Epigenetic control of Tcrb gene rearrangement

Seminars in Immunology 22 (2010) 330–336

Contents lists available at ScienceDirect

Seminars in Immunology journal homepage: www.elsevier.com/locate/ysmim

Review

Epigenetic control of Tcrb gene rearrangement Salvatore Spicuglia a,b,c , Aleksandra Pekowska a,b,c , Joaquin Zacarias-Cabeza a,b,c , Pierre Ferrier a,b,c,∗ a

Centre d’Immunologie de Marseille-Luminy, Inserm, UMR-S 631, F-13009 Marseille, France CNRS, UMR 6102, F-13009 Marseille, France c Université de la Méditerranée, UM 631, F-13009 Marseille, France b

a r t i c l e

i n f o

a b s t r a c t

Keywords: V(D)J recombination Allelic exclusion Tcrb Epigenetics

V(D)J recombination assembles antigen receptor genes from germline V, D and J segments during lymphocyte development. In ␣␤T-cells, this leads to the subsequent expression of T-cell receptor (TCR) ␤ and ␣ chains. Generally, V(D)J recombination is closely controlled at various levels, including cell-type and cell-stage specificities, order of locus/gene segment recombination, and allele usage to mediate allelic exclusion. Many of these controls rely on the modulation of gene accessibility to the recombination machinery, involving not only biochemical changes in chromatin arrangement and structural modifications of chromosomal organization and positioning, but also the refined composition of the recombinase targets, the so-called recombination signal sequences. Here, we summarize current knowledge regarding the regulation of V(D)J recombination at the Tcrb gene locus, certainly one for which these various levels of control and regulatory components have been most extensively investigated. © 2010 Elsevier Ltd. All rights reserved.

1. Introduction

to determining the developmental order of gene rearrangement events and allelically excluded expression of a productively (inframe) rearranged Tcrb gene. During the past 15 years, this locus has served as a tractable model to study these controls, in particular via the utilization of gene targeting technologies for the generation and analysis of >23 genomic deletions (‘knockout’) or replacements (‘knockin’) of different scales [7]. Notably, these studies led to a better appreciation of the hierarchical organization of transcriptional cis-regulatory elements (enhancers and germline promoters) and their impact on the control of V(D)J recombination at the Tcrb locus via the modulation of the chromatin and chromosomal structures [8,9]. How this is achieved at the molecular level is at the focus of continuing and intense investigations. Here, we review relevant aspects of the control of V(D)J recombination at the Tcrb locus, including the structural characteristics and organization of TcrbRSs, as well as the dynamics of discrete epigenetic marking along the locus and changes in chromosomal conformation and spatial localization within the nucleus. We discuss how these aspects could influence Tcrb gene rearrangement; and eventually, be integrated into a dynamical model of allelic exclusion at this locus.

B and T lymphocytes form the adaptive arm of the immune system in jawed vertebrates, which can specifically respond to an astounding number of foreign antigens. This property largely depends on V(D)J recombination events at antigen receptor (AR)encoding loci [1,2]. There are seven AR loci, comprising the immunoglobulin heavy (IgH) and light (Igk and Igl) chain loci expressed in B-cells and the T-cell receptor (Tcra, Tcrb, Tcrd and Tcrg) loci expressed in T-cells. For V(D)J recombination to occur, the presence of the lymphoid-specific proteins RAG1 and RAG2 and the ubiquitously expressed DNA repair factors from the nonhomologous end joining (NHEJ) pathway are required. Regulation of these factors’ recruitment to and function at their genomic targets – primarily effected at the level of chromosomal accessibility – ensures B- or T-cell lineage-specificity of V(D)J recombination, dictates the temporal order of Ig or TCR rearrangements, and allows allelic exclusion at certain AR genes ([3–5]; for recent reviews, see Ref. [6]). In particular, the Tcrb locus is subjected to distinct levels of controls of gene expression, which collectively contribute

2. Tcrb gene rearrangement: overview Abbreviations: AR, antigen receptor; DN, double-negative; DP, double-positive; E␤, Tcrb enhancer; FISH, fluorescence in situ hybridization; GT, germline transcription; Ig, immunoglobulin; kb, kilobases; Pol, RNA polymerase; RS, recombination signal sequence; Tcr, T-cell receptor; TF, transcription factor. ∗ Corresponding author at: Dept. of Immunology, Marseille Luminy (CIML), Parc Scientifique & Technologique de Luminy, Case 906, 13288 Marseille Cedex 9, France. Tel.: +33 491 26 9435; fax: +33 491 26 9430. E-mail address: [email protected] (P. Ferrier). 1044-5323/$ – see front matter © 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.smim.2010.07.002

T-cell development is characterized by temporally regulated expression and rearrangement of the various Tcr genes; with Tcrd/g/b and Tcra gene rearrangements being carried out at two distinct stages of thymic-cell development along the ␥␦ and ␣␤ Tcell differentiation pathways, respectively. In the case of the ␣␤ T-cell pathway, VDJ assembly at the Tcrb locus is initiated and

S. Spicuglia et al. / Seminars in Immunology 22 (2010) 330–336

achieved in CD4− CD8− double-negative (DN) thymocytes (more precisely the so-called DN2 and DN3 sub-compartments; see Ref. [10]); and proceeds stepwise with D␤-to-J␤ occurring first prior to V␤-to-DJ␤ joining (NB: a description of Tcrd and Tcrg gene rearrangements and impact on ␥␦ and ␣␤ T-cell development is beyond the scope of this article; dedicated reviews can be found in Refs. [11,12]). Expression of a functionally (in-frame VDJ-C) rearranged Tcrb gene leads thymocytes to successfully pass ␤-selection and differentiate into CD4+ CD8+ double-positive (DP) cells, in which Tcra gene expression and recombination in turn take place. Afterwards, TCR␣␤-expressing cells might be further selected into mature CD4+ or CD8+ single-positive (SP) T-cells (reviewed in Ref. [13]). In the mouse germline, the approximately 700-kilobases (kb) Tcrb locus lies on the long arm of chromosome 6 and consists of a large (∼425-kb) 5 region containing 22 functional V␤ gene segments as well as 13 additional V␤ pseudogenes, and a shorter (∼25-kb) 3 region comprised of a duplicated cluster of D␤–J␤–C␤ gene segments (Fig. 1 and Ref. [14]). A separate V␤ gene segment (V␤14), lying in the opposite transcriptional direction, is located at the 3 end of the locus. As determined by the type and orientation of their flanking RSs, recombination of all 5 V␤, D␤ and J␤ gene segments is deletional, except that of V␤14 which occurs by inversion. Non-rearranged V␤ and D␤ segments are linked with upstream promoters that are thought to dictate the regional initiation of germline transcription (GT), and thus V(D)J recombination, in a developmentally regulated fashion (reviewed in Ref. [8]). Besides each of the long-known V␤-associated promoter and the D␤1-flanking promoter (the latter also referred to as pD␤1), two promoters located on either sides of the D␤2 gene segment have been lately characterized [15,16]. A single transcriptional enhancer (E␤) lies between the C␤2 coding region and V␤14, which is critical for this locus’s functional activation including GT and V(D)J recombination (detailed further below). The formation of a complete VDJ␤ variable region places the promoter of the rearranged V␤ segment within the E␤ activation area, thus significantly enhancing the transcriptional activity of the newly assembled V␤DJ␤ unit. Recently, putative additional cis-regulatory regions were identified within the V␤–DJ␤ intervening region [17] and across the J␤ gene segments [18], although their function and hierarchical relation (i.e., relative to the above mentioned cis-regulatory elements) with respect to Tcrb gene expression and recombination are not clear yet. The partitioned organization of the Tcrb locus and structure of the particular RSs impose a number of constraints to the regulation of V(D)J recombination at these particular sites.

