- Email: [email protected]
Loss of PTEN expression by mouse fibroblasts results in lung fibrosis through a CCN2-dependent mechanism
Communication
Loss of PTEN expression by mouse fibroblasts results in lung fibrosis through a CCN2-dependent mechanism
Sunil K. Parapuram 1, 3, 4, † , Katherine Thompson 1, † , Matthew Tsang 2, † , James Hutchenreuther 2 , Christian Bekking 1 , Shangxi Liu 1 and Andrew Leask 1, 2 1 - Department of Dentistry, Schulich School of Medicine and Dentistry, University of Western Ontario, Dental Sciences Bldg., London, ON, Canada, N6A 5C1 2 - Physiology and Pharmacology, Schulich School of Medicine and Dentistry, University of Western Ontario, Dental Sciences Bldg., London, ON, Canada, N6A 5C1 3 - Department of Ophthalmology, Lawson Health Research Institute, 268 Grosvenor Street, London, Ontario, Canada, N6A 4V2 4 - Department of Pathology, Lawson Health Research Institute, 268 Grosvenor Street, London, Ontario, Canada, N6A 4V2
Correspondence to Andrew Leask: [email protected]. http://dx.doi.org/10.1016/j.matbio.2015.01.017 Edited by R. Iozzo
Abstract Elevated adhesive signaling promotes fibrosis. Protein phosphatase and tensin homologue (PTEN) dephosphorylates focal adhesion kinase and suppresses the activation of Akt and hence suppresses adhesive signaling. Loss of PTEN expression is associated with lung fibrosis, but whether PTEN expression by type I collagen-expressing cells controls lung fibrosis is unclear. Here, we use mice expressing tamoxifen-dependent cre recombinase expressed under the control of a COL1A2 promoter/ enhancer and mice harboring floxed-PTEN and/or floxed-CCN2 alleles to assess whether loss of PTEN expression by type I collagen producing cells results in lung fibrosis in a CCN2-dependent fashion. In vivo, loss of PTEN expression resulted in the overexpression of both collagen type I and the pro-adhesive matricellular protein connective tissue growth factor (CTGF/CCN2). However, α-smooth muscle actin expression was unaffected. Loss of CCN2 expression by lung fibroblasts rescues this phenotype; i.e.., mice deficient in both PTEN and CCN2 in collagen type I-expressing cells do not develop significant collagen deposition in the lung. PTEN expression by collagen type I-expressing cells controls collagen deposition; therapeutic strategies blocking CCN2 may be of benefit in blocking excessive collagen deposition in fibrosis. © 2015 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Introduction Fibrosis is one of the largest groups of diseases for which there is no generally agreed upon therapy. Lung fibrosis [e.g. idiopathic pulmonary fibrosis (IPF) and Fibrosing Alveolitis Complicating Systemic Sclerosis] is associated with severe morbidity [1,2]. Fibrosis is caused by the excessive production of extracellular matrix (ECM) notably collagen type I by fibroblasts present within connective tissue. Understanding the fundamental mechanisms underlying how fibroblasts contribute to lung fibrosis is therefore of high clinical relevance.
Fibroblasts present in fibrotic lesions are characterized by elevated adhesion to ECM and adhesive signaling including focal adhesion kinase (FAK)/ PI3K/Akt phosphorylation [3–5]. The phosphatase and tensin homologue (PTEN), a dual protein/lipid phosphatase which dephosphorylates FAK and suppresses the activation of PI3K-Akt signaling, is reduced in fibroblasts in IPF patients [6]. Reduced PTEN expression is seen in SSc skin fibroblasts, and genetic deletion of PTEN in skin fibroblasts results in progressive dermal fibrosis [7]. However, whether loss of PTEN expression by lung fibroblasts is sufficient to result in lung fibrosis is unclear.
0945-053/© 2015 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Matrix Biol. (2015) 43, 35–41
36
CCN2 is required for lung fibrosis
It has been hypothesized that the matricellular protein CTGF (connective tissue growth factor/ CCN2) is a central mediator of fibrosis and may be
a novel antifibrotic therapeutic approach [8]. Conditional CCN2 knockout mice were recently used to show that CCN2 was required for skin fibrosis in the
A
50µm
50µm
B
C
D
Fig. 1 (legend on next page).
37
CCN2 is required for lung fibrosis
Fig. 2. Loss of PTEN expression by lung fibroblasts results in enhanced CCN2 production. Control mice (Pten fl/fl) and mice deleted for PTEN (Pten −/−) or PTEN and CCN2 (Pten −/−; CCN2 −/−) cells expressing the COL1A2 promoter/enhancer were generated as described in Methods. Sections were stained with anti-CCN2 antibody. Note the presence of CCN2 expressing cells in Pten −/− compared to control or Pten −/−; Ccn2 −/− mice. N = 3. Representative images are shown.
bleomycin model of skin and for the fibrosis caused by the loss of PTEN expression in skin fibroblasts [9,10]. Moreover, a neutralizing antibody reduced lung fibrosis caused by bleomycin [11]. However, bleomycin is an imperfect model of SSc [12]; whether CCN2 mediates connective tissue deposition in the lung caused by loss of PTEN expression is unclear. In this report, we assess whether loss of PTEN expression by lung fibroblasts results in excessive collagen deposition in the lung and whether this phenotype is mediated by CCN2.
