Hypoxia and Stretch Regulate Intercellular Communication in Vascular Smooth Muscle Cells Through Reactive Oxygen Species Formation
Objective— We hypothesized that the alterations in vasomotor tone and adaptive remodeling responses that occur in the circulation because of hypoxia were dependent on changes in cell to cell communication through regulation of gap junction protein expression and function. Consequently, we studied the amount, distribution, and permeability of the principal vascular smooth muscle cell (VSMC) gap junction protein, connexin43, in rat aortic cultures exposed to oxygen partial pressures of 150 or 15 mm Hg.
Methods and Results— Immunohistochemical staining, immunoblot assays, and Northern blot analyses demonstrated that connexin43 expression was reversibly increased in hypoxic cultures. As a result, hypoxic cells exhibited greater intercellular communication as determined by fluorescence recovery after photobleaching experiments. Using a fluorogenic substrate, hypoxic VSMCs showed increased reactive oxygen species generation, which could be prevented by the glutathione peroxidase mimic ebselen and the mitochondrial complex I inhibitor rotenone but not with the redox-sensitive thiol pyrrolidine dithiocarbamate. The rise in connexin43 expression attributable to hypoxia could be attenuated by ebselen and rotenone treatment. Interestingly, the previously reported induction of connexin43 expression by tensile stretch was also contingent on oxidative activity.
Conclusions— Hypoxia and stretch increased gap junctional intercellular communication in VSMCs attributable to enhanced connexin43 expression initiated by reactive oxygen species formation.
The coordination and propagation of motor responses in blood vessels are brought about by neurohormonal modulation, ion channel conductance, and intercellular communication.1–3 Although acute vasomotion leads to circulatory homeostasis through the maintenance of local vessel tone, sustained motor responses initiate structural remodeling of the tissue.2,4,5 Because vasomotor and arterial remodeling responses require extensive signaling between the endothelium and medial smooth muscle layers, as well as between adjacent smooth muscle cells, we have been studying cell to cell communication in the vasculature.1,2
See page 1707 and cover
The structure of the arterial circulation can be adjusted in both the circumferential and radial direction, resulting in changes to vessel diameter or wall thickness. For instance, vascular tissue responds to variations in the mechanical forces imposed on it by blood pressure (ie, tensile stretch) and blood flow (ie, shear stress) with rapid adjustments to vasomotor tone and with changes in tissue geometry when alterations in these forces persist.6 These responses are specific for the type of hemodynamic force that is imposed on the vessel wall. Whereas an increase in tensile stretch causes vasoconstriction in the short term and arterial thickening in the long term, an increase in blood flow induces acute vasodilation and chronic enlargement of arterial diameter when the time-averaged shear force is maintained.4
Vascular remodeling can also result from exposure to low levels of environmental oxygen. Chronic hypoxia causes pulmonary vasoconstriction and muscularization of precapillary resistance arterioles in the lung.7 Paradoxically, hypoxia results in vasodilation and structural enlargement of many systemic blood vessels. In addition, remodeling after acute changes in vasomotor tone occurs in response to vessel injury, as seen in postangioplasty restenosis, graft vasculopathy, and hyperlipidemia-induced atherosclerosis.4
Whereas the means by which mechanical forces, hypoxia, and vascular injury are transduced into cellular signals is not fully understood; it seemingly involves the cytoskeleton, cellular junctions, ion channels, transmembrane receptors, second messengers, signaling proteins, and reactive oxygen species (ROS) generation.4,5,8,9 In any case, the transition from vasomotor to structural remodeling responses most likely involves expression of many genes that affect cell proliferation and death as well as connective tissue elaboration and organization.6
Because arterial remodeling attributable to chronic changes in oxygen tension likely involves extensive cell to cell communication, we reasoned that gap junctions would play a significant role in mediating early remodeling responses analogous to those demonstrated for mechanical forces.4 Gap junctions are transmembrane channels that directly link the cytoplasm of neighboring cells, thereby allowing for low-resistance electrical coupling between cells in excitable tissues. The channels are formed when a hemichannel, or connexon, in one cell docks to another in an adjacent cell. Six connexin proteins associate to form the hemichannel, and at least 20 connexin isoforms have been predicted to exist in vertebrates.10,11 Connexons composed of different isoforms exhibit distinct electrophysiological properties with respect to conductance, voltage dependence, and permeability. Although vascular tissues are known to express multiple connexins, connexin43 (Cx43, gap junction protein α1) is the predominant isoform expressed in the smooth muscle layers of major arteries.
