V(D)J Recombination: A Paradigm for Studying Chromosome Interactions in Mammalian Cells

It has been known for some time that chromatin modifications can condense or open up a locus to the transcriptional machinery and other factors [1]. In recent years, however, the study of gene regulation has ventured beyond the molecular level to explore the effects of nuclear organization [2]. Chromosomes can acquire conformations that are more or less conducive to specific activities; for example, the formation of loops can bring widely separated enhancers and promoters into contact. Beyond chromatin and chromosome conformation, there is also the possibility of moving a chromosome within the nuclear space. A locus that is to be silenced might move toward the nuclear periphery or pericentromeric heterochromatin, both of which are considered repressive regions [3–6]. Chromosomes can also interact with each other to regulate gene expression, even in mammalian cells (until recently, this phenomenon had been thought to be confined largely to plants, yeast and flies) [7]. Some valuable insights into chromosomal interactions in mammalian cells have come through studies of variable (V), diversity (D) and joining (J) segments in V(D)J recombination, which requires precise spatiotemporal control to avoid aberrant chromosome rearrangements. In this brief editorial, I would like to focus on what we‘ve learned (so far) from studying the various chromosomal interactions used by developing lymphocytes to orchestrate this very complex process.

Developing B and T cells must produce receptors capable of binding to virtually any antigen. To generate such a varied receptor repertoire with limited genetic material, the lymphoid-specific proteins, RAG1 and RAG2, together select V, D and J segments arrayed along a given locus, excise the intervening DNA, and repair the chromosome again by joining the segments to one another through the nonhomologous end-joining pathway. Thus, V(D)J recombination creates an entirely novel gene that will, in turn, encode a unique antigen receptor. The process itself, although relatively straightforward, is governed by multiple interdependent layers of regulation that turn out to be quite complex.

First, the gene segments are arrayed along different chromosomes in seven loci, some of which recombine in B cells (the immunoglobulin or Ig loci) and others in T cells (the T cell receptor or Tcr loci). Although the recombination machinery is the same in both lineages, only the Ig loci recombine in B cells, and only the Tcr loci combine in T cells. This restriction is known as lineage specificity. Second, the order of the stepwise rearrangement process is such that D and J must recombine before a V segment can be joined to the new DJ combined segment. The concept of ordered rearrangement also applies to the loci as a whole, which undergo recombination in a precise sequence: in B cells, for example, Igh is rearranged first before rearrangement of Igk,and Tcrb is rearranged prior to Tcra in T cells. Third, there is allelic exclusion – although there are two alleles in each cell, only one can be productively rearranged, so that each lymphocyte expresses a receptor which has specificity to a single antigen. The key to coordinating these three major regulatory controls is to make the loci inaccessible to the RAG recombinase except during very specific developmental windows [8].

Chromosome self-association: looping & contraction

Some of the antigen receptor loci are quite long – the Ig heavy chain (Igh) locus, for example, spans approximately 3 Mb in the mouse. How can the RAG proteins synapse V and D segments that are so far apart? Several years ago, Kosak et al. noticed that, early in B-cell development, the Ighlocus appears to contract [3]. We and others subsequently established that the chromatin at the Ighlocus forms loops [9,10]; Jhunjhunwala et al. recently used high-precision epifluorescence microscopy and mathematical modeling to determine that the V genes are folded into 1 Mb compartments that are gathered into a rosette of loops connected by linkers [11]. By contrast, the DJ region is able to move about more freely in search of contacts with V genes. This ‘multiloop subcompartment‘ topology has implications for the folding of chromatin in general.

The Igh locus is one of the largest loci, but not the only one that contracts and decontracts. We discovered that Igk, Tcrb and Tcra all undergo reversible locus contraction during recombination [10,12]. We also found that Pax5 triggers this conformational change in Igh so that mid and distal V gene segments, which can be more than 2 Mb away from the D segments, are brought within recombining distance [13]. The factors governing the looping or contracting of the other loci are yet to be identified. Intrachromosomal interactions contribute to the regulation of many genes; it is likely that both unique and common components collaborate to bring about these interactions. For example, cohesin, which mediates sister chromatid cohesion, is involved in loop formation within the Ifng locus [14].

