WNT1-inducible signaling protein–1 mediates pulmonary fibrosis in mice and is upregulated in humans with idiopathic pulmonary fibrosis (original) (raw)
Enhanced expression of WISP1 in proliferating ATII cells in experimental lung fibrosis. We initially characterized primary ATII cells in lung fibrosis by investigating the morphology and proliferative capacity of freshly isolated ATII cells from mice subjected to bleomycin-induced lung fibrosis, as well as from time-matched, saline-treated control mice. A similar purity was observed when isolating ATII cells from control or bleomycin-treated mouse lungs (95% ± 3% of pro-surfactant protein C–positive [SPC-positive] and α-SMA–negative cells) (Figure 1A). Morphological analysis revealed the expression of the epithelial marker proteins SPC, tight junction protein 1 (TJP1), and E-cadherin (ECAD) (Figure 1B; for secondary antibody controls, see Supplemental Figure 1; supplemental material available online with this article; doi:10.1172/JCI33950DS1) as well as occludin (OCCL) and pan-cytokeratin (panCK) (Supplemental Figure 1B) in both cell isolations. ATII cells isolated from the lungs of bleomycin-treated mice, however, demonstrated a significant increase in cell proliferation, as assessed by Ki67 staining and [3H]thymidine incorporation (186%–225% of control ATII cells, 95% CI) (Figure 1, C and D). In accordance with these observations, ATII cells from bleomycin-treated mice exhibited increased mRNA levels of the proliferation markers Ki67, cyclin G1 (Ccng1), and Ccnb2, when compared with time-matched, saline-treated mice (Figure 1E). To uncover potential gene regulatory networks driving increased ATII cell proliferation, we next performed whole genome microarray analysis comparing gene signatures of freshly isolated ATII cells from bleomycin-treated mice with those from saline-instilled mice. As depicted in Figure 2 and Supplemental Figure 2, several gene families were differentially expressed in ATII cells obtained from fibrotic lungs, the details of which are outlined in the Supplemental Data. In accordance with our initial observations, ATII cells isolated from fibrotic mouse lungs demonstrated a remarkable upregulation of proliferative mediators and/or markers, such as oncogenes and cell cycle–associated genes (Supplemental Figure 2). Furthermore, the ATII cell gene expression profile also indicated an enrichment of inflammatory stimuli and proinflammatory cytokines in experimental lung fibrosis, suggesting that, at least in the mouse, this is part of the alveolar epithelial cell response to fibrogenic stimuli.
Enhanced proliferative capacity of ATII cells in experimental lung fibrosis. The purity (A) and phenotype (B) of ATII cell isolations from saline- or bleomycin-treated (Bleo) mice, 14 days after instillation, was analyzed by immunofluorescence staining. ATII cells were fixed directly after isolation (cytocentrifuge preparations) and stained with antibodies against the ATII cell marker SPC (original magnification, ×40, scale bar: 10 μm) or the (myo)fibroblast marker α-SMA (insets, magnification, ×10) (A) or fixed after 24 hours of attachment and subsequently stained with antibodies against SPC, TJP1, or ECAD (original magnification, ×40) (B). (C) Double immunostaining for panCK (green) and Ki67 (red) was performed in primary ATII cells from saline- or bleomycin-treated mice, 14 days after instillation (original magnification, ×40). Nuclei were visualized by DAPI staining (insets; original magnification, ×40). All stainings are representative of at least 3 independent experiments. (D) ATII cell proliferation was analyzed in primary cells isolated from mice 7 or 14 days after instillation with bleomycin by [3H]thymidine incorporation. Data are presented as fold change in [3H]thymidine incorporation compared with saline-instilled controls (n = 10 per group). (E) mRNA levels of the proliferation markers Ki67, Ccng1, and Ccnb2 were analyzed by qRT-PCR using primary ATII cells and plotted as log-fold increase (ΔΔCt) of mRNA levels in bleomycin- versus saline-treated mice, 14 days after instillation (n = 6 each). Results are presented as mean ± SEM; *P < 0.05, **P < 0.02.
