CST Is Epistatic With Shieldin to Limit DNA Double-Strand Break End Resection and Promote Repair During Igh Class Switch Recombination

Abstract

Downstream of 53BP1-RIF1 lies the Shieldin (SHLD) protein complex, which comprises MAD2L2/REV7, SHLD3, SHLD2, and SHLD1, and the CTC1-STN1-TEN1 (CST) complex. During immunoglobulin heavy-chain (Igh) class switch recombination (CSR), 53BP1-RIF1-SHLD promotes productive end-joining by limiting resection of activation-induced cytidine deaminase (AID)-generated DNA double-strand break (DSB) ends. The precise role of the CST complex and its interplay with SHLD during CSR is however elusive. Here, we established AID-inducible B cell lines deficient for CTC1, SHLD1, or both and analyzed CSR in these cells. We show that stimulated CTC1-deficient B cells are defective for IgM-to-IgA class switching, accumulate Igh chromosome breaks and translocations, and display increased end-resection and micro-homology usage at switching sites, demonstrating that CTC1 is essential to suppress alternative end-joining during CSR. We show that CTC1 and SHLD1 are epistatic in preventing exacerbated DNA end resection and genetic instability during CSR. Moreover, using a complementation approach in Shld1 knockout splenic B cells, we show that a SHLD1 mutant defective in CST binding (SHLD1^ΔLDLP) is fully proficient for IgM-to-IgG1, IgG2b, IgG3, and IgA class switching, thus demonstrating that the SHLD1-CTC1 interaction through this motif is dispensable for CST and SHLD functions in promoting CSR.

Keywords: CST; Class switch recombination; DSB repair; Shieldin.

FIGURE 1

SHLD1‐CTC1 interaction via the SHLD1 LDLP motif is dispensable for CSR. (A) Schematic of the experimental workflow was done using Biorender. (B) Representative FACS plots of IgG1 CSR assay in primary B cells of the indicated genotype. IgG1⁺ and Ig^neg cells are highlighted with black lines and dashed black lines respectively. (C–F) Percentages of IgG1⁺ (C), IgG2b⁺ (D), IgG3⁺ (E), and IgA⁺ (F) cells in controls and Shld1^−/− primary B cells complemented with empty vector (EV) or SHLD1 variants. (G–J) Percentages of nonproductive IgM^low IgG1⁻ (G), IgM^low IgG2b⁻ (IgG3⁻) (H), IgM^low IgG3⁻ (IgG2b⁻) (I) and IgM^low IgA⁻ (J) fractions in controls and Shld1^−/− primary B cells complemented with SHLD1 variants. Bars represent the mean ± SD. n = 4–7 for WT; n = 4–6 for Shld1^−/− ; n = 4–5 for WT+EV, Shld1^−/− +EV, Shld1^−/− + SHLD1^WT, Shld1^−/− + SHLD1^L20A, and Shld1^−/− + SHLD1^ΔLDLP. ***p < 0.001; **p < 0.01; *p < 0.05. Mann–Whitney test.

FIGURE 2

CST is required for CSR and is epistatic with SHLD. (A) Schematic of the experimental workflow was done using Biorender. (B) Doubling time for untreated cells (‐CIT) and stimulated cells (+CIT) of the indicated genotype. Bars represent the mean ± SD. (C) Representative FACS plots of IgA CSR assay in the indicated CH12F3 clones. IgA⁺ and Ig^neg cells are highlighted with black lines and dashed black lines respectively. (D) Percentages of IgA⁺ CH12F3 cells with and without in vitro stimulation for 3 days for the indicated genotypes. (E) Percentages of IgM⁻ IgA⁻ CH12F3 cells with and without stimulation for 3 days for the indicated genotypes. Bars represent the mean ± SD of at least three independent experiments. Two clones for WT (CH12F3 and Ct34), 3 clones for Stn1^−/− (St11, St17, St20), 8 clones for Ctc1^−/− (Ct10, Ct11, Ct21, Ct105, Ct131, Ct137, Ct160, Ct168), 2 clones for Shld2^−/− (F11, F12), 3 clones for Shld1^−/− (O6, S8, S25), and 4 clones for Shld1^−/− Ctc1^−/− (CtS25‐4, CtS25‐6, CtS25‐10, CtS8‐4) were used. ***p < 0.001, Mann–Whitney test. CIT: anti‐CD40, IL‐4, TGF‐β.

FIGURE 3

CTC1 and SHLD1 are epistatic in promoting DSB repair and limiting DSB end resection during CSR. (A) Representative images of metaphase spread from WT cells (left panel) and Ctc1^−/− cells (right panel, yellow triangles point to a broken chromosome at the Igh locus), as quantified in (B), scale bars, 3 µm. (B) Quantification of Igh breaks (white bar) and translocations (grey bar) in metaphases of WT, Ctc1^−/− , Shld1^−/− and Shld1^−/− Ctc1^−/− cells with or without stimulation for 3 days. Bars represent means ± SD. See also Table S2. ***p < 0.001; **p < 0.01; *p < 0.05, Student t‐test. (C) Schematic representation of the murine Igh locus. Arrows indicate primers used for the long‐range PCR. (D) Long‐range PCR analysis of Sμ to Sα rearrangements in WT, Ctc1^−/−, Shld1^−/− , and Shld1^−/− Ctc1^−/− stimulated CH12F3 cells (Replicate 1). Il2 gene PCR was used as a loading control. See also Figure S5. (E) Length distribution of mapped PacBio reads from long‐range PCR. Samples from replicates 1 and 2 were pooled per genotype. See also Figure S5. ***p < 0.001; mixed‐model regression test. (F) Analysis of breakpoints localization. Graphs represent the percentage of junctions. Sμ and Sα switch regions are colored light grey and light green respectively. Coordinates on mm10 are indicated. The total number of unique junctions analyzed for each genotype is indicated in brackets. (G) Percentage of resected Sμ–Sα junctions. A junction was categorized as resected when the breakpoint was mapped outside of either Sμ or Sα as defined in (F). Bars represent mean ± SEM. **p < 0.01, Student t‐test. (H) Percentage of Sμ–Sα direct junctions. Bars represent mean ± SEM. **p < 0.01, Student t‐test. (I) Percentage of Sμ–Sα junctions with ≥3 bp microhomologies (MH). Bars represent mean ± SEM. *p < 0.05, Student t‐test. (J) MH usage distribution of Sμ–Sα junctions. Graphs represent mean ± SEM. WT, n  =  3; Ctc1^−/− , n = 6; Shld1^−/− , n = 5; Shld1^−/− Ctc1^−/− , n = 6. The total number of unique junctions analyzed for each genotype is indicated in brackets.