Abstract
Background: B-cell receptor (BCR) signaling is crucial for normal B-cell development and adaptive immunity. In chronic lymphocytic leukemia (CLL), the malignant B cells display many features of normal mature B lymphocytes, including the expression of functional B-cell receptors (BCRs). Cross talk between CLL cells and the microenvironment in secondary lymphatic organs results in BCR signaling and BCR-driven proliferation of the CLL cells. This critical pathomechanism can be targeted by blocking BCR-related kinases (BTK, PI3K, spleen tyrosine kinase) using small-molecule inhibitors. Among these targets, Bruton tyrosine kinase (BTK) inhibitors have the highest therapeutic efficacy; they effectively block leukemia cell proliferation and generally induce durable remissions in CLL patients, even in patients with high-risk disease. By disrupting tissue homing receptor (i.e., chemokine receptor and adhesion molecule) signaling, these kinase inhibitors also mobilize CLL cells from the lymphatic tissues into the peripheral blood (PB), causing a transient redistribution lymphocytosis, thereby depriving CLL cells from nurturing factors within the tissue niches. Summary: The clinical success of the BTK inhibitors in CLL underscores the central importance of the BCR in CLL pathogenesis. Here, we review CLL pathogenesis with a focus on the role of the BCR and other microenvironment cues. Key Messages: (i) CLL cells rely on signals from their microenvironment for proliferation and survival. (ii) These signals are mediated by the BCR as well as chemokine and integrin receptors and their respective ligands. (iii) Targeting the CLL/microenvironment interaction with small-molecule inhibitors provides a highly effective treatment strategy, even in high-risk patients.
Introduction
Chronic lymphocytic leukemia (CLL) is characterized by the expansion of monoclonal mature B lymphocytes expressing CD5 and CD23 in the blood, bone marrow (BM), and secondary lymphatic organs (SLO, i.e., lymph nodes, spleen). CLL cells retain many functional properties of normal B cells, including key signaling pathways such as the B-cell receptor (BCR) and its downstream signaling cascade that normally organizes B-cell selection and proliferation as part of adaptive immune responses [1]. CLL cell survival and proliferation are dependent on signals from the microenvironment in SLO. Direct in vivo measurements demonstrated that the SLO are the primary site of CLL cell proliferation [2, 3]. When removed from the supportive microenvironment, CLL cells undergo spontaneous apoptosis ex vivo unless co-cultured with bone marrow stromal cells (BMSC) [4, 5] or monocyte-derived nurse-like cells (NLC) [6]. The clinical efficacy of kinase inhibitors that disrupt BCR signaling and other CLL cell interactions with the tissue microenvironment further corroborates the central importance of BCR signaling and the tissue microenvironment in CLL pathogenesis [7].
The entire CLL clone consists of a small fraction of proliferating CLL cells, which are primarily found in SLO, and a vast majority of resting B cells [8]. CLL cells from these compartments are distinct in their morphology, gene expression, and immunophenotype. The proliferative compartment in SLO forms characteristic “proliferation centers (PC),” also called “pseudofollicles” that are a typical finding in CLL histopathology. CLL cell proliferation can be quantified in patients by heavy (deuterated) water labeling, ranging between 0.1 to greater than 1% of the entire clone per day [9]. Small resting lymphocytes, the progeny of the proliferating compartment, accumulate in the blood and tissues (SLO, marrow) and traffic between these compartments, in an active process of migration that is organized by chemokine receptors and adhesion molecules, which are expressed on the CLL cell surface, and respective tissue ligands (Table 1).
