Molecular pathogenesis of chronic lymphocytic leukemia (original) (raw)
Karyotypic investigations revealed the association of CLL with del13q14, trisomy 12, del11q22-q23, and del17p13 (refs. 1, 2, 49, 50, and Figure 2). The advent of next-generation sequencing (NGS) technologies, coupled with gene copy number analysis, has enabled exploration of the CLL genome (45, 46, 51, 52), uncovering genetic lesions that recurrently target this leukemia (Figure 2).
Genetic lesions of CLL at different phases of the disease. The frequency (in parentheses) of genetic lesions is shown for CLL at presentation and for two different types of CLL progression: chemorefractoriness without evidence of histologic transformation, and histologic transformation to RS. The two types of CLL progression follow distinct molecular pathways in terms of type and frequency of genetic lesions. Genetic lesions that occur at CLL presentation but are not enriched at CLL progression are indicated in gray.
The CLL genome is largely devoid of the chromosomal translocations and aberrant somatic hypermutations that are involved in several B cell non-Hodgkin lymphomas (B-NHLs) (1, 2, 49, 50, 53). These observation are consistent with a post-GC or GC-independent derivation of CLL because B-NHL–associated translocations are caused by errors during somatic hypermutation and class switch recombination, two mechanisms that are active in GC B cells (23).
NGS studies have further elucidated the genomic complexity of CLLs and have shown that the average number of non-silent mutations (i.e., mutations that alter the protein sequence) per case is 10–12 at diagnosis, whereas the average number of copy number abnormalities is approximately two (45, 46, 51, 52). The order of magnitude of lesions detected in the coding genome of CLL appears considerably lower than in common epithelial cancers (54). Among hematologic malignancies, the complexity of the CLL genome is also on the lower side, markedly smaller than DLBCL and multiple myeloma, and comparable to some acute leukemias (55–58).
Del13q14 is the most frequent alteration and occurs in 50%–60% of cases (2, 49, 50). Because this lesion is found at a similar frequency in MBL and is often detectable as a single lesion (11, 13), this alteration may represent an early event in the disease. The minimal deleted region (MDR) of del13q14 contains the deleted in leukemia-1 (DLEU1) and DLEU2 genes, which code for noncoding transcripts, as well as the MIR15A/MIR16A microRNAs, which map within intron 4 of DLEU2 (59, 60). Expression of DLEU2 and MIR15A/MIR16A is also downregulated in CLL in the absence of the del13q14 deletion, possibly as a consequence of epigenetic mechanisms (61, 62). The DLEU2 locus generates a long non-coding RNA that has only one known function, to serve as the primary RNA from which the mature miR-15a/miR-16a microRNAs are processed (59–61, 63). Based on animal models, MIR15A/MIR16A are considered the main candidate tumor suppressor genes involved with del13q14, although a role is likely for additional genetic elements including DLEU2 (refs. 64, 65, and see below). In normal cells, these two microRNAs appear to inhibit expression of multiple genes, including BCL2, the cyclins CCND1 and CCND3, and cyclin-dependent kinase 6 (CDK6) (66–71). Consequently, deletion of MIR15A/MIR16A abrogates this inhibitory effect and favors the constitutive cycling of B cells. In addition, miR-15a/miR-16a may also participate in a microRNA/tumor protein p53 (microRNA/TP53) feedback circuitry associated with CLL pathogenesis and prognosis (72).
Trisomy 12 occurs in approximately 15% of CLL cases and is thought to alter the gene dosage of one or more proto-oncogenes, although the precise molecular mechanism of this alteration is not currently understood (2, 49). Del11q22-q23 in most cases affects the ataxia telangiectasia mutated (ATM) gene, the deficiency of which causes genomic instability and, if inherited in the germline, a predisposition for lymphoid malignancies (73). In CLL, ATM is also affected by mutations that disrupt its function (74–77). In a fraction of cases carrying del11q22-q23, the ATM gene appears to be intact and the deletion targets the BIRC3 gene, which is in close proximity to ATM (ref. 78 and see below).
In approximately 5%–10% of untreated CLL patients, del17p13 disrupts the TP53 tumor suppressor gene (49, 79–87). Many cases of CLL with del17p13 display inactivation of the second TP53 allele by point mutation (80–87). Altered TP53 function due to 17p deletion and/or TP53 gene mutation is an important predictor of chemorefractoriness (see below) and is associated with reduced survival (81–87). From a diagnostic standpoint, it should be noted that a fraction of CLLs display TP53 mutations in the absence of del17p13. Such cases are not recognized by FISH analysis, and their identification requires mutation analysis of the TP53 gene (88).
