Introduction

Chronic lymphocytic leukemia (CLL) is the most common type of leukemia in Western countries and is characterized by an increase in the amount of CD5+CD19+ cells in peripheral blood and lymphoid organs [1, 2]. The molecular pathogenesis of the disease has not yet been elucidated [3]. Chromosomal deletions such as 17p, 13q, and 11q can be seen in some patients and are associated with worse or better clinical outcomes [4]. There are two staging systems, Rai and Binet, to predict treatment requirements for CLL [5]. However, the clinical course of CLL is highly variable and difficult to predict [6].

Follicular helper T cells (TFH) are a subgroup of CD4+ T lymphocytes with CD3+CD4+CXCR5+ phenotype and are characterized by bcl-6 expression [7, 8]. They stimulate B lymphocytes in the lymph nodes, with cytokines and surface molecules such as CD40L [9]. This interaction induces activation-induced cytidine deaminase (AID) expression which is a key enzyme for somatic hyper mutation (SHM) and class switch recombination (CSR) in B cells [9,10,11]. Recent studies have shown that AID may have a role in mutations and translocations in non-Ig genes [12,13,14]. It was shown that high AID expression may be seen in CLL patients, and it might be associated with chromosomal deletions in 17p, 13q, and 11q [6, 10, 15,16,17]. Similarly, increased CD3+CD4+CXCR5+ TFH cells were seen in CLL patients [18].

Follicular cytotoxic T (TFC) cells are a new subset of CD8+ T cells expressing CXCR5 [19]. CD8+CXCR5+ TFC cells localize in follicles of lymph nodes and eliminate the infected or cancerous B cells and TFH cells by perforin and granzyme expression [19,20,21]. TFC cells also regulate the antibody response, but their role is conflicting [19]. Although its positive effect on antibody response was shown [22], other studies reported vice versa [23, 24].

Co-stimulatory and co-inhibitory receptors such as ICOS, OX-40, and PD-L1, respectively, constitute secondary signals on which immune cells rely to induce or inhibit an immune response, determining the functional outcome of T cell receptor (TCR) signaling. ICOS/ICOSL and OX-40/OX-40L signaling between TFH and B cells have a role in controlling the long-lived humoral immune response [25]. ICOS promotes the TFH cells’ recruitment into the germinal center and helps to the selection of high-affinity B cells, while OX-40 induces the TFH cells to stimulate the functions of B cells [25]. Recent studies about cancer or autoimmunity showed that targeting the ICOS/ICOSL and OX-40/OX-40L molecules might prevent the TFH-B interaction and reduce antibody affinity and also prevent to functions of TFH cells [26, 27]. PD-1 and PD-L1 have a pivotal role to inhibit the immune response, and also play an important role in cancer [28]. The new therapeutic molecules that target the interaction of PD-1 and PD-L1 are used in cancer therapy [29]. Additionally, the literature about the role of ICOS, OX-40, and PD-L1 receptors of TFH and, especially TFC cells, in CLL is limited. Herein, the expression of ICOS, OX-40, and PD-L1 of TFC and TFH cells in CLL patients was investigated.

Material and methods

Study population

Peripheral blood samples were collected from 34 CLL patients and 19 healthy subjects. All subjects underwent a complete physical examination, and they met the CLL/IW 2008 diagnostic criteria, showing the characteristic immunophenotypic profile for CD5 and CD19 expression. Twenty-three men and 11 women with CLL with a mean age of 67.09 (range, 52–78) years were included. The patients’ clinical and cytogenetic features are summarized in Table 1. Ten men and 9 women were included as healthy subjects; their mean age was 53.86 (range, 48–60) years. Written informed consent was obtained according to the Helsinki declaration and the study was approved by the local ethics committee.

Table 1 Clinical and cytogenetic features of the patients
Full size table

Flow cytometry analyses of OX-40, ICOS, and PD-L1 expression of TFH and TFC

Peripheral blood mononuclear cells (PBMCs) were isolated from heparinized blood samples by density gradient centrifugation using Ficoll-Paque (Histopaque-1077; Biochrom, Cambridge, UK) at 2100 × rpm for 25 min. The buffy coat was washed twice with phosphate-buffered saline (PBS) and suspended in a complete cell culture medium RPMI containing 10% fetal bovine serum (Sigma Chem. Co., Germany), 1% L-glutamine (Sigma Chem. Co., Germany), 1% anti-mycotic, and antibiotic solutions (Sigma Chem. Co., Germany).

