
DOI: 10.1161/01.RES.0000024689.07590.C2
1
2
Circulation Research
June 28, 2002
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
carotid body, the neuroepithelial body, and the fetal adrenomedullary cells.4 Inhibition of these O2-sensitive K⫹ channels
leads to depolarization, opening of L-type Ca2⫹ channels,
Ca2⫹ influx, and vasoconstriction. The obligatory role of the
L-type Ca2⫹ channel is suggested by the fact that blockers and
agonists of this channel inhibit8 and enhance9 HPV, respectively. On the other hand there are reports suggesting that it is
release of intracellular Ca2⫹, rather than influx of extracellular
Ca2⫹, that initiates HPV.10 This controversy may relate to
confusion regarding the pool of Ca2⫹ that initiates rather than
sustains HPV and perhaps to the model used to study HPV
(rings versus intact lungs).
There is growing consensus that the vascular O2 sensors are
redox based. A variety of sensors have been proposed,
including NADPH oxidase,11 NADH oxidase,12 and the
ETC.6 It is likely that there is a diversity of sensors among
different O2-sensitive tissues and species. The role of
gp91phox-containing NADPH oxidase in O2 sensing was
recently challenged, at least in the lung13 and the carotid
body,14 although the role of novel oxidases (NOX) in HPV
has not been assessed. The role of mitochondria as O2 sensors
in the pulmonary circulation was proposed 15 years ago5 and
is supported by several recent publications.6,15,16 Their role in
O2 sensing appears to be widespread, because ETC inhibitors
mimic hypoxia in several other O2-sensing organs, such as the
carotid body17 and adrenal medullary cells.18 The acute
hemodynamic effects of hypoxia and ETC inhibitors, unlike
anoxia, do not relate to ATP depletion.19 Rather, we speculate
that mitochondria alter vascular tone by regulating the production of diffusible redox mediators, such as AOS and
peroxides. H2O2, with its long diffusion radius, can modify
targets in the cytoplasm and membrane (guanylate cyclase
and K⫹ channels), and thereby regulate tone.2 In addition,
mitochondria are now recognized as important regulators of
intracellular Ca2⫹.20
We hypothesized that the opposing responses of the pulmonary and systemic vascular beds to hypoxia are, at least in
part, due to differences in mitochondria function. Although
mitochondrial diversity has been shown among neurons21 and
myocardial cells,22,23 and between organs (pigeon heart versus rat liver mitochondria24) potential diversity of mitochondrial function and its link to tone has not been reported so far
in the vasculature.
Materials and Methods
Perfused Lung-Kidney Model
In this model, the isolated lungs and kidneys are perfused in series
with a shared, blood-free solution as previously described.25 The
flow rate was kept constant in the lung (28 mL/min) and the kidney
(10.5 mL/min) using separate perfusion pumps. Changes in PA and
RA pressure were recorded in response to hypoxia or drugs.
Tissue Baths
Resistance PA and RA rings (4 to 5th order), denuded of endothelium, were studied as previously described.26
Patch Clamping
Whole-cell patch-clamp recordings were performed in freshly isolated PASMCS and RASMCs, as previously described.27
Chemiluminescence
Lucigenin (10⫺5 mol/L) enhanced chemiluminescence, a measure of
the production of AOS, was measured in vascular rings or isolated
mitochondria using a Packard 1900CA Liquid Scintillation Analyzer
as described.6 The amount of mitochondria studied was normalized
for protein content. Both arteries and mitochondria were studied
immediately after isolation.
H2O2 Measurement
H2O2 production in vascular rings and mitochondria at baseline and
in response to hypoxia and ETC blockers was measured using the
AmplexRed H2O2 assay kit (according to the manufacturer’s instructions) as well as DCFH-DA (2⬘,7⬘-dichlorofluorescin diacetate)
assay, both from Molecular Probes. Background as well as drug
autofluorescence was subtracted. Significant autofluorescence was
seen with 10 ␮mol/L (but not 1 ␮mol/L) aqueous sodium cyanide
(see online Figure 1, which can be found in the online data
supplement available at http://www.circresaha.org).
Mitochondria isolation was performed and respiration studied
using standard methodology.28,29 Lung and Kidney mitochondria
were kept at 4°C and studied immediately. A protein assay (BioRad)
was performed on the isolated mitochondria, and the pellets were
diluted accordingly to achieve equal mitochondrial protein loading.
O2 consumption was measured with an O2 electrode using a MacLab
D/A converter (AD Instruments). The maximum change in the rate
of respiration in response to the proximal ETC complex substrates
(glutamate 10 mmol/L for complex I and succinate 2.5 mmol/L for
complex II) or substrate plus inhibitor (rotenone 5 ␮mol/L, antimycin A 50 ␮mol/L) was then calculated.
Mitochondrial Quality Control
To determine whether the quality of mitochondria was similar
between the lung and kidney preparations (ie, to exclude possible
differential damage of the one versus the other preparation during
isolation), we used 2 standard quality control techniques: the
NADH-supported respiration in intact versus lysed mitochondria and
measurement of the respiratory control ration (RCR).28,29
Immunoblotting
Immunoblotting of ETC proteins from isolated mitochondria and
MnSOD from isolated arteries was performed as previously
described.27
Confocal Microscopy
Mitochondrial membrane potential (⌬⌿m) was studied in first
passage cultured PASMCs and RASM, using two different dyes that
are widely used in studying ⌬⌿m, JC-1 and TMRM (tetramethylrhodamine methyl-ester perchlorate). Cells were loaded with either
JC-1 (1 ␮mol/L) or TMRM (20 nmol/L) for 30 minutes (37°C). JC-1
is a cationic fluorescent dye that exhibits potential-dependent mitochondrial accumulation, in that it fluoresces green when uptaken in
depolarized mitochondria, whereas it dimerizes and fluoresces red
when uptaken in hyperpolarized mitochondria.30,31 TMRM-loaded
mitochondria exhibit stronger red fluorescence with hyperpolarization.32 ⌬⌿m was compared between PASMC and RASMC mitochondria at baseline and in response to hypoxia. All the imaging
parameters were kept identical between the two preparations.
An expanded Materials and Methods section can be found in the
online data supplement available at http://www.circresaha.org.
Results
Rotenone Mimics Hypoxia Constricting the
Pulmonary While Dilating the Renal Circulation
Rotenone (5 ␮mol/L) and hypoxia both constrict the pulmonary, whereas they dilate the renal vascular bed (Figures 1A
and 1B). This difference is independent of nitric oxide and
prostaglandins because the experiments were performed in
the presence of inhibitors of the endothelial nitric oxide
Michelakis et al
Mitochondrial Diversity in the Vasculature
3
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Figure 1. Opposing actions of rotenone and hypoxia on the renal vs the pulmonary circulation. A and B, Both rotenone (5 ␮mol/L) and
hypoxia (PO2 40 mm Hg) constrict the pulmonary and dilate the renal circulation in the presence of L-NAME and meclofenamate. 4-AP
(5 mmol/L) and angiotensin II (AII, 0.1 ␮g bolus, total perfusate 50 mL) constrict both vascular beds. C, Rotenone mimics hypoxia, constricting PA rings while dilating RA rings (both denuded of endothelium), whereas 4-AP and phenylephrine (PE) constrict both vessels.
synthase, L-NG-nitroarginine methylester (L-NAME, 5⫻10⫺5
mol/L) and cyclooxygenase (meclofenamate, 1.7⫻10⫺5 mol/L).
