How Chromatin Scanning Generates Diverse Antibody Repertoires: New Insights

By Paul Guttry

Researchers in the laboratory of Frederick Alt of the Howard Hughes Medical Institute and Program in Cellular and Molecular Medicine (PCMM) at Children's Hospital Boston made another groundbreaking discovery at the nexus of chromatin biology and immunology. Their work, published today in Nature (Dai et al., “Loop extrusion mediates physiological Igh locus contraction for RAG scanning”), showed that physiological deregulation of the WAPL chromatin-maintenance factor allows developing B lymphocytes to scan linearly through very long chromosomal loops to find and join gene segments that form diverse antibody repertoires.

Recently the Alt group discovered that V(D)J recombination and IgH class switch recombination (CSR), two discrete mechanisms of programmed genetic rearrangement in lymphocytes, both critical for adaptive immunity, are mechanistically driven by a basic process associated with management of the architecture of chromosomal genomes: chromatin loop extrusion (“Chromatin loops unlock antibody class switching”).

V(D)J recombination is initiated by the lymphocyte-specific RAG endonuclease. RAG creates antibody diversity by mediating recombination of germline-encoded V, D, and J gene segments of immunoglobulin heavy chain (IgH) gene loci into V(D)J exons that encode the antigen (pathogen)-binding regions of IgH chains. RAG has two active sites that bind specific target DNA recombination signal sequences (RSSs) that flank V, D, and J segments. Bona fide RSSs have a long DNA code that allows a complementary pair to be cleaved by RAG, when each is bound to one of the two RAG active sites.

The binding of two RSS sequences , which are embedded over long chromosomal distances, to RAG was long thought to occur by diffusion-like mechanisms. But that view changed five or so years ago with discoveries by the Alt lab of linear RAG chromatin scanning, which harnesses a process that forms long chromatin loop domains.

The concept of genomic chromatin looping was introduced by a number of labs some years ago, along with the notion that such a process would increase the likelihood that distant portions of chromosomes could physically interact, facilitating cooperative processes such as transcription among distant regulatory elements.

The CTCF chromatin looping factor plays a wide range of cellular/genetic roles, mostly related to regulating the 3D chromatin domains; e.g., as an insulator to prevent unwanted interactions between genes. Indeed, many chromatin loop domains are anchored by CTCF-binding elements (CBEs) bound by CTCF.

Notably, the CBE loop anchors that form the majority of these domains are most frequently in convergent orientation (pointing at each other), even though they can be mega-bases apart in the linear genome, separated by CBE elements in other orientations, i.e., facing different directions. Based on this finding, the groups of E. Lieberman-Aiden (Baylor College of Medicine) and L. Mirney (MIT) proposed the intriguing cohesin-mediated loop extrusion model to explain how such widely separated sequences could find each other in an orientation-dependent manner. More recently, others demonstrated that cohesin can extrude loops of naked (non-chromatinized) DNA in vitro, supporting this proposed cohesin activity.

Contemporaneously with the loop-extrusion models, the Alt lab discovered that RAG linearly explores convergent CBE-anchored chromatin loops in non-antigen receptor loci from ectopically introduced recombination centers. For this work, they employed a tremendously sensitive V(D)J recombination assay they invented to discover that RAG utilized, at low levels, hundreds of cryptic RSSs (sequences containing a subset of RSSs) along such domains to form aberrant RAG-initiated V(D)J recombination-related joining events with an initiating bona fide RSS.

They further found that the orientation of initiating bona fide RSSs programmed RAG activity linearly in one direction or the other, over distances up to Mbs. This was revealed by long stretches of cryptic RSS-mediated joins, until convergent CBE loop anchors were reached; exploration terminated there, and deletion of those convergent loop anchors allowed RAG to continue exploring the now-enlarged domains.

Moreover, the cryptic RSSs used were nearly always in convergent orientation with the bona fide RSS to which they joined.

Parallels between Alt lab findings regarding RAG chromatin exploration and the cohesin-mediated loop extrusion models laid a foundation for the Alt lab discovery of fundamental roles of cohesin-mediated loop extrusion in lymphocyte-specific recombination processes.

Let us focus more closely on IgH locus V(D)J recombination, which proceeds developmentally according to strict rules.  The Vs, Ds, and Js are encoded in order in separate clusters in the IgH chromosomal locus.  Upon binding a JH-recombination signal sequence (RSS) within a chromatin-based recombination center (RC), RAG scans upstream chromatin presented by cohesin-mediated loop extrusion for D-RSSs to initiate DJH formation.  Importantly for our story, RAG hunts specifically for convergent D-RSSs, meaning those directionally “face to face” with the JH-RSS. The Alt lab created a V(D)J recombination video that is extremely helpful in visualizing this process (V(D)J recombination animation).

The DJH intermediate now becomes a new recombination center poised for joining to upstream VHs, which are all oriented to join by deletion. However, the mechanism by which VHs embedded within the upstream 2.4-Mb VH locus access the DJH recombination center remained a major question.

