Transforming growth factor-β plays divergent roles in modulating vascular remodeling, inflammation, and pulmonary fibrosis in a murine model of scleroderma

Abstract

The efficacy and feasibility of targeting transforming growth factor-β (TGFβ) in pulmonary fibrosis and lung vascular remodeling in systemic sclerosis (SSc) have not been well elucidated. In this study we analyzed how blocking TGFβ signaling affects pulmonary abnormalities in Fos-related antigen 2 (Fra-2) transgenic (Tg) mice, a murine model that manifests three important lung pathological features of SSc: fibrosis, inflammation, and vascular remodeling. To interrupt TGFβ signaling in the Fra-2 Tg mice, we used a pan-TGFβ-blocking antibody, 1D11, and Tg mice in which TGFβ receptor type 2 (Tgfbr2) is deleted from smooth muscle cells and myofibroblasts (α-SMA-Cre^ER;Tgfbr2^flox/flox). Global inhibition of TGFβ by 1D11 did not ameliorate lung fibrosis histologically or biochemically, whereas it resulted in a significant increase in the number of immune cells infiltrating the lungs. In contrast, 1D11 treatment ameliorated the severity of pulmonary vascular remodeling in Fra-2 Tg mice. Similarly, genetic deletion of Tgfbr2 from smooth muscle cells resulted in improvement of pulmonary vascular remodeling in the Fra-2 Tg mice, as well as a decrease in the number of Ki67-positive vascular smooth muscle cells, suggesting that TGFβ signaling contributes to development of pulmonary vascular remodeling by promoting the proliferation of vascular smooth muscle cells. Deletion of Tgfbr2 from α-smooth muscle actin-expressing cells had no effect on fibrosis or inflammation in this model. These results suggest that efforts to target TGFβ in SSc will likely require more precision than simply global inhibition of TGFβ function.

systemic sclerosis (SSc, also known as scleroderma) is a chronic and systemic autoimmune disease characterized by immunologic abnormalities, vasculopathy, and fibrosis in skin, lungs, and other internal organs (1). Pulmonary arterial hypertension (PAH) and pulmonary fibrosis, two major lung complications of SSc, are the leading causes of death in SSc patients (32). Patients with SSc-PAH have a worse prognosis than patients with idiopathic PAH or PAH related to other connective tissue diseases (6). Despite significant research efforts, mechanisms underlying these manifestations of SSc remain poorly understood, and therapeutics are severely limited. The absence of models fulfilling the full clinical picture of SSc has hindered progress in the development and testing of therapies for this disease (1). Although several animal models have been reported to recapitulate some aspects of SSc, few mouse models have the three prominent features of SSc: immune system disorders, vasculopathy, and fibrosis (35).

Fos-related antigen 2 (Fra-2) transgenic (Tg) mice have these three important pathological features (9). Immune cell infiltration precedes other lung pathological phenotypes and is followed by marked intimal hyperplasia and obliteration of pulmonary arteries and pulmonary fibrosis. Thus these mice provide an opportunity to explore the effects of potential therapeutic interventions in a complex phenotype characterized by pulmonary inflammation, fibrosis, and vasculopathy. Fra-2 is upregulated in human fibrotic lung tissue of idiopathic and SSc-associated pulmonary fibrosis (9) and in skin tissue of SSc patients (30), suggesting potential clinical significance of Fra-2 in SSc. Because the vasculopathy and fibrosis in the lungs of Fra-2 Tg mice become apparent almost simultaneously and develop in parallel, there might be some common molecular pathways regulating vasculopathy and fibrosis.

Transforming growth factor-β (TGFβ) is a major fibrogenic cytokine and a central mediator of fibrosis in multiple organs. Several lines of in vitro and in vivo evidence suggest that TGFβ also plays a key role in the development of tissue fibrosis in SSc (37). In this context, multiple drugs to inhibit the TGFβ pathway in SSc have been tested (38). In contrast, the significance of TGFβ in the pulmonary vasculopathy of SSc remains poorly understood. It has been suggested that TGFβ signaling could contribute to vascular pathology. TGFβ signaling has been shown to contribute to the differentiation, proliferation, and homeostatic functions of vascular smooth muscle cells (VSMCs) and endothelial cells (12). Mice lacking TGFβ receptor 2 (TGFβR2), specifically in VSMCs or endothelial cells, showed vascular defects in the yolk sac and embryonic lethality, suggesting that TGFβ in both cell types is essential for vascular development (5). TGFβ signaling is also involved in the pathogenesis of human idiopathic PAH. Mutations in the bone morphogenetic protein (BMP) type 2 receptor (BMPR2), a TGFβ1 receptor superfamily member that can oppose the effects of TGFβ, have been identified in 70% of cases of familial PAH, as well as in 10–40% of cases of idiopathic PAH (23). Several recent studies showed that TGFβ is upregulated in the lungs of experimental models of PAH and that blocking TGFβ reduces the severity of PAH (13, 22, 41). Therefore, TGFβ signaling and the molecular pathways that regulate TGFβ activity and signaling could be key factors in the link between pulmonary vasculopathy and fibrosis in SSc.