In contrast, both vascular beds constrict to angiotensin II (10⫺5
mol/L) and 4-aminopyridine (4-AP, 5 mmol/L), a Kv channel
blocker. The order of hypoxia, rotenone, and 4-AP challenges
varied.
Further evidence that the opposing response to hypoxia and
rotenone is intrinsic to the SMCs and independent of the
endothelium comes from tissue bath experiments. In endothelium-denuded vessels, rotenone (5 ␮mol/L) mimics hypoxia again, constricting PAs and dilating RAs, whereas 4-AP
(5 mmol/L) constricts both RAs and PAs (Figure 1C).
Rotenone Mimics Hypoxia Inhibiting Kⴙ Current
in PASMCs While Activating it in RASMCs
Whole-cell patch clamping in freshly isolated SMCs from
resistance PAs and RAs shows that rotenone inhibits outward
K⫹ current (Ik) in PASMCs while it activates Ik in RASMCs
(Figure 2). The effects of both hypoxia and rotenone were
rapid and occurred within 5 minutes.
Production of AOS and H2O2 at Baseline and in
Response to Hypoxia and ETC Inhibitors Is
Differentially Regulated Between PAs and RAs
We measured baseline AOS production (during normoxia) as
well as the effects of hypoxia and ETC inhibitors in isolated,
denuded, resistance PA and RA rings using 3 independent
methods. (1) Lucigenin-enhanced chemiluminescence, which
preferentially detects superoxide anion levels,33 shows that
baseline AOS levels are significantly higher in the PAs
compared with the RAs, even after incubation with diphenyliodonium (DPI, 1 ␮mol/L), an inhibitor of NAD(P)H
oxidase34 (Figure 3A). Rotenone’s effects on the production
of AOS parallel those of hypoxia. They both decrease AOS
production in isolated PA rings, confirming previous studies
on whole lung preparations.6 In contrast, hypoxia and rotenone increased AOS production in the RAs (Figure 3A),
confirming studies in coronary arteries.35 (2) The AmplexRed
assay, which is specific for H2O2, shows that, concordant with
the chemiluminescence data, the production of H2O2 is higher
in the PAs versus the RAs at baseline (Figure 3B). Once
again, a proximal ETC inhibitor (antimycin A, 50 ␮mol/L)
mimics hypoxia, and they both decease the production of
H2O2 in the PAs. However, neither hypoxia nor antimycin A
alter the production of H2O2 in the RA (Figure 3B). (3)
DCFH-DA fluorescence (Figure 3C), which like AmplexRed
preferentially measures H2O2, once again decreased H2O2 in
PAs but did not affect the RA H2O2 production. The proximal
ETC (complex I) blocker rotenone once again mimicked
hypoxia and decreased the H2O2 production in the PAs,
although it did not have any effects on the RAs. The complex
III blocker myxothiazol had a significant trend to inhibit H2O2
production in the PAs but did not reach statistical significance
(P⬍0.057). Cyanide (1 ␮mol/L) did not affect H2O2 production in either artery. We observed a significant autofluorescence of the aqueous sodium cyanide solution itself at the
dose of 10 ␮mol/L, which prevented us from studying it at
this dose (see online Figure 1).
Lung Mitochondria Are Less Active and Produce
More AOS and H2O2 Than Kidney Mitochondria
The fact that the PAs have higher AOS than the RAs, even in
the presence of an NADPH oxidase inhibitor, suggests that
mitochondria might be the source of this differentially regulated production of AOS. Metabolically active mitochondria
are known to produce less AOS than less active mitochon-
4
Circulation Research
June 28, 2002
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Figure 2. Opposing effects of rotenone and hypoxia on PASMC vs RASMC Ik. Both hypoxia and rotenone inhibit whole cell PASMC Ik,
whereas they both increase RASMC Ik.
dria.20 We isolated lung and kidney mitochondria and studied
respiration driven by substrates that enter the ETC proximally
at the complexes I and II (glutamate and succinate; for results
with glutamate alone versus succinate and glutamate, see the
online data supplement). We show that lung mitochondria
have significantly lower rates of respiration compared with
kidney mitochondria (Figure 4A). Rotenone and antimycin A
inhibit respiration in both lung and kidney mitochondria,
although the lung mitochondria are more sensitive to both
inhibitors than are kidney mitochondria. The percent inhibi-
Figure 3. Differences between the PA and RA production of AOS. A, Lucigenin-enhanced chemiluminescence at normoxic baseline is
much higher in denuded PA rings than in denuded RA rings (‡P⬍0.001), even in the presence of the NAD(P)H oxidase inhibitor DPI
(*P⬍0.05). Both rotenone and hypoxia (PO2 40 mm Hg) decrease AOS in the PA (†P⬍0.01) and increase AOS production in the RA
(**P⬍0.05). B, PAs produce more H2O2 than RAs (‡P⬍0.01). Both hypoxia and antimycin A decrease H2O2 production in the PA
(†P⬍0.05) but not the RA. C, PAs produce more H2O2 than RAs. Both rotenone and hypoxia decrease H2O2 production in the PA but
not in the RA. Myxothiazol and cyanide do not alter H2O2 production in either vessel (*P⬍0.05 compared with normoxic PAs; †P⬍0.01
compared with normoxic PAs).
Michelakis et al
Mitochondrial Diversity in the Vasculature
5
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Figure 4. PA mitochondria are less active than RA mitochondria. A, Baseline respiration is greater in isolated kidney vs lung mitochondria. Rotenone and antimycin A inhibit respiration significantly in both organs (†P⬍0.01, *P⬍0.05). Substrates⫽glutamate (10 mmol/
L)⫹succinate (2.5 mmol/L). B, Mitochondrial preparations are ⬎90% pure as seen by electron microscopy. A few lysosomes are seen.
D, Quality assessment by 2 assays (respiratory control ratio and the ratio of NADH-supported respiration in sonicated/intact mitochondria) shows that the condition of mitochondria from the lung and kidney is similar.
tion in respiration by rotenone for the lung and kidney respectively is ⫺87⫾6 and ⫺53⫾5 and for antimycin A is ⫺96⫾3
and ⫺68⫾15, respectively. The purity of our preparations is
shown by electron microscopy, where mostly intact mitochondria and a few lysosomes are seen (Figure 4B).
