Diversification of Antibodies: From V(D)J Recombination to Somatic Exon Shuffling.

Antibodies that gain specificity by a large insert encoding for an extra domain were described for the first time in 2016. In malaria-exposed individuals, an exon deriving from the leukocyte-associated immunoglobulin-like 1 (LAIR1) gene integrated via a copy-and-paste insertion into the immunoglobulin heavy chain encoding region. A few years later, a second example was identified, namely a dual exon integration from the leukocyte immunoglobulin-like receptor B1 (LILRB1) gene that is located in close proximity to LAIR1. A dedicated high-throughput characterization of chimeric immunoglobulin heavy chain transcripts unraveled, that insertions from distant genomic regions (including mitochondrial DNA) can contribute to human antibody diversity. This review describes the modalities of insert-containing antibodies. The role of known DNA mobility aspects, such as genomic translocation, gene conversion, and DNA fragility, is discussed in the context of insert-antibody generation. Finally, the review covers why insert antibodies were omitted from the past repertoire analyses and how insert antibodies can contribute to protective immunity or an autoreactive response.

Seq2scFv: a toolkit for the comprehensive analysis of display libraries from long-read sequencing platforms.

Antibodies have emerged as the leading class of biotherapeutics, yet traditional screening methods face significant time and resource challenges in identifying lead candidates. Integrating high-throughput sequencing with computational approaches marks a pivotal advancement in antibody discovery, expanding the antibody space to explore. In this context, a major breakthrough has been the full-length sequencing of single-chain variable fragments (scFvs) used in in vitro display libraries. However, few tools address the task of annotating the paired heavy and light chain variable domains (VH and VL), which is the primary advantage of full-scFv sequencing. To address this methodological gap, we introduce Seq2scFv, a novel open-source toolkit designed for analyzing in vitro display libraries from long-read sequencing platforms. Seq2scFv facilitates the identification and thorough characterization of V(D)J recombination in both VH and VL regions. In addition to providing annotated scFvs, translated sequences and numbered chains, Seq2scFv enables linker inference and characterization, sequence encoding with unique identifiers and quantification of identical sequences across selection rounds, thereby simplifying enrichment identification. With its versatile and standalone functionality, we anticipate that the implementation of Seq2scFv tools in antibody discovery pipelines will efficiently expedite the full characterization of display libraries and potentially facilitate the identification of high-affinity antibody candidates.

Large B-cell lymphomas with CCND1 rearrangement have different immunoglobulin gene breakpoints and genomic profile than mantle cell lymphoma.

Mantle cell lymphoma (MCL) is genetically characterized by the IG::CCND1 translocation mediated by an aberrant V(D)J rearrangement. CCND1 translocations and overexpression have been identified in occasional aggressive B-cell lymphomas with unusual features for MCL. The mechanism generating CCND1 rearrangements in these tumors and their genomic profile are not known. We have reconstructed the IG::CCND1 translocations and the genomic profile of 13 SOX11-negative aggressive B-cell lymphomas using whole genome/exome and target sequencing. The mechanism behind the translocation was an aberrant V(D)J rearrangement in three tumors and by an anomalous IGH class-switch recombination (CSR) or somatic hypermutation (SHM) mechanism in ten. The tumors with a V(D)J-mediated translocation were two blastoid MCL and one high-grade B-cell lymphoma. None of them had a mutational profile suggestive of DLBCL. The ten tumors with CSR/SHM-mediated IGH::CCND1 were mainly large B-cell lymphomas, with mutated genes commonly seen in DLBCL and BCL6 rearrangements in 6. Two cases, which transformed from marginal zone lymphomas, carried mutations in KLF2, TNFAIP3 and KMT2D. These findings expand the spectrum of tumors carrying CCND1 rearrangement that may occur as a secondary event in DLBCL mediated by aberrant CSR/SHM and associated with a mutational profile different from that of MCL.

Combining mutation and recombination statistics to infer clonal families in antibody repertoires.

B-cell repertoires are characterized by a diverse set of receptors of distinct specificities generated through two processes of somatic diversification: V(D)J recombination and somatic hypermutations. B-cell clonal families stem from the same V(D)J recombination event, but differ in their hypermutations. Clonal families identification is key to understanding B-cell repertoire function, evolution, and dynamics. We present HILARy (high-precision inference of lineages in antibody repertoires), an efficient, fast, and precise method to identify clonal families from single- or paired-chain repertoire sequencing datasets. HILARy combines probabilistic models that capture the receptor generation and selection statistics with adapted clustering methods to achieve consistently high inference accuracy. It automatically leverages the phylogenetic signal of shared mutations in difficult repertoire subsets. Exploiting the high sensitivity of the method, we find the statistics of evolutionary properties such as the site frequency spectrum and dN/dS ratio do not depend on the junction length. We also identify a broad range of selection pressures spanning two orders of magnitude.