In this study we treated Fra-2 Tg mice with a blocking monoclonal antibody that targets all TGFβ isoforms and found that this treatment reduced the severity of lung vasculopathy. In contrast, the global inhibition of TGFβ did not ameliorate the severity of pulmonary fibrosis and exacerbated lung inflammation. Conditional deletion of TGFβR2 from smooth muscle cells resulted in a similar improvement in lung vasculopathy but did not change the severity of pulmonary fibrosis or inflammation. This study suggests that TGFβ contributes to vasculopathy in this model of SSc but that inhibition of TGFβ does not prevent or reverse fibrosis, perhaps because this intervention dramatically increases pulmonary inflammation.

MATERIALS AND METHODS

Mouse.

Fra-2 Tg mice were provided by Erwin F. Wagner (Spanish National Cancer Center, Madrid, Spain) (9). To determine the specific roles of TGFβ signaling in smooth muscle in vivo, we generated TGFβ receptor type 2 (Tgfbr2)^flox/flox mice (19) that also expressed α-SMA-Cre^ER (39). Ai14 (Rosa26-LSL-tdTomato) mice were obtained from the Jackson Laboratory (24). The genotypes of these transgenic mice were determined by PCR analysis using tail DNA. All the mice were on a pure C57BL/6 background or were backcrossed at least five times to C57BL/6. Mice were maintained in specific-pathogen-free conditions in the Animal Barrier Facility of the University of California, San Francisco. All animal experiments were approved by the Institutional Animal Care and Use Committee of the University of California, San Francisco.

Measurement of hydroxyproline.

Hydroxyproline was measured as previously described (14). Briefly, mouse lungs were incubated in 12 N HCl at 110°C for 18 h. Aliquots of the samples reconstituted in distilled water were added to 1.4% chloramine-T in 10% isopropanol and 0.5 M sodium acetate. Erlich's solution was added, and the samples were incubated at 60°C for 10 min. Absorbance at 562 nm was measured and adjusted according to standard curves.

Bronchoalveolar lavage fluid analysis.

Tracheas were cannulated, and lungs were lavaged three times with 1 ml of PBS. The number of cells in the lavage fluid were counted using a hemocytometer. The bronchoalveolar lavage (BAL) fluid was processed in a cytocentrifuge (Cytospin) at 500 g for 5 min, and differential leukocyte counts were performed on a smear stained with Hema3 (Fisher Scientific, Pittsburgh, PA).

Single-cell tissue dissociation for flow cytometry.

Tissue dissociation was conducted as described previously (33). Briefly, lung tissues were excised, minced with scissors, and then digested in Dulbecco's modified Eagle's medium (Invitrogen) containing Liberase (0.13 IU/ml; Sigma-Aldrich, St. Louis, MO) at 37°C for 20 min. Single-cell suspensions were then prepared using a gentleMACS dissociator (Miltenyi Biotec) as described in the manufacturer's instructions. The cell suspension was passed through 100- and 40-μm cell strainers and suspended in fluorescein-activated cell sorting (FACS) buffer (PBS supplemented with 2% fetal bovine serum). After live/dead staining with 4′,6-diaminido-2-phenylindole (DAPI; Sigma-Aldrich), live single cells with reporters were sorted using a cell sorter (FACSARIA, BD Biosciences).

Quantitative RT-PCR.

Total RNA was isolated from lung tissues or cells using an RNeasy kit (Qiagen, Venlo, The Netherlands). cDNA was analyzed by SYBR Green RT-PCR with a thermocycler (model 7900HT, Applied Biosystems) and normalized to Gapdh or 18S expression. The primers were as follows: Gapdh [AGG TCG GTG TGA ACG GAT TTG (forward) and GGG GTC GTT GAT GGC AAC A (reverse)], 18S [ACG GAA GGG CAC CAC CAG G (forward) and CAC CAA CTA AGA ACG GCC ATG C (reverse)], Tgfbr2 [TTA ACA GTG ATG TCA TGG CCA GCG (forward) and AGA CTT CAT GCG GCT TCT CAC AGA (reverse)], and Bmpr2 [TTG GGA TAG GTG AGA GTC GAA T (forward) and TGT TTC ACA AGA TTG ATG TCC CC (reverse)].

Western blotting.

Lung tissues and cells were lysed in buffer supplemented with protease and phosphatase inhibitors as previously described (29). Lysates containing equal amounts of protein were subjected to SDS-PAGE and transferred to Immobilon-P membranes (Millipore, Billerica, MA). The membranes were probed using primary antibodies [phosphorylated Smad3 (p-Smad3) (21), phosphorylated Smad1/5 (p-Smad1/5) (21), and GAPDH (17) (Cell Signaling Technology, Danvers, MA) and BMPR2 (16) (Proteintech, Rosemont, IL)] followed by peroxidase-conjugated secondary antibodies. Immunoreactive bands were visualized using ECL Western blotting substrate (PerkinElmer, Wellesley, MA). For densitometry, blots were analyzed using ImageJ software.

