Main

CHD with increased pulmonary blood flow is often associated with the development of pulmonary hypertension (1). After birth, as pulmonary vascular resistance normally decreases, the presence of a systemic to pulmonary communication generates an increase in pulmonary blood flow. This abnormal postnatal hemodynamic state results in progressive structural and functional abnormalities of the pulmonary vascular bed (2,3). Our animal model of pulmonary hypertension, developed by inserting an aortopulmonary vascular graft in the late-gestational fetal lamb (4–6), may help elucidate the mechanisms involved. Postnatally, these shunt lambs have increased pulmonary blood flow and pressure (4). In addition, they display vascular remodeling typical of pulmonary hypertension secondary to CHD, characterized by increased medial wall thickness of the small pulmonary arteries and abnormal extension of muscle to peripheral pulmonary arteries (1,4,7).

Recent studies have demonstrated that shunt lambs display abnormal signaling by several growth factors mitogenic for vascular smooth muscle, including ET-1 (6), transforming growth factor β-1 (TGF β-1) (8), and vascular endothelial growth factor (VEGF) (9). Another potential contributor to vascular remodeling is FGF-2. FGF-2 displays mitogenic effects in the early proliferation of SMC (10,11) and in neointimal thickening (12) following vascular injury. Furthermore, a progressive increase in FGF-2 protein within the smooth muscle layer of pulmonary arteries was demonstrated in a rat model of monocrotaline-induced pulmonary hypertension (13). In addition, elevated FGF-2 protein levels were detected in the urine and plasma of patients with pulmonary arterial hypertension, suggesting the involvement of FGF-2 in the SMC proliferation characteristic of this disease (14).

Vascular cells within arterial walls are subjected to biomechanical forces, including cyclic stretch generated by pulsatile blood flow. EC are also subjected to shear stress due to blood flow, and fluid dynamic models of intact blood vessels suggest that SMC also experience shear stress as a result of interstitial flow (15). Cyclic stretch increased FGF-2 mRNA in PASMC (16) and shear stress increased the release of FGF-2 from aortic SMC (17) and aortic EC (18). Since these biomechanical forces are increased in shunt pulmonary arteries due to elevated pulmonary blood flow, we hypothesized that similar to monocrotaline-induced and adult pulmonary hypertension, increased FGF-2 signaling may be associated with pulmonary vascular remodeling in shunt lambs. Thus, the purpose of this study was to monitor FGF-2 expression in shunt lambs, to determine the effects of cyclic stretch and shear stress on FGF-2 expression in PASMC and PAEC, and to determine the effects of exogenous FGF-2 on PASMC proliferation.

MATERIALS AND METHODS

Surgical preparation and care.

Twenty mixed-breed Western pregnant ewes (137–141 d gestation, term = 145 d) were operated on under sterile conditions to insert an 8.0-mm Gore-tex vascular graft (2 mm length; W.L. Gore, Milpitas, CA) between the ascending aorta and main pulmonary artery of the fetus as previously described (4). Unoperated twin fetuses served as controls, because we have previously shown that unoperated and sham-operated control lambs have similar physiology and morphology (4). Lambs were delivered spontaneously, and at 4 wk of age were killed by an intravenous injection of pentobarbital sodium (Euthanasia CII; Central City Medical, Union City, CA) followed by bilateral thoracotomy. All procedures and protocols were approved by the Committee on Animal Research of the University of California, San Francisco.

Tissue/RNA/protein preparation.

The heart and lungs were removed en bloc. Two to three gram sections from each lobe of the lung and sections from third- to fifth-generation intralobar pulmonary arteries and veins, with inside diameters of 0.5–2.0 mm, were removed. These tissues were snap-frozen in liquid nitrogen and stored at –80°C until used. RNA and protein were isolated from snap-frozen lung tissue and analyzed by RNAase protection and Western blotting as described below. Tissue was prepared for immunohistochemistry as previously described (9).

RNAase protection assay.

Antisense single-stranded riboprobes were synthesized by in vitro transcription from a plasmid containing an ovine FGF-2 cDNA fragment, and RNAse protection assays were performed on total RNA prepared from peripheral lung tissue as previously described (5,19). Also included was a probe for 18S to serve as a control for the amount of input total RNA and the recovery of protected probe fragments. In all experiments, sufficient counts were added such that the FGF-2 and 18S riboprobes were always in molar excess.

Western blot analysis.

Twenty micrograms of protein extracts were analyzed by Western blot analysis as previously described (5,19). Membranes were incubated at room temperature with polyclonal antiserum raised against FGF-2 (sc-79, 1:200, Santa Cruz Biochemicals, Santa Cruz, CA) then incubated with a goat anti-rabbit IgG-horseradish peroxidase conjugate. After washing, chemiluminescence was used to detect the protein bands.