The mitochondria isolation process could theoretically
have resulted in preferential damage of one preparation
versus the other, complicating their comparison. We therefore
used 2 different assays to ensure that the quality of the
mitochondria is similar in both preparations. Both preparations have similar RCRs (Figure 4C). These RCRs are lower
that the RCRs from heart mitochondria (⬎10 in the literature29 and 11.4⫾0.4, n⫽3 in our hands), which is expected
because blood vessels are metabolically less efficient than the
myocardium. They show, however, that the quality of our
preparations is similar. This was also confirmed by another
assay, the ratio of NADH-driven respiration in sonicated
versus intact mitochondria. Intact good quality mitochondria
are impermeable and do not respond to NADH, whereas
damaged “leaky” mitochondria (sonicated) respond to NADH
with increased respiration.28 The ratio of NADH-supported
respiration in sonicated/intact mitochondria is similar between the lung and kidney (Figure 4C).
We measured AOS and H2O2 production in lung and kidney
mitochondria preparations with similar amounts of mitochondria, based on protein assays, and under identical conditions
(Figures 5A and 5B). Lung mitochondria produce more AOS
and H2O2 than the kidney mitochondria at baseline. The same
proximal ETC substrate (glutamate⫹succinate) was used in both
assays. The near complete elimination of the signal in the
AmplexRed assay by catalase (10 000 U) confirms the specific-
ity of the assay for H2O2 (Figure 5B). Antimycin A significantly
inhibits the lung AOS production, but it does not alter the kidney
mitochondria AOS production (Figure 5A). In contrast, the distal
ETC inhibitor cyanide (10 ␮mol/L) does not alter AOS production in either preparation, confirming previous reports.6 These
data are consistent with complexes I and III being the major sites
of AOS production within the ETC.20
We then measured the expression of ETC complexes I and
III (where most of the production of AOS occurs20) in PAs
and RAs using immunoblotting. Although the ETC complexes consist of several subunits each, the commercially
available subunits that we studied suggest that complexes I
and III are expressed more in the RAs than the PAs (Figure
5C). Similar loading of the gels was confirmed by the
Ponceau staining. In addition, using the same immunoblots
(see online Figure 3B), we show that PAs express much more
MnSOD, suggesting that the lower expression of proximal
ETC complexes in this vessel is not a nonspecific finding or
artifact of loading.
PASMC Mitochondria Are More Depolarized
Than RASMC Mitochondria
We studied mitochondrial function within cultured PASMCs
and RASMCs (first passage) loaded with 2 different dyes that
are widely used in the measurement of ⌬⌿m, JC-1 and
TMRM, using confocal microscopy (Figure 6).
With JC-1, mitochondria from RASMCs (n⫽44) have
significantly higher ⌬⌿m than mitochondria from PASMCs
(n⫽28) under identical loading and imaging conditions (Figure 6A). Extensive filamentous networks of mitochondria
throughout the cytoplasm are shown in the representative
6
Circulation Research
June 28, 2002
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Figure 5. Lung mitochondria produce more AOS than kidney mitochondria, perhaps due to differences in the expression of ETC complexes I and III. A, Lucigenin-enhanced chemiluminescence is higher at baseline in the lung vs kidney mitochondria (*P⬍0.05). Antimycin, but not cyanide, decrease AOS production in the lung but not in the kidney (†P⬍0.05). B, Catalase-sensitive production of H2O2 is
higher in the lung vs kidney mitochondria at baseline (*P⬍0.05). Glutamate⫹succinate were used as substrates in both A and B. C,
Expression of complex I and III (immunoblotting) is less in the PA.
pictures in Figure 6A. When the SMCs were superfused with
a hypoxic solution, the PASMC mitochondria (n⫽30) hyperpolarized (decreased green/red ratio), whereas the RASMC
mitochondria (n⫽40) depolarized.
In agreement with the JC-1 data, TMRM-loaded mitochondria from RASMCs (n⫽33) have significantly higher ⌬⌿m
compared with mitochondria from PASMC, loaded and
imaged under identical conditions (Figure 6B).
Figure 6. ⌬⌿m is more negative in RASMC mitochondria vs PASMC. A, JC-1–loaded PASMCs and RASMCs. Representative confocal
images are shown in the left (⫻40). Mean data for baseline ⌬⌿m, expressed as the ratio of the relative fluorescent units in the green
and red channels, are shown on the right. PASMC mitochondria are more depolarized than the RASMC mitochondria. Hypoxia hyperpolarizes PASMC mitochondria, whereas it depolarizes RASMC mitochondria. B, TMRM-loaded RASMC mitochondria exhibit stronger
red fluorescence, suggesting they are more hyperpolarized than PASMC mitochondria loaded and imaged under identical conditions
(*P⬍0.05, **P⬍0.0001).
Michelakis et al
Discussion
We report that physiologically significant mitochondrial diversity exists between the renal and the pulmonary circulations. This diversity exists both at baseline and in response to
ETC inhibitors and hypoxia. We show that mitochondrial
function is different not only between the PAs and RAs but
between the lung and kidney as well, perhaps reflecting
differences in the overall redox environment of the two
organs. We propose a model by which this diversity might at
least in part explain the differences of the two circulations in
both their redox state and the response to hypoxia. Our data
are supported by a multitude of techniques, including perfused organs, tissue baths, patch clamping, measurement of
AOS by 3 different techniques, mitochondrial respiration,
immunoblotting, and dynamic confocal imaging using 2
different ⌬⌿m-sensitive dyes.
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Baseline Differences Between the PA/Lung and
RA/Kidney Mitochondria
Under identical experimental conditions, lung mitochondria
have slower respiratory rates (Figure 4A) than kidney mitochondria. Quality control studies, showed that there was no
differential damage during the isolation procedure (Figures
4B and 4C). Direct imaging of vascular SMCs showed that
PASMC are more depolarized than RASMC mitochondria
(Figures 6A and 6B). In addition, and perhaps related to these
differences, lung mitochondria produce more AOS and H2O2
than kidney mitochondria at baseline (Figure 4A and 4B).
Although antimycin inhibited respiration in both preparations
(Figure 4A), the AOS production in the lung is sensitive to
antimycin, whereas the kidney AOS production is antimycininsensitive (Figure 5A). Early work by Boveris and Chance24
showed similar differences in the antimycin sensitivity between the pigeon heart and rat liver mitochondria H2O2
production. This suggests that the regulation of AOS production within the pETC might be different among organs or
vessels.
Redox Potential and the AOS Production Are
Different in the PAs Compared With the RAs
We used 3 different techniques and showed that freshly
isolated PAs have higher AOS and H2O2 production at
baseline compared with the RAs (Figure 3). We also showed
that hypoxia and proximal, but not distal ETC inhibitors,
significantly decrease AOS production in the PA. Several
groups have suggested that proximal ETC function is important in HPV.15,16 However, these authors reported that AOS
were increased by hypoxia and proximal ETC inhibition,
whereas most groups find hypoxia decreases AOS.6,7,36,37 A
strength of our study is that AOS were measured in both
mitochondria and in vessels using 3 different assays. The
results of these assays were all concordant. In addition, we
studied AOS production is freshly isolated arteries, which
may be more physiological than the studies of AOS in
cultured PASMCs, as performed by others.15 Even after
inhibiting NADPH oxidase, there is residual AOS production,
which is greater in the PA than the RA (Figure 3A),
consistent with the concept that mitochondrial diversity exists
Mitochondrial Diversity in the Vasculature
7
between PAs and RAs, beyond any difference in NADPH
oxidase that may exist.