Immunohistochemistry and immunofluorescence.

Mouse paraffin sections were stained by indirect immunohistochemistry as described previously with minor modifications (36). Briefly, deparaffinized and rehydrated tissue sections were heated in citrate buffer (10 mM, pH 6.0) and blocked with 10% normal serum and 0.5% bovine serum albumin. Sections were incubated with anti-α-smooth muscle actin (α-SMA; clone 1A4, Sigma-Aldrich) overnight at 4°C and subsequently incubated with biotin anti-mouse immunoglobulin antibody (Vector Laboratories, Burlingame, CA) for 20 min. Signals were visualized using avidin-biotin peroxidase (Vectastain Elite kit) with 3,3′-diaminobenzidine (Sigma). For immunofluorescence staining, Alexa 488- or Cy3-conjugated anti-α-SMA (clone 1A4, Sigma-Aldrich) was applied to sections for 1 h at room temperature. Other primary antibodies were as follows: rabbit anti-Ki67, rabbit anti-p-Smad3 (Abcam, Cambridge, UK), and rabbit anti-cleaved caspase-3 (Cell Signaling Technology). For frozen tissue sections, mouse lungs were inflated through the trachea and fixed in 4% paraformaldehyde overnight at 4°C. Thereafter, 4% paraformaldehyde was replaced with PBS containing 20% sucrose, and the sections were snap-frozen in OCT embedding compound (Tissue Tek) and stored at −80°C until use. Sections (10 μm) were cut using a cryostat microtome (Leica, Deerfield, IL), placed onto gelatin-coated slides, and air-dried. Sections were blocked for 30 min at room temperature in blocking buffer (PBS containing 5% serum, 0.5% bovine serum albumin, and 0.1% Triton X-100) and then stained with the primary antibodies.

Evaluation of pulmonary vascular remodeling in mouse lung.

Expansion of the vascular smooth muscle compartment was evaluated by immunohistochemical staining of α-SMA (Sigma-Aldrich). The α-SMA-positive area was evaluated in tissue sections as shown in Fig. 1E. Briefly, twenty 20- to 80-μm-diameter peripheral pulmonary arteries were randomly chosen from each blinded lung section, and the α-SMA-positive area was calculated and normalized to total vessel area (percent α-SMA-positive area) using ImageJ software. Mean percent α-SMA-positive area values were calculated for each mouse. After calculation of the mean value, blinded lung sections were unblinded to be categorized into an appropriate treatment group.

Fig. 1.

Vascular remodeling, fibrosis, and inflammation in lungs from Fos-related antigen 2 (Fra-2) transgenic (Tg) mice. A: hematoxylin-eosin (H and E)- and Sirius Red-stained lungs from wild-type (WT) and Fra-2 Tg mice at 14 wk of age. Marked vessel wall thickness and resultant narrowing of the lumen were observed in lungs from Fra-2 Tg compared with WT mice (arrowheads). In Sirius Red-stained images, collagen deposition (red area) was prominent around vessels and in the subpleural area of lungs from Fra-2 Tg mice. Scale bar = 100 μm. B: hydroxyproline content of lungs from WT and Fra-2 Tg mice at 14 wk of age. Values are means ± SE; n = 7–11 in each group. C: cell counts and representative images of bronchoalveolar lavage fluid from 15- to 18-wk-old WT and Fra-2 Tg mice. Values are means ± SE; n = 4–5 in each group. *P < 0.05 vs. WT. Arrowheads indicate increased number of granulocytes in lavage from Fra-2 Tg mice. MΦ, macrophage; Lym, lymphocyte; Gra, granulocyte. D: representative α-smooth muscle actin (α-SMA)-stained (brown) lung sections from WT and Fra-2 Tg mice at 14 wk of age. PA, pulmonary artery; Br, bronchus. Scale bar = 10 μm. E: schema for evaluation of pulmonary vascular remodeling. Blinded lung sections from WT or Fra-2 Tg mice were stained with Cy3-conjugated α-SMA antibody (left). α-SMA fluorescent signals were adjusted by setting lower and upper threshold values in ImageJ software (middle). Thereafter, α-SMA-positive area (X) was calculated and divided by total vessel area (X + Y) for normalization (right). F: proportion of α-SMA-positive area in total vessel area. Each dot represents mean value of α-SMA-positive area/total vessel area calculated from 20 individual vessels from each mouse. G: α-SMA (green) and Ki67 (red) immunostaining of lung sections from 14-wk-old WT and Fra-2 Tg mice. DAPI, 4′,6-diaminido-2-phenylindole. H: proportion of Ki67-positive cells in vascular smooth muscle cells (VSMCs) was significantly increased in lungs from Fra-2 Tg compared with WT mice. I: α-SMA, CD31, and cleaved caspase-3 immunostaining of representative lung sections from WT and Fra-2 Tg mice at 14 wk of age. Scale bar = 25 μm.

