BCR activated CLL B cells use both CR3 (CD11b/CD18) and CR4 (CD11c/CD18) for adhesion while CR4 has a dominant role in migration towards SDF-1

Chronic lymphocytic leukaemia (CLL) is the most common leukaemia in the western world. In previous studies, various proportion of patients was found to carry CD11b+ or CD11c+ B cells whose presence was an unfavourable prognostic factor. The exact mechanism however, how these receptors contribute to the pathogenesis of CLL has not been revealed so far. Here we analysed the role of CD11b and CD11c on B cells of CLL patients in the adhesion to fibrinogen and in the migration towards stromal cell derived factor-1 (SDF-1) and studied the role of CR4 in the adherence of the CD11c+ B cell line BJAB. We observed that both CR3 and CR4 mediate adhesion of the malignant B cells. Moreover, we found, that CR4 was strongly involved in the migration of the leukemic cells towards the chemoattractant SDF-1. Our data suggest that CR3 and CR4 are not only passive markers on CLL B cells, but they might contribute to the progression of the disease. Since the role of SDF-1 is prominent in the migration of CLL cells into the bone marrow where their survival is supported, our findings help to understand how the presence of CD11c on leukemic B cells can worsen the prognosis of chronic lymphocytic leukaemia.


Introduction
The role of complement receptors CR3 (CD11b/CD18, also known as Mac-1, α M β 2 ) and CR4 (CD11c/CD18, also designated as p150,95; α X β 2 ) are known to be involved in actin linked functions such as phagocytosis, adhesion or migration. These two receptors of the β 2 -integrin family have earlier been suggested to mediate overlapping functions, however, we propose that they rather should be considered as "non-identical twins" [1], because of the functional segregation between them. Namely, studying human monocytes, macrophages and dendritic cells these β 2 -integrins. Blood derived B cells were cultured in the presence of IL-2 and anti-IgG/A/ M F(ab') 2 antibody, to ensure survival of the malignant cells. CD11b and CD11c expression was measured by flow cytometry directly after isolation and after 3-day culture. We found, that although unstimulated CLL B cells already express CD11c, in most of the cases the expression of this β 2 -integrin can be further stimulated by BCR activation, similarly to that found on B cells of healthy donors [11] (Fig 2). While all of the examined patients' B cells expressed CD11c at each time point, the expression of CD11b varied by patients and time. As control, the expression of CD11b and CD11c on B cells isolated from the blood of healthy donors is shown. Further characteristics of receptor expression and function on healthy B cells are demonstrated in our previous paper [11].

Expression of α5β1 and αIIbβ3 on B cells of CLL patients
Beside CR3 and CR4 other integrins are also able to bind fibrinogen, such as α5β1 (CD49e/ CD29), αvβ3 (CD51/CD61), or αIIbβ3 (CD41/CD61) [23]. As we analysed the function of CR3 and CR4 using one of their natural ligands, fibrinogen, we found important to test whether the above listed integrins are also present on the B cells of the CLL patients. We detected CD41a and CD49e on the B cells of Patient 8, while none of these integrins were expressed by the unstimulated B cells of healthy donors (Table 1)  (blue) and CR4 (red) expression was measured by flow cytometry on B cells of eight CLL patients directly after isolation and on the 3 rd day of cell culture. As controls, B cells of healthy donors were investigated (in that case one representative histogram is shown of four independent experiments with similar results). On the histograms, the abscissa refers to the measured fluorescence intensity, while the ordinate refers to cell count. Expression was ranked positive if relative mean fluorescence intensity (RMFI) was higher than 150%.

Both CR3 and CR4 mediate adhesion of CLL B cells
Since CD11c was found to play a key role in the adhesion of activated B cells of healthy donors [11], as well as in the case of the Burkitt-like lymphoma cell line BJAB (Fig 1), we set out to investigate whether CR4 has a similar function on CLL B cells. Since in some cases CLL B cells were found to express CR3 as well, we also analysed the role of this β 2 -integrin in the adhesion to fibrinogen. We found, that blocking the function of either CR3 (blue) or CR4 (red) with specific antibodies significantly decreased the adherence of CLL B cells compared to the control samples (grey) (Fig 3). We used as negative control samples, where cells treated with isotype control antibody were let to adhere to a surface without fibrinogen coat (black). As seen in Fig  3, CR4 contributes to the adhesion of CLL B cells to fibrinogen, similarly to that observed in the case of BJAB cell line (Fig 1) and activated tonsillar B cells [11]. Moreover, we found that CR3, when expressed by CLL B cells, is also able to fulfil this function. We also tested how the simultaneous blocking of CD11b and CD11c affects the adhesion of the cells and found that the presence of anti-CD11b significantly augments the inhibitory effect of anti-CD11c. (S1 Fig). Interestingly, while both CD41a and CD49e were previously shown to mediate adhesion to fibrinogen on platelets [24] and on endothelial cells [25], the effect of CD41a and CD49e specific antibodies was not significant in the adhesion of CLL B cells (S2 Fig).

