• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Dev Cell. Author manuscript; available in PMC Feb 1, 2010.
Published in final edited form as:
Dev Cell. Feb 2009; 16(2): 303–313.
doi:  10.1016/j.devcel.2008.12.003
PMCID: PMC2664622
NIHMSID: NIHMS97009

Increased PDGFRα Activation Disrupts Connective Tissue Development and Drives Systemic Fibrosis

SUMMARY

PDGF signaling regulates the development of mesenchymal cell types in the embryo and in the adult, but the role of receptor activation in tissue homeostasis has not been investigated. We have generated conditional knockin mice with mutations in PDGFRα that drive increased kinase activity under the control of the endogenous PDGFRα promoter. In embryos, increased PDGFRα signaling leads to hyperplasia of stromal fibroblasts that disturbs normal smooth muscle tissue in radially patterned organs. In adult mice, elevated PDGFRα signaling also increases connective tissue growth, leading to a progressive fibrosis phenotype in multiple organs. Increased PDGFRα signaling in an Ink4a/Arf-deficient genetic background leads to accelerated fibrosis, suggesting a new role for tumor-suppressors in attenuating fibrotic diseases. These results highlight the role of PDGFRα in normal connective tissue development and homeostasis, and demonstrate a pivotal role for PDGFRα signaling in systemic fibrosis diseases.

Keywords: PDGF, Fibrosis, Connective tissue, Scleroderma

INTRODUCTION

Platelet derived growth factors (PDGF) exert their biological effects through the binding and activation of two receptor tyrosine kinases, PDGFRα and β. Ligand binding to PDGFRα/β on the cell surface induces receptor oligomerization and tyrosine phosphorylation, which activates a number of downstream signal transduction pathways, including Ras/MAP kinases, PI3 kinase/AKT, and PLC/PKC pathways. The cellular responses to PDGF signaling include proliferation, survival, migration, and control of differentiation, and PDGFRα signaling serves critical functions during embryo development (Andrae et al., 2008; Hoch and Soriano, 2003; Soriano, 1997), however the consequences of increased signaling in development and in the adult are unknown. PDGFRα-activating mutations might be expected to disrupt some or all of the tissues where developmental functions have been previously identified. However most mesenchymal progenitors in the early embryo express PDGFRα, and therefore increased signaling could have more broad effects and has the potential to suggest new developmental roles of PDGFRα.

PDGF receptors are tightly regulated by an auto-inhibitory allosteric conformation resulting in very low basal activity in the absence of ligand. Constitutively active versions of PDGF receptors carrying point mutations that relieve auto-inhibition were first generated and characterized in tissue culture (Irusta and DiMaio, 1998). More recently, a small percentage (~5%) of gastrointestinal stromal tumors (GIST) have been found to carry small deletions, insertions, and point mutations in PDGFRα that result in constitutive activity (Heinrich et al., 2003; Hirota et al., 2003). In GISTs there is a strong mutational bias towards PDGFRα mutations in the activation loop of the kinase domain (82–93%), with mutations in the juxtamembrane domain being the second most frequent site of mutation (7–13%)(Corless et al., 2005). Two weakly-activating PDGFRα mutations have been associated with familial cases of GIST or GIST variants (Chompret et al., 2004; de Raedt et al., 2006), however, developmental defects have yet to be linked to strongly activating mutations in this receptor.

Aberrant PDGFR signaling also has been implicated in diverse fibrotic conditions (Andrae et al., 2008) where fibroblasts proliferate and deposit excessive connective tissue matrix, leading to progressive scarring and organ dysfunction. Systemic sclerosis, or scleroderma, is an autoimmune disease afflicting an estimated 300,000 people in the United States. The disease is characterized by chronic autoimmune reactions, microvascular restriction, and widespread fibrosis of the skin and other organs including the gastrointestinal system, skeletal muscle, heart, kidney, and lungs, frequently resulting in organ failure and death (Samter, 1988). Although the initiating events for this disease are unknown, and no single model explains all scleroderma-related pathology, it was recently shown that systemic sclerosis patients circulate auto-activating antibodies for PDGFRα and β in their serum, suggesting that systemic autoantibody stimulation of PDGFR-expressing fibroblasts could be involved in the disease (Baroni et al., 2006). Experimental activation of TGFβ signaling has been shown to recreate fibrotic pathology in the skin (Sonnylal et al., 2007), and this may act in part by increasing the level of PDGF signaling (Bonner, 2004; Yamakage et al., 1992). Nevertheless, there is currently no animal model representing a PDGF component of the disease (Christner and Jimenez, 2004).

To investigate the in vivo consequences of increased PDGFRα signaling and to gain new understanding of the developmental biology of this receptor and the consequences for disease in adulthood, we created activatable alleles with higher intrinsic kinase activity at the PDGFRα locus. Two separate mouse lines were generated for conditional expression of mutant PDGFRα with either juxtamembrane or kinase domain mutations that mimic somatic mutations identified in human GIST. The resulting phenotypes were similar in type but differed in severity, thus providing novel insight into the mutation bias seen in human GIST. The developmental phenotypes resulting from either mutation demonstrate an important role for PDGFRα in the balanced expansion of mesenchymal cells during connective tissue development. In adult animals, increased PDGFRα activation also leads to connective tissue hyperplasia and to progressive, chronic fibrosis in many organs. Finally, we find that loss of the Ink4a/Arf tumor suppressor synergizes with aberrant PDGFRα signaling in both tumorigenesis and the development of fibrotic disease.