The number of Ki67 and α-SMA double-positive cells in blood vessels was calculated and divided by the total number of α-SMA-positive cells. A minimum of five fields from each lung were randomly acquired, and an average of 150 α-SMA-positive pulmonary artery smooth muscle cells were analyzed. For evaluation of TGFβ signaling in lung tissues, DAPI-positive nuclei in smooth muscle cells were traced, and mean signal intensity was calculated using ImageJ software. Ten random fields from each section were analyzed.

Statistical analysis.

Values are means ± SE. The significance of differences between two sample means was determined by two-tailed Student's t-test or Mann-Whitney U-test. P < 0.05 was considered statistically significant. Statistical analyses were carried out using SPSS 17.0 for Windows (SPSS, Chicago, IL).

RESULTS

Fra-2 Tg mice spontaneously develop pulmonary fibrosis, vascular remodeling, and granulocyte infiltration in alveoli.

As reported in several previous studies, we found that Fra-2 Tg mice spontaneously develop pulmonary vascular remodeling, pulmonary inflammation, and pulmonary fibrosis. Beginning at 10 wk of age, mutant mice developed pronounced pulmonary arterial vascular remodeling, characterized by markedly increased wall thickness and resultant narrowing of the arterial lumen (Fig. 1A). Subsequently, they developed pulmonary fibrosis, which became prominent and easily detectable histologically by 14 wk of age (Fig. 1A). At 14 wk of age, hydroxyproline content in the lungs of Fra-2 Tg mice was significantly increased compared with littermate wild-type (WT) mice (Fig. 1B), supporting the evidence of significant fibrotic changes in lungs from Fra-2 Tg mice. Pulmonary fibrosis was prominent in the perivascular area but was also observed in the peripheral, subpleural lung (Fig. 1A).

Fra-2 Tg mice also showed augmented inflammation in their lungs. Histological analysis revealed areas of inflammatory cell infiltration, most prominently around the distal vasculature and airways, in the lungs of Fra-2 Tg mice starting at 6–10 wk of age (Fig. 1A). In accordance with this finding, we observed significant increases in total cell numbers, macrophages, lymphocytes, and, especially, granulocytes in BAL fluid from Fra-2 Tg mice (Fig. 1C).

α-SMA staining revealed significant expansion of the vascular smooth muscle compartment in peripheral pulmonary arteries of lungs from Fra-2 Tg mice (Fig. 1, D–F). Furthermore, the proportion of Ki67-positive smooth muscle cells was significantly increased in lungs from Fra-2 Tg mice compared with WT controls (Fig. 1, G and H). To evaluate apoptosis, we stained lungs with antibody to cleaved caspase-3 but could not detect apoptotic cells in the vascular wall of Fra-2 Tg mice or WT controls, whereas cleaved caspase-3 was easily detected in the inflammatory cells surrounding blood vessels in Fra-2 Tg mice (Fig. 1I). These data suggest that increased vascular smooth muscle and luminal narrowing in lungs from Fra-2 Tg mice are principally due to the overproliferation of VSMCs.

Inhibition of TGFβ signaling exacerbates inflammation but does not ameliorate pulmonary fibrosis in lungs of Fra-2 Tg mice.

TGFβ is well known as a prominent driver of tissue fibrosis. Although several lines of evidence suggest that disrupted TGFβ signaling could contribute to tissue pathology in SSc, the efficacy and safety of inhibiting TGFβ signaling in SSc remain controversial (38). To evaluate the possible contributions of TGFβ signaling to the lung pathologies observed in Fra-2 Tg mice, we employed the pan-TGFβ-neutralizing monoclonal antibody 1D11, which inhibits TGFβ1, -2, and -3 (7). Because this antibody is not very potent and has a short half-life in vivo, we administered 1D11 (n = 8) or control mouse IgG1 (n = 8) intraperitoneally (30 mg·kg⁻¹·day⁻¹) for 4 wk from 10 to 14 wk of age, when fibrosis and vascular remodeling become apparent in the lungs of Fra-2 Tg mice (Fig. 2A). One of eight Fra-2 Tg mice treated with 1D11 died on the 18th day of the treatment, while all the mice in the IgG group survived the treatment course. This administration protocol diminished p-Smad3, a direct downstream product of TGFβ signaling, in total lung homogenates of Fra-2 Tg mice (Fig. 2B). Contrary to our expectations, however, pharmacological blockade of TGFβ did not ameliorate lung fibrosis in Fra-2 Tg mice. At 14 wk of age, no apparent differences in the severity of fibrosis were observed histologically (Fig. 2C) or by hydroxyproline measurements (Fig. 2D) between the two treatment groups.

Fig. 2.

