In Vivo Labeling of Dividing Chronic Lymphocytic Leukemia B Cells Suggests That CXCR4 and CD5 Define the Clonal Subfraction That Recently Proliferatead and Emigrated from a Solid Tissue

Abstract B-cell chronic lymphocytic leukemia (B-CLL) is due, in most of the cases, to the slow progressive invasion of phenotypically homogeneous monoclonal B cells. Several studies are clarifying that B-CLL cells preferentially proliferate in the context of lymphoid solid tissues where they receive accessory signals (chemokines, cytokines, cell to cell contact). Deuterium (2H) incorporation into newly synthesized DNA can be used as a direct in vivo measure of cell division. Using this technique in B-CLL, it has been found that although the fraction of cells dividing each day is small (0.1 – ~2% of the clones), the absolute number of proliferating cells is sizeable (~109 – 1012/day). The study also elucidated a relationship between proliferation and disease activity. By using this same approach, we have tried to refine our understanding of this small percentage of proliferating cells. Thirteen patients drank deuterated water (2H2O) for 6–12 weeks (“labeling period”) and were followed for 24 weeks (“washout period”). Blood was drawn during and at the end of labeling (1st, 2nd time points) and at the end of washout (3rd time point). Based upon proliferation kinetics patients could be divided into 2 groups (Group A and Group B). Four/13 patients (Group A) showed unusual kinetic curves with a strikingly delayed appearance of labeled cells in the blood as compared to Group B patients. This suggested that B-CLL cells of these patients were retained longer at the site of proliferation. Therefore, the two groups were compared for expression of a panel of chemokine receptors (CCR1, CCR2, CCR4, CCR5, CCR7, CXCR1, CXCR3, CXCR4, and CXCR5) by flow cytometry. Only CXCR4 was found at significant higher densities in Group A compared to Group B (P = 0.002). CXCR4 expression was then used to identify B-CLL sub-populations enriched in cells displaying Ki67 and MCM6, two cell cycle related markers. We observed a relationship between CXCR4 and CD5 where CXCR4dim cells tended to be CD5bright and vice-versa with the majority of the cells falling in an intermediate category (CXCR4intCD5int). The CXCR4dimCD5bright subpopulation contained significantly more Ki67+ and MCM6+ cells. Based on this observation, B-CLL cells of 9 patients were sorted into CXCR4brightCD5dim, CXCR4intCD5int, CXCR4dimCD5bright and 2H was measured for each fraction. A hierarchy of 2H enrichment in the fractions was found: CXCR4dimCD5bright > CXCR4intCD5int >CXCR4brightCD5dim. These differences were statistically significant during labelling period (1st and 2nd time point) but completely disappeared at the end of washout (3rd time point). The 2H enrichment measured in the unfractionated clones at the 1st, 2nd and 3rd time point was similar (0.45±0.15 vs. 0.53±0.16 vs. 0.45±0.07, P = ns), suggesting a redistribution of label among the fractions, presumably due to a change of phenotype over time. In vitro studies have suggested that CXCR4 has a crucial role in trafficking and survival of B-CLL cells, and it is emerging as a powerful predictor for bone marrow infiltration and clinical evolution. CXCR4 surface expression is downregulated after encountering its ligand SDF-1. Our data indicate a relationship between CXCR4 levels and B-CLL cells kinetics. Moreover, since CLL cell subpopulations defined by relative CXCR4 and CD5 levels differ in the number of proliferating cells, we suggest that these markers ca be used to define cells that have emigrated recently (CXCR4dimCD5bright) or much earlier (CXCR4brightCD5dim) from the solid tissue in which they divided. The results strongly suggest, in vivo, the important role of the stroma and SDF-1-producing cells for B-CLL cell proliferation providing indications for disease control strategies.