The viability of PBMCs isolated from heparinized blood samples was determined by trypan blue dye exclusion. PBMCs (1 × 06) were stimulated with Cell Activation Buffer without Brefeldin A (500 ×) [(Biolegend, San Jose, CA, USA) containing PMA (40.5 μM) and ionomycin (669.3 μM) in DMSO] for 4 h. Activated PBMCs were washed, then labeled with anti-human CD3 PE.CY7 (HIT3a clone), anti-human CD4 APC (OKT-4 clone), anti-human CD8 APC.CY7 (SK-1 clone), anti-human CXCR5 FITC (J252D4 clone), anti-human PD-L1 PerCP.Cy5.5 (B7-H1 clone), anti-human OX-40 PE (ACT35 clone), and anti-human ICOS-BV421 (C398-4A clone) were incubated for 20 min at room temperature. All mAbs were purchased from Biolegend (San Jose, CA, USA). PBMCs were washed with PBS, and data were collected by NovoCyte (Agilent Technologies, USA).

Flow cytometry analyses of OX-40L, ICOSL, and PD-L1 expression of B cells

PBMCs were labeled with anti-human CD19 APC (HIB19 clone), anti-human CD5 APC.CY7 (L17F12 clone), anti-human PD-L1 PerCP.Cy5.5 (B7-H1 clone), anti-human OX-40L PE (11C3.1 clone), and anti-human ICOSL-PECY7 (2D3 clone) were incubated for 20 min at room temperature. All mAbs were purchased from Biolegend (San Jose, CA, USA). PBMCs were washed with PBS, and data were collected by NovoCyte (Agilent Technologies, USA).

Sorting

PBMCs were labeled with anti-human CD3 PE.CY7 (HIT3a clone), anti-human CD5 PECY5 (UCHT2 clone), anti-human CD8 APC.CY7 (SK-1 clone), anti-human CXCR5 FITC (J252D4 clone), anti-human CD19 APC (SJ25C1 clone), and anti-human CD56 AlxF700 (809,022 clone) incubated for 20 min at room temperature. CD5+CD19+ B-CLL cells, CD3+CD8+CXCR5− T cell, CD3+CD8+CXCR5+ TFC, and CD3−CD56+ NK cells were sorted from PBMCs by FACSAria II (BDBiosciens, USA).

B-CLL-CD8/TFC/NK cells co-culture

Sorted B-CLL cells were co-cultured with CD3+CD8+CXCR5− T cell, CD3+CD8+CXCR5+ TFC, and CD3−CD56+ NK in different well (effector/target ratio: 1/4). Effector cells (2 × 104 CD8, TFC, or NK cells) were added in a 96-well plate, then B-CLL cells were added in a well. For effector cells activation, cell activation buffer without Brefeldin A (500 ×) [(Biolegend, San Jose, CA, USA) containing PMA (40.5 μM) and ionomycin (669.3 μM) in DMSO] were added for 4 h. After incubation, B-CLL cells were washed and labeled with Viability Dye-BV510 for 20 min, then analyzed by NovoCyte (Agilent Technologies, USA).

Statistical analysis

Non-parametric variables were compared using the non-parametric Mann–Whitney U test to analyze associations in quantitative data. P values less than 0.05 were considered statistically significant. All calculations were performed using GraphPad Instat version 5.03 (GraphPad Software Inc., San Diego, CA, USA). Pearson’s correlation coefficient was used to determine the association between OX40, ICOS, and PD-L1 levels.