Inhibition of transforming growth factor-β (TGFβ) signaling does not ameliorate pulmonary fibrosis but exacerbates inflammation. A: schema of 1D11 treatment. Fra-2 Tg mice were treated with 1D11 or IgG1 (30 mg·kg⁻¹·day⁻¹ ip) for 4 wk. BAL, bronchoalveolar lavage. B: Western blot and densitometric analysis showing decreased phosphorylated Smad3 (p-Smad3) in lungs after 4 wk of treatment with 1D11. C: hematoxylin-eosin- and Sirius Red-stained representative lung sections from 14-wk-old Fra-2 Tg mice treated with 1D11 or control IgG. Scale bar = 100 μm. D: hydroxyproline content in the lungs was significantly increased in Fra-2 Tg compared with WT mice, but no apparent differences were observed between 1D11 and IgG treatment groups. NS, not significant. E: number of total cells, macrophages (MΦ), lymphocytes (Lym), and granulocytes (Gra) in bronchoalveolar lavage fluid from WT and Fra-2 Tg mice at 12 wk of age. Fra-2 Tg mice treated with 1D11 showed increased cell numbers compared with those treated with control IgG.

On the other hand, the severity of lung inflammation was significantly increased in the lungs of Fra-2 Tg mice treated with 1D11, as detected from hematoxylin-eosin-stained sections (Fig. 2C) and by quantification of cells obtained by BAL (Fig. 2E). These data suggest that global blockade of TGFβ signaling in Fra-2 Tg mice results in exacerbation of lung inflammation but no detectable reduction in pulmonary fibrosis. Because the exaggerated inflammation caused by TGFβ inhibition might be expected to exacerbate pulmonary fibrosis, we cannot rule out off-setting effects of this treatment on inflammation and fibrosis.

Inhibition of TGFβ signaling ameliorates pulmonary vascular remodeling in Fra-2 Tg mice.

It has been reported that TGFβ signaling is upregulated in the pulmonary vasculature of PAH patients (3). We did not observe significant differences in p-Smad3 signals in total lungs between Fra-2 Tg and WT mice (Fig. 2B). By quantifying p-Smad3 from multiple arterial smooth muscle cells, however, we detected a small, but significant, increase in TGFβ signaling in the lungs of Fra-2 Tg mice that was decreased by treatment with 1D11 (Fig. 3, A and B). In addition, blocking TGFβ signaling by 1D11 ameliorated the expanded α-SMA-positive area in the pulmonary arteries of lungs from Fra-2 Tg mice, despite the increased inflammation in the lungs (Fig. 3C). These findings suggest that increased TGFβ signaling is at least partially responsible for the development of pulmonary vascular remodeling in Fra-2 Tg mice.

Fig. 3.

Global inhibition of TGFβ signaling ameliorates pulmonary vasculopathy in Fra-2 Tg mice. A: representative lungs from WT and Fra-2 Tg mice treated with control IgG or 1D11 and immunostained with p-Smad3 and α-SMA. Scale bar = 25 μm. B: signal intensity of p-Smad3 in VSMCs. C: severity of vasculopathy assessed by measurement of α-SMA-positive area in the vessels (see materials and methods). Values are means ± SE of 20 vessels from an individual mouse lung. *P < 0.05, **P < 0.01.

Loss-of-function mutations of BMPR2, a member of the TGFβ receptor superfamily, is a common cause of familial PAH; therefore, BMP signaling is believed to be crucial for normal homeostasis of the pulmonary artery intima (26, 28). Immunostaining of lung sections revealed that p-Smad1/5, directly downstream of BMPR2 signaling, was localized and easily detected in vascular cells of WT mice (Fig. 4A). Expression of p-Smad1/5 was also detectable in the vasculature of Fra-2 Tg mice (Fig. 4A), but it was decreased compared with WT mice (Fig. 4B). Expression of BMPR2 (immediately upstream of Smad1/5) was significantly decreased in the lungs of Fra-2 Tg mice at protein and mRNA levels (Fig. 4, B and C). This reduction in BMPR2 expression was not affected by TGFβ blockade with 1D11 (Fig. 4, B and C). Since BMP signaling has been shown to be functionally antagonistic to TGFβ signaling in a number of biological systems (15, 26), it is conceivable that the TGFβ-dependent vascular remodeling we observed in Fra-2 Tg mice is a consequence of this reduction in BMPR2.

Fig. 4.

Bone morphogenetic protein receptor type 2 (BMPR2) expression is decreased in lungs from Fra-2 Tg mice. A: representative lungs from 14-wk-old WT and Fra-2 Tg mice immunostained with p-Smad1/5, α-SMA, and CD31. In top row, only 2nd antibody was added without primary anti-p-Smad1/5 antibodies as a control. B: immunoblots of total lung lysates from 14-wk-old WT and Fra-2 Tg mice treated with control IgG or 1D11 for 4 wk. Mouse brain lysate is shown as positive control of BMPR2 expression. **P < 0.01 (by Student's t-test). C: relative mRNA expression of BMPR2 in the total lungs from 12-wk-old WT and Fra-2 Tg mice treated with IgG or 1D11 for 2 wk. Values are means ± SE; n = 4 in each group. *P < 0.05 (by Student's t-test).

