Chronic lymphocytic leukemia cells acquire regulatory B-cell properties in response to TLR9 and CD40 activation

Abstract

Circulating chronic lymphocytic leukemia (CLL) cells share phenotypic features with certain subsets of regulatory B-cells (Bregs). The latter cells have been reported to negatively regulate immune cell responses, mostly by provision of IL-10. The purpose of the current study was to identify and delineate Breg properties of CLL cells. B-cells and T-cells were obtained from the peripheral blood of untreated CLL patients diagnosed according to the 2008 Guidelines of the International Workshop on Chronic Lymphocytic Leukemia. Co-culture assays were used to examine the ability of CLL cells to suppress autologous T-cell immune responses. IL-10 potency of CLL cells was assessed following stimulation with activators of the toll-like receptor 9 (TLR9) or CD40 and was correlated with the inhibitory activity of the cells. TLR9-activated CLL cells were found to increase the frequency of CD4⁺CD25^hiFOXp3⁺ regulatory T-cells (Tregs) and to inhibit autologous CD4⁺ T-cell proliferation. This signaling cascade proved to control IL-10 generation in CLL cells, which in turn promoted the inhibition of T-cell proliferation by CLL cells. However, CD40 activation of CLL cells, while exhibiting a similar ability to augment Treg frequency, did not either affect IL-10 generation or T-cell proliferation. In conclusion, CLL cells demonstrate a unique clonal quality of adopting Breg properties which promote modulation of T-cell characteristics. TLR9 appears to be a potent activator of regulatory abilities in CLL cells, possibly contributing to preferential immune escape of TLR9-responsive cells.

Introduction

In recent years, important breakthroughs have been made in the understanding of disease-specific mechanisms and development of treatment strategies in chronic lymphocytic leukemia (CLL); however, it is still considered an incurable disease [1]. CLL cells are a malignant B-cell clone presenting with distinctive features of immune suppression, such as increased levels of regulatory T-cells (Tregs) [2], an aberrant cytokine secretion profile with enhanced IFNγ and IL-4 generation [3], and defective immune synapse formation [4, 5]. Immune suppression in CLL is also represented by hypogammaglobulinemia [6], suggested to result from decreased survival and activity of plasma cells [7].

Collectively, this immuno-compromised environment is associated with vulnerability to infections caused by a spectrum of pathogens, contributing to the morbidity and mortality in this disease [6].

CLL cells display a unique phenotype resembling several subtypes of B-cells with regulatory properties (Bregs), mostly reminiscent of B10-pro cells, based on their expression of CD5, CD24 and CD27 [6, 8–12] and their ability to secrete IL-10 upon exposure to immune stimulation. Likewise, CLL cells share features with immuno-suppressive B1 cells that have also been suggested to be their normal counterpart [13, 14].

Both B1 and B10 populations are enriched with Bregs, that are reported to be involved in immune modulation [15], constraining T-cell activity [16], stimulating inhibitory NK cells, attenuating monocyte and DC generation of inflammatory cytokines [17], and suppressing macrophage function [18]. Breg inhibitory activities, specifically those of B1 and B10 cells, are mediated by IL-10 secretion [16–18].

The role of Bregs in attenuating cancer-specific effector cell responses has been demonstrated in animal models of solid tumors. The suggested mechanisms included IL-10-mediated reduction in T-cell and NK-cell reactivity [19] as well as conversion of CD4⁺ T-cells to Tregs [20].

Data in hematologic malignancies are less abundant; but a study conducted on a model of Burkitt-like lymphoma suggested that Bregs might be involved in protecting the malignant cells from clearance [21].

Bregs require various non-antigen- as well as antigen-specific signals for maturation and activation, including Toll-like receptors (TLRs), CD40 ligand (CD40L), and the B-cell receptor (BCR) [9, 22, 23]. Dual CD40 and TLR9 stimulation has been shown to induce the highest levels of IL-10 competent Bregs [9].

