Endogenous STAT3-Activated Wnt5a Provides CLL Cells with Survival Advantage

Receptor tyrosine kinase orphan receptor (ROR)-1 is cell surface protein expressed during fetal development. Unlike normal somatic cells chronic lymphocytic leukemia (CLL) express high levels of ROR1. The ROR1 ligand Wnt5a, expressed on bone marrow mesenchymal stromal cells, stimulates ROR1 signaling via β-catenin-dependent and -independent pathways. However whether and how Wnt5a interacts with circulating lymph node derived CLL cells is currently unknown.

In CLL cells the signal transducer and activator of transcription (STAT)-3 is constitutively phosphorylated on serine residues. We have previously reported that phosphoserine STAT3 activates ROR1 expression. Because the Wnt5a gene promoter harbors γ-interferon activation sequence (GAS)-like elements, known as putative STAT3 binding sites, we postulated that phosphoserine STAT3 binds the Wnt5a promoter and activates the Wnt5a gene in CLL cells.

To test our hypothesis we first sought to determine whether CLL cells express Wnt5a. Using Western immunoblotting we detected high levels of Wnt5a protein in circulating CLL cells from 6 randomly selected patients and very low levels in normal B cells. Then, using circulating CLL cells from 6 different patients we immunoprecipitated Wnt5a and detected Wnt5a protein in the immunoprecipitate by Western immunoblotting. To further validate these findings we performed flow cytometry and confocal microscopy. Both assays confirmed that both ROR1 and Wnt5a are expressed on the surface of CLL cells.

Then, to determine whether STAT3 binds the Wnt5a gene promoter, we performed chromatin immunoprecipitation (ChIP). We found that STAT3 co-immunoprecipitated the STAT3-target genes C-Myc, ROR1, VEGF, STAT3 as well as Wnt5a. Sequence analysis identified four putative STAT3-bindings sites spanning within 600 base pairs upstream the Wnt5a start codon. To determine which putative binding site in the Wnt5a promoter binds STAT3 we designed four different probes. Using ChIP we found that STAT3 protein co-immunoprecipitated all four STAT3-putative binding sites at high affinity, and an electromobility shift assay (EMSA) of CLL cells from 5 different patients confirmed these data.

To test whether STAT3 activates the Wnt5a gene we transfected MM1 cells with Wnt5a promoter fragments attached to a luciferase promoter gene, added IL-6 to activate STAT3, and measured the transfected cells' luciferase activity. We found that longer fragments harboring more STAT3 binding sites had a higher luciferase activity, confirming our previous findings. Then, by using STAT3 short hairpin (sh)-RNA, we silenced the expression of STAT3 and found that STAT3-shRNA significantly reduced STAT3 and Wnt5a mRNA and protein levels. Finally, we transfected CLL cells with Wnt5a-siRNA and found that Wnt5a-siRNA significantly increased CLL cells' apoptosis rate. Taken together these data suggest that STAT3 induces the transcription of Wnt5a and that Wnt5a provides CLL cells with survival advantage.

Recent reports showed that miR-15a and miR-16-1, located on the long arm of chromosome 13, inhibit ROR1 expression. Therefore we performed Western immunoblotting using CLL cells from patients in whom trisomy (tri)-12 or deletion (del)-13q were the sole cytogenetic abnormalities. In agreement with previously published data, we found that ROR1 levels were lower in CLL cells with tri12 than in CLL cells with del13q. However, unlike ROR1, Wnt5a levels were significantly higher in CLL cells with tri12 than in CLL cells with del13q, indicating that high levels of endogenous Wnt5a ligand might compensate for low levels of the ROR1 receptor.

In conclusion: STAT3 induces the expression of both ROR1 and its ligand Wnt5a. In that way STAT3 activates a micoenvironment-independent pro-survival pathway in CLL cells.