Molecular Mechanisms of DNA Sequence Selectivity in V(D)J Recombination

Abstract

Antigen receptor (AgR) diversity is central to the ability of adaptive immunity in jawed vertebrates to protect against pathogenic agents. The production of highly diverse AgR repertoires is initiated during B and T cell lymphopoiesis by V(D)J recombination, which assembles the receptor genes from component gene segments in a cut-and-paste recombination reaction. Recombination activating proteins, RAG1 and RAG2 (RAG1/2), catalyze V(D)J recombination by cleaving adjacent to recombination signal sequences (RSSs) that flank AgR gene segments. Previous studies defined the consensus RSS as containing conserved heptamer and nonamer sequences separated by a less conserved 12 or 23 base-pair spacer sequence. However, many RSSs deviate from the consensus sequence, and the molecular mechanism for semiselective V(D)J recombination specificity is unknown. The modulation of chromatin structure during V(D)J recombination is essential in the formation of diverse AgRs in adaptive immunity while also reducing the likelihood for off-target recombination events that can result in chromosomal aberrations and genomic instability. Here we review what is presently known regarding mechanisms that facilitate assembly of RAG1/2 with RSSs, the ensuing conformational changes required for DNA cleavage activity, and how the readout of the RSS sequence affects reaction efficiency.

Figure 1

(A) Schematic representation of a simplified AgR light chain locus undergoing V(D)J recombination. With upstream, dark-blue V-gene segments flanked by pink 23-RSSs, and downstream red J-genes flanked by light blue 12-RSSs. The V-genes, J-genes, and accompanying RSSs are brought together by a multilayered mechanism. RAG1/2 recognizes and binds to 12-RSS and 23-RSS. RAG1/2 proceeds to cleave between the RSSs and the gene segments, creating two DSBs at both bound RSSs. Coding-ends are first joined by NHEJ to create a coding joint, and then, RSSs are joined by NHEJ machinery, creating a signal joint on an excised circular DNA fragment. (B) Cryo-EM structure of the nicked 12-RSS (from PDB: 6CIJ). The coding flank is in black. The 12-RSS is colored by conserved motifs: orange = heptamer, teal = spacer, and green = nonamer.

Figure 2

(A) Diagram of mouse RAG1 domains. (I) The topologically independent domains of RAG1 are shown in dark blue. These include the central noncore domain, CND; zinc dimerization domain, ZDD; nonamer binding domain, NBD; central domain; and C-terminal domain. (II) Core RAG1 regions are shown in dark pink (389–1008 aa) and noncore RAG1 (residues 1–388 and 1009–1040) regions in dark teal. (III) Functional domains of RAG1 are shown in light purple. These include the RING finger domain, RING; Zinc Finger A domain, ZFA; Dimerization and DNA binding domain, DDBD; pre-RNase A domain, PreR; Catalytic RNase H, RNH; Zinc binding and coordinating domains, ZnC2 and ZnH2; C-terminal domain, CTD. The DDE active site residues are shown in black below the RAG1 bar diagram. The ribbon diagram of the ZDD (PDB: 1RMD) is shown below the RAG1 bar diagram (at left). The ZDD subunits are pink or purple, with zinc ions shown as red spheres. The NBD (PDB: 3JBW) is shown to the right. Each NBD subunit is colored pink or purple and is shown in complex with the RSS nonamer (shown in light gray). (B) Diagram of mouse RAG2 domains. (I) Topologically independent domains of RAG2 are shown in dark blue. These include the 6-blade ß propeller and the Plant Homeodomain, PHD. (II) Core (residues 1–352) and noncore regions of RAG2 (residues 353–527) are shown in dark pink and dark teal, respectively. (III) Functional domains of RAG2 are shown in light purple. These include the Acidic Hinge and the Nuclear Localization Signal, NLS. Residue T490 is shown in black below the RAG2 bar diagram. The ribbon diagram of the 6-bladed ß propeller structure of core RAG2 (PDB: 4WWX) is shown at left below the RAG2 bar diagram. Each blade’s surface is colored red, orange, yellow, green, blue, and purple. To the right is the ribbon diagram of RAG2’s PHD in complex with an H3K4me3 peptide (PDB: 2V89). The H3 peptide backbone and side chains are shown in a stick depiction with transparent spheres overlaid. The trimethylated lysine 4 residue is highlighted with pink spheres.

Figure 3

(A) A cartoon model of the RAG1/2 heterotetramer (PDB: 6CIJ) bound to a 12- and 23-RSS. The two RAG1 subunits are in pink and purple, and the two RAG2 subunits are colored teal and green. Both RSSs are shown in black. Partial structures for HMGB1, bound to the 23-RSS spacer, are shown in red. (B) A zoomed-in view of the nicked 23-RSS in the active site of RAG1 (PDB: 6CIJ). The RAG1 protein is colored in panel A, and the 23-RSS heptamer and coding flank are colored orange and black, respectively. A divalent cation bound in the active site is represented by a red sphere.

Figure 4

(A) A schematic workflow of the SARP-seq assay with a randomized 12-RSS sequence library inserted into the extrachromosomal V(D)J recombination plasmid substrate. PCR incompatible primer orientations are shown as bent black lines. RAG1/2 binds and cleaves the constant 23-RSS and the library of 12-RSSs with each sequence being recombined at different activity levels. Recombined plasmid products have a primer orientation that is compatible with PCR. Next-generation sequencing (NGS) of the recombined 12/23 RSS signal joints reveals differential 12-RSS utilizations. The relative recombination efficacy of each 12-RSS is determined by differential read count analysis. (B) Pie chart showing the relative efficacy of RSS heptamer motifs at positions 5–7, where R and Y are purine and pyrimidine, respectively. (C) The ranges of minor groove widths of 12-RSSs over the time scale of the MD simulations, which demonstrate minor groove width variability of the RYR in purple (consensus – CACAATG|ATAC) and YRY in pink (anticonsensus – CACTTAT|GTAC) motifs at heptamer positions 5–7. The pie chart and plots shown in panels B and C, respectively, are representative data from Hoolehan et al.