Nurse-like cells control the activity of chronic lymphocytic leukemia b cells via galectin-1

CROCCI DIEGO; MORANDE PABLO; DERGAN DYLON S ; BORGE MERCEDES; TOSCANO MARTA; STUPIRSKY J; BEZARES FERNANDO; SANCHEZ AVALOS JULIO; NARBAITZ MARINA; GAMBERALE ROMINA; RABINOVICH GABRIEL; GIORDANO MIRTA

LEUKEMIA

NATURE PUBLISHING GROUP

Lugar: Londres; Año: 2013 vol. 27 p. 1413 - 1416

0887-6924

Chronic lymphocytic leukemia (CLL) cells are preferentially activated in so-called proliferation centers frequently found in lymph nodes and bone marrow from CLL patients1. In these ?privileged? sites leukemic cells establish close contact with a variety of cell types that provide long-term support for their survival and progression. In addition, CLL cells favor the establishment of immunosuppressive microenvironments by altering the cytokine milieu2. Galectin 1 (Gal1), an endogenous b-galactoside-binding lectin found at sites of inflammation and tumor growth, displays pro-survival activity on malignant cells as demonstrated for CD45RA(-) primary myeloma cells3. Moreover it controls tumor cell proliferation and invasiveness and plays key roles in tumor-immune escape by dampening T cell-mediated immunity4. In Hodgkin lymphoma Gal1 is over-expressed in Reed-Sternberg cells, is a predictive biomarker of disease progression and is responsible for creating the Th2/regulatory T cell?skewed microenvironment typical of this lymphoproliferative disease5, 6. These unique characteristics of Gal1 prompted us to investigate its potential role in CLL biology.   We first assessed the expression of Gal1 in peripheral blood and bone marrow samples from CLL patients using qPCR analysis, flow cytometry and immunohistochemistry. We found that monocytes in peripheral blood (Fig 1.A and B) and stromal and myeloid cells in bone marrow biopsies (Fig 1.C) are the main sources of Gal1. CLL cells do not express Gal1, but they are able to bind Gal1 in a dose-dependent manner (Fig 1.D). This effect was glycan-specific, as addition of the disaccharide lactose, but not sucrose inhibited binding of Gal1 to CLL cells (Fig 1.D).  In the presence of leukemic cells, monocytes from CLL patients can differentiate in vitro into large, adherent cells that protect the leukemic clone from spontaneous and drug-induced apoptosis. These so-called nurse-like cells (NLC) reside in lymphoid tissues where they presumably deliver pro-survival and stimulating signals to CLL cells7. To determine whether Gal1 secreted by myeloid cells can influence leukemic B cells responses, we knocked down Gal1 synthesis in NLC. For this purpose, we differentiated NLC from peripheral blood CLL samples as previously described8, removed non-adherent cells (> 90% leukemic B cells) and transduced adherent NLC with retrovirus expressing Gal1-specific short hairpin RNA (shRNA-gal1) or a scrambled control shRNA ( shRNA-scr) (Fig 1.E, F). After 6 h of incubation, transduced NLC were thoroughly washed and non-adherent cells were incorporated to the plates for further co-culture. Thereafter, we evaluated in the leukemic clone expression of the activation markers CD80, CD86 and CD25, production of IL-10 as a prototypical anti-inflammatory cytokine and synthesis of CCL3 and CCL4 as key chemokines responsible for the recruitment of monocytes and T lymphocytes to lymphoid tissues. We found that blockade of Gal1 in NLC impaired the expression of activation markers in CLL cells suggesting that the presence of endogenous Gal1 in myeloid cells is required for full stimulation of the leukemic clone (Suppl Fig 1). Blockade of Gal1 also  decreased mRNA and protein levels of IL-10 and mRNA levels of CCL3 in CLL cells, without affecting those of CCL4 (Fig 1.G,H). While previous reports showed that recombinant Gal1 induces the release of IL-10 from activated T lymphocytes and dendritic cells9, there is still no information on its effects on CCL3. Both, IL10 and CCL3, are relevant in CLL pathogenesis. Their serum concentrations are increased in CLL patients compared to healthy subjects11, 12 and, more importantly, they correlate with shorter time-to-first treatment (TTFT) and survival13. Since the B cell receptor (BCR) pathway plays a crucial role in the survival, proliferation and trafficking of CLL cells (cita), we evaluated whether Gal1 was able to modulate BCR signaling in these cells. We found that, in the presence of Gal1, suboptimal concentrations of anti-IgM were able to activate BCR pathway as assessed by ERK and Syk phosphorylation in CLL cells (Fig 1.J), indicating that Gal1 decreases the threshold of BCR activation probably through the formation of lattices as proposed for other galectins (cita).  While these results suggest that Gal1 secreted by NLC may exert a direct effect on the leukemic clone, we also found that knocking down Gal1 diminished the expression of BAFF and APRIL in NLC (Fig 1.I). Although not evaluated in leukemic cells, BAFF is able to enhance CD86 and induce the secretion of IL10 in resting B cells15. Hence, Gal1 secreted by NLC might directly or indirectly influence CLL activity through glycan-dependent binding to these cells or through modulation BAFF and/or APRIL secretion. Finally, we determined the concentration of Gal1 in plasma from 49 CLL samples and 40 age-matched healthy donors. Clinical features of CLL patients are depicted in Supplementary Table 1. We found that plasma concentration of Gal1 was significantly increased (p<0.0001) in CLL patients compared to healthy subjects (Fig. 2.A). When we discriminated CLL patients in high and low risk groups according to the expression of CD38 and ZAP-70 on CLL cells, we observed a trend (although not statistically significant) to increased levels of Gal1 in plasma from patients expressing one or both prognostic markers (Fig.2.B) compared to the double-negative group. Similarly, we observed that patients in Binet A staging had about half the concentration of Gal1 in plasma compared to patients in Binet C (208 vs 517 ng/ml, n=25 vs 7). Finally, we analyzed the expression of Gal1 in bone marrow biopsies from patients with stable and progressive disease. As depicted in Fig. 2.C, E we found both an increased number of cells expressing Gal1 and a higher expression of this lectin in bone marrow samples from patients with progressive disease. In agreement with previous reports, we also observed increased numbers of CD68+ cells in samples from patients with progressive disease (Fig. 2D, E). These data indicate that Gal1 is associated with poor outcome in CLL. Collectively, our findings suggest that Gal-1 secreted by accompanying myeloid cells (i.e. NLC, macrophages and dendritic cells) contributes to stimulate the activity of CLL cells and may help to establish the appropriate microenvironmental conditions for leukemic progression. From a therapeutic standpoint, our study suggests that selective manipulation of Gal-1 expression or signaling in NLC may be able to influence CLL differentiation and survival, a critical effect with implications in the design of novel anti-leukemic therapies.