Interestingly, although lucigenin-enhanced chemiluminescence was increased in RAs in response to hypoxia and
proximal ETC inhibitors, the amount of H2O2 production was
not altered. This might be perhaps due to lower levels of
MnSOD in the RASMCs. MnSOD catalyzes the formation of
H2O2 from superoxide, the major AOS detected by chemiluminescence. Indeed, we showed that both mRNA and expressed protein for MnSOD are significantly lower in the
RAs (online data supplement results and online Figure 3).
The differences in AOS production in PAs versus RAs
parallel the differences in AOS production in lung versus
kidney mitochondria. This might be related to the fact that
there might be a more generalized difference in the redox
environment between the 2 organs that affects their vasculature as well. In vivo, the lungs and the resistance PAs are
exposed to alveolar oxygen levels, which are higher than
those in arterial blood perfusing the rest of the organs like the
kidneys. This might have induced redox differences. In fact
we show that the levels of reduced glutathione (GSH) in the
lungs in normoxia are much higher than those in the kidney
(online Figure 3). The findings that the lungs have a more
oxidized redox potential than the kidneys and their importance are discussed in the online data supplement.
ETC Function Is Inversely Related to AOS
Production: Lessons From Human and Animal
Disease Models
The difference between PA and RA mitochondria is analogous to the comparison of mitochondria from healthy subjects
versus patients with ETC complex I deficiency. Fibroblasts
from complex I– deficient patients have increased
mitochondrial-derived AOS production when compared with
controls.38,39 In addition, their fibroblasts have impaired
mitochondrial function, as reflected by depolarized ⌬⌿m.38,39
Furthermore, superoxide-induced MnSOD is upregulated in
these patients.38 The normoxic PASMCs, like the fibroblasts
from complex I deficiency patients, have (compared with the
kidney) decreased levels of complex I (Figure 5C), decreased
mitochondrial respiration (Figure 4A), and depolarized ⌬⌿m
(Figure 6A). Lung mitochondria also make more superoxide
and H2O2 than the kidney mitochondria (Figures 5A and 5B).
The higher oxidative stress of PAs is associated with an
apparently homeostatic induction of MnSOD expression and
elevated GSH levels (online Figure 3). Furthermore, the fact
that the RAs are relatively “deficient” in both GSH and
MnSOD, both forms of oxidant defense, is in agreement with
the finding that mice that genetically lack MnSOD also have
30% less GSH when compared with wild-type mice.40
How do these differences in ETC function and AOS
production relate to the opposing response of the PA and RA
to hypoxia? Important clues come from the striking similarities in the effects of ETC inhibitors and hypoxia discussed
subsequently.
Proximal ETC Inhibitors Mimic the Opposing
Effects of Hypoxia in PAs Versus RAs
We show that proximal ETC inhibitors and hypoxia inhibit
PASMC Ik and constrict the Pas, whereas they activate
8
Circulation Research
June 28, 2002
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Figure 7. Proposed sensors, mediators, and effectors in vascular O2 sensing.
RASMC Ik and dilate the renal circulation (Figures 1 to 2).
These inhibitors are the only class of drugs, to our knowledge, that mimic hypoxia in both circulations. Other important modulators of HPV do not mimic hypoxia in that they
constrict both the pulmonary and systemic circulations. For
example, endothelin-1 constricts both PAs41 and RAs42 and
inhibits Kv currents depolarizing both the PASMCs43 and
RASMCs.44 Similarly, the Kv blocker 4-AP constricts both
vessels (Figure 1). This suggests that the mechanism of the
opposing effects of hypoxia on the 2 circulations is intrinsic
to the mitochondrial ETC, upstream from the Kv channels.
Furthermore, because these opposing effects persist in the
absence of endothelium (Figures 1A through 1C) and are
present in isolated SMCs (Figures 2 and 6), it suggests that
they are intrinsic to the vascular SMCs and not the
endothelium.
Vascular SMC Mitochondrial ETC Is an
O2 Sensor
Our findings are in agreement with several recent studies that
suggest a central role of mitochondria ETC and redox
mediators in HPV,6,15,16 as previously proposed.5 Based on
these studies, a potential mechanism for vascular O2 sensing
includes (Figure 7) an O2 sensor (mitochondrial ETC or
vascular oxidases) that regulates the production of a diffusible redox factor in proportion to O2 (oxygen radicals and
peroxides, collectively known as AOS). Of the various
species, H2O2 is an attractive candidate mediator because it is
more stable than oxygen radicals, such as superoxide, it is
freely diffusible, and it can affect several cellular mechanisms
involved in tone control, including O2-sensitive K⫹ channels
(effectors; for review, see Archer at al45).
Our study supports many of the known features of vascular
O2 sensing: the high level of H2O2 production at baseline in
the PAs could explain the relatively vasodilated (lowpressure) state of the pulmonary, compared with the systemic,
circulation. Tonically produced AOS/H2O2 from the proximal
ETC can diffuse to the cytoplasm and cause K⫹ channel
activation, SCM hyperpolarization, decreased opening of the
voltage-gated Ca2⫹ channels and decreased [Ca2⫹]i levels and
vasodilatation (Figure 7). During acute hypoxia, the tonic
production of these vasodilators (AOS/H2O2) decreases (Figure 3), thus promoting Ik inhibition (Figure 2) and vasoconstriction (Figure 1), ie, HPV. In contrast, a hypoxia-induced
increase in AOS (Figure 3A) would increase Ik46 and promote
RA vasodilatation.
We acknowledge that more than one O2-sensing mechanism might take place in one or both circulations at the same
time. Thus, novel vascular oxidases could also contribute to
redox signaling.
The concept of mitochondrial diversity in the vasculature is
new and requires further study as it is likely relevant to other
aspects of vascular function. In addition to its role in vascular
O2 sensing, the diversity concept may have implications for
apoptosis, vascular wall remodeling, or ischemia/reperfusion
injury. It remains to be shown whether this mitochondrial
diversity is due to a genetic difference intrinsic to the
mitochondria or a result of the different redox environment
that the PAs are exposed to, compared with the RAs.
Acknowledgments
Drs Michelakis and Archer are both supported by the Canadian
Institutes for Health Research, The Heart and Stroke Foundation of
Alberta, and the Alberta Heritage Foundation for Medical Research.