Results Loss of PTEN expression by lung fibroblasts results in collagen deposition in the lung To identify cell types in the lung in which the COL1A2 promoter/enhancer used for our studies is expressed, we generated mice harboring an allele which permits the tamoxifen-dependent expression of cre recombinase under the control of the COL1A2 promoter/enhancer and an allele that either allows
Fig. 1. Loss of PTEN expression by type I collagen producing cells in the lung results in enhanced collagen deposition. (A) Reporter mice expressing a tamoxifen-inducible cre recombinase under the control of a COL1A2 promoter/enhancer [19] and either a β-galactosidase gene expressed only when cre is active (COL1A2-lacZ) or a green fluorescent protein (GFP) gene expressed only when cre is active were generated. Three week old mice were injected with tamoxifen, and 60 days later localization of reporter expression was detected (with X-gal, or with anti-GFP antibody, as indicated). (Left hand and middle panels) Tissue was counterstained with anti-α-SMA antibody (to detect smooth muscle cells) or anti-S100A4 antibody (to detect fibroblasts). Note that some α-SMA expressing cells and some S100A4-expressing cells were positive for COL1A2 promoter/enhancer expression. Tissue was counterstained with DAPI to detect nuclei. (Right hand panel) Tissue was counterstained with anti-PTEN antibody. Note that some PTEN-expressing cells were positive for COL1A2 promoter/enhancer expression. Tissue was counterstained with DAPI to detect nuclei. (B) Control mice (Pten fl/fl) and mice deleted for PTEN (Pten −/−) in cells expressing the COL1A2 promoter/enhancer were generated as described in Methods. Lung sections were stained with α-PTEN antibody and counterstained with DAPI. Representative images are shown. Fig. 1B: N = 3 (avg of 3 fields per 3 lungs), *** = p b 0.001, Student's t-test, +/− SD. (C) Control mice (Pten fl/fl) and mice deleted for PTEN (Pten −/−) in cells expressing the COL1A2 promoter/enhancer were generated as described in Methods and subjected to trichrome staining. A representative stained section is shown. N = 6. (D) Control mice (Pten fl/fl) and mice deleted for PTEN (Pten −/−) in cells expressing the COL1A2 promoter/enhancer were generated as described in Methods. Measurement of COL1A1 and COL1A2 mRNA expression relative to 18S controls confirmed elevation of collagen in PTEN-deficient mice (*, p b 0.05, N = 4). Note that PTEN expression was reduced (p = 0.06, N = 4), but CCN2 and α-SMA mRNA expression in whole lung was not appreciably altered.
38
CCN2 is required for lung fibrosis
Fig. 3. Loss of CCN2 rescues elevated collagen production and increased Akt phosphorylation seen in PTEN-deficient mice. Control mice (Pten fl/fl) and mice deleted for PTEN (Pten −/−) or PTEN and CCN2 (Pten −/−; CCN2 −/−) in fibroblasts were generated as described in Methods. (A) Trichrome staining. N = 6. A representative stained section is shown. (B) Collagen ELISA. Equal amounts of protein from tissue samples from lungs of control mice (Pten fl/fl) (N = 6) and mice deleted for PTEN in fibroblasts (Pten −/−) (N = 5) or PTEN and CCN2 (Pten −/−; CCN2 −/−) (N = 6) in fibroblasts were subjected to an ELISA that detects total collagen (** = p b 0.01). (C). Sections were stained with anti-α-SMA and anti-phospho-Akt antibodies. Tissue was counterstained with DAPI to detect nuclei. Note the presence of cells positive for phospho-Akt in Pten −/− compared to control or Pten −/−; Ccn2 −/− mice. N = 3. Representative images are shown. Note the absence of expansion of α-SMA expressing cells (i.e. the induction of myofibroblasts) in PTEN-deficient mice. N = 3 (avg of 3 fields per 3 lungs), *** = p b 0.001, one-way ANOVA, +/− SD.
39
CCN2 is required for lung fibrosis
the expression of beta-galactosidase in cells expressing cre (COL1A2-lacZ) or green fluorescent protein in cells expressing cre (COL1A2-GFP). COL1A2-lacZ mice were injected with tamoxifen, and assessed for β-galactosidase (using X-gal) expression 60 days post-cessation of tamoxifen injection; tissue sections were counterstained with α-smooth muscle actin (SMA) antibody to detect smooth muscle cells. Reporter activity was found in some α-SMA expressing cells around blood vessels but mostly within fibroblasts (Fig. 1A). To confirm this observation, COL1A2-GFP mice were also injected with tamoxifen, and assessed for GFP expression (using an anti-GFP antibody) expression 60 days post-cessation of tamoxifen injection; tissue was counterstained with anti-α−SMA antibody and anti-S100A4 antibody (to detect fibroblasts). Reporter activity was found again in some α-SMA expressing cells around blood vessels and in S100A4-expressing fibroblasts (Fig. 1A). Reporter activity was localized to a subset of cells endogenously expressing PTEN (Fig. 1A). To begin to assess whether loss of PTEN expression by collagen type I-expressing cells resulted in lung fibrosis, mice deleted (PTEN −/−) or not (PTEN fl/fl) for PTEN in type I collagen (COL1A2) producing cells were generated and genotyped as described previously [7] and in Methods. Loss of PTEN expression around blood vessels and bronchioles and in most fibroblasts in the lung was confirmed using an anti-PTEN antibody (Fig. 1B). PTEN-deficient mice showed fibrosis revealed as visualized by trichrome stain (Fig. 1C). Lungs of PTEN-deficient mice, compared to their wild-type counterparts, also showed statistically significantly elevated expression of Col1A1 and Col1A2 mRNA (Fig. 1D). PTEN-deficient mice showed reduced expression of Pten mRNA (p = 0.06, N = 4) (Fig. 1D). However, when whole lung was examined, CCN2 and α-SMA RNA expression was not altered (Fig. 1D). Loss of CCN2 rescues the fibrotic phenotype of PTEN-deficient lung It was possible that loss of PTEN expression, although not resulting in alterations of CCN2 and α-SMA expression when whole lung was examined, might cause induction of CCN2 or α-SMA protein expression in some areas of the lung. To begin to assess whether the excess collagen deposition caused by loss of PTEN expression in the lung could be caused by CCN2, we first showed that loss of PTEN expression resulted in increased CCN2 expression in some cells around vessels and in some fibroblasts (Fig. 2). We then generated mice deficient for PTEN and CCN2 in fibroblasts, as previously described [10]. Examination of mice deleted for both PTEN and CCN2 confirmed loss of
CCN2 (Fig. 2). Intriguingly, loss of CCN2 rescued the excess collagen production seen in PTEN-deficient animals (Fig. 3A, B). Similarly, loss of PTEN, in a CCN2-dependent fashion, resulted in an increase in p-Akt-positive cells (Fig. 3C). Intriguingly, α-SMA expression was not appreciably altered in PTEN-deficient mice, indicating that although loss of PTEN in collagen type I-expressing cells resulted in increased collagen levels, this induction was not accompanied by the increased appearance of myofibroblasts. These results suggest that loss of PTEN, in a CCN2-dependent fashion, is sufficient to result in increased collage deposition in the lung.