3. Structural attributes impinging on Tcrb gene rearrangement 3.1. Recombination signal sequences (RSs) Despite their overall conservation, RSs in general, and TcrbRSs in particular, exhibit marked sequence variations compared to the canonical RSs [14,19–22]. As the RS itself, together with a few nucleotides of immediately flanking sequences within the adjacent gene coding segment, represents the loading platform for the RAG1/RAG2 recombinase [23], such structural variability naturally impacts the effectiveness of rearrangement of the given gene segment. Actually, the Tcrb genomic repertoire indeed reflects this subtle interplay between the RSs and the flanking coding sequences [24–26]. Moreover, nucleotide sequences of some VH and V␤ RSs cause them to recombine less efficiently relative to RSs (such as the D␤-RSs) the sequences of which better match the RS consensus [25,27]; and J␤-RSs appear highly heterogeneous within each cluster, displaying only a few conserved nucleotides,

331

which may account for their apparent low efficacy in focusing RAGmediated cleavage [27]. Besides, at the Tcrb locus, a spectacular example on how RSs can influence, and may in fact dictate, V(D)J recombination profiles was illustrated by the discovery of a novel paradigm referred to as ‘beyond the 12/23’, or B12/23 [28]. Bypassing the usual 12/23 rule, B12/23 restriction allows D␤-12RSs, but not J␤-12RSs, to efficiently target V␤-23RSs for rearrangement, making direct V␤-to-J␤ joining practically inexistent within the endogenous locus—a phenomenon that (i) depends uniquely on the RAG1/2 apparatus (i.e., with no participation of any extra lymphoidspecific factor); (ii) relies entirely on the RSs’ nucleotide sequences; and (iii) occurs at or prior to RAG1/2-mediated DNA coupled cleavage of V(D)J recombination (reviewed in Ref. [29]). No matter how remarkable the B12/23 restriction, its actual impact on the regulation of V(D)J recombination at the Tcrb locus and basic timing of ordered D␤-to-J␤ and V␤-to-DJ␤ rearrangements is not clear yet. In this respect, additional insight into this phenomenon came from the unexpected findings that the transcription factor (TF) c-Fos, a component of the AP1 transcriptional complex, not only binds the 3 D␤1-23RS but also directs the recruitment of the RAG1/2 core component of the VDJ recombinase at this site, thereby enabling D␤-to-J␤ joining [30]. Further, c-Fos deficiency results in decreased Tcrb gene recombination and the detection of V␤–D␤1 rearrangements, suggesting a disruption of ordered recombination [30]. Independently, we also demonstrated that single-strand DNA nicks (the very initial step of RAG1/2mediated DNA cleavage at a RS) are primarily observed in vivo at both the 3 - and 5 D␤-RSs, and in a E␤-dependent way, implying a primary role for these sites during the course of Tcrb gene recombination, possibly via recombinase anchoring and the nucleated capture of their respective V␤ and J␤ RS partners (provided the D␤ RSs were first made accessible via E␤ function(s) on chromatin structure, Ref. [27]). In the latter study, we further demonstrated that the presence of the 3 D␤1-23RS impedes DNA cleavage at the adjacent 5 D␤1-12RS in some way, hence providing potential molecular hints as to how and why D␤1–J␤ assembly occurs first, followed by the joining of a V␤ gene segment whose capture by an activated (i.e., RAG1/2 loaded) 5 D␤1-12RSS would thus require prior excision of the 3 D␤1-23RS via deletional D–J recombination. Accumulating evidence that D␤-to-J␤ rearrangement initially involves the D␤1–J␤1 cluster during T-cell development [31–33], adds further consistency to this scenario in terms of the regulation of Tcrb ordered rearrangement. However, following analysis of the rearrangements involving D␤1–J␤1.1 genomic sequences outside of their native Tcrb domain, it has been argued that RAG deposition on 3 D␤ RSSs may not be sufficient to direct ordered Tcrb rearrangements; and that the location of the D␤–J␤ segments relative to their germline promoters and/or the E␤ enhancer may instead be critical for directing the assembly of endogenous V␤s through DJ␤ intermediates [34]. In any case, these results collectively indicate that RSs, especially D␤-RSs, impose significant constraints on Tcrb gene assembly, the effects of which are felt beyond those restricting chromosomal accessibility to the VDJ recombinase. At a minimum, their asset in driving the action of the recombination apparatus at this locus guarantees the utilization of an intermediate D␤ gene segment, hence the expression of a fully diversified TCR␤ repertoire [35]. 3.2. Chromatin accessibility and epigenetics Pioneering studies by Alt and collaborators [36–37] led to the concept that locus-specific control and temporal-ordering of V(D)J recombination primary involve the modulation of locus and/or gene segment accessibility to a common VDJ recombinase (accessibility model). Since then, results from countless experiments have confirmed this model (reviewed in Refs. [1,9,38]), including the

332

S. Spicuglia et al. / Seminars in Immunology 22 (2010) 330–336

seminal findings that both the lineage-specificity and temporalordering of gene rearrangement is reflected in in vitro recombinase cleavage of RSs flanking Ig and TCR gene segments within chromatin in isolated nuclei [39]. Generally speaking, chromatin at gene segments/loci undergoing V(D)J rearrangement displays criteria that indeed correlate with an open, ‘non-compacted’ configuration, as opposed to those found alongside recombination-inert regions ([40,41]; lately reviewed in Ref. [42]). As for the Tcrb locus, for example, several molecular parameters synonymous with euchromatin (GT; lack of CpG methylation; enrichment in histone H3/H4 acetylation and H3K4 methylation; accessibility to restriction enzymes; diminished nucleosome abundance) have been linked both locally and in a stage-specific way with D␤-to-J␤ recombination events [33,41,43–47]; but, from the DN-to-DP thymic-cell transition onwards, lack of GT and decreased H3/H4ac predominated along chromosomal regions comprising non-rearranged 5 V␤ genes [43,48–51]. In the same line, via insertion of a D␤–J␤ recombination substrate into the endogenous V␤14 gene segment, it has recently been shown that the local chromatin environment imparts lineage and developmental stage-specific accessibility for recombination upon the inserted reporter [52].

