A Combination of Cytokines Rescues Highly Purified Leukemic CLL B-Cells from Spontaneous Apoptosis In Vitro

B-chronic lymphocytic leukemia (B-CLL), the most common human leukemia, is characterized by predominantly non-dividing malignant mature CD5+ B lymphocytes with an apoptosis defect. Various microenvironmental stimuli confer a growth advantage on these leukemic cells and extend their survival in vivo. Nevertheless, when cultured in vitro, CLL B-cells rapidly die from apoptosis. Certain cytokines may extend the survival capacity of CLL B-cells in vitro and individual anti-apoptotic effects of several cytokines have been reported. The potential cumulative effect of such cytokines has not been studied. We therefore investigated the effects on CLL B-cells survival in vitro of humoral factors, polyclonal lymphocyte activators and a combination of cytokines known for their anti-apoptotic effects. Purified CLL B-cells were cultured in the presence or absence of various soluble molecules and the leukemic cell response was assessed in terms of viability. Apoptotic cell death was detected by flow cytometry using annexinV and 7-amino-actinomycin. The survival of CLL B-cells in vitro was highly variable. When tested separately, cytokines (IL-2, -6, -10, -12, -15, -21, BAFF and APRIL) improved CLL B cell survival moderately; in combination, they significantly enhanced survival of these cells, even up to 7 days of culture. We also report that humoral factors from autologous serum are important for survival of these malignant cells. Our findings support the concept that the CLL microenvironment is critical and suggest that soluble factors may contribute directly to the prolonged survival of CLL B-cells. Therefore, the combination of cytokines we describe as providing strong resistance to apoptosis in vitro might be used to improve the treatment of CLL.


Introduction
B-chronic lymphocytic leukemia (B-CLL), the most common human leukemia, involves the accumulation of predominantly quiescent CD5+ B cells driven by the proliferation of a subpopulation [1]. In addition to this atypical proliferative profile, the B-CLL cells exhibit a strong dependence on cellular and cytokine components of their microenvironment, making their manipulation ex vivo complex and resulting in biased findings [2]. Indeed, B-CLL can affect diverse tissues by the infiltration of malignant cells into blood compartments, bone marrow and lymph nodes. The B-CLL cells isolated from these tissues exhibit heterogeneity in their transcriptional, phenotypic, proliferative and apoptotic profiles dictated by the local microenvironments [3,4,5,6]. Studies of these tissues and the extrapolation from our knowledge of normal B cell biology [7] suggest that the lymph node is the seat of B-CLL pathogenesis. However, for reasons of inaccessibility and ethics, almost all relevant scientific studies have addressed peripheral blood leukemic cells, only partially reflecting B-CLL physiopathology. Research is further impeded by the absence of a cellular model for B-CLL.
The most explicit demonstration of the influence of the microenvironment on B-CLL is the spontaneous apoptosis of B-CLL cells cultured ex vivo. This spontaneous apoptosis can be prevented by the presence of accessory cells and cytokines [2,8]. Several modulators of the survival of CLL cells have been identified and include soluble factors and cell-cell interactions [9,10]. Nurse-like cells [11], bone marrow stromal cells [12,13], T cells [14,15], dendritic cells [16], accessory leukocytes [17], the leukemic cells themselves [18] and the cytokines derived from them: BAFF and APRIL [19], IL-6 [20], IL-10 [21], IL-2 [22], IL-4 [23], IL-21 [24], IL-12 [25] and IL-15 [26,27], have all been suggested to act as pro-survival factors or growth factors in CLL. Nevertheless, it is unclear whether all the cytokines that promote viability in vitro also play a significant role in vivo; it is consequently relevant to study the effect of humoral components on CLL B-cell survival [28,29]. Antigenic stimulation is the initial molecular event triggering the expansion of B lymphocytes. Therefore, in view of the role of the B-cell receptor in normal B cell physiology, it is possible that the antigenic stimulation of B-CLL cells is an oncogenic event leading to the proliferation and survival of a leukemic clone [30]. Indeed, signaling via the B-cell receptor may activate proliferation and survival pathways in CLL cells [31].
These various stimuli are provided by diverse components of the microenvironment where CLL cells accumulate, and are believed to extend the survival capacity of B-CLL cells in vitro. A major goal in CLL therapy is both to induce apoptosis in leukemic cells and to overcome their resistance to apoptosis. The CLL microenvironment makes an important contribution to increasing resistance to apoptosis. In vitro studies evaluating the efficacy of treatment for CLL can help identify possible therapeutic strategies [32]. We  therefore investigated the effects of a combination of cytokines and the effects of humoral factors and antigenic stimulation on apoptosis of B-CLL cells in vitro. Partial re-creation in vitro of the B-CLL microenvironment, including in particular leukemic cell survival signals, is a potentially promising approach for testing existing and new treatments to overcome resistance to apoptosis.

