Abstract

The aim of this study was to investigate the role of microenvironmental interactions in the regulation of CD20 expression, since down-modulation of CD20 in the context of immune niches would be a straightforward explanation for the stroma-mediated rituximab resistance described by us and others (Mraz et al., BJH, 2011; Buchner et al., BJH, 2010; Marquez et al., BJH, 2015). Firstly, we co-cultured primary CLL cells with the bone marrow stromal cell line HS-5 for 24/48/72 hrs and analyzed CD20 surface expression by flow cytometry. Suprisingly, CD20 levels were significantly up-regulated (P=0.03) on CLL cells co-cultured with HS-5 in comparison to CLL cells cultured on plastic. The changes in CD20 expression levels are likely to play a direct role in microenvironmental interactions and especially BCR signaling in immune niches (Uchida et al., Int Immunol, 2004). To test this hypothesis we assessed the CD20 expression on CLL cell populations defined according to CXCR4 and CD5 levels. The low-level surface expression of chemokine receptor CXCR4 and high-level expression of activation molecule CD5 can characterize B cells that have recently exited the lymph nodes (Calissano et al., Mol Med, 2011). CXCR4dim CD5bright cells had ~2-fold higher surface CD20 expression when compared to CXCR4bright CD5dim CLL cells circulating in the blood of the same patient for a longer time (P<0.0001, N=21) and CD20 expression clearly decreased with the transition of CLL cells from CXCR4dim to CXCR4bright together with the decrease of CD5 expression. Further, sorted CXCR4dim CD5bright CLL cells had also ~2 times higher CD20 mRNA expression (P=0.002, N=8) suggesting that changes in CD20 expression were due to changes in its mRNA levels rather than a surface modulation. This led us to hypothesize that CXCR4/SDF-1 is directly implicated in CD20 regulation. Indeed, CLL cells treated with SDF-1α (CXCL12), a ligand for CXCR4 which is produced by HS-5, up-regulated surface CD20 (P=0.03). We also observed that CLL cells with higher expression of CD20 had higher levels of surface Ig (P=0.0015, N=12) and this was coupled with higher responsiveness to BCR crosslinking with anti-IgM as measured by calcium flux (P=0.005). Therefore, we investigated the effect of ibrutinib on CD20 levels as ibrutinib is a BCR-signaling pathway inhibitor and prevents CLL cells from responding to microenvironmental stimuli, including chemokine signaling. We observed that treatment of CLL cells with ibrutinib in vitro resulted in CD20 down-modulation (P<0.001). The levels of CD20 in CLL cells were also strongly down-modulated in samples obtained before vs. during ibrutinib treatment of CLL patients in vivo (pre- ibrutinib vs. day 15 and week 5/12, N=8; P=0.02, P=0.01, P=0.07, respectively). These data suggest that a reduction of the target antigen (CD20) is not the cause for rituximab resistance in the context of immune niches, and the sustained activity of ibrutinib and rituximab combinations in clinical trials (Burger et al., Lancet Oncol, 2014) suggests that ibrutinib has other mechanisms that allow for rituximab efficacy. We focused on the regulation of anti-apoptotic molecules, namely Mcl1, since that was shown to protect from rituximab-induced apoptosis and to be rapidly down-modulated by rituximab infusion in vivo (Byrd et al., Blood, 2002; Hussain et al., Clin Cancer Res, 2007). Indeed, we observed higher levels of Mcl1 mRNA in the CXCR4dim CD5bright CLL subpopulation (N=8, P=0.04), and in CLL cells co-cultured with stromal cells. Importantly, Mcl1 was down-regulated after ibrutinib treatment both in vitro and in ibrutinib treated patients.

Conclusion: Microenvironmental interactions up-regulate CD20 expression in CLL B cells through the CXCR4/SDF-1 axis, and the ibrutinib-mediated impairment of microenvironmental interactions down-regulates CD20 expression. This study reveals a novel regulation of CD20 levels in the context of immune niches, which has important implications for CD20-targeting antibodies and the use of BCR-inhibitors in combination.

Supported by: SoMoPro II-no.4SGA8684 (cofinanced by the European Union and the South-Moravian Region); NGS-PTL(306242); EHA Fellowship award; IGA MZ CR NT11218-6/2010; MUNI/A/1180/2014; CZ.1.05/1.1.00/02.0068; G.P. is a city of Ostrava scholarship holder. #G.P. and V.S. contributed equally to this study. contact: marek.mraz@email.cz

Disclosures

Mayer:Janssen: Research Funding. Davids:Genentech: Other: ad board; Pharmacyclics: Consultancy; Janssen: Consultancy.

Author notes

*

Asterisk with author names denotes non-ASH members.