Constituents of the CLL microenvironment
Receptors/ligands . | Function . | Prognostic impact . |
---|---|---|
BCR | Main driver of CLL cell proliferation and survival; activation is mediated through exogenous (auto) antigens or autonomous BCR activation | IGHV mutation status distinguishes indolent from more active disease courses [10] |
CXCR4 | Surface receptor for CXCL12; mediates the migration and chemotaxis of CLL cells | CXCR4 expression levels allow the distinction of quiescent CXCR4bright/CD5dim and proliferating CXCR4dim/CD5bright CLL cells [11] |
CXCR5 | Surface receptor for CXCL13 expressed on CLL cells; mediates chemotaxis and actin polymerization | - |
CCL3/4 | Secreted by CLL cells to attract T cells and monocytes to the microenvironment | Serum CCL3 levels serve as a prognostic marker for time to treatment [12] |
VLA-4 | Expressed on CLL cells; facilitates CLL cell adhesion through binding of fibronectin or VCAM-1 | VLA-4 expression levels on CLL cells predict time to treatment and overall survival [13] |
Receptors/ligands . | Function . | Prognostic impact . |
---|---|---|
BCR | Main driver of CLL cell proliferation and survival; activation is mediated through exogenous (auto) antigens or autonomous BCR activation | IGHV mutation status distinguishes indolent from more active disease courses [10] |
CXCR4 | Surface receptor for CXCL12; mediates the migration and chemotaxis of CLL cells | CXCR4 expression levels allow the distinction of quiescent CXCR4bright/CD5dim and proliferating CXCR4dim/CD5bright CLL cells [11] |
CXCR5 | Surface receptor for CXCL13 expressed on CLL cells; mediates chemotaxis and actin polymerization | - |
CCL3/4 | Secreted by CLL cells to attract T cells and monocytes to the microenvironment | Serum CCL3 levels serve as a prognostic marker for time to treatment [12] |
VLA-4 | Expressed on CLL cells; facilitates CLL cell adhesion through binding of fibronectin or VCAM-1 | VLA-4 expression levels on CLL cells predict time to treatment and overall survival [13] |
Cellular Components of the CLL Microenvironment
NLCs (CLL-Associated Macrophages)
The first evidence that myeloid cells are contributing to the pathogenesis of CLL came from the in vitro observation that blood monocytes, in the presence of CLL cells, spontaneously differentiate into large adherent feeder cells termed NLC which protect CLL cells from apoptosis [6, 14]. NLC secrete the chemokines CXCL12 and CXCL13, activating CLL cells via the corresponding chemokine receptors CXCR4 and CXCR5, respectively [6, 15], thereby contributing to the survival and tissue homing of CLL cells. In addition, NLC also express other CLL survival factors, such as B-cell activating factor (BAFF) and a proliferation-inducing ligand (APRIL) [16, 17]. Gene expression profiling of CLL cells revealed that NLC activate BCR and NF-κB signaling [18], recapitulating in vitro the gene expression signature of CLL cells isolated from CLL lymph nodes [2]. NLC express antigens such as vimentin and calreticulin that can directly crosslink the BCR and activate BCR signaling in CLL cells [19, 20]. They also carry high levels of CD31 and plexin-B1, which deliver survival and proliferation signals to CD38+/CD100+ CLL cells [21]. Co-culture with NLC induces BCR signaling-dependent transcription and secretion of the chemokines CCL3 and CCL4 by CLL cells that can attract CLL-supportive immune cells, especially T cells and monocytes, into the tissue microenvironment [18, 22]. In addition, NLC secrete other cytokines including IL-6, IL-8, and IL-10 that promote CLL cell proliferation and/or survival.
Based on gene expression profiles, surface markers, and cytokine expression, NLC resemble M2-skewed tumor-associated macrophages (TAM). NLC express high levels of the scavenger receptor CD163, a marker of M2 tissue macrophages, along with CD68. In the tissues (lymph nodes, BM), NLC form cell membrane extensions for intimate contact with CLL cells, which are disrupted by BTK inhibitor therapy with ibrutinib [23]. Increased numbers of CD163+ or CD68+ cells in LN are associated with active disease requiring treatment and shorter overall survival [24], and high density of CD163+ cells in BM correlated with disease persistence with ibrutinib therapy [25], corroborating the importance of NLC in CLL. Further evidence that NLC are the key players in the CLL microenvironment was generated in CLL mouse models. In the Eµ-TCL1 mouse model, monocytes and M2-skewed macrophages co-localize to sites where CLL cells accumulate, such as in the peritoneal cavity, spleen, and BM. Targeted deletion of macrophage migration inhibitory factor (MIF) leads to reduced numbers of macrophages in the spleen and BM and significantly delays the development of CLL in Eµ-TCL1 mice [26]. Accordingly, depletion of monocytes and macrophages significantly delayed the development of CLL and improved survival in several mouse models [27, 28]. Furthermore, monocytes and macrophages facilitate leukemia engraftment in xenotransplantation models of CLL [29, 30].
Mesenchymal Stromal Cells
Mesenchymal stromal cells (MSC) are well known for supporting the growth and maturation of normal hematopoietic progenitor cells (HPC). These cells are primarily found in the BM, where they are called BMSC, but also are a part of the stromal cell network in normal [31] and malignant SLO, for example, in CLL patients [32, 33]. The initial evidence that MSC are important for CLL cell survival came from studies that demonstrated the rescue of CLL cells from spontaneous and drug-induced apoptosis when co-cultured with BMSC [34]. The high affinity of CLL cells for BMSC is illustrated by pseudo-emperipolesis, the spontaneous migration of CLL cells beneath and underneath BMSC layers in vitro [35]. BMSC constitutively secrete stromal cell-derived factor-1 (SDF-1, now designated CXCL12), a tissue homing chemokine that attracts CLL cells via the corresponding chemokine receptor CXCR4, which CLL cells express highly on their cell surface [18] (Fig. 1). Besides attracting CLL cells into the tissues, CXCL12 also has modest pro-survival activity and is an important factor in BMSC-mediated protection of CLL cells from spontaneous or chemotherapy-induced apoptosis [6, 36]. Adhesion molecules are also highly important for interactions between CLL cells, MSC, and other cells in the tissue microenvironment. For example, VLA-4 integrins (CD49d/CD29) on CLL cells can bind to VCAM-1 or vimentin, both present in the tissue microenvironment and on MSC, and binding subsequently activates CLL cells and rescues them from apoptosis via Bcl-2 and Bcl-XL up-regulation. The importance of VLA-4 in CLL is further emphasized by its robust prognostic impact; high VLA-4 levels on CLL cells correlate with increased homing to the BM and inferior clinical outcome [37].