TLR activation has also been suggested to be involved in the pathogenesis of B-cell malignancies through apoptosis blockage and immune modulation [24]. Notably, TLR dysregulation and polymorphisms have been shown to increase the risk of developing B-cell lymphoproliferative disorders in a murine model, which could be attributed to elevated Breg levels [25, 26]. A limited number of studies have addressed the issue of Breg activity of CLL cells, demonstrating their ability to induce FOXp3 in CD4⁺CD25⁻ T-cells [27], and inhibit TNFα secretion from macrophages [8]. CLL cells are an example of a malignant B-cell clone with consistent expression of TLR9 [28]. These cells also demonstrate a unique response pattern to TLR9 agonist stimulation, as reflected by marked upregulation of activation molecules and time-dependent increased apoptosis upon an initial period of proliferation [28, 29]. We hypothesized that CLL B-cells could manipulate the immune system, without involvement of non-malignant B-cells as a “third party”, leading to the obstruction of responses against both invading pathogens and the cancerous cells themselves. The current study aimed to explore in vitro the ability of CLL cells to behave as Bregs in modulating T-cell responses and assess the contribution of stimulators, such as TLR9 and CD40, to the establishment of this activity.

Methods

Study population

This study was conducted between 2010 and 2014 upon approval by the Institutional Review Board of the Rambam Health Care Campus (approval no. 0015-IRB). Forty four patients with CLL, defined according to the National Cancer Institute Working Group criteria [30], were recruited to the study upon obtaining their informed consent. Untreated patients with all Rai/Binet stages of disease were eligible. Blood samples were derived from 35 patients at diagnosis and 9 in relapse. Patient demographics and clinical characteristics are shown in Table 1. The immunoglobulin variable region heavy chain (IgVH) mutational status was assessed in CLL cells of 9 patients only; 6 were found to be unmutated. Fluorescence in situ hybridization (FISH) analysis was available for 70% of the patient cohort and was abnormal in 48% of the patients tested. Blood samples of eight healthy donors were used as controls.

Table 1.

Patient characteristics

Parameter	CLL patients n = 44 (%)
Median age (range)	62 (35–80)
Gender (male/female), %	57/43
White blood cell × 109/L, (range)	23.84 (6-426)
Clinical staging, % of patients
A	17 (38)
B	15 (34)
C	12 (28)
CD38% (cutoff 20%), no. of patients (%)
Positive	24 (54)
Negative	20 (46)
LDH, no. of patients (%)
Normal (< 225)	33 (75)
High (> 225)	11 (25)
Β2 microglobulin, % of patients
Normal	5 (11)
High	29 (66)
Not available	10 (23)
Cytogenetics (FISH)
Del17p	1 (2)
Del11q	4 (9)
Del 13q	7 (16)
Trisomy12	3 (7)
No abnormality	16 (36)
Not available	13 (30)

Study design

B-cells obtained from peripheral blood were assessed for their responsiveness to CD40L and TLR9 agonist, as well as to an inhibitor of the up-stream intracellular TLR-associated kinase IRAK4 (interleukin-1 receptor-associated kinase 4). Changes in the CLL cells and their effects on co-cultured autologous T-cells were evaluated. Autologous T-cell proliferation in response to non-antigen dependent stimulation, and Treg induction were used as models for the assessment of Breg-induced T-cell inhibition.

Purification of cell populations

PBMCs were prepared from whole blood by density centrifugation over Ficoll-Hypaque gradients (Sigma-Aldrich, St. Louis, MO). CD4⁺ cells were positively selected from PBMCs with anti-human CD4 microbeads (Miltenyi Biotech, Bergisch Gladbach, Germany). Subsequently, B-cells were negatively selected using The EasySep™ Human B Cell Enrichment Kit (STEMCELL Technologies, BC, Canada). Purity of both subsets was > 90%.