Immunohistochemistry.

Lung tissue samples (snap frozen and stored at –80°C) were embedded in Tissue-Tek O.C.T Compound (Sakura Finetek USA Inc., Torrance, CA), cryosectioned at 5 μm, and stored at –80°C until needed. Immunohistochemistry was performed using a 1:200 dilution of the anti-FGF-2 antibody (ab8880, Abcam Inc., Cambridge, MA) as described previously (9). Slides were imaged using an Olympus IX51 (Leeds Precision Instruments, Inc., Minneapolis, MN) and quantified using the Image Pro Plus software (MediaCybernetics, Silver Spring, MD).

Serum FGF-2 quantification.

Systemic arterial blood was prepared as described previously (20). FGF-2 protein levels in serum samples were determined using a quantitative sandwich enzyme immunoassay technique (Quantikine, R & D Systems, Minneapolis, MN). The OD of each well was read at 450 nm, with wavelength correction at 540 nm, using a Multiscan MCC/340 microplate reader (Labsystem, Helsinki, Finland).

Cell culture.

Primary cultures of PASMC from 4-wk old sheep that had not previously been surgically operated were isolated by the explant technique as described previously (21). Identity was confirmed as PASMC by immunostaining (>99% positive) with antibodies against α-smooth muscle actin, calponin, and caldesmon. This was taken as evidence that cultures were not contaminated with fibroblasts or with endothelial cells. Primary cultures of PAEC were isolated as described previously (22). Endothelial cell identity was confirmed by their typical cobblestone appearance, contact inhibition, specific uptake of DiI-Ac-LDL (Molecular Probes, Eugene, OR), and positive staining for von Willebrand factor (DAKO, Carpinteria, CA).

All cultures for subsequent experiments were maintained in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% FCS (Hyclone Laboratories, Logan, UT), antibiotics (MediaTech, Herndon, VA), and antimycotics (MediaTech) at 37°C in a humidified atmosphere with 5% CO2–95% air. Cells were used between passages 3 and 10.

Promoter analysis.

A plasmid containing human FGF-2 promoter DNA from –1800 to +314 bp (relative to the transcription start site) fused to a luciferase reporter gene was a gift from Dr. M. Stachowiak and Dr. J. Moffett. Cells were co-transfected at 70% confluence with 4 μg of plasmid DNA and 0.1 μg of Renilla luciferase internal control vector on a 10 cm2 tissue culture plate using Effectine (QIAGEN, Valencia, CA) according to the manufacturer's instructions. After 24-h cells were split onto 6-well plates, maintained in DMEM containing 1% serum and 10 μg/mL heparin, and left static or subjected to cyclic strain or laminar shear stress for 8–24 h as described below. Luciferase activity of 20 μL protein extracts was determined using the Dual-Luciferase Reporter Assay System (Promega, Madison, WI) and a Femtomaster FB12 luminometer (Zylux, Oak Ridge, TN).

Cyclic stretch and fluid shear stress.

Cells were maintained in DMEM containing 1% serum and 10 μg/mL heparin on 6-well BioFlex plates coated with collagen type I (FlexCell) for 24 h, then subjected to biaxial cyclic stretch using the FlexCell 3000 Strain Unit. Plates were placed on a loading station and stretched by applying an oscillatory vacuum to the underside of the membranes. Cells were stretched at a frequency of 1 Hz with 20% amplitude for 8–24 h in accordance with a previous study (16).

Cells were maintained in DMEM containing 1% serum and 10 μg/mL heparin on 6-well tissue culture plates (Costar) for 24 h before initiating shear stress experiments. A cone-plate viscometer was designed and built such that it accepts 6-well plates (22). This allowed the PASMC and PAEC monolayer to be subjected to a radially constant fluid shear stress at laminar flow rates representing levels of shear stress within physiologic parameters. Typical physiologic shear stress in the major human arteries is in the range of 5–20 dynes/cm2 (23). Thus, we imparted a shear stress of 20 dynes/cm2 for 8–24 h to mimic the upper limit of the physiologic range.

FGF-2 quantification by ELISA.

Intracellular and extracellular FGF-2 protein levels were determined using a quantitative sandwich enzyme immunoassay technique (Quantikine). Cell extracts were adjusted for protein content then further diluted 1:100 before assay. To determine FGF-2 released, 100 μL of undiluted medium was added per well of the precoated microplate. Samples and FGF-2 standards were assayed in duplicate for each assay according to the manufacturer's instructions. The OD of each well was read at 450 nm, with wavelength correction at 540 nm, using a Multiscan MCC/340 microplate reader (Labsystems).