RESULTS

Derivation of Knockin Mice

Previously, we described a generic knockin vector that drives expression of a PDGFRα cDNA from the endogenous promoter while simultaneously disrupting expression of the endogenous PDGFRα gene (Klinghoffer et al., 2001). To explore the consequences of mutant PDGFRα signaling, we modified this vector by inserting a lox-stop-lox cassette so that transcription is blocked until Cre-recombination removes the stop cassette (Figure 1A). In the absence of Cre-recombination, knockin mice are heterozygous for PDGFRα and phenotypically normal (Soriano, 1997). We generated two lines of knockin mice harboring different activating mutations that have been repeatedly isolated from human tumors (Figure 1B). The first mutation (D842V), designated αK, is located in the kinase domain and is thought to interfere with the inactive conformation of the ATP-binding pocket, which leads to constitutive activity. It is the single most common PDGFRα mutation found in human GISTs (Corless et al., 2005). The second mutation (V561D), designated αJ, is the most common PDGFRα juxtamembrane domain mutation found in GISTs. This mutation is thought to lead to constitutive activity by disrupting inhibitory contacts between the juxtamembrane and kinase domains that are important for full auto-inhibition (Hubbard, 2004). Correctly targeted ES cell clones were identified by Southern blot analysis (Figure 1C and Figure S1) and used to derive germline chimeras for the αK and αJ strains.

Figure 1
Constitutive and Inducible Signaling by PDGFRα Mutant cDNA Knockins

PDGFRα+/(S)K and PDGFRα+/(S)J heterozygous mice that carried the inactivating lox-stop-lox cassette, designated (S), were then crossed to Cre-expressing strains to obtain embryos or mice with activated alleles. We initially tested the activated alleles by crossing with the Meox2-Cre line, which drives recombination in the epiblast and therefore all cell lineages of the embryo proper (Tallquist and Soriano, 2000). Western blots of PDGFRα+/+, PDGFRα+/K and PDGFRα+/J embryo lysates showed similar expression levels of the receptor in all lines (Figure 1D). Immunoprecipitation of PDGFRα followed by phosphotyrosine Western blot indicated that mutant receptors were hyperphosphorylated, reflecting their increased activation (Figure 1D).

Constitutive and Inducible PDGFRα Signaling in Mutant Fibroblasts

PDGFRα activating mutations like D842V and V561D allow high basal kinase activity without the addition of ligand (Heinrich et al., 2003; Hirota et al., 2003), but it is unclear if these isoforms can be further stimulated by ligand. To determine both basal activity and examine the potential for superactivation in the presence of ligand, we cultured fibroblasts from mutant embryos and examined PDGFRα phosphorylation. After 48 hours of serum starvation, PDGFRα+/K and PDGFRα+/J fibroblasts retained elevated basal phosphorylation of PDGFRα compared to wild type cells (Figure 1E), consistent with the earlier work. Additionally, PDGFRα from mutant cells became hyperphosphorylated after a pulse of PDGF-AA (Figure 1E). This result indicates that PDGFRα+/K and PDGFRα+/J cells remain responsive to their cognate ligands, with stronger activation of the αK mutant receptor compared to αJ (Figure S2). To test the contribution of the wild type receptor in this response, we derived PDGFRα−/K fibroblasts expressing only αK receptors. In these cells a pulse of PDGF-AA still induced elevated tyrosine phosphorylation of αK receptors (Figure 1F). It was not possible to generate PDGFRαK/K or PDGFRαJ/J mutant cell lines because PDGFRα+/K and PDGFRα+/J pups suffer neonatal and juvenile lethality (see below), precluding further crosses. Because mutant PDGFRα molecules can respond to exogenous ligand, we conclude that they localize to the plasma membrane and retain dimerization capabilities. This suggests that developmental responses relying on a directional source of ligand, such as chemotaxis by migrating fibroblasts, may still function in the presence of mutant receptors.

We also measured the phosphorylation state of downstream signaling molecules to see if increased signaling altered specific pathways. There was little difference in the phosphorylation status of Erk1/2 on Thr202/Tyr204 between mutant and wild type fibroblasts, with all cell lines displaying low basal phosphorylation and prominent stimulation by PDGF-AA after 10 minutes. Similar results were seen for Akt phosphorylation on Ser473 under basal conditions or in response to ligand. In contrast, PLCγ1 was constitutively phosphorylated on Tyr783 in mutant fibroblasts, and this modification appeared to be stronger in PDGFRα+/K cells compared to PDGFRα+/J cells (Figures 1G, S2). Based on these observations we predict that some biological processes could remain unperturbed by activating Pdgrα mutations as long as differential PLCγ1 activation is not critical.

PDGFRα+/K Embryos Exhibit Defects in Lung Development, Edema, and Aberrant Skeletal Growth

We examined embryonic and neonatal viability in PDGFRα+/K and PDGFRα+/J mice because human congenital defects have not yet been linked to strongly activating mutations in PDGFRα. Genotyping of offspring from crosses between PDGFRα+/(S)K (or PDGFRα+/(S)J) and Meox2-Cre parents at postnatal day 7–14 indicated no PDGFRα+/K survivors (0/132 offspring) and a lower than predicted frequency of PDGFRα+/J survivors (44/220 offspring = 20%). However, PDGFRα+/K and PDGFRα+/J embryos were recovered at the expected frequency until the time of birth (Table S1). PDGF-A signaling through PDGFRα has previously been shown to regulate the spreading and proliferation of smooth muscle progenitors during postnatal alveolar development (Boström et al., 1996). Histological analysis of lungs at E18.5 showed a failure of terminal sack morphogenesis, indicating that respiratory failure was a likely cause of rapid death in PDGFRα+/K neonates. Instead of accomplishing the final stage of embryonic lung development, PDGFRα+/K embryos only reached the canalicular stage, and exhibited a ~1.6-fold increase in mesenchymal cell density and widespread PDGFRα expression when compared to PDGFRα+/J and wild type lungs (Figures 2A,B and S3). The PDGFRα+/K lung phenotype is similar to that reported for transgenic overexpression of PDGF-A under the surfactant protein C promoter (Li and Hoyle, 2001) except that PDGFRα+/K lungs were not larger than controls (data not shown). Loss of PDGFRα signaling is known to affect somite-derived structures, resulting in smaller embryos with subepidermal edema, as well as aberrant growth or fusion of rib and vertebral elements (Soriano, 1997). PDGFRα+/J embryos appeared normal in all of these aspects, but PDGFRα+/K embryos were notably larger from E12.5 – E16.5 with uncondensed mesenchyme and edema of the torso (Figure 2C,D). During the final days of embryogenesis the edema was normalized with the development of enlarged lymphatic structures (data not shown). Preparations of wild type and PDGFRα+/K skeletons at E18.5 revealed a wide, unsegmented sternum (Figure 2E,F) that was likely the result of incomplete contact between ribs and unfused sternal bands at E16.5 (Chen, 1953) (Figure 2G,H). We conclude that the increased severity of the PDGFRα+/K phenotype relative to that seen in PDGFRα+/J mutants may reflect a threshold of PDGF signaling that is only achieved by the more potent kinase domain mutation.