Composition of the CLL microenvironment. CLL cells rely on a multitude of signals from the tissue microenvironment in SLO to promote their survival and proliferation. These signals are received through CLL surface receptors that are engaged by corresponding microenvironmental ligands, including signals from the BCR, CD40, chemokine receptors (CXCR4, CXCR5), and adhesion molecules, such as integrins, and TNF-family receptors including BAFF-R, TACI, and BCMA. BTK inhibitors inhibit several of these signaling cascades and effectively disrupt the crosstalk between CLL cells and the microenvironment, thereby depriving them from essential survival factors, and disrupting tissue homing signals, resulting in mobilization of tissue CLL cells from SLO into the PB.
Composition of the CLL microenvironment. CLL cells rely on a multitude of signals from the tissue microenvironment in SLO to promote their survival and proliferation. These signals are received through CLL surface receptors that are engaged by corresponding microenvironmental ligands, including signals from the BCR, CD40, chemokine receptors (CXCR4, CXCR5), and adhesion molecules, such as integrins, and TNF-family receptors including BAFF-R, TACI, and BCMA. BTK inhibitors inhibit several of these signaling cascades and effectively disrupt the crosstalk between CLL cells and the microenvironment, thereby depriving them from essential survival factors, and disrupting tissue homing signals, resulting in mobilization of tissue CLL cells from SLO into the PB.
Given the multitude of surface molecules and downstream signaling pathways activated by contact between CLL cells and MSC, it is no surprise that MSC induce substantial changes in CLL cell gene and protein expression. Co-culture with BMSC leads to substantial up-regulation of the T-cell leukemia/lymphoma 1 (TCL1) proto-oncogene in CLL cells [38]. MSC induce up-regulation of ZAP-70, CD38, and CD49d, which are markers associated with more aggressive clinical behavior. Contact with MSC also induces CLL cell resistance to chemotherapeutic agents, a mechanism to explain why residual disease after chemotherapy-based regimen typically is more predominant in the tissues than in the blood. CLL cells not only benefit from the contact with tissue MSC, they also, vice versa, activate and reshape MSC. Through the release of exosomes and via direct cell-cell contact, CLL cells can induce protein kinase C beta II (PKCβII) expression and nuclear factor kappa B (NF-κB) pathway activation in MSC, inducing an inflammatory cancer-associated fibroblast (CAF) phenotype. Activated stromal cells, in turn, secrete pro-tumorigenic cytokines, such as IL-1 and IL-15 which further propagate this symbiotic relation between CLL and MSC. In an elegant series of experiments, vom Stein et al. [39] recently demonstrated that the LYN kinase, which is overexpressed in lymph node stromal cells from CLL patients, induces the inflammatory CAF phenotype via modulation of cytokine secretion and extracellular matrix (ECM) components. In turn, such activated CAF effectively promoted CLL cell survival.
T Cells
During normal adaptive immune responses, T cells migrate to germinal centers (GCs) for providing help to antigen-selected B cells and constitute 5–20% of the GC. The GC response depends on CD4+ T cells and CD40, as well as the costimulatory molecules CD28 and ICOS, and the cytokines IL-4 and IL-21 [40]. GC B cells compete for contact/help by these T cells, resulting in the selection and expansion of B cells carrying B-cell receptors (BCRs) with high affinity to the antigen. The chemokines CCL3 and CCL4, secreted by B cells in response to BCR triggering [41], attract T cells and seem to play an important role in this selection process, a mechanism which also seems to play a role in CLL [18, 42].
Normal GC reactions are regulated by specialized CD4+ T helper cell subpopulations, especially T follicular helper (Tfh) cells, which provide the majority of T cell help to GC B cells by activating B cells via CD40 ligand and IL-21. Tfh cells are characterized by high expression of the CXCR5 chemokine receptor, also highly expressed on CLL cells [15, 43], which allows normal B and Tfh cells to home to the SLO. Tfh cells also express the inducible co stimulator molecule (ICOS), programmed cell death protein-1 (PD-1), transcriptional repressor B-cell lymphoma 6 (BCL6), and produce IL-21. Tfh maturation stages can be distinguished based on expression levels of CXCR5 and PD-1. For example, Tfh cells which exit GCs are less activated and become memory Tfh cells, with low or absent PD-1 expression. Another important T helper cell subset in SLO is the regulatory T (Treg) cell which regulates immune tolerance and homeostasis and express the transcription factor FOXP3.