The establishment of a euchromatin configuration and associated histone marks (e.g., H3K4ac, H3K4me, . . .) at V(D)J rearranging loci raises the possibility that a specific histone code acts to restrict V(D)J recombination in vivo. However, to date, no particular combination of histone marks has been exclusively associated with the AR loci (Refs. [41,53]; but see below). Regardless of these concerns, the question still remains as to how active epigenetic marks are established through the recombining gene segments and associated RSs? One possibility could be that GT, either sense or antisense, impinges on epigenetic marking before AR V(D)J assembly [33,54–56]; although the functional significance, if any, of antisense transcription at a given Ig/TCR locus or cluster still requires thorough investigation (also see Refs. [57–58]). By performing ChIP-on-chip experiments for H3K4me2 post-translational tagging throughout the Tcrb locus, we found that this mark of euchromatin is highly enriched at the DJ␤ clusters as compared to the V␤ regions [32]. Interestingly, H3K4me2 was not confined to the Tcrb cisregulatory elements, but spread over the entire DJC␤ transcription units in a E␤-dependent manner (similar observation was made for additional marks of an active chromatin including H3K4me3 and H3/H4ac; our unpublished results), suggesting a close link between

Fig. 1. Genomic organization of the mouse Tcrb locus. The V, D, J and C gene segments are designed following the conventional ImMunoGeneTics (IMGT; http://imgt.cines.fr/) nomenclature (top) and corresponding published names (bottom; for V␤ segments only). Thick black and green lines represent functional and non-functional V␤ gene segments, respectively; and grey boxes represent trypsinogen genes. Thin black lines are for D gene segments; white and dashed boxes for J and C gene segments, respectively. An enlargement of the 3 region is shown, where the E␤ enhancer (red oval) and transcriptional D␤ and V␤14 promoters (black ovals) are also figured. The table in the bottom refers to published knockout (ko) and knockin (ki) alleles of the E␤ enhancer, and the resulting phenotypes.

S. Spicuglia et al. / Seminars in Immunology 22 (2010) 330–336

RNA polymerase (Pol) II-mediated transcription and epigenetic tagging at these particular sites of the Tcrb locus. In this context, it is remarkable that binding of each RAG1 and RAG2 proteins occurs in a highly focal manner to a small region of active chromatin precisely encompassing Tcrb J and proximal D gene segments, in a developmental stage- and lineage-specific manner [59]. Likely, chromatin opening at E␤-proximal D–J␤ regions authorizes the formation of these small RAG-bound regions, now referred by Ji and colleagues to as ‘recombination centers’. Does euchromatin and associated histone marks at V(D)J rearranging loci play a more elaborate role (i.e., besides accessibility) in targeting recombinase activity to these particular sites? Insight into this issue was provided by recent studies demonstrating that the PHD finger domain of RAG2 binds with a high affinity to histone H3 trimethylated at K4 (H3K4me3) [60–62]. Moreover, H3K4me3 appears to also stimulate purified RAG enzymatic activity at both the nicking and hairpinning steps of V(D)J recombination [63]. Likewise, the N-terminal part of RAG1 contains a RING domain that preferentially interacts directly with and promotes monoubiquitylation of histone H3, an activity that could play a role in regulating the joining phase of chromosomal V(D)J recombination [64]. Overall, these data support the notion of a significant and direct impact of the chromosomal environment on V(D)J recombinase tethering and enhancement of catalytic activity. Complementary to these new concepts, the work from the Schatz laboratory mentioned above also provided evidence that whereas RAG1 binds specifically to AR gene segments in a cell-type and stage-specific manner, RAG2 displays a much larger chromosomal binding landscape, in fact interacting with H3K4me3-enriched regions genome-wide [59]. Because most cryptic RSs commonly recognized to be aberrantly utilized in chromosomal translocation in T-cell leukemia, also appear to be located nearby H3K4me3-enriched domains in normal T-cells [63], these findings may link translocation frequency with the chromatin landscape in cells undergoing V(D)J recombination. 3.3. Spatial organization within the nucleus In recent years, it became obvious that chromosomal movements and interactions in the nucleus play a crucial role in gene regulation [66]. Accordingly, studies using fluorescence in situ hybridization (FISH) have revealed large scale locus contraction and chromosomal looping as novel processes that may be involved in the developmental regulation of V(D)J recombination at AR loci (reviewed in Ref. [42]). In particular, biallelic contraction of the Tcrb 5 V and D–J–C chromosomal domains was observed in DN thymocytes, but appeared to be reversed at the next DP developmental stage [65]. Likely, the reversible juxtaposition of the 5 V␤ genes next to the D␤J␤C␤ gene clusters first contributes to facilitate V–DJ recombination while locus decontraction, in separating the two remote 5 V␤ and D␤J␤C␤ domains, may also prevent these rearrangements to continue in DP thymocytes (hence maintaining allelic exclusion, see Section 5). Of note however, locus decontraction may not be the primary process to avoid V␤ gene recombination in DP cells given all the evidence that V␤ gene segments (including V␤14) adjacent to a V␤DJ␤–C␤ rearranged domain maintain a configuration of chromosomal accessibility in DP thymocytes, yet are apparently not subjected to further rearrangement with available DJ␤ units [10,34,50]. In the same line, the multiple, independently rearranging IgH loci found in the earliest vertebrates possessing an adaptive immune system (e.g., sharks and skates) may bypass the need for reversible contraction but nonetheless display allelic exclusion in B lymphocytes [66]. In this context, FISH analyses of the subnuclear organization of AR genes have, in addition, revealed that the repositioning/tethering of these loci to the nuclear periphery results in gene transcriptional repres-