RAG1/2 induces double-stranded DNA breaks at non-Ig loci in the proximity of single sequence repeats in developing B cells.

In developing B cells, V(D)J gene recombination is initiated by the RAG1/2 endonuclease complex, introducing double-stranded DNA breaks (DSBs) in V, D, and J genes and resulting in the formation of the hypervariable parts of immunoglobulins (Ig). Persistent or aberrant RAG1/2 targeting is a potential threat to genome integrity. While RAG1 and RAG2 have been shown to bind various regions genome-wide, the in vivo off-target DNA damage instigated by RAG1/2 endonuclease remains less well understood. In the current study, we identified regions containing RAG1/2-induced DNA breaks in mouse pre-B cells on a genome-wide scale using a global DNA DSB detection strategy. We detected 1489 putative RAG1/2-dependent DSBs, most of which were located outside the Ig loci. DNA sequence motif analysis showed a specific enrichment of RAG1/2-induced DNA DSBs at GA- and CA-repeats and GC-rich motifs. These findings provide further insights into RAG1/2 off-target activity. The ability of RAG1/2 to introduce DSBs on the non-Ig loci during the endogenous V(D)J recombination emphasizes its genotoxic potential in developing lymphocytes.

RAG1 and RAG2 non-core regions are implicated in leukemogenesis and off-target V(D)J recombination in BCR-ABL1-driven B-cell lineage lymphoblastic leukemia.

The evolutionary conservation of non-core RAG regions suggests significant roles that might involve quantitative or qualitative alterations in RAG activity. Off-target V(D)J recombination contributes to lymphomagenesis and is exacerbated by RAG2' C-terminus absence in Tp53-/- mice thymic lymphomas. However, the genomic stability effects of non-core regions from both Rag1c/c and Rag2c/c in BCR-ABL1+ B-lymphoblastic leukemia (BCR-ABL1+ B-ALL), the characteristics, and mechanisms of non-core regions in suppressing off-target V(D)J recombination remain unclear. Here, we established three mouse models of BCR-ABL1+ B-ALL in mice expressing full-length RAG (Ragf/f), core RAG1 (Rag1c/c), and core RAG2 (Rag2c/c). The Ragc/c (Rag1c/c and Rag2c/c) leukemia cells exhibited greater malignant tumor characteristics compared to Ragf/f cells. Additionally, Ragc/c cells showed higher frequency of off-target V(D)J recombination and oncogenic mutations than Ragf/f. We also revealed decreased RAG cleavage accuracy in Ragc/c cells and a smaller recombinant size in Rag1c/c cells, which could potentially exacerbate off-target V(D)J recombination in Ragc/c cells. In conclusion, these findings indicate that the non-core RAG regions, particularly the non-core region of RAG1, play a significant role in preserving V(D)J recombination precision and genomic stability in BCR-ABL1+ B-ALL.

Addressing Technical Pitfalls in Pursuit of Molecular Factors That Mediate Immunoglobulin Gene Regulation.

The expressed Ab repertoire is a critical determinant of immune-related phenotypes. Ab-encoding transcripts are distinct from other expressed genes because they are transcribed from somatically rearranged gene segments. Human Abs are composed of two identical H and L chain polypeptides derived from genes in IGH locus and one of two L chain loci. The combinatorial diversity that results from Ab gene rearrangement and the pairing of different H and L chains contributes to the immense diversity of the baseline Ab repertoire. During rearrangement, Ab gene selection is mediated by factors that influence chromatin architecture, promoter/enhancer activity, and V(D)J recombination. Interindividual variation in the composition of the Ab repertoire associates with germline variation in IGH, implicating polymorphism in Ab gene regulation. Determining how IGH variants directly mediate gene regulation will require integration of these variants with other functional genomic datasets. In this study, we argue that standard approaches using short reads have limited utility for characterizing regulatory regions in IGH at haplotype resolution. Using simulated and chromatin immunoprecipitation sequencing reads, we define features of IGH that limit use of short reads and a single reference genome, namely 1) the highly duplicated nature of the DNA sequence in IGH and 2) structural polymorphisms that are frequent in the population. We demonstrate that personalized diploid references enhance performance of short-read data for characterizing mappable portions of the locus, while also showing that long-read profiling tools will ultimately be needed to fully resolve functional impacts of IGH germline variation on expressed Ab repertoires.