Because smooth muscle cells are the primary cell type found in the vessel media and are nominally the final effectors of vasomotor tone and structural remodeling responses,1,12 we tested whether vascular smooth muscle cells (VSMCs) exhibit oxygen tension-sensitive expression of Cx43. We found that hypoxic smooth muscle cell cultures had increased Cx43 mRNA and protein levels that resulted in greater gap junctional intercellular communication. Furthermore, we demonstrated that inhibition of enhanced ROS production during hypoxia prevented the increase in Cx43 expression and cell to cell communication. Intriguingly, the previously reported rise in Cx43 protein induced by mechanical stretch was also found to be mediated through the formation of ROS.6
Animal procedures received institutional approval and were conducted according to the NIH guidelines. Wistar rat thoracic aorta VSMCs were isolated and cultured as described earlier.6 Confluent passage 4 cultures were exposed to a normobaric partial pressure of oxygen (Po2) of 150 or 15 mm Hg for various times in either a standard culture incubator or a hypoxia chamber set to 2.2% oxygen (Coy Laboratory Products). Oxygen levels were reduced in the humidified chamber using 95% N2/5% CO2 replacement and verified by dissolved gas measurements on culture media using a Stat Profile phOx Plus analyzer (Nova Biomedical). Other VSMCs were subjected to mechanical strain (20% static stretch) as previously specified.6,13 Where applicable, cells were subjected to a 10-minute pretreatment with 10 μmol/L pyrrolidine dithiocarbamate (PDTC) (Sigma), 50 μmol/L ebselen (2-phenyl-1,2-benzisoselenazol-3[2H]-one) (Calbiochem), or 10 μmol/L rotenone (Sigma) before treatment with either hypoxia or stretch in the presence of these compounds.
Fixed VSMCs were stained with 10 μg/mL anti-Cx43 monoclonal antibody (Chemicon) and detected with a 1:200 dilution of anti-mouse Alexa 568-conjugated secondary antibody mixed with 4 U/mL Alexa 488-phalloidin (Molecular Probes). Slides were visualized using a BioRad MRC1024 confocal microscope as described previously.14
Immunoblotting was performed as described before.15 Cx43 was detected with 2 monoclonal antibodies to distinguish between total protein and unphosphorylated protein. The Chemicon antibody (MAB3067) was used at a concentration of 1 μg/mL, whereas the Zymed antibody (13-8300) was used at 0.5 μg/mL. Primary antibodies were detected with horseradish peroxidase-labeled anti-mouse secondary antibodies (1:2500) and the ECL kit (Amersham). Densitometric analysis was performed using the public domain Image software developed by the NIH and Scion Corporation. Integrated area measurements were analyzed as described earlier.15
Northern Blot Analysis
Total RNA isolation, electrophoresis, and transfer to nitrocellulose were performed as previously described.6 The rat Cx43 cDNA clone16 and GAPDH cDNA clone (Ambion) were radioactively labeled and hybridized to RNA as detailed earlier.6
Fluorescence Recovery After Photobleaching
VSMCs were grown on bovine plasma fibronectin (Sigma)-coated 25-mm No. 1 glass coverslips and loaded for 10 minutes with 2 μmol/L calcein AM (Molecular Probes) in phenol red-free medium 199 (Gibco).6,17 After rinsing the cells 3 times, coverslips were mounted in a Sykes-Moore chamber (Bellco) and images were digitally acquired with a MicroMAX 1300YHS CCD camera (Princeton Instruments) at a ×100 magnification using a Zeiss Axiovert 35 fluorescent microscope fitted with standard excitation and emission filters and a 75-W xenon arc lamp. Cells in the center of the acquired fields were photobleached for 30 seconds using a ×63 Zeiss Achroplan objective lens (numerical aperture [NA] 0.75; working distance [WD] 0 to 1.5 mm). Images were subsequently obtained at 1-minute intervals with a ×10 Achroplan (NA 0.25) objective lens. Average fluorescence intensity region measurements were made using MetaMorph 6.0 software (Universal Imaging).
