Non-canonical functions of lymphoid-specific proteins AID and RAG1: Understanding mechanism, regulation, and implications in genomic instability

Abstract

Recombination-activating genes (RAGs) and Activation-induced cytidine deaminase (AID) are lymphoid cell-specific proteins that play crucial roles in the development of adaptive immunity. Diversification of the immune system is initiated by a carefully orchestrated DNA rearrangement process, V(D)J recombination, in the Immunoglobulins (Ig) and T-cell receptor (TCR) genes during lymphocyte development. The lymphoid-specific recombinases, RAG1 and RAG2, mediates the process of V(D)J recombination to generate antigen receptor diversity. All B lymphocytes that successfully complete V(D)J recombination express IgM on surface and migrate to the secondary lymphoid organs, where they are activated after encountering antigens. Activated B cells further undergo other genetic alterations, namely somatic hypermutation (SHM), class-switch recombination (CSR) and gene conversion (GC), carried out by a B cell-specific enzyme: Activation-induced cytidine deaminase to further expand the immunoglobulin repertoire. On the other hand, the development, differentiation, and maturation of T lymphocytes are known to be dependent on transcription factors like GATA-3, c-MYB, members of the E2A/HEB family, BCL11B etc. Among these, BCL11B is a key transcription factor known to play a role in the survival and differentiation of T cells during early stages of T cell development in both mice and humans. Deletion of Bcl11b confers natural-killer cell-like properties to T cells, resulting in increased proliferation and survival of immature T lymphocytes. Several studies have reported that BCL11B is very frequently mutated in human T-cell acute lymphoblastic leukemia (T-ALL) patients. These mutations which are significantly more frequent in exon 4 of BCL11B disrupt the structure of the zinc finger domains and prevent the binding of the transcription factor to DNA, blocking the transition from Double negative 2-Double negative 3 (DN2-DN3) stage of T cell development. However, the mechanism behind the generation of such mutation hotspots in BCL11B is yet to be deciphered. In the first part of the study, we have investigated the molecular mechanism of fragility of BCL11B in T-ALL. The high frequency of C to T or G to A conversion at the AID-hotspot motifs (DGWY/WRCH/WRC) in the deregulated gene prompted us to investigate the role of the enzyme, AID behind the BCL11B gene mutagenesis. Although AID primarily involves in somatic hypermutation and class-switch recombination of the Ig genes in activated B cells, its aberrant expression is reported across B cell lymphomas/leukemias and T cell lymphomas, suggesting that AID could play a role in tumorigenesis. Furthermore, AID transgenic mice spontaneously develop T cell lymphomas and micro-adenomas in the lung and harbour frequent point mutations in the TCR and c-MYC genes. In the present study, we show that AID is aberrantly expressed in T-ALL patients and cell lines. Endogenous expression of AID generates a signature mutation pattern specifically in BCL11B fragile region in T-ALL cells. To establish the role of AID in the mutagenesis of BCL11B, we generated AID-knockout (KO) in one of the T-ALL cell lines, MOLT4 using CRISPR-Cas9 technology. Cloning and DNA sequencing of the clones confirmed mutations in AID exon 2 at the target sequence in the knockout cells. Loss of AID expression was confirmed using immunoblotting. Chromatin immunoprecipitation studies revealed the specific binding of AID to BCL11B fragile region in exon 4 which was significantly decreased in the AID-KO cells. Interestingly, when both the wild type (WT) and AID-KO cells were grown in culture for a long period of time to assess the generation of de novo mutations in BCL11B fragile region, we observed a reduced mutation load in the AID knockouts compared to WT cells. Further, overexpression of AID in the knockout cells led to a two-fold increase in AID-induced hypermutation (C to T and G to A conversion), indicating that AID play a significant role in mutating BCL11B exon 4. When BCL11B fragile region was further analysed by isolating genomic DNA from lymphocytes of healthy individuals, we did not observe AID-induced mutations in this region. The action of AID requires single-stranded region in the genome, generated by transcription bubbles or stem-loop structures which are natural targets of this enzyme. Bioinformatic analysis revealed the presence of G4 sequence motifs, inverted repeats and mirror repeat sequences in the exon 4 of BCL11B of human and