Increased mRNA expression of Wisp1 and WNT signaling components in ATII cells isolated from bleomycin-treated mice. (A) ATII cell gene expression profiles were analyzed by whole genome expression analysis using RNA from freshly isolated ATII cells from saline- or bleomycin-treated mouse lungs 14 days after administration. Red and green indicate increased and decreased gene expression levels, respectively, in ATII cells isolated from bleomycin- versus saline-treated mice. Columns represent individual samples, including dye-swap experiments. Selected genes are represented in rows. Detailed description of the whole genome expression analysis is given in the Supplemental Data. (B) Confirmation of microarray results was performed for selected genes in freshly isolated ATII cells (n = 6), as well as in ATII cells 72 hours after isolation (n = 3) by qRT-PCR, as indicated. The following genes were analyzed: secreted frizzled-related protein 1 (Sfrp1), inhibin beta A (Inhba), found in inflammatory zone 1 (Fizz1), secreted phosphoprotein 1 (Spp1), Wisp1, cadherin 16 (Cdh16), and potassium voltage-gated channel subfamily E member 2 (Kcne2). Results are presented as mean ± SEM; **P < 0.02 for all bars, compared with ATII cells isolated from saline-treated mice. (C and D) The mRNA levels of the WNT target gene Mmp7, the WNT ligands Wnt1, Wnt2, Wnt3a, Wnt7b, and Wnt10b (C), the receptors frizzled 1 (Fzd1), Fzd2, and Fzd4, and the intracellular signal transducers Ctnnb1, Gsk3b, and Tcf4 (D) were assessed in ATII cells isolated from bleomycin- and saline-treated mice (n = 6 each) by qRT-PCR. Results are presented as mean ± SEM. *P < 0.05, **P < 0.02.
Differentially expressed transcripts also included genes that have previously been reported to be upregulated in bleomycin-induced lung fibrosis and IPF, including Spp1 (16, 17), Timp1 (18), Sfrp1 (19), and Pai1 (20). To confirm these findings, we further investigated the gene expression profiles in an independent set of freshly isolated and short term–cultured (48 hours) ATII cells by quantitative RT-PCR (qRT-PCR) (Figure 2B), and the findings were essentially the same as in the microarray analysis. Of interest, the expression of genes of the WNT signaling pathway (Wnt10a, Sfrp1, Tcf4, Ccnd1) was upregulated in ATII cells during bleomycin-induced lung fibrosis. In particular, expression of Wisp1, a member of the recently described CCN family of secreted signaling molecules (21, 22), was highly upregulated (Figure 2B).
Increased expression of WNT/β-catenin signaling molecules in lung epithelial cells during experimental lung fibrosis. The WNT family of highly conserved secreted growth factors is essential to organ development and known to determine epithelial cell fate (23, 24). The canonical WNT signaling pathway, or β-catenin–dependent pathway, regulates gene transcription by stabilization of β-catenin. Upon WNT stimulation, receptor activation leads to glycogen synthase kinase–3β (GSK-3β) phosphorylation, thereby preventing β-catenin phosphorylation by GSK-3β. As a result, β-catenin accumulates, translocates to the nucleus, and regulates target gene expression via interaction with the T cell–specific transcription factors (TCFs) (23, 24).
As recently demonstrated, this pathway is expressed and operative in adult lung epithelium in IPF (25, 26). To further elucidate whether WNT/β-catenin activation is an early event in experimental lung fibrosis, as indicated by our initial gene expression analysis, we sought to quantify the mRNA expression of canonical WNT/β-catenin signaling components in ATII cells isolated from the lungs of bleomycin- or saline-treated mice. The investigated WNT ligands were variably expressed in ATII cells, and Wnt1, Wnt2, Wnt7b, and Wnt10b mRNA levels were markedly upregulated, whereas Wnt3a was significantly downregulated (Figure 2C). The common WNT receptors frizzled 1–4 (Fzd1–4), as well as the intracellular signal transducers Gsk3b, Ctnnb1, and Tcf4, were expressed in ATII cells, with a relatively high abundance of Ctnnb1. Fzd1 and Gsk3b were significantly upregulated in ATII cells of bleomycin-treated mice (Figure 2D). To further confirm cell-specific expression of the WNT/β-catenin signaling molecules, we identified the cell types capable of WNT secretion and signaling by immunohistochemistry of WNT1, CTNNB1, and GSK-3β (Figure 3). All signal components were largely expressed in bronchial and alveolar epithelium, with enhanced staining of alveolar epithelial cells in bleomycin-treated mouse lungs in early and advanced stages of lung fibrosis (Figure 3, arrows).