Deletion of TGFβ signaling from α-SMA-expressing cells did not change the severity of pulmonary fibrosis or inflammation in Fra-2 Tg mice.

Inhibition of all TGFβ function has therapeutic limitations because of the pleiotropic role of this growth factor in normal tissue homeostasis, cancer, and inflammation. 1D11 dramatically augmented pulmonary inflammation in lungs from Fra-2 Tg mice, and this increased inflammatory response could have exacerbated the severity of pulmonary fibrosis. In an attempt to examine the role of TGFβ independent from effects on inflammation in Fra-2 Tg mice, we generated Tg mice in which TGFβR2 is specifically deleted in smooth muscle cells by tamoxifen treatment of mice homozygous for a conditional allele of Tgfbr2 (Tgfbr2^flox/flox) that also expressed α-SMA-Cre^ER in the presence or absence of the Fra-2 transgene.

Since TGFβ signaling is crucial for normal vascular development and homeostasis, we started tamoxifen treatment at 8 wk of age (Fig. 5A). To evaluate the specificity and efficiency of Cre-mediated recombination in response to our tamoxifen treatment regimen, we crossed α-SMA-Cre^ER mice with Ai14 (Rosa-CAG-LSL-tdTomato-WPRE) mice in which TdTomato expression is regulated by Cre-induced recombination (24). At 1 wk after tamoxifen treatment (2 mg·mouse⁻¹·day⁻¹ for 3 days), we examined Tdtomato expression in lung sections. Tdtomato was specifically detected in vascular and airway smooth muscle cells, and Tdtomato expression closely matched α-SMA antibody staining (Fig. 5B). Moreover, Acta2, which is expressed in smooth muscle, was highly expressed in Tdtomato-positive cells isolated from the lungs of α-SMA-Cre^ER;Ai14^flox/+ mice, but not in Tdtomato-negative cells (Fig. 5, C and D). Tgfbr2 expression in the isolated Tdtomato-positive cells was significantly decreased in the Tgfbr2 mutant mice (α-SMA Cre^ER;Tgfbr2^flox/flox;Ai14^flox/+) compared with Tgfbr2 WT mice (α-SMA-Cre^ER;Tgfbr2^+/+;Ai14^flox/flox; Fig. 5E). These data suggest that, in this line, Tgfbr2 expression is specifically and efficiently decreased in α-SMA-expressing cells.

Fig. 5.

Specific deletion of TGFβ receptor type 2 (TGFβR2) from α-SMA-expressing cells did not affect inflammation or fibrosis in lungs from Fra-2 Tg mice. A: vascular remodeling, fibrosis, and inflammation in mice treated with tamoxifen intraperitoneally every 2 wk from 8 to 12 wk of age. B: efficiency of Cre recombination 2 wk after tamoxifen treatment in lung sections from α-SMA-Cre^ER;Ai14^flox/+ mice. TdTomato expression was specifically detected in vascular and airway smooth muscle cells. Scale bar = 25 μm. C: isolation of tdTomato-expressing cells from α-SMA-Cre^ER;Ai14^flox/+ mice. Mouse lungs were dissociated with collagenase-containing solution (Liberase). Among DAPI-negative singlet live cells, phycoerythrin (PE)-positive (P3) or -negative cells (P4) were sorted by PE and Pacific Blue laser. D: Acta2 expression normalized by GAPDH expression. Acta2 expression was significantly higher in cells isolated from the PE-positive (P3) than PE-negative (P4) population. E: semiquantitative PCR to determine Tgfbr2 expression in isolated tdTomato-positive smooth muscle cells. 18S expression was used as a control. F: cell counts in bronchoalveolar lavage fluids. Numbers of macrophages (MΦ), lymphocytes (Lym), and granulocytes (Gra) were similarly increased in lungs from Fra-2 Tg mice with and without TGFβR2 deletion in smooth muscle cells. All the mice are Tgfbr2^flox/flox and 14 wk of age (n = 5–9 in each group). G: hydroxyproline content in left lungs from WT and Fra-2 Tg mice at 14 wk of age. Fra-2 Tg lungs showed significantly increased hydroxyproline content, but this was not ameliorated by deletion of TGFβR2 from α-SMA-expressing cells (n = 4–16 in each group).

BAL cell numbers were comparable in Fra-2 Tg mice with and without conditional TGFβR2 deletion (Fig. 5D), suggesting that TGFβR2 deletion in α-SMA-expressing cells did not affect the severity of inflammation in the lungs of Fra-2 Tg mice. The severity of pulmonary fibrosis induced by overexpression of Fra-2 was also unaffected by loss of TGFβR2 in α-SMA-expressing cells (Fig. 5E).

TGFβ signaling in smooth muscle cells contributes to vascular remodeling in Fra-2 Tg mice.

