Tenascin-C (TN-C) expression and matrix metalloproteinase (MMP) activity are induced within remodeling pulmonary arteries (PAs), where they promote cell growth. Because pulmonary vascular disease in children with congenital heart defects is commonly associated with changes in pulmonary hemodynamics, we hypothesized that changes in pulmonary blood flow regulate TN-C and MMPs. To test this, we ligated the left PAs of neonatal pigs. After 12 wk, we evaluated the levels of TN-C and MMPs in control and ligated lung tissue. Modifying pulmonary hemodynamics increased TN-C mRNA and protein expression, MMP activity, and the DNA-binding activity of Egr-1, a transcription factor that has been shown to activate TN-C expression. To link MMP-mediated remodeling of the extracellular matrix to increased TN-C expression and Egr-1 activity, porcine PA smooth muscle cells were cultivated either on denatured type I collagen, which supported TN-C expression and Egr-1 activity, or on native collagen, which had the opposite effect. These data provide a framework for understanding how changes in pulmonary blood flow in the neonate modify the tissue microenvironment and cell behavior.
- pulmonary vascular disease
altered pulmonary blood flow due to left-to-right shunts commonly occurs in children with congenital heart defects (21). If surgical repair is not performed at an early age, this may lead to irreversible pulmonary vascular disease, characterized by remodeling of the arterial extracellular matrix (ECM), and the onset of vascular smooth muscle cell (SMC) growth and migration. An examination of graded lung biopsy specimens isolated from children with congenital heart defects shows that the induction of the ECM glycoprotein tenascin-C (TN-C) accompanies progressive pulmonary vascular changes (12). This finding is significant because TN-C colocalizes with proliferating SMC in vivo and actively supports their growth in tissue culture (5-7, 13, 15). Furthermore, treatment of vascular SMC cultures with TN-C antisense oligonucleotides (5) and gene knockout studies (7) using antisense-ribozyme constructs expressed in TN-C-producing adult rat hypertrophied pulmonary arteries (PAs) demonstrates that ablation of TN-C gene expression limits SMC growth, survival, and the severity of pulmonary vascular lesions.
As well as increased TN-C biosynthesis, activation of matrix metalloproteinases (MMPs) represents another hallmark of pulmonary and other vascular diseases (1-3, 8, 10, 18, 19, 27, 30). That MMPs play a key role in the development of vascular pathobiology has been demonstrated both in tissue culture and in vivo experiments using synthetic and endogenous MMP inhibitors, which prevent or attenuate SMC migration, ECM biosynthesis, vascular hypertrophy, and neointimal formation (1-3, 8, 10, 18, 19, 27, 30). Of particular relevance to the present study is the finding that inhibition of MMP activity in cultured SMCs and PAs suppresses TN-C gene expression, whereas the active forms of these proteinases promote TN-C expression (6, 7, 13). In light of these and other studies showing that TN-C and MMPs control the course of vascular disease, numerous investigations have been made to identify the extrinsic and intrinsic factors that control their expression.
In addition to soluble growth factors and cytokines, biomechanical factors have recently been shown to regulate TN-C and MMPs. For example, systemic-like hemodynamics enhances monocrotaline-induced pulmonary vascular remodeling, which is characterized in its early stages by increased expression of TN-C and increased activity of MMP-2 and MMP-9 (6, 7, 11, 26). Similar to TN-C, MMPs are also induced and activated in models of human vein graft stenosis and after physical injury of arteries (1-3, 8, 10, 19). Furthermore, our previous tissue culture studies (6, 13) show that mechanically stress unloading rat PA SMCs and intact rat PAs maintained on type I collagen gels not only suppresses the expression and gelatinolytic activity of MMP-2 and MMP-9, but also the expression of TN-C.
