Regulation of Vascular Connexin43 Gene Expression by Mechanical Loads

From The Toronto Hospital Research Institute (B.L.L., D.B.C.), the Samuel Lunenfeld Research Institute (S.J.L.), Mount Sinai Hospital, and the Department of Laboratory Medicine and Pathobiology (D.B.C., B.L.L.), the Department of Physiology (D.B.C., S.J.L.), and the Department of Obstetrics and Gynecology (S.J.L., B.L.L.), the University of Toronto (Canada).

Correspondence to B. Lowell Langille, PhD, CCRW 1–836, The Toronto Hospital (General Division), 200 Elizabeth St, Toronto, Ontario M5G 2C4, Canada. E-mail lowell.langille{at}utoronto.ca

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Abstract—Vascular tissues respond to changes in the mechanical forces imposed on them with changes in vasomotor tone in the short term and with structural remodeling in the long term. Since these responses involve intercellular communication, we have investigated regulation of the gap junction proteins, connexin26 (Cx26), connexin37 (Cx37), connexin40 (Cx40), and connexin43 (Cx43), by mechanical loads. Results were compared with parallel experiments on c-fos and GAPDH. Twenty percent stretch of cultured vascular smooth muscle cells caused a 3-fold increase in Cx43 mRNA levels by 2 hours. Cx26 was expressed at low levels but failed to respond to stretch, and Cx37 and Cx40 were not detected. c-fos mRNA levels increased after 30 minutes of stretch, whereas GAPDH mRNA did not change. Protein levels of Cx43 increased by 4 hours and remained elevated for 16 hours. Nuclear run-on experiments confirmed that Cx43 and c-fos were transcriptionally regulated by stretch. New protein synthesis was not a requirement for the stretch-induced rise in Cx43 expression, since mRNA levels were unaffected by treatment with cycloheximide. To examine transcriptional control of Cx43, stretched and unstretched vascular smooth muscle cells were transfected with a variety of promoter-reporter gene constructs. Cx43 sequences extending from within exon 1 (+162) to -1686 in the 5'-flanking region were coupled to the chloramphenicol acetyl transferase reporter gene. Deletions from the 5' end of these sequences differentially regulated reporter gene expression and indicated multiple potential regulatory sites. In particular, a putative activator protein-1 site at the -42 to -48 region was required for basal reporter activity. None of the promoter constructs revealed stretch sensitivity, indicating that the site of transcriptional control by stretch lies outside the -1686 to +162 region. Finally, Cx43 mRNA levels were assessed in cultured endothelial cells subjected to laminar shear stress of 15 dynes/cm². Cx43 mRNA levels increased by 4-fold at 1 hour and remained elevated for the duration of shear force. In conclusion, both mechanical strain and fluid shear stress caused increased expression of the gap junction protein Cx43.

Key Words: vascular remodeling • gap junction • cell stretch • shear stress

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Vascular tissues respond to changes in the mechanical loads imposed on them by blood pressure (tensile stretch) and blood flow (shear stress) with acute adjustments in vasomotor tone and with structural remodeling when variations in these loads persist. These responses are specific for the type of mechanical loads that are imposed on the vessel wall. For example, increases in tensile stretch prompt the acute vasoconstriction that underlies myogenic autoregulation of peripheral blood flows¹ and the chronic arterial thickening that is associated with hypertension.² In contrast, increases in shear stress induce the acute vasodilation³ and chronic remodeling that result in enlarged resting arterial diameter.⁴

The transition from acute vasomotor to chronic remodeling responses involves the expression of many genes that affect cell proliferation, cell death, and connective tissue elaboration and reorganization. For example, the expression of arterial elastin and collagen is sensitive to mechanical loads,⁵ as is the expression of PDGF,⁶ transforming growth factor-ß1,⁷ fibroblast growth factor,⁸ and many other genes.⁹ We reasoned that tissue remodeling as a result of changes in mechanical load probably involves extensive cell-to-cell communication; therefore, mechanical forces may affect expression and formation of gap junctions.

Gap junctions are intercellular channels that provide a route for small molecules and ions to pass from one cell to the next.¹⁰ A gap junction is formed when a hemichannel, or connexon, in one cell couples to another in an adjacent cell to create an aqueous pore between the two cells. Six protein monomers known as connexins associate to form the connexon, and at least 14 homologous connexin isoforms have been characterized in the rat.¹¹ Connexons composed of these different connexins exhibit distinct electrophysiological properties that permit the regulated flow of electrical current between cells.^{10 12} Currently, three connexins, Cx37, Cx40, and Cx43, have been identified in vascular smooth muscle and endothelium.^{11 13}

Smooth muscle cells, the primary cell type found in the media of many arteries, are largely responsible for effecting remodeling responses. Since there is good evidence that the expression of genes that are important to tissue remodeling is directly sensitive to the stretch of these cells, we tested whether vascular smooth muscle cells grown on a distensible substrate exhibited stretch-sensitive expression of connexins. In contrast, most evidence indicates that remodeling induced by altered blood flow is initiated by the vascular endothelium. These cells signal to underlying smooth muscle cells that effect the tissue elaboration and reorganization that achieves remodeling.¹⁴ Consequently, we tested whether monolayers of cultured vascular endothelial cells grown in a parallel-plate flow chamber exhibited shear-sensitive connexin expression.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Cell Culture
Thoracic aortic smooth muscle cells were isolated from 10-week-old male Wistar rats (Charles River Laboratories, Saint-Constant, Quebec, Canada) by enzymatic dispersion. Once the arteries had been cleared of blood, fat, connective tissue, and adventitia, they were filled with 0.1% collagenase (Cooper Biomedical) and 0.2% trypsin (Difco) in PBS (pH 7.4). After a 40-minute incubation, cells were washed from the lumen of the arteries with a flamed Pasteur pipette. The first wash contained both endothelial cells and smooth muscle cells, whereas the second and third eluates contained only smooth muscle cells. Smooth muscle cells were cultured in medium 199 (GIBCO) with 10% fetal bovine serum (GIBCO), 1% penicillin/streptomycin (GIBCO), and 1% Fungizone (GIBCO) and were passaged at 75% confluence up to four times. Purity of cultures was confirmed by immunohistochemical staining with monoclonal anti–-smooth muscle cell actin antibody (Sigma Chemical Co) at a 1:400 dilution. The primary antibody was detected with a 1:20 dilution of Texas Red–conjugated affinity-purified donkey anti-mouse IgG (heavy and light chain specific) secondary antibody as described by the manufacturer (Jackson ImmunoResearch). Total cell numbers were determined by adding Hoechst 33258 at a final concentration of 1 µg/mL to the secondary antibody cocktail. By passage 2, the cultures were deemed to consist entirely of smooth muscle cells.

