Genetic Risk Variants for Class Switching Recombination Defects in Ataxia-Telangiectasia Patients

Ataxia-telangiectasia (A-T), also known as Louis-Bar syndrome (OMIM #208,900), is an autosomal recessive disorder caused by mutations in the ataxia telangiectasia mutated (ATM) gene encoding a serine/threonine-protein kinase (ATM) [1, 2]. A-T patients exhibit a broad range of clinical manifestations, including progressive cerebellar ataxia, oculocutaneous telangiectasia, variable immunodeficiency, radiosensitivity, and susceptibility to malignancies [3, 4]. Other phenotypes such as infections, pulmonary diseases, insulin-resistant diabetes, growth failure, gonadal atrophy, cutaneous abnormality, and metabolic and cardiovascular disease have also been reported in these patients [3, 5‐9]. The ATM protein plays a major role in DNA double-strand breaks (DSB) repair, cell cycle regulation, and genomic stability [10, 11]. Furthermore, ATM plays important role in B and T cell development (particularly in antigen receptor rearrangement) and class switching recombination (CSR) in mature B cells [12, 13].

Overall, A-T patients manifest significantly variable clinical and immunological features without genotype–phenotype correlation, involving modifying factors. Based on serum immunoglobulins (Ig) profile, patients with A-T could be assigned to one of the following subgroups: normal Ig level, IgA deficiency, hypogammaglobulinemia, and hyper IgM (HIgM) phenotype known as Ig CSR defect (CSR-D) [14‐17]. The most frequent immunodeficiencies in A-T individuals are related to IgG2 and IgA deficiencies [15, 18]. On the other hand, a minority of A-T patients present HIgM phenotype, who manifest low switched Igs (IgG, IgA, and IgE) with normal or increased IgM [19, 20]. In our previous study performed in 2017, we showed that about 20% of A-T patients show Ig CSR-D [21]. Generally, A-T patients with HIgM experience a more severe course of the disease leading to a lower quality of life and shorter survival [22].

Some previous studies have confirmed the role of ATM in the CSR mechanism [12, 13, 23]; however, the causative pathogenesis of CSR-D phenotype compared to patients with normal CSR (CSR-N) in A-T patients remains unclear. T cell abnormality and absence of germinal center activation due to cellular defect has been proposed, which was failed when compared between CSR-D and CSR-N A-T patients [24]. It has been hypothesized that the type or the location of ATM mutations may be the cause of CSR defect in some A-T patients, but the observation of different CSR phenotypes in patients with the same mutations falsified this anticipation [16]. On the other hand, other genetic factors could be involved in CSR-D in A-T patients. Nevertheless, no data is available to determine the molecular level on the modification of ATM activity by other signaling proteins. Towards a better understanding of the phenomenon of CSR, we classified our A-T patients into two groups based on CSR status and compared the genotype of the two groups by whole-exome sequencing (WES). In this study, for the first time, we investigated variations in genes other than ATM that might be attributed to CSR-D phenotype in A-T patients. The majority of the variants we found have known roles in the CSR mechanism, suggesting them as potential candidates for further investigation in the future.

The clinical features of A-T are complex and multi-systemic, including neurological abnormalities, oculocutaneous telangiectasia, recurrent infections, immunodeficiencies, and susceptibility to cancers [3]. In this study, all A-T patients exhibited ataxia and telangiectasia as the main clinical features, but there was no significant difference between the onsets of these manifestations in the two groups. Moreover, we did not observe a significant difference between the onsets of other clinical manifestations in the two groups. Recurrent infections are the most common manifestation associated with immunodeficiency in A-T patients and a major factor for early age morbidity and mortality [5, 15, 36]. We found significantly increased episodes of infections (especially respiratory tract infections) in the CSR-D group compared to the CSR-N group. Since environmental factors should be considered the main modifying factor, the female frequency was higher in previously reported cases with CSR-D compared to CSR-N [21, 22, 37], as was observed in the current assay as well.