Despite this knowledge, remarkably little is known about the way in which changes in pulmonary hemodynamics control TN-C and MMPs. To address this, we analyzed TN-C expression and MMP activity using an established in vivo model of altered pulmonary blood flow in which the left PA of neonatal pigs is surgically ligated. In addition, we determined whether the DNA-binding activity of Egr-1 is also increased with altered pulmonary blood flow. This particular transcripton factor was chosen because it has already been shown to respond to changes in tissue biomechanics, and it is capable of transactivating the TN-C gene promoter (11, 16, 24). Finally, to provide a functional link between MMP-mediated remodeling of the ECM and the induction of TN-C expression and Egr-1 activity, we carried out tissue culture experiments using primary porcine PA SMCs cultivated either on native type I collagen or on the denatured form of this substrate. Whereas nonmodified native type I collagen suppressed TN-C and Egr-1, the denatured form of this substrate supported these activities in an extracellular regulated kinase (ERK1/2) mitogen-activated protein kinase (MAPK)-dependent fashion. Collectively, our studies provide a molecular framework to explain how elevated blood flow may control the composition of the ECM and SMC behavior in pulmonary vascular disease.
MATERIALS AND METHODS
Male and female neonatal Yorkshire swine underwent left PA (LPA) ligation at a mean age of 6.1 + 0.1 days and a mean weight of 2.2 + 0.2 kg. Briefly, the pericardium was opened and the LPA doubly ligated at its origin, and the ductus arteriosus was occluded with a hemoclip. The thoracotomy was repaired in layers, and the animals were allowed to recover. At 12 wk of age, a group of ligated and age-matched control animals underwent a hemodynamic study that included measurements of PA and left atrial pressures and PA flow (cardiac output). Pressure and flow were collected at 200 Hz for 10 s with the ventilator placed at 3 mmHg continuous positive pressure. Pulmonary vascular resistance was calculated using the following formula PVR = (PAP − LAP)/CO, where PVR is pulmonary vascular resistance in Woods Units (WU), PAP is mean arterial pressure (mmHg), LAP is mean left atrial pressure (mmHg), and CO is cardiac output (l/min). Another group of 12-wk-old ligated and age-matched control animals was used for collection of tissues to be used in the subsequent studies outlined below. For this, animals were euthanized with a 40-meq intravenous bolus KCl while under deep anesthesia. Right lower lobe lung tissue that was derived from ligated and age-matched control animals was either flash-frozen in liquid nitrogen or placed in sterile phosphate-buffered saline (PBS) for immediate use. A minimum of three control and three experimental animals was used for each of the studies outlined below. All animal experiments were performed with approval of the Animal Care Committee, Children's Hospital of Philadelphia.
Generation of porcine TN-C and Egr-1 cDNA probes.
Oligonucleotide primers for TN-C (5′-CATCGTGACAGAGGTGACGGAAGA-′3 and 5′-GTGGCCACCCTGGCGCTGACAGGA-3′) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (5′-ACGGCAAGTTCCACGGCACAGT-′3 and 5′-CTCCAGGCGGCAGGTCAGAT-3′) were designed based on that published porcine sequences for these genes. The primers for Egr-1 (5′-GTTTCACGTCTTGGTGCCTTTTGTGTG-3 and 5′TAGACTTTTGAGAGTTGAGATT-′3) were designed based on identical sequences for human and mouse Egr-1. Total RNA was extracted from cultured cells and right lower lobe tissue isolated from 12-wk control and LPA-ligated animals (25). Total RNA (10 μg) was reversed transcribed in a reaction containing 5× synthesis buffer, 10 mM dNTPs, 0.1 mol/l 1,4-dithiothreitol (DTT), and 200 units of Superscript II (GIBCO BRL). Polymerase chain reaction (PCR) was performed with the use of an Expand PCR System (Boehringer Mannheim) with 1 μg of template cDNA and 0.1 μg of each primer in a total volume of 100 μl. Purified PCR products of an appropriate size (TN-C, 260 bp; GAPDH, 593 bp; and Egr-1, 200 bp) were gel eluted (Quiagen) and ligated into a pCR blunt vector (Invitrogen). Clones were selected by restriction analysis, and sequences were verified by automated sequencing.
RNase protection assays.
Total RNA was extracted from 12-wk control and ligated right lower lobe tissue. Riboprobes for TN-C and GAPDH were generated from a linearized pCR blunt plasmid clone using Maxiscript reagents (Ambion) and a T7 promoter site. For the RNase protection assay, Hybspeed kit reagents (Ambion) were used as per the manufacturer's instructions. Briefly, 10 μg of total RNA from 12-wk control and LPA-ligated lungs was incubated with 32P-labeled riboprobes for TN-C and GAPDH in hybridization buffer at 68°C for 20 min. Samples were then digested with A/T1 RNase, precipitated, and electrophoresed on 5% acrylamide gel. Gels were dried, exposed to a phosphoimager, and radioactive bands for TN-C (250 bp) and GAPDH (270 bp) were analyzed densitometrically with the use of Imagequant software (Molecular Dynamics). Data are expressed as a ratio of TN-C/GAPDH.