mice. Fragile region of exon 4 had majority of the non-B DNA motifs and hence was used for further studies. Sodium bisulfite modification assay revealed the single-strandedness of BCL11B fragile region at the chromosomal DNA level, indicating the propensity of these sequences to fold into non-B DNA structures. Using various biochemical and biophysical assays like Electrophoretic mobility shift assay (EMSA), Circular Dichroism (CD), DMS protection assay, P1 nuclease assay etc. we determined the formation of non-B DNA structures like parallel intramolecular G-quadruplex structure in the template strand, triplex DNA structure in the non-template strand and hairpin DNA in both the strands. Using overexpressed and purified GST-AID, we show that AID preferentially binds to such non-B DNA over linear ssDNA. Taken together, our results suggest that binding of AID to the single-stranded region due to the formation of non-B DNA in BCL11B exon 4 and subsequent deamination of unpaired cytosines results in a U:G mismatch. When repaired erroneously, deleterious point mutations/deletion/insertion are generated in the coding region of BCL11B gene, resulting in loss of its function, contributing towards leukemiagenesis. The aberrant expression of AID in T cell leukemia and the generation of thymic lymphomas in AID-transgenic mice prompted us to determine the endogenous expression of AID across the developmental stages of mouse and human thymocytes. Normalized count matrix was obtained from the publicly available NCBI GEO dataset (GSE139242) for the human T cell stages. RNA-seq dataset of 42 T cell samples from early T cell progenitors to terminally differentiated T helper cell subsets of the mouse were analyzed. The result revealed a significant expression of Aid in the double negative stages of mouse T cells and a moderate Aid expression in CD4 and CD8 population of human T cells. Interestingly, Bcl11b is highly expressed and transcriptionally active in the early T cell developmental stages, specifically the double-negative and double-positive stages. To determine the expression of Aid in the T lymphocytes of both young and adult mice, we sorted CD3+ T cells from mice thymus, wherein we observed a low but consistent expression of Aid in T cells of both the age groups. Sequencing of the transcript confirmed its identity. We further demonstrated that AID could bind to Bcl11b exon 4 in vivo, owing to the single-strandedness of this region in the genome of mouse thymocytes. Interestingly, nucleotide alterations in Bcl11b exon 4 were observed only in adult mice (7-8 months old) compared to young mice (2-3 weeks old). Whether the observed mutations are indeed generated due to Aid activity in this region warrants further investigation. In the second part of the study, we have explored the regulation of RAG1 activity by microRNAs (miRNAs), in addition to its regulation by transcription factors, enhancers etc. The discovery that RAGs can act on non-B DNA structures and cryptic RSS, that are abundant in the genome, significantly increases the spectrum of illegitimate recombinase activity by RAGs which can lead to the generation of genomic instability and cancer. Therefore, cells must possess stringent regulatory mechanisms for RAG expression to ensure genomic stability. Restriction of RAG expression to early stages of lymphocyte development, allowing unequal access to transcription factors/coactivators, controlling RAG expression through chromatin remodelling are some of the mechanisms operating in lymphocytes to prevent the occurrence of chromosomal translocation and development of B cell leukemias and lymphomas. microRNAs represent an additional layer of gene regulation that affects different arms of the living system, including the mammalian immune system. They are a class of small, non- coding RNAs, around 20-25 nucleotides in length, that fine-tune gene expression by either degradation of target mRNAs or inhibiting their protein expression following its binding to the 3’UTR region of its target gene. Although miRNAs regulating B cell development, maturation etc. are well explored, regulation of RAG1 expression and function in lymphoid cells by microRNAs have only been recently reported. Using in silico, ex vivo and in vivo studies, recently we identified miR-29c as a novel and direct regulator of RAG1 which modulates V(D)J recombination during B cell development. Interestingly, unlike miR-29c, knockout of miR-29a, a member of the miR-29 family, contributed to poor survival, impaired immunity, thymic involution etc. in mice. Since miR-29a shared the exact same seed sequence as that of miR-29c, we investigated