Background:Bruton's tyrosine kinase (BTK) is a critical component of the BCR pathway, but BTK and/or BTK-related enzymes such as ITK, Tec, BMX and RLK are also expressed in T, NK and myeloid cells, all of which are important sources of CLL-associated immune dysfunction. Whether BTK and/or BTK-related enzymes within non-malignant immune cells have any effects on CLL progression and immune function is not known. This is highly relevant translationally, as newer generations of BTK inhibitors selectively target BTK while sparing BTK-related enzymes in other cells. As the effects of constitutively active BCR and BTK targeting can be adequately modelled in the TCL-1 and TCL-1xXID mouse models (Woyach et al. Blood 2014), we used these preclinical tools to investigate if BTK expression in non-malignant immune cells modulates CLL progression in vivo. Our aims were: (1) to characterize differences in phenotype and function of T, NK and myeloid cells in leukemia-free and leukemic WT and BTK-inactive mice; (2) to assess BTK-related PD-L1 expression; and (3) to evaluate the in vivo relevance of BTK in non-malignant immune cells to CLL progression.Methods:Single-cell suspension cells were obtained from spleens from CLL-free matched WT B6 and BTK-inactive XID mice (n=11 each) and leukemic TCL-1xXID and TCL-1 mice (n=10 each). Flow cytometry was used to characterize the phenotype of B, T, NK and myeloid cells and myeloid-derived suppressor cells (MDSCs), PD-L1 expression, and T-cell degranulation. For survival experiments, WT B6 (n=29) and XID (n=27) mice were injected with 5x106 splenocytes from leukemic TCL1 donors and monitored daily. For in vivo depletion of monocytes/ macrophages, WT and XID (both n=12) mice received clodronate i.p. every 3 days from day 1 after injection of 3x107 splenocytes from leukemic TCL1 donors, and were euthanized after 36 days of treatment. RT qPCR was conducted on BTK-deficient XLA cell lines.Results:To explore whether BTK expression in non-malignant immune cells affects CLL development in vivo, we first conducted adoptive transfer experiments into WT and BTK-inactive XID mice. While WT mice had a median survival of 52 days after CLL development, survival was almost twice as long (90 days) in XID mice, indicating that BTK inactivation in non-leukemic cells prolongs survival. To identify whether this could be explained by differences in immune cell populations, we next characterized phenotype and function of T, NK and myeloid cells in leukemia-free and leukemic BTK functional and BTK-inactive mice. This comparison revealed that CLL-induced T-cell changes were broadly recapitulated in BTK-inactive mice, while significantly more immature (p=0.0206) and mature (p=0.008) NK cells were maintained. In the myeloid compartment, BTK activity did not affect the overall quantity of CD11b+ myeloid cells, but led to significant qualitative CLL-associated changes: while mice with functional BTK exhibited a significant CLL-driven loss of neutrophils (p=0.0002) and expansion of F4/80+CD11b+ macrophages (p=0.0001), leukemic BTK-inactive mice largely maintained these populations, suggesting anti-tumor properties of these cells. We next depleted myeloid cells with presumed BTK-associated anti-tumor properties in CLL-engrafted mice using clodronate. After 36 days of treatment, differences in spleen weights, CLL load in spleen and blood, and percentage of proliferating CLL cells were lost between WT and XID mice (all p>0.05), indicating that monocytes/ macrophages in XID mice indeed have in vivo anti-tumor functions. Neutrophils were expanded in both WT (median 44.7% of CD11b+ cells) and XID mice (43.7% vs. 6.1% in non-depleted mice). Interestingly, PD-L1 expression as a surrogate marker for tumor-associated immune suppression was generally significantly increased on antigen-presenting cells from BTK-inactive leukemic mice (CLL cells p=0.0002; inflammatory monocytes p=0.0014; patrolling monocytes p=0.001; MDSCs p=0.0201). RT qPCR experiments using XLA cell lines with modified BTK expression revealed that absence of BTK was associated with significantly higher PD-L1 RNA expression (p=0.0075), indicating a direct mechanistic link between BTK and PD-L1.Conclusions: Our data indicate that BTK in myeloid and NK cells directly mediates CLL development and progression in vivo. Targeting BTK in those cells in addition to CLL cells might provide substantial clinical benefits. DisclosuresWoyach:Acerta: Research Funding; Karyopharm: Research Funding; Morphosys: Research Funding.