Total RNA (15 μg per lane) isolated as described above was separated by electrophoresis on a 1.0% agarose formaldehyde gel and transferred onto a H+ nylon membrane (Amersham). Membranes were hybridized with 32P-random labeled cDNA probes for TN-C and Egr-1 at 106 cpm/ml of Hybridization Solution (Stratagene) for 1 h at 68°C. Membranes were washed with 2× standard sodium citrate (SSC) containing 0.1% sodium dodecyl sulfate (SDS) at room temperature for 2 × 15 min, followed by a wash in 0.1 × SSC containing 0.1% SDS at 68°C for 30 min. Membranes were exposed to X-ray film at −70°C. Northern analysis was carried out in duplicate and was used solely to evaluate the TN-C mRNA isoforms produced.
Expression of TN-C and Egr-1 was determined by immunoblotting 20 μg of whole cell lysates or nuclear extracts with appropriate antibodies using Western analysis. Briefly, samples were loaded onto 4–12% polyacrylamide gels (Bio-Rad). To confirm that equal protein was loaded, duplicate SDS-PAGE gels were run in parallel with experimental gels and then stained with Coomassie blue. Proteins were transferred to Immobilon polyvinyl fluoride membranes (Millipore), and these were blocked for 1 h at 37°C in wash buffer composed of 10 mM Tris (pH 7.5), 100 mM NaCl, and 0.1% Tween 20, supplemented with 5% milk protein (Bio-Rad). To detect TN-C protein, membranes were sequentially incubated with mouse monoclonal anti-TN-C antisera (diluted 1:100; GIBCO BRL), followed by incubation with horseradish peroxidase-conjugated goat anti-mouse secondary antibody (GIBCO BRL) diluted 1:5,000 in wash buffer. To detect Egr-1 protein, blots were incubated with a rabbit polyclonal IgG (Santa Cruz), followed by incubation with a horseradish peroxidase-conjugated goat anti-rabbit secondary antibody. All membranes were then rinsed in wash buffer for 3× 15 min, and TN-C and Egr-1 proteins were visualized by enhanced chemiluminescence (ECL kit; Amersham) before exposure to Kodak X-Omat film.
To determine the relative levels of TN-C expression and its pattern of localization, lung tissue from 12-wk control and LPA-ligated animals was assessed using standard immunoperoxidase methods. Briefly, sections of right lower lobe lung were fixed in 10% formalin for 24 h at 4°C. Fixed tissues were subsequently dehydrated in graded ethanols and incubated in xylene before they were embedded in paraffin. The 7-μm-thick sections were deparaffinzed in xylene and sequentially rehydrated in ethanol (100%, 70%, and 50%). Sections were then washed for 3× 15 min in PBS and incubated in pronase-PBS (1 mg/ml) for 15 min at room temperature. After 3× 15-min washes with PBS, the sections were incubated overnight at 4°C with a mouse monoclonal antibody against human TN-C (GIBCO BRL) diluted 1:100 in PBSA (PBS plus 1% bovine serum albumin). Deposition of TN-C was detected using horseradish peroxidase-conjugated goat anti-mouse IgG, followed by exposure to a peroxidase substrate solution (Vectastain). All sections were counterstained with eosin.
Gelatin substrate zymography.