In patients with CLL, peripheral blood T-cell numbers are increased and oligoclonal [44], and BM and LN are also enriched in T cells, predominantly with CD4+ helper cells, constituting a substantial portion of the lymphoid infiltrate. T cells localize both inside and outside of proliferation centers [45‒48]. When investigating T-cell subsets in CLL, several studies demonstrated increased frequencies of regulatory T cells (Tregs), interleukin 17-secreting T cells (Th17), and follicular T helper cells (Tfh) in PB [49]. T-cell populations in SLO from CLL patients remain incompletely defined. Generally, untreated CLL patients have increased numbers of CD4+ T cells in their PB, and CLL lymph nodes also are densely infiltrated by T helper cells [42, 50] which are in intimate contact with proliferating CLL cells [48] and express CD40L [47]. On the other hand, immunosurveillance by T cells is compromised in CLL, due to, among others, defective cytoskeletal signaling and immune synapse formation with CLL cells due to immunosuppressive signaling [51]. CLL T cells express high levels of exhaustion markers, including programmed cell death protein-1 (PD-1); accordingly, CLL cells express high levels of PD-L1. Additionally, CLL T cells demonstrate increased expression of the inhibitory receptor cytotoxic T-lymphocyte-associated protein 4 (CTLA-4). Besides the well-recognized functional deficiencies in the T-cell compartment in CLL patients, there is also growing evidence that specialized T helper cells, which are expanded in untreated patients, support the growth of the CLL clone in the SLO. The requirement for activated CD4+ T cells for survival and growth of CLL cells in patient-derived xenograft (PDX) models was among the first robust evidence for this model [29]. This observation was extended by demonstrating that in vitro pre-activation of CLL T cells with CD3/CD28 beads and IL-2 resulted in robust transient CLL cell expansion in vivo [52].
TCR sequencing in CLL patients documented a restricted TCR gene repertoire, with an oligoclonal T-cell expansion and shared T-cell receptor clonotypes among different patients, suggesting T-cell selection and expansion by restricted antigens, possibly on CLL cells. This is supported by experiments in which T cells isolated from CLL patients were stimulated with autologous CLL cells, resulting in a clonal T-cell expansion. Similarly, in the CLL Eμ-TCL1 mouse model, a restricted TCR repertoire was found [53]. Vice versa, when CLL patients receive effective therapy (with ibrutinib), the TCR repertoire diversifies, along with deletion of high-frequency clones and normalization of T-cell counts [54]. In addition, in a recent study in patients with stereotyped BCR immunoglobulins, TCR repertoire analyses demonstrated remarkable repertoire skewing and T-cell oligoclonality, further supporting the concept of co-evolution of CLL and supportive T-cell clones [44]. Most studies about T helper cells in CLL have been based on PB samples, even though, as discussed, their primary role appears to be in the SLO. One study recently reported, based on lymph node aspirates, about enrichment in Tfh and Treg in CLL lymph nodes, as compared to PB, and their depletion from the PB after treatment with obinutuzumab plus venetoclax or with ibrutinib plus venetoclax [55]. We recently reported that co-culture of CLL T cells in the NLC model resulted in the expansion of Tfh cells based on PD-1, BCL6, and ICOS expression [56]. Tregs, which promote immune tolerance, also expanded in NLC co-cultures. T-cell receptor gene repertoire analyses confirmed the clonal expansion of T helper cells, with an enrichment of TR clonotypes commonly expanded also in primary CLL samples. Multicolor confocal microscopy revealed that Tfh, but not Tregs co-localize with proliferating CLL cells in CLL lymph node sections [56].
Follicular Dendritic Cells
Follicular dendritic cells (FDC) are an important component of normal lymphoid follicles, where they present antigens to selected B cells. However, the role of FDC in CLL remains ill-defined, based on variable results when staining CLL tissues for FDC and detection of FDC only during early-stage CLL [57]. A FDC cell line protected CLL cells from apoptosis in vitro [58]. In the Eµ-TCL1 mouse model, CXCR5-expressing leukemia cells migrate towards CXCL13-secreting FDC to gain access to the growth-promoting follicle microenvironment [59], and, in turn, lymphotoxin alpha beta (LTαβ) secretion by CLL cells promoted CXCR13 production by FDC [59].
BCR signaling in CLL
In the lymph node microenvironment, BCR signaling is activated in the CLL cells [2], resulting in leukemia cell proliferation [3]. Signaling downstream of the BCR activates upstream kinases (LYN, spleen tyrosine kinase [SYK], BTK, PI3K). Among these, BTK emerged as the most successful therapeutic target. Downstream signaling via ERK and NFκB results in transcriptional activation, for example, of the T-cell chemo-attractants CCL3 and CCL4, which are secreted at high concentrations by BCR-activated CLL [18] and function as plasma biomarkers when targeting BTK [60] or PI3K [61]. The evidence that CLL cell proliferation is BCR-driven is based on the following findings: first, the pseudo-follicular architecture [62] and engagement of proliferating CLL cells with activated T helper cells [47, 48] and specifically with Tfh [56] replicates the cellular architecture of normal B-cell follicles. Second, SLO are the primary site of CLL cell proliferation, as demonstrated directly in patients after labeling of CLL cells with deuterated (heavy) water [3]. Third, gene expression studies demonstrated that BCR and NF-κB signaling are exclusively activated in CLL cells isolated from SLO but not in blood or BM CLL cells [2]. Fourth, inhibition of BCR signaling with kinase inhibitors, such as ibrutinib, results in immediate and almost complete abrogation of CLL cell proliferation [63]. As with normal B cells, additional costimulatory signals are necessary for CLL cell expansion, which appear to derive from T cells and macrophages/NLC, as discussed earlier.