Next, we evaluated the specific roles of TGFβ signaling in the pulmonary VSMCs of Fra-2 Tg mice. Conditional deletion of TGFβR2 starting at 8 wk of age did not change the gross appearance of pulmonary arteries from that of Fra-2 and WT mice at 14 wk of age (data not shown), and no difference was observed in smooth muscle cell area of pulmonary arteries compared with Cre-negative controls (Fig. 6, A and B). On the other hand, in Fra-2 Tg mice, deletion of TGFβ signaling from smooth muscle cells diminished the abnormal expansion of smooth muscle cell area (Fig. 6, A and B). In accordance with this finding, the proportion of Ki67-positive cells was significantly decreased in pulmonary artery smooth muscle cells of Fra-2 Tg mice following deletion of TGFβR2 in these cells (Fig. 6C). These data suggest that direct TGFβ signaling in pulmonary artery smooth muscle cells significantly contributes to the overproliferation of these cells and consequent vasculopathy in the lungs of Fra-2 Tg mice.

Fig. 6.

TGFβ signaling in smooth muscle cells contributes to development of vasculopathy in Fra-2 Tg mice. A: representative immunostaining of pulmonary arteries from WT and Fra-2 Tg mice with/without α-SMA Cre^ER (all the mice are Tgfbr2^flox/flox and 14 wk of age) with α-SMA and Ki67. Scale bar = 25 μm. B: α-SMA-positive area in each vessel normalized by total vessel area. Values are means ± SE of 20 different vessels randomly chosen from blinded lung sections of individual mice (n = 6–16 in each group). C: proportion of Ki67-positive cells in pulmonary artery smooth muscle cells calculated as described in materials and methods. Values are means ± SE from an individual mouse (n = 5–9 in each group).

DISCUSSION

In the present study we found that the occlusive vasculopathy in Fra-2 Tg mice could be partially prevented by systemic treatment with a monoclonal antibody (1D11) that binds to and inhibits the function of all three mammalian TGFβ isoforms (7). Protection by TGFβ inhibition was associated with a significant reduction in the increased VSMC proliferation in these mice. Our observation that the protection from vasculopathy and the decrease in smooth muscle cell proliferation could be recapitulated by conditional deletion of TGFβR2 from smooth muscle cells further supports the important role of TGFβ in this process and suggests that protection by TGFβ inhibition is due to a reduction in TGFβ signaling in VSMCs. Our finding that Fra-2 overexpression was associated with a significant reduction in BMPR2, a well-described antagonist of TGFβ function (15) that is frequently mutated in patients with familial PAH (28), provides a possible explanation for augmented effects of TGFβ in this model and suggests that our findings could have clinical relevance. Reich et al. (30) suggested that Fra-2 might be a downstream mediator of platelet-derived growth factor (PDGF) and TGFβ in the fibroblasts in SSc. In addition, Maurer et al. (25) also reported that PDGF receptor signaling pathways were increased in the lungs of Fra-2 Tg mice and that the tyrosine kinase inhibitor nilotinib ameliorated vasculopathy and fibrosis in these mice. These findings, along with our results, suggest that multiple growth factors, including TGFβ and PDGF, coordinately contribute to the development of lung pathologies in Fra-2 Tg mice.

However, our finding that systemic TGFβ inhibition also caused a marked increase in the granulocytic pulmonary inflammation that characterizes pulmonary pathology in Fra-2 Tg mice highlights the risks of global inhibition of TGFβ for the treatment of diseases characterized by ongoing tissue inflammation (2, 27, 34, 40), a concern that might be relevant for treatment of SSc. Because enhanced pulmonary inflammation might be expected to worsen pulmonary fibrosis (4, 10, 18), our failure to see any protective effect of TGFβ inhibition on pulmonary fibrosis in Fra-2 Tg mice needs to be interpreted with caution and does not rule out a contribution of TGFβ to fibrosis in this model.

Fra-2 Tg mice with conditional deletion of TGFβR2 in smooth muscle cells were protected from vasculopathy, but the degree of pulmonary inflammation in these mice was the same as that in Fra-2 Tg mice with intact TGFβ signaling. These findings allow us to conclude that TGFβ directly affects vascular smooth muscle proliferation and contributes to vasculopathy independent of effects on pulmonary inflammation. Because the α-SMA promoter can also drive Cre expression in a subpopulation of pathological fibroblasts, we had hoped that this line might also allow us see a role for TGFβ signaling in the pulmonary fibrosis in these mice. However, our recent work (33) and work by others (31) showed that the majority of collagen-producing fibroblasts in the lungs of mice treated with bleomycin do not express α-SMA and that the system we used to drive Cre expression did not result in Cre-mediated recombination in most of the collagen-producing cells (33). Therefore, we are again unable to definitively determine from our current work whether TGFβ signaling in collagen-producing fibroblasts significantly contributes to pulmonary fibrosis in Fra-2 Tg mice. Furthermore, inducible Fra-2 expression just after birth (Rosa-rtTA;tetO-Fra-2 on a C57BL/6 background) also results in pulmonary vascular remodeling, but not pulmonary inflammation or fibrosis (personal communication, Erwin Wagner), providing further support for distinct mechanisms underlying vasculopathy and fibrosis in this model.