Lung tissue from control and LPA-ligated animals was homogenized at +4°C in protein lysis buffer containing 10 mM sodium phosphate, pH 7.2, 150 mM sodium chloride, 1% Triton X-100, and 0.1% SDS. Homogenates were centrifuged at 5,000 rpm for 5 min and thereafter stored at −70°C. Protein concentrations were detected with bicinchoninic acid protein assay reagents (Pierce). Samples were mixed 3:1 in sample buffer (250 mM Tris · HCl, pH 6.8, 10% SDS, 4% sucrose, and 0.01 bromphenol blue) and then electrophoresed under nonreducing conditions in a 10% agarose gel containing 1 mg/ml of gelatin (Sigma). After electrophoresis, SDS was eluted from the gel in 2.5% Triton X-100 for 30 min at room temperature. The gel was then incubated in substrate buffer (50 mM Tris · HCl, pH 8.0 containing 5 mM CaCl2 and 1 μM ZnCl2) at 37°C for 16 h. After staining with Coomassie blue R-250, gelatin-degrading enzymes were identified by their ability to clear the substrate at their respective molecular weights. To demonstrate that the gelatinolytic bands represent MMPs, gels were incubated in 1 mM EDTA (which inhibits activity of these enzymes). To determine whether higher molecular weight gelatinolytic bands represent the proforms of the MMPs studied, samples were incubated withp-aminophenylmercuric acetate (AMPA) for 1 h before gel loading. This treatment converts the proform of MMPs to its lower molecular weight active form. To determine whether MMPs in lung tissue are equivalent in molecular size and activity to MMP-9 (92 kDa gelatinase) and MMP-2 (72 kDa gelatinase), a purified MMP-9/MMP-2 standard was also separated on substrate gels.
Gel electromobility shift assays.
To generate porcine lung and SMC tissue culture nuclear extracts, samples were processed using a dounce homogenizer in low-salt buffer composed of 0.6% NP40, 150 mM NaCl, 10 mM HEPES, pH 7.9, 1 mM EDTA, and 0.5 mM/l phenylmethylsulfonyl fluoride (PMSF) before centrifugation at 5,000 rpm for 5 min to pellet the nuclei. Nuclei were resuspended in high-salt buffer (25% glycerol, 20 mM HEPES, pH 7.9, 420 mM NaCl, 1.2 mM MgCl2, 0.2 mM EDTA, 0.5 mM DTT, 0.5 mM PMSF, 2 mM benzamide, together with 5 μg/ml each of pepstatin, leupeptin, and aprotinin) and incubated on ice for 20 min. Samples were then centrifuged briefly at 13,000 rpm, and supernatants were collected, quantified by bicinchoninic acid methods (Pierce), and frozen in liquid nitrogen.
Double-stranded Egr-1 consensus (5′-GGATCCAGCGGGGGCGAGCGGGGGCGA-3′) or mutant probes (5′-GGATCCAGCTAGGGCGAGCTAGGGCGA-′3) were end-labeled with [γ-32P]ATP. Nuclear protein was incubated with 20,000 cpm of 32P-labeled consensus or mutant probe for 30 min in a total volume of 10 μl in a binding buffer consisting of a final concentration of 10 mM HEPES, pH 7.9, 3 mM MgCl2, 50 mM NaCl, 10% glycerol, 0.5 mM DTT, 0.5 mM EDTA, and 1 μg of poly(dI-dC). For supershift assays, nuclear proteins were preincubated with a rabbit polyclonal antibody for Egr-1 diluted 1:10 (Santa Cruz, Santa Cruz, CA) before addition of radiolabeled Egr-1 oligonucleotide. For competition reactions, a 50-fold excess of unlabeled Egr-1 oligonucleotide was added to reaction mixtures 10 min before addition of labeled oligonucleotide. All samples were resolved on 5% nondenaturing acrylamide gels in 0.25 mol/l Tris-borate-EDTA (TBE) buffer. The gel was vacuum dried and exposed to Kodak X-Omat film.
Cell culture studies.
Primary PA SMC were isolated from normal 4- to 12-wk porcine PAs using an established method (11). Cultures were routinely maintained in M-199 with 10% fetal bovine serum (FBS) and penicillin-streptomycin (GIBCO BRL). Cells were passaged by trypsinization using 0.05% trypsin-EDTA. For experimentation, SMC were plated at a density of 1.76 × 104cells/cm2 and were initially cultivated for 15–18 h on native or heat-denatured type I collagen (2.48 mg/ml) in M-199 plus 2% FBS, as described previously (14). Thereafter, SMC were maintained for an additional 24 h in M-199 plus 0.1% BSA-fetal calf serum until harvested for experiments. For MEK inhibitor studies, 2 μM PD-98059 (Calbiochem) was included in the medium 24 h after plating for a total of 24 h. An equivalent volume of DMSO served as a control vehicle. All experiments were performed in triplicate.
Immunofluorescent detection of Egr-1 protein.