B-cell activating molecules

B-cells were activated with either TLR9 agonist CpG-ODN 2006 (CpG-oligodeoxynucleotide; Enzo Life Sciences, NY, USA) at 1 µM or soluble CD40L (R&D systems, MN, USA) at 5 µg/ml.

Inhibitor molecules

To test the dependence of Breg properties on TLR9 signaling, an IRAK1/4 inhibitor (Sigma-Aldrich), preventing up-stream TLR9-associated signaling, was added daily to B-cell cultures at concentrations of 2 and 5 µM.

FACS antibodies

The expression of cell surface antigens and intracellular proteins was analyzed by FACS. Cell acquisition and analysis were performed using FlowJo software (Treestar, OR, USA) on a FACS Calibur instrument (BD Bioscience, CA, USA).

The antibodies were conjugated to either phycoerythrin (PE), Allophycocyanin (APC), or Brilliant Violet 421 (BV). Cells were stained with the following antibodies: CD4-APC (clone RTA-P4), CD19-BV (clone HIB19), (Biolegend Ltd. CA, USA), CD25-BV, intracellular FOXp3—PE (clone PCH101), (eBioscince CA, USA) and IL-10—PE (clone JES3-19F1) (Biolegend, CA, USA). Isotype controls were IgG1 (clone X40-FITC, clone X40-PE) (BD Biosciences, CA, USA); clone MOPC-21-APC (Biolegend, CA, USA).

Intracellular staining

IL-10 expression was assessed in B-cells, cultured for 48 h in 5% fetal calf serum, 1% glutamine, and 1% penicillin/streptomycin solution, at 37 °C and 5% CO₂. Cells were placed in 24-well plates (1 × 10⁶/well) with activating or inhibitory molecules at the above-mentioned concentrations. Prior to harvesting, 40 ng/ml of Phorbol myristate acetate (PMA) and 1 µg/ml of ionomycin (Sigma-Aldrich, St. Louis, MO, USA) were added for 5 h, with 2 µM/ml of GolgiStop™ added for the final 0.5 h. Cells were then co-stained for surface antigens and subsequently permeabilized with a saponin-based reagent (cytofix-cytoperm fixation permeabilization kit, BD Biosciences), followed by intracellular staining. FOXp3 expression was evaluated after permeabilizing cells with fixation/permeabilization buffer (eBioscience, CA, USA).

Cytokine array

CLL cells were analyzed for a panel of cytokines with the RAYBIO cytokine array G3 series (RayBiotech, GA, USA) according to the manufacturer’s instructions following 48 h stimulation with 1 µM of ODN or 5 µg/ml CD40L. Signals were visualized using a fluorescence laser scanner. Inductions and reductions in signal intensity are defined as biologically significant if fold changes are > 1.5 or < 0.65. The chosen array detects: ENA-78, G-CSF, GM-CSF, GRO, GRO α, I-309, IL-1α, IL-1β, IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-8, IL-10, IL-12 p40/p70, IL-13, IL-15, IFN-Ƴ, MCP-1, MCP-2, MCP-3, M-CSF, MDC, MIG, MIP-1δ, RANTES, SCF, SDF-1, TARC, TGF-β1, TNF-α, TNF-β, EGF, IGF-1, Angiogenin, Oncostatin M, Thrombopoietin, VEGF-A, PDGF-β, and Leptin.

Small interfering RNA (siRNA) cell transfection

5 × 10⁶ CLL cells were suspended in 100 µl of nucleofector solution (Human B cell Nucleofector™ kit, Ge Healthcare, Dharmacon RNAi & Gene Expression, CO, USA) and 100 nM of IL-10 siRNA (Dharmacon siGENOME SMART pool, Ge Healthcare, Dharmacon RNAi & Gene Expression, CO, USA) or control non-targeting RNA were added. Transfection efficacy for the CLL cells was assessed with a GFP (green fluorescent protein) plasmid. Cells were placed in a sterile electroporation cuvette and subjected to high voltage, in an Amaxa nucleofector (Amaxa Biosystems, Cologne, Germany) at an optimized setting. After electroporation, in order not to subject the cells to further stress, they were immediately plated out using pre-warmed growth media supplemented with 10% FCS, for further investigation.