Figure 2
Phenotypic Analysis of PDGFRα+/K Embryos

Increased PDGFRα Signaling Disrupts the Balanced Development of Connective Tissue and Smooth Muscle

Although PDGFRα+/J mutants did not present a lung phenotype and most survived until weaning, they all developed bilateral hydronephrosis and distention of the proximal ureter (Figure 3A–C). Enlargement of the renal pelvis was apparent in PDGFRα+/J and PDGFRα+/K embryos as early as E16.5 (Figure 3C and data not shown). Urine transport from the kidney to the bladder requires smooth muscle contraction in the ureter; smooth muscle defects can result in functional obstruction of the ureter, which has been found to account for hydronephrosis in other mouse models (Mendelsohn, 2006). A normal ureter is composed of a specialized urothelium surrounded by smooth muscle and two rings of stromal cells that generate connective tissue (Figure 3D,G). In PDGFRα+/K and PDGFRα+/J mutants, normal patterning was apparent when visualized by staining with antibody for smooth muscle actin (αSMA). However, stromal cells were hyperplastic and smooth muscle cells were poorly condensed (Figure 3E,F). We next examined PDGFRα−/− embryos to see if this receptor might be required in ureter development. PDGFRα−/− ureters developed smooth muscle and connective tissue populations, but were slightly smaller than wild type controls (Figure S4), suggesting partial compensation by other signals in normal development. Smooth muscle differentiation commenced at the expected time (E15.5) in PDGFRα+/J and PDGFRα+/K ureters, as determined by αSMA expression, and serial sections of mutant urogenital tracts at E15.5 or E18.5 did not reveal obstructions of the lumen (data not shown). Instead, stromal hyperplasia appeared to disrupt the formation of a tight smooth muscle layer, leading to a non-functional ureter. In surviving PDGFRα+/J mice, the excessive stroma eventually differentiated as a thick sheath of collagenous connective tissue (Figure 3G,H). Connective tissue hyperplasia thus represents a novel developmental mechanism for functional obstruction of the urinary tract.

Figure 3
Hyperplasia of Radially Patterned Gut and Ureter Mesenchyme in PDGFRα+/J and PDGFRα+/K Embryos

We also noticed selective hyperplasia of stromal cells throughout the digestive system, but most strikingly in the esophagus and stomach (Figure 3I–K and data not shown). Radial patterning of the digestive system and urinary tract involves conserved epithelial-mesenchymal signaling (Mendelsohn, 2006), but a role for PDGF in this process has not been previously identified. We used Tbx18 as a specific marker to identify the uretral mesenchyme (Figure 4A)(Airik et al., 2006) and compared the expression of PDGF genes. PDGFA ligand was expressed in the epithelium of developing ureters and gut at E12.5, and PDGFC ligand was expressed in both the epithelium and a sub-population of mesenchyme (Figure 4B,C). PDGFRα, the principal receptor for these ligands, was broadly expressed in mesenchyme, including renal mesenchyme (Figure 4D). This pattern of PDGF ligand and receptor expression persisted in wild type ureters from E13.5–E15.5 (Figure 3E–G and data not shown), and a similar pattern was seen in the E12.5 and later digestive system (Figure 4H–J and data not shown). Primary fibroblasts from PDGFRα+/J and PDGFRα+/K embryos have increased growth potential when cultured in vitro (Figure S3), and we therefore expected to find a proliferation difference underlying the hyperplasic phenotypes. Proliferation was not significantly increased at E13.5 in the PDGFRα+/K ureter compared to the wild type (Figure 4K–M), but in the esophagus the proliferation difference was significantly increased (Figure 4O–Q). Even at E13.5 there were many more mesenchymal cells per section of PDGFRα+/K ureter than in wild type sections (Figure 4N), suggesting that, in addition to proliferation, enhanced migration could endow the developing ureter with a larger initial complement of uretral mesenchyme. Consistent with the potential involvement of such a mechanism, PDGFRα+/K cells showed increased migration towards PDGF-AA compared to wild type fibroblasts (Figure 4S). Therefore, αK receptors with high basal phosphorylation are not blinded to a directional source of ligand and increased chemotaxis could contribute to the hyperplastic phenotypes.

Figure 4
Mesenchymal Hyperplasia During Early Stages of Radially Patterned Organogenesis

Widespread Fibrosis of Skin and Internal Organs Following Mosaic PDGFRα Activation in Adult Mice

Based on the connective tissue phenotypes we observed in the embryo, and to explore the potential role of PDGF signaling in adult fibrotic conditions, we induced mosaic activation of αJ and αK alleles using ROSA26-CreER mice that ubiquitously express a tamoxifen-inducible Cre enzyme (Badea et al., 2003). We first generated cohorts by crossing PDGFRα+/K or PDGFRα+/J mice with homozygous ROSA26CreER/CreER mice to obtain mutant and wild type animals each with one copy of ROSA26-CreER. We regulated the timing of activation by treating with a single pulse of tamoxifen (Tam, see methods) and achieved recombination efficiencies ranging between 10% and 30% (average 24% +/−7.2%), as measured by real-time PCR at 5 days post injection. To visualize this effect in situ we injected Tam into ROSA26-CreER mice also heterozygous for the ROSA26R-lacZ allele (Soriano, 1999). Frozen sections from these ROSA26CreER/RlacZ mice showed mosaic β-galactosidase expression in nearly all tissues (e.g. gut shown in Figure 5B).