Cultures of A10 cells, a fetal rat smooth muscle cell line, were provided by Dr A.I. Gotlieb (The Toronto Hospital Research Institute) and maintained on Falcon 3003 100-mm culture dishes. Porcine thoracic aortas were obtained from a local slaughterhouse, and endothelial cells were purified and maintained as described by Rosenthal and Gotlieb.¹⁵ Plates of neonatal rat cardiomyocytes and adult rat hepatocytes were provided by Dr J. Tsoporis (The Toronto Hospital).

Mechanical Loading
Five cell culture–stretching apparatuses, based on a prototype provided by Dr F. Lyall,¹⁶ were manufactured at the Center for Biomaterials, University of Toronto. The system allows static stretch to be imposed on cells growing on 75-mmx120-mmx0.5-mm sheets of silicone rubber (Altec) by use of a threaded drive. The uniaxial static stretch system was assembled as previously described¹⁶ and sterilized by ethylene oxide gas for 16 hours. Before the addition of 1 to 2x10⁶ vascular smooth muscle cells to each apparatus, the sheets were treated with 0.1 mL of 0.1% bovine plasma fibronectin (Sigma) and 0.1 mL bovine serum. Cells were detached from Falcon culture plates using 0.05% trypsin and 0.53 mmol/L EDTA (GIBCO) and plated onto the unstretched fibronectin-coated silicone sheets mounted in the stretch apparatus. The cells were allowed to proliferate to confluence for 5 days before a static 20% stretch was applied for up to 16 hours. For some experiments, cycloheximide was diluted with culture media to a final concentration of 20 µg/mL and added to cultures before stretch. Cells were collected into 4°C PBS using a Costar cell lifter for subsequent analyses.

Endothelial cells were subjected to laminar fluid shear stress in a parallel-plate flow chamber¹⁷ (manufactured by Strite Industries) that was perfused by gravity feed from a glass reservoir system (Leslie Scientific Glass) exactly as described previously.¹⁷ Porcine endothelial cells at passages 3 to 8 were grown to confluence on untreated autoclaved glass slides (Corning), and a constant shear stress of 15 dynes/cm² was imposed on the cells for up to 24 hours. The cells were collected by scraping with a Costar cell lifter.

Northern Blot Analysis
Total RNA was isolated from smooth muscle cells and endothelial cells,¹⁸ and then electrophoresis, transfer to nitrocellulose membrane (Schleicher & Schuell), cross-linking, and hybridization steps were performed as described by Cowan et al.¹⁹ Equal quantities of total RNA were loaded into each lane before electrophoresis and Northern blot analyses. The rat cDNA clone (G2A) for Cx43 was provided by Beyer et al,²⁰ the rat c-fos cDNA was obtained from Curran et al,²¹ and the pTRI-GAPDH rat cDNA clone was purchased from Ambion Laboratories. The probes for rat Cx26, Cx37, and Cx40 have been previously described.²² Purified insert DNA was labeled using [-³²P]dCTP (ICN Biomedicals) and the NEBlot random primer kit (New England BioLabs). Blots were washed in 0.1% SDS and 0.1x SSC at 22°C for 60 minutes, with changes in wash solution every 5 minutes. Nitrocellulose was exposed to Kodak X-Omat AR film at -80°C for up to 2 weeks. Assays were repeated three or more times.

Western Blot Analysis
Protein extracts were obtained from cultures by rinsing the cells with ice-cold PBS once and cold TBS (pH 7.4) once. Cells were resuspended in cold lysis buffer (150 mmol/L NaCl, 20 mmol/L Tris-HCl [pH 7.5], 1 mmol/L EDTA [pH 8.0], 0.5% sodium deoxycholate, 1% Nonidet P-40, 10 µg/mL aprotinin, 10 µg/mL leupeptin, and 1 µmol/L phenylmethylsulfonyl fluoride), and cellular debris was removed by centrifugation at 12 000g for 10 minutes at 4°C. Supernatants were stored at -80°C. Denaturing discontinuous SDS–polyacrylamide gel electrophoresis was performed on a 10% Laemmli gel essentially as described by Gallagher.²³ Electrophoresis and transfer to nitrocellulose membranes (Schleicher & Schuell) was carried out using a Bio-Rad Mini-Protean II apparatus according to Gallagher.²³ Cx43 protein was detected using a 1:1000 dilution of Cx43 monoclonal antibody (Chemicon) and an enhanced chemiluminescence kit (ECL, Amersham) according to the manufacturer's protocols. The primary antibody was raised against amino acids 252 to 272 in the C-terminal portion of Cx43. A horseradish peroxidase–linked secondary anti-mouse antibody (Amersham) was used at a final dilution of 1:2500, and the film was exposed for 45 minutes. Experiments were replicated three times.

Nuclear Run-on Assays
Pelleted nuclei were isolated according to Andrews and Faller²⁴ and used immediately for transcription synthesis rate analyses as described by Cowan et al.¹⁹ The concentration of nuclei was determined using a model ZF Coulter counter (100-µm orifice, 0.5-mA aperture, 0.707 amplification, and 14 threshold). For transcription, 3.5x10⁷ nuclei were resuspended in 16% glycerol, 20 mmol/L Tris-HCl (pH 8.0), 5 mmol/L MgCl₂, 150 mmol/L KCl, 1 mmol/L ATP, 1 mmol/L CTP, 1 mmol/L GTP (Pharmacia), 100 µCi of [-³²P]UTP (ICN Biomedicals), and 100 U RNAguard (Pharmacia) and incubated at 26°C for 40 minutes. Newly transcribed RNA was isolated as described by Kedzierski and Porter²⁵ and hybridized to 1 µg cold denatured cDNA immobilized on nitrocellulose. Denatured insert cDNA was used for slot-blot fixation according to the instructions provided by Schleicher & Schuell. Prehybridization, hybridization, and wash steps were identical to those described for Northern analysis.¹⁹ Equal amounts of labeled RNA were used for stretched and unstretched samples at each time point. Three independent assays were performed for each time point.