References
1. Vadula MS, Kleinman JG, Madden JA. Effect of hypoxia and norepinephrine on cytoplasmic free Ca2⫹ in pulmonary and cerebral arterial
myocytes. Am J Physiol. 1993;265:L591–L597.
Michelakis et al
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
2. Wolin MS. Interactions of oxidants with vascular signaling systems.
Arterioscler Thromb Vasc Biol. 2000;20:1430 –1442.
3. Reeve HL, Weir EK, Nelson DP, Peterson DA, Archer SL. Opposing
effects of oxidants and antioxidants on K⫹ channel activity and tone in
vascular tissue. Exp Physiol. 1995;80:825– 834.
4. Weir EK, Archer SL. The mechanism of acute hypoxic pulmonary vasoconstriction: the tale of two channels. FASEB. 1995;9:183–189.
5. Archer S, Will J, Weir E. Redox status in the control of pulmonary
vascular tone. Herz. 1986;11:127–141.
6. Archer SL, Huang J, Henry T, Peterson D, Weir EK. A redox based
oxygen sensor in rat pulmonary vasculature. Circ Res. 1993;73:
1100 –1112.
7. Freeman BA, Crapo JD. Hyperoxia increases oxygen radical production
in rat lungs and lung mitochondria. J Biol Chem. 1981;256:10986 –10992.
8. McMurtry I, Davidson B, Reeves J, Grover R. Inhibition of hypoxic
pulmonary vasoconstriction by calcium antagonists in isolated rat lungs.
Circ Res. 1976;38:99 –104.
9. McMurtry I. Bay K8644 potentiates and A23187 inhibits hypoxic vasoconstriction in rat lungs. Am J Physiol. 1985;249:H741–H746.
10. Robertson TP, Hague D, Aaronson PI, Ward JP. Voltage-independent
calcium entry in hypoxic pulmonary vasoconstriction of intrapulmonary
arteries of the rat. J Physiol. 2000;525(suppl 3):669 – 680.
11. Mohazzab-H K, Fayngersh R, Kaminski P. Potential role of NADPH
oxidoreductase-derived reactive O2 species in calf pulmonary arterial
PO2-elicited responses. Am J Physiol. 1995;269:L637–L644.
12. Mohazzab-H K, Wolin M. Properties of a superoxide anion-generating
microsomal NADH oxidoreductase, a potential pulmonary artery PO2
sensor. Am J Physiol. 1994;267:L823–L831.
13. Archer SL, Reeve HL, Michelakis EM, Puttagunta L, Waite R, Nelson
DP, Dinauer MC, Weir EK. O2-sensing is preserved in mice lacking the
91 kD subunit of Nicotinamide Adenine Dinucleotide Phosphate Oxidase,
NAD(P)H oxidase. Proc Natl Acad Sci U S A. 1999;96:7944 –7949.
14. Obeso A, Sanz-Alfayate G, Agapito MT, Gonzalez C. Significance of
ROS in oxygen chemoreception in the carotid body chemoreception:
apparent lack of a role for NADPH oxidase. Adv Exp Med Biol. 2000;
475:425– 434.
15. Waypa GB, Chandel NS, Schumacker PT. Model for hypoxic pulmonary
vasoconstriction involving mitochondrial oxygen sensing. Circ Res.
2001;88:1259 –1266.
16. Leach RM, Hill HM, Snetkov VA, Robertson TP, Ward JP. Divergent
roles of glycolysis and the mitochondrial electron transport chain in
hypoxic pulmonary vasoconstriction of the rat: identity of the hypoxic
sensor. J Physiol. 2001;536:211–224.
17. Mulligan E, Lahiri S, Storey BT. Carotid body O2 chemoreception and
mitochondrial oxidative phosphorylation. J Appl Physiol. 1981;51:
438 – 446.
18. Mojet MH, Mills E, Duchen MR. Hypoxia-induced catecholamine
secretion in isolated newborn rat adrenal chromaffin cells is mimicked by
inhibition of mitochondrial respiration. J Physiol. 1997;504:175–189.
19. Buescher P, Perse D, Pillai R, Litt M, Mitchell M, Sylvester JT. Energy
state and vasomotor tone in hypoxic pig lungs. J Appl Physiol. 1991;70:
1874 –1881.
20. Duchen MR. Contributions of mitochondria to animal physiology: from
homeostatic sensor to calcium signalling and cell death. J Physiol. 1999;
516:1–17.
21. Pettus EH, Betarbet R, Cottrell B, Wallace DC, Madyastha V,
Greenamyre JT. Immunocytochemical characterization of the mitochondrially encoded ND1 subunit of complex I (NADH: ubiquinone oxidoreductase) in rat brain. J Neurochem. 2000;75:383–392.
22. Palmer JW, Tandler B, Hoppel CL. Biochemical properties of subsarcolemmal and interfibrillar mitochondria isolated from rat cardiac muscle.
J Biol Chem. 1977;252:8731– 8739.
23. Weinstein ES, Benson DW, Fry DE. Subpopulations of human heart
mitochondria. J Surg Res. 1986;40:495– 498.
24. Boveris A, Chance B. The mitochondrial generation of hydrogen
peroxide. Biochem J. 1973;134:707–716.
Mitochondrial Diversity in the Vasculature
9
25. Hampl V, Weir EK, Archer SL. Endothelium-derived nitric oxide is less
important for basal tone regulation in the pulmonary than the renal
circulation of the adult rat. J Vasc Med Biol. 1994;5:22–30.
26. Michelakis E, Weir E, Nelson D, Reeve H, Tolarova S, Archer S.
Dexfenfluramine elevates systemic blood pressure by inhibiting
potassium currents in vascular smooth muscle cells. J Phamacol Exper
Ther. 1999;291:1143–1149.
27. Michelakis ED, Weir EK, Wu X, Nsair A, Waite R, Hashimoto K,
Puttagunta L, Knaus HG, Archer SL. Potassium channels regulate tone in
rat pulmonary veins. Am J Physiol Lung Cell Mol Physiol. 2001;280:
L1138 –L1147.
28. Darley-Usmar VM, Rickwood D, Wolson MT. Mitochondria: A Practical
Approach. Oxford, UK: IRL Press; 1987.
29. Nedergaard J, Cannon B. Overview: preparation, and properties of mitochondria from different sources. Methods Enzymol. 1979;55:3–33.
30. Reers M, Smiley ST, Mottola-Hartshorn C, Chen A, Lin M, Chen LB.
Mitochondrial membrane potential monitored by JC-1 dye. Methods
Enzymol. 1995;260:406 – 417.
31. Salvioli S, Ardizzoni A, Franceschi C, Cossarizza A. JC-1, but not
DiOC6(3) or rhodamine 123, is a reliable fluorescent probe to assess delta
psi changes in intact cells: implications for studies on mitochondrial
functionality during apoptosis. FEBS Lett. 1997;411:77– 82.
32. Nicholls DG, Ward MW. Mitochondrial membrane potential and
neuronal glutamate excitotoxicity: mortality and millivolts. Trends
Neurosci. 2000;23:166 –174.
