Leaky severe combined immunodeficiency and aberrant DNA rearrangements due to a hypomorphic RAG1 mutation

Abstract

The RAG1/2 endonuclease initiates programmed DNA rearrangements in progenitor lymphocytes by generating double-strand breaks at specific recombination signal sequences. This process, known as V(D)J recombination, assembles the vastly diverse antigen receptor genes from numerous V, D, and J coding segments. In vitro biochemical and cellular transfection studies suggest that RAG1/2 may also play postcleavage roles by forming complexes with the recombining ends to facilitate DNA end processing and ligation. In the current study, we examine the in vivo consequences of a mutant form of RAG1, RAG1-S723C, that is proficient for DNA cleavage, yet exhibits defects in postcleavage complex formation and end joining in vitro. We generated a knockin mouse model harboring the RAG1-S723C hypomorphic mutation and examined the immune system in this fully in vivo setting. RAG1-S723C homozygous mice exhibit impaired lymphocyte development and decreased V(D)J rearrangements. Distinct from RAG nullizygosity, the RAG1-S723C hypomorph results in aberrant DNA double-strand breaks within rearranging loci. RAG1-S723C also predisposes to thymic lymphomas associated with chromosomal translocations in a p53 mutant background, and heterozygosity for the mutant allele accelerates age-associated immune system dysfunction. Thus, our study provides in vivo evidence that implicates aberrant RAG1/2 activity in lymphoid tumor development and premature immunosenescence.

Figure 1

Impaired lymphocyte development in RAG1-S723C homozygous mice. Flow cytometric analyses were performed on RAG1^+/+, RAG1^+/S723C, and RAG1^S723C/S723C littermates, as indicated, at 5 weeks of age. (A) Thymocytes and lymph node cells were stained with αCD4 and αCD8 antibodies. (B) Bone marrow and splenocytes were stained with antibodies against the indicated cell-surface markers.

Figure 2

PCR amplification of endogenous rearrangements in RAG1-S723C homozygous developing lymphocytes. (A) TCRβ rearrangements. Genomic DNA from purified DN thymocytes from RAG1^+/+ (WT) and RAG1^S723C/S723C (S723C) mice and a nonrearranging tissue (tail) was PCR amplified to detect Dβ1 to Jβ1 and Dβ2 to Jβ2 rearrangements. For Dβ1 to Jβ1 rearrangements, genomic DNA was first digested with the XbaI restriction endonuclease. Control PCR amplification of a nonrearranging locus was performed to normalize levels of input DNA. Rearrangements were detected by Southern blot hybridization using an oligonucleotide probe. Experiments (A-C) were repeated 3 times with genomic DNA samples from 3 different sets of mice; representative results are shown. (B) IgH rearrangements. Genomic DNA from sorted pro- and pre-B-cell populations from RAG1^+/+ (WT) and RAG1^S723C/S723C (S723C) mice and a nonrearranging tissue (kidney) was PCR amplified to detect D_H to J_H rearrangements. (C) Endogenous TCRδ signal joints. Levels of extrachromosomal signal joints formed between TCR Dδ2 and Jδ1 RSSs were detected by PCR amplification of genomic thymocyte DNA isolated from RAG^+/+ (WT) and RAG1^S723C/S723C (S723C) mice and a nonrearranging tissue (tail). Vertical lines have been inserted to indicate repositioned gel lanes.

Figure 3

LM-PCR analyses of signal ends in RAG1-S723C developing lymphocytes. Three-fold serially diluted linker ligated genomic DNA isolated from sorted (A) DN thymocytes and (B) sorted pro-B bone marrow cells from RAG1^+/+, RAG1^+/S723C, and RAG1^S723C/S723C were used in PCR amplification reactions for detection of signal ends at the indicated loci. PCR amplification of a nonrearranging locus was performed as a normalization control. C indicates non–linker-ligated wild-type genomic DNA. (C) Location of LM-PCR products within the IgH J_H locus. Δ bp indicates number of nucleotides deleted 5′ of the J_H RSS depicted at the top of the columns; triangles, RSSs; and arrows, RAG1/2 cleavage sites. LM-PCR analyses were repeated 3 times on genomic DNA isolated from at least 3 independently sorted DN thymocyte and pro-B-cell samples. Representative results are shown.

