Phosphorylation at S2053 in Murine (S2056 in Human) DNA-PKcs Is Dispensable for Lymphocyte Development and Class Switch Recombination This information is current as of October 1, 2021. Wenxia Jiang, Verna M. Estes, Xiaobin S. Wang, Zhengping Shao, Brian J. Lee, Xiaohui Lin, Jennifer L. Crowe and Shan Zha J Immunol published online 17 May 2019

http://www.jimmunol.org/content/early/2019/05/16/jimmun Downloaded from ol.1801657

Supplementary http://www.jimmunol.org/content/suppl/2019/05/16/jimmunol.180165 Material 7.DCSupplemental http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

by guest on October 1, 2021 • Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2019 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published May 17, 2019, doi:10.4049/jimmunol.1801657 The Journal of Immunology

Phosphorylation at S2053 in Murine (S2056 in Human) DNA-PKcs Is Dispensable for Lymphocyte Development and Class Switch Recombination

Wenxia Jiang,* Verna M. Estes,* Xiaobin S. Wang,*,† Zhengping Shao,* Brian J. Lee,* Xiaohui Lin,* Jennifer L. Crowe,*,† and Shan Zha*,‡,x

The classical nonhomologous end-joining (cNHEJ) pathway is a major DNA double-strand break repair pathway in mammalian cells and is required for lymphocyte development and maturation. The DNA-dependent kinase (DNA-PK) is a cNHEJ factor that encompasses the – () heterodimer and the large DNA-PK catalytic subunit (DNA-PKcs). In mouse models, loss of DNA-PKcs (DNA-PKcs2/2) abrogates end processing (e.g., hairpin opening), but not end-ligation, whereas expres- sion of the kinase-dead DNA-PKcs protein (DNA-PKcsKD/KD) abrogates end-ligation, suggesting a kinase-dependent structural function of DNA-PKcs during cNHEJ. Lymphocyte development is abolished in DNA-PKcs2/2 and DNA-PKcsKD/KD mice because Downloaded from of the requirement for both hairpin opening and end-ligation during V(D)J recombination. DNA-PKcs itself is the best- characterized substrate of DNA-PK. The S2056 cluster is the best-characterized autophosphorylation site in human DNA- PKcs. In this study, we show that radiation can induce phosphorylation of murine DNA-PKcs at the corresponding S2053. We also generated knockin mouse models with alanine- (DNA-PKcsPQR) or phospho-mimetic aspartate (DNA-PKcsSD) substitutions at the S2053 cluster. Despite moderate radiation sensitivity in the DNA-PKcsPQR/PQR fibroblasts and lymphocytes, both DNA-PKcsPQR/PQR and DNA-PKcsSD/SD mice retained normal kinase activity and underwent efficient V(D)J recombination and class switch recombi- http://www.jimmunol.org/ nation, indicating that phosphorylation at the S2053 cluster of murine DNA-PKcs (corresponding to S2056 of human DNA-PKcs), although important for radiation resistance, is dispensable for the end-ligation and hairpin-opening function of DNA-PK essential for lymphocyte development. The Journal of Immunology, 2019, 203: 000–000.

ymphocyte development requires ordered assembly and includes ATM and ATR kinases (1). DNA-PKcs and ATM are subsequent modifications of the Ag receptor loci through both activated by DNA DSBs in vivo and share many common programmed DNA double-strand breaks (DSBs). Although substrates, including histone H2AX, KAP1, and DNA-PKcs

L by guest on October 1, 2021 the initiation of these DNA breaks is lymphocyte specific, these itself, which contribute to their critical and redundant functions in physiological DSBs activate the DNA damage response and are embryonic development, lymphocyte-specific rearrangement, repaired by the ubiquitously expressed classical nonhomologous and DNA repair (2–6). end-joining (cNHEJ) pathway. DNA-dependent protein kinase The cNHEJ pathway is one of the two best-characterized DNA (DNA-PK) is a vertebrate-specific cNHEJ factor. DNA-PK ho- DSB repair pathways in mammalian cells. As its name implies, loenzyme includes the evolutionarily conserved DNA-binding cNHEJ ligates two DNA ends together, independent of sequence Ku70andKu80(KU86inhuman)heterodimer(referredtoas homology. KU initiates cNHEJ by binding to the dsDNA ends, KU together) and the vertebrate-specific large catalytic subunit which in turn recruits and activates DNA-PKcs (7). Among other (DNA-PKcs). DNA-PK also belongs to the PI3 kinase–related functions, DNA-PK holoenzyme further recruits and activates serine/threonine protein kinase (PI3KK) family, which also another vertebrate-specific cNHEJ factor, Artemis endonuclease,