Increased epithelial expression of WNT/β-catenin signaling components in experimental lung fibrosis. Immunohistochemical staining for CTNNB1, GSK-3β, and WNT1 was performed on whole-lung sections of saline- or bleomycin-treated mice 7 or 14 days after bleomycin application, as indicated. The arrows indicate distinct alveolar epithelial cells. Stainings are representative of 2 independent experiments using at least 3 different bleomycin- or saline-treated lung tissues.
Active WNT/β-catenin signaling in vivo during the development of experimental lung fibrosis. TOPGAL reporter mice were used next to localize the activation of the WNT/β-catenin pathway in vivo in experimental lung fibrosis. The detailed treatment scheme is outlined in Supplemental Figure 3. Mice were treated orotracheally with either recombinant WNT3A, to demonstrate the capability of the lung to activate WNT/β-catenin signaling (Figure 4A, top row), or bleomycin, to induce lung fibrosis (Figure 4A, bottom row). As depicted, bronchial and alveolar epithelial cells routinely stained for β-gal in response to WNT3A or bleomycin. Examination of mouse lungs harvested at different time points after a single administration of bleomycin revealed an activation of WNT/β-catenin signaling as early as 5 days after the initial injury, with distinct bronchial and alveolar epithelial cells responding to WNT activation (Figure 4A). The epithelial nature of cells with active WNT signaling was further confirmed by colocalization of β-gal and the ATII cell marker SPC and the clara cell–specific protein (CCSP), respectively (Figure 4B). Increased expression of the WNT target genes Ccnd1 and Wisp1 upon WNT3A stimulation in primary ATII cells in vitro further confirmed these results (Figure 4C).
Activation of WNT/β-catenin signaling in vivo during experimental lung fibrosis. TOPGAL reporter mice were treated orotracheally with WNT3A or bleomycin, as described in detail in Methods. Supplemental Figure 3 depicts the treatment scheme. (A) X-gal staining of β-gal activity in lungs from WNT3A- and vehicle-treated mice (top row) or bleomycin- and saline-treated mice (bottom row). Pictures are representative of at least 2 independent experiments using at least 4 different lung tissues for each condition. (B) X-gal, SPC, and clara cell–specific protein (CCSP) protein expression in serial whole-lung sections from bleomycin-treated TOPGAL reporter mouse was assessed by immunohistochemistry. (C) Primary ATII cells were stimulated with WNT3A (100 ng/ml), and the mRNA levels of Ctgf, Wisp1, and Ccnd1 were analyzed by qRT-PCR (n = 4 for each) at the indicated time points and plotted as log-fold increase (ΔΔCt) of mRNA levels in WNT-stimulated versus unstimulated cells. Results are presented as mean ± SEM; *P < 0.05, **P < 0.02.
Increased expression of WISP1 in ATII cells in vitro and in vivo in experimental and idiopathic pulmonary fibrosis. Based on the evidence that (a) the WNT target WISP1 was one of the most highly regulated genes in ATII cells isolated from fibrotic mouse lungs and (b) active WNT signaling is present in lung fibrosis, we focused our further studies on WISP1 as a potential novel mediator and amenable therapeutic target. WISP1 is a member of the CCN family of matricellular proteins, which consist of CYR61/CCN1, connective tissue growth factor/CCN2 (CTGF/CCN2), NOV/CCN3, WISP1/CCN4, WISP2/CCN5, and WISP3/CCN6 (21, 27, 28). Except for CCN5, all CCN proteins comprise 4 conserved cysteine-rich modular domains. They act through binding to specific integrin receptors and heparin sulfate proteoglycans or modulating the activities of other growth factors and cytokines, thereby triggering a variety of cell functions, such as mitosis, adhesion, and migration of multiple cell types (27). CCN family members have been associated with different developmental and disease processes; however, little is known about WISP1 and WISP2.