A-T patients show variable cellular and humoral immune abnormalities [38]. During the last decade, several A-T cases have been reported in whom Ig CSR defect has been implicated [16, 19, 39, 40]. Some previous studies reported that about 10% of the A-T patients present with the HIgM/CSR-D phenotype [19, 41]. In contrast, our previous study showed that the frequency of CSR defect in Iranian A-T patients was higher than in other populations (21.2%) [21]. Previous studies have shown that the CSR junctions in cells of A-T patients are aberrant, indicating a role for ATM in the final steps of CSR, including DNA end modification, repair, and joining [42, 43], which may suggest ATM as a player in the CSR process. However, the majority of A-T patients with ATM mutations do not demonstrate CSR defects. Thus, the cause of this Ig profile in A-T is not entirely understood. The current study supports the notion that A-T patients with hyper IgM level, normal T cell subsets, low ability to IgE switching with stimulation of CD40L and IL-4, and having abrogated IgA memory B-cells or plasmablasts in their periphery can be classified as CSR defects; however, other hypothetical mechanisms including selective apoptosis of IgA, IgG and IgE but not IgM would be an alternative mechanism for this phenomenon (A-T patients with Ig deficiency).

Comparing the genotype of CSR-D and CSR-N patients, in the current study, we proposed a new understanding of the abovementioned immunological defect. First, we rejected the hypothesis that the type, zygosity, or the location (affected domain) of ATM mutations may be the cause of CSR-D in A-T patients [16], as we observed similar mutation distributions in the two groups (Table S1 and Figure S4). Moreover, evaluation of the exomes of the two groups in an unsupervised manner revealed 1645 variants with significant allelic differences between the CSR-D and CSR-N groups. These variants represented enrichment terms such as antigen processing and presentation pathway, plasma membrane, autoimmunity, allograft rejection, graft-versus-host disease, and malignancy.

We found several variants in the antigen processing pathway to be associated with the CSR phenotype under study. Antigen processing and presentation is a complex process in which many molecules and proteins are involved [44]. Once a B-cell receptor (BCR) recognizes a particular antigen, proteasomes degrade the antigen, and subsequently, peptide fragments were presented at the cell surface through MHC class II molecules [45]. MHC class II molecules are highly polymorphic and normally expressed only on professional antigen-presenting cells such as B cells, dendritic cells, and mononuclear phagocytes. CD4⁺ T cells specific for this antigen initiate a cascade allowing the cognate B cell activation. In particular, one of the most important interactions for humoral immune responses is the engagement of CD40 molecule on B cells to the CD40 ligand on follicular helper T cells [45, 46]. At this point, the activated B cells can either differentiate into plasmablasts or get recruited into a specialized region, called germinal centers (GCs) [47]. In the GC, B cells are targeted by clonal expansion, somatic hypermutation, affinity maturation, and CSR, eventually forming antibody-secreting plasma cells [48, 49]. It has been previously suggested that antigen processing and presentation are indirectly related to the quantity and quality of Ig class switching. For instance, patients with CD40/L deficiencies, known as HIgM syndrome, display an impaired production of IgG, IgA, IgE, and normal or elevated levels of IgM [50]. It would therefore not be surprising to find other genes in this pathway particularly HLA-DRB5 to play a role in the CSR mechanism.

Several studies have reported that A-T patients with CSR-D present with a more severe course of the disease leading to a lower quality of life at earlier ages and shorter survival than other A-T patients [21, 40, 51]. Moreover, A-T is a genomic instability syndrome leading to an extremely high incidence of malignancies (10–25%) [52‐54]. Leukemia and lymphoma account for 85% of all malignancies in A-T patients in childhood [55]. However, adults are susceptible to both lymphoid tumors and various types of solid tumors including breast, liver, gastric, and esophageal carcinomas [56]. Based on our results, it seems that A-T patients with CSR-D harbor additional genomic variants mainly associated with various solid tumors such as kidney, liver, melanoma, breast, and endometrial cancers.