*Institute for Cancer Genetics, Vagelos College of Physicians and Surgeons, Columbia cells for lymphocyte development and function. W.J. generated the DNA-PKcsSD and University, New York, NY 10032; †Graduate Program of Pathobiology and Molecular DNA-PKcsPQR mice. X.S.W., J.L.C., Z.S., and B.J.L. established the HTGTS assay Medicine, Vagelos College of Physicians and Surgeons, Columbia University, New York, and performed the high-throughput genomic translocation sequence analyses of the NY 10032; ‡Division of Pediatric Oncology, Hematology and Stem Cell Transplantation, class switch recombination junctions for this study. X.L. analyzed the V(D)J recom- Department of Pediatrics, Vagelos College of Physicians and Surgeons, Columbia Univer- bination junction in the DNA-PKcsSD and DNA-PKcsPQR mice. sity, New York, NY 10032; and xDepartment of Pathology and Cell Biology, Vagelos The high-throughput genomic translocation sequence data reported in this paper have College of Physicians and Surgeons, Columbia University, New York, NY 10032 been deposited in the Omnibus database (https://www.ncbi.nlm.nih. ORCIDs: 0000-0002-5097-0975 (W.J.); 0000-0002-8265-0474 (X.L.); 0000-0002- gov/geo/) under accession numbers GSE117628 and GSE129895. 6568-1818 (S.Z.). Address correspondence and reprint requests to Dr. Shan Zha, 1130 St. Nicholas Received for publication December 21, 2018. Accepted for publication April 22, Avenue, RM503B, Institute for Cancer Genetics, Columbia University, New York, 2019. NY 10032. E-mail address: [email protected] This work is in part supported by National Institutes of Health (NIH)/National The online version of this article contains supplemental material. Cancer Institute (NCI) Grants R01CA158073, R01CA184187, P01CA174653, and Abbreviations used in this article: cNHEJ, classical nonhomologous end-joining; CSR, R01CA226852. S.Z. is the recipient of a Leukemia Lymphoma Society Scholar class switch recombination; DNA-PK, DNA-dependent protein kinase; DNA-PKcs, Award. W.J. was supported by NIH/NCI Grant T32-CA09503. J.L.C. was supported by DNA-PK catalytic subunit; DP, double positive; DSB, double-strand break; ES, embry- NIH/NCI F31CA183504-01A1 and Clinical and Translational Science Award/NIH onic stem; GL, germline; HTGTS, high-throughput genomic translocation sequence; IR, Grant TL1 TR000082. This research was funded in part through NIH/NCI Cancer ionizing radiation; MEF, murine embryonic fibroblast; MH, microhomology; NeoR, Center Support Grant P30CA013696 to Herbert Irving Comprehensive Cancer Center neomycin-resistant; WT, wild-type. of Columbia University. W.J., V.M.E., S.Z. designed experiments. S.Z. wrote the paper with help from W.J. Copyright Ó 2019 by The American Association of Immunologists, Inc. 0022-1767/19/$37.50 and V.M.E. W.J. and V.M.E. analyzed the DNA-PKcsPQR, DNA-PKcsSD mouse and