Communication between alleles: homologous pairing

At the Igh locus, D–J recombination occurs on both alleles before V–DJ recombination can begin. Clearly, this requires that the V region remains inaccessible to the recombinase while the D and J regions are in an open chromatin configuration, but the mechanics of this differential accessibility are enigmatic. Reasoning that the beginning of recombination on both alleles prior to allelic exclusion could indicate communication between alleles, we examined the positions of the alleles at each stage of B-cell development to trace their movements, if any, as different loci undergo recombination. We discovered that homologous alleles form pairs at the stage they recombine [15]. Pairing at the different stages is mediated by RAG protein expression; even a catalytically inactive RAG enzyme can induce pairing. Cross-talk between distinct regions of the locus at different stages of rearrangement – remember that part of the locus is deleted by recombination, so an insulator could be removed with the rest of the DNA – could help ensure that recombination takes place in the proper order, so that D–J rearrangement occurs on both alleles before V gene rearrangement begins. In fact, we found that Igh pairing occurs during D–J recombination and then again during V–DJ recombination; in the latter case, pairing is dependent on Pax5 and, we suspect, an alteration in locus conformation.

Homologous pairing also plays a part in initiating allelic exclusion [15,16]. The introduction of DNA breaks on one of the paired alleles induces the other allele to move towards pericentromeric heterochromatin. Transient repositioning of the uncleaved allele to this repressive compartment of the nucleus is likely to inhibit further RAG-mediated cleavage while the first allele is being repaired; in the absence of the DNA damage signaling protein ATM, the two alleles remain euchromatic and cleavage occurs on both. Chromosomal translocations also occur, in accordance with previous studies of ATM-deficient mice [17–19]. ATM recruited to the site of the break somehow triggers this movement to heterochromatin, and further studies will be needed to elucidate this effect.

Pairing also contributes to allelic exclusion during Igk recombination, but unlike Igh, Igk pairing repositions one allele at pericentromeric heterochromatin before cleavage. However, ATM functions similarly at both loci it helps maintain the location of the uncleaved allele at pericentromeric heterochromatin and inhibits simultaneous biallelic cleavage. Why is Igk treated differently than Igh? Perhaps this is because the repositioning of Igk at pericentromeric heterochromatin performs a dual function in regulating both Igh and Igk loci, as we shall see in the next section.

Communication between different loci

How is rearrangement coordinated at the pre-B cell stage so that Igh rearrangement finishes when Igk rearrangement starts? At the Igh locus, loss of intrachromosomal connections and altered locus accessibility occur as a result of an inter-chromosomal interaction with pericentromerically located Igk alleles. Igk (located on chromosome 6) directs the unrearranged Igh allele (located on chromosome 12) to pericentromeric heterochromatin, where the two loci associate, mediated by the 3´ enhancer (3´Ek) of the Igk locus. Repositioning of the unrearranged Igh locus at pericentromeric heterochromatin limits access to RAG proteins, while decontraction physically prevents further V gene rearrangement [20].

In summary, intrachromsomal associations enable synapse formation between widely separated gene segments; interactions between homologous alleles mediate ordered rearrangement and help initiate allelic exclusion. Finally, the Igk locus interacts with the Igh locus to coordinate the transition from Igh to Igk rearrangement and thus from one stage of B-cell development to the next. In mammalian cells, there are only a few reports of interactions between homologous or heterologous loci, but to find the sorts of transient interactions we describe here took careful analysis of consecutive developmental stages in wild-type and gene-targeted mice. We anticipate that the regulation of many genes will turn out to involve passing connections. Genome-wide techniques, such as chromosome conformation capture, will broaden our knowledge in this area; such unbiased approaches should educate us about even the most unlikely partnerships.

Acknowledgements

The author would like to thank V L Brandt for helpful discussions and editorial input.

Financial & competing interests disclosure

The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.