333

sion [42]. FISH-based studies applied to the analysis of the Tcrb locus in DN cells yielded controversial results though, suggesting either mono-allelic recruitment to the heterochromatin subnuclear compartment [65], or fairly high level of biallelic association with the nuclear lamina and pericentromeric compartments [67]. The reason(s) for the discrepancy between the two studies is not clear at the moment. Recently, in order to gain further dynamical insight into the spatial organization of an AR locus (in this case, the IgH locus), Jhunjhunwala et al. [68] used spectral high-precision epifluorescence microscopy assisted by computer simulations of potential configurations of chromatin structure. Comparison of the spatial distance distributions between 12 genomic markers that span the entire IgH locus to computer simulations of alternative chromatin arrangements and further trilateration and triplepoint angle measurements indicated compartmentalization and striking conformational changes involving VH and DH –JH elements during early B-cell development, the entire repertoire of VH regions appearing to juxtapose to the DH elements, mechanistically permitting long-range genomic interactions to occur with relatively high frequency [68]. Whether other AR loci are structured and behave in a similar fashion during lymphoid cell development remains to be investigated. At the Tcrb locus, it is conceivable that distant V␤ and DJ␤ regions likewise assemble into flexible chromosomal domains, which could recurrently juxtapose in early developing T-cells. It was further suggested that, once poised for V(D)J recombination, the folding of a chromatin fiber may be stabilized at AR loci by the action of “bridging” factors such as CTCF [42,69]. The DNA binding profile of CTCF along the IgH locus in B-cells tends to support this assumption [69,70].

4. Enhancer and promoter function on V(D)J recombination and chromatin remodeling at the Tcrb gene Targeted deletion of transcriptional enhancers at AR loci generally resulted in drastically reduced levels of V(D)J cis-recombination and associated gene expression at the knocked out allele [9]. Notably, the stringent T-cell phenotype (i.e., lack of ␣␤T-cells) and Tcrb gene rearrangement defect reported in E␤-deleted (E␤−/− ) mice made these animals an excellent model system to investigate the role of transcriptional enhancers of AR genes in regulating recombinational accessibility (Fig. 1, bottom insert). Indeed, detailed analysis of DN T-cells from the E␤−/− mice provided compelling evidence that E␤ orchestrates chromatin remodeling within the proximal D␤–J␤–C␤ domains via the activation of the germline promoters flanking the D␤ gene segments [33,44,46,47]. In sharp contrast, E␤ deletion seems not to affect expression and chromatin structure of both the distal 5 V␤ and proximal 3 V␤14 gene-containing domains [44]. Similar experimental approaches to analyze the V(D)J recombinational and chromatin remodeling functions of, respectively, the pD␤1 promoter and V␤13 promoter indicated that each element normally acts in cis to effect local chromatin accessibility and rearrangement of the associated gene segment without affecting those activities at neighboring D␤2–J␤2 or V␤ gene segments [47,71,72]. Complementary molecular analysis of E␤-deleted thymocytes further implied that this element critically contributes to the assembly of a functional nucleoprotein complex at pD␤1, including the loading of discrete TFs, as well as that of the Pol II and TBP components of the basal transcription machinery [46]. Such a dedicated process likely involves a physical interaction between both cis-regulatory elements, possibly contributing to the formation of a stable holocomplex [47]. Interestingly, opposite to what was observed at E␤-deleted alleles, chromosomal accessibility of the J␤1 gene segments is relatively unaffected at pD␤1-deleted alleles implying that E␤ exerts both pD␤1-dependent and -independent

334

S. Spicuglia et al. / Seminars in Immunology 22 (2010) 330–336

chromatin remodeling functions [47]. Whether E␤ may also contribute to additional regulatory processes at the Tcrb locus during rearrangement and T-cell development, such as for example locus reversible contraction, spatial organization or allelic exclusion, still needs thorough investigation. In this connection, intriguingly, we recently found that replacement of the endogenous E␤ by a conserved core enhancer fragment (E␤169 ) resulted in an extremely stochastic phenotype (in terms of GT and Tcrb gene rearrangement) when driving D␤ promoter activation in DN thymocytes, and a biased and variegated phenotype (in terms of V␤ gene segment usage and TCR␤ chain expression) when driving V␤ promoter activation in more mature T-cells (Ref. [32] and discussion herein). Ultimately, determining the enhancer precise molecular function(s) in chromatin remodeling vs. another type(s) of chromosomal process may require mutational/deletional analyses of the discrete TF binding sites lying within the E␤ regulatory modules. 5. Tcrb allelic exclusion and dynamics of Tcrb gene rearrangement 5.1. Allelic exclusion: basic features In developing lymphocytes, allelic exclusion restricts the assembly of a productively rearranged (in-frame) variable exon by V(D)J recombination to only one allele at certain AR genes. Thus, ␣␤Tcells commonly display a Tcrb genotype made of a combination of one in-frame VDJ assembled allele and either one out-of-frame VDJ assembled allele or one partially (DJ) rearranged allele [73]. The still partly understood phenomenon of allelic exclusion was proposed to involve two separate steps: (i) initiation, during which rearrangement is initiated at only one of the two homologous AR alleles; and (ii) maintenance, whereby rearrangement on the second allele is prevented by expression of an appropriate AR chain [74]. Here, we focus on recent studies aimed to decipher the basic rules that sustain the primary (initiation) phase of allelic exclusion; and could possibly rely on (some of) the molecular attributes and/or processes evoked above. Experimental results (see below) led to suggest either a deterministic (instructive) or a stochastic (probabilistic) mode of AR gene activation for V(D)J recombination to explain asynchronous rearrangement at two homologous AR alleles in individual cells. Briefly, in the former scenario, AR alleles are proposed to randomly display distinct physical properties due to differential epigenetic marking acquired during development such that one will be preferentially used for initial rearrangement. Instead, stochastic models argue that allele dissociation results from inter-allelic competition and, usually, a low probability of locus activation for V(D)J recombination. Whatever the cause, an in-frame V(D)J rearranged outcome eventually leads to the prohibition of further AR gene rearrangement via a signal conveyed from the pre-TCR or pre-BCR receptor built from the newly synthesized TCR or Ig polypeptide—a feedback inhibition that can explain maintenance of allelic exclusion. Indeed, the concept of feedback inhibition of V-to-(D)J rearrangement agrees with the ∼60–40% ratio of VDJ/DJ and VDJ/VDJ rearranged Tcrb or IgH alleles typically observed in mature T and B lymphocytes, respectively [74]. 5.2. Deterministic vs. stochastic models: experimental assets The mechanism(s) that might be involved in instigating allelic exclusion remains largely unknown. According to current understanding, it may – at V, D, J containing AR loci – preferentially impinge on the latter step of V-to-DJ gene assembly [74]. As mentioned, FISH-based studies have initially revealed a preferential mono-allelic association of Ig loci with pericentromeric hete-