Results

Clinical features of patients

Thirty-four CLL patients were evaluated and 67.6% of them were male (Table 1). The patients’ clinical features are summarized in Table 1. All but 5 CLL patients were currently not under treatment and not taking any medicine at the time of blood sampling. Five patients were receiving venetoclax treatment and none of them were receiving any BTK inhibitor like ibrutinib, before venetoclax treatment. None of the patients had any findings of acute or chronic infection. According to Binet stage, 44% of patients had Binet A (n = 15), 26% of patients had Binet B (n: 9), and 30% of patients had Binet C (n = 10) disease. Ten (30%), 10 (30%), 6 (18%), 4 (11%), and 4 (11%) patients had Rai stage 0, 1, 2, 3, and 4 disease, respectively. Seven patients have chromosomal deletion with del 13q, del11q, or del 17p.

Increased TFH and TFC cells in CLL

TFH and TFC cells activate B lymphocytes in secondary lymphoid organs via surface molecules and interleukins (IL). These stimulations are important for SHM and CSR in B lymphocytes. Recent studies have focused on the indirect roles of these cells in chromosomal deletions observed in CLL patients. In this study, CD3+CD4+CXCR5+ (TFH) and CD3+CD8+CXCR5+ (TFC) cells were analyzed in PBMCs by flow cytometry. Lymphocytes were gated on SSC/FSC dot-plot per tube, then CD3+ cells were gated in SSC/CD3 dot-plot. CD4+CXCR5+ (TFH) and CD8+CXCR5+ (TFC) in CD3+ cell population were then determined (Fig. 1a and 2a, respectively). The median, minimum, maximum, and p values of all patient groups are shown in Table 2.

Fig. 1
figure 1

Gating strategy of the CD3+CD4+CXCR5+ TFH cells. CD3+ cells were gated in lymphocyte gate, CD4+ T cells were gated in CD3+ cells. The CXCR5+ in CD4+ T cells were chosen TFH cells (a). The graph represents percentages of TFH cells in healthy subjects (n = 16) and CLL patients (n = 34) according to status of treatment, deletion, CD38 expression status, Rai and Binet stating systems (b). (*p < 0.05, **p < 0.001, ***p < 0.0001; according to healthy control)

Full size image
Fig. 2
figure 2

Gating strategy of the CD3+CD8+CXCR5+ TFC cells. CD3+ cells were gated in lymphocyte gate, CD8+ T cells were gated in CD3+ cells. The CXCR5+ in CD8+ T cells were chosen TFC cells. (a). The graph represents percentages of TFC cells in healthy subjects (n = 16) and CLL patients (n = 34) according to status of treatment, deletion, CD38 expression, Rai and Binet stating systems (b). (*p < 0.05, **p < 0.001, ***p < 0.0001; according to healthy control)

Full size image
Table 2 TFH, TFC, ICOS, OX-40, and PD-1L expression [median(min–max)-p value]. (*p < 0.05, **p < 0.001, ***p < 0.0001; according to healthy control)
Full size table

TFH cells were increased in CLL patients compared to healthy subjects (p < 0.0001) (Fig. 1b). TFH cells were also increased in untreated and treated patient groups compared to healthy subjects (p = 0.0002 and p = 0.001, respectively), but there was no difference between untreated and treated patients. According to CD38 or chromosomal deletion status, elevated TFH cells were obtained in CD38 negative, CD38 positive, deletion negative, and deletion positive patient groups compared to healthy subjects (p = 0.001, p = 0.006, p = 0.004, and p = 0.005, respectively).

According to Rai and Binet staging systems, high TFH cells were detected in all stages of patients compared to healthy subjects (Rai 0, p < 0.0006; Rai 1 p = 0.02; Rai 2, p = 0.01; Rai 3, p = 0.03; Rai 4, p = 0.01, Binet A, p = 0.0007, Binet B, p = 0.01 and Binet C, p = 0.005), but there were no differences between diverse stage patient groups of Rai or Binet (Fig. 1). On the other hand, there were no differences in the context of TFH cell percentages between CLL groups regarding deletions, CD38 expression, treatment status, or Rai and Binet stages.