We next analyzed the expression of all CCN family members in vivo in mice subjected to bleomycin-induced lung fibrosis. Of the 6 CCN family members, Wisp1 mRNA exhibited the highest fold increase in lung homogenates during bleomycin-induced lung fibrosis (Figure 5A). WISP1 protein localized to ATII cells in vivo, as documented by immunohistochemistry, and increased expression in lung homogenates was demonstrated by Western blot analysis (Figure 5B). In support of this finding, Wisp1 mRNA exhibited the highest fold upregulation of all CCN family members in isolated ATII cells, but not primary fibroblasts, isolated from bleomycin-treated mouse lungs (Figure 5C), underscoring that WISP1 originates from ATII cells during lung fibrosis. WISP1 expression was increased at the protein level in isolated ATII cells, as documented by coimmunofluorescence staining of WISP1 and ECAD (Figure 5D).
Increased WISP1 expression in ATII cells in vitro and in vivo in experimental lung fibrosis. (A) Time-course analysis of CCN family member gene expression was performed using qRT-PCR of lung homogenates harvested 7, 14, or 21 days after administration of bleomycin. Respective mRNA levels were plotted as log-fold change (ΔΔCt) of mRNA levels in bleomycin- versus time-matched saline-treated mice (n = 4) and are presented as mean ± SEM. (B) WISP1 protein expression was assessed using immunohistochemical staining of whole-lung sections of bleomycin- or saline-treated mice 14 days after application (upper panels) and Western blot analysis in total protein lysates (lower panels). Recombinant mouse WISP1 protein served as a positive control; β-actin served as a loading control. Data are representative of at least 2 independent experiments using 6 (Saline) or 5 (Bleo) different lung tissues each. (C) The mRNA levels of all CCN family members were determined by qRT-PCR in primary ATII cells (black bars, n = 6) or primary mouse fibroblasts (mFb; white bars, n = 4) isolated from the lungs of saline- or bleomycin-treated mice 14 days after administration. Results are plotted as log-fold change (ΔΔCt) of mRNA levels in bleomycin-derived versus saline-derived cells and are presented as mean ± SEM. (D) WISP1 protein expression was assessed using double immunostaining for ECAD (green) and WISP1 (red) of primary ATII cells from saline- or bleomycin-treated mice, respectively (original magnification, ×40). Nuclei were visualized by DAPI staining (inset; original magnification, ×40). Data are representative of at least 3 independent experiments.*P < 0.05, **P < 0.02.
We next investigated whether increased WISP1 expression was also evident in human lung tissues derived from IPF patients. To this end, we analyzed the mRNA levels of all CCN family members in lung specimens obtained from IPF (usual interstitial pneumonia [UIP] pattern) or control (transplant donors) patients. With the exception of WISP3, all CCN family members were expressed in human lungs (Figure 6A). WISP1 demonstrated the lowest overall lung mRNA expression but the greatest difference in expression in IPF compared with donor lung homogenates. Furthermore, increased expression of WISP1 mRNA was detectable in septae obtained by laser-assisted microdissection from IPF compared with donor lungs (Figure 6B). qRT-PCR analysis of primary human ATII cells and fibroblasts further revealed that WISP1, and to a lesser extent CTGF, was highly upregulated in ATII cells but not in primary fibroblasts obtained from IPF patients (Figure 6C). Consistently, WISP1 localized to hyperplastic, proliferating ATII cells in close proximity to epithelial lesions and fibroblast foci in IPF (Figure 6D; for antibody control, see Supplemental Figure 4), as assessed by staining of WISP1 and phospho–histone H3 in serial sections. WISP1 protein expression was increased in tissue samples from IPF patients, as determined by Western blot analysis (Figure 6E). Importantly, increased expression of WISP1 was specific for IPF, whereas in other lung disorders, such as nonspecific interstitial pneumonia (NSIP) and chronic obstructive pulmonary disease (COPD), no regulation of WISP1 mRNA was observed (Figure 6F).