Recently, whole exome sequencing studies have revealed recurrent genetic lesions that affect genes implicated in different biological pathways of potential pathogenetic relevance for CLL (45, 46, 51, 52, 89). These genes include NOTCH1, splicing factor 3b, subunit 1 (SF3B1), BIRC3, and myeloid differentiation primary response gene 88 (MYD88).
Mutational activation of the Notch pathway. NOTCH1 encodes a ligand-activated transcription factor that regulates several downstream pathways important for cell growth control and is affected by activating mutations in 60% of T-lineage acute lymphoblastic leukemias (90). In CLL, the frequency of NOTCH1 mutations at the time of diagnosis is approximately 10% (45–47, 91). Conversely, mutations of NOTCH1 are very rare in MBL (92). NOTCH1 mutations at CLL diagnosis preferentially occur among U-CLLs and cluster with trisomy 12 (45–47, 93, 94). NOTCH1 mutations identify a high-risk subgroup of patients showing poor survival comparable to that associated with TP53 abnormalities and exert a prognostic role that is independent of widely accepted risk factors, as confirmed in multiple consecutive series from different institutions (45–47).
NOTCH1 mutations in CLL are highly consistent among cases and disrupt the C-terminal PEST domain that normally limits the intensity and duration of NOTCH1 signaling (45–47). Removal of the PEST domain Results in impaired NOTCH1 degradation and accumulation of an active NOTCH1 isoform (90). One recurrent mutation (c.7544_7545delCT) accounts for approximately 80% of all NOTCH1 mutations (45–47) and can be rapidly detected by a simple PCR-based strategy, providing a potential approach for a first-level screening of NOTCH1 alterations (47). Although the precise role of NOTCH1 activation in CLL pathogenesis is still under scrutiny, the relevance of these mutations is highlighted by the distinctive GEP signature of CLL carrying NOTCH1 mutations (46, 94).
Mutations of the SF3B1 gene. SF3B1 is a critical component of both major (U2-like) and minor (U12-like) spliceosomes, which enact the precise excision of introns from pre-mRNA (95). At diagnosis, SF3B1 is mutated in 5%–10% of CLLs (51, 52, 89). Conversely, SF3B1 mutations are exceptional in MBL (96). The pathogenicity of SF3B1 mutations in CLL is supported by the clustering of these mutations in evolutionarily conserved hotspots localized within HEAT domains (51, 52, 89, 97). Although the precise mechanistic aspects of SF3B1 mutations are still under investigation, the observation that SF3B1 regulates the alternative splicing program of genes controlling cell cycle progression and apoptosis points to a potential contribution of SF3B1 mutations in modulating tumor cell proliferation and survival (51, 52). Across the spectrum of mature B cell neoplasias, mutations of SF3B1 are selectively restricted to CLL, whereas these same mutations are highly recurrent in myelodysplastic syndromes (98). At CLL diagnosis, SF3B1 mutations predict reduced survival independent of other clinical and biological risk factors (51, 52, 89).
BIRC3 gene inactivation. BIRC3, along with TRAF2 and TRAF3, cooperates in the same protein complex that negatively regulates MAP3K14, an activator of the non-canonical pathway of NF-κB signaling (99). All BIRC3 mutations in CLL are predicted to disrupt the C-terminal RING domain, which is essential for proteasomal degradation of MAP3K14 by BIRC3 (78). Consistently, CLLs harboring BIRC3 disruption display constitutive NF-κB activation (78). At CLL diagnosis, BIRC3 disruption associates with unfavorable clinical and genetic features and predicts poor outcome independent of other risk factors (78). Among B cell neoplasms, mutations of BIRC3 appear to be specific for CLL and for splenic marginal zone lymphoma (78, 100).
Other mutations revealed by NGS studies. Whole exome sequencing studies have revealed mutations of several other genes in CLL, albeit at low frequency (45, 46, 51, 52). Of these, MYD88 is mutated in 3%–5% of CLL at diagnosis (46). MYD88 is a critical adaptor molecule of the TLR complex that is also mutated in other B cell malignancies, namely DLBCL and marginal zone lymphoma (100, 101). In CLL, mutations of MYD88 target specific hotspots also implicated in other B cell malignancies (46, 101, 102).