Oxidative Activity Measurements
The 6-carboxy-2′,7′-dichlorodihydrofluorescein diacetate, di(acetoxymethyl ester) (H2DCFDA-AM) (Molecular Probes) isomer was loaded into VSMCs at 5 μmol/L in the presence or absence of PDTC, ebselen, and rotenone. Cells were carefully harvested with a rubber policeman and analyzed directly with a Packard Fusion 3.02 fluorescent plate reader or fixed with 4% paraformaldehyde in PBS pH 7.4 and then stained with a 1:200 dilution of an anti-smooth muscle actin-Cy3 conjugate (Sigma) and 300 nmol/L 4′, 6-diamidino-2-phenylindole, dihydrochloride (DAPI) (Molecular Probes) before mounting and viewing with a Zeiss fluorescent microscope.
Cx43 Expression in Hypoxic VSMCs
The effect of hypoxia on cultured VSMC Cx43 protein distribution and quantity was determined by immunohistochemical staining (Figure 1) and immunoblot analysis (please see the online supplement, available at http://atvb.ahajournals. org). Confluent monolayers of cells (A through C) exhibited a large increase in Cx43 staining (D through F) when exposed to hypoxia (B, E, and H). Staining of filamentous actin is depicted in green (A through C, G through I), and Cx43 staining is shown in red (D through F, G through I). Specifically, cells grown continually at a Po2 of 150 mm Hg (A, D, and G) were compared with those cultured at a Po2 of 15 mm Hg for 24 hours (B, E, and H). Hypoxic VSMCs reoxygenated at 150 mm Hg for an additional 24 hours (C, F, and I) had a similar amount of staining as control VSMCs. In similar experiments on cardiomyocyte and cardiac fibroblast cultures, there were no appreciable increases in Cx43 staining in response to hypoxia (not shown). Although the photomicrographs depict a projection of a series of optical sections, analysis of individual sections revealed that much of the observed Cx43 staining seemed to be intracellular, although there was no colocalization of Cx43 with filamentous actin (G through I).
Immunoblots of Cx43 protein levels in response to hypoxia from 0 to 24 hours are shown in Figure I of the online supplement (available at http://atvb.ahajournals.org). Equal protein concentrations were loaded in each lane, and protein quality was assessed by staining identical gels with Coomassie Brilliant Blue R-250 (not shown). The top panel shows total Cx43 protein, whereas the bottom panel shows only the unphosphorylated protein. Hypoxia induced the protein (P<0.001) by 4 hours, and the amount of Cx43 remained elevated for at least 24 hours (compare lane 1 to lanes 2 through 5). As well, there was no readily apparent difference in the ratio of phosphorylated protein to unphosphorylated Cx43 (lanes 1 through 5). These observations were confirmed by densitometry and statistical analyses.15
Cx43 mRNA levels in response to Po2s of 150 mm Hg (lane 1), 15 mm Hg for 24 or 48 hours (lanes 2 and 3), or 15 mm Hg for 24 hours followed by reoxygenation at 150 mm Hg for 4 (lane 4), 8 (lane 5), and 24 (lane 6) hours are depicted in Figure 2A. Cx43 mRNA was considerably increased in the low-oxygen environment (compare lane 1 with lanes 2 and 3) and, on return to a Po2 of 150 mm Hg, the levels returned to baseline by 8 hours. We also found that Cx43 mRNA levels were increased at 1 and 2 hours after the introduction of cultures into the hypoxic environment and that increased expression persisted for at least 48 hours (not shown). Results were compared with GAPDH mRNA levels (B) and total rRNA levels (C), neither of which were altered in response to changing O2 levels.