CR4 dominates over CR3 in the migration of CLL B cells towards SDF-1
Since the migration of leukemic cells plays an important role in the pathogenesis of the disease, we set out to analyse the involvement of CR3 and CR4 of CLL B cells in the migration through fibrinogen-coated and PLL-PEG blocked transwell membrane. We found that while the CD11b blocking antibody had no significant effect on migration (Fig 4A), inhibiting CR4 with a CD11c specific antibody resulted in a significant decrease in the number of migrated CLL-B cells ( Fig 4B). This suggests that the capacity of the two β 2 -integrins is not identical, and CR4 dominates over CR3 in the migration of CLL-B cells. In the assays we used SDF-1 as  chemoattractant, which is a key contributor to CLL pathomechanism, directing the malignant cells to the bone marrow, and providing survival signal to them [26,27]. As negative control, we performed the assay in the absence of SDF-1 and/or using a transwell membrane which was not coated with fibrinogen, only blocked with PLL-PEG.

Discussion
Complement receptors CR3 and CR4 are widely expressed on myeloid cells, where CR3 has a dominant role in phagocytosis [2] while CR4 prevails over CR3 in adhesion to fibrinogen [3]. Though these complement receptors, which belong to the family of β 2 -integrins have earlier been thought to carry out overlapping functions, we suggest to consider them rather as "nonidentical twins" [1], because of the functional segregation between them. Regarding their expression and role on B lymphocytes, we found recently that a proportion of B cells begin to express CD11c after BCR mediated activation, while CD11b could not be detected on B cells of healthy donors. The vast majority of CD11c + human B lymphocytes belong to memory B cell subsets, and they use CR4 to mediate adhesion and migration on fibrinogen covered surfaces [11]. It is also known, that integrins are often overexpressed in malignant B cells, and the anatomical distribution of different B-cell lymphomas can be partially explained by the profile of the adhesion molecules expressed on their surface [28]. Amongst others, β 2 -integrins have been also detected on various B cell lymphomas, including diffuse large B-cell lymphoma, mantle cell lymphoma, Hodgkin's lymphoma, CLL or Burkitt's lymphoma [28]. While CR3 and CR4 are reported to appear also on B cells of patients suffering from Hodgkin's lymphoma [12], hairy cell leukaemia [13] or CLL [4,[13][14][15][16][17][18][19], the role of these complement receptors on malignant B cells is still in question. Since CR4 was found to be involved in the adhesion and migration of activated B cells of healthy donors [11], we assumed that it retains its function on malignant B cells and may even serve as an active driver of the disease.
To answer this question, first we studied BJAB, the EBV-negative Burkitt lymphoma cell line, which expresses CD11c (Fig 1A). We found that CR4 maintains its function in adhesion to fibrinogen, proving its role also on this model of malignant B cells (Fig 1B). While cell lines can serve mainly as model, the results obtained provided important information regarding the function of CD11c and encouraged us to perform further experiment using the B cells of CLL patients.
Out of the mentioned CR3 and CR4 bearing B cell malignancies, CLL is the most common leukaemia in the western world. This disease is characterized by the monoclonal expansion of dysfunctional CD5-, CD19-and CD23-positive B cells in the blood, secondary lymphoid tissues and bone marrow, where they crowd out healthy blood cells [21]. BCR-dependent activation is a key factor in the pathogenesis of CLL, however, BCR-mediated responses are heterogeneous depending on numerous factors and vary from case to case [29]. One of these factors is the IGVH mutational status of the CLL cells, which divides the disease to mutated (M-CLL) and unmutated (U-CLL) cases. The outcome of antigen engagement in U-CLL differs from that in M-CLL [30], with U-CLL tending to have a lower affinity for antigen than affinity-matured M-CLL, which is related with the generally more aggressive behavior of unmutated cases [29]. However, the difference is not absolute, and antigen-responding B cells are apparently present in both U-CLL and M-CLL [30], presumably leading to proliferation and/or survival [29,31].
Regarding CR3 and CR4 expression, varying percentage of patients were found to carry CD11c + B cells, namely 21% [16], 26% [14], 27% [15], 40% [17], 49% [13] or even 89% [19]. The prevalence of CD11b varied between 20% [15] and 66% [16]. The presence of either CR3 or CR4 was found to be an unfavourable prognostic factor in B-CLL as the expression of both β 2 -integrins correlates with the pattern of bone marrow infiltration [14]. However, the exact mechanism of how these receptors contribute to the pathomechanism of CLL has not been revealed so far. Since CLL B cells display a homogeneous memory phenotype [22], we assumed, that CR4 may exert a similar function as on activated memory B cells of healthy donors [11]. Based on our earlier findings we focused on the adhesion and migration of CLL B cells.