Transient Transfections and Reporter Gene Assays
Cultured smooth muscle cells were plated onto stretch chambers and allowed to attach and proliferate for 5 days. Twenty micrograms of the CAT constructs²⁶ or control (Promega) was combined with 5 µg of RSV–ß galactosidase plasmid and introduced into cells using the CaPO₄ precipitation method.²⁷ After 16 hours, the cells were washed twice with 37°C PBS, fresh medium was added, and the chamber was subjected to a 20% static stretch. Cells were collected 2, 4, and 8 hours after imposition of stretch and assayed for CAT and ß-galactosidase activities.^{28 29} For each CAT assay, equal amounts of cytosolic proteins were analyzed, and the reactions were allowed to proceed for 1 hour. Three independent transfections were performed for 2, 4, and 8 hours.³⁰ The different samples yielded uniform ß-galactosidase activity.

Densitometric and Statistical Analyses
Integrated volumes of unsaturated autoradiograms from Northern and Western blots as well as nuclear run-on assays were determined using a Molecular Dynamics computing densitometer (model 300A) and ImageQuant software (version 3.3). Data have been expressed as fold increase (mean±SD) compared with control data after normalization to matched GAPDH values.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
mRNA Levels in Stretched Vascular Smooth Muscle Cells
The effect of a 20% static stretch of early-passage rat thoracic aorta smooth muscle cells on Cx43, Cx26, c-fos, and GAPDH mRNA levels is shown in Figure 1A. Cx43 message was increased after 30 minutes and 1 hour of stretch, reaching levels that were triple the control values by 2 hours (Figure 1B). Cx43 mRNA then declined to levels found in unstretched cultures after 4 hours (Figure 1). Cx43 message was also detected in the A10 smooth muscle cell line (Figure 1A, lane 1), and Cx43 mRNA levels were slightly elevated in the cultures grown on silicone sheets covered with fibronectin (Figure 1A, lane 3) compared with cells grown on standard culture plates (Figure 1A, lane 2).

View larger version (32K): [in this window] [in a new window]   Figure 1. A, Representative Northern blot analyses of Cx43, Cx26, c-fos, and GAPDH in smooth muscle cells (SMCs) stretched by 20% for various lengths of time. RNA from the A10 SMC line (lane 1) was used for comparison with primary culture SMCs (lanes 2 to 7). Lanes 1 and 2 contain total RNA from cultures not exposed to stretch; lanes 3 to 7 represent RNA derived from stretched SMCs (10 µg per lane). Electrophoresis of samples on a 1.0% formaldehyde-agarose gel, hybridization, and wash steps were performed using standard procedures. B, Fold increases in Cx43 mRNA levels as a result of stretch of SMCs (mean±SD). All densitometric values have been normalized to matched GAPDH measurements and then expressed as a ratio of normalized values to mean mRNA level in unstretched cultures. N indicates the number of cultures.

Cx26 mRNA was detected in both A10 and primary smooth muscle cells; however, levels were relatively low and did not change in response to stretch (Figure 1A, panel 2). Cx37 and Cx40 mRNA were not detected in these cells by Northern analysis (data not shown).

mRNA for c-fos was not detected in unstretched cells; however, high levels of c-fos transcripts were detected after 30 minutes of stretch (Figure 1A, panel 3), and longer exposures of the x-ray film revealed c-fos mRNA after 1 hour of stretch. In additional experiments, even greater induction of c-fos mRNA levels was observed after 15 minutes of stretch of smooth muscle cells (data not shown). These data replicate observations of Lyall et al,¹⁶ who previously reported that stretch induces c-fos expression in smooth muscle, confirming that our system yields results similar to theirs.

GAPDH mRNA levels were relatively constant when primary smooth muscle cells were subjected to static stretch (Figure 1A). Lane 1, bottom panel, shows less GAPDH message than the other lanes because less A10 total RNA was available for analysis because of poor growth of these cells on fibronectin-treated silicone sheets. This underloading of A10 RNA (Figure 1A, lane 1) is, at least in part, responsible for less Cx43 signal being detected (top panel).

Cx43 Protein Levels in Stretched Smooth Muscle Cells
Western blot analysis of Cx43 was performed to determine whether increases in Cx43 mRNA levels due to stretch caused a rise in protein concentration (Figure 2). The mouse anti-Cx43 monoclonal antibody reacted well with rat protein extracts and recognized a single band. Initially, stretch of smooth muscle cells caused no apparent change in Cx43 protein concentration; however, Cx43 levels began to increase after 4 hours of stretch (Figure 2, lane 7). By 6 hours, Cx43 protein concentration was increased by 7-fold (7.01±1.51), and it remained elevated at similar levels for at least 16 hours. Figure 2, lane 1, contains fetal rat cardiomyocyte protein extract (positive control); lane 2 has protein derived from adult rat hepatocyte cultures (negative control). As expected, Cx43 protein was found in the cardiomyocyte extract and not in the hepatocyte lysates.

View larger version (41K): [in this window] [in a new window]   Figure 2. Western blot analysis of Cx43 in stretched smooth muscle cells. Neonatal rat cardiomyocyte culture (C, lane 1) and adult rat hepatocyte culture (H, lane 2) total cellular protein served as positive and negative controls for Cx43 expression. Lanes 3 to 11 contain protein from smooth muscle cells subjected to 20% uniaxial mechanical strain for various periods of time. All lanes contain 20 µg protein immobilized on nitrocellulose membranes and blocked with nonfat milk. Cx43 was detected with a primary monoclonal antibody (Chemicon), a horseradish peroxidase–conjugated secondary anti-mouse IgG antibody, and an enhanced chemiluminescence kit (ECL, Amersham). All wash steps were performed in 1x TBS–Tween 20 (0.1%).

Effect of Smooth Muscle Cell Stretch on Transcriptional Synthesis Rates
Nuclear run-on experiments established that both Cx43 and c-fos are transcriptionally induced at 30 minutes and 1 hour after stretch of vascular smooth muscle cells (compare static versus stretched at each time in Figure 3). Cx43 mRNA synthesis rates 0.5 hour after stretch were approximately tripled (fold increase, 3.15±0.55), and c-fos was also induced at 0.5 hour. After 1 hour of stretch, Cx43 mRNA synthesis was twice the rate found in controls (fold increase, 2.01±0.20), but at 2 and 4 hours after the imposition of stretch, Cx43 and c-fos mRNA synthesis rates had returned to baseline levels. GAPDH showed no significant change in either stretched or unstretched cultures at any time.

