Why is the Immunoglobulin Heavy Chain Gene Mutation Status a Prognostic Indicator in Chronic Lymphocytic Leukemia?

Abstract

The immunoglobulin heavy chain gene (IgHV) mutation status correlates with the clinical outcome of patients with chronic lymphocytic leukemia (CLL) treated with chemoimmunotherapy. Why the survival rate of patients with unmutated IgHV is worse than that of patients with mutated IgHV is unknown. CLL cells with unmutated IgHV were thought to originate from naïve B lymphocytes whereas CLL cells with mutated IgHV were thought to arise from B cells that have undergone somatic hypermutation (SHM). Cell surface protein expression profile and gene expression studies showing that all CLL cells, regardless of their IgHV mutation status, are of post-germinal center origin, negated this hypothesis. We hereby propose that all CLL cells undergo SHM and their proliferation rate determines their IgHV mutation status. DNA breaks, accumulated during SHM, are restored by various DNA repair mechanisms. In rapidly dividing cells DNA breaks are repaired by the efficient high-fidelity homology-directed DNA repair apparatus whereas in slowly dividing cells by the inefficient low-fidelity non-homology end-joining repair mechanism. Accordingly, a low IgHV mutation rate is found in rapidly dividing cells whereas a high mutation rate is typically found in slowly dividing cells. Thus, CLL cells’ proliferation rate determines the IgHV mutation status and patients’ clinical outcome.

B-cell chronic lymphocytic leukemia (CLL) is characterized by the accumulation of leukemic lymphocytes that co-express CD5, CD19, and CD23. The clinical course of CLL is extremely variable. One-third of CLL patients experience an indolent disease, with a life expectancy similar to that of age-matched healthy individuals. Other patients have a benign course lasting for 5 to 10 years after which the disease rapidly progresses, whereas in approximately 10% of patients have an aggressive disease, with a median survival of 1 to 3 years from the time of diagnosis [1].

The immunoglobulin heavy chain variable region gene (IgHV) mutation status has been an established prognostic factor in patients with CLL. A cutoff of 2% deviation or 98% sequence identity to germline in IgHV sequence has been routinely used as a prognostic marker [2]. Several studies demonstrated that patients with mutated IgVH respond better to chemoimmunotherapy [3], whereas patients with unmutated IgHV have shorter time to treatment, shorter time to next treatment, inferior response to chemotherapy/ chemoimmunotherapy, higher chemotherapy resistance rate, and lower survival rate [4-6]. Therefore, although the significance of IgHV mutation status in patients treated with novel agent is less clear [7], the new guidelines from the international workshop on chronic lymphocytic leukemia (iwCLL) recommend testing for IgHV mutation status at baseline in all patients with CLL [8].

Although more than 20 years have passed since this observation was first reported, it is still unknown why the the IgHV mutation status predicts outcome of CLL patients treated with chemoimmunotherapy.

It was proposed that CLL is not one disease but two separate entities that arise from B cells at different stages of maturation [9]. Because the genomic arrangement of CLL with unmutated IgHV is similar to that of naïve B cells that undergo somatic recombination, it was postulated that IgHV-unmutated CLL cells arise from naïve, pre-germinal center B cells and that CLL cells with mutated IgHV originate from post-germinal center normal B cells that underwent somatic hypermutation (SHM) after antigen recognition [5,10].

However, recent data do not support this hypothesis. A retrospective analysis of the survival and progression-free survival of 535 CLL patients treated with chemoimmunotherapy showed that the percent deviation of the immunoglobulin heavy chain gene variable region sequence from the germline sequence (IgHV%) is a continuous variable. CLL patients’ survival correlated with a continuous increase of IgHV% deviation. The higher the percent deviation, the better was treatment outcome, suggesting that IgHV% is a continuous, not a dichotomized prognostic variable [11]. Furthermore, the CLL cell gene profile is independent of the IgVH mutation status. It is closely related to that of post-germinal center memory B cells [12] and is similar to the gene profile of CD5+ normal B cells [13]. In addition, CLL cells from all patients, regardless of IgHV mutation status, express CD27, a surface protein that is expressed on memory B cells [14]. Also, both IgHV-mutated and -unmutated CLL cells produce a functional activation-induced deaminase (AID), an enzyme that plays a key role in SHM in normal B lymphocytes [15].