Figure 4

Significant impairment in B- and T-cell development in older RAG1-S723C heterozygous mice. Flow cytometric analyses were performed on RAG1^+/+, RAG1^+/S723C, and RAG1^S723C/S723C mice at 12 months of age. (A) Thymocytes and splenocytes were stained with αCD4 and αCD8 antibodies. (B) Bone marrow cells and splenocytes were stained with antibodies against the indicated cell-surface markers.

Figure 5

Rearrangements in RAG1-S723C thymocytes and RAG1-S723C/p53 double-mutant thymic lymphomas. (A) TCRγV3S1 (chr 13) to TCRβJ2S7 (chr 6) interchromosomal trans-rearrangements were PCR amplified from genomic thymocyte DNA isolated from mice of the indicated genotypes (top panel). S723C indicates RAG1^S723C/S723C; Kid, kidney. Products corresponding to trans-rearrangements were detected by Southern blot analysis. Levels of rearrangements were normalized to PCR amplification of a nonrearranging locus (middle panel). Representative results are shown. Vertical lines have been inserted to indicate repositioned gel lanes. PCR products corresponding to TCRγV3-TCRβJ2 rearrangements in RAG1-S723C thymocytes were subcloned and sequenced (bottom panel). Coding sequences for TCRγV3 and TCRβJ2 are shown in boxes (bold); nucleotides added (center column) include P nucleotides (underlined) and N nucleotides; number of clones of each sequence are indicated. (B) Genomic DNA isolated from RAG1-S723C/p53 double-mutant lymphomas was digested with EcoRI then analyzed by Southern blot. Dβ1 to Jβ1 rearrangements were detected using a probe located 5′ of the Dβ1 segment (diagram). Individual tumors are indicated, top. C indicates wild-type thymus DNA; Kid, nonrearranging tissues; GL, unrearranged, germline band; and R, D-Jβ1 rearrangements. Amounts of input DNA were normalized to a nonrearranging locus (LR8). (C) PCR amplification of Dβ1 to Jβ1, Dβ2 to Jβ2, and Vβ8 to DJβ1 rearrangements were performed as described in Figure 2A using genomic DNA (250 ng) isolated from RAG1-S723C/p53 double-mutant lymphomas (tumor numbers indicated, top), wild-type thymocytes (C), and a nonrearranging tissue (tail). Input DNA levels were normalized to a nonrearranging locus (ATM). Vertical lines have been inserted to indicate repositioned gel lanes.

Figure 6

RAG1-S723C/p53 double-mutant mice are predisposed to thymic lymphomas with chromosomal translocations. (A) Survival of a cohort of control (RAG1^+/+, RAG1^+/S723C, RAG1^S723C/S723C, n = 39), p53^−/− (RAG1^+/+p53^−/−, RAG1^+/S723Cp53^−/−, n = 35), and RAG1^S723C/S723Cp53^−/− (n = 31) mice was observed for a period of 35 weeks. Kaplan-Meier survival curves were compared using the 2-tailed log-rank test with a 95% confidence interval (RAG1^S723C/S723Cp53^−/− vs p53^−/− cohorts, P = .017; RAG1^S723C/S723Cp53^−/− vs control cohorts, P < .001). (B) RAG1^S723C/S723Cp53^−/− thymic lymphomas harbor chromosomal translocations and fusions. DAPI staining (left panels) and SKY analysis (right panels) of tumors S542 and S922 containing clonal chr 15;15 and 12;12 short-arm fusions and t(14;8) translocations, respectively. Arrows indicate translocated chromosomes.