Intercellular Communication in Hypoxic VSMCs
Because we noticed much of the Cx43 protein in hypoxic VSMCs was located intracellularly rather than on the cell surface, we sought to determine whether the increases in Cx43 expression correlated with enhanced intercellular communication. To assess gap junction function, we used a modified fluorescence recovery after photobleaching (FRAP) technique18,19 on confluent monolayers of cells. VSMCs were loaded with the fluorescein derivative calcein (molecular weight, 622.54), and a subset of these cells was subjected to photobleaching. The ability of calcein to transfer from the unbleached cells to adjacent bleached cells through gap junction channels was determined. The top part of Figure 3 shows fluorescent photomicrographs of representative VSMCs grown at a Po2 of 150 mm Hg continuously (A through C) or 15 mm Hg for 24 hours (D through F). Panels A and D show 2 fields of calcein-stained cells before photobleaching (B and E) and 2 minutes after recovery (C and F). Note that each row of panels (A through C or D through F) depicts the same field of cells. The bottom part of Figure 3 shows a graphical representation of the average fluorescence intensity (n=8) of VSMCs grown at a Po2 of 150 mm Hg (black bars) or 50 mm Hg (white bars). Recovery of fluorescence in the hypoxic cultures was significantly greater (P<0.0001) than that of normoxic VSMCs, with dye transferred between immediately adjacent cells (first-order cells) as well as more distant cells (second-order cells). There was only slightly more recovery after 2 minutes, because most fluorescence was recovered within the first minute (not shown). Control experiments included calcein-loaded, photobleached isolated cells (ie, not connected by gap junctions) and confluent VSMCs loaded with 5 μmol/L CellTracker CM-DiI. Cells lacking neighbors were not able to recover fluorescence, and cells labeled with the membrane tracer DiI demonstrated an inability to transfer fluorescence from one cell to the next (not shown). In addition, photobleached cells were confirmed as viable, because they could repeatedly be loaded with calcein AM and subjected to photobleaching and recovery.
ROS Formation in Hypoxic VSMCs
The effect of hypoxia on VSMC oxidative activity was investigated using the intracellular probe H2DCFDA-AM. Figure II of the online supplement shows photomicrographs depicting fluorescent emission resulting from interaction of oxidized H2DCFDA with reactive oxidant species (green) after exposure to a Po2 of 150 mm Hg (A) or a Po2 of 15 mm Hg for 1 (B), 2 (C), and 4 hours (D). The cells were cross-stained for nuclei with DAPI (blue) and smooth muscle actin using an anti-α-smooth muscle actin-Cy3 antibody (red). Figure 4 graphically depicts DCF fluorescence from VSMCs (n=3) cultured at a Po2 of 150 mm Hg (control) or 15 mm Hg for 1, 2, and 4 hours. Results are also presented for 1-hour hypoxia treatment in the presence of 50 μmol/L ebselen, 10 μmol/L rotenone, or 10 μmol/L PDTC. Increased ROS production was seen at 1 and 2 hours after the onset of hypoxia (P<0.001 for both). The rise in DCF fluorescence in hypoxic cultures could be abolish by treatment with ebselen (P<0.001) and rotenone (P<0.001) but not with PDTC. In fact, PDTC significantly increased ROS generation (P<0.001) in cells subjected to 1 hour of hypoxia.
ROS Inhibition and Cx43 Expression
To ascertain if oxidative activity in VSMCs was associated with gap junction protein α1 expression, we examined Cx43 levels in hypoxic cells treated with specific inhibitors of various ROS by immunoblotting (Figure 5A). Simultaneously, we investigated the ability of these same inhibitors to prevent the previously reported increase in VSMC Cx43 expression in response to mechanical strain (Figure 5B).6 In both instances, the total and unphosphorylated protein levels were examined in cells treated with a Po2 of 15 mm Hg for 6 hours or 20% static stretch for the same length of time (Figured 5A and 5B, top and bottom panels, respectively). Densitometric data were analyzed for statistical significance.15 The expected increase (P<0.001) in Cx43 protein levels for hypoxic and stretched VSMCs was detected (compare lanes 1 and 2 in Figures 5A and 5B, respectively), and this response could be inhibited (P<0.001) by both ebselen and rotenone in either treatment group (lanes 4 and 5). PDTC treatment, on the other hand, had little effect (P>0.05) on Cx43 protein levels in hypoxic or stretched cells (compare lanes 2 and 3 in Figures 5A and 5B). In untreated cells, PDTC seemed to increase unphosphorylated Cx43 protein levels (lane 6); however, this observation was not statistically significant. Alternatively, ebselen and rotenone were both found to depress (P<0.05) baseline Cx43 expression with little effect on state of protein phosphorylation (lanes 7 and 8). FRAP experiments confirmed that ebselen and rotenone (but not PDTC) reduced (P<0.001) cell to cell dye transfer in hypoxic VSMCs (not shown).