Because normal memory B cells are phenotypically similar to CLL cells and because the SHM molecular apparatus is present both in IgHV-mutated and -unmutated CLL cells, it is reasonable to assume that, similar to normal antigen-experienced memory B cells, all CLL cells undergo SHM. However, unlike in normal memory B cells, in CLL cells the IgHV gene is “mutated” in only half of the patients. In normal B cells, SHM is a two-step process. The first step is gene transcription, during which a cytosine of single-stranded DNA undergoes deamination by AID and transforms to uracil. The second step is DNA replication, in which the DNA replication machinery recognizes uracil as thymidine and the original C:G base pair is converted to T:A [16]. Deamination, followed by mutation, occurs frequently, even in the absence of AID, and transcription errors are usually corrected in non-germinal center B cells by the base excision repair mechanism using DNA polymerase I. However, the DNA of germinal center B cells is repaired by the error-prone Polɳ; as a result, IgHV gene mutations occur at a million-fold higher rate than do spontaneous mutations in other somatic cells [16]. Because hypermutated DNA segments are fragile and tend to break [17], double-strand DNA breaks (DSBs), which accumulate during SHM near the newly formed hypermutated segments [18], activate repair mechanisms.

DSBs repair mechanisms are cell cycle-dependent and largely divided into two types: homology-directed repair and non-homology end-joining (NHEJ) repair [19]. Homology-directed repair requires a template of a homologous chromosome and operates only in dividing cells during the S/G2 phase of the cell cycle. Homology-directed repair is a high-fidelity mechanism that restores DNA continuity and corrects DNA sequence mutations that were introduced during SHM [19]. In proliferating cells that transition through the S/G2 phase of the cell cycle, a template for homology directed-repair is readily available and constantly utilized. CLL cells that proliferate at a relatively high rate utilize this mechanism more frequently than CLL cells that proliferate at a low rate or do not proliferate at all. This high-fidelity DNA repair mechanism eliminates SHM and, as a result, the level of mutated IgHV will be low and will be defined as “unmutated”. The low-fidelity NHEJ repair mechanism uses a specialized ligase to join two DNA ends. This mechanism does not require a template and operates throughout the cell cycle [19]. The NHEJ DNA damage-repair apparatus, recruited mainly in non- or slowly proliferating cells, does not repair DNA mutations that occur during SHM and introduces additional mutations by not correcting the loss of damaged nucleotides at the DNA break site [20]. Therefore, in non- or slowly proliferating CLL cells, in which DSBs are repaired by NHEJ, the level of mutated IgHV will be high and defined as “mutated”.

Thus, the IgHV mutation status of CLL cells is a result of a physiological recruitment of different DNA repair mechanisms during SHM. In proliferating CLL cells, high-fidelity homology-directed repair reduces IgHV mutations to the level of “unmutated”, whereas in non- or slowly-proliferating CLL cells, low-fidelity NHEJ repair does not reduce or even increases IgHV mutation rate resulting in “mutated” IgHV, as is commonly found in normal memory B cells (Figure 1).

Figure 1.

The proposed model of IgHV mutation repair: A. CLL cells proliferate at various rates illustrated here by a proliferation axis. B. CLL cells originating from the post-germinal center undergo somatic hypermutation catalyzed by activation induced deaminase (AID). A similar IgHV mutation rate (pink dotes) occurs in all DNA strands irrespective of their position along the proliferation axis. C. Two different mechanisms are recruited to repair the DNA breaks: the non-homology end-joining (NHEJ) repair, recruited by non- or slowly-proliferating cells and the homology-directed repair (HDR), recruited in rapidly proliferating cells. Because HDR is operative in rapidly proliferating cells, the higher the proliferation rate the more frequently HDR is used. D. In non- or slowly-proliferating CLL cells, the low-fidelity NHEJ repair mechanism does not reduce or even increases IgHV mutation rate. Conversely, in rapidly proliferating CLL cells, the high-fidelity HDR efficiently reduces IgHV mutation rate. E. Accumulation of IgHV mutations along the proliferation axis depends on the recruitment of NHEJ Vs. HDR.

Recent studies support the hypothesis that the IgHV mutation status is cause by the CLL cell proliferation rate. Expanded proliferation centers, consisting of cells with a high proliferation rate, were characteristically found in lymph node biopsies of patients with unmutated but not mutated IgHV [21]. In addition, comprehensive heavy water labeling studies, measuring the proliferation rate of CLL cells, demonstrated an statistically significant association between increased proliferation rate and decreased IgVH mutation status [22].

We hereby propose that CLL cells’ IgHV gene mutation rate is determined by the type of the DNA repair mechanism utilized during SHM. In CLL cells that proliferated at a high rate, a high-utility repair mechanism is engaged; as a result, the IgHV mutation level is low (or “unmutated”). Conversely, in CLL cells that proliferated at a low rate, a low-utility repair mechanism is engaged; as a result, the IgHV mutation level is high (or “mutated”). Thus, the IgHV mutation status does not directly affect the disease course or treatment outcome it rather depends on proliferation rate of CLL cells and.