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1801657 2 LYMPHOCYTES DEVELOP WITHOUT DNA-PKcs S2056 PHOSPHORYLATION which processes the DNA ends (e.g., hairpin opening) (8). Finally, human cells (26). Since then, phosphorylation at the S2056 has the end-ligation is completed by the evolutionarily conserved been widely used as the marker for DNA-PKcs activation. In cNHEJ factors (Ligase4/XRCC4/XLF complex) with the help of Chinese hamster ovary cells expressing human DNA-PKcs, KU. Two new cNHEJ factors, PAXX and CYREN/MRI, were phosphorylation of DNA-PKcs at the S2056 and T2069 clus- recently identified and interact with KU directly (9–13). Although ters alone and in combination were found to be important for not essential for cNHEJ in otherwise wild-type (WT) cells, they radiation sensitivity, repair pathway choices, and release of are essential for end-ligation in XLF-deficient cells (9–13). DNA-PKcs from DNA breaks (26, 32–35). In mouse models, Because cNHEJ functions throughout the cell cycle, it is the alanine substitutions in the T2609 cluster (DNA-PKcs3A/3A) leads predominant repair pathway in nonproliferating cells (e.g., to neonatal lethality associated with p53-dependent bone marrow postmitotic neurons). Complete loss of KU, Ligase4, or Xrcc4 failure, which also abolished lymphocyte development (36, 37). abrogates end-ligation and leads to postmitotic neuronal apoptosis But the biological consequence of S2056 phosphorylation re- and, in the case of Ligase4 or Xrcc4-deficiency, p53-dependent mains unknown. In fact, whether endogenous murine DNA-PKcs embryonic lethality (14). Loss of DNA-PKcs or Artemis abrogates canbephosphorylateduponirradiationhasnotbeendemon- hairpin opening, but not end-ligation. Accordingly, DNA-PKcs2/2 strated. In this study, we developed a customized Ab and showed and Artemis2/2 mice were born of normal size at the expected ratio for the first time, to our knowledge, that murine DNA-PKcs (15–17). But in contrast to the normal development of DNA-PKcs2/2 can be phosphorylated at S2053 (corresponding to the S2056 mice, mice expressing the kinase-dead DNA-PKcs protein (DNA- in human DNA-PKcs) upon irradiation. Using mouse models PKcsKD/KD) die embryonically with severe end-ligation defects with knockin alanine-substitution (DNA-PKcsPQR) or phospho- (18), indicating that the presence of catalytic-inactive DNA-PKcs mimetic mutations (DNA-PKcsSD)attheS2056cluster,we blocks end-ligation, whereas lack of DNA-PKcs does not. further show that although moderately sensitive to radiation, Downloaded from In developing lymphocytes, the cNHEJ pathway is exclusively DNA-PKcsPQR/PQR and DNA-PKcsSD/SD lymphocytes are pro- required for the assembly of the Ag receptor gene products via ficient for both the hairpin opening and end-ligation necessary V(D)J recombination. RAG initiates V(D)J recombination by for development and maturation, even when measured by generating a pair of blunt 59-phosphorylated signal ends and a pair sensitive junctional sequencing analyses. Moreover, the end of covalently-sealed hairpin-coding ends. The two signal ends can joining in DNA-PKcsPQR/PQR or DNA-PKcsSD/SD cells does not be ligated directly to form a signal joint. The coding end hairpin require ATM kinase activity and, by extension, compensatory http://www.jimmunol.org/ has to be opened by Artemis with the help of DNA-PKcs before phosphorylation of DNA-PKcs by ATM. Thus, in contrast to the coding ends can be ligated together to form a coding joint, the T2609 cluster, the DNA-PKcs S2056 cluster phosphorylation which encodes the V region of the Ig and TCR and is is largely dispensable for cNHEJ during lymphocyte development ultimately required for lymphocyte development (8). Thus, loss of in mice. DNA-PKcs (null) prevents coding end hairpin opening and causes T– and B–severe combined immunodeficiency in animal models Materials and Methods KD/KD (15, 16). In DNA-PKcs B cells, both signal joints and coding Generation and characterization of the mouse models joints cannot be formed because of end-ligation defects (18). 2 2 2 2 DNA-PKcs / and Atm / mice have been described previously (15, 38). by guest on October 1, 2021 Mature B lymphocytes also undergo additional gene rearrangement– The DNA-PKcsSD allele and DNA-PKcsSD/SD embryonic stem (ES) cells class switch recombination (CSR) at the IgH chain gene to were also described before (18). The DNA-PKcsPQR mutation substitutes generate Abs of different isotypes and, therefore, different effector five serine residues within the S2053 (corresponding to S2056 in human functions. End-ligation during CSR is also mediated by the cNHEJ DNA-PKcs) cluster with alanine using the same targeting arms as before pathway. In the absence of cNHEJ (i.e., XRCC4 deficiency), a (Fig. 1, Supplemental Fig. 1A). Briefly, the pEMC-based targeting con- struct was used to insert a neomycin-resistant (NeoR) cassette in intron 45, subset of CSR can be completed by the alternative end-joining about∼600 bp upstream from the S2053A mutation site in exon 46 of the pathway that preferentially uses microhomologies (MHs) at the murine DNA-PKcs (Prkdc) locus. The 59 arm is∼3.5kbandthe39 arm junction (19, 20). Consistent with DNA-PKcs not being essential is∼4.9 kb. The sequence (∼665 bp) with all mutations was custom for end-ligation, DNA-PKcs2/2 B cells have only moderate defects synthesized by Genewiz and integrated into the 39 arm (Fig. 1A, KD/KD Supplemental Fig. 1A). CSL3 ES cells (129/sv background) were targeted in CSR (21, 22), whereas DNA-PKcs mature B cells carrying by electroporation and screened via Southern blotting (SacI digestion, preassembled IgH/IgL display severe CSR defects, such as in with the 59 probe generated by PCR using primers 59-CAG AAA TGA 2 2 Xrcc4 / B cells (23). Nevertheless, high-throughput sequenc- ACA ATT CCT G-39 and 59-CAG TTG TAT GCA CAA TGA-39)(Fig.1C). ing analyses showed that the residual CSR in both DNA-PKcs2/2 The expected germline (GL) band is∼5.5 kb, and the targeted band ∼ and DNA-PKcsKD/KD B cells preferentially use MH at the is 7.0 kb (with the insertion of NeoR) (Fig. 1C). The correct clones were confirmed with a 39 probe generated by PCR with primers (59- junctions (23). GGT TGG TGT GAC GGA GTT TT-39,59-GCT GAG GCT CTT GAA How DNA-PKcs kinase activity regulates end-ligation and CT-39). Three out of six targeted clones were sequenced and verified to end processing is still not fully understood. Loss of KU rescues have the desired DNA-PKcsPQR mutations. Two were injected for GL FLIP/FLIP the embryonic lethality of DNA-PKcsKD/KD mice, and deletion of transmission. The chimeras were bred into the Rosa26 mice (The Jackson Laboratory; catalog no. 003946) to remove the NeoR cassette. Tail the KU80 C terminus that is necessary for the recruitment of PQR/PQR SD/SD KD/KD DNA from DNA-PKcs and DNA-PKcs mice were PCR DNA-PKcs partially restores end-ligation in DNA-PKcs amplified and sequenced to confirm the desired mutations (Supplemental cells, indicating that the catalytically inactive DNA-PKcs protein Fig. 1A, 1B). Genotyping was performed with primers (59-TTT CCT physically blocks end-ligation at the DNA ends. DNA-PKcs is CTC AAG TAC CAC A-39 and 59-AAA TTA TGT CAG GGA TTT the best-characterized substrate of DNA-PK (24). Several DNA AGA-39) flanking the flippase recognition target site left after NeoR deletion. The genotyping primers are marked in Fig. 1B as solid black damage-induced phosphorylation sites have been identified on arrows. The PCR product corresponding to the WT allele is 280 bp. The DNA-PKcs, including S2056, T2609, and S3590 (25–28). Whereas PCR product corresponding to the PQR or SD allele is 387 bp. A set of the T2609 clusters can be phosphorylated by DNA-PKcs itself representative genotyping results are included in Supplemental Fig. 1C. PQR/PQR SD/SD (25), ATM (29, 30), or ATR (31), DNA-PKcs–specific inhibitors The DNA-PKcs and DNA-PKcs mice used in this study were all of the 129/Sv backgrounds. All animal work was conducted in a or a catalytically inactive DNA-PKcs mutation abolished the specific pathogen–free facility and all the procedures were approved phosphorylation of human DNA-PKcs at the S2056 cluster, by Institutional Animal Care and Use Committee at Columbia University establishing S2056 as the bona fide autophosphorylation site in Medical Center. The Journal of Immunology 3

