Molecular analysis of γ1, γ3, and α class switch recombination junctions in APOBEC3-deficient mice using high-throughput sequencing

Activation-induced deaminase (AID) is required for immunoglobulin (Ig) class switch recombination (CSR), in which the constant (C)_μ gene of IgM is substituted with C_γ, C_ε, or C_α, thereby generating IgG, IgE, and IgA antibodies, respectively, with new effector functions but the same antigenic specificity.¹ AID targets specific DNA switch (S) regions preceding C regions except for C_δ.² S_μ is usually the donor region, while S_γ,ε,α are the acceptor regions. AID deaminates C into U on single-stranded DNA by targeting the WRCY (W = A/T, R = A/G, and Y = C/T) hot motif and, to a lesser extent, the SYC (S = G/C, Y = C/T) cold motif.^3,4 AID is a member of the apolipoprotein B editing complex (APOBEC) family. Among APOBEC genes, a family of evolutionarily conserved cytidine deaminases, APOBEC3 is implicated in diverse cell functions including innate immunity against retroviruses.⁴ The DNA-editing APOBEC3 enzymes have recently attracted attention due to their involvement in cancer and potential applications in gene editing.^5–7 While a single copy of each APOBEC3 gene is present in rodents, seven copies of each APOBEC3 gene are found in humans. For example, APOBEC3A is expressed predominantly in monocytes, while high levels of APOBEC3G are found in B-cells. Both APOBEC3A and APOBEC3G expression is elevated in response to interferon-α.⁸ CSR is mediated by only AID attack, but some studies have suggested the putative involvement of APOBEC in the cleavage of some σ_δ and S_α DNAs.^9,10 Furthermore, B-cell agonists that stimulate CSR upregulated both AID and APOBEC3G.¹¹ APOBEC3G was also found to mutate V_H genes during viral infection.^12,13 APOBEC3 may contribute to a basal level of CSR, but its involvement was largely undetectable when a few dozen CSR junctions were reported after cloning and subsequent sequencing by the Sanger method. We recently reported a new computational tool (CSReport) for the automatic analysis of CSR junctions sequenced by high-throughput sequencing.¹⁴ The structural profiles (blunt, microhomology, large homology, or junctions with insertions) and locations of breakpoints within AID hotspots and other motifs can be analyzed for not just a few dozen but rather several hundreds or thousands of junctions. This computational tool has already been used to analyze the phenotypes of rare S_μ–S_δ junctions during IgD CSR^9,15 and compare the structural profiles of S_μ–S_γ1/S_μ–S_γ3/S_μ–S_α junctions in uracil DNA glycosylase-deficient mice,¹⁶ IgH E_μ enhancer-deficient mice,¹⁷ and wild-type (wt) mice.¹⁸ In this study, we used high-throughput sequencing to analyze the molecular signature of S_μ–S_γ1/S_μ–S_γ3/S_μ–S_α junctions in mice expressing human APOBEC3A or APOBEC3G in an APOBEC3 knockout background for the first time.

Mice were housed and procedures were conducted in agreement with European Directive 2010/63/EU on the use of animals for scientific purposes applied in France as the “Décret n°2012-118 du 1^er février 2013 relatif à la protection des animaux utilisés à des fins scientifiques”. Accordingly, the present project (APAFiS≠13855) was authorized by the “Ministère de l’Education Nationale, de l’Enseignement Supérieur et de la Recherche” and reviewed by the Ethics Committee of the University of Limoges (France). APOBEC3-deficient mice¹⁹ expressing human APOBEC3A (hAPOBEC3A) or hAPOBEC3G²⁰ and C57BL6 wt mice (both types of mice were housed in a conventional animal facility) were used. B-splenocytes (1 × 10⁶ cells/ml) were cultured for 4 days in RPMI 1640 with 10% fetal calf serum and 5 μg/ml LPS with (for IgG₁ CSR) or without (for IgG₃ CSR) 20 ng/ml IL-4 or 5 ng/ml TGFβ (for IgA CSR) (PeproTech, Rocky Hill, NJ).^21,22 Splenocytes were stained with several labeled antibodies and analyzed on a Fortessa LSR2 (Beckman Coulter). B-cell splenocyte DNA was extracted, and S_μ–S_γ1, S_μ–S_γ3, and S_μ–S_α junctions were quantified as previously reported.¹⁴ Ion proton sequencing was performed in the “GénoLim platform” at the Limoges University (France). Sequenced reads were mapped to S_μ,_γ1,_γ3,_α regions using the BLAST algorithm. The computational tool developed for the experiments performed junction assembly, identified breakpoints in S regions, identified junction structures (blunt, microhomology, large homology, or junctions with insertions) and output a statistical summary of the identified junctions.