Cell proliferation assays.

Increases in PASMC number in response to exogenous FGF-2 were determined as described previously (21). Cells were seeded onto 96-well plates (Costar, Corning, NY) at approximately 25% confluence, allowed to adhere for at least 18 h, then synchronized by serum starvation for a further 24 h. Cells were then treated with 0–100 ng/mL recombinant human FGF-2 (Invitrogen, Carlsbad, CA) in serum-free medium. After 72 h, viable PASMC number was determined using the Cell Titer 96 AQueous One Solution kit (Promega), Twenty microliters of reagent was added directly to cells in 100 μL medium, and, following a 2-h incubation period at 37°C, the absorbance at 492 nm was read using a Labsystems Multiskan EX plate reader (Fisher Scientific, Pittsburgh, PA).

Statistical analysis.

Unless otherwise stated, values are expressed as mean ± SD. Comparisons between treatment groups were made by the unpaired t test using the GB-STAT or GraphPad Prism version 4.00 for Windows (GraphPad Software, San Diego, CA) software programs. A p < 0.05 was considered statistically significant.

RESULTS

Analysis of FGF-2 mRNA by RNAse protection assays showed increased expression in shunt lambs relative to age-matched control animals (Fig. 1 A). Densitometric analysis of lower exposure autoradiographs to those presented in Figure 1 A indicated an 80% increase in shunt FGF-2 mRNA normalized to 18S relative to control animals (Fig. 1 B).

Figure 1
figure 1

RNAase protection assays for FGF-2 (bFGF) on lung tissue from 4-wk-old lambs with increased (shunt) and normal (control) pulmonary blood flow and pressure. Representative RNAse protection assay (A) showing that FGF-2 mRNA expression is increased in shunt lambs. A molecular weight ladder with sizes in nucleotides is presented in the left lane. “tRNA” is tRNA plus bFGF probe; “probe alone” is digested bFGF probe; “bFGF” is undigested bFGF probe; “18S” is undigested 18S probe; “Shunt” is shunt RNA digested with bFGF and 18S probes; “Control” is control RNA digested with bFGF and 18S probes. 18S RNA band indicates equal RNA loading. Densitometric values for relative FGF-2 mRNA (normalized to 18S RNA and to controls) from four control and four shunt lambs were determined (B). n = 4 shunt and 4 control animals. Values are means ± SD. *p < 0.05, shunt vs control.

Full size image

The increase in FGF-2 mRNA was consistent with a significant increase in FGF-2 protein in lung tissue extracts from shunt animals as detected by Western blotting (120% higher than controls, Fig. 2, A and B).

Figure 2
figure 2

Western blot for FGF-2 from shunt and control lung tissue. Representative Western blot (A) showing that FGF-2 protein is increased in shunt lambs relative to controls. Densitometric values for relative FGF-2 protein from four control and four shunt lambs were determined (B). n = 4 shunt and 4 control animals. Values are means ± SD. *p < 0.05, shunt vs control.

Full size image

The localization of FGF-2 protein expression in shunt and control lambs was evaluated by immunohistochemistry using peripheral lung tissue sections. FGF-2 staining was evident in all animals, with strong expression in the vessels (Fig. 3, A and B). Quantification of vessel-associated fluorescence demonstrated that FGF-2 protein was increased by 180% in shunt lamb vessels relative to controls (Fig. 3 C). Omission of either the primary or secondary antibody resulted in undetectable signals indicating lack of autofluorescence (data not shown).

Figure 3
figure 3

Immunohistochemistry to determine FGF-2 protein distribution and levels in lung tissue sections prepared from 4-wk-old shunt and control lambs. Representative images of control (A) and shunt (B) lung tissue stained with an anti-FGF-2 antibody. Calibration bars = 100 μm. Airways are indicated by “AW.” Vessels are indicated by arrows. (C) Image Pro Plus software analysis was used to quantify FGF-2 protein in vessels. Shunts: n = 7 animals, 70 vessels; controls: n = 3 animals, 27 vessels. Values are mean ± SD. *p < 0.05 shunt vs control.

Full size image

ELISA indicated an increase in circulating FGF-2 protein in shunt animals, with serum concentrations more than 4-fold higher in shunts relative to controls (Fig. 4).

Figure 4
figure 4

ELISA to quantify FGF-2 protein in serum from 4-wk-old control and shunt lambs. n = 4 shunt and 4 control animals. Values are mean ± SEM. *p < 0.05 shunt vs control.