View larger version (30K): [in this window] [in a new window]   Figure 3. Nuclear run-on transcription analyses of Cx43, c-fos, and GAPDH in stretched versus unstretched smooth muscle cell cultures. Nuclei were freshly isolated from smooth muscle cells cultured either with or without 20% stretch for various times. Approximately 3.5x107 nuclei were used to generate [-32P]UMP–labeled mRNA as described in "Materials and Methods." Panels 1 to 4 (left to right) are representative films of slot blots containing immobilized "cold" cDNA insert hybridized against 5x106 cpm total labeled nascent RNA. Wash steps were performed in 0.1x SSC and 0.1% SDS as detailed previously.19

Effect of Cycloheximide on Increases in Cx43 mRNA Due to Stretch
To determine whether translation of proteins was needed for induction of Cx43 mRNA transcription as a result of mechanical strain, smooth muscle cells were subjected to cycloheximide treatment at the onset of stretch or in unstretched cultures. After 2 hours, total RNA was collected and assessed for Cx43 and GAPDH mRNA. As anticipated, stretch of smooth muscle cells caused an increase in Cx43 mRNA and no change in GAPDH mRNA compared with no stretch (static cultures) (Figure 4). Cycloheximide treatment resulted in no significant change in the Cx43 stretch response, since essentially identical increases in Cx43 message levels in stretched cells were observed with or without treatment with cycloheximide (fold increases in mRNA levels were 4.77±0.48 and 4.09±0.58, respectively). As well, cycloheximide did not affect GAPDH levels in stretched and unstretched cultures.

View larger version (51K): [in this window] [in a new window]   Figure 4. Northern analysis of vascular smooth muscle cells (VSMCs) treated with 20 µg/mL cycloheximide (cyclohex). Lanes 1 and 2 contain 10 µg total RNA from unstretched VSMCs. Lane 2 was treated with cyclohex for 2 hours. Lanes 3 and 4 contain 10 µg total RNA from cells simultaneously stretched by 20%. Lane 4 was treated with cyclohex for 2 hours. After capillary transfer to nitrocellulose, baking at 80°C in vacuo for 1.5 hours, and prehybridization for 2 hours, membranes were probed with an excess of radiolabeled Cx43 probe. Hybridization and wash steps followed standard procedures.19

Reporter Gene Analyses of the Cx43 5'-Flanking Region in Stretched SMCs
A variety of Cx43 5'-flanking region/CAT reporter gene constructs²⁶ were transfected into smooth muscle cells. Stretched or unstretched cells were assessed for CAT activity at 2, 4, and 8 hours after imposition of stretch (Figure 5). The greatest CAT activity was observed in the cells transfected with pCx1686-CAT (lanes 1 and 9). The pCx75-CAT construct was the only other plasmid that established an appreciable level of enzyme activity (lanes 5 and 13). Interestingly, the 2-bp substitution in the putative AP-1 site (-46 and -47 were modified from GC to TT) that was present in pCx75M-CAT completely abolished activity (lanes 6 and 14), presumably as a result of a cis-acting DNA element residing at this genomic location. Mechanical stretch produced no significant effect on reporter gene activity.

View larger version (48K): [in this window] [in a new window]   Figure 5. Reporter gene analysis of the Cx43 5'-flanking region in unstretched versus stretched smooth muscle cells. The Cx43 genomic structure and reporter gene constructs are shown above a typical autoradiogram. CAT activity was measured 2, 4, and 8 hours after imposition of stretch and has been adjusted for ß-galactosidase activity and protein concentration. There were no significant differences between activity measurements at the three time points after imposition of stretch, and three transfections were performed at each time point. Cx43 5'-flanking region deletion constructs (top to bottom) correspond to lanes 1 to 6 (unstretched cultures) and lanes 9 to 14 (stretched cultures), respectively. Arrows indicate a 2-bp substitution in a putative AP-1 cis-acting sequence located at -46 and -47 with respect to the transcriptional start site (+1). pCAT-control (+) and pCAT-basic (-) vectors (Promega) were used as positive and negative controls (lanes 7 and 8), respectively. Chl. indicates [14C]chloramphenicol; ac.C., monoacetylated forms of chloramphenicol. Restriction enzyme sites, exons 1 and 2, and the 3' untranslated (UTR) region are shown on the genomic schematic diagram at the top.

mRNA Levels in Endothelial Cells Exposed to Shear Stress
Cx43 and GAPDH mRNA levels were determined in endothelial cells exposed to laminar fluid shear stress at 15 dynes/cm² for up to 24 hours (Figure 6). Lane 1 contained total RNA from smooth muscle cells in static cultures. Cx43 and GAPDH mRNA levels were comparable in smooth muscle cell versus endothelial cell cultures. For endothelial cells subjected to shear stress, Cx43 message levels were increased at 1 hour after initiation of fluid flow and remained elevated at 2, 4, 8, and 24 hours of shear (Figure 6, top panel) (fold increases over controls were 3.65±1.10, 4.42±1.43, 4.01±1.53, 3.96±0.38, 2.81±0.74, and 1.15±0.22, respectively). A minor reduction in Cx43 mRNA at 15 minutes of shear and a perceptible drop in signal intensity at 24 hours of shear stress were also observed. In contrast, GAPDH mRNA levels displayed no appreciable shear sensitivity (Figure 6, bottom panel).

View larger version (62K): [in this window] [in a new window]   Figure 6. Northern blot analysis of Cx43 and GAPDH in endothelial cells subjected to shear stress for various times. Total vascular smooth muscle cell (VSMC) RNA (lane 1) was run beside endothelial cell RNA (lanes 2 to 10) for comparison. Lanes 1 and 2 contain RNA from static cultures; lanes 3 to 10 contain RNA from endothelial cells subjected to fluid shear stress at 15 dynes/cm2 in a parallel-plate flow chamber for up to 24 hours. Electrophoresis, transfer, immobilization, prehybridization, probe preparation, hybridization, and wash steps were performed according to standard procedures.19 This figure is representative of three independent experiments.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
Connexins are expressed very early in development, and their expression is both tissue specific and developmentally regulated. These observations have led to the inference that cell-cell communication via gap junctions is important in the morphogenesis and developmental remodeling of many tissues.^{31 32 33 34} In the cardiovascular system, developmental remodeling of the heart depends on expression of specific connexins; mice with null mutations of Cx43, the primary gap junctional protein expressed by myocardium, survive through embryonic and fetal development with apparently quite normal cardiac contractile activity, presumably because upregulation of alternate gap junctional proteins permits coordinated contraction of the myocardium, but they die hours after birth as a result of a severe obstructive defect of the right ventricular outflow tract.³⁵ Expression of several connexins by vascular tissues has been described, and resulting gap junction formation apparently allows coordinated contractile activity in the blood vessel wall^{36 37} ; however, the role of gap junctions in arterial remodeling has not been investigated. Since communication among vascular cells is a prerequisite for coordinated tissue remodeling and because changes in contractile activity are precursors of the structural remodeling of the arterial wall that results from chronically altered blood pressure and blood flow rates,³⁸ we hypothesized that mechanical forces affect expression of gap junction proteins.