the regulation of RAG1 by miR-29a and its impact on the immunity of mice. While overexpression of miR-29a in pre-B cell line, Nalm6 led to reduced expression of RAG1, inhibition of miR-29a using miRNA-specific inhibitor resulted in its enhanced expression in B cells. A significant downregulation of RAG1 expression at the protein level was also observed in miR-29a stable overexpression Nalm6 cells. On the other hand, enrichment of 3'-UTR of RAG1 within the cells led to enhanced expression of RAG1. The generation of mature miR-29a transcripts inside the cell following transfection with pre-miR-29a construct, pUR2-8, revealed the endogenous processing of the pre-miRNAs by the RISC (RNA-induced silencing) complex. Therefore, to determine its association with the Argonaute protein, we performed native RNA immunoprecipitation from bone marrow cells of mice and Ago HITS-CLIP (GEO: GSE137071) wherein we observed a strong interaction of Ago2 protein with miR-29a and RAG1 3’ UTR. Consistent with this, we observed generation of mature miR-29a transcripts in mice after intramuscular injection with the pre-miR-29a construct. This further led to a decrease in RAG1 protein expression in bone marrow cells of mice highlighting the significance of this regulation in the physiological context. However, there was no major alteration in T and B cell population, or other blood parameters of mice. Expression of GFP gene in bone marrow cells of mice post intramuscular delivery of pmaxGFP plasmid reiterates the delivery of plasmid DNA to bone marrow cells via the intramuscular route. Analysis of miR-29a and RAG1 levels across developmental stages of B cells in mice revealed a strong negative correlation between them. The early pre-B cells exhibited high RAG1 and low miR-29a expression, whereas robust miR-29a and low RAG1 level was observed in mature B cells. Interestingly, several reports suggest an impaired immune response in miR-29a knockout mice which die as early as 7 months after birth. The early death of these animals could be possibly attributed to increased DNA breaks and apoptosis caused due to unchecked expression of RAG1. Ubiquitous expression of Rag genes in mice is known to cause severe lymphopenia, growth retardation and a severe block in both B and T cell lymphopoiesis, a phenotype reminiscent of those reported for mice deficient in DNA double-strand break repair. Furthermore, we determined the level of miR-29a and RAG1 in lymphoid cell lines where we observed a negative correlation between them. Analysis of RNA-seq datasets from T-ALL and B-CLL patients also revealed a negative correlation between miR-29a and RAG1, underlying the possibility of using it as a biomarker and therapeutic target in cancer. Our results highlight the significance of miR-29a mediated regulation of RAG1 activity during B cell development in mice and its possible impact on immunoglobulin diversity, immune response, oncogenesis, and its potential use in cancer therapeutics. Further, this suggests potential role of miR-29a mediated regulation of RAG1 in maintaining the genomic stability in lymphoid cells. Thus, the first part of the study reveals that in addition to its physiological role, illegitimate action of AID outside the Ig loci could lead to mutations and deregulation of a key T cell transcription factor, BCL11B, essential for the survival and proliferation of the T cells. We show that AID is primarily responsible for generating point mutations in BCL11B exon 4, which is important for its binding to DNA and subsequent functional activity in T cells. Interestingly, we also find a consistent AID expression in the early stages of mice T cell development, which is in the same window where Bcl11b is expressed. From our study, it appears that AID binding to BCL11B exon 4 could be the initiating point of genomic fragility in this region, which warrants further investigation. In the second part of the thesis, we report the post-transcriptional regulation of RAG1 by a microRNA, miR-29a in B cells of mice and humans and its implications on different immunological parameters when pre-miR-29a is overexpressed in mice. Our study is consistent with observed defects in the survival of B and T cells in miR-29a knockout mice (van Nieuwenhuijze et al., 2017). Further, an inverse correlation of miR-29a with RAG1 in leukemia patients suggests the potential use of this microRNA in cancer therapeutics. Taken together, our study suggests that tight regulation of RAG1 and AID expression is crucial to prevent frequent DNA breaks, deleterious mutations, and genomic rearrangements mediated by these lymphoid-specific proteins inside the cells