Discussion In this report, we showed that loss of PTEN expression resulted in collagen deposition in the lung in a CCN2-dependent fashion. These data support the notion that alterations in PTEN expression contribute to fibrogenesis in vivo [6,7] and that CCN2 is a key mediator of collagen production in fibrosis [8]. Fibroblasts from patients with IPF have low membrane-associated PTEN expression due to low-expression of caveolin-1 [6]. Also, deletion of PTEN in lung epithelial cells results in injury to lung epithelial cells and subsequent fibrosis [13]. Our previous studies provide evidence that CCN2 expression is induced downstream of loss of PTEN expression and is required for dermal fibrosis caused by loss of PTEN expression [7,10]. The current data indicate that a similar mechanism occurs in our model of collagen deposition in the lung and provides strong support that CCN2 elevation in lung fibrosis may represent a novel therapeutic approach to reduce excessive collagen deposition in these diseases [8,14,15]. It is interesting to note that this involvement of CCN2 occurred even though CCN2 expression appeared to be induced at the protein level in relatively few cells; when whole lung was examined, loss of PTEN did not cause an increase in the overall levels of CCN2 mRNA. CCN2 is also involved with matrix deposition in the invertebral disc; loss of CCN2 exacerbates age-dependent progressive disc degeneration [16– 18]. Future efforts will be expended to assess if loss of PTEN in collagen type I-expressing cells in the lung results in progressive lung fibrosis. Conversely, our studies strongly indicate an important homeostatic role for PTEN in maintaining the extracellular matrix environment in the lungs, as well as the skin. The mechanisms that result in its decrease in IPF are being deciphered [6]; however, CCN2 represents a downstream mediator of at least certain features of fibrosis and may therefore be a suitable target for anti-fibrotic drug intervention.
40
Materials and methods Animal studies
CCN2 is required for lung fibrosis
Santa Cruz). After primary antibody incubation, sections were then washed with PBS and incubated with appropriate fluorescent secondary antibodies (Jackson Immunoresearch) 1 h at room temperature. Sections were washed with PBS, mounted using 4′,6-diamidino-2phenylindole (DAPI) and photographed using a Zeiss fluorescence microscope and Northern Eclipse software (Empix, Missassagua, ON, Canada). Alternatively, signal was developed using the Vectastain ABC Kit (Vector Laboratories) and ImmPACT DAB Peroxidase (HRP) Substrate (Vector Laboratories). When indicated, average fluorescence intensity (three fields/lung) was calculated using Northern Eclipse software (Empix, Missassagua, ON, Canada).
Mice deleted in fibroblasts for PTEN or both PTEN and CCN2 were generated essentially as previously described [7,10]. Briefly, loxP-Pten mice were obtained from The Jackson Laboratory; loxP-Ccn2 mice and mice expressing tamoxifen-dependent Cre recombinase under the control of fibroblast-specific regulatory sequence from the pro-alpha 2 (I) collagen gene were previously described [7,10,19]. Mice hemizygous for Cre and homozygous for loxP-Pten were administered tamoxifen (1 mg/mouse for 5 days) or corn oil at 21–24 days of age to generate mice deleted for the Pten gene specifically in fibroblasts or control mice, respectively. Mice hemizygous for Cre and homozygous for loxP-Pten and loxP-Ccn2 were similarly used to generate mice deleted for PTEN and CCN2. Deletion of PTEN and/or CCN2 was tested by PCR genotyping as described previously [7,10]. Mice were sacrificed by CO2 euthanasia at time points indicated. Thirty-sixty days post-cessation of tamoxifen injection [7,10], lung samples were collected for RT-PCR, histology, immunohistochemistry and for collagen assay. When indicated, mice hemizygous for an allele allowing the expression of tamoxifen-dependent Cre recombinase under the control of fibroblast-specific regulatory sequence from the pro-alpha 2 (I) collagen gene and for an allele allowing the expression of β-galactosidase or green fluorescent protein (GFP) in cells expressing active, nuclear cre recombinase (Jackson Labs) were used. All animal protocols were given ethical approval by the University of Western Ontario's Animal Care Committee.
Lung tissue collected from mice was frozen down at − 80 °C in RNAlater stabilization solution (Invitrogen). Extraction of RNA was done using Trizol (Invitrogen) and chloroform (Sigma) method and 40 ng of RNA samples was run in triplicate. RNA was reverse transcribed and amplified in a 15 μl reaction using TaqMan Assays on Demand primers, 6-carboxyfluroscein-labeled TaqMan MGB probe and Reverse Transcriptase qPCR One-step Mastermix (Quanta, VWR). ABI Prism 7900 HT sequence detector (Perkin-Elmer-Cetus, Vaudreuil, QC) was used according to the manufacturer's instructions to detect amplified sequences. Expression values were standardized to control values from 18S primers using the ΔΔCt method. Statistical analysis on three independent experiments was done using Student's t-test on GraphPad Prism.