FIGURE 1. Generation of the DNA-PKcsPQR/PQR and DNA-PKcsSD/SD mouse models. (A)Diagramof the human S2056 (mouse S2053) cluster of DNA-PKcs. (B) The targeting scheme of DNA-PKcsPQR and DNA-PKcsSD allele. Murine DNA-PKcs locus (top), targeting vectors (second row), targeted allele (third row), and the neo-deleted alleles (DNA-PKcsPQR and DNA-PKcsSD, bottom). The NeoR cassette was inserted at∼600 bp upstream from the S2053D mutation site. Not all exons on the arm were marked. P is the PstI site, and S is the SacI site. The 59 and 39 probes are marked as thick black lines. The exons and flippase recognition target sites are shown as solid boxes and open triangles, respectively. The genotyping primers are marked as solid black arrows. A set of representative Downloaded from genotyping results is shown in Supplemental Fig. 1C. The map is not drawn to scale. This clone contains the NeoR cassette, which is subsequently deleted upon breeding with the constitutive ROSAFLIP/FLIP allele (The Jackson Laboratory; 003946). The open box in- dicates the mutated exon (A2053-PQR), and the filled- C box indicates the WT exon (S2053). ( ) Southern blot http://www.jimmunol.org/ analyses of SacI-digested DNA from DNA-PKcs+/+ and DNA-PKcsPQR targeted ES cells, blotted with the 59 probe. WT (7.0 kb) and targeted (5.5 kb) clone (GL). (D) Breeding frequency to obtain DNA-PKcsPQR and DNA-PKcsSD mice from heterozygous crossing. The p value was calculated based on the Fisher exact test. (E) The body weight of adult DNA-PKcsPQR/PQR, DNA-PKcsSD/SD, and control mice. The bar represents the average and SE. The p value was calculated using a Student t test. There is no significant difference (p . 0.05) by guest on October 1, 2021 between any two genotype groups.