TFC cells were also increased in CLL patients compared to healthy subjects (p < 0.0001) (Fig. 2b). High TFC cells were detected both in untreated and treated patient groups compared to healthy subjects (p = 0.0003 and p = 0.004, respectively). According to CD38 or chromosomal deletion status, TFC cells in CD38 negative, CD38 positive, deletion negative, and deletion positive patient groups were found to be significantly higher than healthy subjects (p = 0.004, p = 0.001, p = 0.0004, and p = 0.02, respectively). However, no differences were observed between patient groups. TFC cells were increased in all stage of patients except Rai II compared to healthy controls (Rai 0, p < 0.0003; Rai 1 p = 0.01; Rai 3, p = 0.008; Rai 4, p = 0.05; Binet A, p = 0.0003; Binet B, p = 0.05 and Binet C, p = 0.001). But there were no differences, like TFH, in TFC cells between CLL groups discriminated by the presence of deletions, CD38 expression, treatment, or Rai and Binet stages.

In our study, the ratio of TFH and TFC cells was performed in PBMCs after a 4-h stimulation with cell activation buffer. It is known that the stimulation condition might affect the size and granularity of the cells. Therefore, the levels of TFH and TFC cells in stimulated PBMC may not be compatible with the ratio of the same cells in whole blood samples. However, the absolute number of TFH and TFC cells was calculated using lymphocyte counts in whole blood. The absolute number of TFH and TFC cells in patients was significantly increased compared to healthy controls (p = 0.008 and p < 0.0001, respectively) (Supplement 1). TFH cells were positively correlated with B-CLL cells and [R = 0.458, (p = 0.0009)] TFC cells [R = 0.771, (p < 0.0001)].

Increased activator and decreased inhibitor molecule expression on TFH and TFC cells in CLL patients

Co-stimulatory and co-inhibitory receptors constitute secondary signals on which immune cells rely on induce or inhibit an immune response, determining the functional outcome of TCR signaling. For the SHM and CSR activation of B cells, the balance of co-stimulatory and co-inhibitory receptors on TFH and TFC cells is critical. Thus, ICOS, OX-40, and PD-L1 expression of TFH and TFC cells after treating PBMC samples with PMA and ionomycin for 4 h were analyzed.

High -ICOS and low-PD-L1 expression on TFH cells of CLL

High-ICOS and low-PD-L1 expression were determined in TFH cells of CLL patients compared to healthy controls (p = 0.02 and p < 0.0001, respectively) (Fig. 3b). However, there were no differences in OX-40 expression on TFH cells. There were no differences in ICOS, OX-40, PD-L1 expression of TFH cells for chromosomal deletions, CD38 expression, or treatment status (Fig. 3). On the other hand, there is a significant decrease in PD-L1 expression of TFH cells in untreated patients (p < 0.0001), but it was found to be similar to healthy control levels in patients treated with Venetoclax (p > 0.05). Increased ICOS and decreased PD-L1 expression on TFH cells were detected in patients with Rai 0 (p = 0.01 and p = 0.001, respectively), Rai I (p = 0.05 and p = 0.001, respectively), Rai II (p > 0.05 and p = 0.01, respectively), Rai III (p > 0.05 and p = 0.05, respectively), Rai IV (p > 0.05 and p = 0.01, respectively), Binet A (p = 0.05 and p = 0.0004, respectively), and Binet B (p = 0.03 and p = 0.0005, respectively) and Binet C (p > 0.05 and p = 0.005, respectively) compared to healthy controls (Fig. 3). OX-40 expression was positively correlated with TFH cells levels [R = 0.504, (p = 0.002)]. Additionally, OX-40 expression was positively correlated with ICOS expression [R = 0.660, (p < 0.0001)] and PD-L1 expression [R = 0.649, (p < 0.0001)]. Although PD-L1 expression increased in patients, this high expression of PD-L1 remains lower than in healthy controls. No differences were found between patients and healthy controls for MFI values of ICOS, OX-40, or PD-L1 expression on TFH cells.