In order to isolate the mechanical loads imposed by blood pressure and blood flow, we used two in vitro devices that subject cell cultures to either uniform stretch or laminar fluid shear stress.^{16 17} We chose to examine the effects of steady rather than pulsatile forces because we are interested in arterial remodeling in response to changes in physical forces, which appears to depend on time-averaged forces rather than on the pulsatile components of these forces. For example, flow-induced remodeling of rabbit carotid arteries correlated strongly with changes in mean flow rate but was not influenced by selective changes in the pulsatile components of flow.⁴ Also, arterial thickening in response to hypertension appears to normalize the mean tensile stress imposed on the arterial wall.²

We imposed static stretch on smooth muscle cell cultures grown on a silicone substrate. This is an artificial environment for cells that normally grow in a three-dimensional matrix; however, the model has several positive features. First, cultures grown under similar conditions exhibited stretch that was proportional to that of the substrate, and they were not injured by stretch.^{39 40} Second, a 20% strain is within the range expected for physiological changes in mean blood pressure.⁴¹ Third, unlike commercially available stretch apparatuses,⁴² the device used for the present study delivered a uniform stretch to all smooth muscle cells grown on the silicone.¹⁶ The parallel-plate flow chamber that we used to impose shear stress on endothelium has been extensively characterized,^{9 17 43 44 45 46} and we used it to generate a level of shear stress (15 dynes/cm²) that is physiological for large mammalian arteries.⁴⁷

We have demonstrated dramatic induction of Cx43 mRNA in smooth muscle cells exposed to stretch and in endothelial cells exposed to shear stress. In the cell stretch model,¹⁶ mechanical stimulation of Cx43 protein levels was preceded by increases in mRNA. Furthermore, nuclear run-on assays indicated that Cx43 expression was controlled at the level of transcription. Although Cx26 was also expressed in smooth muscle cells, expression was not affected by mechanical forces. Cx37 and Cx40 were not detectable, possibly because of the low level of expression of these isoforms in cultured aortic smooth muscle cells.^{37 48 49 50 51} In previous reports, Cx40 was detected in arteriolar smooth muscle cells using immunofluorescence,⁵² whereas others reported expression of Cx40 in endothelium but not in smooth muscle by using similar techniques.^{48 49} Cx37 expression has been detected in endothelium but not in smooth muscle.^{53 54 55}

The transcriptional induction of Cx43 occurred rapidly after imposition of stretch, with RNA synthesis rates, detected by nuclear run-on, increasing at 30 minutes and 1 hour. Message levels increased by 1 and 2 hours, and elevation of protein levels began at 4 hours and persisted for at least 16 hours. The waning of mRNA levels for Cx43 after 4 hours (Figure 1A, top panel, lane 7) may reflect a transient response to sustained stretch; however, it may occur because the smooth muscle cells reorganize substrate adhesion complexes so that the cells become off-loaded with time despite persistent stretch of substrate. The prolonged increase in Cx43 protein levels despite an early decline in mRNA level indicates that stretch may stabilize the protein in addition to enhancing transcription from the gene, since usual estimates of protein half-life range from 1.5 to 5 hours.^{56 57}

To characterize regulation of the Cx43 gene in vascular cells, we performed transfection experiments to localize cis-acting DNA sequences. The -1686 to +162 region of the Cx43 gene and constructs created from this fragment by deletion were fused to the CAT reporter gene²⁶ so that variations in activity of CAT would indicate potential sites of transcriptional regulation in this 5'-flanking region. We found that only the -1686 to +162 and -75 to +162 regions of the Cx43 promoter drove substantial activity of the CAT reporter in vascular smooth muscle cells. These findings indicate that positive control sequences are located between -368 and -1686 and within 75 bases from the transcriptional start site (Figure 5). Substitution of 2 bp at -46 and -47 (GC to TT) abolished reporter gene activity, indicating that a positive control region is located at this putative AP-1 site.²⁶ Interestingly, Geimonen et al⁵⁸ have found an analogous AP-1 site in the human Cx43 gene. Induction of Cx43 in uterine smooth muscle cells by activation of protein kinase C is regulated through this AP-1 site.⁵⁸ Also noteworthy is our finding that the addition of 29 bp of promoter to the –75 construct (pCx104-CAT) can completely prevent CAT activity, indicating that a negative cis-acting element(s) resides at this location.

Although transcription could be driven by the -1686 to +162 region of the Cx43 gene, this expression lacked any sensitivity to stretch. It is unlikely that we missed a rise in CAT activity due to the transient nature of the Cx43 stretch response, because activity measurements were made at 2, 4, and 8 hours after imposing stretch. Because CAT is a relatively stable protein, any differences between stretched and unstretched cultures as a result of a rise in reporter gene expression for a period of 1 or 2 hours (as seen for Cx43 mRNA) would be apparent at one of these times. We infer that transcriptional control due to stretch-sensitive mechanisms is conveyed by DNA sequences outside this 5'-flanking region, although we cannot rule out cooperative interactions with sites within the -1686 to +162 region.

The rise in Cx43 gene expression as a result of stretch was independent of the translation of other proteins, since treatment of smooth muscle cells with the ribosomal blocker, cycloheximide, during imposition of mechanical load had no effect on the increase in Cx43 mRNA. This result precludes involvement of some immediate-early gene products such as c-fos in the transcriptional control of Cx43 by stretch, since it is generally accepted that c-fos protein is synthesized de novo on stimulation.²¹ Similar to Cx43 expression, the transcriptional stretch response of the c-fos gene occurred without the need for de novo protein synthesis⁵⁹ and followed essentially the time profile observed in all studies to date.^{16 39 59} Presumably, a stretch-responsive transcription factor or complex was already present in the cell, poised for activation or import to the nucleus.