Derivation of murine embryonic fibroblasts and radiation blotted for indicated Abs. The phosphorylation-specific Ab (polyclonal sensitivity assays rabbit) recognizing murine DNA-PKcs (catalog no. BL15933) at the S2053 clusters was developed in collaboration with the Bethyl Laboratories using Murine embryonic fibroblasts (MEFs) were derived from E13.5–E14.5 Bethyl’s standard phosphorylation-specific Ab pipeline and was used at embryo from timed breeding. The individual embryo was manually 1:500 dilution. dissected. Cells were segregated using trypsin (Life Technologies) and cultured in DMEM (Life Technologies) supplemented with 15% FBS. Lymphocyte development and CSR analyses For radiation sensitivity assays, the early passage MEFs (passage 1 or 2) Single-cell suspensions were prepared from thymus, bone marrow, spleen, were plated on 96-well plates (2000 cells per well) in quadruplication. The and lymph node from young adult mice. Splenocytes were treated with RBC plates were subjected to ionizing radiation (IR) at 1, 2, or 5 Gy. The lysis buffer (Lonza ACK Lysis Buffer) for 1–2 min at room temperature. cellularity was determined by CyQUANT Direct Cell Proliferation Assay Approximately 1 3 105 cells were stained using fluorescence-conjugated (Invitrogen) 4 d after irradiation. To measure radiation sensitivity in Abs and analyzed by flow cytometry (13, 18). The following Ab mixtures B cells, purified splenic B cells were activated for CSR (see below), were used for B cells: FITC anti-mouse CD43 (BioLegend; 553270), PE irradiated within 1hr after purification in triplicates or quadruplicates, and Goat anti-mouse IgM (Southern Biotech; 1020-09), PE-Cyanines5 anti- the cell numbers were determined by an automatic cell counter (Invitrogen Hu/Mo CD45R (B220) (eBioScience; 15-0452-83), and allophycocyanin– Countess) 4 d later. Relative survival was calculated using the cell counts anti-mouse TER119 (BioLegend; 116212), For T cells, the following of the unirradiated group as the denominators. Ab mixtures were used: PE rat anti-mouse CD4 (BioLegend; 557308), The development and use of a phosphorylation-specific Ab FITC anti-mouse CD8a (BioLegend; 100706), PE/Cy5 anti-mouse CD3e (eBioscience; 15-0031-83), and allophycocyanin–anti-mouse TCRb (BD For Western blotting, the cells were lysed with SDS-containing total protein Pharmingen 553174). We used the following gate strategies in order when lysis buffer (150 mM NaCl, 10 mM Tris-HCl [pH7.4], 0.1% SDS, 0.1% applicable: 1) the dead cells and debris were excluded based on high-side Triton-100, 1% deoxycholate, and 5 mM EDTA) and sonicated to disrupt scatter and low-forward scatter; 2) for bone marrow and spleen (with genomic DNA. The protein concentration was measured with the DC significant and sometimes variable erythrocytes after RBC lysis), the Protein Assay reagent (Bio-Rad Laboratories). Approximately 100 mg total Ter119+ population (mostly erythrocytes) was gated out; 3) the remaining protein (unless otherwise indicated) was electrophoresed on a 6% poly- cells were analyzed for B or T cell specific markers as shown in Figs. 3 acrylamide gel, transferred to PVDF membrane (with 10% methanol), and and 4. The endogenous V(D)J recombination junctions were analyzed as 4 LYMPHOCYTES DEVELOP WITHOUT DNA-PKcs S2056 PHOSPHORYLATION previously described (13). DNA from purified splenic B cells was ampli- and DNA-PKcsSD/SD mice were born at Mendelian ratios (Fig. 1D), fied via nesting PCR, first with JH_DQ52 (59-TGA TAG GCA CCC AAG are of normal size (Fig. 1E, Supplemental Fig. 1D), and are fertile TAC ACT A-39) and JH4 Int (59-CCT CTC CAG TTT CGG CTG AAT in both males and females, suggesting that phosphorylation at the CC-39) and then with the same JH_DQ52 primer and JH4E (59-AGG CTC TGA GAT CCC TAG ACA G-39). The∼536-bp amplicon corresponding to S2056 cluster is not essential for embryonic development and, by the coding joint between DQ52-JH4 was isolated and cloned into pGEM-T extension, the end-ligation required in postmitotic neurons. (Promega) and sequenced. For the CSR assay, CD43 splenocytes were purified with magnetic beads Radiation-dependent phosphorylation of S2053 in 2 (MACS; Miltenyi Biotec) and cultured (5 3 105 cells ml 1) in RPMI 1640 murine DNA-PKcs (Life Technologies) supplemented with 15% FBS and 2 mgml21 anti-CD40 (BD Biosciences) plus 20 ng ml21 of IL-4 (R&D Systems). Cells were Despite the sequence similarity (Fig.1A),thecommercialAbs maintained daily at 1 3 106 cells ml21 and collected for flow cytometry developed against S2056 phosphorylated human DNA-PKcs were with directly conjugated Abs (FITC anti-IgG1; BD Pharmingen, and PE not able to recognize the corresponding phosphorylation (S2053) Cy5 anti-B220; eBioscience) after gating out dead cells and debris. Flow in murine DNA-PKcs (Supplemental Fig. 1E). This cannot be solely cytometry was performed on a FACSCalibur flow cytometer (BD Biosci- ences), and data were processed using the FlowJo V10 software package. explained by the nearly 20-fold lower abundance of DNA-PKcs B cell radiation sensitivity is detailed in the radiation sensitivity section protein in mouse cells (MEFs) relative to human counterparts above. (Hela) (Fig. 2A, Supplemental Fig. 1E) because commercial Abs against human DNA-PKcs T2609 phosphorylation can ef- High-throughput genomic translocation sequence of fectively recognize corresponding phosphorylation in murine cells CSR junctions (Supplemental Fig. 1E). We noted that there is a proline-to-arginine High-throughput genomic translocation sequence (HTGTS) was performed difference of 3 aa after the conserved S2053 (2056 in human as described previously (23, 39). Briefly, genomic DNA was collected from DNA-PKcs), which could potentially affect Ab recognition. Another Downloaded from activated B cells after 4 d, sonicated (Diagenode Bioruptor), and amplified with an Sm specific biotinylated primer (59-/5BiosG/CAGACCTGG- possibility is that the percentage of DNA-PKcs phosphorylated at GAATGTATGGT-39) and nested primer (59-CACACAAAGACTCTGG- S2053 might be much lower than that at T2609, which could also ACCTC-39). AflII was used to remove GL (nonarranged) sequences. further compromise detection. We reasoned that a phosphorylation- Because all mice are of pure 129 background, the IgH switch region specific Ab developed against pS2053 of murine DNA-PKcs might (from JH4 to the last Ca exon, chr12 114, 494, 415-114, 666, 816) of be able to recognize both human and murine DNA-PKcs phos- the C57/BL6 based mm9 was replaced by the corresponding region in the http://www.jimmunol.org/ AJ851868.3 129 IgH sequence (1415966–1592715) to generate the mm9sr phorylation. In collaboration with the Bethyl Laboratories, we (switch region replacement) genome, and the sequence analyses were developed a novel Ab using phosphorylated murine DNA-PKcs performed as detailed previously (39, 40). The best-path searching al- peptide (corresponding to pS2053) as an immunogen. Our re- gorithm [related to YAHA (41)] was used to identify optimal sequence sult shows that radiation can induce phosphorylation of murine alignments from Bowtie 2–reported top alignments (alignment score .50). The reads were filtered to exclude mis-priming, GL (unmodified), DNA-PKcs at S2053. Importantly, radiation-induced S2053 phos- sequential joints, and duplications. To plot all the S region junctions, phorylation is not detectable in DNA-PKcs null cells and is abolished including those in the repeats but unequivocally mapped to an individual in DNA-PKcsPQR/PQR cells, supporting the specificity of the Ab switch region, we combined the ones filtered by a mappability filter but (Fig. 2B). As predicted, this Ab can effectively detect phosphor- unequivocally mapped to S regions with “good” reads passing both the ylation of human DNA-PKcs at S2056, which is significantly re- mappability (both deduplicated) filters (39). MHs are defined as regions by guest on October 1, 2021 of 100% homology between the bait and prey-break site. Insertions are duced in the presence of a DNA-PKcs–specific inhibitor (NU7441) defined as regions containing nucleotides that map to neither the bait nor (Fig. 2B). Notably, radiation-induced phosphorylation of murine prey-break site. Blunt junctions are considered to have no MHs or in- DNA-PKcs at S2053 was not significantly decreased by either ATM- sertions. The HTGTS data reported in this paper have been deposited in specific inhibitor (KU95533) or DNA-PKcs inhibitor (NU7441) the Gene Expression Omnibus (https://www.ncbi.nlm.nih.gov/geo/) under accession numbers GSE117628 and GSE129895. alone (Fig. 2B), suggesting potential redundancy. Nevertheless, the results, demonstrate for the first time, to our knowledge, that endogenous murine DNA-PKcs can be phosphorylated at S2053 Results PQR PQR/PQR SD/SD upon irradiation and confirm that the DNA-PKcs allele leads DNA-PKcs and DNA-PKcs mice to the expression of DNA-PKcs that cannot be phosphorylated The human DNA-PKcs S2056 phosphorylation cluster contains at S2053. two SQ sites (human 2041/2056, corresponding to murine S2038/ Normal lymphocyte development in DNA-PKcsPQR/PQR and 2053) and three other serine residues (S2029, S2053, and S2055 SD/SD in human, corresponding to S2026, S2050, and S2052 in murine DNA-PKcs mice DNA-PKcs) (Fig. 1A) (24, 33, 34). To interrogate the impact of The cNHEJ pathway is exclusively required for the somatic as- DNA-PKcs S2053 cluster phosphorylation, we generated a mouse sembly of the Ig and TCR gene products via V(D)J recombination. model with alanine substitutions at all five potential phosphor- Next, we examined lymphocyte development in DNA-PKcsPQR/PQR ylation targets (all serine residues) within the S2053 cluster and DNA-PKcsSD/SD mice. The weight and total cellularity of the (DNA-PKcsPQR corresponding to human S2056) (Fig. 1A–C). The lymphoid organs (thymus and spleens) of DNA-PKcsPQR/PQR and desired mutations were confirmed by sequencing the tail DNA from DNA-PKcsSD/SD mice were indistinguishable from age-matched DNA-PKcsPQR/PQR mice and the cDNA from DNA-PKcs+/PQR mice control DNA-PKcs+/+ littermates (Supplemental Fig. 2A, 2B). (Supplemental Fig. 1A). The equal presentation of WT and mutated Flow cytometry analyses showed that the number and frequency bases in the cDNA derived from DNA-PKcs+/PQR cells suggests that of T cell progenitors and mature T cells are indistinguishable the DNA-PKcsPQR allele is expressed normally. In parallel, we also in DNA-PKcsPQR/PQR, DNA-PKcsSD/SD, and littermate control injected the previously generated ES cells with a phospho-mimetic DNA-PKcs+/+ mice (Fig. 3A, 3C, Supplemental Fig. 2A, 2B). mutation at the best-characterized S2056 (S2053D in mouse) Sequential rearrangements of the TCRa locus in CD4+CD8+ (DNA-PKcsSD) (18) for GL transmission (Fig. 1A, Supplemental double positive (DP) thymocytes are coupled with both pos- Fig. 1B). Complete loss of the end-ligation function of the cNHEJ itive and negative selections. Relatively moderate V(D)J recom- pathway (e.g., in Ligase42/2, Xrcc42/2,orDNA-PKcsKD/KD)leads bination defects in ATM- or 53BP1-deficient mice result in reduced to severe postmitotic neuronal apoptosis and Trp53-dependent late surface TCRb and its coreceptor CD3 levels on the DP thymocytes embryonic lethality in mice (18). In this regard, both DNA-PKcsPQR/PQR (Fig. 3A) (42, 43) and a significant reduction of the mature single The Journal of Immunology 5

