The failure of B cells to induce non-canonical MYD88 splice variants correlates with lymphomagenesis via sustained NF-κB signaling

Gain-of-function mutations of the TLR adaptor and oncoprotein MyD88 drive B cell lymphomagenesis via sustained NF-{kappa}B activation. In myeloid cells, sustained TLR activation and NF-{kappa}B activation lead to the induction of inhibitory MYD88 splice variants that restrain prolonged NF-{kappa}B activation. We therefore sought to investigate whether such a negative feedback loop exists in B cells. Analyzing MYD88 splice variants in normal B cells and different primary B cell malignancies, we observed that MYD88 splice variants in transformed B cells are dominated by the canonical, strongly NF-{kappa}B-activating isoform of MYD88 and contain at least three novel, so far uncharacterized signaling-competent splice isoforms. TLR stimulation in B cells unexpectedly reinforces splicing of NF-{kappa}B-promoting, canonical isoforms rather than the MyD88s, a negative regulatory isoform that is typically induced by TLRs in myeloid cells. This suggests that an essential negative feedback loop restricting TLR signaling in myeloid cells at the level of alternative splicing, is missing in B cells, rendering B cells vulnerable to sustained NF-{kappa}B activation and eventual lymphomagenesis. Our results uncover MYD88 alternative splicing as an unappreciated promoter of B cell lymphomagenesis and provide a rationale why oncogenic MYD88 mutations are exclusively found in B cells.