Increased WISP1 expression in ATII cells in vitro and in vivo in IPF. (A) mRNA levels of the CCN family members were analyzed by qRT-PCR using lung homogenates derived from donor or IPF lung explants (n = 10 each). (B) mRNA levels of WISP1 and CTGF were analyzed by qRT-PCR in microdissected septae from donor or IPF lungs (n = 5 each). Results in A and B are plotted as relative mRNA levels (ΔCt) and presented as mean ± SEM. (C) mRNA levels of WISP1 (white bars) and CTGF (black bars) were determined by qRT-PCR in primary human ATII cells (n = 4) or fibroblasts (n = 3) isolated from donor or IPF lung tissue. Results are plotted as log-fold increase (ΔΔCt) of mRNA levels in IPF-derived versus donor-derived cells and presented as mean ± SEM. (D) WISP1 protein expression in sections from control or IPF lung specimens was assessed by immunohistochemistry. Arrows indicate extracellular WISP1 staining. WISP1 and phospho–histone H3 (Phospho H3) protein expression in serial whole-lung sections from IPF patients was assessed by immunohistochemistry (bottom row). Data are representative of at least 2 independent experiments using at least 4 different lung tissues from IPF specimens. (E) WISP1 protein expression was determined in total protein lysates from donor or IPF lung tissue using Western blot analysis. Lamin A/C was used as a loading control. Data are representative of at least 2 independent experiments using 6 different lung tissues for donor and IPF specimens. (F) mRNA levels of WISP1 were analyzed by qRT-PCR using lung homogenates derived from IPF (n = 6), nonspecific interstitial pneumonia (NSIP; n = 4), or chronic obstructive pulmonary disease (COPD; n = 6) lung explants. Results are plotted as log-fold increase (ΔΔCt), compared with control lungs (transplant donor), and are presented as mean ± SEM. *P < 0.05, **P < 0.02.
Increased ATII cell proliferation and profibrotic marker release in response to WISP1. To begin to delineate the functional contribution of WISP1, we next assessed the effect of recombinant WISP1 on ATII cells. WISP1 treatment exerted a strong proliferative effect on primary ATII cells (154%–220%, 95% CI), which was more pronounced than that of CTGF or keratinocyte growth factor (KGF) (Figure 7, A and B). Similarly, WISP1 treatment led to increased proliferation of A549 cells (Supplemental Figure 5A). In contrast, ATII cells obtained from bleomycin-treated animals were not responsive to WISP1 stimulation (bleomycin, 186%–213% vs. bleomycin plus WISP1, 199%–215%) (Figure 7A). Since these cells secreted higher amounts of WISP1 (Figure 5), thereby driving a proliferative response, we sought to neutralize WISP1 using 2 different approaches: As depicted in Figure 7C, WISP1 antagonism using neutralizing antibodies attenuated the increased baseline proliferation of fibrotic ATII cells (bleomycin plus α-WISP1, 103%–148%). Second, these results were confirmed using an siRNA against Wisp1 (mRNA and protein knockdown efficiency are shown in Supplemental Figure 5, B and C). The knockdown of Wisp1 led to decreased proliferation of primary ATII cells isolated from bleomycin- and saline-treated mouse lungs, as analyzed by cell counting (Figure 7D) and [3H]thymidine incorporation (Supplemental Figure 5D), respectively.