Fig. 3
figure 3

Representative dot-plots of OX-40, ICOS, and PD-L1 expression on TFH cells in unstained tube, CLL patients, and healthy subjects (a). The graph represents percentages OX-40, ICOS, and PD-1L expression of TFH cells in healthy subjects (n = 16) and CLL patients (n = 34) according to status of treatment, deletion, CD38 expression, Rai and Binet stating systems (b). (*p < 0.05, **p < 0.001, ***p < 0.0001; according to healthy control)

Full size image

High-ICOS and low-PD-L1 expression on TFC cells of CLL

Similarly, to TFH, high-ICOS and low-PD-L1 expression were determined in TFC cells of CLL patients compared to healthy controls (p = 0.007 and p = 0.03, respectively) (Fig. 4b). But additionally, increased OX-40 levels in TFC cells of CLL patients were observed (p = 0.006). There were no differences in ICOS, OX-40, and PD-L1 expression of TFC cells for chromosomal deletions, CD38 expression, or treatment status (Fig. 4). On the other hand, there is a high ICOS and OX-40, low-PD-L1 expression of TFC cells in untreated patients, but the level of these molecules was found to be similar to healthy control levels in patients treated with venetoclax (untreated: p = 0.009, p = 0.004, and p = 0.007, respectively; treated p > 0.05, for all parameters). According to Rai and Binet staging systems, increased ICOS and OX-40 expression on TFC cells were detected only in Rai 0 (p = 0.003 and p = 0.002, respectively), Rai I (p = 0.04 and p = 0.05, respectively), Binet A (p = 0.002 and p = 0.002, respectively), and Binet B (p = 0.05 and p > 0.05, respectively) stage patients compared to healthy controls (Fig. 4). Low-PD-L1 expression was detected only in Binet B (p = 0.003), Rai I (p = 0.008), and Rai II (p = 0.05) stage patients. OX-40 and PD-L1 expression was positively correlated with TFC cells levels [R = 0.316, (p = 0.05) and R = 0.387, (p = 0.02), respectively]. Additionally, OX-40 expression was positively correlated with ICOS expression [R = 0.753, (p < 0.0001)] and PD-L1 expression [R = 0.749, (p < 0.0001)]. Similarly, ICOS expression was positively correlated with PD-L1 expression [R = 0.405, (p = 0.01)]. Although PD-L1 expression increased in patients, this increase remains lower than in healthy controls. No differences were found between patients and healthy controls for MFI values of ICOS, OX-40, or PD-L1 expression in TFC cells.

Fig. 4
figure 4

Representative dot-plots of OX-40, ICOS, and PD-L1 expression on TFC cells in, unstained tube, CLL patients, and healthy subject (a). The graph represents percentages OX-40, ICOS, and PD-1L expression of TFC cells in healthy subjects (n = 16) and CLL patients (n = 34) according to status of treatment, CD38 expression, deletion, Rai and Binet stating systems (b). (*p < 0.05, **p < 0.001, ***p < 0.0001; according to healthy control)

Full size image

High OX-40L, ICOSL, and PD-L1 expression on B Cells of CLL

Because of increased OX-40 and ICOS expression of TFH and TFC cells in CLL patients, these receptor ligands OX-40L and ICOSL of B cells were analyzed in PBMCs. Although the OX-40L, ICOSL, and PD-L1 expression of B cells in patients did not differ significantly compared to healthy subjects, the mean fluorescence intensity (MFI) levels of OX-40L, ICOSL, and PD-L1 in B cells in patients significantly increased compared to healthy subjects (p = 0.0002, p = 0.001, and p < 0.0001, respectively) (Fig. 5). There were no differences in ICOSL, OX-40L, and PD-L1 expression of B cells for chromosomal deletions, CD38 expression, or treatment status (Fig. 4). According to Rai and Binet staging systems, increased OX-40L and ICOSL expression on B cells were detected in Rai II (p = 0.003 and p = 0.002, respectively), Rai IV (p = 0.01), Binet A (p = 0.002 and p = 0.02, respectively), Binet B (p = 0.03 and p = 0.05, respectively), and Binet C (p = 0.001 and p = 0.0007, respectively) stage patients compared to healthy controls (Fig. 4). Increased PD-L1 expression were detected all Rai and Binet B (p = 0.0009, p = 0.0006, p = 0.001, p = 0.01, p = 0.001, p = 0.0001, p = 0.0005, p < 0.0001, respectively) stage patients compared to healthy controls. OX-40L expression in B cells was positively correlated with ICOSL expression [R = 0.568, (p = 0.0004)].