Molecular characterization of V(D)J rearrangements in immature acute leukemias.

Early T-cell Precursor Acute Lymphoblastic Leukemia (ETP-ALL), T-Lymphoid/Myeloid Mixed Phenotype Acute Leukemia (T/M-MPAL), and Acute Myeloid Leukemia with minimal differentiation (AML-M0) are immature acute leukemias (AL) that present overlapping T-cell lymphoid and myeloid features at different degrees, with impact to disease classification. An interesting strategy to assess lymphoid lineage commitment and maturation is the analysis of V(D)J gene segment recombination, which can be applied to investigate leukemic cells in immature AL. Herein, we revisited 19 ETP-ALL, 8 T/M-MPAL, and 12 AML-M0 pediatric patients to characterize V(D)J rearrangement (V(D)J-r) profiles associated with other somatic alterations. V(D)J-r were identified in 74 %, 25 %, and 25 % of ETP-ALL, T/M-MPAL, and AML-M0, respectively. Forty-six percent of ETP-ALL harbored ≥ 3 V(D)J-r, while there was no more than one V(D)J-r per patient in AML-M0 and T/M-MPAL. TCRD was the most rearranged locus in ETPALL, but it was not rearranged in other AL. In ETP-ALL, N/KRAS mutations were associated with absence of V(D)J-r, while NF1 deletion was most frequent in patients with ≥ 3 V(D)J-r. Relapse and death occurred mainly in patients harboring one or no rearranged locus. Molecular characterization of V(D)J-r in our cohort indicates a distinct profile of ETP-ALL, compared to T/M-MPAL and AML-M0. Our findings also suggest that the clinical outcome of ETP-ALL patients may be affected by blast cell maturity, inferred from the number of rearranged TCR loci.

Molecular basis for differential Igk versus Igh V(D)J joining mechanisms.

In developing B cells, V(D)J recombination assembles exons encoding IgH and Igκ variable regions from hundreds of gene segments clustered across Igh and Igk loci. V, D and J gene segments are flanked by conserved recombination signal sequences (RSSs) that target RAG endonuclease1. RAG orchestrates Igh V(D)J recombination upon capturing a JH-RSS within the JH-RSS-based recombination centre1-3 (RC). JH-RSS orientation programmes RAG to scan upstream D- and VH-containing chromatin that is presented in a linear manner by cohesin-mediated loop extrusion4-7. During Igh scanning, RAG robustly utilizes only D-RSSs or VH-RSSs in convergent (deletional) orientation with JH-RSSs4-7. However, for Vκ-to-Jκ joining, RAG utilizes Vκ-RSSs from deletional- and inversional-oriented clusters8, inconsistent with linear scanning2. Here we characterize the Vκ-to-Jκ joining mechanism. Igk undergoes robust primary and secondary rearrangements9,10, which confounds scanning assays. We therefore engineered cells to undergo only primary Vκ-to-Jκ rearrangements and found that RAG scanning from the primary Jκ-RC terminates just 8 kb upstream within the CTCF-site-based Sis element11. Whereas Sis and the Jκ-RC barely interacted with the Vκ locus, the CTCF-site-based Cer element12 4 kb upstream of Sis interacted with various loop extrusion impediments across the locus. Similar to VH locus inversion7, DJH inversion abrogated VH-to-DJH joining; yet Vκ locus or Jκ inversion allowed robust Vκ-to-Jκ joining. Together, these experiments implicated loop extrusion in bringing Vκ segments near Cer for short-range diffusion-mediated capture by RC-based RAG. To identify key mechanistic elements for diffusional V(D)J recombination in Igk versus Igh, we assayed Vκ-to-JH and D-to-Jκ rearrangements in hybrid Igh-Igk loci generated by targeted chromosomal translocations, and pinpointed remarkably strong Vκ and Jκ RSSs. Indeed, RSS replacements in hybrid or normal Igk and Igh loci confirmed the ability of Igk-RSSs to promote robust diffusional joining compared with Igh-RSSs. We propose that Igk evolved strong RSSs to mediate diffusional Vκ-to-Jκ joining, whereas Igh evolved weaker RSSs requisite for modulating VH joining by RAG-scanning impediments.