Increased ATII cell proliferation in response to WISP1. (A) The effects of WISP1 (1 μg/ml), CTGF (2.5 ng/ml), or keratinocyte growth factor (KGF; 10 ng/ml) on primary mouse ATII cell proliferation were assessed by [3H]thymidine incorporation and presented as relative proliferation, compared with unstimulated ATII cells isolated from saline-treated mice (control [Ctr]) (n = 10); **P < 0.02. (B) The effect of WISP1 (1 μg/ml, 24 hours) on the proliferation of primary ATII cells was assessed by coimmunostaining of Ki67 (red) and TJP1 (green) (original magnification, ×40). Nuclei were visualized by DAPI staining (blue). (C) The effects of neutralizing α-WISP1 antibodies (20 μg/ml α-WISP1) or preimmune serum (IgG control), each applied 30 minutes prior to the addition of WISP1, were analyzed by [3H]thymidine incorporation. Data are presented as relative proliferation, compared with unstimulated ATII cells isolated from saline-treated mice (n = 5); **P < 0.02 versus control; #P < 0.02 versus WISP1 stimulation; ##P < 0.02 versus control Bleo. (D) Proliferation of ATII cells subjected to scrambled (Scr) or Wisp1 siRNA (siW1) treatment (150 nM each) was assessed by cell counting 24 hours after treatment. Data are presented as mean ± SEM; ‡P < 0.02, Bleo versus saline; ¶P < 0.02, siRNA versus scrambled. The efficiency and specificity of WISP1 knockdown by siRNA treatment were investigated by qRT-PCR and Western blot analysis (see Supplemental Figure 5, B and C).
EMT in response to WISP1. EMT is the reversible phenotypic switching of epithelial to fibroblast-like cells and has recently gained recognition as a possible mechanism underlying the increase in the (myo)fibroblast pool that occurs in pulmonary fibrosis (29, 30). It has been demonstrated that TGF-β represents a main inducer and regulator of EMT in multiple organ systems (30, 31), but little is known about other cytokines or mediators that are able to induce EMT during lung fibrosis. Here, we show that WISP1 treatment of primary ATII cells led to decreased mRNA levels of Tjp1, Cdh1, and Ocln but elevated mRNA levels of Fsp1 an Acta2, as analyzed by qRT-PCR, indicating that WISP1 is a potent inducer of EMT in ATII cells in vitro (Figure 8A). The induction of EMT was corroborated by immunofluorescence staining, which revealed an increase in α-SMA–positive cells (Figure 8B, left panel), as well as α-SMA and TJP1 double-positive cells (mesenchymal and epithelial markers, respectively; Figure 3B, middle panel, and Supplemental Figure 6) in response to WISP1. This was abrogated by neutralizing antibodies against WISP1 (Figure 8B, right panel, and Supplemental Figure 6). Furthermore, treatment with WISP1 led to enhanced migration of ATII cells, which is associated with the process of EMT (Figure 8C). WISP1 treatment of ATII cells rapidly induced the expression of promigratory genes, such as Mmp7 and Mmp9, as well as the previously identified mediators in pulmonary fibrosis Pai1 and Spp1 (Figure 8D). This strongly suggests that WISP1 not only is causally involved in ATII cell hyperplasia but also induces increased expression and secretion of profibrotic mediators, thereby perpetuating the process of lung fibrosis. Finally, the potential of ATII cells to undergo EMT in vivo was supported by qRT-PCR analysis of freshly isolated ATII cells, which revealed a gain of mesenchymal marker expression and a loss in epithelial cell marker expression in ATII cells isolated from fibrotic mouse lungs (Figure 8E).
EMT of ATII cells in response to WISP1. (A) Primary mouse ATII cells were stimulated with WISP1 (1 μg/ml, 12 hours), and the mRNA levels of the EMT marker genes Tjp1, Cdh1, Ocln, Fsp1, Vim, and Acta2 were analyzed by qRT-PCR (n = 5 for each). (B) Primary ATII cells were stimulated with WISP1 (1 μg/ml, 12 hours) in the absence or presence of neutralizing α-WISP1 antibodies or preimmune serum (IgG control). EMT was assessed by immunofluorescence detection of α-SMA expression (left panels, original magnification, ×10) and colocalization of α-SMA (green) and TJP1 (red) (middle and right panels; original magnification, ×40). Nuclei were visualized by DAPI staining. (C) The migration of ATII cells in response to WISP1 was determined in a Boyden chamber assay; TGF-β1 (2 ng/ml) was used as a positive control. Data are presented as the mean ± SEM of 2 independent experiments performed in triplicate. (D) Primary ATII cells were stimulated with WISP1 (1 μg/ml, 12 hours), and the mRNA levels of the metalloproteinases Mmp2, Mmp7, and Mmp9 and the profibrotic marker genes Pai1 and Spp1 were analyzed by qRT-PCR (n = 5 for each) and plotted as log-fold increase (ΔΔCt) of mRNA levels in WISP1-stimulated versus unstimulated cells. (E) The mRNA levels of the EMT marker genes Tjp1, Cdh1, Ocln, Fsp1, Vim, and Acta2 were determined by qRT-PCR in primary ATII cells isolated from saline- or bleomycin-treated mice 14 days after administration (n = 6). All qRT-PCR results are presented as mean ± SEM. **P < 0.02, *P < 0.05.
