We report that the receptor tyrosine kinase-like orphan receptor 2 (ROR2), a transmembrane protein that participates in Wnt signalling, is frequently repressed by aberrant promoter hypermethylation in human colon cancer. In addition, we demonstrate that the epigenetic-dependent loss of ROR2 can promote tumour growth in colon cancer cells. This effect can be mediated, at least in part, by the canonical Wnt signalling pathway, since restoration of ROR2 activity in DLD1 colon cancer cells induced a decrease in β-catenin/TCF-dependent transcription.
ROR2 is reported to have oncogenic properties in other tumour types such as oral cancer , renal cancer  and osteosarcoma ; it is frequently overexpressed in these tumours, and its suppression in renal cancer cells inhibits cell migration and growth in orthotopic xenograft models . In osteosarcoma cells, suppressed expression of ROR2 (or its extracellular effector, WNT5A) inhibits cell invasiveness and decreases invadopodium formation .
These lines of evidence suggest that the role of ROR2 in cancer is complex and that it can either promote or suppress tumour formation, depending on tumour type and molecular context. A possible explanation for this complexity could be the multifaceted role of ROR2 in the Wnt signalling pathway. On the one hand, ROR2 can mediate WNT5A-dependent activation of JNK, a member of the non-canonical Wnt pathway in mice [9, 21], while on the other, ROR2 can govern the WNT5A-dependent inhibition of canonical Wnt signalling downstream of β-catenin stabilisation . Depending on the tumour type, ROR2 signals can therefore show a preference for β-catenin/TCF-dependent genes or for non-canonical Wnt pathways. Two lines of evidence support these possibilities: first, the pro-tumourigenic role of ROR2 in renal cancer and osteosarcoma is thought to be mediated by activation of the non-canonical Wnt signalling kinase JNK [13, 14], and second, in colon cancer cells with constitutive Wnt signalling activity, restoration of ROR2 activity increased the inhibition of β-catenin reporter genes (our data and ).
As ROR2 is reported to mediate the WNT5A-dependent inhibition of β-catenin-TCF-regulated genes in human embryonic cells [8, 15], we tested whether WNT5A mediated the canonical Wnt-dependent role of aberrant epigenetic ROR2 repression in colon cancer. We evaluated the effect of increased or decreased extracellular WNT5A levels on cell growth and canonical Wnt pathway status in ROR2-expressing DLD1 cells and controls. Although modulation of extracellular WNT5A levels affected cell growth, this effect was little influenced by ROR2 expression. The canonical Wnt pathway did not appear to be the effector of WNT5A modulation in these cells. Our data thus suggest that the canonical Wnt-dependent role of ROR2 hypermethylation in colon cancer cells does not necessarily require WNT5A. As WNT5A might have canonical and non-canonical effects, WNT5A/ROR2 could have a role in colon cancer through non-canonical Wnt signalling within certain molecular contexts. This pathway is frequently altered in other tumour types [12–14]. Our results nonetheless do not rule out a role for WNT5A in colon cancer through the canonical Wnt signalling pathway; indeed, WNT5A inhibits β-catenin/TCF-dependent transcription in HCT116 colon cancer cells .
As proposed for ROR2, WNT5A might also have either a tumour-promoting or-suppressing role. Many studies report a tumour-suppressing effect, and it is downregulated in a number of cancers such as colorectal and ductal breast cancer, neuroblastoma and leukaemia (reviewed in ). As we show here for ROR2, WNT5A repression in colon and haematopoietic tumours is mediated by aberrant promoter hypermethylation [19, 20]. In contrast, WNT5A also has a tumour-promoting role in cases such as non-small-cell lung cancer, melanoma, breast, gastric, pancreatic and prostate cancers (reviewed in ). As ROR2 and WNT5A functions are intimately associated--indeed, ROR2 knockout mice phenocopy most of alterations seen in WNT5A knockout mice --ROR2 and WNT5A might also have parallel or complementary roles in cancer. ROR2/WNT5A upregulation could be advantageous to cancers driven by non-canonical Wnt signalling, while their epigenetic downregulation would benefit tumours, such as colon and haematopoietic cancers, that are driven by canonical Wnt signalling.