Fig. 2A), suggesting even expression of ∼50% of DNA-PKcs-PQR polypeptide is sufficient to support lymphocyte development in vivo. Finally, sequence analyses of the nonselective DQ52 to JH4 junctions from B cells derived from DNA-PKcsPQR/PQR and DNA-PKcsSD/SD mice did not find excess nucleotide insertion or deletion (Fig. 3D, Supplemental Tables I, II). Together, these data indicate that phosphorylation at the S2053 (S2056 in human) cluster is not required for the end-ligation and hairpin opening that are necessary for V(D)J recombination and early lymphocyte development. End-ligation during CSR is also unaffected in DNA-PKcsPQR/PQR and DNA-PKcsSD/SD Bcells Mature B cells undergo CSR at the IgH locus to generate Abs with different effector functions. During CSR, a DSB in the upstream m switch region (Sm) is joined to another DSB in a downstream switch region (e.g., Sg1). The end-ligation during CSR is also mediated by the cNHEJ pathway. Consistent with significant ligation defects, DNA-PKcsKD/KD mature B cells have severe CSR defects like Xrcc42/2 B cells (23). In cNHEJ hypomorphic Downloaded from XLF-deficient B cells, chromosomal V(D)J recombination is not significantly affected, but CSR is reduced by nearly 50% (44), suggesting CSR might be more sensitive to mild cNHEJ deficiency. Thus, we analyzed CSR of purified splenic B cells derived from DNA-PKcsPQR/PQR, DNA-PKcsSD/SD mice, and control DNA-PKcs+/+ littermates. In vitro stimulation with anti-CD40 and IL-4 induced http://www.jimmunol.org/ robust CSR to IgG1 and expression of surface IgG1 in∼30% of DNA-PKcs+/+, DNA-PKcsPQR/PQR, and DNA-PKcsSD/SD B cells by day 4 (Figs. 4A, 5C, Supplemental Fig. 2C), suggesting normal CSR. Moreover, the kinetics of CSR to IgG1 is also not affected in DNA-PKcsPQR/PQR and DNA-PKcsSD/SD B cells (Fig. 4C). In the absence of cNHEJ (i.e., Xrcc4-deficiency), a subset of CSR can be completed by the alternative end-joining pathways that preferentially use MH at the junction (19, 20). We used HTGTS by guest on October 1, 2021 (40) to analyze thousands of CSR junctions between the 59 of Sm and its joining partner genome wide, including both internal de- letion (Sm–Sm) and potentially productive switching (Sm–Sg1). In a recent study (23) using HTGTS, we uncovered a prominent shift FIGURE 2. DNA-PKcsPQR allele abolished radiation-induced phosphor- A to MH, extensive erosion within the switch region, and a relative ylation at S2053 of murine DNA-PKcs. ( ) Western blotting analyses show 2/2 that the abundance of DNA-PKcs in human cells (Hela) is between 10- and increase in inversions (marked by red/blue ratio) in DNA-PKcs 50-fold higher than that of MEFs. Anti–DNA-PKcs Ab is used at 1:2000 B cells with only moderate CSR defects (21, 22) (Fig. 5A). dilution (Invitrogen). (B) The phosphorylation-specific Ab developed for Thus, we performed HTGTS analyses of CSR junctions from 2 2 S2053 of murine DNA-PKcs can reliably detect radiation-induced (20 Gy, DNA-PKcsPQR/PQR and control DNA-PKcs / B cells. Although after 1 h) DNA-PKcs phosphorylation at S2053 in murine DNA-PKcs and the results confirmed S region erosion and increased MH usage in S2056 in human DNA-PKcs. DNA-PKcs2/2 B cells, no change in MH usage (Fig. 5A) or Sm junction distribution (Fig. 5B) was noted in DNA-PKcsPQR/PQR positive versus immature DP T cell ratio (Fig. 3B). Yet, surface B cells. Together, these results indicate that phosphorylation at expression of TCRb/CD3 in DP cells and the single positive/DP the S2056 cluster, unlike expression of the kinase-dead DNA-PKcs ratio are unaffected in DNA-PKcsPQR/PQR and DNA-PKcsSD/SD or loss of DNA-PKcs, does not affect cNHEJ during V(D)J mice (Fig. 3A, 3B). Likewise, the frequency of immature pro-B recombination or CSR. (CD43+B220+IgM2), pre-B (CD432B220+IgM2), newly generated naive B (IgM+B220low), and recirculating B (IgM+B220hi)cellsin S2053 cluster phosphorylation does not affect DNA-PK activity the bone marrow from DNA-PKcsPQR/PQR and DNA-PKcsSD/SD or DNA damage response in activated B cells mice are also comparable with those of DNA-PKcs+/+ mice (Fig. 4A). DNA end-ligation in murine B cells with alanine substitutions at the Successful V(D)J recombination at the IgH locus is required for the T2609 clusters (DNA-PKcs3A/3A) is hypersensitive to ATM kinase transition from pro-B to pre-B cells. Thus, DNA-PKcs2/2 mice inhibition, suggesting ATM might phosphorylate other substrates, with hairpin-opening defects, demonstrate a severe blockade in including other sites on DNA-PKcs, to promote end-ligation in pro-B to pre-B cell transition, evidenced by a drop of the pre-B/ DNA-PKcs3A/3A cells (45). Next, we determined whether the pro-B ratio in the bone marrow (Fig. 4B). In contrast, the ratios S2056 mutation affects DNA-PKcs kinase activity and whether of bone marrow–derived pre-B/pro-B cells are comparable in the end-ligation in DNA-PKcsPQR/PQR cells depends on compensa- DNA-PKcsPQR/PQR, DNA-PKcsSD/SD, and the DNA-PKcs+/+mice tory phosphorylation by ATM kinase. IR-induced phosphorylation (Fig. 4B). Furthermore, lymphocyte development and number are of KAP1 and H2AX, two substrates shared by DNA-PK and ATM also normal in DNA-PKcsPQR/- mice with compound heterozygosity (4–6), is not different in DNA-PKcsPQR/PQR and DNA-PKcsSD/SD of the PQR allele and a true null allele (Figs. 3C, 4B, Supplemental B cells compared with DNA-PKcs+/+ controls, both in the presence 6 LYMPHOCYTES DEVELOP WITHOUT DNA-PKcs S2056 PHOSPHORYLATION Downloaded from http://www.jimmunol.org/ by guest on October 1, 2021