Patients
Leukemic lymphocytes were obtained from the peripheral blood of 15 B-CLL patients (9 women and 6 men, with median age 74 (54-80) years), diagnosed according to the international guidelines (Table S1). All patients provided written informed consent. All studies involving patient samples were approved by the local ethics committee (Comité de Protection des Personnes (CPP) Nord-Ouest, France). All cases involved clonal expansion of small lymphocytes with high nucleus/cytoplasm ratios that co-expressed CD19, CD5, and CD23. None of the patients were being treated at the time of the analysis.

Cell viability
Cell viability was measured by flow cytometry using annexin Vphycoerythrin (PE) and 7-amino-actinomycin (7-AAD) staining. Cells were harvested and resuspended in 100 ml annexin Vbinding buffer containing 5 ml Annexin V-PE and 5 ml 7-AAD staining solutions (BD Biosciences), incubated for 15 minutes at RT, and analyzed in a FACSCantoII or FACSAriaII flow cytometer (BD Biosciences). FlowJo software (Tree Star) was used for data analysis. AnnexinV-7AAD-double-negative cells were counted as viable. The results were confirmed by determining changes in forward light scattering properties associated with dead cells, which were smaller than viable cells.

Statistical analysis
Prism 5 software (GraphPad Software) was used for statistical analyses. The statistical significance of differences between groups was determined using Student's t test or the Wilcoxon test, as appropriate; P values less than 0.05 (0.01) were considered (highly) statistically significant, * P,0.05, ** P,0.01.

Autologous serum (AS) protects CLL B-cells from apoptosis in vitro
To determine the effects of AS on CLL B-cells survival we incubated cells in the presence of 10% of AS. Complete medium and medium supplemented with 10% heterologous serum (HS) were used as controls ( Figure 1). The percentages of viable CLL Bcells after 24 h were highly significantly greater in the presence of AS (63.7%613.4%) than HS (46.1%611.1%) or in complete medium (41.3%613.8%) After 48 h of incubation, the difference between AS (51.7%623.3%) and complete medium (34.2%620.7%) remained significant, and the viability in HS conditions was intermediate (41.4%614.2%; not significant) ( Figure 1). After 72 h of culture, viability in the presence of AS was not significantly higher than in the other two conditions ( Figure 1).

PMA increases survival of CLL B-cells in vitro
We tested for the effects of antigenic stimulation on CLL B-cell survival. Cells were cultured with the following polyclonal stimulators: PMA, PHA, ionomycin and LPS ( Figure 3A). In the presence of 1 mg/ml of PMA, the number of viable CLL B-cells (70.7%620.9%) was significantly higher than that in complete medium (32.2%621%). LPS had no major effect on CLL B-cell survival (35.1%628.4%). Surprisingly, each ionomycin and PHA   significantly induced apoptosis of CLL B-cells: the percentages of viable cells were 17.6%616.6% and 6.6%65.9%, respectively. We analyzed the dose effect relationships for PMA, PHA, and ionomycin. In our experiment, the pro-survival effect of PMA and the pro-apoptotic effect of ionomycin appeared to be independent of dose used ( Figure 3B, 3C). By contrast, PHA significantly increased apoptosis from a dose of 1 mg/ml ( Figure 3D). CLL Bcells were treated with a combination of PMA and IL-4 to test for any cumulative effect on survival ( Figure 3F): the effect of this combination on CLL B-cells survival was no greater than either of the individual effects.

PMA, IL-4 and Cc have lasting pro-survival effects
To assess whether the pro-survival effects of PMA, IL-4 and Cc were sustained, apoptosis was checked after 24 h, 48 h, 72 h and 168 h (Figure 4). After 168 h, the percentages of viable cells were 30.7%625.3% in the presence of PMA, 47.5%629.5% in that of IL-4 and 36%629.7% in that of Cc, all significantly higher than the control value of 12.6%621.8% in complete medium. Changes in viable cell number are represented in Figure 4B. In the complete medium controls, the increases of apoptosis over time caused a substantial reduction in cell number. By contrast, the diminished apoptosis of PMA-, IL-4-and cytokine combinationtreated cells resulted in a significant preservation of the numbers of viable cells ( Figure 4B).