33. Li Y, Zhu H, Kuppusamy P, Roubaud V, Zweier J, Trush M. Validation
of lucigenin (bis-N-methylacridinium) as a chemilumigenic probe for
detecting superoxide anion radical production by enzymatic and cellular
systems. J Biol Chem. 1998;273:2015–2023.
34. Cross A, Jones O. The effect of the inhibitor diphenylene iodonium on the
superoxide generating system of neutrophils. Biochem J. 1986;237:
111–116.
35. Kaminski PM, Wolin MS. Hypoxia increases superoxide anion production from bovine coronary microvessels, but not cardiac myocytes, via
increased xanthine oxidase. Microcirculation. 1994;1:231–236.
36. Paky A, Michael J, Burke-Wolin T, Wolin M. Endogenous production of
superoxide by rabbit lungs: effects of hypoxia or metabolic inhibitors.
J Appl Physiol. 1993;74:2868 –2874.
37. Burke-Wolin T, Wolin MS. H2O2 and cGMP may function as an O2
sensor in the pulmonary artery. J Appl Physiol. 1989;66:167–170.
38. Pitkanen S, Robinson BH. Mitochondrial complex I deficiency leads to
increased production of superoxide radicals and induction of superoxide
dismutase. J Clin Invest. 1996;98:345–351.
39. Barrientos A, Moraes C. Titrating the effects of mitochondrial complex I
impairment in the cell physiology. J Biol Chem. 1999;274:16188 –16197.
40. Williams MD, Van Remmen H, Conrad CC, Huang TT, Epstein CJ,
Richardson A. Increased oxidative damage is correlated to altered mitochondrial function in heterozygous manganese superoxide dismutase
knockout mice. J Biol Chem. 1998;273:28510 –28515.
41. Wong J, Vanderford PA, Winters JW, Chang R, Soifer SJ, Fineman JR.
Endothelin-1 does not mediate acute hypoxic pulmonary vasoconstriction
in the intact newborn lamb. J Cardiovasc Pharmacol. 1993;22:
S262–S266.
42. Naicker S, Bhoola KD. Endothelins: vasoactive modulators of renal
function in health and disease. Pharmacol Ther. 2001;90:61– 88.
43. Shimoda LA, Sylvester JT, Sham JS. Inhibition of voltage-gated K⫹
current in rat intrapulmonary arterial myocytes by endothelin-1. Am J
Physiol. 1998;274:L842–L853.
44. Gordienko DV, Clausen C, Goligorsky MS. Ionic currents and endothelin
signaling in smooth muscle cells from rat renal resistance arteries. Am J
Physiol. 1994;266:F325–F341.
45. Archer SL, Weir EK, Reeve HL, Michelakis E. Molecular identification
of O2 sensors and O2-sensitive potassium channels in the pulmonary
circulation. Adv Exp Med Biol. 2000;475:219 –240.
46. Gebremedhin D, Bonnet P, Greene AS, England SK, Rusch NJ, Lombard
JH, Harder DR. Hypoxia increases the activity of Ca2⫹-sensitive K⫹
channels in cat cerebral arterial muscle cell membranes. Pflügers Arch.
1994;428:621– 630.
Diversity in Mitochondrial Function Explains Differences in Vascular Oxygen Sensing
Evangelos D. Michelakis, Vaclav Hampl, Ali Nsair, XiCheng Wu, Gwyneth Harry, Al Haromy,
Rachita Gurtu and Stephen L. Archer
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Circ Res. published online June 6, 2002;
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2002 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7330. Online ISSN: 1524-4571
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circres.ahajournals.org/content/early/2002/06/06/01.RES.0000024689.07590.C2.citation
Data Supplement (unedited) at:
http://circres.ahajournals.org/content/suppl/2002/06/17/90.12.1307.DC1
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation Research can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further information
about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation Research is online at:
http://circres.ahajournals.org//subscriptions/
MS #3695-R2 ONLINE SUPPLEMENT
Mitochondrial Diversity in the Vasculature
On-line Supplement
Methods
The rats were euthanized with pentobarbital overdose before the lungs and kidneys were removed. All
protocols were approved by the University of Alberta animal ethics committee.
The perfused lung-kidney model. In this model the isolated lungs and kidneys are perfused in series
with a shared, blood-free solution as previously described. 1 Under anesthesia (pentobarbital, 50mg/kg,
IP) the abdomen was opened and dual lumen cannulas were advanced into the right RA and PA. The
pulmonary circuit drained into a heated reservoir via a left atrial cannula. The lungs were suspended by
a tracheal cannula and ventilated. The right kidney was removed and placed in a heated Petri dish. The
flow rate was kept constant in the lung (28ml/min) and the kidney (10.5ml/min) using separate
perfusion pumps. Changes in PA and RA pressure were recorded in response to hypoxia (PO2
40mmHg, PCO2 40mmHg, pH 7.4) or drugs. Hypoxia was induced by ventilating the lungs with a
hypoxic gas (2.5% O2, 5% CO2, balance N2).
Tissue baths. Resistance PA and RA rings (4-5th order), denuded of endothelium, were studied in 4mL tissue baths as previously described.
2
Rings were constricted with phenylephrine (10-5M) but
failed to relax to Acetylcholine (Ach, 10-7,6M). This confirmed their denudation since in the absence of
functional endothelium, Ach either fails to relax or causes a slight vasoconstriction. Optimal resting
tension, defined as the tension at which maximal constriction to KCl 80mM occurred, was 0.8g for the
PA and 1g for the RA.
1
MS #3695-R2 ONLINE SUPPLEMENT
Patch clamping: Whole cell patch-clamp recordings were performed as previously described 3. The
pipette solution contained (in mM): KCl 140, MgCl2 1.0, HEPES 10, EGTA 5, Glucose 10, pH 7.2.
The chamber containing the cells was perfused (2 ml/min) with a solution containing (in mM) NaCl
145, KCl 5.4, MgCl2 1.0, CaCl2 1.5, HEPES 10, glucose 10, pH 7.4 (extracellular solution). Cells
were voltage-clamped at a holding potential of -70 mV and currents were evoked by steps from –130
mV to +50mV, as shown in Figure 2. The normoxic and hypoxic solution perfusate reservoirs were
bubbled with 20% and 0% O2, respectively (plus 3.5% CO2/balance N2), producing normoxic and
hypoxic PO2s in the cell chamber of 120 and 40 mm Hg, respectively. PCO2 was 40 mm Hg, and pH
was 7.40 to 7.41 under both conditions. Drugs were dissolved in the normoxic solution. The pH of 4aminopyridine (4-AP) was adjusted to 7.4 before dissolved in the normoxic solution.
Pipette resistances ranged from 1 to 4 M after Sylgard coating and fire-polishing, with series
resistance typically compensated by 90% to 95%. Data were recorded and analyzed using pCLAMP
6.02 software (Axon Instruments, Foster City, CA). Whole-cell capacitance was measured using the
automated function of the Axopatch amplifier. Whole-cell currents for each cell were divided by the
cell's capacitance, giving a measure of current density (in picoamperes per picofarad).