Immunohistochemistry and assessment
Total collagen assay
Lung tissue sections (0.5 μm) were cut using a microtome (Leica, Richmond Hill, ON, Canada), collected on Superfrost Plus slides (Fisher Scientific, Ottawa, ON, Canada), de-waxed in xylene and rehydrated by successive immersion in descending concentrations of alcohol. When indicated, tissue was stained with Harris's hematoxylin and eosin or Trichrome. Sections were also subjected for immunofluorescence staining. Sections are boiled in citric acid pH 6 for antigen retrieval. Tissue sections were blocked by incubation with 5% BSA, 0.1% Triton X-100 in phosphate-buffered saline (PBS) for 1 h and then incubated with primary antibodies under humidified conditions at 4 °C overnight. Primary antibodies used were: anti-PTEN (1:1000 dilution, Sigma), anti-SMA (1:500, Sigma), anti-GFP (1:100 Santa Cruz), anti-p-Akt (1:1000, Cell Signaling Technology), anti-S100A4 (1:800, Cell Signaling Technology) and anti-CCN2 (1:100,
To quantify collagen, a Total Collagen Kit that detects hydroxyproline was performed in tissue hydrolysates, as described by the manufacturer (Quickzyme). Tissue underwent complete hydrolysis by adding 100 μL/10 mg of 12 M HCl and boiling at 95 °C for 20 h. Samples were diluted 10 × in 4 M HCl and 35 μL of the diluted sample was incubated with 75 μL buffer supplied by the kit for 20 min in a 96 well plate. Next, 75 μL of detecting reagent was added to each well and incubated at 60 °C for 1 h. A series of standards with known concentrations of 300, 200, 100, 50, 25, 12.5, 6.25 and 0 μg/mL were run with the tissue samples under the same conditions. Absorbance at 570 nm was determined (iMark™ microplate absorbance reader, BioRad) and compared to the standard curve. Total collagen values obtained were subject to statistical analysis using one way ANOVA and Tukey's post-hoc test on GraphPad Prism.
Real-time PCR of lung tissue
41
CCN2 is required for lung fibrosis
Author contributions SK KT JH, CB, MT and SL designed, performed experiments and analyzed data. AL SK KT JH CB MT and SL wrote and edited the paper.
Competing interests None
Acknowledgments This work is supported by grants from the Canadian Institutes of Health Research (MOP-77603) and the Scleroderma Society of Ontario (to AL) (MOP-119410). SKP was supported by the Canadian Scleroderma Research Group. MT was supported by a Master's student scholarship from NSERC. Received 12 November 2014; Received in revised form 22 January 2015; Accepted 23 January 2015 Available online 30 January 2015 Keywords: CCN2; ctgf; PTEN; Lung fibrosis; Scleroderma; Matricellular proteins; CCN family † Co-first authors.
References [1] du Bois RM. Strategies for treating idiopathic pulmonary fibrosis. Nat Rev Drug Discov 2010;9:129–40. [2] Fan MH, Feghali-Bostwick CA, Silver RM. Update on scleroderma-associated interstitial lung disease. Curr Opin Rheumatol 2014;26:630–6. [3] Garneau-Tsodikova S, Thannickal VJ. Protein kinase inhibitors in the treatment of pulmonary fibrosis. Curr Med Chem 2008;15:2632–40. [4] Lagares D, Busnadiego O, García-Fernández RA, Kapoor M, Liu S, Carter DE, et al. Inhibition of focal adhesion kinase prevents experimental lung fibrosis and myofibroblast formation. Arthritis Rheum 2012;64:1653–64.
[5] Shi-wen X, Thompson K, Khan K, Liu S, Murphy-Marshman H, Baron M, et al. Focal adhesion kinase and reactive oxygen species contribute to the persistent fibrotic phenotype of lesional scleroderma fibroblasts. Rheumatology (Oxford) 2012;51:2146–54. [6] Xia H, Khalil W, Kahm J, Jessurun J, Kleidon J, Henke CA. Pathologic caveolin-1 regulation of PTEN in idiopathic pulmonary fibrosis. Am J Pathol 2010;176:2626–37. [7] Parapuram SK, Shi-wen X, Elliott C, Welch ID, Jones H, Baron M, et al. Loss of PTEN expression by dermal fibroblasts causes skin fibrosis. J Invest Dermatol 2011; 131:1996–2003. [8] Leask A. Possible strategies for anti-fibrotic drug intervention in scleroderma. J Cell Commun Signal 2011;5:125–9. [9] Liu S, Shi-wen X, Abraham DJ, Leask A. CCN2 is required for bleomycin-induced skin fibrosis in mice. Arthritis Rheum 2011;63:239–46. [10] Liu S, Parapuram SK, Leask A. Fibrosis caused by loss of PTEN expression in mouse fibroblasts is crucially dependent on CCN2. Arthritis Rheum 2013;65:2940–4. [11] Ponticos M, Holmes AM, Shi-wen X, Leoni P, Khan K, Rajkumar VS, et al. Pivotal role of connective tissue growth factor in lung fibrosis: MAPK-dependent transcriptional activation of type I collagen. Arthritis Rheum 2009;60: 2142–55. [12] Artlett CM. Animal models of scleroderma: fresh insights. Curr Opin Rheumatol 2010;22:677–82. [13] Miyoshi K, Yanagi S, Kawahara K, Nishio M, Tsubouchi H, Imazu Y, et al. Epithelial Pten controls acute lung injury and fibrosis by regulating alveolar epithelial cell integrity. Am J Respir Crit Care Med 2013;187:262–75. [14] Allen JT, Knight RA, Bloor CA, Spiteri MA. Enhanced insulinlike growth factor binding protein-related protein 2 (connective tissue growth factor) expression in patients with idiopathic pulmonary fibrosis and pulmonary sarcoidosis. Am J Respir Cell Mol Biol 1999;21:693–700. [15] Sato S, Nagaoka T, Hasegawa M, Tamatani T, Nakanishi T, Takigawa M, et al. Serum levels of connective tissue growth factor are elevated in patients with systemic sclerosis: association with extent of skin sclerosis and severity of pulmonary fibrosis. J Rheumatol 2000;27: 149–54. [16] Bedore J, Leask A, Séguin CA. Targeting the extracellular matrix: matricellular proteins regulate cell-extracellular matrix communication within distinct niches of the intervertebral disc. Matrix Biol 2014;37:124–30. [17] Bedore J, Sha W, McCann MR, Liu S, Leask A, Séguin CA. Impaired intervertebral disc development and premature disc degeneration in mice with notochord-specific deletion of CCN2. Arthritis Rheum 2013;65:2634–44. [18] Tran CM, Shapiro IM, Risbud MV. Molecular regulation of CCN2 in the intervertebral disc: lessons learned from other connective tissues. Matrix Biol 2013;32:298–306. [19] Ponticos M, Abraham D, Alexakis C, Lu QL, Black C, Partridge T, et al. Col1a2 enhancer regulates collagen activity during development and in adult tissue repair. Matrix Biol 2004;22:619–28.