FIGURE 3. Normal T lymphocyte development in DNA-PKcsPQR/PQR and DNA-PKcsSD/SD mice. (A) Representative flow cytometry analyses of T cell development in the thymus and spleen of DNA-PKcs+/+, DNA-PKcsPQR/PQR, DNA-PKcsSD/SD, as well as DNA-PKcsSD/SDATM2/2 and ATM2/2 mice. (B) The ratio between the sum of CD4+ or CD8+ single T cells and CD4+CD8+ DP immature T cells in the thymus from mice of different genotypes. The bars represent the average and SD of each genotype (n $ 3 mice). Student t test was used to calculate the p value. There is no significant difference (n.s.) between WT and all DNA-PKcs mutants. **p , 0.01 between WT and Atm2/2 controls. (C) The relative cell number of the different T cell populations in the thymus and spleen. The data represent the cellularity as the percentage of the age-matched WT controls analyzed at the same time. More than three biological repeats were performed for each genotype. The DNA-PKcsPQR/- mice are compound heterozygous for the PQR and the null (-) alleles. The bar represents the average and SE. (D) Comparison of the number of nucleotide deletions and insertions (including both N- and P- elements) in de novo coding joints formed in splenic B cells from DNA-PKcs+/+, DNA-PKcsPQR/PQR,,andDNA-PKcsSD/SD mice. Each dot represents a unique joint (n =19for DNA-PKcs+/+, n = 16 for DNA-PKcsPQR/PQR, and n = 20 for DNA-PKcsSD/SD). The junction sequences are presented in Supplemental Tables I–III (p . 0.05 for both insertion and deletion and for both genotypes). or absence of ATM kinase inhibitor (KU55933) (Fig. 5D). This is development (2, 3) and CSR (6) because of their overlapping consistent with in vitro findings (26) and indicates that S2056 substrates. Consistent with normal kinase activity in DNA-PKcsSD/SD phosphorylation does not affect DNA-PK kinase activity. ATM mice and the lack of dependence on ATM kinase activity, and DNA-PK have critical redundant functions during embryonic DNA-PKcsSD/SDAtm2/2 mice were born at the expected ratio The Journal of Immunology 7