Initially, it seemed reasonable that c-fos may be controlling the Cx43 stretch response. This gene is mechanically responsive,^{16 39 59} and its expression increases before the rise in Cx43 mRNA.⁵⁸ Moreover, sequences resembling an AP-1 binding site were present in the Cx43 proximal promoter region (-49 to -42), and c-fos localized to the nucleus of cardiac cells subjected to stretch.³⁹ Although the putative AP-1 site in the Cx43 proximal promoter revealed transcriptional activity that could be abolished with a 2-bp substitution, this site was not stretch sensitive. This finding also makes involvement of c-fos in the Cx43 stretch response improbable, unless it functions in concert with sites outside the -1686 to +162 region.

We have not examined the mechanism of Cx43 regulation by shear stress; however, shear and stretch may both impose the same mechanical loads (tension) on subcellular structures⁹ ; therefore, a common signaling pathway may drive stretch-induced Cx43 expression in smooth muscle and shear stress–induced increases in Cx43 mRNA levels in endothelial cells. A candidate target protein for activation in a mechanically sensitive signal cascade is the transcription factor, nuclear factor-B, which binds to a DNA sequence known as the shear stress–responsive element, which drives both shear- and stretch-induced expression of the PDGF-B chain in vascular endothelium.⁶⁰ This sequence, however, is not found in the immediate 5'-flanking region of the Cx43 gene.^{26 58} Furthermore, multiple cis elements that mediate transcriptional regulation by shear stress have now been identified.^{61 62}

The robust nature of the Cx43 response to both stretch and shear implies that this gap junction protein is important in communicating alterations in mechanical environment in both endothelial cells and smooth muscle cells, and it may impart functional changes to these cells. Several investigators have proposed that changes in gap junctional activity contribute to the increased vascular reactivity that is observed in hypertension.^{48 63 64} Haefliger et al⁶⁴ recently showed that aortic Cx43 levels were increased during the early phases of deoxycorticosterone acetate–salt and two-kidney/one-clip hypertension in rats. Our data indicate that this increased Cx43 expression may be a direct result of increased arterial distension associated with early hypertension. A resulting increase in vascular reactivity and vasomotor tone could contribute to the increased total peripheral resistance that characterizes most forms of hypertension. There is also preliminary evidence that vasomotion secondary to altered shear stress can take 24 hours to go to completion.⁴ Over this time scale, modulation of connexin expression and gap junction activity could participate in adjusting vasomotor tone.

Finally, there is good evidence that gap junctions are important in local vasomotor controls in the microvasculature. Localized vasomotor stimuli elicit responses that propagate along microvessels to sites far distant from the initial stimulus,⁶⁵ and inhibitors of gap junctional communication block this propagation.⁶⁶ Enhanced coupling of cells via gap junctions in regions of the microvasculature that experience persistent increases in blood flow or persistent increases in local intravascular pressures may amplify this signal propagation and thereby influence vasomotor reactivity at sites of increased metabolic activity.

In summary, we have found that expression of the gap junctional protein, Cx43, is sensitive to mechanical forces in vascular smooth muscle and endothelium. Sensitivity of expression is controlled at the transcriptional level and does not depend on the early expression of other proteins. Using CAT reporter constructs, we detected potential sites of transcriptional regulation in the region extending from -1686 to +162 of the 5'-flanking region of the Cx43 gene; however, these sites were not sensitive to stretch, and we infer that transcriptional responses to stretch depend on sites outside this region.

	Selected Abbreviations and Acronyms

 

AP-1 = activator protein-1 CAT = chloramphenicol acetyl transferase Cx (with number) = connexin PDGF = platelet-derived growth factor TBS = Tris-buffered saline

	Acknowledgments

 
This study was supported by grant GR-13299 from the Medical Research Council of Canada. Dr Cowan was the recipient of a Medical Research Council of Canada Fellowship, and Dr Langille is a Career Investigator with the Heart and Stroke Foundation of Ontario. The donation of cell culture materials by Dr A.I. Gotlieb and Dr J. Tsoporis is gratefully acknowledged.