ROS Formation in Stretched VSMCs
Because ebselen and rotenone attenuated the induction of Cx43 expression in response to mechanical loading,6 we measured oxidative activity in stretched VSMC cultures using H2DCFDA-AM (Figure 6). There was a significant increase in ROS generation (P>0.001) in stretched cells. The rise in DCF fluorescence attributable to mechanical strain could be reduced to levels lower than seen in controls with ebselen and rotenone treatment. In contrast, PDTC enhanced the production of ROS as a result of VSMC stretch (P<0.01) (compare stretch to stretch plus PDTC). There was no difference in oxidative activity between cells cultured on standard tissue culture dishes versus those cultured on unstretched fibronectin-coated silicone membranes.
We have demonstrated a reversible increase in Cx43 expression and gap junctional intercellular communication in rat thoracic aorta VSMCs cultured in a low oxygen tension environment. Other investigators have shown Cx43 expression can increase (ie, in the carotid body and petrosal ganglion)20 or decrease (ie, in cultured astrocytes)21 in response to hypoxia; however, the underlying mechanism of this regulation remains unknown. Remarkably, by suppressing the hypoxia-induced elevation in ROS formation in smooth muscle cells by either a mitochondrial complex I inhibitor (rotenone) or a glutathione peroxidase mimic (ebselen), the ensuing rise in Cx43 protein levels and gap junction channel permeability could be largely prevented. Although our findings suggest that VSMC intercellular communication is controlled by mitochondrially derived ROS, we recognize other important sources of reactive oxygen intermediates exist in the vasculature (eg, NAD[P]H oxidase) and that these may contribute to regulation of cell to cell communication.
Similarities in the magnitude and timing of Cx43 induction during hypoxia and that reported earlier for tensile stretch6 indicate that these 2 stimuli may share a common mechanism of activation. Comparable to the hypoxic cultures, the increased Cx43 protein levels observed on static mechanical strain of VSMCs was attenuated by treating stretched cells with either rotenone or ebselen. In addition, stretched VSMCs were found to have higher oxidative activity compared with static cultures. Combined, these results indicated that regulation of intercellular communication by hypoxia and mechanical strain was mediated through the amplification of ROS formation in VSMCs.
Examination of individual optical sections from confocal images revealed that much of the Cx43 protein translated in response to hypoxia seemed to be distributed intracellularly rather than localized to the cell surface. The perinuclear pool of Cx43 may have represented proteins localized to the endoplasmic reticulum, Golgi apparatus, and transport vesicles destined for the cell surface or possibly contained within compartments originating from the plasma membrane that could either be recycled to the surface or degraded.22–24 Despite this observation, functional assays (ie, FRAP analyses) revealed that the increase in Cx43 expression was associated with improved intercellular coupling.
For the FRAP experiments, we used a cell-permeant version of the dye calcein, which is colorless and nonfluorescent until it is hydrolyzed by esterases within VSMCs. The intensely fluorescent product, calcein, was small enough to pass through gap junction channels and well-retained in cells because of its charge.25,26 Most of the calcein fluorescence was detected in the region surrounding the VSMC nuclei, presumably because the cytoplasmic volume was the greatest in that region. Dye transfer was apparent between fluorescent donor (unbleached) cells immediately adjacent to nonfluorescent recipient (bleached) cells (first-order transfer) as well as between more distant nonfluorescent cells (second-order transfer) located in the middle of the bleached field and not in direct contact with unbleached cells. We attributed the higher rate of dye transfer in hypoxic cultures to the corresponding increase in Cx43 expression because it was by far the most highly expressed gap junction protein present in these cells1,6,12 and induced by exposure to low oxygen tension.