RNA extraction and real-time polymerase chain reaction (qPCR)

Electroporation success was validated through IL-10 mRNA expression. RNA was isolated from transfected CLL cells with the total RNA Aurum RNA kit (Biorad, Hercules, CA, USA) according to the manufacturer’s instructions, followed by single-strand cDNA synthesis with the RevertAid First Strand cDNA Synthesis kit (Thermofisher scientific, MA, USA). QPCR was performed using SYBR SensiMix (Bioline, London, UK). The IL-10 gene was amplified with the following primers: forward: GACTTTAAGGGTTACCTGGGTTG; reverse: TCACATGCGCCTTGATGTCTG, and normalized against GAPDH.

T-cell proliferation assay

CD4⁺ T-cells were stained with carboxyfluorescein diacetate succinimidyl ester (CFSE) using the standard protocol (Molecular probes, Invitrogen, OR, USA). After washing, labeled cells were stimulated with anti-CD3 (10 µl/ml), anti-CD28 (5 µl/ml) antibodies (Biolegend), and IL-2 (0.1 ng/ml, Peprotech Asia).

Immediately following stimulation, CFSE stained CD4⁺ T-cells (1 × 10⁵/well) were co-cultured in 96-well plates with freshly isolated autologous CLL cells at a ratio of 1:1 in 5% fetal calf serum, 1% glutamine, and 1% penicillin/streptomycin solution, at 37 °C and 5% CO₂, with or without B-cell activating molecules CpG-ODN or CD40L. Each condition was assessed in a quadruplicate of wells; the final volume in each well was 200 μl. After 5 days of incubation, cells were collected and proliferation cycles were analyzed by FACS. Inhibited cells, retained in proliferation cycle 0, were referred to as the “non-proliferating pool”, and cells reaching advanced proliferation cycles (≥ 3) were defined as the “proliferating pool”.

In cell contact dependence experiments, T-cells were placed in 24-well plates (5 × 10⁵/well), and B-cells were loaded at a 1:1 ratio at the upper chamber of a 0.4 µM pore insert (Millicell^® cell culture plate inserts, EMD Millipore, Darmstadt, Germany) allowing only flow of small soluble molecules between the chambers. The final volume in each well was 1 ml.

Treg assay

CD4⁺ T-cells (5 × 10⁵/well), supplemented with 0.01 ng/ml IL-2, were co-cultured in 24-well plates with freshly isolated autologous CD19⁺ CLL B-cells at a ratio of 1:1, in 5% fetal calf serum, 1% glutamine, and 1% penicillin/streptomycin solution at 37 °C and 5% CO₂ for 5 days in the presence or absence of 1 µM ODN or 5µ g/ml CD40L. The final volume in each well was 1 ml. Induction of Tregs was defined as a > 10% change in the percentage of CD4⁺CD25^hiFOXp3^+/hi.

Statistical analysis

Data from two tested groups were compared using Student’s t test. P values < 0.05 were considered significant.

Results

TLR9-activated CLL cells inhibit autologous T-cell proliferation

CLL cells were activated with either the TLR9 agonist ODN or CD40L and co-cultured with autologous T-cells stimulated with anti-CD3, CD28 antibodies and IL-2. The ability of activated CLL cells to inhibit T-cell proliferation was examined. A significantly lower number of T-cells reached advanced proliferation cycles when co-cultured with ODN-activated CLL cells compared to untreated cells (30.3 vs. 46.3%, p < 0.01). This phenomenon was not observed following CD40L activation (43 vs. 46.3%, NS) (Fig. 1a, b). Notably, ODN-activated CLL cells were found to inhibit T-cell proliferation in 81% (N = 27) of the tested patient samples.

Fig. 1.