Two CBEs between the VH and DH loci are key elements of a regulatory region termed IGCR1, which was shown about 10 years ago by the Alt lab to insulate the VH portion of the locus from the D and JH portion to allow the latter to recombine first.  How this barrier could be developmentally neutralized to provide RC access to the VHs was unknown until recent work in the Alt lab. 

The majority of the 109 VHs were known to lie in a 2.4-Mb region containing numerous CTCF-bound CBEs that could also impede cohesin-mediated loop extrusion.  Thus, the mechanism by which VHs embedded within that upstream region access the DJH recombination center remained a major, perplexing question. 

Primary progenitor B cells in the bone marrow were known to undergo a physical VH locus contraction, of unknown mechanism, capable of bringing VHs into proximity to the DJH recombination center.  The Alt lab hypothesized that cohesin-mediated loop extrusion may contributes to both VH locus contraction and distal VH scanning through a progenitor B cell-specific mechanism that would involve large-scale neutralization of the 100s of potential CBE impediments across the 2.4-Mb IgH locus.  As this represented a significant departure from standard thinking in the developmental and chromatin biology fields, researchers in the Alt lab decided to test it using several different approaches. 

For decades the Alt lab has used Abelson murine leukemia virus (“v-Abl”)-transformed progenitor B cell lines to study V(D)J recombination.  Upon arrest in the G1-cell cycle stage, those cells express RAG and undergo robust D-to-JH rearrangement, but little VH-to -DJH rearrangement; their distal VH locus is uncontracted and so cannot interact with the recombination center.  The Alt lab surmised that these cells could be an ideal system in which to assess whether neutralizing CBE impediments via depleting the CTCF factor could activate VH locus contraction and/or long-range RAG scanning.

By depleting CTCF in v-Abl lines, the Alt lab  provided proof-of-principle that down-modulation of CBE impediments activates locus contraction and long-range VH recombination ("Chromatin regulation enables generation of diverse antibodies").  Thus, the next challenge, which is directly addressed in the current Nature study, was to determine 1) whether RAG scanning underlies long-range VHs rearrangement in primary bone marrow progenitor B cells and; if so, 2) the mechanism by which CBE impediments are neutralized during progenitor B-cell development.

In the current Nature Article, Haiqiang Dai and his colleagues in the Alt lab addressed this question by creating a 2.4-Mb VH locus inversion in mice. They found that this inversion abrogated rearrangements involving both bona fide VH-RSSs and normally convergent VH locus cryptic RSSs, even though VH locus contraction still occurred.  Moreover, this large-scale inversion activated utilization of cryptic VH-locus RSSs normally in opposite orientation.  Together, these findings clearly demonstrated that in normal primary pro-B cells, recombination center-bound RAG can linearly scan segments across the long VH locus, indicating that upstream CBE impediments must be down-regulated.  

Equally remarkably, the current Nature Article further indicated that when the VH locus was inverted (which inactivated the robust bona fide VHRSS targets within it), RC-center based RAG scanning proceeded through the VH locus and then through multiple upstream convergent-CBE domains to the telomere.  This latter finding yielded the insight that primary pro-B cells must broadly deregulate CBE impediments, beyond those of the IgH locus, to extend recombination center-based RAG scanning across the VH locus.

The group then searched for potential down-regulation of cohesin complex factors in progenitor B cells, as opposed to v-Abl transformed pro-B cells, which lack VH locus contraction and RAG-scanning.  This search revealed a likely culprit to be the wings apart-like (WAPL) cohesin complex unloading factor, based on its low expression, compared to other cohesin complex factors, in primary progenitor B cells versus v-Abl-transformed progenitor B cell lines.  A contemporaneous study by the group of M. Busslinger (IMP, Vienna reached related conclusions (Hill et al. Nature, 2020).

To confirm that WAPL down-regulation activates RAG scanning and locus contraction, Dai et al. depleted WAPL in v-Abl-pro-B cells, which provided direct evidence that down-modulation of WAPL levels activates both processes.  Moreover,  when the VH locus was inverted in WAPL-depleted v-Abl lines, they saw all the same effects on locus contraction and RAG-scanning activity they had found in primary pro-B cells with an inverted VH locus. 

Dai et al. speculated that the mechanism by which WAPL down-regulation achieves such broad deregulation of loop-extrusion-mediated VH locus contraction and RAG-scanning likely relates to prior findings by others — that WAPL-depletion in non-lymphoid cells makes it more likely that loops will be extruded past CTCF-bound CBE anchors.  If less cohesin is unloaded, more cohesin remains on the chromosome, increasing the chances that dynamic CBE obstacles will be passed by when openings occur. 

This study also provided evidence that, after WAPL-down-regulation, remaining CBE activity that persists, as well as local  transcription, can slow down scanning to help direct RAG activity to many VHs along the scanning path opened up by WAPL depletion.

Finally, the Dai et al. paper intriguingly suggests that much more is to be learned. They found that WAPL levels sufficient to impede IgH VH locus contraction and VH utilization allow normal patterns of V(D)J recombination across the equally long Igk light-chain locus, suggesting potentially differential mechanisms for mediating long-range V(D)J recombination in the IgH and Igk loci . Ongoing studies will further elucidate this new, long-range chromatin-based mechanism of developmental gene regulation.