rochromatin suggesting that a deterministic mechanism may in this way direct the inactivation of the non-functional allele ([75], and references therein). This has been further supported by the proposal that the IgH and Ig␬ loci frequently associate in pre-B-cell nuclei; and that the two alleles of the IgH and Ig␬ loci may also be paired at some point during B-cell differentiation in a stage-specific way ([76]; even though reassessment of the data initially presented in this study led to a less definite view, see corrigendum). Whether similar processes also occur at Tcrb alleles during T-cell development remains to be shown. However, in a separate study, Tcrb alleles were reported to instead associate – via their 5 V␤ domain – at a high frequency and presumably stochastically with repressive pericentric heterochromatin and nuclear lamina, such that only 5% of nuclei showed two alleles free of both compartments against approximately 60% still showing potentially repressive dual interactions [67]. In the latter study, evidence has also been provided suggesting that V␤–DJ␤ rearrangement occurs more effectively on non-associated alleles, thought the analysis did not permit the determination of dynamical figures, e.g., how frequently associated alleles may separate from the repressive compartments. Based on an experimental system consisting in the introduction of the GFP marker into the Ig␬ locus, it has been argued that limiting TFs randomly and infrequently activate this locus ([77]; <5% of pre-B-cells apparently expressed the marker). Such a stochastic effect has also been implied as potentially impacting on Tcrb allelic expression and rearrangement as both processes may be perturbed by gene dosage and/or deficiency of TFs such as the Helix–Loop–Helix protein E47 [78]. However, recent investigations questioning the validity of the GFP system in supplying evidence for probabilistic gene expression at Ig␬ alleles [79,80] cast doubt on these figures. In fact, as with the V␤ locus in pro-T-cells [81], Igk GT – when detected – seems to occur biallelically in single pre-B-cells [79,80,82]. Still from a stochastic point of view, allelic exclusion at the Tcrb locus might rely, at least in part, on processes that would make V␤-to-DJ␤ rearrangement less efficient as compared to D␤-to-J␤ rearrangement or, due to increased complexity, be a factor(s) that introduces important noise during the former event, therefore contributing to dissociate allelic assembly at this latter step. Increased intricacy of V␤-to-DJ␤ rearrangement as compared to that of D␤-to-J␤ may, in a non-mutually exclusive manner, involve one of the following features, some of them already mentioned earlier: (i) the partitioning of distant genomic regions via developmentally regulated chromosomal looping [65]; (ii) the break-up of V␤-privileged connection to nuclear repressive compartments [67]; (iii) overriding the relatively ineffective V␤ RSs [25,26]; and perhaps (iv) supplanting of inter-V␤ antagonisms for productive coupling with a DJ␤ unit [17,52]. In this context, it is notable that the unrearranged V␤ domains were found to be over-methylated and to frequently associate with repressive compartments in DN nuclei [44,48,67], even though paradoxically the V␤ region is biallelically expressed at this stage [81]. Moreover, engineered mice with a preassembled D␤1J␤1.1 gene segment exhibited normal Tcrb allelic exclusion [83]. 5.3. Dynamic modeling of Tcrb rearrangements and allelic exclusion Whether assuming a deterministic or a stochastic process to explain allelic exclusion, it remains unclear as to how recombination proceeds to the opposite allele in the relatively frequent cases of an out-of-frame V(D)J initial assembly. Moreover, these models fall short in terms of properly depicting the production, besides the predominant subpopulations of VDJ/DJ and VDJ/VDJ allelically excluded cells (≥95% of total ␣␤T-cells, on average), of a few atypical cells (≤5% of total ␣␤T-cells) carrying productively

S. Spicuglia et al. / Seminars in Immunology 22 (2010) 330–336

rearrangements on the two alleles (allelically included), or carrying a germline, non-rearranged allele in addition to the productively rearranged allele. These concerns, coupled with the data implying V␤ gene biallelic activation in DN cell nuclei [81], prompted us to consider a dynamical scenario for modeling Tcrb gene recombination in individual cells, which is not tied into a strictly sequential mode of inter-allelic activation for V(D)J recombination. Studies in which mathematical modeling was shown to fit the distribution of subsets of TCR or BCR-expressing T and B lymphocytes strengthen evidence for the effectiveness of stochastic-based formalisms at simulating developmental milestones in lymphoid cells [84–87]. Likewise, we thus applied a dynamical approach to model Tcrb gene recombination in an effort to comprehend the stochastic and regulated premises of allelic exclusion within a developmental scheme that would integrate all the observed cell-subsets at once [88]. Our study used a “continuous-time Markov chain”-based modeling whereby essential steps in the biological procedure (D–J and V–DJ rearrangements, and feedback inhibition) evolve independently on the two Tcrb alleles in every single cell while displaying random modes of initiation and duration. While keeping allele crosstalk to a minimum, the model comprehensively satisfied the allelic exclusion landmark; and accurately predicted experimentally compatible ratios of cell cohorts comprised of all reported Tcrb genotypes, suggesting that these basic precepts truly capture the nature of the underlying controls at the Tcrb locus. Contrary to the common notion, this dynamical, stochastic-based view of allelic exclusion does not necessarily hinge on poorly efficient, mono-allelic (and strictly sequential) recombination, as long as the inhibiting control becomes effective during the time lag otherwise separating V(D)J assembly at opposite sites. Our model thus may reveal concepts awarding a discrete chromosomal system with the property to display robust allelic exclusion at minimal regulatory cost, while keeping the opportunity to maximize genetic diversity via dual allele usage. 6. Concluding remarks Sophisticated mechanisms appear to underly the regulation of Tcrb gene recombination. New concepts, when first described at other AR loci, require further investigation regarding their generalization to the Tcrb locus. Notably, at this locus, it will be important to decipher the precise link between GT, chromatin arrangements and chomosomal organization and connection to nuclear sub-compartments; as well as the mechanisms behind the establishment and enforcement of allelic exclusion. Significant progress may come from thorough characterization of epigenetics, as well as nucleoprotein complexes bound at various domains of the Tcrb locus, ideally in a developmental order, using high-throughput technologies (ChIP-chip, ChIPseq) [32,69,78,89,90]. Given the complexity of this model system where the molecular and cellular outcome may not at first be obvious, mathematical simulations could help in connecting hypotheses with experimental observations. Competing interests The authors declare that they have no competing interests. Acknowledgements We thank members of the PF laboratory for helpful discussions; and in particular, Pierre Cauchy for critical reading and his corrections. We apologize to those whose valuable contributions to this field were only referenced in the context of published reviews, due to space limitation. Work in the PF laboratory is supported by institutional grants from Inserm and the CNRS, and by spe-

335

cific grants from the “Fondation Princesse Grace de Monaco”, the “Association pour la Recherche sur le Cancer” (ARC), the “Agence Nationale de la Recherche” (ANR), the “Institut National du Cancer” (INCa), and the Commission of the European Communities. AP was supported by a Marie Curie Research Training Network (RTN) fellowship, and is now supported by a fellowship from the “Fondation pour la Recherche Médicale” (FRM). JZ-C is supported by a fellowship from the ANR.