Only ODN-stimulated CLL B-cells inhibit T-cell proliferation independently of cell contact. CD4⁺ T-cells pre-stained with CFSE were stimulated with anti-CD3 (10µ g/ml), anti-CD28 (5µ g/ml), and IL-2 (0. 1 ng/ml). T-cells (1 × 10⁵/well) were immediately co-cultured in 96-well plates with autologous B-cells (1 × 10⁵/well), either untreated or activated with 1 µM of ODN or 5 µg/ml CD40L. After 5 days, T-cell proliferation was analyzed by FACS. Data demonstrate the mean percentage of T-cells reaching advanced proliferation cycles (≥ 3 cycles). a Representative histograms and pie charts of respective amount and percentage of T-cells in each proliferation generation with either unstimulated or ODN-activated CLL B-cells. b T-cell proliferation under various B-cell stimulatory conditions: patient samples. c T-cell proliferation under various B-cell stimulatory conditions: healthy donor samples d To assess T-cell proliferation under non-contact conditions, T-cells (5 × 10⁵/well in 24-well plates) were seeded in the bottom chamber of a 0.4 µm transwell insert and CLL B-cells (5 × 10⁵/well) were loaded to the upper chamber without direct contact. gen generation

In healthy donor blood samples, no significant difference in the number of T-cells reaching advanced proliferation cycles was found between the various B-cell conditions [43.6% (n = 8), 44.6% (n = 8), and 52.2% (n = 5) for naturally occurring B-cells, ODN-activated, and CD40L-activated cells, respectively; NS] (Fig. 1c).

Inhibition of T-cell proliferation is not contact-dependent

Since it is unknown whether soluble or surface molecules are involved in T-cell inhibition by CLL cells, the impact of cellular contact was evaluated. No significant differences in the percentage of T-cells reaching advanced proliferation cycles were observed between contact and non-contact conditions both for untreated and ODN-activated CLL cells (57.6 vs. 54.2% and 41.1 and 42.2%, respectively). Namely, the level of proliferating T-cells decreased significantly under both contact and non-contact conditions following ODN activation, with 0.71 (± 0.14) and 0.73 (± 0.19)-fold reductions in T-cell proliferation, respectively (N = 7, p < 0.01) (Fig. 1d).

CLL cells induce the production of autologous Tregs

The potential of CLL B-cells to influence Treg frequency of autologous CD4⁺ T-cells was examined. Following incubation of CLL B-cells with purified CD4⁺ T-cells, the percentage of CD25^hiFOXp3^+/hi Tregs was assessed (Fig. 2a). Untreated CLL cells demonstrated the ability to increase the frequency of Tregs (1.1 ± 0.8 fold increase, p < 0.05); however, activation by ODN or CD40L was more potent, resulting in 1.59 ± 0.6-fold and 1.5 ± 0.8-fold increases, respectively (p < 0.01) (Fig. 2b). Elevated Treg frequencies were observed in 68% of ODN-stimulated and 80% of the CD40L-stimulated co-cultures (N = 15).

Fig. 2.

ODN and CD40L activated B-cells increase Treg frequency among autologous CD4⁺ T-cells. Autologous CD4⁺ T-cells (5 × 10⁵/well), supplemented with 0.01 ng/ml IL-2, were co-cultured in 24-well plates for 5 days with CD19⁺ enriched CLL B-cells (5 × 10⁵/well) in the presence or absence of 1 µM ODN or 5 µg/ml CD40L. Following co-culture the cells were harvested and stained for CD4/CD25/FOXp3. a Representative dot plot of gated CD4 cells. The population of CD25^hiFOXp3^hi Tregs is encircled. b Data represent Treg (CD25⁺FOXp3^+/hi) percentage (± SD) among the CD4⁺ T-cell population (**p < 0.01, *p < 0.05)

IL-10 expression and secretion in CLL cells are enhanced by ODN

Given the role of IL-10 in mediating Breg activity, intracellular expression of this cytokine by CLL cells following stimulation was assessed by FACS. Following ODN activation, a significant increase in IL-10 expression was observed (3.3 ± 2 fold; N = 16, p < 0.01), whereas CD40L had no essential impact on intracellular IL-10 levels (N = 10) (Fig. 3a).