Because VSMCs have previously been shown to generate reactive oxygen intermediates in response to hypoxia,27–30 we investigated whether oxidant activity may regulate Cx43 expression in a manner similar to that described for various inflammatory proteins such as vascular cellular adhesion molecule-1, monocyte chemoattractant protein-1, and E-selectin.9 Initially, the effect of hypoxia on VSMC oxidative activity was investigated using the nonfluorescent leuco dye H2DCFDA-AM, which can freely enter cells and be oxidized to yield fluorescent dichlorofluoroscein (DCF). The oxidation to the parent dye can occur through interaction with several ROS, including H2O2, hydroxyl radicals, and nitric oxide, but not through interaction with superoxide radicals. The lack of DCF reactivity with superoxide, however, does not preclude the possibility that this radical is generated by hypoxia in VSMCs, because O2− · can be reduced by either the cytosolic copper/zinc or mitochondrial manganese forms of superoxide dismutase to yield H2O2 and molecular oxygen. Consequently, dichlorofluoroscein may be considered a sensitive but not specific indicator of overall cellular oxidant activity.27–30
We have shown that VSMCs formed significantly more reactive oxygen intermediates after 1 and 2 hours at a Po2 of 15 mm Hg compared with cells cultured at a Po2 of 150 mm Hg or for longer periods of time (ie, >4 hours) in the hypoxic environment. The rapid, hypoxia-mediated rise in DCF fluorescence could be almost completely prevented by pretreatment with ebselen and rotenone but not with PDTC. As a matter of fact, PDTC significantly increased DCF emission but had no effect on Cx43 expression. These apparently contradictory effects were likely the result of prooxidant activity at low concentrations (10 μmol/L) rather than the more commonly referenced antioxidant effects seen at higher concentrations (ie, ≥100 μmol/L).31,32 In contrast, stretched aortic myocyte cultures caused production of ROS as measured by DCF fluorescence analogous to that observed for hypoxic VSMCs. Although it has been reported that mechanical stretch of VSMCs enhances ROS production through NAD[P]H oxidase,27–29,33,34 the fact that peroxide production could be prevented with rotenone indicates that mitochondrial ROS generation30 contributed significantly toward regulating gap junctional intercellular communication.
In summary, our findings support the notion that coordinated conduction of vasomotor responses attributable to a change in oxygen concentration or hemodynamic forces is dependent on gap junctional intercellular communication. As a result, electrical impulses could be propagated in both the transverse and longitudinal direction of the vascular wall, thereby orchestrating acute vasomotor and chronic remodeling responses throughout the blood vessel.
This work was supported by the NIH grants HL68915, HL52589, and HL63095.
- Received June 30, 2003.
- Accepted August 21, 2003.
Christ GJ, Spray DC, El-Sabban M, Moore LK, Brink PR. Gap junctions in vascular tissues: evaluating the role of intercellular communication in the modulation of vasomotor tone. Circ Res. 1996; 79: 631–646.
Segal SS. Cell-to-cell communication coordinates blood flow control. Hypertension. 1994; 23: 1113–1120.
Lagaud G, Karicheti V, Knot HJ, Christ GJ, Laher I. Inhibitors of gap junction’s attenuate myogenic tone in cerebral arteries. Am J Physiol. 2002; 283: H2177–H2186.
Davis MJ, Hill MA. Signaling mechanisms underlying the vascular myogenic response. Physiol Rev. 1999; 70: 387–423.
Cowan DB, Lye SJ, Langille BL. Regulation of vascular connexin43 gene expression by mechanical loads. Circ Res. 1998; 82: 786–793.
Weir EK, Hong Z, Porter VA, Reeve HL. Redox signaling in oxygen sensing by vessels. Respir Physiol Neurobiol. 2002; 132: 121–130.
Kunsch C, Medford RM. Oxidative stress as a regulator of gene expression in the vasculature. Circ Res. 1999; 85: 753–766.
Nielsen PA, Beahm DL, Giepmans BN, Baruch A, Hall JE, Kumar NM. Molecular cloning, functional expression, and tissue distribution of a novel human gap junction-forming protein, connexin-31.9: interaction with zona occludens protein-1. J Biol Chem. 2002; 277: 38272–38283.
Wang HZ, Day N, Valcic M, Hsieh K, Serels S, Brink PR, Christ GJ. Intercellular communication in cultured human vascular smooth muscle cells. Am J Physiol. 2001; 281: C75–C88.