References [1] Jung D, Alt FW. Unraveling V(D)J recombination: insights into gene regulation. Cell 2004;116:299–311. [2] Schatz DG. V(D)J recombination. Immunol Rev 2004;200:5–11. [3] Sleckman BP. Lymphocyte antigen receptor gene assembly: multiple layers of regulation. Immunol Res 2005;32:253–8. [4] Perlot T, Alt FW. Cis-regulatory elements and epigenetic changes control genomic rearrangements of the IgH locus. Adv Immunol 2008;99:1–32. [5] Schlissel MS. Epigenetics drives RAGs to recombination riches. Cell 2010;141:400–2. [6] Ferrier P. V(D)J recombination. Adv Exp Med Biol 2009. Landes Bioscience. [7] Bonnet M, Ferrier P, Spicuglia S. Molecular genetics at the T-cell receptor ␤ locus: insights into the regulation of V(D)J recombination. Adv Exp Med Biol 2009;650:116–32. [8] Krangel MS. Mechanics of T cell receptor gene rearrangement. Curr Opin Immunol 2009;21:133–9. [9] Thomas LR, Cobb RM, Oltz EM. Dynamic regulation of antigen receptor gene assembly. Adv Exp Med Biol 2009;650:103–15. [10] Jackson AM, Krangel MS. Turning T-cell receptor ␤ recombination on and off: more questions than answers. Immunol Rev 2006;209:129–41. [11] Garbe AI, Von Boehmer H. TCR and Notch synergize in ␣␤ versus ␥␦ lineage choice. Trends Immunol 2007;28:124–31. [12] Taghon T, Rothenberg EV. Molecular mechanisms that control mouse and human TCR␣␤ and TCR␥␦ T cell development. Semin Immunopathol 2008;30:383–98. [13] Singer A, Adoro S, Park JH. Lineage fate and intense debate: myths, models and mechanisms of CD4- versus CD8-lineage choice. Nat Rev Immunol 2008;8:788–801. [14] Glusman G, Rowen L, Lee I, Boysen C, Roach JC, Smit AF, et al. Comparative genomics of the human and mouse T cell receptor loci. Immunity 2001;15:337–49. [15] McMillan RE, Sikes ML. Differential activation of dual promoters alters D␤2 germline transcription during thymocyte development. J Immunol 2008;180:3218–28. [16] McMillan RE, Sikes ML. Promoter activity 5 of Dbeta2 is coordinated by E47, Runx1, and GATA-3. Mol Immunol 2009;46:3009–17. [17] Bassing CH, Whitlow S, Mostoslovasky R, Yang-Iott K, Ranganath S, Alt FW. V␤ cluster sequences reduce the frequency of primary V␤2 and V␤14 rearrangements. Eur J Immunol 2008;38:2564–72. [18] Khor B, Mahowald GK, Khor K, Sleckman BP. Functional overlap in the cisacting regulation of the V(D)J recombination at the TCR␤ locus. Mol Immunol 2009;46:321–6. [19] Ramsden DA, Baetz K, Wu GE. Conservation of sequence in recombination signal sequence spacers. Nucleic Acids Res 1994;22:1785–96. [20] Feeney AJ, Goebel P, Espinoza CR. Many levels of control of V gene rearrangement frequency. Immunol Rev 2004;200:44–56. [21] Gerstein RM, Lieber MR. Coding end sequence can markedly affect the initiation of V(D)J recombination. Genes Dev 1993;7:1459–69. [22] Lee AI, Fugmann SD, Cowell LG, Ptaszek LM, Kelsoe G, Schatz DG. A functional analysis of the spacer of V(D)J recombination signal sequences. PLoSBiol 2003;1:E1. [23] Lieber MR, Chang CP, Gallo M, Gauss G, Gerstein R, Islas A. The mecanism of V(D)J recombination: site-specificity, reaction fidelity and immunologic diversity. Semin Immunol 1994;6:143–53. [24] Livak F, Burtrum DB, Rowen L, Schatz DG, Petrie HT. Genetic modulation of T cell receptor gene segment usage during somatic recombination. J Exp Med 2000;192:1191–6. [25] Liang HE, Hsu LY, Cado D, Cowell LG, Kelsoe G, Schlissel MS. The “dispensable” portion of RAG2 Is necessary for efficient V-to-DJ rearrangement during B and T Cell development. Immunity 2002;17:639–51. [26] Cowell LG, Davila M, Yang K, Kepler TB, Kelsoe G. Prospective estimation of recombination signal efficiency and identification of functional cryptic signals in the genome by statistical modeling. J Exp Med 2003;197:207–20. [27] Franchini DM, Benoukraf T, Jaeger S, Ferrier P, Payet-Bornet D. Initiation of V(D)J recombination by D␤-associated recombination signal sequences: a critical control point in TCR␤ gene assembly. PLoS One 2009;4:e4575. [28] Bassing CH, Alt FW, Hughes MM, D’Auteuil M, Wehrly TD, Woodman BB, et al. Recombination signal sequences restrict chromosomal V(D)J recombination beyond the 12/23 rule. Nature 2000;405:583–6. [29] Tillman RE, Wooley AL, Hughes MM, Khor B, Sleckman BP. Regulation of T-cell receptor ␤-chain gene assembly by recombination signals: the beyond 12/23 restriction. Immunol Rev 2004;200:36–43.