Fig. 3.

ODN enhances IL-10 generation in CLL cells. Isolated CD19⁺ CLL B-cells (1 × 10⁶/well) were cultured for 48 h with or without either 1 µM of ODN or 5 µg/ml CD40L. a Cells were harvested and stained for intracellular IL-10 and examined by FACS. The data represent an average fold change in the expression of IL-10 following activation. b Culture media were processed according to the RayBio cytokine array protocol. The presented fluorescent values depict the concentration of IL-10 released to the medium per patient sample (N = 3)

Notably, IL-10 was the only cytokine whose secretion repeatedly increased following ODN activation (> 1.65-fold increase; N = 3) in a cytokine array, while none of the 41 additional examined cytokines displayed consistent baseline secretion to the medium or reproducible responses to ODN. Similar to FACS analysis, the cytokine array failed to reveal alterations in IL-10 or other cytokine levels following CD40L stimulation (Fig. 3b).

IL-10 generation is regulated by TLR9 signaling in CLL cells

To explore whether ODN-induced IL-10 generation and secretion in CLL cells was directly controlled by the TLR9 signaling cascade, the vital up-stream intracellular TLR-associated kinase IRAK1/4 was inhibited. The percentage of CLL cells expressing IL-10 decreased in a dose-dependent manner from 5.88 (± 2.01) with ODN activation to 5.18 (± 2.01) following low-dose (2 µM) inhibition (N = 5, p < 0.05) and to 4.83 (± 1.93) following high-dose (5 µM) inhibition of IRAK1/4 signaling (N = 5, p < 0.01) (Fig. 4).

Fig. 4.

IRAK1/4 inhibition leads to a dose-dependent decrease in IL-10 generation. CD19⁺ isolated CLL cells (1 × 10⁶/well) were cultured with ODN (1 µM) for 48 h in the presence or absence of an IRAK1/4 inhibitor at 2 µM and 5µM and analyzed by FACS for intracellular IL-10. Data represent the percent of CLL B-cells expressing IL-10 (**p < 0.01)

IL-10 generated by CLL cells contributes to the inhibition of T-cell proliferation

Since ODN-activated CLL cells were shown to inhibit T-cell proliferation in a non-contact-dependent manner, and IL-10 emerged as a unique CLL inhibitory cytokine, its significance as a mediator of T-cell inhibition was assessed by knocking down IL-10 expression. SiRNA transfection efficacy, assessed with a GFP plasmid, was found to be 28 (± 6.1)% (data not shown). Transfection of CLL cells with IL-10 siRNA resulted in a 25% reduction in IL-10 mRNA expression in the cells (N = 4, p < 0.05) (Fig. 5a). IL-10 siRNA transfection diminished the inhibitory effect of CLL cells on proliferation of co-cultured autologous T-cells, demonstrating a 21% reduction in the number of cells retained in proliferation cycle 0. In the control co-culture, 12.4% (± 2.3) of stimulated CD4⁺ T-cells remained in G0, whereas in the presence of IL-10-depleted CLL cells, only 9.8% (± 1.87) of CD4⁺ T-cells remained in the non-proliferating pool (N = 8, p < 0.05) (Fig. 5b). These data were previously published in part in our abstract presented at the 58th ASH Annual Meeting, December 3–6, 2016, San Diego CA, USA [31].

Fig. 5.