Loss of PTEN expression by mouse fibroblasts results in lung fibrosis through a CCN2-dependent mechanism
Sunil K. Parapuram 1, 3, 4, † , Katherine Thompson 1, † , Matthew Tsang 2, † , James Hutchenreuther 2 , Christian Bekking 1 , Shangxi Liu 1 and Andrew Leask 1, 2 1 - Department of Dentistry, Schulich School of Medicine and Dentistry, University of Western Ontario, Dental Sciences Bldg., London, ON, Canada, N6A 5C1 2 - Physiology and Pharmacology, Schulich School of Medicine and Dentistry, University of Western Ontario, Dental Sciences Bldg., London, ON, Canada, N6A 5C1 3 - Department of Ophthalmology, Lawson Health Research Institute, 268 Grosvenor Street, London, Ontario, Canada, N6A 4V2 4 - Department of Pathology, Lawson Health Research Institute, 268 Grosvenor Street, London, Ontario, Canada, N6A 4V2
Correspondence to Andrew Leask: [email protected]. http://dx.doi.org/10.1016/j.matbio.2015.01.017 Edited by R. Iozzo
Abstract Elevated adhesive signaling promotes fibrosis. Protein phosphatase and tensin homologue (PTEN) dephosphorylates focal adhesion kinase and suppresses the activation of Akt and hence suppresses adhesive signaling. Loss of PTEN expression is associated with lung fibrosis, but whether PTEN expression by type I collagen-expressing cells controls lung fibrosis is unclear. Here, we use mice expressing tamoxifen-dependent cre recombinase expressed under the control of a COL1A2 promoter/ enhancer and mice harboring floxed-PTEN and/or floxed-CCN2 alleles to assess whether loss of PTEN expression by type I collagen producing cells results in lung fibrosis in a CCN2-dependent fashion. In vivo, loss of PTEN expression resulted in the overexpression of both collagen type I and the pro-adhesive matricellular protein connective tissue growth factor (CTGF/CCN2). However, α-smooth muscle actin expression was unaffected. Loss of CCN2 expression by lung fibroblasts rescues this phenotype; i.e.., mice deficient in both PTEN and CCN2 in collagen type I-expressing cells do not develop significant collagen deposition in the lung. PTEN expression by collagen type I-expressing cells controls collagen deposition; therapeutic strategies blocking CCN2 may be of benefit in blocking excessive collagen deposition in fibrosis. © 2015 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Introduction Fibrosis is one of the largest groups of diseases for which there is no generally agreed upon therapy. Lung fibrosis [e.g. idiopathic pulmonary fibrosis (IPF) and Fibrosing Alveolitis Complicating Systemic Sclerosis] is associated with severe morbidity [1,2]. Fibrosis is caused by the excessive production of extracellular matrix (ECM) notably collagen type I by fibroblasts present within connective tissue. Understanding the fundamental mechanisms underlying how fibroblasts contribute to lung fibrosis is therefore of high clinical relevance.
Fibroblasts present in fibrotic lesions are characterized by elevated adhesion to ECM and adhesive signaling including focal adhesion kinase (FAK)/ PI3K/Akt phosphorylation [3–5]. The phosphatase and tensin homologue (PTEN), a dual protein/lipid phosphatase which dephosphorylates FAK and suppresses the activation of PI3K-Akt signaling, is reduced in fibroblasts in IPF patients [6]. Reduced PTEN expression is seen in SSc skin fibroblasts, and genetic deletion of PTEN in skin fibroblasts results in progressive dermal fibrosis [7]. However, whether loss of PTEN expression by lung fibroblasts is sufficient to result in lung fibrosis is unclear.
0945-053/© 2015 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Matrix Biol. (2015) 43, 35–41
36
CCN2 is required for lung fibrosis
It has been hypothesized that the matricellular protein CTGF (connective tissue growth factor/ CCN2) is a central mediator of fibrosis and may be
a novel antifibrotic therapeutic approach [8]. Conditional CCN2 knockout mice were recently used to show that CCN2 was required for skin fibrosis in the
A
50µm
50µm
B
C
D
Fig. 1 (legend on next page).
37
CCN2 is required for lung fibrosis
Fig. 2. Loss of PTEN expression by lung fibroblasts results in enhanced CCN2 production. Control mice (Pten fl/fl) and mice deleted for PTEN (Pten −/−) or PTEN and CCN2 (Pten −/−; CCN2 −/−) cells expressing the COL1A2 promoter/enhancer were generated as described in Methods. Sections were stained with anti-CCN2 antibody. Note the presence of CCN2 expressing cells in Pten −/− compared to control or Pten −/−; Ccn2 −/− mice. N = 3. Representative images are shown.
bleomycin model of skin and for the fibrosis caused by the loss of PTEN expression in skin fibroblasts [9,10]. Moreover, a neutralizing antibody reduced lung fibrosis caused by bleomycin [11]. However, bleomycin is an imperfect model of SSc [12]; whether CCN2 mediates connective tissue deposition in the lung caused by loss of PTEN expression is unclear. In this report, we assess whether loss of PTEN expression by lung fibroblasts results in excessive collagen deposition in the lung and whether this phenotype is mediated by CCN2.