Fig. 5
figure 5

Representative dot-plots of OX-40L, ICOSL, and PD-L1 expression on B cells in, unstained tube, CLL patients, and healthy subject (a). The graph represents percentages OX-40L, ICOSL, and PD-1L expression of B cells in healthy subjects (n = 16) and CLL patients (n = 34) according to status of treatment, CD38 expression, deletion, Rai and Binet stating systems (b). (*p < 0.05, **p < 0.001, ***p < 0.0001; according to healthy control)

Full size image

Cytotoxic effect of TFC cells

TFC, NK, and CD8 cells from CLL patients (n = 6) were cultured with B-CLL cells to compare the cytotoxic effects of TFC cells with NK and CD8 cells. After incubation, the viability rates of B-CLL cells were analyzed. There were no differences between the cytotoxic effects of TFC cells and CD8 or NK cells. These results showed that TFC cells are as effective as NK or CD8 cells to lyses of B-CLL cells (Fig. 6).

Fig. 6
figure 6

Gating strategy of the CD5+CD19+ B-CLL cells, CD3+CD8+CXCR5+ TFC, CD3+CD8+CXCR5− T cells, and CD3−CD56+ NK cells for sorting (a). Representative dot-plots of viability of B-CLL cells with CD3+CD8+CXCR5+ TFC, CD3+CD8+CXCR5− T cells, and CD3−CD56+ NK cells (b, c)

Full size image

Discussion

CLL is characterized by the accumulation of small, mature-appearing CD5+ B cells in blood, bone marrow, and lymphoid tissues and has a highly variable clinical course [30]. IgVH mutation and chromosomal deletions such as del17p, del13q, and del11q have prognostic impact [31, 32]. The molecular mechanism of these genetic changes has not yet been elucidated. Knock-out mice studies have specifically shown that constitutive and ubiquitous AID expression induces c-myc/IgH chromosome translocations that are responsible for the development of plasmacytomas [13, 33]. Other studies showed that AID acts as a mutator in BCR-ABL-1 transformed acute leukemia cells [34], and is associated with imatinib resistance [35]. Therefore, AID has a potent role in mutation in non-Ig genes [12, 16]. Current studies about CLL suggested that AID enzyme may be responsible for these genetic changes [6, 10, 36, 37].

TFH and B cell interaction is required for AID expression [38]. Our previous data and different studies showed that CLL patients have high TFH cells [18, 39, 40]. TFH cells have co-stimulatory and co-inhibitory receptors like OX-40, ICOS, and PD-1L [41]. OX-40 and ICOS are expressed by activated CD4+ and CD8+ T cells, and they have been shown to be an important co-stimulator for antibody responses and germinal center reactions [42,43,44]. Recent studies showed that elevated TFH cells had increased expression of PD-1, ICOS, and IL-21 were seen in CLL patients [45,46,47]. Similarly, TFH cells were increased in CLL patients and, TFH cells had high activator receptor ICOS expression and lower PD-L1 expression in CLL patients compared to healthy controls in this study. These findings showed that TFH cells of CLL patients might be more potent and active than TFH cells of healthy subjects. Co-culture experiment in CLL indicated that there is a positive correlation between the ratio of TFH cells and B-CLL cell proliferation and high expression of various cytokines and co-stimulatory molecules in TFH cells might be associated with B-CLL cell proliferation [45]. To our results, OX-40L and ICOSL expression of B cells were increased in CLL patients. These results indicated that more potent, and active TFH cells might interact with more B cells, and stimulate AID expression more extensively on B cells, possibly contributing to the high-AID expression observed in CLL patients. Additional studies comparing the potential role of CLL and healthy TFH cells to induce AID expression on B cells are needed to clarify the pathogenesis. However, there was no difference was obtained in TFH cell ratio between Rai and Binet staging groups. It was seen that increased TFH cells might be an independent factor of the staging systems.