NUDCD3 deficiency disrupts V(D)J recombination to cause SCID and Omenn syndrome.

Inborn errors of T cell development present a pediatric emergency in which timely curative therapy is informed by molecular diagnosis. In 11 affected patients across four consanguineous kindreds, we detected homozygosity for a single deleterious missense variant in the gene NudC domain-containing 3 (NUDCD3). Two infants had severe combined immunodeficiency with the complete absence of T and B cells (T -B- SCID), whereas nine showed classical features of Omenn syndrome (OS). Restricted antigen receptor gene usage by residual T lymphocytes suggested impaired V(D)J recombination. Patient cells showed reduced expression of NUDCD3 protein and diminished ability to support RAG-mediated recombination in vitro, which was associated with pathologic sequestration of RAG1 in the nucleoli. Although impaired V(D)J recombination in a mouse model bearing the homologous variant led to milder immunologic abnormalities, NUDCD3 is absolutely required for healthy T and B cell development in humans.

Two recurrent types of IGH::5' BCL2 breakpoints representing cytogenetic ins(14;18)(q32;q21q21) and t(14;18)(q32;q21), mediated by the VDJ and class switch recombination processes, respectively.

We describe two types of IGH::BCL2 breakpoints involving the 5' region of BCL2 (5' BCL2). One was ins(14;18)(q32;q21q21) observed in 2 follicular lymphoma (FL) cases, in which IGH was cleaved at 3' of IGHD and 5' of IGHJ and BCL2 was cleaved at 5' BCL2 and downstream regions, and a 281- or 201-kilobase pair fragment containing the BCL2 protein-coding sequences was invertedly inserted into IGH. In another type observed in 2 FL and 2 chronic lymphocytic leukemia (CLL) cases, breakage and reunion occurred within the switch region associated with IGHM (Sµ) and 5' BCL2, creating IGH Sµ::5' BCL2 fusion sequences on der(18)t(14;18)(q32;q21). The former is considered to be mediated by VDJ-recombination, while the latter by the class switch recombination process. There were no particular features in FL or CLL cases with IGH::5' BCL2 breakpoints compared with those with t(14;18)(q32;q21)/IGH::BCL2 involving the 3' breakpoint cluster regions.

Non-canonical functions of enhancers: regulation of RNA polymerase III transcription, DNA replication, and V(D)J recombination.

Enhancers are the key regulators of other DNA-based processes by virtue of their unique ability to generate nucleosome-depleted regions in a highly regulated manner. Enhancers regulate cell-type-specific transcription of tRNA genes by RNA polymerase III (Pol III). They are also responsible for the binding of the origin replication complex (ORC) to DNA replication origins, thereby regulating origin utilization, replication timing, and replication-dependent chromosome breaks. Additionally, enhancers regulate V(D)J recombination by increasing access of the recombination-activating gene (RAG) recombinase to target sites and by generating non-coding enhancer RNAs and localized regions of trimethylated histone H3-K4 recognized by the RAG2 PHD domain. Thus, enhancers represent the first step in decoding the genome, and hence they regulate biological processes that, unlike RNA polymerase II (Pol II) transcription, do not have dedicated regulatory proteins.

The endogenous Mtv8 locus and the immunoglobulin repertoire.

The vast diversity of mammalian adaptive antigen receptors allows for robust and efficient immune responses against a wide number of pathogens. The antigen receptor repertoire is built during the recombination of B and T cell receptor (BCR, TCR) loci and hypermutation of BCR loci. V(D)J recombination rearranges these antigen receptor loci, which are organized as an array of separate V, (D), and J gene segments. Transcription activation at the recombining locus leads to changes in the local three-dimensional architecture, which subsequently contributes to which gene segments are utilized for recombination. The endogenous retrovirus (ERV) mouse mammary tumor provirus 8 (Mtv8) resides on mouse chromosome 6 interposed within the large array of light chain kappa V gene segments. As ERVs contribute to changes in genomic architecture by driving high levels of transcription of neighboring genes, it was suggested that Mtv8 could influence the BCR repertoire. We generated Mtv8-deficient mice to determine if the ERV influences V(D)J recombination to test this possibility. We find that Mtv8 does not influence the BCR repertoire.

Evolutionary preservation of CpG dinucleotides in RAG1 may elucidate the relatively high rate of methylation-mediated mutagenesis of RAG1 transposase.