In the present study, all the analysed patients' B cells were positive for CR4, while only few patients carried CR3. Since we performed the functional studies using fibrinogen as ligand, and no systematic studies were done so far on the appearance of other fibrinogen-binding integrins (α5β1, αvβ3, and αIIbβ3), we tested their expression on CLL B cells, and as control, on normal B cells. Unstimulated B cells of healthy donors did not express any of these molecules, but after 3 days of BCR stimulation they began to express CD49e, which is in line with the findings of Ballard et al., who demonstrated VLA-5 expression on tonsillar and peripheral blood B lymphocytes stimulated by Staphylococcus aureus Cowan (SAC) [32]. In the case of CLL B cells we detected CD41a and CD49e on the freshly isolated B cells of Patient 8 (Table 1). Moreover, both CD41a and CD49e expression increased further after three days of the BCR stimulus.
Analyzing the contribution of CD41a and CD49e as well as of CR3 and CR4 to the adhesion to fibrinogen of malignant B cells we found, that both CR3 and CR4 were involved (Fig 3), while neither CD41a nor CD49e contributed significantly to this function of CLL B cells (S2 Fig). Importantly, CR3 and CR4 together were responsible for approximately 70% of fibrinogen-dependent adhesion in the case of the studied patients (S1 Fig), suggesting that other receptors also participate in this process. One of the candidates is ICAM-1 (CD54), which is known to interact with fibrinogen and is expressed at high levels in CLL cells [19,33].
The stromal cell-derived factor-1 (SDF-1, also known as CXCL-12) has at least two major effects on CLL B cells; it causes migration towards stromal cells and provides survival signals as well [26]. The proliferating compartment of CLL exists in the bone marrow and lymph nodes, where the stromal microenvironment provides anti-apoptotic and pro-survival signals [27]. Moreover, increasing evidence suggests that the stromal microenvironment contributes to resistance to a wide variety of treatments. It has been observed that although therapies are often effective at killing CLL cells in the blood, residual cells causing disease remain in the bone marrow and lymph nodes [27], where the tumor microenvironment has been shown to promote chemoresistance [34]. For this reason, we analyzed the participation of CR3 and CR4 in the migration of CLL B cells towards SDF-1. We found that CR4 strongly contributes to the SDF-1 dependent migration of CLL B cells (Fig 4B) in contrast to CR3 (Fig 4A). Knowing the essential role of the stromal microenvironment in the pathomechanism of CLL, the involvement of CR4 to the migration towards SDF-1 vindicates the association of CD11c expression with bone marrow infiltration.
In conclusion, we revealed a mechanism, how CR3 and CR4 might contribute to the progression of chronic lymphocytic leukemia. Overcoming adhesion-mediated resistance is important in developing new therapies, as proven in the case of the integrin VLA-4. Natalizumab, the blocking anti-VLA-4 antibody, has been demonstrated to decreases B lymphocyte adherence to stroma and thereby partially control stromal protection toward rituximab and cytotoxic drugs [35]. We suggest that a similar effect can be expected from anti-CR4 and maybe even from anti-CR3 antibodies, which may even complete the partial effect of natalizumab and other treatments.

Patients
Blood samples of patients diagnosed with chronic lymphocytic leukaemia were obtained from the Department of Internal Medicine and Oncology of Semmelweis University with their clinical data summarized in Table 2. None of these patients required therapy at the time of the study, they were monitored using a watch and wait strategy. The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Hungarian Medical Research Council Scientific and Research Committee (ETT TUKEB, permission number: 21655-1/2016/ EKU). As controls, peripheral blood B lymphocytes were isolated from buffy coat obtained from healthy donors provided by the Hungarian National Blood Transfusion Service. Written informed consent was provided for the use of blood samples according to the Helsinki Declaration.

Isolation of B cells
Peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll-Hypaque (GE Healthcare, Chicago, IL, USA) density gradient centrifugation from patients' EDTA-treated venous blood. Patients' B cells were purified by negative selection using the Miltenyi B-CLL Cell Isolation Kit (Miltenyi Biotec, Bergisch Gladbach, Germany) achieving >97% purity, verified by CD19 expression. B cells of healthy donors were isolated from PBMC using the Pan B Cell Isolation Kit (Miltenyi Biotech). In some experiments BJAB, an African EBV-negative Burkittlike lymphoma cell line obtained from American Type Culture Collection (ATCC; U.S.A.) was used.