336

S. Spicuglia et al. / Seminars in Immunology 22 (2010) 330–336

[30] Wang X, Xiao G, Zhang Y, Wen X, Gao X, Okada S, et al. Regulation of Tcr␤ recombination ordering by c-Fos-dependent RAG deposition. Nat Immunol 2008;9:794–801. [31] King AG, Kondo M, Scherer DC, Weissman IL. Lineage infidelity in myeloid cells with TCR gene rearrangement: a latent developmental potential of proT cells revealed by ectopic cytokine receptor signaling. Proc Nat Acad Sci USA 2002;99:4508–13. [32] Bonnet M, Huang F, Benoukraf T, Cabaud O, Verthuy C, Boucher A, et al. Duality of enhancer functioning mode revealed in a reduced TCR␤ gene enhancer knockin mouse model. J Immunol 2009;183:7939–48. [33] Kondilis-Mangum HD, Cobb RM, Osipovich O, Srivatsan S, Oltz EM, Krangel MS. Transcription-dependent mobilization of nucleosomes at accessible TCR gene segments in vivo. J Immunol 2010;184:6970–7. [34] Yang-Iott KS, Carpenter AC, Rowh MA, Steinel N, Brady BL, Hochedlinger K, et al. TCR␤ feedback signals inhibit the coupling of recombinationally accessible V␤14 segments with DJ␤ complexes. J Immunol 2010;184:1369–78. [35] Hughes MM, Yassai M, Sedy JR, Wehrly TD, Huang CY, Kanagawa O, et al. T cell receptor CDR3 loop length repertoire is determined primarily by features of the V(D)J recombination reaction. Eur J Immunol 2003;33:1568–75. [36] Yancopoulos GD, Alt FW. Developmentally controlled and tissue-specific expression of unrearranged V H gene segments. Cell 1985;40:271–81. [37] Yancopoulos GD, Blackwell TK, Suh H, Hood L, Alt FW. Introduced T cell receptor variable region gene segments recombine in pre-B cells: evidence that B and T cells use a common recombinase. Cell 1986;44:251–9. [38] Krangel MS. T cell development: better living through chromatin. Nat Immunol 2007;7:687–94. [39] Stanhope-Baker P, Hudson KM, Shaffer AL, Constantinescu A, Schlissel MS. Cell type-specific chromatin structure determines the targeting of V(D)J recombinase activity in vitro. Cell 1996;85:887–97. [40] McMurry MT, Krangel MS. A role for histone acetylation in the developmental regulation of VDJ recombination. Science 2000;287:495–8. [41] Morshead KB, Ciccone DN, Taverna SD, Allis CD, Oettinger MA. Antigen receptor loci poised for V(D)J rearrangement are broadly associated with BRG1 and flanked by peaks of histone H3 dimethylated at lysine 4. Proc Natl Acad Sci USA 2003;100:11577–82. [42] Jhunjhunwala S, van Zelm MC, Peak MM, Murre C. Chromatin architecture and the generation of antigen receptor diversity. Cell 2009;138:435–48. [43] Chattopadhyay S, Whitehurst CE, Schwenk F, Chen J. Biochemical and functional analyses of chromatin changes at the TCR␤ gene locus during CD4−CD8− to CD4+CD8+ thymocyte differentiation. J Immunol 1998;160:1256–67. [44] Mathieu N, Hempel WM, Spicuglia S, Verthuy C, Ferrier P. Chromatin remodeling by the T cell receptor (TCR)-␤ gene enhancer during early T cell development: implications for the control of TCR-␤ locus recombination. J Exp Med 2000;192:625–36. [45] Ryu CJ, Haines BB, Draganov DD, Kang YH, Whitehurst CE, Schmidt T, et al. The T cell receptor ␤ enhancer promotes access and pairing of D␤ and J␤ gene segments during V(D)J recombination. Proc Nat Acad Sci USA 2003;100: 13465–70. [46] Spicuglia S, Kumar S, Yeh JH, Vachez E, Chasson L, Gorbatch S, et al. Promoter activation by enhancer-dependent and -independent loading of activator and coactivator complexes. Mol Cell 2002;10:1479–87. [47] Oestreich KJ, Cobb RM, Pierce S, Chen J, Ferrier P, Oltz EM. Regulation of TCR␤ gene assembly by a promoter/enhancer holocomplex. Immunity 2006;24:381–91. [48] Senoo M, Shinkai Y. Regulation of V␤ germline transcription in RAG-deficient mice by the CD3epsilon-mediated signals: implication of V␤ transcriptional regulation in TCR␤ allelic exclusion. Int Immunol 1998;10:553–60. [49] Tripathi R, Jackson A, Krangel MS. A change in the structure of V␤ chromatin associated with TCR␤ allelic exclusion. J Immunol 2002;168:2316–24. [50] Mathieu N, Spicuglia S, Gorbatch S, Cabaud O, Fernex C, Verthuy C, et al. Assessing the role of the T cell receptor ␤ gene enhancer in regulating coding joint formation during V(D)J recombination. J Biol Chem 2003;278:18101–9. [51] Jackson AM, Krangel MS. Allele-specific regulation of TCR␤ variable gene segment chromatin structure. J Immunol 2005;175:5186–91. [52] Ranganath S, Carpenter AC, Gleason M, Shaw AC, Bassing CH, Alt FW. Productive coupling of accessible V␤14 segments and DJ␤ complexes determines the frequency of V␤14 rearrangement. J Immunol 2008;180:2339–46. [53] Xu CR, Feeney AJ. The epigenetic profile of Ig genes is dynamically regulated during B cell differentiation and is modulated by pre-B cell receptor signaling. J Immunol 2009;182:1362–9. [54] Abarrategui I, Krangel MS. Noncoding transcription controls downstream promoters to regulate T-cell receptor ␣ recombination. EMBO J 2007;26:4380–90. [55] Bolland DJ, Wood AL, Corcoran AE. Large-scale chromatin remodeling at the immunoglobulin heavy chain locus: a paradigm for multigene regulation. Adv Exp Med Biol 2009;650:59–72. [56] Abarrategui I, Krangel MS. Germline transcription: a key regulator of accessibility and recombination. Adv Exp Med Biol 2009;650:93–102. [57] Zhao Y, Dunn-Walters DK, Barone F, Spencer J. Antisense transcripts of V(D)J rearrangements: artifacts caused by false priming? Mol Immunol 2009;46:2357–62. [58] Perlot T, Li G, Alt FW. Antisense transcripts from immunoglobulin heavy-chain locus V(D)J and switch regions. Proc Natl Acad Sci USA 2008;105:3843–8. [59] Ji Y, Resch W, Corbett E, Yamane A, Casellas R, Schatz DG. The in vivo pattern of binding of RAG1 and RAG2 to antigen receptor loci. Cell 2010. [60] Liu Y, Subrahmanyam R, Chakraborty T, Sen R, Desiderio S. A plant homeodomain in RAG-2 that binds hypermethylated lysine 4 of histone H3

[61]

[62]

[63]

[64]

[65]

[66] [67]

[68]

[69]

[70]

[71]

[72]

[73] [74] [75] [76]

[77]

[78]

[79]

[80]

[81] [82] [83]

[84] [85]

[86] [87]

[88]

[89]