For immunofluorescence studies, SMCs were cultured on Tissue-Tek chamber slides coated with 2.48 mg/ml of neutralized native or denatured type I collagen. For detection of Egr-1 protein, cells were fixed in cold 100% methanol for 30 min at −20°C. After preincubation in blocking solution containing 10% normal goat serum and 2% BSA for 1 h at room temperature, slides were incubated overnight at 4°C with a rabbit polyclonal anti-Egr-1 antisera (Santa Cruz), diluted 1:100 in PBS with 1% BSA. Cells were washed in PBS and incubated for 1 h at room temperature with 1:50 dilution of fluoroscein isothiocyanate-conjugated goat anti-rabbit antibody. Nuclei were stained with 4′-6-diamidino-2-ohenylindole, diluted 1:10,000 in PBS. All coverslips were mounted onto glass slides using Antifade reagent (Molecular Probes). Observations and photomicrographs were obtained with a Nikon fluorescent microscope using epifluorescence.
To compare expression levels for Egr-1 and TN-C, autoradiographs were scanned and densitometric values derived. All experiments were performed at least in triplicate and the mean values and standard deviation calculated. Differences between groups were established using Student's t-test. A P value of <0.05 was taken to indicate statistically significant differences between groups.
Increased pulmonary blood flow upregulates TN-C expression.
To determine whether increased pulmonary blood flow regulates TN-C mRNA and protein expression, we analyzed right lung tissue retrieved from control and experimental pigs at 12 wk after ligation of the LPAs of 6-day-old neonatal animals. The lung tissue examined in this study was identical to that in which we had already assessed the effects of PA ligation on animal weight and in which we had previously performed hemodynamic and morphometric studies (28). Briefly, we showed that the mean weight for control and ligated animals was 27.6 and 12.4 kg, respectively. The mean baseline PA pressure was 24.1 ± 3.0 mmHg in the ligated group (n = 12) and 20.8 ± 1.9 mmHg in the control group (n = 12;P > 0.05). Cardiac output was 1.7 ± 0.1 l/min in the ligated group and 3.11 ± 0.3 l/min in the controls (P < 0.001), whereas vascular resistance was 13.2 ± 2.2 and 5.8 ± 0.8 WU in the ligated and control groups, respectively (P = 0.001).
RNA protection assays with radiolabeled riboprobes for TN-C and GAPDH showed that compared with control animals, TN-C mRNA levels increased ∼2.6-fold (P < 0.05) in right lung tissue after left PA ligation (Fig. 1 A;left and right). Furthermore, Northern blot analysis on total RNA extracted from lung tissue showed that two TN-C isoforms of apparent molecular weight of 7.4 and 6.6 kb were upregulated in ligated animals (Fig. 1 B). Western immunoblot analysis for TN-C protein demonstrated that a significant ∼26-fold (P < 0.01) increase in the production of two corresponding TN-C protein isoforms of 220 and 180 kDa occurred in ligated animals, with the higher molecular weight isoform predominating (Fig. 1 C; left and right).
To determine where TN-C protein is deposited within lung tissue, we carried out immunohistochemical studies. No TN-C immunoreactivity was observed in PAs isolated from control animals (Fig.2), whereas TN-C protein was detected in PAs derived from ligated animal lung tissue (Fig. 2). Sections treated with control IgG were negative, thereby demonstrating the specificity of the TN-C immunostaining (data not shown). Collectively, these data indicate that altered pulmonary blood flow increases TN-C expression, both at the mRNA and protein levels.
Activation of MMPs with elevated pulmonary blood flow.
We then determined whether the expression and/or activity of MMPs is also modulated by changes in pulmonary blood flow. This was assessed using lung tissue extracts and gelatin substrate zymography. Gelatinolytic bands in the 92- and 72-kDa range were present in both control and experimental lung tissue (Fig.3 A). An additional band at ∼68 kDa was detected in the ligated group, but not in the controls (Fig. 3 A). Molecular size estimations of these gelatinases indicate that they likely represent MMP-9 and MMP-2. Also the finding that EDTA prevents gelatin substrate proteolysis indicates that these gelatinases represent MMPs (Fig. 3 B). Treatment of control tissue extracts with AMPA (an agent that promotes MMP activation) resulted in the production of gelatinases of ∼90 and 68 kDa (Fig.3 C), further indicating that the 92- and 72-kDa species represent the proforms of MMP-9 and MMP-2, respectively. In addition, a recombinant MMP-2 and MMP-9 standard comigrated with the proteinases identified in control and ligated tissues, once again indicating that they represent MMP-9 and MMP-2.