Figure 5
Gastrointestinal Fibrosis and Polyp Formation in Adult Mutants

Cohorts of Tam-treated mice expressing wild type, αJ, and αK alleles were allowed to age for up to 76 weeks. Initial evaluation of the mice did not reveal any change in morphology or health status, and all animals grew to the normal size and weight. Starting at 6 months after Tam-treatment, PDGFRα+/K mice began to develop symptoms of intestinal disease and eventually died or had to be sacrificed (Figure 5C). All PDGFRα+/K mice that underwent necropsy (n = 17) presented partial or complete fibrosis of the small intestine and cecum, and histological examination showed increased submucosal connective tissue with a proportional increase in fibroblast-like cells and collagenous matrix (Figure 5D–I). Additionally, in 9/17 PDGFRα+/K mutants a variable number of firm, white polyps ranging in size from 1mm to 10mm were associated with fibrotic GI tract (Figure 5J). Of the PDGFRα+/J mutants that underwent necropsy, 4/9 exhibited intestinal fibrosis and one had polyps (data not shown). GISTs are thought to arise from interstitial cells of Cajal (ICCs), which are pacemaker cells that associate with smooth muscle and express CD117/Kit. Constitutive activation of the Kit receptor tyrosine kinase leads to ICC hyperplasia and serves as a mouse model for GIST (Rubin et al., 2005; Sommer et al., 2003). None of our mice developed ICC hyperplasia, and fibrotic intestine and polyps did not express increased Kit or αSMA, but did increase expression of PDGFRα (Figure 5K–N). This demonstrates a clear difference between the responses of cells that are sensitive to Kit or PDGFRα mutation, even though both RTKs have been implicated in the pathogenesis of GISTs.

In addition to gastrointestinal fibrosis, all PDGFRα+/K and some PDGFRα+/J mutants developed tight skin that adhered strongly to underlying muscle. Histological analysis revealed that this phenotype corresponded to structural changes in all subepidermal connective tissue layers, with an expanded network of dense collagen fibers and a proportional increase in the number of fibroblasts (asterisk in Figure 6B,C). Connective tissue fibrosis was also noted in heart muscle (arrows in Figure 6E), around bronchioles (asterisk in Figure 6G), in skeletal muscle (asterisk in Figure 6I), and in the kidney glomeruli and renal interstitium while leaving vascular smooth muscle unperturbed (Figure S5). As in the digestive system, a milder and slower developing fibrosis phenotype was noted in PDGFRα+/J mutants (quantified for various organs in Figure 6K). Expression of PDGFRα was present in cells residing within regions of fibrosis, and foci of higher expression could also be found, presumably representing active lesions (Figure 6J). Consistent with the in vitro analysis (Figure 1G), fibrotic tissue displayed elevated PLCγ1 phosphorylation (Figure 6L), suggesting a role for PLCγ1/PKC signaling in the observed fibrotic responses.

Figure 6
Systemic Organ Fibrosis in Adult PDGFRα+/J and PDGFRα +/K Mice

Loss of Ink4a/Arf Enhances Tumor Initiation and Fibrosis Driven by PDGFRα

Aberrant PDGF signaling has been associated with several human tumors including sarcomas (Ostman and Heldin, 2007), and we therefore expected some incidence of neoplasia in our mice. However, spontaneous tumors were rare and late to appear in Tam-induced cohorts on a wild type B6/129 background. Among 29 PDGFRα+/K mice, five developed sarcomas between 44 and 76 weeks of age (mean 56 weeks). These sarcomas appeared to arise from the dermis or muscle connective tissue, and presented numerous mitotic figures, pleiomorphic nuclei, little collagen content, and infiltration of surrounding adipose and muscle. PDGFRα was robustly expressed in these sarcomas (Figure S5). None of the PDGFRα+/+ or PDGFRα+/J cohort mice (n = 35, 33 respectively) developed sarcomas by 76 weeks of age.

The above data suggests that increased PDGFRα activation could be sufficient for sarcoma formation in the absence of other experimental gene alterations. Oncogenes often cooperate with additional mutations that disrupt tumor suppressor pathways, such as INK4a/ARF, which is frequently inactivated in malignant and recurrent forms of GIST (Sabah et al., 2004). To examine the role of the tumor suppression response in mutant PDGFRα signaling, we generated further cohorts of Ink4a/Arf and PDGFRα compound mutants. PDGFRα+/+;Ink4a/Arf−/− mice had an average tumor-free survival of >20 weeks (Figure 7A), consistent with previous studies with Ink4a/Arf−/− mice (Serrano et al., 1996). Loss of Ink4a/Arf synergized with increased PDGFRα activation, such that all Tam-treated PDGFRα+/K;Ink4a-Arf−/− mice (n=40) developed PDGFRα-expressing sarcomas of the skin or muscle connective tissue after a latency period of 10–19 weeks (Figure 7A,B). These sarcomas histologically resembled those described earlier in PDGFRα+/K mice (Figure 7C–E and Figure S5). Two PDGFRα+/K;Ink4a/Arf−/− mice also developed tumors of the large intestine that most resembled undifferentiated fibrosarcomas and did not express TMEM16A (Figure S6), a marker for GIST (West et al., 2004). In addition to sarcomas, all PDGFRα+/K;Ink4a/Arf−/− mice developed accelerated fibrosis of the skin and muscle at 10–19 weeks post-Tam that was not evident in similarly aged PDGFRα+/K or Ink4a/Arf−/− mice (Figure 7F–I). Gastrointestinal fibrosis was also often seen in PDGFRα+/K;Ink4a/Arf−/− mice that survived >15 weeks (data not shown). Taken together, this data shows that loss of Ink4a/Arf cooperates with increased PDGFRα activity in sarcoma formation, and furthermore accelerates PDGFRα-driven fibrosis, suggesting a previously unappreciated role for Ink4a/Arf in limiting fibrotic diseases.