The mutational status of the Ig heavy chain variable (IGHV) genes is a powerful prognostic marker in CLL. IGHV mutation analysis distinguishes mutated CLL (M-CLL) with >2% deviation from the IGHV germline sequence, who generally have indolent disease, from patients with unmutated CLL (U-CLL) with ≤2% deviation [10, 64], who present with more active disease. These differences in clinical behavior have been linked to differences in BCR responsiveness. U-CLL clones respond to various (auto) antigens, triggering robust BCR signaling, whereas M-CLL display more selective binding to restricted antigens, and less active BCR signaling (Fig. 2a). Consequently, M-CLL clones remain stable or expand at a slower rate than clones from U-CLL patients. Interestingly, the negative prognostic impact of U-CLL is overcome when patients are treated with the new targeted agents. Long-term follow-up of CLL patients receiving ibrutinib frontline therapy demonstrated that the survival of U-CLL patients was excellent and not different from those with M-CLL [65].
CLL BCR subsets and BCR signaling inhibitors. a CLL patients can be distinguished based on the mutational status of IGHV gene sequences with the variable region of the BCR heavy chains (VH) and categorized into U-CLL or M-CLL, which differ in terms of activation properties and antigen affinity and selectivity, as highlighted in the figure. b Engagement of the BCR results in the activation of the receptor-associated kinases LYN and SYK which further propagate the signal via BTK, PI3K, and PLCg2. This ultimately results in the activation of ERK and NFkB and the modulation of their transcriptional targets. The BCR-associated kinases also serve as important hubs for signals emanating from receptors other than the BCR, such as integrin and chemokine receptors. SYK, PI3K, and BTK are targets for small-molecule kinase inhibitors, including the SYK inhibitor fostamatinib (not approved), the BTK kinase inhibitors ibrutinib, acalabrutinib, zanubrutinib, and pirtobrutinib, and the PI3K inhibitors idelalisib and duvelisib.
CLL BCR subsets and BCR signaling inhibitors. a CLL patients can be distinguished based on the mutational status of IGHV gene sequences with the variable region of the BCR heavy chains (VH) and categorized into U-CLL or M-CLL, which differ in terms of activation properties and antigen affinity and selectivity, as highlighted in the figure. b Engagement of the BCR results in the activation of the receptor-associated kinases LYN and SYK which further propagate the signal via BTK, PI3K, and PLCg2. This ultimately results in the activation of ERK and NFkB and the modulation of their transcriptional targets. The BCR-associated kinases also serve as important hubs for signals emanating from receptors other than the BCR, such as integrin and chemokine receptors. SYK, PI3K, and BTK are targets for small-molecule kinase inhibitors, including the SYK inhibitor fostamatinib (not approved), the BTK kinase inhibitors ibrutinib, acalabrutinib, zanubrutinib, and pirtobrutinib, and the PI3K inhibitors idelalisib and duvelisib.
Originally, the presence or absence of hypermutations in IGHV sequences of CLL patients was interpreted as a sign of antigenic stimulation or lack thereof in M-CLL versus U-CLL, respectively. However, this initial view has been replaced by the model that all CLL cases, regardless of IGHV mutation status, encounter antigen(s) through their BCR, with differential signaling effects. This is supported by the CLL cell surface marker pattern that resembles antigen-experienced B cells [66], with characteristic up-regulation of CD27 and CD23, and down-modulation of CD79b and surface IG [66]. Low-surface IG levels on CLL cells can be recovered to normal, high levels when cultured in vitro, in the absence of antigenic stimulation [67]. Vice versa, sIg levels can be down-modulated by BCR activation in co-culture of CLL cells with NLC, and recover/up-regulated by removal from NLC [68]. BCR cross-linking induces potent pro-survival and proliferative signals in U-CLL cells, while M-CLL are less BCR-responsive [69, 70]. The evidence of antigenic stimulation in CLL, including U-CLL, ignited a quest for putative CLL antigens, particularly in patients with stereotyped BCR, because these clones are most probably selected by antigen. Structural studies demonstrated that, generally, IG from U-CLL clones are polyreactive, resembling natural antibodies [71, 72]; in contrast, the IG from M-CLL have higher specificity and more restricted binding, for example, to fungal antigens [73]. Regarding the putative antigens, of particular interest is CLL BCR binding of molecules on apoptotic bodies, such as non-muscle myosin heavy chain IIA [74] or oxidized LDL [72], a profile reminiscent of natural antibodies engaged in the removal of apoptotic cell debris. Foreign antigens can also be recognized in some instances, including bacterial polysaccharides [72] or fungal glucans [73], and strong associations with viral infections (CMV or EBV) [75] have also been demonstrated in distinct stereotyped subsets. Also relevant to mention, CLL IG may show a rheumatoid factor activity (i.e. auto-reactivity against the FC portion of IgG antibodies), a finding reported more than 50 years ago [76, 77], though more recently associated with distinct stereotyped BCR IG, likely in the context of HCV infection [78].