Results Loss of PTEN expression by lung fibroblasts results in collagen deposition in the lung To identify cell types in the lung in which the COL1A2 promoter/enhancer used for our studies is expressed, we generated mice harboring an allele which permits the tamoxifen-dependent expression of cre recombinase under the control of the COL1A2 promoter/enhancer and an allele that either allows
Fig. 1. Loss of PTEN expression by type I collagen producing cells in the lung results in enhanced collagen deposition. (A) Reporter mice expressing a tamoxifen-inducible cre recombinase under the control of a COL1A2 promoter/enhancer [19] and either a β-galactosidase gene expressed only when cre is active (COL1A2-lacZ) or a green fluorescent protein (GFP) gene expressed only when cre is active were generated. Three week old mice were injected with tamoxifen, and 60 days later localization of reporter expression was detected (with X-gal, or with anti-GFP antibody, as indicated). (Left hand and middle panels) Tissue was counterstained with anti-α-SMA antibody (to detect smooth muscle cells) or anti-S100A4 antibody (to detect fibroblasts). Note that some α-SMA expressing cells and some S100A4-expressing cells were positive for COL1A2 promoter/enhancer expression. Tissue was counterstained with DAPI to detect nuclei. (Right hand panel) Tissue was counterstained with anti-PTEN antibody. Note that some PTEN-expressing cells were positive for COL1A2 promoter/enhancer expression. Tissue was counterstained with DAPI to detect nuclei. (B) Control mice (Pten fl/fl) and mice deleted for PTEN (Pten −/−) in cells expressing the COL1A2 promoter/enhancer were generated as described in Methods. Lung sections were stained with α-PTEN antibody and counterstained with DAPI. Representative images are shown. Fig. 1B: N = 3 (avg of 3 fields per 3 lungs), *** = p b 0.001, Student's t-test, +/− SD. (C) Control mice (Pten fl/fl) and mice deleted for PTEN (Pten −/−) in cells expressing the COL1A2 promoter/enhancer were generated as described in Methods and subjected to trichrome staining. A representative stained section is shown. N = 6. (D) Control mice (Pten fl/fl) and mice deleted for PTEN (Pten −/−) in cells expressing the COL1A2 promoter/enhancer were generated as described in Methods. Measurement of COL1A1 and COL1A2 mRNA expression relative to 18S controls confirmed elevation of collagen in PTEN-deficient mice (*, p b 0.05, N = 4). Note that PTEN expression was reduced (p = 0.06, N = 4), but CCN2 and α-SMA mRNA expression in whole lung was not appreciably altered.
38
CCN2 is required for lung fibrosis
Fig. 3. Loss of CCN2 rescues elevated collagen production and increased Akt phosphorylation seen in PTEN-deficient mice. Control mice (Pten fl/fl) and mice deleted for PTEN (Pten −/−) or PTEN and CCN2 (Pten −/−; CCN2 −/−) in fibroblasts were generated as described in Methods. (A) Trichrome staining. N = 6. A representative stained section is shown. (B) Collagen ELISA. Equal amounts of protein from tissue samples from lungs of control mice (Pten fl/fl) (N = 6) and mice deleted for PTEN in fibroblasts (Pten −/−) (N = 5) or PTEN and CCN2 (Pten −/−; CCN2 −/−) (N = 6) in fibroblasts were subjected to an ELISA that detects total collagen (** = p b 0.01). (C). Sections were stained with anti-α-SMA and anti-phospho-Akt antibodies. Tissue was counterstained with DAPI to detect nuclei. Note the presence of cells positive for phospho-Akt in Pten −/− compared to control or Pten −/−; Ccn2 −/− mice. N = 3. Representative images are shown. Note the absence of expansion of α-SMA expressing cells (i.e. the induction of myofibroblasts) in PTEN-deficient mice. N = 3 (avg of 3 fields per 3 lungs), *** = p b 0.001, one-way ANOVA, +/− SD.
39
CCN2 is required for lung fibrosis
the expression of beta-galactosidase in cells expressing cre (COL1A2-lacZ) or green fluorescent protein in cells expressing cre (COL1A2-GFP). COL1A2-lacZ mice were injected with tamoxifen, and assessed for β-galactosidase (using X-gal) expression 60 days post-cessation of tamoxifen injection; tissue sections were counterstained with α-smooth muscle actin (SMA) antibody to detect smooth muscle cells. Reporter activity was found in some α-SMA expressing cells around blood vessels but mostly within fibroblasts (Fig. 1A). To confirm this observation, COL1A2-GFP mice were also injected with tamoxifen, and assessed for GFP expression (using an anti-GFP antibody) expression 60 days post-cessation of tamoxifen injection; tissue was counterstained with anti-α−SMA antibody and anti-S100A4 antibody (to detect fibroblasts). Reporter activity was found again in some α-SMA expressing cells around blood vessels and in S100A4-expressing fibroblasts (Fig. 1A). Reporter activity was localized to a subset of cells endogenously expressing PTEN (Fig. 1A). To begin to assess whether loss of PTEN expression by collagen type I-expressing cells resulted in lung fibrosis, mice deleted (PTEN −/−) or not (PTEN fl/fl) for PTEN in type I collagen (COL1A2) producing cells were generated and genotyped as described previously [7] and in Methods. Loss of PTEN expression around blood vessels and bronchioles and in most fibroblasts in the lung was confirmed using an anti-PTEN antibody (Fig. 1B). PTEN-deficient mice showed fibrosis revealed as visualized by trichrome stain (Fig. 1C). Lungs of PTEN-deficient mice, compared to their wild-type counterparts, also showed statistically significantly elevated expression of Col1A1 and Col1A2 mRNA (Fig. 1D). PTEN-deficient mice showed reduced expression of Pten mRNA (p = 0.06, N = 4) (Fig. 1D). However, when whole lung was examined, CCN2 and α-SMA RNA expression was not altered (Fig. 1D). Loss of CCN2 rescues the fibrotic phenotype of PTEN-deficient lung It was possible that loss of PTEN expression, although not resulting in alterations of CCN2 and α-SMA expression when whole lung was examined, might cause induction of CCN2 or α-SMA protein expression in some areas of the lung. To begin to assess whether the excess collagen deposition caused by loss of PTEN expression in the lung could be caused by CCN2, we first showed that loss of PTEN expression resulted in increased CCN2 expression in some cells around vessels and in some fibroblasts (Fig. 2). We then generated mice deficient for PTEN and CCN2 in fibroblasts, as previously described [10]. Examination of mice deleted for both PTEN and CCN2 confirmed loss of
CCN2 (Fig. 2). Intriguingly, loss of CCN2 rescued the excess collagen production seen in PTEN-deficient animals (Fig. 3A, B). Similarly, loss of PTEN, in a CCN2-dependent fashion, resulted in an increase in p-Akt-positive cells (Fig. 3C). Intriguingly, α-SMA expression was not appreciably altered in PTEN-deficient mice, indicating that although loss of PTEN in collagen type I-expressing cells resulted in increased collagen levels, this induction was not accompanied by the increased appearance of myofibroblasts. These results suggest that loss of PTEN, in a CCN2-dependent fashion, is sufficient to result in increased collage deposition in the lung.
