Background/Purpose: While canonical Wnt/beta-catenin signaling has been identified as a
core pathway of fibrosis in SSc, non-canonical Wnt signaling pathways have not
yet been analyzed. A subset of Wnt proteins can activate non-canonical Wnt
pathways that are independent of beta-catenin by binding to different cell
surface receptors to modulate cellular functions such as proliferation and
differentiation.

In this study we
aimed to characterize the role of Wnt5a in fibrotic diseases.

Methods: The
expression of non-canonical Wnt ligands was analyzed by qPCR, IF and Western
blot in patients with SSc, IPF and healthy controls as well as in experimental
fibrosis. The effect of fibroblast-specific (col1a2; CreER) as ubiquitous (ubiquitin;
CreER) knockout of Wnt5a was evaluated in bleomycin-induced fibrosis and in
experimental cGvHD. Wnt5a-induced signaling was analyzed in vitro and in vivo
using small molecule inhibitors, siRNA mediated knockdown, Western blots and
reporter assays.

Results: The mRNA and protein
levels of Wnt5a, but not other non-canonical Wnt ligands, were increased in
skin and lungs of patients with SSc and IPF as compared to healthy volunteers. Wnt5a
was also upregulated in murine models of fibrosis. Co-staining with cell-type
specific markers identified fibroblasts as a major cellular source of Wnt5a in
fibrotic diseases. Wnt5a induced myofibroblast differentiation and collagen
release in vitro and induced pulmonary and dermal fibrosis in vivo.
In contrast, inducible fibroblast-specific knockout of Wnt5a ameliorated
bleomycin-induced fibrosis and experimental cGvHD. The effects of
fibroblast-specific knockout were comparable to those in mice with inducible
ubiquitous knockout. The pro-fibrotic effects of Wnt5a were independent of
Wnt/beta-catenin and Wnt/Calcium signaling, but were mediated by Wnt/PCP
signaling with phosphorylation of JNK and cJun. Consistently, pharmacologic or
genetic inactivation of JNK or cJun abrogated the pro-fibrotic effects of Wnt5a
in vitro and in vivo. The stimulatory effects of Wnt5a on
fibroblasts were biphasic. Western Blots and SBE-reporter assays demonstrated
that, at early timepoints, Wnt5a activated JNK, which cross-phosphorylated
Smad3 to stimulate the expression of TGFbeta/Smad-dependent genes. This early
effect was completely blocked by inactivation of JNK. At later timepoints,
Wnt5a induced the expression of TGFbeta to upregulated pro-fibrotic genes. In
contrast to the early effects, the delayed effects of Wnt5a were blocked by
neutralization of TGFbeta or by TGFbetaRI inhibition.

Conclusion: We characterize Wnt5a as a novel mediator of fibroblast activation in
fibrotic diseases. Wnt5a is upregulated in SSc and IPF and induces fibroblast
activation and tissue fibrosis. The pro-fibrotic effects of Wnt5a are
independent of canonical Wnt signaling and are mediated by JNK-induced
activation of Smad3 at early time points and transcriptional upregulation of
TGFbeta at later time points.