The concept of BCR activation through exogenous (auto) antigens in CLL has been challenged by an alternative model, activation through autonomous BCR signaling [79], which is due to interactions between two adjacent BCR molecules mediated by CDR3 recognition. This phenomenon appears to be CLL-specific and the internal epitopes involved in the BCR interaction are distinct for different IGs [80]. How this autonomous BCR stimulation plays a role in predisposing to CLL transformation or progression or integrates with classic antigenic stimulation in CLL remains to be determined. In one recently postulated scenario, classical antigen-induced BCR signaling is required for CLL cell growth and proliferation in SLO, whereas cell autonomous signals exclusively derived from the IgM-BCR ensure CLL cell survival outside of their supportive microenvironment [81]. Despite the central importance of the BCR signaling in CLL pathogenesis, it is surprising that the extensive leukemic genome sequencing efforts did not report about pathogenic variants in molecules of the upstream BCR signaling pathway that are targeted by kinase inhibitors in CLL (Fig. 2b). The notable exception are mutations detected in the BTK and PLCG2 genes identified in patients relapsing on therapy with covalent [82, 83] or non-covalent [84] BTK inhibitors.
Stereotyped BCRs in CLL
The discovery of virtually identical (stereotyped) BCR sequences in unrelated CLL patients solidified the concept of common antigens driving CLL development and progression. Approximately 30% of patients can be grouped into CLL subsets based on shared immunoglobulin sequence motifs [85], with distinct clinical behavior and functional differences in terms of BCR activation [86]. Subset #2, for instance, is associated with an aggressive clinical course and poor treatment response irrespective of the underlying IGHV mutational status, whereas subset #4 is characterized by a more indolent disease phenotype [87]. Stereotyped BCRs in CLL were first described in 2003, when cases with BCRs encoded by the IGHV3-21 gene were found to carry highly homologous or identical amino acid sequences within the variable heavy complementarity determining region 3 (VH CDR3) and exhibited highly restricted usage of the IGLV3-21 gene [88]. Soon thereafter, additional cases with stereotyped BCR IG were identified in M-CLL and U-CLL using other IGHV genes [71, 89‒91]. After these original publications, it was established that a large fraction of patients with CLL can be assigned to different subsets with distinct, stereotyped VH CDR3 sequences [85, 91‒94]: some of these subsets are more common and thus named “major” subsets [95]. Interestingly, only six IGHV genes (IGHV1-69, IGHV1-3, IGHV1-2, IGHV3-21, IGHV4-34, and IGHV4-39) account for approximately 80% of the major stereotyped subsets [93, 95]. Moreover, certain IGHV genes (e.g., IGHV3-21, IGHV1-69, IGHV1-2) are over-represented while others (e.g. IGHV3-7, IGHV3-23, IGHV3-30) are under-represented among the stereotyped fraction of CLL [93]. More recently, after sequencing approximately 30,000 patients with CLL, Agathangelidis et al. [95] reported that 41% of patients expressed stereotyped BCR IG. This is remarkable considering the extremely low probability of randomly finding identical VH CDR3 sequences in different B cells, which is in the range of 1:10−16 to 1:10−18 [96].
Chemokines and Their Receptors
CXCR4 (CD184)
CXCR4 is expressed at high levels on the surface of PB CLL cells [35], and mediates CLL cell chemotaxis, migration across vascular endothelium, actin polymerization, and migration beneath and underneath CXCL12-secreting BMSC [35]. CXCL12 also has a pro-survival effect on CLL cells [6, 16, 97] which is not surprising, given that CXCL12 initially was characterized as pre-B-cell growth-stimulating factor (PBSF) [98]. CXCR4 surface expression is regulated by its ligand CXCL12 (previously called stromal cell-derived factor-1/SDF-1) via receptor endocytosis [35], with down-regulation of surface CXCR4 on tissue CLL cells by CXCL12 present at high levels in the tissues. This characteristic can be used to distinguish tissue (lymphatic tissue- and marrow-derived) from blood CLL cells, which express low or high CXCR4 levels, respectively [2, 35]. Proliferating, Ki-67+ CLL cells from marrow and lymphatic tissue displayed significantly lower levels of CXCR4 and CXCR5 than non-proliferating CLL cells [99]. In vivo deuterium (2H) labeling of CLL cells revealed that patients with higher CXCR4 expression on their CLL cells had delayed appearance of newly produced CD38+ cells in the blood and increased the risk for lymphoid organ infiltration and poor outcome [11]. These 2H studies also revealed intraclonal heterogeneity of CXCR4 expression, with an enrichment of CLL cells expressing lower CXCR4 surface levels in the CD38+/CD5bright fraction, along with increased 2H incorporation [11]. Co-culture with BMSC down-regulates CXCR4 on CLL cells due to the receptor internalization. Interestingly, this mechanism of CXCR4 receptor endocytosis is the basis for distinguishing recently divided from older quiescent CLL cells in the PB [11] and the tissue compartments [3]. The proliferative fraction contains CLL cells which recently exited the SLO into the PB with a characteristic surface signature of CXCR4dimCD5bright cells. Vice versa, CXCR4brightCD5dim cells are a quiescent CLL cell population destined for re-circulating from PB back into the tissue compartment. Heavy water labeling demonstrated significantly increased labeling of CXCR4dimCD5bright cells in patients [3].