Figure 7
Loss of Ink4a/Arf Accelerates PDGFRα-driven Tumorigenesis and Fibrosis

DISCUSSION

In this study we explored the role of mouse PDGFRα by targeting mutant cDNAs with higher intrinsic kinase activity to the PDGFRα locus; the effects of increased PDGFRα activation are thereby limited to cells that normally express the receptor. To mimic separately the consequence of germline or somatic mutation, we activated cDNA expression early in embryogenesis with Meox2-Cre, or in a mosaic fashion in older mice using a tamoxifen-inducible method. From the results of these experiments, we conclude that a major biological outcome of increased PDGFRα signaling is the activation of cellular programs that generate connective tissue in both the embryo and adult.

The ureters and digestive tract are radially patterned tubes composed of concentric layers of connective tissue and smooth muscle surrounding a specialized epithelium and a lumenal space. Cells lining the lumen express hedgehog signals from E11.5 onward to regulate mesenchymal proliferation and patterning (Ramalho-Santos et al., 2000; Yu et al., 2002). We found that PDGFRα and its principal ligands were also expressed during this time in a pattern that could affect epithelial-mesenchymal signaling during development of these organs. Increased PDGFRα signaling affected stromal fibroblasts, which became hyperplastic and generated fibrotic overgrowths. The smooth muscle lineage was not directly affected, but we identified situations where smooth muscle structure was disrupted by connective tissue overgrowth. Because PDGF ligands were expressed in the prospective urothelium and mutant fibroblasts were more responsive in a chemotaxis assay, PDGF signaling might regulate the initial migration of metanephric or other mesenchyme toward the distal ureteric bud. Fibroblasts isolated from mutant embryos exhibited both constitutive and inducible Pdgrα signaling, which is also consistent with enhanced chemotaxis towards a source of ligand in vivo. Finally, mesenchymal progenitors in the ureter and esophagus were more proliferative when they expressed mutant PDGFRα. We conclude that increased proliferation and enhanced migration likely account for the stromal hyperplasia seen in the developmental setting.

Two different PDGFRα-activating point mutations were tested in this study, and we hypothesized that different biological outcomes could result from these mutations because analogous mutations in the closely-related KIT receptor predict different spectrums of neoplasia in humans: juxtamembrane domain mutations produce GISTs, while kinase domain mutations are commonly found in lymphomas, leukemias, and testicular tumors (Corless et al., 2004). A biological difference between the two PDGFRα mutations was clearly revealed in our models by the increased severity of mesenchymal defects and fibrosis in PDGFRα+/K mice as compared to PDGFRα+/J mice. Unlike human KIT mutations however, our PDGFRα mutations generated a similar spectrum of phenotypes with differing severity. Also, the increased severity of αK-driven phenotypes is concordant with the observed bias towards PDGFRα D842V in human tumors. While neither of the PDGFRα mutations we tested has been reported as a germline mutation in human families, the V561D mutation was identified in a single individual exhibiting gastrointestinal stromal tumors plus multiple recurrent fibroid polyps (Pasini et al., 2007), similar to the polyps that we observed in this study.

A few sarcomas appeared in our mutant mice with increased PDGFRα activation, but only after long latency, implying other gene alterations might cooperate in tumorigenesis. One of the most frequently mutated loci in human cancers is INK4a/ARF (Ruas and Peters, 1998), and previous studies in mice have shown a synergistic effect on neoplasia when oncogenic RTK signaling is combined with loss of Ink4a/Arf (Sharp et al., 2002). More specifically, loss of Ink4a/Arf has been shown to facilitate glioblastoma progression in a mouse model based on PDGF-PDGFR autostimulation (Dai et al., 2001; Hesselager et al., 2003). When we induced PDGFRα mutations in an Ink4a/Arf-deficient background to investigate cooperative effects in tumorigenesis, sarcoma formation was accelerated. Furthermore, gastrointestinal polyps were never seen to progress to malignancy in PDGFRα+/K or PDGFRα+/J mice with wild type Ink4a/Arf, but two gastrointestinal tumors arose in young PDGFRα+/K;Ink4a/Arf−/− mice. These tumors appeared to be Kit-negative and most resembled poorly differentiated fibrosarcomas, although there is a paucity of specific markers for GISTs that do not express Kit. Because GISTs are generally diagnosed in older patients, it remains possible that PDGFRα-driven GISTs would eventually appear in the mice if fibrosis and other sarcomas did not shorten the animal’s lifespan.

Increased PDGFRα signaling in both the embryo and in the adult led to connective tissue hyperplasia and increased extracellular matrix deposition. In the adult, this resulted in fibrosis with a striking phenotypic overlap with the human autoimmune disease, systemic sclerosis. Skin tightening was observed with age due to excessive connective tissue deposition, and furthermore the skeletal muscle, heart, lung, kidney, and intestine became scarred with infiltrates of matrix-producing fibroblasts. Interestingly, we noted that experimental loss of Ink4a/Arf dramatically accelerated the development of fibrotic lesions, suggesting that cells are able to limit the effects of PDGFRα signaling through an Ink4a/Arf-dependent pathway. This property may relate to previously known tumor suppression roles of Ink4a/Arf, which is known to respond to stress pathways and hyperproliferative signals by enhancing the growth suppressing activities of p53 and p16/Rb. It is clear that cell senescence plays an important role in suppressing tumor growth in vivo; however it was recently shown that acute fibrosis in chemically treated murine livers can be limited by a senescence response (Krizhanovsky et al., 2008). Therefore, PDGFRα hyperproliferative signaling could lead to Ink4a/Arf-dependent senescence of activated fibroblasts, thereby suppressing both tumorigenesis and fibrosis.