Discussion In this report, we showed that loss of PTEN expression resulted in collagen deposition in the lung in a CCN2-dependent fashion. These data support the notion that alterations in PTEN expression contribute to fibrogenesis in vivo [6,7] and that CCN2 is a key mediator of collagen production in fibrosis [8]. Fibroblasts from patients with IPF have low membrane-associated PTEN expression due to low-expression of caveolin-1 [6]. Also, deletion of PTEN in lung epithelial cells results in injury to lung epithelial cells and subsequent fibrosis [13]. Our previous studies provide evidence that CCN2 expression is induced downstream of loss of PTEN expression and is required for dermal fibrosis caused by loss of PTEN expression [7,10]. The current data indicate that a similar mechanism occurs in our model of collagen deposition in the lung and provides strong support that CCN2 elevation in lung fibrosis may represent a novel therapeutic approach to reduce excessive collagen deposition in these diseases [8,14,15]. It is interesting to note that this involvement of CCN2 occurred even though CCN2 expression appeared to be induced at the protein level in relatively few cells; when whole lung was examined, loss of PTEN did not cause an increase in the overall levels of CCN2 mRNA. CCN2 is also involved with matrix deposition in the invertebral disc; loss of CCN2 exacerbates age-dependent progressive disc degeneration [16– 18]. Future efforts will be expended to assess if loss of PTEN in collagen type I-expressing cells in the lung results in progressive lung fibrosis. Conversely, our studies strongly indicate an important homeostatic role for PTEN in maintaining the extracellular matrix environment in the lungs, as well as the skin. The mechanisms that result in its decrease in IPF are being deciphered [6]; however, CCN2 represents a downstream mediator of at least certain features of fibrosis and may therefore be a suitable target for anti-fibrotic drug intervention.
40
Materials and methods Animal studies
CCN2 is required for lung fibrosis
Santa Cruz). After primary antibody incubation, sections were then washed with PBS and incubated with appropriate fluorescent secondary antibodies (Jackson Immunoresearch) 1 h at room temperature. Sections were washed with PBS, mounted using 4′,6-diamidino-2phenylindole (DAPI) and photographed using a Zeiss fluorescence microscope and Northern Eclipse software (Empix, Missassagua, ON, Canada). Alternatively, signal was developed using the Vectastain ABC Kit (Vector Laboratories) and ImmPACT DAB Peroxidase (HRP) Substrate (Vector Laboratories). When indicated, average fluorescence intensity (three fields/lung) was calculated using Northern Eclipse software (Empix, Missassagua, ON, Canada).
Mice deleted in fibroblasts for PTEN or both PTEN and CCN2 were generated essentially as previously described [7,10]. Briefly, loxP-Pten mice were obtained from The Jackson Laboratory; loxP-Ccn2 mice and mice expressing tamoxifen-dependent Cre recombinase under the control of fibroblast-specific regulatory sequence from the pro-alpha 2 (I) collagen gene were previously described [7,10,19]. Mice hemizygous for Cre and homozygous for loxP-Pten were administered tamoxifen (1 mg/mouse for 5 days) or corn oil at 21–24 days of age to generate mice deleted for the Pten gene specifically in fibroblasts or control mice, respectively. Mice hemizygous for Cre and homozygous for loxP-Pten and loxP-Ccn2 were similarly used to generate mice deleted for PTEN and CCN2. Deletion of PTEN and/or CCN2 was tested by PCR genotyping as described previously [7,10]. Mice were sacrificed by CO2 euthanasia at time points indicated. Thirty-sixty days post-cessation of tamoxifen injection [7,10], lung samples were collected for RT-PCR, histology, immunohistochemistry and for collagen assay. When indicated, mice hemizygous for an allele allowing the expression of tamoxifen-dependent Cre recombinase under the control of fibroblast-specific regulatory sequence from the pro-alpha 2 (I) collagen gene and for an allele allowing the expression of β-galactosidase or green fluorescent protein (GFP) in cells expressing active, nuclear cre recombinase (Jackson Labs) were used. All animal protocols were given ethical approval by the University of Western Ontario's Animal Care Committee.
Lung tissue collected from mice was frozen down at − 80 °C in RNAlater stabilization solution (Invitrogen). Extraction of RNA was done using Trizol (Invitrogen) and chloroform (Sigma) method and 40 ng of RNA samples was run in triplicate. RNA was reverse transcribed and amplified in a 15 μl reaction using TaqMan Assays on Demand primers, 6-carboxyfluroscein-labeled TaqMan MGB probe and Reverse Transcriptase qPCR One-step Mastermix (Quanta, VWR). ABI Prism 7900 HT sequence detector (Perkin-Elmer-Cetus, Vaudreuil, QC) was used according to the manufacturer's instructions to detect amplified sequences. Expression values were standardized to control values from 18S primers using the ΔΔCt method. Statistical analysis on three independent experiments was done using Student's t-test on GraphPad Prism.