Full size image

PASMC and PAEC isolated from 4-wk-old lambs were subjected to 20% cyclic stretch at 1 Hz and 20 dyn/cm2 laminar shear stress to mimic physiologic levels of these biomechanical forces during pulsatile blood flow (23,24). Using an 1800 bp FGF-2 promoter fragment fused to a luciferase reporter gene we first determined the time-dependent effects of biomechanical forces on FGF-2 promoter activity. In PASMC, 24 h of stretch and shear increased luciferase activity by 77% and 44%, respectively, relative to static controls, with no significant change detectable after 8 h (Fig. 5 A). In PAEC, promoter activity peaked after 8 h of stretch and shear (33% and 39%, respectively) before decreasing to basal levels by 24 h (Fig. 5 B).

Figure 5
figure 5

Luciferase assays to determine the activity of an 1800 bp FGF-2 promoter fragment in PASMC (A) and PAEC (B) exposed to 8 h and 24 h stretch (▪) and shear (□); n = 4. Values are mean ± SD. *p < 0.05 vs static cells.

Full size image

Intracellular FGF-2 protein levels were quantified from cells exposed to stretch and shear using an ELISA-based assay. Significant increases in intracellular FGF-2 protein were detected in PASMC after 8 and 24 h of stretch and shear (Fig. 6 A). At 24 h, stretch and shear increased levels by 43% and 178%, respectively (Fig. 6 A). Intracellular FGF-2 levels were also elevated in PAEC, with values increasing by 59% and 175% after 24-h stretch and shear, respectively (Fig. 6 B).

Figure 6
figure 6

ELISA to determine FGF-2 protein content in cell extracts from PASMC (A) and PAEC (B) exposed to 8 and 24-h stretch (▪) and shear (□); n = 4. Values are mean ± SEM. *p < 0.05 vs static cells.

Full size image

FGF-2 released by cells exposed to stretch and shear was quantified by performing ELISA assays on the extracellular media. Eight and 24 h of shear significantly increased FGF-2 protein detected in the media surrounding PASMC by 15.9- and 13.6-fold, respectively, relative to static controls (Fig. 7 A). Conversely, 8-h stretch significantly increased FGF-2 release by 5.9-fold, before decreasing to a statistically insignificant level of 1.9-fold by 24 h (Fig. 7 A). In PAEC, shear increased FGF-2 released by 5.0-fold and 7.5-fold after 8 h and 24 h, respectively (Fig. 7 B), while stretch increased FGF-2 released by 1.9-fold at 8 h before decreasing to basal levels by 24 h (Fig. 7 B).

Figure 7
figure 7

ELISA to determine FGF-2 protein content in extracellular media of PASMC (A) and PAEC (B) exposed to 8 and 24 h stretch (▪) and shear (□); n = 4. Values are mean ± SEM. *p < 0.05 vs static cells.

Full size image

We next determined the dose-dependent effects of exogenous FGF-2 on PASMC proliferation. At 72 h, 100 pg/mL FGF-2 increased viable PASMC number by 30% relative to serum-free medium (Fig. 8), while optimum growth medium containing 10% serum increased viable cell number by 92% (Fig. 8).

Figure 8
figure 8

Viable cell assays to determine increases in PASMC number in response to exogenous FGF-2; n = 4. Values are mean ± SD. *p < 0.05 vs serum free.

Full size image

DISCUSSION

We have previously reported an underlying dysregulation of growth factor expression in 4-wk-old shunt lambs, including increases in ET-1 (6), TGF-β1 (8), and VEGF (9) signaling. The present study identifies increased FGF-2 expression as another potential contributor to abnormal PASMC growth in the shunt model of pulmonary hypertension.

Our in vitro data suggest that biomechanical forces, resulting from increased pulmonary blood flow, may play a role in the up-regulation of FGF-2 expression within shunt pulmonary arteries. The increase in FGF-2 promoter activity in PASMC after 24 h of stretch is in agreement with a previous study demonstrating an 80% increase in FGF-2 mRNA in ovine PASMC exposed to 24 h of cyclic stretch (16). Our data further add to the existing knowledge of FGF-2 expression in SMC by demonstrating a role for shear stress in its regulation. Vascular SMC are normally exposed to shear stress due to interstitial flow driven by transmural pressure gradients (15), and to our knowledge this is the first study to provide evidence of increased FGF-2 transcription in vascular SMC in response to shear stress. The effects of biomechanical forces on endothelial FGF-2 expression appear to depend on the location of the cells within the vasculature. In bovine aortic EC, FGF-2 mRNA levels transiently increased after 6 h shear, while cyclic stretch had no effect (25). Conversely, increased FGF-2 mRNA was detected in human umbilical vein EC after 24 h of shear stress (26), while in vivo exposure of rat mesenteric small arteries to elevated flow failed to alter endothelial FGF-2 mRNA at any time over 7 d (27). Our data suggest that in the pulmonary circulation, stretch and shear both transiently increase FGF-2 promoter activity in PAEC.