Increased pulmonary blood flow leads to sustained Egr-1 expression and activity.
The TN-C gene promoter is highly conserved between species and contains a consensus binding site that interacts with Egr-1, a zinc finger transcription factor that is known to be activated by changes in biomechanical force (see Ref. 11 for review). To determine whether altered pulmonary hemodynamics modulates Egr-1, we first carried out Northern analysis with total RNA isolated from control and ligated lung tissue. No significant differences (P = 0.375) in the steady-state levels of an ∼3.9-kDa Egr-1 mRNA were observed between controls and the experimental group (Fig.4 A). Also, the levels of Egr-1 protein (82 kDa) in control or ligated animals did not change significantly (P = 0.439), as determined by Western immunoblot analysis (Fig. 4 B). It should be noted, however, that Egr-1 protein in the ligated animal group migrated at a higher molecular weight, indicating that it may have been activated at the posttranslational level, potentially via phosphorylation (Fig.4 B; left).
To determine whether the DNA-binding activity of Egr-1 is affected by changes in pulmonary blood flow, we carried out electrophoretic mobility shift assays by using nuclear extracts from control and ligated animals combined with a radiolabeled double-stranded oligonucleotide that represents a consensus binding site for Egr-1 protein. These experiments showed that altered pulmonary blood flow promotes the formation of a DNA-nuclear protein complex, which was present at low or nondetectable levels in control animals (Fig.5). Moreover, complex formation was increased by the inclusion of increasing amounts of nuclear protein. The specificity of complex formation between the Egr-1 consensus-binding site and nuclear protein(s) was demonstrated by incubating DNA:nuclear protein complexes with an Egr-1-specific antisera, which resulted in a mobility supershift in ligated animal tissue samples (Fig. 5). Thus chronic changes in pulmonary blood flow support interactions of Egr-1 with its consensus-binding site.
Cultivating primary porcine PA SMC on denatured collagen mimics flow-dependent induction of TN-C and Egr-1 DNA binding activity.
Having shown that that changes in pulmonary blood flow increase MMP activity, TN-C expression, and the DNA-binding activity of Egr-1 in vivo, we explored the hypothesis that flow-mediated activation of MMPs supports SMC TN-C expression and Egr-1 activity via its effect on the structure of type I collagen. To test this, primary porcine PA SMCs were cultivated either on native or heat-denatured type I collagen, an ECM substrate that mimics MMP-proteolyzed type I collagen (13,14). On native type I collagen, SMCs formed three-dimensional aggregated networks composed of stellate-shaped cells (Fig.6 A), which expressed low levels of two TN-C protein isoforms of 220 and 180 kDa, as determined by Western immunoblot analysis (Fig. 6 B). In contrast, SMC cultivated on denatured collagen adopted a spread morphology (Fig.6 A) and expressed high levels of the 220-kDa TN-C isoform and lower levels of the 180-kDa protein (Fig. 6 B).
In concert with increased TN-C protein expression, Western immunoblotting using nuclear extracts revealed that Egr-1 protein accumulates within nuclei of SMCs cultivated on denatured collagen, but not on native collagen (Fig.7 A). Consistent with this, electrophoretic mobility shift assays showed that nuclear protein binding to a radiolabeled Egr-1 consensus sequence oligonucleotide is suppressed by native collagen, but is promoted by denatured collagen (Fig. 7 B). In contrast, a mutant Egr-1 oligonucleotide was unable to form a complex with nuclear extracts derived from SMC maintained on native type I collagen (data not shown).
Because TN-C promoter activity in SMCs is dependent on ERK1 and ERK2 MAPKs (19), we tested the effects of blocking ERK1 and ERK2 activity (phosphorylation) on the nuclear localization and DNA binding activity of Egr-1. Immunofluorescence microscopy demonstrated that treatment of SMCs cultivated on denatured type I collagen with PD-98059 (a MEK1 inhibitor) prevented accumulation of Egr-1 in the nucleus, whereas nuclei of cells treated with medium alone (not shown) or with the DMSO control vehicle contained Egr-1 protein (Fig.7 C). In keeping with the idea that phosphorylation and nuclear localization of Egr-1 may be required for binding to target sequences, supershift assays with a radiolabeled Egr-1 consensus oligonucleotide and Egr-1 antisera demonstrated that inhibition of ERK1 and ERK2 activity with PD-98059 suppresses the formation of DNA-Egr-1 protein complexes in SMC cultivated on denatured collagen (Fig.7 D). Together, these data demonstrate that alterations in type I collagen structure and concomitant changes in SMC adhesion modulate the localization and activity of Egr-1 as well as expression of TN-C in a manner that recapitulates the control that increased pulmonary flow exerts over these factors in vivo.