Chemiluminescence: Lucigenin (10-5M) enhanced chemiluminescence, a measure of the production of
AOS, was measured using a Packard 1900CA Liquid Scintillation Analyzer as described 4. Vascular
o
rings or isolated mitochondria were placed in a 2 ml heated cuvette (37 C) and photon counts were
displayed at 0.1 second intervals. An equal weight of resistance RA and PA rings was used in each
2
MS #3695-R2 ONLINE SUPPLEMENT
experiment. The amount of mitochondria studied was normalized for protein content. Both arteries and
mitochondria were studied immediately after isolation. Background luminescence was subtracted from
the recorded values and readings after 3 minutes of hypoxia or drugs were reported. The preparation
was made hypoxic via bubbling with a hypoxic (PO2~40mmHg) versus normoxic (PO2~120mmHg)
solution (balanced CO2 in each case). Careful attention was made to preserve physiological pH and
temperature.
H2O2 measurement.
The AmplexRed™ H2O2 assay (Molecular Probes, Eugene, Oregon): The assay is based on the
detection of H2O 2 using 10-acetyl-3,7-dihydroxyphenoxazine (AmplexRed reagent), which in the
presence of horseradish peroxidase, reacts with H2O2 with a 1:1 stoichiometry to produce the highly
fluorescent resorufin-1 (absorption and fluorescence emission maxima of approximately 563nm and
587nm). Sections of vascular tissue (1-3mg) were incubated in vented 2ml capped Eppendorf tubes in
0.5ml of Krebs’ buffer (119mM NaCl, 3.52mM KCl, 1.17mM MgSO4, 1.18mM KH2PO4, 3.2mM
CaCl2, pH 7.4) plus 0.5ml of AmplexRed reagent. Hypoxic (PO2~40mm Hg) or normoxic
(PO2~120mm Hg) gas was bubbled through the incubation medium. Similar experiments were
performed with freshly isolated lung and kidney mitochondria, using the isolation media described
below.
The DCF assay: Production of H2O2 by isolated RA or PAs was measured by oxidation of 10µM
2',7'-dichlorofluorescin diacetate (DCF) (Molecular Probes, Eugene OR) in Krebs buffer at 37°C over
3
MS #3695-R2 ONLINE SUPPLEMENT
a sixty minute interval. Aqueous sodium cyanide was used at 1µM concentrations, myxothiazol
100ng/ml (in ethanol) and rotenone at 10µM (in DMSO) (Sigma Aldrich, Oakville ON.) Drugs were
tested under normoxia. Fluorescence resulting from oxidation of DCF was detected using a
spectrofluorometer, (Spectra Max Gemini XS, Molecular Probes, Eugene OR) with excitation at
495nm and emission at 530nm. Background was corrected for by running controls containing the
appropriate drug. Relative Fluoresence (RF) was calculated as follows (FU=fluorescence units):
Sample fluorescence (S)= sample FU60min – sample FU0min
Control fluorescence (C) control FU60min-control FU0min
RF = S - C/Tissue wt (mg)
Mitochondria isolation and respiration were studied using standard methodology.
5,6
Tissue was
homogenized in 15ml of isolation medium (250mM Mannitol, 5mM HEPES pH 7.2, 0.5mM EGTA,
0.1% fatty acid free BSA) in a homogenizer at 4°C. Debris was precipitated by two rounds of
centrifugation at 600xg for 5min at 4°C in a Sorvall superspeed centrifuge (DuPont Instruments New
Town CT USA). The mitochondrial fraction was recovered from the supernatant by centrifugation at
10,000xg, washed twice, and the pellet resuspended in 250mM sucrose with 0.1% BSA. Mitochondria
o
were kept at 4 C and studied within 3 hours from isolation. A protein assay (BioRad Hercules, CA
USA) was performed on the isolated mitochondria and the pellets were diluted accordingly to achieve
equal mitochondrial protein loading. Mitochondria were placed in an air-tight 2ml chamber in a heated
water bath with a tight control of temperature at 37oC. The chamber was stirred with a magnetic bar.
4
MS #3695-R2 ONLINE SUPPLEMENT
O2 consumption was measured with an O2 electrode using a MacLab D/A converter (AD Instruments,
Mountain View, CA). A buffer (1.5ml) was added containing (mM) KCl (45), potassium acetate (25),
EGTA (0.1), MgCl2 (1) and HEPES (10) (pH7.2) and respiration was measured for 10 minutes. The
maximum change in the rate of respiration in response to the proximal ETC complex substrates
(glutamate 10mM and succinate 2.5mM or substrate plus inhibitor (rotenone 5µM, antimycin A 50µM)
was then calculated.
We used succinate and glutamate together because we observed a significant additive effect in
mitochondrial respiration. This cocktail is commonly used in the studies of the proximal electron
transport chain function as in the classic paper by Boveris and Chance 7. Because the respiratory rates
of these mitochondrial are lower than other metabolically more active tissues (heart, liver) this increase
was helpful for our experiments. We have observed however that whether we use succinate alone or
glutamate alone, the kidney mitochondria have higher respiratory rates than the lung mitochondria
(online supplement figure 1).
Mitochondrial quality control: To determine whether the quality of mitochondria was similar
between the lung and kidney preparations we used two standard quality control techniques, NADHsupported respiration in intact versus lysed mitochondria and measurement of the respiratory control
ration (RCR)
5,6
. Mitochondrial integrity was assessed by calculating the ratio of NADH (10mM)-
supported respiration by sonicated versus untreated extracts. Sonication was performed in 5 second
bursts x7. Respiratory control ratio (RCR), a sensitive index of mitochondrial integrity, was
5
MS #3695-R2 ONLINE SUPPLEMENT
determined by calculating the ratio of state 4 (200µM ADP) to state 3 respiration (substrate). The
buffer contained 300mM sucrose, 1mM EGTA, 5mM MOPS pH 7.2, 5mM KH2PO4, 0.1% fatty acid
free BSA.
Immunoblotting. Immunoblotting of proteins from isolated mitochondria was performed as
previously described. 8 Monoclonal antibodies against MnSOD and the nuclear encoded ETC subunits
of Complex I (NADH ubiquinol oxidoreductase, 39 KDa subunit) and Complex III (ubiquinol
cytochrome c oxidoreductase, subunit 1) were used (Molecular Probes, Eugene, OR). The supplier's
recommended antibody dilutions were followed.
RT-PCR. Total RNA was isolated from homogenized rat PA and RAs using a QIAGEN RNeasy mini
Kit (Missisauga, ON, Canada) and 2 µgs of RNA was reverse transcribed using QIAGEN Omniscript
reverse transcriptase as described.
3
The primer for MnSOD was designed based on rat cloned
sequences from GeneBank. The PCR product was sequenced and confirmed the specificity of the
product.