The mechanisms that mediate the regulation of FGF-2 expression in PASMC and PAEC by biomechanical forces are currently unknown. We detected temporal differences in FGF-2 expression between PASMC and PAEC exposed to stretch and shear suggesting that cell-specific mechanisms including increased RNA and protein stability may be involved. In addition, stretch and shear may regulate FGF-2 expression in the same cell type via different pathways. For example, a shear stress–responsive element within the platelet-derived growth factor-B (PDGF-B) promoter was necessary for increased expression in bovine aortic EC exposed to shear (28), but not to stretch (29). Additional studies are therefore warranted to characterize the transcriptional, posttranscriptional, and posttranslational regulation of FGF-2 expression in PASMC and PAEC in response to stretch and shear.

Biomechanical forces may also exert indirect effects on FGF-2 expression and signaling in shunt lambs. TGF-β1 expression was increased in 4-wk-old shunt animals relative to controls (8), which may be due in part to increased TGF-β1 expression and release by PASMC exposed to stretch (30). Furthermore, TGF-β1 expression was enhanced by increased flow and shear stress in vivo (31), and TGF-β1 has been shown to up-regulate FGF-2 in vascular SMC (32). Together, these data suggest that induction of TGF-β1 expression and release by biomechanical forces may stimulate FGF-2 expression in PASMC. Similar to patients with pulmonary arterial hypertension (14), our results show that 4-wk-old shunt lambs exhibit increased FGF-2 serum levels. This may be due, in part, to the stimulation of expression and release of FGF-2 from pulmonary vascular cells exposed to biomechanical forces. Several studies have demonstrated FGF-2 autoregulation (33), raising the possibility of sustained increases in FGF-2 expression in shunt pulmonary vascular cells via positive feedback mechanisms. While many other factors may be involved, these data illustrate the potentially complex regulation of FGF-2 expression in shunt lambs.

Circulating FGF-2 may trigger PASMC growth by binding extracellular receptors, whereas increases in intracellular FGF-2 in PASMC exposed to stretch and shear can potentially stimulate cell proliferation by activating nuclear-localized receptors (33). FGF-2 may further contribute to pulmonary hypertension and vascular remodeling via its interaction with other growth factor signaling pathways. Neutralization of FGF-2 reduced VEGF release from vascular SMC (34), whereas FGF-2 was found to increase ET-1 subtype A receptor (ETA) expression in PASMC (35). ETA mRNA and protein levels were increased in shunt lambs (6), and ET-1 stimulated the proliferation of fetal PASMC via ETA receptor-mediated signaling (21). These data raise the possibility that FGF-2 may exert an additional indirect mitogenic effect on PASMC growth by up-regulating ETA receptor expression. In addition, this is likely to enhance the contribution made by ET-1-mediated vasoconstriction to pulmonary hypertension in shunt lambs. Further studies are needed to characterize the interaction between FGF-2 and ET-1 signaling pathways in shunt pulmonary arteries.

This study has focused on the role of biomechanical forces, although other factors may also contribute to abnormal FGF-2 expression in shunt lambs. Pulmonary arterial oxygen saturation was significantly increased in shunts due to the mixing of aortic and pulmonary arterial blood (4) (63.3 ± 11.4% shunts vs 42.0 ± 5.8% controls, p < 0.05; Dr. Jeffrey Fineman, unpublished observations). Although it is not yet known whether this increase in oxygen saturation is sufficient to stimulate FGF-2 expression in shunt lambs, levels of FGF-2 mRNA and protein were increased in the lungs of mice exposed to 80% O2 for 4 d (36). Furthermore, exposure to 60% O2 for 14 d increased ET-1–dependent PDGF-beta receptor expression in a newborn rat model of pulmonary hypertension with vascular remodeling (37). These data suggest that increased pulmonary arterial oxygen saturation may also contribute to abnormal FGF-2 and ET-1 signaling in shunt lambs.

Overall, the above data demonstrate the potential complex mechanisms that mediate abnormal FGF-2 signaling, and highlight the many roles that FGF-2 may play in contributing to pulmonary vascular remodeling in shunt lambs.