The B-cell calcium sensor predicts progression of chronic lymphocytic leukemia

Identifying mechanisms responsible for the clinical heterogeneity of chronic lymphocytic leukemia (CLL) is important to develop better treatments for this disease. Variations in responsiveness to immunoreceptor signaling may be responsible for differences in proliferation of CLL cells in vivo.1 Accordingly, we examined the status of the B-cell calcium sensor (CaoS) in primary CLL cells, as it responds to extracellular calcium (Cao2+) fluctuations by modulating subsequent signal transduction through immunoreceptors.2 In contrast to normal B cells, nearly half (23/51) of the CLL samples examined (with approval from the Sunnybrook Health Sciences Center Research Ethics Board) did not release intracellular calcium (Cai2+) in response to CaCl2 (labeled Cao2+ non-responders) (Table 1; Figure 1a). This impaired CaoS activity was not due to decreased stores of Cai2+ in the endoplasmic reticulum, as the Ca2+ ATPase inhibitor, thapsigargin, was able to mobilize Cai2+ in these cells (not shown). While normal B-cells mobilized Cai2+ in response to as little as 250 M CaCl2, Cao2+ non-responder CLL cells remained insensitive to doses as high as 1.5 mM (above which, calcium was toxic) (not shown). These results suggested that the defective responses to Cao2+ were not due to reduced expression or binding affinity of the CaoS. Tyrosine phosphorylation in response to CaCl2 was also impaired in these cells (Figure 1b), implying that the defect resided at the level of the CaoS.2 As CLL patients could be classed into two groups, depending on the responsiveness of their tumor cells to Cao2+, CaoS status was correlated to other clinical parameters and conventional biological prognostic factors3 (Table 1). Strikingly, intact Cao2+-sensing was associated with increased Rai stage, serum 2-microglobulin levels, CD38 expression and need for treatment, along with significantly shorter lymphocyte doubling times (LDTs) (Table 1), all of which are markers of more aggressive disease.3 Increases in Cao2+ responsiveness over time were observed in the CLL cells of six patients, associated with changes in clinical course such as shorter lymphocyte doubling times (LDTs) and need to institute chemotherapy (not shown). We considered that a functional CaoS might confer an aggressive phenotype by allowing CLL cells to respond more strongly to microenvironmental signals, such as antigens, cytokines and Toll-like receptor agonists.1 Consistent with this hypothesis, changes in CD83 surface expression 4 h after stimulation by phorbol esters2 (which mimic aspects of antigenic signaling4) were much stronger than responses by Cao2+ non-responsive CLL cells (Figure 1c). To further address this hypothesis, following a brief period of culture in Ca2+ (to trigger the Cao2+ sensor) or in Ca2+-free media (to prevent Cao2+ sensor activation), CLL cells were activated with interleukin-2 and the Toll-like receptor-7 agonist, S28690, in Ca2+-containing AIM-V media (to allow proper functioning of downstream signaling pathways that depend on calcium release-activated calcium channels2). In addition, the effect of Cao2+ responsiveness on survival was studied by subjecting cells to prolonged culture in saline in the presence or absence of Ca2+. Only in CLL cells with intact Cao2+ sensing did exposure to Ca2+ increase the expression of tumor necrosis factor- and phosphorylated signal transducers and activators of transcription proteins (used to indicate responsiveness to interleukin-2 and S28690)2 (Figures 1d and e) or survival under harsh culture conditions (Figure 1f). To help explain impaired Cao2+ responsiveness of CLL cells, the stress-activated protein kinase, JNK,5 was studied, as agents that activate this pathway (such as the protein kinase-C agonists, PMA and Bryostatin, ultraviolet light or heat shock) inhibited Cao2+ sensing in normal B cells (Figures 2a and b). In contrast to normal B cells, JNK was often phosphorylated (consistent with activation) in CLL samples, and expression of phosphorylated 46 and 54 kDa JNK isoforms correlated inversely with Cao2+ responsiveness (Figures 2c and d). The specific JNK inhibitor, SP600125,5 restored Cao2+-induced Cai2+ release (Figure 2e) as well as transcription and translation of CD83 (Figures 2f and g), a downstream target of the B-cell CaoS.2 These results support a possible inhibitory effect of activated JNK on the CaoS in indolent CLL cells and an association of intact Cao2+ sensing with the absence of JNK-mediated inhibition in aggressive cells, perhaps reflecting exposure of CLL cells in vivo to immunoreceptor ligands that can activate JNK,1 as well as impaired JNK functioning that is known to accompany cancer progression.6 These results are consistent with recent findings that high-risk CLL cells respond more strongly to activation of immunoreceptors, including the B-cell receptor7 and Toll-like receptors.8 In the absence of molecular identification of the CaoS and direct measurement of Cao2+ perturbations in CLL microenvironments in vivo, it is unclear whether Cao2+ responsiveness in vitro has physiological relevance or is simply a manifestation of the enhanced signaling properties of more aggressive CLL cells. Although larger, prospective studies are required to confirm the observations, particularly regarding the possible role of JNK, our results suggest that the ability of CLL cells to respond to Cao2+ perturbations may be informative with respect to clinical outcome. For example, a 'switch' to Cao2+ responsiveness may herald disease progression and a need to institute chemotherapy. Potentially, CaoS inhibitors might allow aggressive forms of CLL, with poor prognosis, to be converted to more indolent types with a better outcome. This study was supported by the Ontario Institute of Cancer Research (OICR), Canadian Institutes of Health Research (CIHR), National Cancer Institute of Canada (NCIC) with funds from the Terry Fox Foundation, Leukemia and Lymphoma Society of Canada (to DS), and studentships from the Natural Sciences and Engineering Research Council of Canada (NSERC) (to CH) and NCIC (to JT).