Recombination-activating gene 1 (RAG1) is a vital player in V(D)J recombination, a fundamental process in primary B cell and T cell receptor diversification of the adaptive immune system. Current vertebrate RAG evolved from RAG transposon; however, it has been modified to play a crucial role in the adaptive system instead of being irreversibly silenced by CpG methylation. By interrogating a range of publicly available datasets, the current study investigated whether RAG1 has retained a disproportionate level of its original CpG dinucleotides compared to other genes, thereby rendering it more exposed to methylation-mediated mutation. Here, we show that 57.57% of RAG1 pathogenic mutations and 51.6% of RAG1 disease-causing mutations were associated with CpG methylation, a percentage that was significantly higher than that of its RAG2 cofactor alongside the whole genome. The CpG scores and densities for all RAG ancestors suggested that RAG transposon was CpG denser. The percentage of the ancestral CpG of RAG1 and RAG2 were 6% and 4.2%, respectively, with no preference towards CG containing codons. Furthermore, CpG loci of RAG1 in sperms were significantly higher methylated than that of RAG2. In conclusion, RAG1 has been exposed to CpG mediated methylation mutagenesis more than RAG2 and the whole genome, presumably due to its late entry to the genome later with an initially higher CpG content.

MicroRNA, miR-501 regulate the V(D)J recombination in B cells.

The stringent regulation of RAGs (Recombination activating genes), the site-specific endonuclease responsible for V(D)J recombination, is important to prevent genomic rearrangements and chromosomal translocations in lymphoid cells. In the present study, we identify a microRNA, miR-501, which can regulate the expression of RAG1 in lymphoid cells. Overexpression of the pre-miRNA construct led to the generation of mature miRNAs and a concomitant reduction in RAG1 expression, whereas inhibition using anti-miRs resulted in its enhanced expression. The direct interaction of the 3'UTR of miR-501 with RAG1 was confirmed by the reporter assay. Importantly, overexpression of miRNAs led to inhibition of V(D)J recombination in B cells, revealing their impact on the physiological function of RAGs. Of interest is the inverse correlation observed for miR-501 with RAG1 in various leukemia patients and lymphoid cell lines, suggesting its possible use in cancer therapy. Thus, our results reveal the regulation of RAG1 by miR-501-3p in B cells and thus V(D)J recombination and its possible implications on immunoglobulin leukemogenesis.

Intra- and interchromosomal contact mapping reveals the Igh locus has extensive conformational heterogeneity and interacts with B-lineage genes.

To produce a diverse antibody repertoire, immunoglobulin heavy-chain (Igh) loci undergo large-scale alterations in structure to facilitate juxtaposition and recombination of spatially separated variable (VH), diversity (DH), and joining (JH) genes. These chromosomal alterations are poorly understood. Uncovering their patterns shows how chromosome dynamics underpins antibody diversity. Using tiled Capture Hi-C, we produce a comprehensive map of chromatin interactions throughout the 2.8-Mb Igh locus in progenitor B cells. We find that the Igh locus folds into semi-rigid subdomains and undergoes flexible looping of the VH genes to its 3' end, reconciling two views of locus organization. Deconvolution of single Igh locus conformations using polymer simulations identifies thousands of different structures. This heterogeneity may underpin the diversity of V(D)J recombination events. All three immunoglobulin loci also participate in a highly specific, developmentally regulated network of interchromosomal interactions with genes encoding B cell-lineage factors. This suggests a model of interchromosomal coordination of B cell development.

Three-way contact analysis characterizes the higher order organization of the Tcra locus.

The generation of highly diverse antigen receptors in T and B lymphocytes relies on V(D)J recombination. The enhancer Eα has been implicated in regulating the accessibility of Vα and Jα genes through long-range interactions during rearrangements of the T-cell antigen receptor gene Tcra. However, direct evidence for Eα physically mediating the interaction of Vα and Jα genes is still lacking. In this study, we utilized the 3C-HTGTS assay, a chromatin interaction technique based on 3C, to analyze the higher order chromatin structure of the Tcra locus. Our analysis revealed the presence of sufficient information in the 3C-HTGTS data to detect multiway contacts. Three-way contact analysis of the Tcra locus demonstrated the co-occurrence of the proximal Jα genes, Vα genes and Eα in CD4+CD8+ double-positive thymocytes. Notably, the INT2-TEAp loop emerged as a prominent structure likely to be responsible for bringing the proximal Jα genes and the Vα genes into proximity. Moreover, the enhancer Eα utilizes this loop to establish physical proximity with the proximal Vα gene region. This study provides insights into the higher order chromatin structure of the Tcra locus, shedding light on the spatial organization of chromatin and its impact on V(D)J recombination.