The development and progression of pulmonary vascular disease is influenced by multiple factors, including hemodynamics (21,28). In this study, we showed that TN-C expression, MMP activity, and the DNA-binding activity of Egr-1 are supported by changes in pulmonary blood flow in neonatal pigs. Although it is possible that such changes arise due to elevated PA pressure, it is also likely that increased pulmonary blood flow is a major contributory factor. This conclusion is because the ligated animals were about twice the size of the controls. If cardiac output of the ligated animals is doubled to account for the size difference, it is almost exactly the same as the controls. In addition, in the ligated animals, blood flow is directed to only one lung, not both, as is the case for the controls. Thus the flow to the right lung is elevated and almost twice that of the right lung in the controls.
The results of this study led us to directly explore the mechanistic links between flow-induced changes in type I collagen structure that occur in response to MMPs and to relate this to changes in Egr-1 activity and TN-C expression. To this end, we showed that modifying cultured SMC interactions with type I collagen recapitulates flow-dependent increases in Egr-1 binding activity and TN-C expression that are observed in vivo. Moreover, we showed that nuclear localization and DNA binding activity of Egr-1 in SMC cultivated on denatured collagen is dependent on ERK1 and ERK2 MAPKs. Figure8 shows a hypothetical scheme summarizing these ideas.
Mechanical signals sensed by the ECM and associated cell adhesion receptors are known to initiate intracellular pathways that modulate cell adhesion components used to survey and interact with this mechanical microenvironment. For example, cyclic stretch of human umbilical endothelial cells increases β3-integrin mRNA levels, whereas cyclic strain of human umbilical endothelial cells leads to reorganization of β1-, α5-, and α2-integrins (reviewed in Ref. 25). Mechanical strain has been shown to induce TN-C in neonatal rat cardiac myocytes (29), whereas chick embryo fibroblasts and vascular SMC cultured on attached type I collagen substrates produce higher levels of TN-C than those cultivated on mechanically relaxed collagen gels (11, 13). Physical loading of rat ulnae leads to early increases in osteoblast TN-C expression, and recent in vivo studies in fibroblasts and chondrocytes within the osteotendinous junction show that removal of mechanical stress suppresses TN-C (reviewed in Ref.11). Pressure also affects the distribution of TN-C during scar formation in the epidermis and during ex vivo remodeling of porcine arteries (reviewed in Ref. 11). Combined with the present findings, it is clear that TN-C is regulated by mechanical factors. This notion is also in keeping with gene microarray studies showing that out of 5,000 genes examined, uniaxial stretch of vascular SMC produced increases in expression of more than 2.5-fold in only three genes, one of which was TN-C (9).
Increases in both TN-C mRNA and protein expression were produced in right lung tissue after left PA ligation. TN-C mRNA was also detected in lung tissue from control animals, yet the relative levels of TN-C protein compared with mRNA were significantly higher. This suggests that the stability or translation of TN-C mRNA is suppressed by normal blood flow, or alternatively that translation is increased with elevated blood flow. In keeping with this idea, Chicurel and co-workers (4) have demonstrated that applying mechanical forces to cells through integrins promotes accumulation of mRNA and ribosomes to focal adhesions, thereby enhancing mRNA translation near the site of signal reception. Increased catabolism of TN-C protein by ECM-degrading enzymes may have also accounted for the differences between TN-C mRNA and protein levels in control animals. This is less likely, though, given that MMP activity was greater in lung tissue derived from the ligated animal group. Northern and Western immunoblot analyses showed that two TN-C mRNA and protein isoforms are produced in developing lung tissue, with the larger isoform predominating, regardless of hemodynamics. Although alternative splicing of TN-C mRNA may account for the appearance of these different isoforms, the functional significance of this, vis-a-vis pulmonary vascular development and disease, remains to be determined.