TFC cells defined as a new subgroup of CD8+ T cells have the ability to migrate to inside or proximal part of the B follicle by expressing CXCR5 chemokine receptor and to secrete enzymes such as perforin and granzyme, eliminating infected or malignant cells [19, 20]. It has been suggested that TFC cells can carry out immune surveillance in B cell follicles to eliminate infected and cancerous cells [48]. TFC cells may also prevent B cell malignancy. Recent studies showed that TFC cells have role to eliminate infected cell in many infectious diseases like HIV, EBV [49], and many malignant diseases such as hepatocellular carcinoma and pancreatic tumors [50, 51]. There were not many studies about the role of TFC cells in CLL pathogenesis.

A recent study focused on the role of TFC cells in malignancy indicated that CXCR5+PD-1+CD8+ T cells were increased effector differentiation in CLL patients, and CXCR5+PD-1+CD8+ T cells produced high IFN-γ and TNF-α after stimulation [52]. Similarly, increased TFC cells were found in CLL patients, and TFC cells had elevated expression of activator receptor OX-40 and ICOS expression and low expression of PD-1L on TFC cells in CLL patients. Like TFH cells, high amounts, more potent and active TFC cells were seen in CLL. We showed that ICOS and OX-40 expression on TFC cells is increased in patients without chromosomal deletion, but it was similar to healthy controls in patients with chromosomal deletion. Increased activity of TFC cells might be more prominent to eliminate malignant B cells and TFH cells in the lymph nodes of patients without chromosomal deletion and could then inhibit B cell-TFH interaction that is required for AID expression. This might provide indirect protection to the mutator effect of AID in non-Ig genes of B cells. Some studies indicated that TFC cells had decreased perforin and granzyme levels compared to CXCR5− classical CD8+ T cells; therefore, cytotoxicity of TFC cells was reduced in the acute phase of infection [19], but some studies showed that IL-21 expressing TFC cells do not express perforin and granzyme [53]. To analyze the cytotoxic effect of TFC cells compared to NK and CD8+ T cells in our study, B-CLL cells were co-cultured with TFC, NK, and CD8+ T cells. To our data, TFC cells have the same cytotoxic effect on B-CLL cells as NK and CD8+ T cells. In this study, TFC cells were analyzed in peripheral blood, not in the lymph nodes. Lymph nodes of CLL cases are not sampled until Richter transformation is suspected or if there is no treatment resistance. The studies that plan to investigate the role of TFC cells in the lymph node could contribute to the biology of CLL.

At the same time, it has been reported that TFC cells expressing ICOSL inhibit antibody response and autoimmunity by suppressing TFH cells, while TFC cells expressing CD40L and secreting IL-21 increase antibody responses and CSR by stimulating B cells [19]. Increased TFC cells in CLL patients might express ICOSL and IL-21 and more stimulate B cells to enhance AID expression. Future studies focusing on the functions of TFC cells with different phenotypes will reveal the role of these cells in the pathogenesis of CLL more clearly.

BCL2 is an anti-apoptotic protein pathologically overexpressed on the CLL cells. Venetoclax, a BCL2 inhibitor, is an active drug even in CLL patients with high genetic risk factors [54]. Decreased PD-L1 expression on TFH cells in untreated patients, increased ICOS and OX-40 expression and decreased PD-1L expression on TFC cells were detected in untreated patients, all three parameters were similar to healthy controls in patients with venetoclax treatment. Our findings have suggested that venetoclax might affect the expression of co-activator/inhibitor receptors on TFH and TFC cells, resulting in changes in functional activity. Even there was no difference was obtained in TFC cells ratio between Rai and Binet staging groups, increased TFC cells in total CLL patients would be independent of the staging system.

Further studies with larger numbers of patients may allow identification of the role of TFH and TFC cells in CLL pathogenesis. Increasing the number of advanced-stage patients will be more efficient for comparing early and advanced-stage patients. Elevated CD3+CD4+CXCR5+ TFH and CD3+CD8+CXCR5+ TFC cells with high OX-40 and ICOS activator and low PD-1L inhibitor receptor expression in CLL patients might have a role in the pathogenesis of CLL. Further studies about the functional and cytotoxic role of TFC will shed light.