We also identified several variants in the selected DSBs repair pathway with association with the CSR phenotype in our study. DSBs are potentially lethal lesions occurring as a result of exposure to exogenous agents such as radiation and certain chemicals [10]. DSBs also occur as intermediates in various biological events, such as V(D)J recombination and efficient CSR [57, 58]. The most common pathways used to repair DSBs are non-homologous end joining (NHEJ) and homologous recombination (HR) [59]. Generally, an early event during the DSB response is the activation of ATM protein, leading to rapid phosphorylation of several proteins involved in DNA repair, cell cycle checkpoint, and transcription regulation. Based on the importance of the DSB repair pathway in CSR and the defect of this pathway in A-T patients, we evaluated genes that are involved in DSB repair pathways, and we observed multiple variants significantly different between the two A-T groups. Remarkably, among these variants, p.T241M variant of the XRCC3 gene, the coding protein involved in homology-directed repair, considered to be a risk factor observed exclusively in the CSR-D group, and in every single case in this group. Indeed, NHEJ and alternative end-joining (A-EJ, using homology-directed repair) are the main pathways involved in the repair of CSR breaks. However, some findings demonstrate that although AID-induced breaks are repaired primarily in the G1 checkpoint by the NHEJ pathway, Igh DSBs that escape repair or have defects in NHEJ can persist into the S phase, where they are considerably resected and become substrates for homology-directed repair using microhomology in the S regions (A-EJ) [60‐62]. It seems that A-EJ-mediated repair of IGH breaks that failed NHEJ-mediated CSR attempts would restore an intact IGH allele for the next round of AID targeting and CSR. In fact, these findings suggest that A-EJ contributes to the repair of CSR-related DSBs [60‐63]. The main component involved in A-EJ is usually XRCC1 to recruit LIG3; however, recent studies observed XRCC1 independent microhomology-mediated A-EJ with a tight connection of PARP1 and XRCC3 [64, 65]. On the other hand, a few studies reported a role for HR in proliferation and genome stability in early B cell development [66, 67]. Caddle et al. [66] have demonstrated that HR, with a major role of XRCC3, is essential for the promotion of lymphocyte differentiation or maturation. The study showed that the functions of XRCC2, a homolog member of the RECA/RAD51-related protein family that participates in HR, in early B cell development seem to differ from its roles in mature and activated B cells. Indeed, defective HR leads to the accumulation of AID-induced DSBs at both IGH and non-IGH loci suggesting that high fidelity repair of AID-inflicted breaks is required for the B cell genome integrity [60, 68]. These were the first results to implicate HR as an important pathway with a defined role in the adaptive immune system. In fact, XRCC2 was transcriptionally upregulated after B cell activation [67]. Therefore, it seems that there is an interrelationship among B cell activation, immunoglobulin class switching, and HR which is more essential in the context of NHEJ monogenic defects. In our study, we found a relationship between XRCC3, as another member of the RECA/RAD51-related protein family, and class switching recombination. Previous studies in cutaneous malignant melanoma and head and neck cancer have also reported this variant as a potential risk factor in DSB repair [69, 70]. Thus, it is postulated that, along with ATM mutations, this variant of the XRCC3 gene plays an important role in CSR defect in A-T patients, which needs to be proved by functional studies in the future. In general, possible roles for homologous recombination, in either normal B cells development or immunodeficiency, remain controversial.

In case–control genetic studies, aggregation analysis is often used as a suitable method to identify genes associated with diseases of interest, even if variants found within them are heterogeneous in nature and position. We found that some variants of two genes were involved in DSB repair, and seven genes related to antigen presentation were significantly different between our two study groups. This highlights the importance of these two mechanisms in CSR. Overall, it seems that the variants of FANCM and MDC1 genes are highly important due to their effect on normal CSR mechanisms. Furthermore, our findings confirm the importance of variants of seven genes involved in antigen processing and presentation, especially HLA-DRB5, in normal CSR. However, it is not clear which genetic variant has larger effects on the CSR mechanism, and further studies are required to elucidate this question with a higher sample size. Apparently, important causal variants in the CSR mechanism cannot be identified until functional validation assays are performed. Our study calls for further investigations of the effect of the identified variants involved in DSB response and antigen processing and presentation pathways at functional levels in A-T patients. Moreover, evaluation of other CSR defect diseases in the same pathway including MRE11 and NBN deficiencies and their severity may empower this observation in future studies.

To date, no evidence is available at the molecular level on potential modification of ATM activity by other signaling proteins, and interaction partners of the ATM protein are not completely recognized [71]. A comprehensive understanding of this field is required for characterizing the pathogenesis of A-T and other ATM-related diseases such as cancer. Research in this area offers a new horizon to increase our knowledge regarding ATM signaling and phenotypic diversity of patients, and perhaps these findings would be helpful in the management and prognostic estimation of the disease.

Given that similar mutations in the ATM gene result in different clinical phenotypes, including different immunological profiles in A-T patients, additional genetic alternations are thought to play important roles in A-T disease outcomes. In the present study, the relationship between the genotype of A-T patients and the CSR defect phenotype was investigated for the first time. Our findings showed that in addition to the ATM gene variants, variants in genes related to this process could help explain CSR defects in A-T patients. Further research at the functional level is required to complete and confirm the findings conclusively.