T-cell proliferation is mediated by IL-10 derived from CLL cells. CLL cells were pulsed with 100 nM IL-10 siRNA or non-targeting control RNA (control siRNA) prior to stimulation with 1 µM ODN. a Following 48 h of stimulation, IL-10 mRNA expression was assessed. b Alternatively, pulsed CLL cells (1 × 10⁵/well) were immediately co-cultured with CFSE stained autologous CD4⁺ T-cells (1 × 10⁵/well) for 5 days in the presence or absence of 1 µM of ODN. Cells were collected and proliferation cycles were analyzed by FACS. Data represent mean percentages of non-proliferating (proliferation cycle 0) T-cells co-cultured with the control and the IL-10 siRNA transfected CLL cells (*p < 0.05)

Discussion

Cancer cells engage multiple mechanisms to ensure persistence. A malignant B-cell clone that adopts Breg characteristics could potentially tailor immune cell responses by inhibiting effector cells and thus gain a survival advantage. While CLL cells bear phenotypic resemblance to various Breg subsets, it remains unknown whether they employ Breg-specific mechanisms to diminish anti-cancer activity, hence enabling disease propagation, as well as to diminish antimicrobial immune activity and to promote infection [6].

The current study evaluated a potential ability of CLL cells to function as Bregs, focusing on CLL regulatory effects on autologous T-helper cells, and the contribution of Breg-activating molecules to the acquisition of these properties. Of note, despite the fact that the examined autologous T-cells had been pre-conditioned in vivo in an immuno-suppressive environment, the use of these cells was likely to diminish alloreactivity, enabling the assessment of cellular responses in a “controlled” environment.

CLL cells co-cultured with autologous T-cells were found to inhibit T-cell proliferation and induce the generation of CD4⁺CD25⁺FOXp3^hi Tregs, features attributed to Bregs. The ability of a non-malignant Breg subset (e.g., CD25⁺CD27⁺CD86⁺CD1d⁺IL-10⁺TGFβ⁺) to inhibit T-cell proliferation and induce FOXp3 expression in Tregs has been previously reported [11]. In addition, there is limited evidence, suggesting that CLL cells can influence the T-cell phenotype by converting allogeneic CD4⁺CD25⁻FOXp3⁻ to CD4⁺CD25⁻FOXp3⁺, which represents a non-classic Treg subtype [27]. To the best of our knowledge, the present study has been the first to demonstrate that CLL cells induce a classic Treg phenotype from autologous CD4⁺T-cells and inhibit T-cell proliferation, confirming the regulatory behavior of CLL cells.

The results of the present study have demonstrated no effect of healthy donor circulating B-cells on T-cell proliferation inhibition, as they are likely to represent a mixture of B-cell subsets, with different inhibitory and stimulatory capacities. The comprehensive inhibitory nature exhibited by CLL cells in this study may represent either superior inhibitory abilities of the malignant cells over their “normal” counterparts [8, 32], or an effect associated with a relatively greater “load” of inhibitory cells in CLL samples.

Our findings showed that while CLL cells possessed regulatory abilities in an unstimulated state, the environmental signals activating the TLR9 and CD40 pathways enhanced these properties in a stimulant-specific manner. Namely, TLR9 activation of CLL cells resulted both in a significant increase in the Treg percentage and a pronounced inhibition of T-cell proliferation. At the same time, CD40 activation led to a less intense Treg induction, while not affecting T-cell proliferation, suggesting that TLR9 activation was a more potent Breg stimulus for CLL cells than CD40L. CLL cells were shown to express higher levels of TLR9 than their normal CD5⁺CD19⁺ counterparts [29], possibly explaining their marked responsiveness to TLR9 activation, as compared to other malignant and non-malignant B-cells [28]. Indeed, CD40/CD40L signaling was reported to be weak in CLL [33]. Nevertheless, the fact that, in our study, CLL cells could enhance their regulatory behavior in response to both TLR9 and CD40 signals confirmed the resourcefulness of these cells in maintaining regulatory abilities.