In a study of the effects of the TGFβ-blocking antibody CAT-192, Denton et al. (8) did not show significant clinical benefit at any of the end points measured in patients with SSc. However, they were unable to determine whether the dose of CAT-192 used was sufficient to inhibit TGFβ signaling in relevant cells (8). In the present study we used quite high concentrations of the TGFβ-blocking antibody (30 mg/kg) and administered it at frequent intervals (daily). We chose this regimen to optimize TGFβ inhibition based on preliminary (unpublished) findings that even this high dose did not completely inhibit TGFβ signaling in mice. Indeed, our findings from p-Smad3 staining in the lungs of 1D11-treated mice showed that while we were able to achieve a partial reduction in TGFβ signaling with this dosing regimen, TGFβ signaling was still significant. The fact that we achieved a similar degree of protection from vasculopathy in mice treated with 1D11 and mice with TGFβR2 deleted in smooth muscle cells suggests that the degree of inhibition we achieved with this aggressive 1D11 treatment regimen was sufficient to inhibit the functional effects of TGFβ on VSMC proliferation. The marked increase in granulocytic inflammation in 1D11-treated Fra-2 Tg mice suggests that this regimen was also sufficient to modulate the suppressive effects of TGFβ on tissue inflammation. However, it is clear that even the aggressive regimen of 1D11 treatment we used was not sufficient to completely inhibit TGFβ signaling in the lungs, and it remains possible that inadequate access of this antibody to pathological lung fibroblasts prevented us from identifying an important role of TGFβ in driving collagen production from these cells.

An important unanswered question raised by the current study is how TGFβ increases VSMC proliferation and contributes to vasculopathy in Fra-2 Tg mice. A recent study using mice in which TGFβR2 was genetically ablated from smooth muscle cells (Myh11-Cre^ER;Tgfbr2^flox/flox) reported overproliferation of aortic smooth muscle and dilation and dissections in the thoracic aorta (20). These findings suggest that TGFβ function in VSMCs can differ at different sites in the vasculature. One clue in the current study that might explain these differential effects is our finding that Fra-2 overexpression leads to a reduction in expression of BMPR2 in the pulmonary vasculature. Since signaling through BMPR2 has been considered to antagonize some of the consequences of TGFβ signaling in a number of cell types, including vascular smooth muscle (15, 26), it is conceivable that even a small increase in expression or activation of TGFβ (e.g., in response Fra-2-induced tissue inflammation or injury as a consequence of inflammation) could be amplified in the setting of reduced BMP signaling. Recently, similar decreases in BMP signaling were reported in other animal models of SSc associated with pulmonary vascular remodeling and in the lungs of SSc patients (11). The previously reported finding that there is no increase in expression of any TGFβ isoform in the lungs of Fra-2 Tg mice would be consistent with the idea that the important role of TGFβ in vascular remodeling in this model could be due to a loss of normal homeostatic inhibitory pathways, rather than an increase in TGFβ production or activity. The fact that loss-of-function mutations in BMPR2 are the most common cause of familial PAH (28) suggests that our findings could have clinical relevance. However, the findings in the current study are simply associative: we have not provided direct functional evidence that a reduction in BMPR2 plays a causative role in the vascular remodeling in Fra-2 Tg mice. Such evidence could be provided, for example, if transgenic overexpression of BMPR2 in smooth muscle could rescue the vasculopathy phenotype in these mice, but such an investigation is beyond the scope of the current study.

In summary, we have shown that systemic blockade of TGFβ with a blocking monoclonal antibody partially protects mice from pulmonary vasculopathy in the Fra-2 Tg model of SSc but that this protection comes at the considerable cost of exaggerated pulmonary inflammation. Furthermore, even high, frequent doses of a TGFβ-blocking antibody have no protective effect on pulmonary fibrosis in this model. Thus our results reinforce the need for caution in attempts to globally block TGFβ to treat diseases characterized by tissue fibrosis in the setting of ongoing inflammation.

GRANTS

This work was supported in part by grants from the Uehara Memorial Foundation, Kanae Foundation for the Promotion of Medical Science (K. Tsujino), the Scleroderma Research Foundation (D. Sheppard), and National Heart, Lung, and Blood Institute Grant HL-53949.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

K.T., N.I.R., A.A., and X.R. performed the experiments; K.T. analyzed data; K.T., N.I.R., and D.S. interpreted the results of the experiments; K.T. prepared the figures; K.T. and D.S. drafted the manuscript; K.T. and D.S. edited and revised the manuscript; K.T., N.I.R., A.A., X.R., and D.S. approved the final version of the manuscript.