Previous investigations have suggested a role for PDGF signaling in the fibrosis of individual organs, including the heart (Ponten et al., 2005), liver (Campbell et al., 2005), skin (Distler et al., 2007), and lung (Abdollahi et al., 2005). Altered TGFβ1 signaling has also been implicated in many fibrotic conditions, and consistent with the model that PDGF acts independently or downstream of TGFβ1, we did not detect changes in Smad2/3 phosphorylation in unstimulated mutant fibroblasts (data not shown). Our work provides in vivo evidence that increased PDGFRα signaling can be sufficient to drive systemic connective tissue disease in the mouse tissues and organ systems that are relevant to human systemic sclerosis. Because systemic autoantibody stimulation of PDGFRs has recently been hypothesized to influence systemic sclerosis (Baroni et al., 2006), it could be informative to use this type of an approach to determine if increased PDGFRβ signaling can contribute to fibrosis in other organs or produce the vascular restriction phenotypes also seen in this disease. This work also establishes an excellent animal model for testing novel therapeutic approaches for blocking aberrant PDGFRα signaling in human systemic sclerosis.

EXPERIMENTAL PROCEDURES

Mice

A detailed description of ES cell targeting and vector design is provided in Supplemental Procedures. Protocols for the derivation of mutant mice were approved by our IACUC committee. Embryos and pups generated with Meox2-Cre were examined in both a mixed B6/129 and pure 129 backgrounds, with identical phenotypes. ROSA26-CreER mice were obtained from the Jackson Laboratories and Ink4/aArf−/− mice were obtained from the NCI mouse repository.

To generate mouse cohorts for survival studies, PDGFRα+/(s)J or PDGFRα+/(s)K heterozygotes were crossed with ROSA26CreER/CreER homozygotes. Tamoxifen (Sigma) dissolved in corn oil (20mg/ml) was administered by intraperitoneal injection under one of two schedules: 1) pregnant females treated with 60mg/kg Tam at embryonic day 18.0 to generate cohort offspring, or 2) cohort offspring treated with 200mg/kg Tam at postnatal day 30. Recombination efficiency in test animals was determined as described in Supplementary Data. Both schedules gave the same phenotypes after similar latency, and therefore mice from both groups are presented together (in Kaplan-Meier survival plots, etc.).

To visualize tamoxifen-induced Cre-recombination in tissues by activation of the ROSA26RlacZ reporter (Soriano, 1999), ROSA26CreER/RlacZ mice were treated with both treatment schedules described above then sacrificed after 5 days. Nearly all tissues exhibited mosaic recombination except the brain.

To generate cohorts on a tumor-susceptible background, PDGFRα+/(S)K;Ink4a/Arf+/− trans-heterozygotes were crossed to ROSA26CreER/CreER;Ink4aArf−/− double homozygotes, and Tam was administered on schedule 1 described above. All cohort mice generated with ROSA26-CreER were examined in a mixed B6/129 background.

Cells

To generate mouse embryonic fibroblasts (MEFS), individual E13.5 embryos were isolated from crosses between PDGFRα+/(S)J or PDGFRα+/(S)K heterozygotes and Meox2+/Cre heterozygotes. Alternatively, Meox2+/Cre;PDGFRα+/− double heterozygotes were used in these crosses to produce PDGFRα−/K MEFs. Additional experimental details are described in Supplementary Data.

Immunoprecipitation and Western Blots

Cleared lysate was quantified for protein concentration (DC Protein Assay, Bio-Rad), normalized across samples, and then immunoprecipitated with a rabbit anti-PDGFRα antibody 27P (a kind gift from Dr. Andrius Kazlauskas) and then purified with Protein-A agarose (Sigma). Immunoprecipitates were separated on 7.5% SDS-PAGE gels, and blotted on membranes with anti-phosphoTyrosine antibody 4G10 (Upstate). After stripping, membranes were re-probed for total PDGFRα. For signaling proteins, cleared and normalized lysates were separated on 12% SDS-PAGE gels and blotted on membranes with rabbit antibodies against phosphoAKT (Ser473), phosphoErk1/2 (Thr202/Tyr204), and phosphoPLCγ1 (and unphosphorylated proteins) according to manufacturers instructions (Cell Signaling).

Tissues and Histology

Embryos and adult tissue were fixed with either 10% buffered formalin (Fisher) or Bouin’s solution (Sigma), paraffin embedded, and sectioned at 8 microns before staining. Fibrosis in various organs was quantified in three mice per genotype by staining with picosirius red, followed by photography and image analysis to measure the percent of collagen per tissue area (ImageJ). For immunohistochemistry,α-SMA antibody (1A4, Sigma) or phospho-PLCγ1 antibody (Cell Signaling), was used with avidin-biotin amplification and DAB staining (ABC kit, Vector labs). For bromodeoxyuridine (BrdU) labeling, we injected 50mg Brdu/kg for a 90-minute pulse. Proliferating cells (BrdU+) were detected with a BrdU antibody at 1:20 dilution (Sigma). Hematoxylin was used as a counterstain for immunohistochemistry. For in situ hybridizations, formalin-fixed tissues were cryoprotected in 20% sucrose solution and cryosectioned at 14-micron thickness. Dig-labelled RNA probes were synthesized with DIG RNA labeling mix (Roche) from cDNA templates (see Table S3 for probe information). Hybridized probes were detected with AP-conjugated anti-Dig Fab fragments (Roche) and developed with NBT/BCIP (Roche). β-galactosidase activity from ROSA26R-lacZ tissue was assayed as previously described (Soriano, 1999) with development at 37°C overnight. Eosin or nuclear fast red were used as counterstain for in situ hybridizations and β-gal assays.