Although, in the current study, neither CD40L nor TLR9 activation of healthy donor B-cells was found to affect autologous T-cell proliferation, normal B-cells had been previously reported to respond to these activators, leading to a variety of inhibitory effects, including T-cell proliferation inhibition [11, 34, 35]. These differences could be related to slight variance in the experimental design of the studies. Moreover, given that CLL B-cells have undergone malignant transformation and may thus utilize signals and pathways differently than normal B-cells, it is possible that full-range CD40 or TLR9 activation in CLL and normal B-cells requires different ligand concentrations. Another possible explanation could be related to the lower number of healthy donor cell co-cultures, used in our study. Since the goal of the current study was to define inherent properties of CLL cells, we decided to focus on malignant B-cell features and to a lesser extent on the comparison between the CLL and normal cell settings.

The current study demonstrated that direct contact between the TLR9-activated CLL cells and the responding T-cells was not required to inhibit T-cell proliferation, indicating that a soluble molecule was responsible for cellular communication in this case. Indeed, the evidence obtained supported the role of IL-10 in executing TLR9-activated regulatory functions, since IL-10 was the main cytokine consistently secreted by CLL cells. Moreover, our results show that IL-10 generation in CLL cells is directly controlled by the TLR9 pathway, since attenuation of signaling through the protein kinase IRAK 1/4, recruited directly by the adaptor protein MYD88, led to decreased IL-10 levels. Our data are in line with publications reporting the generation of IL-10 by non-malignant Bregs [9, 10, 22, 23], and by CLL cells following the combined stimulation by CD40L and a TLR9 agonist [8]. IL-10 secretion has been reported to be more prominent in the IgVH mutated CLL cells, due to lower methylation of variably methylated regions in the IL-10 gene [8, 32]. The present study failed to confirm the role of CD40L in stimulating the production of IL-10 or other cytokines of inhibitory potential in CLL, suggesting that surface and not soluble molecules play an essential part in mediating CD40L-induced Treg generation. A plausible explanation could be related to the fact that, unlike the current study, other studies did not explore the activation of CLL cells by CD40 as a single stimulator, but rather combined it with TLR or BCR agonists [8, 36, 37]. Hence, the role of CD40 as a sole stimulus in promoting Breg activity in CLL cells requires further assessment. Surface immune check-point modulators could be potentially involved in this process, as PD1 has been reported to be upregulated on CLL B-cells following stimulation with CD40L and IL-4 [38].

Furthermore, findings of the present study demonstrate that IL-10- knockdown CLL cells are less effective T-cell inhibitors than IL-10-potent cells. While transfection of CLL cells is known to be limited, the fact that inhibition of T-cell proliferation was still compromised despite the presence of a residual amount of IL-10 confirms the major contribution of this cytokine to the inhibitory activity of CLL cells. Conversely, in normal B-cells, IL-10 was not found to be a key factor in inhibition of T-cell proliferation [34], possibly reflecting once again the difference between malignant and non-malignant cells in their responses to immune stimuli.

In conclusion, CD40L and TLR9 agonists are found to induce Breg properties in CLL cells in a stimulant-dependent manner. These regulatory properties suggest that CLL cells are a clone of Bregs, possibly gaining a survival advantage through the generation of a tolerant T-cell environment. Preferential TLR9 responsiveness, mediated by IL-10, could represent a novel mechanism promoting Breg behavior in CLL cells. Further studies examining the role of combinations of Breg-activating molecules, including those stimulating the B-cell receptor, are warranted to define the full extent of Breg abilities in CLL cells.

Abbreviations

ASH

American Society of Hematology

APC

Allophycocyanin

BCR

B-cell receptor

Bregs

Regulatory B-cells

BV

Brilliant violet

CD40L

CD40 ligand

CFSE

Carboxyfluorescein diacetate succinimidyl ester

CLL

Chronic lymphocytic leukemia

FISH

Fluorescence in situ hybridization

IgVH

Immunoglobulin variable region heavy chain

IRAK4

Interleukin-1 receptor-associated kinase 4

ODN

Oligodeoxynucleotide

qPCR

Quantitative polymerase chain reaction

siRNA

Small interfering RNA

Tregs

Regulatory T-cells

Compliance with ethical standards

Conflict of interest

There are no conflicts to declare.