B-cell antigen receptor (BCR) signaling results in down-modulation of CXCR4 [100, 101], along with enhanced chemotaxis towards CXCL12 and CXCL13, at least in our hands [100]. This may explain why ZAP-70+ CLL cells display increased chemotaxis and survival in response to CXCL12 when compared to ZAP-70-negative CLL cells [97], given that ZAP-70 expression is associated with a higher responsiveness to BCR stimulation [102]. CXCR4 signaling in CLL cells is pertussis toxin-sensitive and induces calcium mobilization, activation of PI3 kinases [35], p44/42 MAP kinases [6], and serine phosphorylation of signal transducer and activator of transcription 3 (STAT3) [36]. CXCR4 signaling can be inhibited by isoform-selective PI3 kinase inhibitors [103], including idelalisib, and inhibitors of SYK [100], and BTK [60], leading to impaired CLL cell migration.
CXCR5
CXCR5 (CD185) is the receptor for CXCL13, a chemokine that regulates lymphocyte homing and positioning within lymph follicles [104]. CXCR5 is expressed by mature B cells, a small subset of T cells, and skin-derived dendritic cells (reviewed in [105]). CXCR5 gene-deleted mice have defective primary follicles and GCs in the spleen and Payer’s patches and lack inguinal lymph nodes [106]. The ligand for CXCR5 is designated CXCL13 and is constitutively secreted by stromal cells in B-cell areas of lymphoid follicles [104, 107]. CXCR5 induces recruitment of circulating naïve B cell to follicles [104, 107] and is responsible for positioning within the GC [108‒111]. In addition, it has been suggested that the primordial function of CXCL13 may be the recruitment of primitive B cells to body cavities for T-independent responses [112]. CLL and MCL cells express high levels of CXCR5 [15, 113]. Stimulation of CLL cells with CXCL13 induces actin polymerization, CXCR5 endocytosis, chemotaxis [114], and prolonged activation of MAPK (ERK 1/2). In CLL, CXCR5 signals through Gi proteins, PI3 kinases, and p44/42 MAPK pathway [15]. CXCL13 mRNA and protein is expressed by NLC in vitro and in vivo [15]. These data suggest that CXCR5 plays a role in positioning and interactions between malignant B cells and CXCL13-secreting stromal cells, such as NLC/LAM in lymphoid tissues.
Chemokines Secreted by Malignant B Cells: CCL3, CCL4
CCL3 and CCL4 are chemo-attractants for monocytes and lymphocytes [115]. CCL3 expression in normal B cells is induced by BCR triggering and CD40 ligand [41, 116, 117], and repressed by Bcl-6 [118]. Activated CLL cells express and secrete CCL3/4 [18, 22, 119] in response to BCR stimulation and in co-culture with NLC [18]. This BCR- and NLC-dependent induction of CCL3/4 is sensitive to inhibition of BCR signaling, using SYK, BTK, and PI3K inhibitors [18, 60, 100, 120]. CLL patients have elevated CCL3/4 plasma levels [18] and plasma levels of CCL3 were strongly associated with established prognostic markers and time to treatment. A multivariable analysis revealed that CCL3, advanced clinical stage, poor risk cytogenetics, and CD38 expression were independent prognostic markers in a cohort of 351 CLL patients [12]. The function of CCL3/4 in CLL remains poorly defined, but based upon the function of B-cell-derived CCL3/4 in normal immune responses, increased CCL3/4 secretion by CLL cells may induce trafficking and homing of accessory cells, particularly of T cells and monocytes to CLL cells in the tissue microenvironments [18, 121].