[90]

is necessary for efficient antigen-receptor-gene rearrangement. Immunity 2007;27:561–71. Matthews AG, Kuo AJ, Ramon-Maiques S, Han S, Champagne KS, Ivanov D, et al. RAG2 PHD finger couples histone H3 lysine 4 trimethylation with V(D)J recombination. Nature 2007;450:1106–10. Ramon-Maiques S, Kuo AJ, Carney D, Matthews AG, Oettinger MA, Gozani O, et al. The plant homeodomain finger of RAG2 recognizes histone H3 methylated at both lysine-4 and arginine-2. Proc Natl Acad Sci USA 2007;104:18993–8. Shimazaki N, Tsai AG, Lieber MR. H3K4me3 stimulates the V(D)J RAG complex for both nicking and hairpinning in trans in addition to tethering in cis: implications for translocations. Mol Cell 2009;34:535–44. Grazini U, Zanardi F, Citterio E, Casola S, Goding CR, McBlane F. The RING domain of RAG1 ubiquitylates histone H3: a novel activity in chromatin-mediated regulation of V(D)J joining. Mol Cell 2010;37:282–93. Skok JA, Gisler R, Novatchkova M, Farmer D, de LW, Busslinger M. Reversible contraction by looping of the Tcr␣ and Tcr␤ loci in rearranging thymocytes. Nat Immunol 2007;8:378–87. Malecek K, Lee V, Feng W, Huang JL, Flajnik MF, Ohta Y, et al. Immunoglobulin heavy chain exclusion in the shark. PLoSBiol 2008;6:e157. Schlimgen RJ, Reddy KL, Singh H, Krangel MS. Initiation of allelic exclusion by stochastic interaction of Tcr␤ alleles with repressive nuclear compartments. Nat Immunol 2008;9:802–9. Jhunjhunwala S, van Zelm MC, Peak MM, Cutchin S, Riblet R, van Dongen JJ, et al. The 3D structure of the immunoglobulin heavy-chain locus: implications for long-range genomic interactions. Cell 2008;133:265–79. Degner SC, Wong TP, Jankevicius G, Feeney AJ. Cutting edge: developmental stage-specific recruitment of cohesin to CTCF sites throughout immunoglobulin loci during B lymphocyte development. J Immunol 2009;182:44–8. Featherstone K, Wood AL, Bowen AJ, Corcoran AE. The mouse immunoglobulin heavy chain V–D intergenic sequence contains insulators that may regulate ordered V(D)J recombination. J Biol Chem 2010;285:9327–38. Whitehurst CE, Schlissel MS, Chen J. Deletion of germline promoter PD␤1 from the TCR␤ locus causes hypermethylation that impairs D␤1 recombination by multiple mechanisms. Immunity 2000;13:703–14. Ryu CJ, Haines BB, Lee HR, Kang YH, Draganov DD, Lee M, et al. The T-cell receptor ␤ variable gene promoter is required for efficient V␤ rearrangement but not allelic exclusion. Mol Cell Biol 2004;24:7015–23. Khor B, Sleckman BP. Allelic exclusion at the TCR␤ locus. Curr Opin Immunol 2002;14:230–4. Mostoslavsky R, Alt FW, Rajewsky K. The lingering enigma of the allelic exclusion mechanism. Cell 2004;118:539–44. Cedar H, Bergman Y. Choreography of Ig allelic exclusion. Curr Opin Immunol 2008;20:308–17. Hewitt SL, Yin B, Ji Y, Chaumeil J, Marszalek K, Tenthorey J, et al. RAG1 and ATM coordinate monoallelic recombination and nuclear positioning of immunoglobulin loci. Nat Immunol 2009;10:655–64. Published erratum appeared in 2009 Nat. Immunol. 2010:1034 and 2010 Nat. Immunol. 2011:2355–2356. Liang HE, Hsu LY, Cado D, Schlissel MS. Variegated transcriptional activation of the immunoglobulin kappa locus in pre-b cells contributes to the allelic exclusion of light-chain expression. Cell 2004;118:19–29. Agata Y, Tamaki N, Sakamoto S, Ikawa T, Masuda K, Kawamoto H, et al. Regulation of T cell receptor ␤ gene rearrangements and allelic exclusion by the helix–loop–helix protein, E47. Immunity 2007;27:871–84. Amin RH, Cado D, Nolla H, Huang D, Shinton SA, Zhou Y, et al. Biallelic, ubiquitous transcription from the distal germline Ig␬ locus promoter during B cell development. Proc Natl Acad Sci USA 2009;106:522–7. Taylor B, Cobb BS, Bruno L, Webster Z, Fisher AG, Merkenschlager M. A reappraisal of evidence for probabilistic models of allelic exclusion. Proc Natl Acad Sci USA 2009;106:516–21. Jia J, Kondo M, Zhuang Y. Germline transcription from T-cell receptor V␤ gene is uncoupled from allelic exclusion. EMBO J 2007;26:2387–99. Singh N, Bergman Y, Cedar H, Chess A. Biallelic germline transcription at the k immunoglobulin locus. J Exp Med 2003;197:743–50. Carpenter AC, Yang-Iott KS, Chao LH, Nuskey B, Whitlow S, Alt FW, et al. Assembled DJ␤ complexes influence TCR␤ chain selection and peripheral V␤ repertoire. J Immunol 2009;182:5586–95. Sepulveda N, Boucontet L, Pereira P, Carneiro J. Stochastic modeling of T cell receptor ␥ gene rearrangement. J Theor Biol 2005;234:153–65. Piper H, Litwin S, Mehr R. Models for antigen receptor gene rearrangement. II. Multiple rearrangement in the TCR: allelic exclusion or inclusion? J Immunol 1999;163:1799–808. Kalmanovich G, Mehr R. Models for antigen receptor gene rearrangement. III. Heavy and light chain allelic exclusion. J Immunol 2003;170:182–93. Mehr R, Shannon M, Litwin S. Models for antigen receptor gene rearrangement. I. Biased receptor editing in B cells: implications for allelic exclusion. J Immunol 1999;163:1793–8. Farcot E, Bonnet M, Jaeger S, Spicuglia S, Fernandez B, Ferrier P. T-cell Receptor (TCR) ␤ allelic exclusion in dynamical models of V(D)J recombination based on allele independence. J Immunol 2010;185:1622–32. Xu CR, Schaffer L, Head SR, Feeney AJ. Reciprocal patterns of methylation of H3K36 and H3K27 on proximal vs. distal IgVH genes are modulated by IL-7 and Pax5. Proc Natl Acad Sci USA 2008;105:8685–90. Malin S, McManus S, Cobaleda C, Novatchkova M, Delogu A, Bouillet P, et al. Role of STAT5 in controlling cell survival and immunoglobulin gene recombination during pro-B cell development. Nat Immunol 2010;11:171–81.