Adhesion assay
Adherence of CLL B cells was assessed on the 3 rd day of the culture. We also measured the adhesive capacity of CD11c positive BJAB cells. Adhesion assay was carried out as described in our previous paper [11]. Namely, before and during the assay cells were incubated with Fcreceptor blocking reagent (Miltenyi Biotec) to avoid Fc-receptor mediated binding of the integrin-specific antibodies. For blocking the function of integrins, cells were treated for 30 minutes at 4˚C with 10 μg/ml of the following antibodies: anti-CD11b (clone ICRF44, IgG1, Biolegend), anti-CD11c (clone BU15, IgG1, ImmunoTools), anti-CD41a (clone HIP8, IgG1, Invitrogen), and anti-CD49e (clone SAM1, IgG2b, Invitrogen). All of the used integrin specific antibodies were previously shown to block the function of the target molecule [10,11,24,36]. As control, isotype matched control antibodies were used (mouse IgG1, clone MOPC-21, Biolegend and mouse IgG2b, clone MPC-11, Biolegend). To ensure that integrins recycled from the cytoplasm are also blocked, the antibodies were not washed out for the assay. Ninety-sixwell CELLview cell culture dish with glass bottom (Greiner Bio-One, Kremsmünster, Austria) was coated with 10 μg/ml fibrinogen (Merck, Budapest, Hungary) for 1 h at 37˚C. After washing with PBS, free surfaces were blocked with 250 μg/ml synthetic copolymer PLL-PEG (ethylene glycol) (PLL-g-PEG, SuSoS AG, Dübendorf, Switzerland) for 1 h at 37˚C. As negative control, we measured the number of adhered cells to PLL-PEG blocked surfaces in the absence of fibrinogen coat. After blocking the cells with antibodies as mentioned above, they were allowed to adhere to the fibrinogen-coated and/or PLL-PEG blocked surfaces for 1 h at 37˚C and 5% CO 2 in 100 μl of medium. After fixing with 2% paraformaldehyde (Sigma-Aldrich) for 10 minutes unbound cells were washed away with PBS and the adherent cells were stained with Draq5 (Biolegend) and phalloidin-Alexa488 (Molecular Probes, Thermo Fisher Scientific, Waltham, MA, USA) containing 0.1% Triton X-100 (Reanal, Budapest, Hungary). Images were taken by an Olympus IX81 laser scanning confocal microscope using the FluoView 500 software. Eight representative fields were scanned in two wells for each treatment and nuclei were counted using ImageJ software.

Migration assay
The measurement of migration was performed on CLL B cells on the 3 rd day of the culture.
Migration assay was carried out as described in our previous paper [11]. Namely, before and during the assay cells were incubated with Fc-receptor blocking reagent (Miltenyi Biotec) to avoid Fc-receptor mediated binding of the specific antibodies. For blocking the function of integrins, cells were treated for 30 minutes at 4˚C with 10 μg/ml of anti-CD11b (clone ICRF44, Biolegend), anti-CD11c (clone BU15, ImmunoTools), or isotype matched control antibodies (mouse IgG1, clone MOPC-21, Biolegend). To ensure that integrins recycled from the cytoplasm are also blocked, the antibodies were not washed out for the assay. The migration assay has been performed using 24 well Transwell plates (polycarbonate membrane with 5.0 μm pore, Corning, NY, USA) towards 100 ng/ml SDF-1α (Thermo Fisher Scientific). Transwell membranes were coated with 100 μg/ml fibrinogen in PBS overnight at 37˚C and masked with 250 μg/ml PLL-PEG for 1 h at 37˚C. Antibody-treated cells were added to the upper chamber of the transwell assay in 100 μl RPMI-1640 medium containing 10% FCS and 50 μg/ml gentamycin, while the lower chamber of the assay contained 100 ng/ml SDF-1α (Thermo Fisher Scientific) diluted in 600 μl of the same medium. As negative control, we measured the number of migrated cells through PLL-PEG masked transwell membranes in the absence of fibrinogen coat and/or the chemoattractant SDF-1. After the cells were allowed to migrate for 4 h at 37˚C and 5% CO 2 , 25 mM EDTA was added to the lower chamber and the upper chamber was removed. Transmigrated cells were collected from the lower chamber, and cell number was counted immediately using a CytoFLEX cytometer (Beckman Coulter Life Sciences).

Statistics
For the functional analyses, we compared each treatment to the appropriate control sample, presented on the graphs as 100%. Statistical tests were performed with GraphPad Prism 6 software, with p<0.05 considered significant.