VLA-4 (CD49d) Adhesion Molecules in CLL
Integrins are a superfamily of heterodimeric glycoproteins, consisting of various α (1 through 11) and β (1 through 6) subunits, whose function is to mediate cell-cell and cell-matrix adhesion in various cell types. The term “integrin” was first proposed in 1986 to describe membrane complexes involved in the transmembrane association between fibronectin (FN) as part of the ECM and the actin cytoskeleton [122]. Integrins are categorized into subfamilies with members sharing a common β subunit pairing with a unique α subunit. β1 integrins are very late activation antigens (VLA) that have the same β1 subunit but various α chains (α 1 through 6) [123]. The α4β1 integrin VLA-4 (CD49d) is a receptor for FN and vascular cell adhesion molecule-1 (VCAM-1/CD106). VLA-4 is expressed on lymphocytes, monocytes, and most other hematopoietic cells (except for neutrophils); VLA-4 is involved in both cell-cell and cell-ECM adhesion and plays a role in lymphocyte trafficking and homing as part of immune surveillance [124], trafficking and homing of other hematopoietic cells, and inflammation. Integrins are highly versatile adhesion molecules; their adhesiveness can rapidly be regulated by the cells on which they are expressed, for example, by chemokine receptor activation [125]. VLA-4 mediates lymphocyte adhesion to the VCAM-1 (CD106), which is expressed on cytokine-activated endothelium. VCAM-1 mediates leukocyte-endothelial cell adhesion and may play a role in the development of artherosclerosis and rheumatoid arthritis. VLA-4 also binds FN, an ECM component expressed on MSC [126], by interacting with at least three FN sites: CS-1 and REDV in the IIICS region and H1 in the HepII region [127]. VLA-4 plays a particularly important role for interactions between normal and malignant hematopoietic cells and the marrow microenvironment. VLA-4 integrins cooperate with chemokine receptors in CLL cell adhesion to stromal cells [35, 128], they cooperate with CD38 [129], and their function can be inhibited by the BTK inhibitor ibrutinib [130]. Moreover, VLA-4 expression on CLL cells has a prognostic impact [13, 131], indicating the relevance of these interactions in vivo. Collectively, these studies indicate that VLA-4 integrins play a key role for adhesion of CLL and other leukemia cells to stromal cells and ECM, and provide a rationale to further explore and target this molecule in CLL.
Targeted Therapies in CLL and Their Relation to the Microenvironment
Normal and malignant B cells contain cytoplasmic kinases that are essential for transmitting signals from surface receptors into the cell, thereby regulating B-cell survival and proliferation. These include SYK, Bruton’s tyrosine kinase (BTK) and PI3 kinase isoforms, especially the delta isoform (PI3Kδ). These kinases are closely related to BCR signaling, as well as the signaling of B-cell chemokine receptors and adhesion molecules, which regulate the trafficking and homing of B lymphocytes. The SYK inhibitor fostamatinib was the first targeted agent to be tested in a clinical trial in patients with different B-cell malignancies including CLL patients [132]. Shortly thereafter, other inhibitors that target BCR signaling, especially BTK became available and were found to be highly effective for treating CLL patients. Building on basic and translational research that emphasized the importance of the BCR [133] and the CLL microenvironment [134] in CLL disease biology, we now can look back at 2 decades of transformative research in CLL, resulting in the approval of drugs that are targeting these key disease pathways. Within a short time, the BTK inhibitors have overturned CLL treatment paradigms, replacing chemotherapy-based treatments for patients with CLL. The efficacy of these kinase inhibitors in CLL is related to a high dependency of CLL cells on external signals derived from the microenvironment, especially the BCR. These developments demonstrate that CLL basic and translational research has come full circle, from studying the CLL microenvironment and the role of the BCR in CLL pathogenesis to effective targeted therapies replacing chemo-(immuno) therapy.
Conclusions
CLL is a paradigmatic malignancy in which a delicate balance between cell-extrinsic triggers from the tissue microenvironment and cell-intrinsic aberrations regulates the growth of the malignant B cells, which in turn determines the clinical behavior of individual patients. Signals from the microenvironment of secondary lymphoid tissues result in downstream signaling that promotes CLL cell growth. Among the surface receptors initiating these signaling cascades, the BCR has been shown to be central for the growth of the malignant B cells. The BCR structure is finely tuned by somatic hypermutation and dictates the signaling response of the BCR in individual patients, which appears to be the basis for the heterogenous clinical behavior in patients with CLL. Moreover, the BCR is key to treating CLL, as shown by the impressive clinical results in patients treated with kinase inhibitors targeting BCR signaling, such as the BTK inhibitors, which have become the first choice of therapy for many patients with CLL. The BCR appears to be the central driver of CLL disease progression, which is supported by the markedly different clinical outcomes of patients with CLL displaying a similar background of genomic aberrations, but different IGHV mutation statuses (M-CLL vs. U-CLL), and the remarkable therapeutic efficacy of BTK inhibitors even in patients with major cell cycle de-regulation due to cell-intrinsic aberrations (e.g., TP53 or ATM defects), underscoring the critical dependence of CLL cells on the BCR and the tissue microenvironment.
Acknowledgments
We apologize that we were not able to discuss and cite a number of additional studies of other investigators that are related to CXCR4 in neoplastic diseases because of space limitation.
Conflict of Interest Statement
The authors have no conflict of interest related to this review article.
Funding Sources
J.A.B. is supported by MD Anderson’s Moon Shot Program in CLL and in part by the MD Anderson Cancer Center Support Grant CA016672.
Author Contributions
S.K. and J.A.B. wrote the manuscript and designed the figures together.