Immunohistochemistry and assessment
Total collagen assay
Lung tissue sections (0.5 μm) were cut using a microtome (Leica, Richmond Hill, ON, Canada), collected on Superfrost Plus slides (Fisher Scientific, Ottawa, ON, Canada), de-waxed in xylene and rehydrated by successive immersion in descending concentrations of alcohol. When indicated, tissue was stained with Harris's hematoxylin and eosin or Trichrome. Sections were also subjected for immunofluorescence staining. Sections are boiled in citric acid pH 6 for antigen retrieval. Tissue sections were blocked by incubation with 5% BSA, 0.1% Triton X-100 in phosphate-buffered saline (PBS) for 1 h and then incubated with primary antibodies under humidified conditions at 4 °C overnight. Primary antibodies used were: anti-PTEN (1:1000 dilution, Sigma), anti-SMA (1:500, Sigma), anti-GFP (1:100 Santa Cruz), anti-p-Akt (1:1000, Cell Signaling Technology), anti-S100A4 (1:800, Cell Signaling Technology) and anti-CCN2 (1:100,
To quantify collagen, a Total Collagen Kit that detects hydroxyproline was performed in tissue hydrolysates, as described by the manufacturer (Quickzyme). Tissue underwent complete hydrolysis by adding 100 μL/10 mg of 12 M HCl and boiling at 95 °C for 20 h. Samples were diluted 10 × in 4 M HCl and 35 μL of the diluted sample was incubated with 75 μL buffer supplied by the kit for 20 min in a 96 well plate. Next, 75 μL of detecting reagent was added to each well and incubated at 60 °C for 1 h. A series of standards with known concentrations of 300, 200, 100, 50, 25, 12.5, 6.25 and 0 μg/mL were run with the tissue samples under the same conditions. Absorbance at 570 nm was determined (iMark™ microplate absorbance reader, BioRad) and compared to the standard curve. Total collagen values obtained were subject to statistical analysis using one way ANOVA and Tukey's post-hoc test on GraphPad Prism.
Real-time PCR of lung tissue
41
CCN2 is required for lung fibrosis
Author contributions SK KT JH, CB, MT and SL designed, performed experiments and analyzed data. AL SK KT JH CB MT and SL wrote and edited the paper.
Competing interests None
Acknowledgments This work is supported by grants from the Canadian Institutes of Health Research (MOP-77603) and the Scleroderma Society of Ontario (to AL) (MOP-119410). SKP was supported by the Canadian Scleroderma Research Group. MT was supported by a Master's student scholarship from NSERC. Received 12 November 2014; Received in revised form 22 January 2015; Accepted 23 January 2015 Available online 30 January 2015 Keywords: CCN2; ctgf; PTEN; Lung fibrosis; Scleroderma; Matricellular proteins; CCN family † Co-first authors.
References [1] du Bois RM. Strategies for treating idiopathic pulmonary fibrosis. Nat Rev Drug Discov 2010;9:129–40. [2] Fan MH, Feghali-Bostwick CA, Silver RM. Update on scleroderma-associated interstitial lung disease. Curr Opin Rheumatol 2014;26:630–6. [3] Garneau-Tsodikova S, Thannickal VJ. Protein kinase inhibitors in the treatment of pulmonary fibrosis. Curr Med Chem 2008;15:2632–40. [4] Lagares D, Busnadiego O, García-Fernández RA, Kapoor M, Liu S, Carter DE, et al. Inhibition of focal adhesion kinase prevents experimental lung fibrosis and myofibroblast formation. Arthritis Rheum 2012;64:1653–64.
[5] Shi-wen X, Thompson K, Khan K, Liu S, Murphy-Marshman H, Baron M, et al. Focal adhesion kinase and reactive oxygen species contribute to the persistent fibrotic phenotype of lesional scleroderma fibroblasts. Rheumatology (Oxford) 2012;51:2146–54. [6] Xia H, Khalil W, Kahm J, Jessurun J, Kleidon J, Henke CA. Pathologic caveolin-1 regulation of PTEN in idiopathic pulmonary fibrosis. Am J Pathol 2010;176:2626–37. [7] Parapuram SK, Shi-wen X, Elliott C, Welch ID, Jones H, Baron M, et al. Loss of PTEN expression by dermal fibroblasts causes skin fibrosis. J Invest Dermatol 2011; 131:1996–2003. [8] Leask A. Possible strategies for anti-fibrotic drug intervention in scleroderma. J Cell Commun Signal 2011;5:125–9. [9] Liu S, Shi-wen X, Abraham DJ, Leask A. CCN2 is required for bleomycin-induced skin fibrosis in mice. Arthritis Rheum 2011;63:239–46. [10] Liu S, Parapuram SK, Leask A. Fibrosis caused by loss of PTEN expression in mouse fibroblasts is crucially dependent on CCN2. Arthritis Rheum 2013;65:2940–4. [11] Ponticos M, Holmes AM, Shi-wen X, Leoni P, Khan K, Rajkumar VS, et al. Pivotal role of connective tissue growth factor in lung fibrosis: MAPK-dependent transcriptional activation of type I collagen. Arthritis Rheum 2009;60: 2142–55. [12] Artlett CM. Animal models of scleroderma: fresh insights. Curr Opin Rheumatol 2010;22:677–82. [13] Miyoshi K, Yanagi S, Kawahara K, Nishio M, Tsubouchi H, Imazu Y, et al. Epithelial Pten controls acute lung injury and fibrosis by regulating alveolar epithelial cell integrity. Am J Respir Crit Care Med 2013;187:262–75. [14] Allen JT, Knight RA, Bloor CA, Spiteri MA. Enhanced insulinlike growth factor binding protein-related protein 2 (connective tissue growth factor) expression in patients with idiopathic pulmonary fibrosis and pulmonary sarcoidosis. Am J Respir Cell Mol Biol 1999;21:693–700. [15] Sato S, Nagaoka T, Hasegawa M, Tamatani T, Nakanishi T, Takigawa M, et al. Serum levels of connective tissue growth factor are elevated in patients with systemic sclerosis: association with extent of skin sclerosis and severity of pulmonary fibrosis. J Rheumatol 2000;27: 149–54. [16] Bedore J, Leask A, Séguin CA. Targeting the extracellular matrix: matricellular proteins regulate cell-extracellular matrix communication within distinct niches of the intervertebral disc. Matrix Biol 2014;37:124–30. [17] Bedore J, Sha W, McCann MR, Liu S, Leask A, Séguin CA. Impaired intervertebral disc development and premature disc degeneration in mice with notochord-specific deletion of CCN2. Arthritis Rheum 2013;65:2634–44. [18] Tran CM, Shapiro IM, Risbud MV. Molecular regulation of CCN2 in the intervertebral disc: lessons learned from other connective tissues. Matrix Biol 2013;32:298–306. [19] Ponticos M, Abraham D, Alexakis C, Lu QL, Black C, Partridge T, et al. Col1a2 enhancer regulates collagen activity during development and in adult tissue repair. Matrix Biol 2004;22:619–28.