Supplementary Material

01

Acknowledgments

We thank Philip Corrin and Alan Wong for expert technical assistance, Micah Greenberg for care of the mouse colony, and Sue Knoblaugh for advice on tumor pathology. We thank our laboratory colleagues and Susan Parkhurst for critical comments on the manuscript. LEO was supported by a fellowship from the American Cancer Society. This work was supported by grants RO1HD24875 and R37HD25326 from the National Institute for Child Health and Human Development to PS.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • Abdollahi A, Li M, Ping G, Plathow C, Domhan S, Kiessling F, Lee LB, McMahon G, Grone HJ, Lipson KE, Huber PE. Inhibition of platelet-derived growth factor signaling attenuates pulmonary fibrosis. J Exp Med. 2005;201:925–935. [PMC free article] [PubMed]
  • Airik R, Bussen M, Singh MK, Petry M, Kispert A. Tbx18 regulates the development of the ureteral mesenchyme. J Clin Invest. 2006;116:663–674. [PMC free article] [PubMed]
  • Andrae J, Gallini R, Betsholtz C. Role of platelet-derived growth factors in physiology and medicine. Genes Dev. 2008;22:1276–1312. [PMC free article] [PubMed]
  • Badea TC, Wang Y, Nathans J. A noninvasive genetic/pharmacologic strategy for visualizing cell morphology and clonal relationships in the mouse. J Neurosci. 2003;23:2314–2322. [PubMed]
  • Baroni SS, Santillo M, Bevilacqua F, Luchetti M, Spadoni T, Mancini M, Fraticelli P, Sambo P, Funaro A, Kazlauskas A, et al. Stimulatory autoantibodies to the PDGF receptor in systemic sclerosis. N Engl J Med. 2006;354:2667–2676. [PubMed]
  • Bonner JC. Regulation of PDGF and its receptors in fibrotic diseases. Cytokine Growth Factor Rev. 2004;15:255–273. [PubMed]
  • Bostrom H, Willetts K, Pekny M, Leveen P, Lindahl P, Hedstrand H, Pekna M, Hellstrom M, Gebre-Medhin S, Schalling M, et al. PDGF-A signaling is a critical event in lung alveolar myofibroblast development and alveogenesis. Cell. 1996;85:863–873. [PubMed]
  • Campbell JS, Hughes SD, Gilbertson DG, Palmer TE, Holdren MS, Haran AC, Odell MM, Bauer RL, Ren HP, Haugen HS, et al. Platelet-derived growth factor C induces liver fibrosis, steatosis, and hepatocellular carcinoma. Proc Natl Acad Sci U S A. 2005;102:3389–3394. [PMC free article] [PubMed]
  • Chen JM. Studies on the morphogenesis of the mouse sternum. III. Experiments on the closure and segmentation of the sternal bands. J Anat. 1953;87:130–149. [PMC free article] [PubMed]
  • Chompret A, Kannengiesser C, Barrois M, Terrier P, Dahan P, Tursz T, Lenoir GM, Bressac-De Paillerets B. PDGFRA germline mutation in a family with multiple cases of gastrointestinal stromal tumor. Gastroenterology. 2004;126:318–321. [PubMed]
  • Christner PJ, Jimenez SA. Animal models of systemic sclerosis: insights into systemic sclerosis pathogenesis and potential therapeutic approaches. Curr Opin Rheumatol. 2004;16:746–752. [PubMed]
  • Corless CL, Fletcher JA, Heinrich MC. Biology of gastrointestinal stromal tumors. J Clin Oncol. 2004;22:3813–3825. [PubMed]
  • Corless CL, Schroeder A, Griffith D, Town A, McGreevey L, Harrell P, Shiraga S, Bainbridge T, Morich J, Heinrich MC. PDGFRA mutations in gastrointestinal stromal tumors: frequency, spectrum and in vitro sensitivity to imatinib. J Clin Oncol. 2005;23:5357–5364. [PubMed]
  • Dai C, Celestino JC, Okada Y, Louis DN, Fuller GN, Holland EC. PDGF autocrine stimulation dedifferentiates cultured astrocytes and induces oligodendrogliomas and oligoastrocytomas from neural progenitors and astrocytes in vivo. Genes Dev. 2001;15:1913–1925. [PMC free article] [PubMed]
  • de Raedt T, Cools J, Debiec-Rychter M, Brems H, Mentens N, Sciot R, Himpens J, de Wever I, Schoffski P, Marynen P, Legius E. Intestinal neurofibromatosis is a subtype of familial GIST and results from a dominant activating mutation in PDGFRA. Gastroenterology. 2006;131:1907–1912. [PubMed]
  • Distler JH, Jungel A, Huber LC, Schulze-Horsel U, Zwerina J, Gay RE, Michel BA, Hauser T, Schett G, Gay S, Distler O. Imatinib mesylate reduces production of extracellular matrix and prevents development of experimental dermal fibrosis. Arthritis Rheum. 2007;56:311–322. [PubMed]
  • Heinrich MC, Corless CL, Duensing A, McGreevey L, Chen CJ, Joseph N, Singer S, Griffith DJ, Haley A, Town A, et al. PDGFRA activating mutations in gastrointestinal stromal tumors. Science. 2003;299:708–710. [PubMed]
  • Hesselager G, Uhrbom L, Westermark B, Nister M. Complementary effects of platelet-derived growth factor autocrine stimulation and p53 or Ink4a-Arf deletion in a mouse glioma model. Cancer Res. 2003;63:4305–4309. [PubMed]
  • Hirota S, Ohashi A, Nishida T, Isozaki K, Kinoshita K, Shinomura Y, Kitamura Y. Gain-of-function mutations of platelet-derived growth factor receptor alpha gene in gastrointestinal stromal tumors. Gastroenterology. 2003;125:660–667. [PubMed]
  • Hoch RV, Soriano P. Roles of PDGF in animal development. Development. 2003;130:4769–4784. [PubMed]