At least two different TN-C gene promoter elements and transcription factors appear to be activated in response to mechanical factors. The induction of TN-C in chick embryo fibroblasts cultured on mechanically stressed type I collagen gels relies on a conserved GAGACC stress response element within the TN-C gene promoter (reviewed in Refs.25 and 26). In this study, however, we have focused on the Egr-1 transcription factor for the following reasons: 1) the TN-C gene promoter contains an element that binds Egr-1, 2) mechanical factors activate Egr-1, and 3) identical to TN-C expression, activation of Egr-1 is dependent on ERK1/2 MAPKs. We now show that chronic increases in pulmonary blood flow in neonatal pigs lead to sustained activation of Egr-1 in the context of increased TN-C production. Furthermore, our tissue culture studies demonstrate that the DNA-binding activity of Egr-1 depends on activated ERK1 and ERK2. The finding that activated Egr-1 is detectable at 12 wk post-LPA ligation is also noteworthy, given that this transcription factor has previously been designated as an immediate and early gene. A recent study (17) in mouse and human atherosclerotic tissues examining the expression of Egr-1, however, also challenges this preexisting notion.
How might MMPs control the activity of Egr-1 and TN-C gene expression? Many of the gene expression events initiated by mechanical signals are identical to those initiated by specific ECM ligands, indicating that the molecular pathways elicited by these factors overlap. For example, both mechanical stretch and cell adhesion to the ECM stimulates activation of focal adhesion kinase (FAK) and ERK MAPKs. Increasing evidence also shows that specific ECM proteins are required to sense mechanical stimuli. Cyclic stretch of neonatal rat vascular SMC on vitronectin, fibronectin, or collagen substrates promotes SMC growth, whereas laminin and elastin are unable to support this function (see Ref. 25 for review). Similarly, specific integrin subunits are required to transduce mechanical signals exemplified by the finding that increased SMC growth in response to strain can be abrogated by RGD peptides or antibodies against β3- and αvβ5-integrin antibodies, but not by antibodies against β1-integrin subunits. Because β3-integrins are upregulated after vascular injury, these tissue culture experiments are likely to be relevant in the pathobiology of vascular disease. Overall, it appears that β3-integrins are well positioned to acts as mechanotransducers; this would require the presence of an appropriate β3-integrin ligand, which can be generated by MMP-mediated proteolysis of type I collagen (13, 14). This is especially pertinent to the present study in which we have shown that the activity of an ∼68-kDa MMP is activated by increased pulmonary blood flow.
The notion that the structure of type I collagen is critical for the induction of TN-C is also supported by our tissue culture studies, which demonstrate that native type I collagen (a β1-integrin ligand) suppresses nuclear translocation and the DNA-binding activity of Egr-1 as well as TN-C expression, whereas heat-denatured type I collagen [a β3-integrin ligand (13, 19)] has the opposite effect. Future cell transfection and antibody blocking studies will be informative in determining whether Egr-1 plays a direct role in upregulating TN-C in vascular smooth muscle cells and whether this is dependent on β3-integrins.
In summary, the present study highlights the dynamic and reciprocal links that exist between mechanical factors and cell adhesion components and demonstrates for the first time how changes in pulmonary blood flow may alter SMC behavior via its effect on both the catabolism and biosynthesis of the arterial ECM by MMPs and TN-C, respectively. In addition, the finding that Egr-1 is activated in a sustained manner in response to alterations in pulmonary blood flow or SMC adhesion in tissue culture offers new mechanistic insights as to how hemodynamic factors and arterial ECM components may cooperate to modify SMC behavior during the development and progression of pulmonary vascular disease.
The authors thank Dr. Frederick Jones for helpful discussions and Dr. Robert Levy for critical reading of the manuscript. We also thank Genevieve Dion for help in preparing the figures.
P. L. Jones was supported by the Florence Murray Award and by American Heart Association National Grant-In-Aid 9950622N. J. W. Gaynor is supported by the W. W. Smith Charitable Trust (Grant H9704) and by the Tommy Martin Memorial Fund.
Address for reprint requests and other correspondence: P. L. Jones, Univ. of Colorado Health Sciences Center, Dept. of Pediatrics, 4200 E. 9th Ave., Box B-131, Denver, CO 80262 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2002 the American Physiological Society