Confocal microscopy. Mitochondrial membrane potential (∆ψm) was studied in first passage cultured
PASMC and RASM, loaded with either JC-1 (1µM) or TMRM (20nM) for 30 minutes (37ºC). JC-1 is
a cationic fluorescent dye that exhibits potential-dependent mitochondrial accumulation and is
considered superior to several other dyes used in the measurement of ∆Ψm.
9,10
Within depolarized
mitochondria JC-1 fluoresces in the green spectra, whereas within hyperpolarized mitochondria, the
dye dimerizes and fluoresces in the red spectra. The green/red fluorescence ratio is therefore used to
6
MS #3695-R2 ONLINE SUPPLEMENT
estimate ∆Ψm and thus mitochondrial function. TMRM-loaded mitochondria exhibit stronger red
fluorescence the more hyperpolarized they are. 11
Confocal images were obtained within 30 minutes using an Axiovert 100M inverted confocal
microscope (Carl Zeiss Canada, North York, Ontario). A 488nm Argon laser was used for excitations
and the resultant red and green fluorescence was quantified using LP 560nm and BP 505-530nm filters
respectively. A custom-made perfusion chamber on the stage of the microscope allowed for perfusion
of the cells with normoxic versus hypoxic solutions (see patch-clamping section above). ∆ψm was
compared between PA and RASMC mitochondria at baseline and in response to hypoxia. All the
imaging parameters (including laser power and gains) were kept identical between the two
preparations.
HPLC. A previously published method was used to measure glutathione (GSH) levels on
homogenized normoxic rat lung and kidney. 12
Statistics: Values are expressed as the mean±SEM. Intergroup differences were measured using a
repeated measures ANOVA with post-hoc analysis using Fisher’s PLSD test (Statview 4.02, Abacus
Concepts). A p<0.05 was considered statistically significant. All drugs are from Sigma unless stated
otherwise.
7
MS #3695-R2 ONLINE SUPPLEMENT
Online supplement Results/Discussion
The PAs have more oxidized redox potential compared to the RAs at baseline.
The data presented in figure 3 of the manuscript suggest that the RAs are in a more reduced state than
the PAs at baseline. We confirmed this by comparing glutathione levels between the 2 arteries at
baseline, using HPLC (online suppl. figure 3). Glutathione is a major antioxidant and it is also needed
for the conversion of H2O2 to H2O by catalase. The higher levels of glutathione in the PAs are an
expected compensation with the higher levels of AOS, including H2O2, in this vessel, compared to the
RAs. Similarly, the mitochondria-specific MnSOD dismutates superoxide to H2O2 and it is known to
be induced by high superoxide levels.
13
The higher levels of superoxide in the PA (figure 3A) are in
agreement with our finding that the levels of MnSOD expression is much higher in the PA versus the
RA (online suppl. Figure 3B). The fact that the mRNA levels for MnSOD are slightly lower in the PAs
than the RAs (online suppl. Figure 3C), suggests that MnSOD protein expression might be induced by
the higher levels of superoxide in the PA at a post-transcriptional level.
Online Supplement Figure legends
Figure 1
The respiratory rates are higher in the kidney compared to the lung mitochondria whether glutamate is
used as substrate alone or in combination with succinate.
8
MS #3695-R2 ONLINE SUPPLEMENT
Figure 2
Using the DCF technique we show that in contrast to the rest of the drugs (and their solvents) used,
10µM (but not 1µM) aqueous solution of sodium cyanide has very high auto-fluorescence, preventing
its use in our experiments at this dose.
Figure 3
A. GSH levels are higher in the lung versus the kidney. (†p<0.001).
B. mRNA levels for MnSOD are higher in the RA than the PA.
C. Conversely, MnSOD protein levels (immunoblotting) are much higher in the PA (representative
trace + normalized data).
References
1.
2.
3.
4.
5.
6.
Hampl V, Weir EK, Archer SL. Endothelium-derived nitric oxide is less important for basal
tone regulation in the pulmonary than the renal circulation of the adult rat. J Vasc Med Biol.
1994;5:22-30.
Michelakis E, Weir E, Nelson D, Reeve H, Tolarova S, Archer S. Dexfenfluramine elevates
systemic blood pressure by inhibiting potassium currents in vascular smooth muscle cells. J
Phamacol Exper Ther. 1999;291:1143-9.
Michelakis ED, Weir EK, Wu X, Nsair A, Waite R, Hashimoto K, Puttagunta L, Knaus HG,
Archer SL. Potassium channels regulate tone in rat pulmonary veins. Am J Physiol Lung Cell
Mol Physiol. 2001;280:L1138-47.
Archer SL, Huang J, Henry T, Peterson D, Weir EK. A redox based oxygen sensor in rat
pulmonary vasculature. Circ Res. 1993;73:1100-1112.
Darley-Usmar VM, Rickwood D, Wolson MT. Mitochondria: A practical Approach. Oxford:
IRL Press; 1987.
Nedergaard J, Cannon B. Overview-Preparation and Properties of Mitochondria From Different
Sources. Methods in Enzymology. 1979;55:3-33.
9
MS #3695-R2 ONLINE SUPPLEMENT
7.
8.
9.
10.
11.
12.
13.
Boveris A, Chance B. The mitochondrial generation of hydrogen peroxide. Biochem J.
1973;134:707-716.
Archer SL, Souil E, Dinh-Xuan AT, Schremmer B, Mercier JC, El Yaagoubi A, Nguyen-Huu
L, Reeve HL, Hampl V. Molecular identification of the role of voltage-gated K+ channels,
kv1.5 and kv2.1, in hypoxic pulmonary vasoconstriction and control of resting membrane
potential in rat pulmonary artery myocytes. J Clin Invest. 1998;101:2319-30.
Reers M, Smiley ST, Mottola-Hartshorn C, Chen A, Lin M, Chen LB. Mitochondrial
membrane potential monitored by JC-1 dye. Methods Enzymol. 1995;260:406-17.
Salvioli S, Ardizzoni A, Franceschi C, Cossarizza A. JC-1, but not DiOC6(3) or rhodamine
123, is a reliable fluorescent probe to assess delta psi changes in intact cells: implications for
studies on mitochondrial functionality during apoptosis. FEBS Lett. 1997;411:77-82.
Nicholls DG, Ward MW. Mitochondrial membrane potential and neuronal glutamate
excitotoxicity: mortality and milivolts. Trends Neurosci. 2000;23:166-74.
Lenton KJ, Therriault H, Wagner JR. Analysis of glutathione and glutathione disulfide in whole
cells and mitochondria by postcolumn derivatization high-performance liquid chromatography
with ortho-phthalaldehyde. Anal Biochem. 1999;274:125-30.
Pitkanen S, Robinson BH. Mitochondrial complex I deficiency leads to increased production of
superoxide radicals and induction of superoxide dismutase. J Clin Invest. 1996;98:345-51.
10