E protein binding at the Tcra enhancer promotes Tcra repertoire diversity.

V(D)J recombination of antigen receptor loci is a highly developmentally regulated process. During T lymphocyte development, recombination of the Tcra gene occurs in CD4+CD8+ double positive (DP) thymocytes and requires the Tcra enhancer (Eα). E proteins are known regulators of DP thymocyte development and have three identified binding sites in Eα. To understand the contribution of E proteins to Eα function, mutants lacking one or two of the respective binding sites were generated. The double-binding site mutant displayed a partial block at the positive selection stage of αß T cell development. Further investigation revealed loss of germline transcription within the Tcra locus at the Jα array, along with dysregulated primary and impaired secondary Vα-Jα rearrangement. Eα E protein binding increases Tcra locus accessibility and regulates TCRα recombination, thus directly promoting Tcra repertoire diversity.

Contribution of the IGCR1 regulatory element and the 3'Igh CTCF-binding elements to regulation of Igh V(D)J recombination.

Immunoglobulin heavy chain variable region exons are assembled in progenitor-B cells, from VH, D, and JH gene segments located in separate clusters across the Igh locus. RAG endonuclease initiates V(D)J recombination from a JH-based recombination center (RC). Cohesin-mediated extrusion of upstream chromatin past RC-bound RAG presents Ds for joining to JHs to form a DJH-RC. Igh has a provocative number and organization of CTCF-binding elements (CBEs) that can impede loop extrusion. Thus, Igh has two divergently oriented CBEs (CBE1 and CBE2) in the IGCR1 element between the VH and D/JH domains, over 100 CBEs across the VH domain convergent to CBE1, and 10 clustered 3'Igh-CBEs convergent to CBE2 and VH CBEs. IGCR1 CBEs segregate D/JH and VH domains by impeding loop extrusion-mediated RAG-scanning. Downregulation of WAPL, a cohesin unloader, in progenitor-B cells neutralizes CBEs, allowing DJH-RC-bound RAG to scan the VH domain and perform VH-to-DJH rearrangements. To elucidate potential roles of IGCR1-based CBEs and 3'Igh-CBEs in regulating RAG-scanning and elucidate the mechanism of the ordered transition from D-to-JH to VH-to-DJH recombination, we tested effects of inverting and/or deleting IGCR1 or 3'Igh-CBEs in mice and/or progenitor-B cell lines. These studies revealed that normal IGCR1 CBE orientation augments RAG-scanning impediment activity and suggest that 3'Igh-CBEs reinforce ability of the RC to function as a dynamic loop extrusion impediment to promote optimal RAG scanning activity. Finally, our findings indicate that ordered V(D)J recombination can be explained by a gradual WAPL downregulation mechanism in progenitor-B cells as opposed to a strict developmental switch.

Origin and evolutionary malleability of T cell receptor α diversity.

Lymphocytes of vertebrate adaptive immune systems acquired the capability to assemble, from split genes in the germline, billions of functional antigen receptors1-3. These receptors show specificity; unlike the broadly tuned receptors of the innate system, antibodies (Ig) expressed by B cells, for instance, can accurately distinguish between the two enantiomers of organic acids4, whereas T cell receptors (TCRs) reliably recognize single amino acid replacements in their peptide antigens5. In developing lymphocytes, antigen receptor genes are assembled from a comparatively small set of germline-encoded genetic elements in a process referred to as V(D)J recombination6,7. Potential self-reactivity of some antigen receptors arising from the quasi-random somatic diversification is suppressed by several robust control mechanisms8-12. For decades, scientists have puzzled over the evolutionary origin of somatically diversifying antigen receptors13-16. It has remained unclear how, at the inception of this mechanism, immunologically beneficial expanded receptor diversity was traded against the emerging risk of destructive self-recognition. Here we explore the hypothesis that in early vertebrates, sequence microhomologies marking the ends of recombining elements became the crucial targets of selection determining the outcome of non-homologous end joining-based repair of DNA double-strand breaks generated during RAG-mediated recombination. We find that, across the main clades of jawed vertebrates, TCRα repertoire diversity is best explained by species-specific extents of such sequence microhomologies. Thus, selection of germline sequence composition of rearranging elements emerges as a major factor determining the degree of diversity of somatically generated antigen receptors.