  • Hubbard SR. Juxtamembrane autoinhibition in receptor tyrosine kinases. Nat Rev Mol Cell Biol. 2004;5:464–471. [PubMed]
  • Irusta PM, DiMaio D. A single amino acid substitution in a WW-like domain of diverse members of the PDGF receptor subfamily of tyrosine kinases causes constitutive receptor activation. Embo J. 1998;17:6912–6923. [PMC free article] [PubMed]
  • Klinghoffer RA, Mueting-Nelsen PF, Faerman A, Shani M, Soriano P. The two PDGF receptors maintain conserved signaling in vivo despite divergent embryological functions. Mol Cell. 2001;7:343–354. [PubMed]
  • Krizhanovsky V, Yon M, Dickins RA, Hearn S, Simon J, Miething C, Yee H, Zender L, Lowe SW. Senescence of activated stellate cells limits liver fibrosis. Cell. 2008;134:657–667. [PMC free article] [PubMed]
  • Li J, Hoyle GW. Overexpression of PDGF-A in the lung epithelium of transgenic mice produces a lethal phenotype associated with hyperplasia of mesenchymal cells. Dev Biol. 2001;239:338–349. [PubMed]
  • Mendelsohn C. Going in circles: conserved mechanisms control radial patterning in the urinary and digestive tracts. J Clin Invest. 2006;116:635–637. [PMC free article] [PubMed]
  • Ostman A, Heldin CH. PDGF Receptors as Targets in Tumor Treatment. Adv Cancer Res. 2007;97:247–274. [PubMed]
  • Pasini B, Matyakhina L, Bei T, Muchow M, Boikos S, Ferrando B, Carney JA, Stratakis CA. Multiple gastrointestinal stromal and other tumors caused by platelet-derived growth factor receptor alpha gene mutations: a case associated with a germline V561D defect. J Clin Endocrinol Metab. 2007;92:3728–3732. [PubMed]
  • Ponten A, Folestad EB, Pietras K, Eriksson U. Platelet-derived growth factor D induces cardiac fibrosis and proliferation of vascular smooth muscle cells in heart-specific transgenic mice. Circ Res. 2005;97:1036–1045. [PubMed]
  • Ramalho-Santos M, Melton DA, McMahon AP. Hedgehog signals regulate multiple aspects of gastrointestinal development. Development. 2000;127:2763–2772. [PubMed]
  • Ruas M, Peters G. The p16INK4a/CDKN2A tumor suppressor and its relatives. Biochim Biophys Acta. 1998;1378:F115–177. [PubMed]
  • Rubin BP, Antonescu CR, Scott-Browne JP, Comstock ML, Gu Y, Tanas MR, Ware CB, Woodell J. A knock-in mouse model of gastrointestinal stromal tumor harboring kit K641E. Cancer Res. 2005;65:6631–6639. [PubMed]
  • Sabah M, Cummins R, Leader M, Kay E. Loss of heterozygosity of chromosome 9p and loss of p16INK4A expression are associated with malignant gastrointestinal stromal tumors. Mod Pathol. 2004;17:1364–1371. [PubMed]
  • Samter M. Immunological diseases. Boston: Little, Brown; 1988.
  • Serrano M, Lee H, Chin L, Cordon-Cardo C, Beach D, DePinho RA. Role of the INK4a locus in tumor suppression and cell mortality. Cell. 1996;85:27–37. [PubMed]
  • Sharp R, Recio JA, Jhappan C, Otsuka T, Liu S, Yu Y, Liu W, Anver M, Navid F, Helman LJ, et al. Synergism between INK4a/ARF inactivation and aberrant HGF/SF signaling in rhabdomyosarcomagenesis. Nat Med. 2002;8:1276–1280. [PubMed]
  • Sommer G, Agosti V, Ehlers I, Rossi F, Corbacioglu S, Farkas J, Moore M, Manova K, Antonescu CR, Besmer P. Gastrointestinal stromal tumors in a mouse model by targeted mutation of the Kit receptor tyrosine kinase. Proc Natl Acad Sci U S A. 2003;100:6706–6711. [PMC free article] [PubMed]
  • Sonnylal S, Denton CP, Zheng B, Keene DR, He R, Adams HP, Vanpelt CS, Geng YJ, Deng JM, Behringer RR, de Crombrugghe B. Postnatal induction of transforming growth factor beta signaling in fibroblasts of mice recapitulates clinical, histologic, and biochemical features of scleroderma. Arthritis Rheum. 2007;56:334–344. [PubMed]
  • Soriano P. The PDGF alpha receptor is required for neural crest cell development and for normal patterning of the somites. Development. 1997;124:2691–2700. [PubMed]
  • Soriano P. Generalized lacZ expression with the ROSA26 Cre reporter strain. Nat Genet. 1999;21:70–71. [PubMed]
  • Tallquist MD, Soriano P. Epiblast-restricted Cre expression in MORE mice: a tool to distinguish embryonic vs. extra-embryonic gene function. Genesis. 2000;26:113–115. [PubMed]
  • West RB, Corless CL, Chen X, Rubin BP, Subramanian S, Montgomery K, Zhu S, Ball CA, Nielsen TO, Patel R, et al. The novel marker, DOG1, is expressed ubiquitously in gastrointestinal stromal tumors irrespective of KIT or PDGFRA mutation status. Am J Pathol. 2004;165:107–113. [PMC free article] [PubMed]
  • Yamakage A, Kikuchi K, Smith EA, LeRoy EC, Trojanowska M. Selective upregulation of platelet-derived growth factor alpha receptors by transforming growth factor beta in scleroderma fibroblasts. J Exp Med. 1992;175:1227–1234. [PMC free article] [PubMed]
  • Yu J, Carroll TJ, McMahon AP. Sonic hedgehog regulates proliferation and differentiation of mesenchymal cells in the mouse metanephric kidney. Development. 2002;129:5301–5312. [PubMed]
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...