Received September 23, 1997; accepted January 30, 1998.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Jackson PA, Duling BR. Myogenic response and wall mechanics of arterioles. Am J Physiol. 1989;257:H1147–H1155.[Abstract/Free Full Text]
2. Folkow B. Hypertensive structural changes in systemic precapillary resistance vessels: how important are they for in vivo haemodynamics. J Hypertens. 1995;13:1546–1559.[Medline] [Order article via Infotrieve]
3. Busse R, Pohl U. Chronic effects of blood flow on the artery wall. In: Frangos JA, ed. Physical Forces and the Mammalian Cell. San Diego, Calif: Academic Press Inc; 1993:223–248.
4. Langille BL. Remodelling of developing and mature arteries: endothelium, smooth muscle and matrix. J Cardiovasc Pharmacol. 1993;21:S11–S17.
5. Sutcliffe MC, Davidson JM. Effect of static stretching on elastin production by porcine aortic smooth muscle cells. Matrix. 1990;10:148–153.[Medline] [Order article via Infotrieve]
6. Hsieh H-J, Li N-Q, Frangos JA. Shear stress increases endothelial platelet-derived growth factor mRNA levels. Am J Physiol. 1991;260:H642–H646.[Abstract/Free Full Text]
7. Ohno M, Cooke JP, Dzau VJ, Gibbons GH. Fluid shear stress induces endothelial transforming growth factor beta-1 transcription and production: modulation by potassium channel blockade. J Clin Invest. 1995;95:1363–1369.
8. Malek AM, Gibbons GH, Dzau VJ, Izumo S. Fluid shear stress differentially modulates expression of genes encoding basic fibroblast growth factor and platelet-derived growth factor B chain in vascular endothelium. J Clin Invest. 1993;92:2013–2021.
9. Davies PF. Flow-mediated endothelial mechanotransduction. Physiol Rev. 1995;75:519–560.[Abstract/Free Full Text]
10. De Leon JR, Buttrick PM, Fishman GI. Functional analysis of the connexin43 gene promoter in vivo and in vitro. J Mol Cell Cardiol. 1994;26:379–389.[Medline] [Order article via Infotrieve]
11. Polacek D, Bech F, McKinsey JF, Davies PF. Connexin 43 gene expression in the rabbit arterial wall: effects of hypercholesterolemia, balloon injury, and their combination. J Vasc Res. 1997;34:19–30.[Medline] [Order article via Infotrieve]
12. Angst BD, Khan LUR, Severs NJ, Whitely K, Rothery S, Thompson RP, Magee AI, Gourdie RG. Dissociated spatial patterning of gap junctions and cell adhesion junctions during postnatal differentiation of ventricular myocardium. Circ Res. 1997;80:88–94.[Abstract/Free Full Text]
13. Little TL, Xia J, Duling BR. Dye tracers define differential endothelial and smooth muscle coupling patterns within the arteriolar wall. Circ Res. 1995;76:498–504.[Abstract/Free Full Text]
14. Langille BL, O'Donnell F. Reductions in arterial diameter produced by chronic decreases in blood flow are endothelium-dependent. Science. 1986;231:405–407.[Abstract/Free Full Text]
15. Rosenthal A, Gotlieb AI. Macrovascular endothelial cells from porcine aorta. In: Piper HM, ed. Cell Culture Techniques in Heart and Vessel Research. Berlin, Germany: Springer-Verlag; 1990:117–129.
16. Lyall F, Deehan MR, Greer IA, Boswell F, Brown WC, McInnes GT. Mechanical stretch increases proto-oncogene expression and phosphoinositide turnover in vascular smooth muscle cells. J Hypertens. 1994;12:1139–1145.[Medline] [Order article via Infotrieve]
17. Frangos JA, McIntire LV, Eskin SG. Shear stress induced stimulation of mammalian cell metabolism. Biotech Bioeng. 1988;32:1053–1060.
18. Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem. 1987;162:156–159.[Medline] [Order article via Infotrieve]
19. Cowan DB, Weisel RD, Williams WG, Mickle DAG. The regulation of glutathione peroxidase gene expression by oxygen tension in cultured human cardiomyocytes. J Mol Cell Cardiol. 1992;24:423–433.[Medline] [Order article via Infotrieve]
20. Beyer EC, Paul DL, Goodenough J. Connexin 43: a protein from rat heart homologous to a gap junction protein from liver. J Cell Biol. 1987;105:2621–2629.[Abstract/Free Full Text]
21. Curran T, Gordon MB, Rubino KL, Sambucetti C. Isolation and characterization of the c-fos (rat) cDNA and analysis of post-translational modification in vitro. Oncogene. 1987;2:79–84.[Medline] [Order article via Infotrieve]
22. Orsino A, Taylor CV, Lye SJ. Connexin 26 and connexin 43 are differentially expressed and regulated in the rat myometrium throughout late pregnancy and with the onset of labor. Endocrinology. 1996;137:1545–1553.[Abstract]
23. Gallagher SR. Analysis of proteins. In: Ausubel SM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA, Struhl K, eds. Short Protocols in Molecular Biology. New York, NY: John Wiley & Sons; 1995:10-5–10-26.
24. Andrews NC, Faller DV. A rapid micropreparation technique for extraction of DNA-binding proteins from limiting numbers of mammalian cells. Nucleic Acids Res. 1991;19:2499.[Free Full Text]
25. Kedzierski W, Porter JC. A novel non-enzymatic procedure for removing DNA template from RNA transcription mixtures. Biotechniques. 1991;10:210–214.[Medline] [Order article via Infotrieve]
26. Chen ZQ, Lefebvre D, Bai XH, Reaume A, Rossant J, Lye SJ. Identification of two regulatory elements within the promoter region of the mouse connexin 43 gene. J Biol Chem. 1995;270:3863–3868.[Abstract/Free Full Text]
27. Kingston RE. Introduction of DNA into mammalian cells. In: Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA, Struhl K, eds. Current Protocols in Molecular Biology. New York, NY: John Wiley & Sons; 1988:9-1-1–9-1-4.
28. Gorman GM, Moffat LF, Howard BH. Recombinant genomes which express chloramphenicol acetyl transferase in mammalian cells. Mol Cell Biol. 1982;2:1044–1051.[Abstract/Free Full Text]
29. Rosenthal N. Identification of regulatory elements of cloned genes with functional assays. Methods Enzymol. 1987;152:704–720.[Medline] [Order article via Infotrieve]
30. Cowan DB, Weisel RD, Williams WG, Mickle DAG. Identification of oxygen responsive elements in the 5'-flanking region of the human glutathione peroxidase gene. J Biol Chem. 1993;268:26904–26910.[Abstract/Free Full Text]
31. Green CR, Bowles L, Crawley A, Tickle C. Expression of the connexin43 gap junctional protein in tissues at the tip of the chick limb bud is related to the epithelial-mesenchymal interactions that mediate morphogenesis. Dev Biol. 1994;161:12–21.[Medline] [Order article via Infotrieve]
32. Nagajski DJ, Guthrie SC, Ford CC, Warner AE. The correlation between patterns of dye transfer through gap junctions and future developmental fate in Xenopus: the consequences of u.v. irradiation and lithium treatment. Development. 1989;105:747–752.[Abstract/Free Full Text]
33. Munari-Silem Y, Guerrier A, Fromaget C, Rabilloud R, Gros D, Rousset B. Differential control of connexin-32 and connexin-43 expression in thyroid epithelial cells: evidence for a direct relationship between connexin-32 expression and histiotypic morphogenesis. Endocrinology. 1994;135:724–734.[Abstract]