Enhanced myofibroblast activation and ECM deposition in response to WISP1. Disturbed epithelial-mesenchymal crosstalk is a hallmark of lung fibrosis (3, 12). We therefore next assessed the effects of WISP1 on lung fibroblasts, the key effector cell type in lung fibrosis. WISP1 treatment led to a significant induction of the ECM components type I collagen α1 (Col1a1) and fibronectin 1 (Fn1), as well as (myo)fibroblast activation markers in NIH 3T3 cells (Figure 9A) and human lung fibroblasts (Figure 9D), as assessed by qRT-PCR of WISP1- and vehicle-treated fibroblasts. Further, WISP1 treatment resulted in a marked increase in collagen production by fibroblasts in vitro (Figure 9, B and E). Moreover, collagen production was similar in WISP1- and TGF-β1–treated cells (219% ± 42% vs. 184% ± 45% for NIH 3T3 fibroblasts and 238% ± 14% vs. 195% ± 11% for human lung fibroblasts, respectively). This was confirmed by immunofluorescence staining of type I collagen in fibroblasts, which demonstrated increased collagen staining in response to WISP1 (Figure 9C). Interestingly, the proliferation of fibroblasts was not affected by WISP1 treatment, independent of serum conditions (Supplemental Figure 7).
Enhanced ECM deposition and myofibroblast marker expression by fibroblasts in response to WISP1. (A) NIH 3T3 fibroblasts were stimulated with WISP1 (1 μg/ml; 6 or 12 hours, as indicated), and the mRNA levels of the ECM components Col1a1, Col1a2, fibronectin (Fn1), and the (myo)fibroblast activation markers Fsp1 and Acta2 were analyzed by qRT-PCR (n = 4). Results are plotted as log-fold increase (ΔΔCt) of mRNA levels in WISP1-stimulated versus unstimulated cells and presented as mean ± SEM. (B) NIH 3T3 fibroblasts were stimulated with WISP1 (1 μg/ml) or TGF-β1 (2 ng/ml) for 24 hours, and total collagen content was quantified using the Sircol collagen assay (n = 6). (C) Fibroblast collagen expression and localization after WISP1 stimulation for 24 hours were assessed by immunofluorescence detection of type I collagen 1 (red). Nuclei were visualized by DAPI staining (blue). Data are representative for at least 3 independent experiments. Original magnification, ×40. (D) Human lung fibroblasts were stimulated with WISP1 (1 μg/ml; 6 or 12 hours, as indicated), and the mRNA levels of the ECM components Col1a1, fibronectin (Fn1), the (myo)fibroblast activation markers Fsp1, Acta2, tenascin C (Tnc), and the tissue inhibitor of matrix metalloproteinases 1 (Timp1) were analyzed by qRT-PCR (n = 3) as described in A. (E) Human lung fibroblasts were stimulated with WISP1 (1 μg/ml) or TGF-β1 (2 ng/ml) for 24 hours, and total collagen content was quantified using the Sircol collagen assay (n = 3).*P < 0.05, **P < 0.02.
In sum, these findings suggest that WISP1 also exerts paracrine effects on lung fibroblasts, further contributing to its profibrotic action and demonstrating its relevance to IPF pathogenesis.