Cowan DB, Noria S, Stamm C, Garcia LM, Poutias DN, del Nido PJ, McGowan FX Jr. Lipopolysaccharide internalization activates endotoxin-dependent signal transduction in cardiomyocytes. Circ Res. 2001; 88: 491–498.
Cowan DB, Poutias DN, Del Nido PJ, McGowan FX Jr. CD14-independent activation of cardiomyocyte signal transduction by bacterial endotoxin. Am J Physiol. 2000; 279: H619–H629.
Beyer EC, Paul DL, Goodenough DA. Connexin43: a protein from rat heart homologous to a gap junction protein from liver. J Cell Biol. 1987; 105: 2621–2629.
Wade MH, Trosko JE, Schindler M. A fluorescence photobleaching assay of gap junction-mediated communication between human cells. Science. 1986; 232: 525–528.
Deleze J, Delage B, Hentati-Ksibi O, Verrecchia F, Herve J-C. Fluorescence recovery after photobleaching. In: Bruzzone R, Giaume C. eds. Connexin Methods and Protocols, Vol 154. Totowa, NJ: Humana Press; 2001: 313–327.
Chen J, He B, Dinger L, Stensaas L, Fidone S. Chronic hypoxia upregulates connexin43 expression in rat carotid body and petrosal ganglion. J Appl Physiol. 2002; 92: 1480–1486.
Jordan K, Solan JL, Dominguez M, Sia M, Hand A, Lampe PD, Laird DW. Trafficking, assembly, and function of a connexin43-green fluorescent protein chimera in live mammalian cells. Mol Biol Cell. 1999; 10: 2033–2050.
Falk MM. Connexin-specific distribution within gap junctions revealed in living cells. J Cell Sci. 2000; 113: 4109–4120.
Tomasetto C, Neveu MJ, Daley J, Horan PK, Sager R. Specificity of gap junction communication among human mammary cells and connexin transfectants in culture. J Cell Biol. 1993; 122: 157–167.
Meda P. Assaying the molecular permeability of connexin channels. In: Bruzzone R, Giaume C. eds. Connexin Methods and Protocols, Vol 154. Totowa, NJ: Humana Press; 2001.
Hishikawa K, Oemar BS, Yang Z, Lüscher TF. Pulsatile stretch stimulates superoxide production and activates nuclear factor-κB in human coronary smooth muscle. Circ Res. 1997; 81: 797–803.
Leopold JA, Loscalzo J. Cyclic strain modulates resistance to oxidant stress by increasing G6PDH expression in smooth muscle cells. Am J Physiol. 2000; 279: H2477–H2485.
Nguyen KT, Frye SR, Eskin SG, Patterson C, Runge MS, McIntire LV. Cyclic strain increases protease-activated receptor-1 expression in vascular smooth muscle cells. Hypertension. 2001; 38: 1038–1043.
Waypa GB, Marks JD, Mack MM, Boriboun C, Mungai PT, Schumacker PT. Mitochondrial reactive oxygen species trigger calcium increases during hypoxia in pulmonary arterial myocytes. Circ Res. 2002; 91: 719–726.
Brennan P, O’Neill LA. 2-mercaptoethanol restores the ability of nuclear factor kappa B (NF-κB) to bind DNA in nuclear extracts from interleukin 1-treated cells incubated with pyrrolidine dithiocarbamate (PDTC): evidence for oxidation of glutathione in the mechanism of inhibition of NF-κB by PDTC. Biochem J. 1996; 320: 975–981.
Chung KC, Park JH, Kim CH, Lee HW, Sato N, Uchiyama Y, Ahn YS. Novel biphasic effect of pyrrolidine dithiocarbamate on neuronal cell viability is mediated by the differential regulation of intracellular zinc and copper ion levels, NF-κB, and MAP kinases. J Neurosci Res. 2000; 59: 117–125.
Grote K, Flach I, Luchtefeld M, Akin E, Holland SM, Drexler H, Schieffer B. Mechanical Stretch Enhances mRNA Expression and Proenzyme Release of Matrix Metalloproteinase-2 (MMP-2) via NAD(P)H Oxidase-Derived Reactive Oxygen Species. Circ Res. 2003; 92: e80–e86.