Flow cytometry experiments indicated that deletion of the APOBEC3 gene had no significant impact on in vitro CSR toward IgG₃, IgG₁, and IgA compared with CSR in wt mice (Fig. 1a). Extracted stimulated B-cell DNA was then used to determine the molecular signatures of S_μ–S_γ3 (IgG₃ CSR), S_μ–S_γ1 (IgG₁ CSR) and S_μ–S_α (IgA CSR) junctions by high-throughput sequencing. The structural profiles (blunt, microhomology, large homology, or junctions with insertions) of 481 S_μ–S_γ3, 4092 S_μ–S_γ1 and 225 S_μ–S_α APOBEC3-deficient junctions and 1394 S_μ–S_γ3, 4163 S_μ–S_γ1, and 711 S_μ–S_α wt junctions are reported in Fig. 1b. No differences in IgG₃, IgG₁, and IgA junctions were found between APOBEC3-deficient mice and wt mice. In addition to the types of junctions, the distributions of IgG₃, IgG₁, and IgA junctions in terms of break positions along S regions were similar between APOBEC3-deficient and wt mice (Fig. 1c). Similar results were found for the locations of breakpoints within TYC APOBEC3 hotspot motifs²³ (displayed along specifically targeted segments within S regions) (Fig. 1d). The frequency of breakpoints found in S_γ3, S_γ1, and S_α was lower than the theoretical frequency for this motif in S_γ3, S_γ1, and S_α regions. In comparison, the frequency of breakpoints within WRCY AID hotspots was similar to the theoretical frequency of this motif in S_γ3, S_γ1, and S_α regions (Fig. 1e). Finally, the effects of hAPOBEC3A and hAPOBEC3G transgenes on murine CSR were explored. A total of 838 S_μ–S_γ3, 3154 S_μ–S_γ1, and 409 S_μ–S_α hAPOBEC3A junctions and 541 S_μ–S_γ3, 2910 S_μ–S_γ1 and 168 S_μ–S_α hAPOBEC3G junctions are reported in Fig. 1b. No differences in IgG₃, IgG₁, and IgA junctions were found between hAPOBEC3A or hAPOBEC3G mice and wt mice.

Fig. 1.

CSR in APOBEC3-deficient mice. a IgG₁, IgG₃, and IgA CSR in APOBEC3-deficient and wt mice. B-splenocytes were cultured with LPS for 4 days with (IgG₁ CSR) or without (IgG₃ CSR) IL-4 or TGFβ (IgA CSR). Cells gated on B220⁺ and/or CD138⁺ cells were labeled with anti-IgM, anti-IgG₃, anti-IgG₁, and anti-IgA antibodies. The mean ± SEM of 4 experiments is reported for each isotype. No significant differences were found (Mann–Whitney U-test). b Structural profiles of S_μ–S_γ3, S_μ–S_γ1, and S_μ–S_α junctions in APOBEC3-deficient, hAPOBEC3A, hAPOBEC3G, and wt mice. Junctions are classified as follows: junction with insertion, blunt junction, junction with microhomology (<4 bp) or large homology (≥4 bp). The results (mean ± SEM) from two to seven mice for each isotype are pooled. No significant differences were found (chi-squared test) compared to wt mice. c Breakpoint locations in S_μ–S_γ1, S_μ–S_γ3 and S_μ–S_α junctions in wt and APOBEC3-deficient mice (same junctions as in b). d Breakpoints in mouse APOBEC3 TYC hotspot motifs in S_μ, S_γ1, S_γ3, and S_α junctions (same junctions as in b). The theoretical levels of TYC breakpoints are highlighted. e Breakpoints in mouse AID WRCY hotspot motifs in S_μ, S_γ1, S_γ3, and S_α junctions (same junctions as in b). The theoretical levels of AID breakpoints are highlighted

Recent studies have attracted attention to the involvement of APOBEC3 enzymes in cancer and gene editing.^5–7 We report in the present study that B-cell γ₁, γ₃, and α CSR was unaffected in mice following deletion of the APOBEC3 gene. Similar results were found in mice expressing hAPOBEC3A and hAPOBEC3G. Breaks in the TYC APOBEC3 hotspot motifs were not affected by APOBEC3 deficiency, suggesting AID-induced breaks (TYC is also a cold AID motif). Compared with wt mice, the types and locations of CSR junctions in the absence of APOBEC3 were no different. These results were highlighted by the analysis of thousands of independent junctions using our computational tool and CSR junctions sequenced by high-throughput sequencing. In conclusion, molecular analysis of CSR junctions demonstrated that APOBEC3 makes no obvious contribution to conventional (IgG1, IgG3, and IgA) CSR.

Author contributions

Y.D. designed the research and obtained grants and the agreement of the Ethics Committee. N.G., H.I., M.F., and S.L. performed the research. F.B., J.C.M., and Y.D. analyzed the data. N.G., H.I., M.F., J.C.M., F.B., and Y.D. wrote the paper.

Competing interests

The authors declare no competing interests.