34. Lo CW. The role of gap junction membrane channels in development. J Bioenerg Biomembr. 1996;28:379–385.[Medline] [Order article via Infotrieve]
35. Reaume A, de Sousa PA, Kulkarni S, Langille BL, Zhu D, Kidder GM, Rossant J. Cardiac malformation in neonatal mice lacking connexin 43. Science. 1995;267:1831–1834.[Abstract/Free Full Text]
36. Segal SS. Cell-to-cell communication coordinates blood flow control. Hypertension. 1994;23:1113–1120.[Abstract/Free Full Text]
37. 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.[Abstract/Free Full Text]
38. Cowan DB, Langille BL. Cellular and molecular biology of vascular remodeling. Curr Opin Lipidol. 1996;7:94–100.[Medline] [Order article via Infotrieve]
39. Sadoshima JI, Jahn L, Takahashi T, Kulik TJ, Izumo S. Molecular characterization of stretch-induced adaptation of cultured cardiac cells: an in vitro model of load-induced cardiac hypertrophy. J Biol Chem. 1992;267:10551–10560.[Abstract/Free Full Text]
40. Wilson E, Sudhir K, Ives HE. Mechanical strain of rat vascular smooth muscle cells is sensed by specific extracellular matrix/integrin interactions. J Clin Invest. 1995;96:2364–2372.
41. Arndt JO, Stegall HF, Wicke HJ. Mechanics of the aorta in vivo: a radiographic approach. Circ Res. 1971;28:693–704.[Abstract/Free Full Text]
42. Vandenburgh HH. Mechanical forces and their second messengers in stimulating cell growth in vitro. Am J Physiol. 1992;262:R350–R355.[Abstract/Free Full Text]
43. Diamond SL, Eskin SG, McIntire LV. Fluid flow stimulates tissue plasminogen activator secretion by cultured human endothelial cells. Science. 1989;243:1483–1485.[Abstract/Free Full Text]
44. Malek AM, Izumo S. Molecular aspects of signal transduction of shear stress in the endothelial cell. J Hypertens. 1994;12:989–999.[Medline] [Order article via Infotrieve]
45. Shyy JY-J, Lin M-C, Han J, Lu Y, Petrime M, Chien S. The cis-acting phorbol ester `12-O-tetradecanoylphorbol 13-acetate'-responsive element is involved in shear stress-induced monocyte chemotactic protein 1 gene expression. Proc Natl Acad Sci U S A.. 1995;92:8069–8073.[Abstract/Free Full Text]
46. Resnick N, Gimbrone MA Jr. Hemodynamic forces are complex regulators of endothelial gene expression. FASEB J. 1995;9:874–882.[Abstract]
47. Langille BL, Gotlieb AI, Kim DW. Vascular tissue response to experimentally altered local blood flow conditions. In: Westerhof N, Gross DR, eds. Vascular Dynamics. New York, NY: Plenum Publishing Corp; 1989:229–235.
48. Bastide B, Neyses L, Ganten D, Paul M, Willecke K, Traub O. Gap junction protein connexin40 is preferentially expressed in vascular endothelium and conductive bundles of rat myocardium and is increased under hypertensive conditions. Circ Res. 1993;73:1138–1149.[Abstract/Free Full Text]
49. Bruzzone R, Haefliger J-A, Gimlich RL, Paul DL. Connexin40, a component of gap junctions in vascular endothelium, is restricted in its ability to interact with other connexins. Mol Biol Cell. 1993;4:7–20.[Abstract]
50. Gourdie RG, Green CR, Severs NJ, Anderson RH, Thompson RP. Evidence for a distinct gap-junctional phenotype in ventricular conduction tissues of the developing and mature avian heart. Circ Res. 1993;72:278–289.[Abstract/Free Full Text]
51. Van Rijen HVM, Van Kempen MJA, Analbers LJS, Rook MB, Van Ginneken ACG, Gros D, Jongsma HJ. Gap junctions in human umbilical cord endothelial cells contain multiple connexins. Am J Physiol. 1997;272:C117–C130.[Abstract/Free Full Text]
52. Little TL, Beyer EC, Duling BR. Connexin 43 and connexin 40 gap junctional proteins are present in arteriolar smooth muscle and endothelium in vivo. Am J Physiol. 1995;268:H729–H739.[Abstract/Free Full Text]
53. Reed KE, Westphale EM, Larson DM, Wang H-Z, Veenstra RD, Beyer EC. Molecular cloning and functional expression of human connexin37, an endothelial cell gap junction protein. J Clin Invest. 1993;91:997–1004.
54. Willecke K, Haubrich S. Connexin expression system: to what extent do they reflect the situation in the animal? J Bioenerg Biomembr. 1996;28:319–326.[Medline] [Order article via Infotrieve]
55. Larson DM, Wrobleski MJ, Sagar GDV, Westphale EM, Beyer EC. Differential regulation of connexin43 and connexin37 in endothelial cells by cell density, growth, and TGF-ß1. Am J Physiol. 1997;272:C405–C415.[Abstract/Free Full Text]
56. Darrow BJ, Laing JG, Lampe PD, Saffitz JE, Beyer EC. Expression of multiple connexins in cultured neonatal rat ventricular myocytes. Circ Res. 1995;76:381–387.[Abstract/Free Full Text]
57. Laird D. The life cycle of a connexin: gap junction formation, removal, and degradation. J Bioenerg Biomembr. 1996;28:311–318.[Medline] [Order article via Infotrieve]
58. Geimonen E, Jiang W, Ali M, Fishman GI, Garfield RE, Anderson J. Activation of protein kinase C in human uterine smooth muscle induces connexin-43 gene transcription through an AP-1 site in the promoter sequence. J Biol Chem. 1996;271:23667–23674.[Abstract/Free Full Text]
59. Komuro I, Kaida T, Shibazaki Y, Kurabayashi M, Katoh Y, Hoh E, Takaku F, Yazaki Y. Stretching cardiac myocytes stimulates protooncogene expression. J Biol Chem. 1990;265:3595–3598.[Abstract/Free Full Text]
60. Khachigian LM, Resnick N, Gimbrone MA Jr, Collins T. Nuclear factor-kappaB interacts functionally with the platelet-derived growth factor B-chain shear-stress response element in vascular endothelial cells exposed to fluid shear stress. J Clin Invest. 1995;96:1169–1175.
61. Shyy JY-J, Li YS, Lin MC, Chen W, Yuan SL, Usami S, Chien S. Multiple cis-elements mediate shear stress-induced gene expression. J Biomech. 1995;28:1451–1457.[Medline] [Order article via Infotrieve]
62. Malek AM, Izumo S. Control of endothelial cell gene expression by flow. J Biomech. 1995;28:1515–1519.[Medline] [Order article via Infotrieve]
63. Tsai ML, Watts SM, Loch-Caruso R, Webb RC. The role of gap junctional communication in contractile oscillations in arteries from normotensive and hypertensive rats. J Hypertens. 1995;13:1123–1133.[Medline] [Order article via Infotrieve]
64. Haefliger J-A, Castillo E, Waeber G, Bergonzelli BE, Aubert J-F, Sutter E, Nicod P, Waeber B, Meda M. Hypertension increases connexin43 in a tissue-specific manner. Circulation. 1997;95:1007–1014.[Abstract/Free Full Text]
65. Segal SS, Damon DN, Duling BR. Propagation of vasomotor responses coordinates arteriolar resistances. Am J Physiol. 1989;256:H832–H837.[Abstract/Free Full Text]
66. Frame MDS, Sarelius IH. Endothelial cell dilatory pathways link flow and wall shear stress in an intact arteriolar network. J Appl Physiol. 1996;81:2105–2114.[Abstract/Free Full Text]