Discussion
Programmed cell death or apoptosis is a common form of cell elimination that can be activated in different cell types in response to a number of physiologically relevant stimuli. In vivo, CLL CD5+ B-cells are arrested in G0 and display enhanced survival, whereas they undergo spontaneous apoptosis in vitro. This suggests the existence of apoptosis-inhibitory factors or survival-support factors in vivo that may be involved in the accumulation of these cells. Such factors may be provided by cells in the CLL microenvironment and by B-CLL cells themselves. It is currently believed that CLL B-cells migrate several times from the peripheral blood to lymph nodes or to bone marrow where they receive signals for their growth and survival. Although the mechanisms are not known, CLL B-cells express a set of chemokine receptors regulating their trafficking between different tissues [33]. Numerous cytokines, including IL-2, IL-4, IL-6, IL-10, IL-12, IL-15, IL-21, BAFF and APRIL individually or in some cases in combination, have been implicated in the regulation of B-CLL cell apoptosis in vitro [8,9,19,20,21,22,24,25,27]. The cellular B-CLL microenvironment is constituted of various types of cells able to produce and secrete all of these cytokines ( Figure 5). However, the effect of the combination of all these cytokines on CLL B-cell apoptosis in vitro has not been studied. Consistent with earlier studies, we observed that IL-4 and BAFF each had significant prosurvival effects. In our experimental conditions, IL-2, IL-6, IL-10, IL-12, IL-15, IL-21 and APRIL failed to enhance cell viability significantly. Nevertheless, we show that a combination of these cytokines significantly increased CLL B-cell survival (Figure 2). for 9 independent experiments. The significance of differences was calculated with the Wilcoxon test: * p,0.05 ** p,0.01. Red asterisks for PMA versus Medium, clear blue asterisks for Cc versus Medium and indigo asterisks for IL-4 versus Medium. doi:10.1371/journal.pone.0060370.g004 The effect of Cc on CLL B-cells was comparable to that of IL-4. This agrees with the findings that soluble factors provided by the microenvironment increase the growth and survival capacity of leukemic cells. Our study clearly indicates that these factors might act in synergy to protect CLL cells from apoptosis. Differential expression of cytokine receptors by CLL B-cells of the same clone may explain the lesser effect on survival of individual cytokines, and, at least, the pro-survival effect of Cc.
By contrast to the well-documented pro-survival effect of phorbol ester on B-CLL cells [34], the effects of other polyclonal activation molecules, and in particular ionomycin, PHA and LPS, are not well documented. Consistent with earlier studies [35], we showed that PMA has a significant pro-survival effect on CLL Bcells, and that this effect did not differ between the doses we tested (Figure 3). This pro-survival effect of PMA is mediated by the activation of the PKC but not MAPK pathway [34]. Unexpectedly, PHA and ionomycin induced apoptosis in CLL B-cells ( Figure 3). No such pro-apoptotic effect of these molecules has been reported before, possibly because these molecules are often used in combination with phorbol ester which may mask this effect (data not shown). However, dose-response experiments showed that ionomycin is pro-apoptotic at all doses tested, and that PHA has no effect on cell survival at concentrations of 0.05 and 0.1 mg/ ml but had a significant pro-apoptotic effect at 1 and 5 mg/ml ( Figure 3). The concentrations of several cytokines including IL-2, IL-4, IL-6, IL-10, IL-12, BAFF and APRIL, is abnormally high in the serum of CLL patients [10,36,37,38]. Coherent with this observation, we show, for the first time, that autologous serum reduces apoptosis of highly purified CLL B-cells (Figure 1). The pro-survival effects of PMA, IL-4 and the cytokine cocktail on CLL B-cells continued for at least 7 days of culture ( Figure 4). These various findings contribute to our understanding of how the B-CLL microenvironment supports the survival of B-CLL cells; in particular, CLL B-cells appear to migrate several times between peripheral blood and lymph nodes/bone marrow where they receive growth and survival signals. It is also now clear that the CLL microenvironment plays a crucial role in resistance of CLL cells to therapy [32,39,40].
In conclusion, our data show that spontaneous apoptosis of B-CLL cells incubated with autologous serum was significantly lower than that of control CLL B-cells (incubated with heterologous serum or complete medium). Moreover, we report that a combination of several cytokines significantly enhances and prolongs the survival of these cells in vitro. Therefore, our findings support the notion that autocrine (B-CLL cell-derived cytokines) and paracrine (IL-2, IL-4, IL-6, IL-10, IL-12, IL-15, IL-21, BAFF and APRIL from T cells, stromal cells, dendritic cells and Nurse-like cells) factors ( Figure 5), and subsequent growth loops established by interactions of such cytokines may make large contributions to the progressive accumulation of B-CLL cells. The combination of cytokines we describe, and that provides substantial resistance to apoptosis in vitro, may be used to improve CLL therapy. Also, the ability to interrupt such cytokine networks in vivo may be exploited to amend existing therapies for B-CLL.