Attenuation of lung fibrosis in vivo by WISP1 inhibition. To assess whether modulation of WISP1 activity represents an effective therapeutic option in lung fibrosis, we depleted WISP1 during bleomycin-induced lung fibrosis using antibodies shown to be effective in neutralizing WISP1 activity (Figure 7C). To this end, we subjected mice to bleomycin-induced lung fibrosis and treated them with repetitive orotracheal applications of α-WISP1 or species-matched preimmune control antibodies. As depicted in Figure 10, mice subjected to WISP1 neutralization exhibited significantly less pulmonary fibrosis and a marked decrease in ECM deposition, as assessed by immunohistochemistry for type 1 collagen (Figure 10A), quantification of total lung collagen (bleomycin plus IgG, 295% ± 17% vs. bleomycin plus α-WISP1, 160% ± 31%, compared with saline-treated controls) (Figure 10B), as well as α-SMA and tenascin C immunostaining (Figure 10C and Supplemental Figure 8, respectively).
WISP1 neutralization in vivo leads to the attenuation of lung fibrosis. (A) Mice were subjected to saline or bleomycin instillation, as described above, and treated either with neutralizing α-WISP1 antibodies or preimmune serum (IgG control) by orotracheal application as described in detail in Methods. Lungs were processed 14 days after bleomycin application for immunohistochemical analysis and stained for type I collagen. (B) Total collagen content in lung homogenates was quantified using the Sircol collagen assay. Results are derived from whole lungs harvested 14 days after saline, bleomycin, bleomycin plus preimmune serum (IgG control), or bleomycin plus neutralizing α-WISP1 antibody instillation by orotracheal application (n = 5 each). Results are presented as mean ± SEM; **P < 0.02, #P < 0.02 versus Bleo + IgG treatment. (C) Indicated lung sections were used for immunohistochemical analysis and stained with α-SMA. Pictures are representative of at least 2 independent experiments using at least 4 different lung tissues for each condition.
These findings were corroborated by the finding that WISP1 neutralization also led to decreased mRNA expression of the profibrotic markers Col1a1, Spp1, Mmp7, and Pai1 (Figure 11B and Supplemental Figure 9A), which is of significance, as we have shown that WISP1 induces the expression of these markers in primary ATII cells (Figure 8D). In addition, WISP1 neutralization resulted in a reversal of EMT marker gene expression in vivo (Figure 11C and Supplemental Figure 9B), which were induced by WISP1 in vitro (Figure 8A). Interestingly, WISP1 neutralization led to a marked decrease in ECM deposition but did affect inflammatory cell influx in response to bleomycin (Figure 11A, right panels).
WISP1 inhibition in vivo leads to decreased profibrotic marker gene expression and increased survival in lung fibrosis. (A) Indicated lung sections were used for immunohistochemical analysis and stained with H&E. (B and C) mRNA levels of the profibrotic marker genes Col1a1, Spp1, Mmp7, Pai1, and Ctgf (B) and the EMT marker genes Tjp1, Cdh1, Ocln, Fsp1, Vim, and Acta2 (C) were analyzed by qRT-PCR (n = 5 each). Results are plotted as log-fold change (ΔΔCt) of mRNA levels in lung specimens 14 days after bleomycin instillation treated with neutralizing α-WISP1 antibodies, compared with lungs treated with preimmune serum (IgG control). Results are presented as mean ± SEM. See Supplemental Figure 9, A and B, for a comparison of all treatment groups. (D) Lung compliance measurements were obtained from mice instilled with saline, bleomycin, bleomycin plus IgG control, or bleomycin plus α-WISP1 antibody (n = 10 for each), 14 days after initial exposure to bleomycin. (E) The survival of mice subjected to neutralizing α-WISP1 or preimmune serum (IgG control) instillations (n = 18 for each) was monitored. Days of antibody instillations are indicated on the x axis. *P < 0.05, **P < 0.02; #P < 0.02 versus Bleo + IgG treatment.
Finally, WISP1 neutralization partially restored normal lung function, as assessed by lung compliance measurements (bleomycin plus IgG, 0.065 ± 0.073 ml/kPa vs. bleomycin plus α-WISP1, 0.09 ± 0.11 ml/kPa, 95% CI; Figure 11D). Most importantly, WISP1 neutralization significantly improved the survival of bleomycin-challenged mice (bleomycin plus IgG, 47% vs. bleomycin plus α-WISP1, 74%; n = 18 for each) (Figure 11E), suggesting a valuable novel approach for the treatment of lung fibrosis.