FIGURE 4. Normal B lymphocyte de- velopment and CSR in DNA-PKcsPQR/PQR and DNA-PKcsSD/SD mice. (A)Representative flow cytometry analyses of the develop- ment in the bone marrow and spleen and +/+

in vitro CSR of B cells from DNA-PKcs , Downloaded from DNA-PKcsPQR/PQR, DNA-PKcsSD/SD,aswell as DNA-PKcsSD/SDATM2/2 and ATM2/2 mice. (B) Quantification of bone marrow–Pre-B (B220 plus IgM–CD432)versus–Pro-B (B220 plus IgM–CD43+) B cell ratio in the bone marrow. More than three biological

repeats were performed for each genotype. http://www.jimmunol.org/ (C) The quantification of CSR kinetics in purified B cells activated in vitro. The bar represents the average and SE. The p value was calculated using a Student t test. **p,0.01. n.s., not significant. by guest on October 1, 2021

(Supplemental Fig. 2D). DNA-PK kinase inhibitor (NU7441, 10 mM), in MEFs, we measured IR sensitivity in DNA-PKcsPQR/PQR B cells ATM kinase inhibitor (KU55933, 7.5 mM), or ATM deletion did not activated for CSR. Activated DNA-PKcsPQR/PQR B cells proliferate further reduce the CSR in DNA-PKcsPQR/PQR or DNA-PKcsSD/SD well upon activation (Fig. 6B), consistent with normal development B cells beyond what was observed in DNA-PKcs+/+ B cells of DNA-PKcsPQR/PQR mice and normal CSR of DNA-PKcsPQR/PQR (Figs. 4A, 5C, Supplemental Fig. 2C). Thus, the S2056 phos- B cells. Yet, in two independent experiments, DNA-PKcsPQR/PQR phorylation does not affect DNA-PK kinase activity or its sensi- B cells displayed moderate yet consistent IR sensitivity (Fig. 6C, tivity to ATM inhibition. Supplemental Fig. 2E). Thus, we conclude that phosphorylation of murine DNA-PKcs at the S2053 cluster promotes radiation resis- S2053 cluster affects the radiation sensitivity of both MEFs tance in both MEFs and B cells. Given the normal lymphocyte de- and activated B cells velopment, the results suggest that S2056 phosphorylation might be S2056 cluster phosphorylation was implicated in IR sensitivity in necessary for the repair of a subset of complex DNA lesions gen- human cells (26). Next, we derived primary MEFs from both erated upon radiation, or perhaps S2056 phosphorylation plays a role DNA-PKcsPQR/PQR and DNA-PKcsSD/SD mice. Although DNA-PKcs2/2 in the response to acute pathological insult but not to physiological MEFs are sensitive to IR, as previously reported (15, 16), DNA double strand breaks during lymphocyte development. DNA-PKcsPQR/PQR MEFs are only moderately sensitive to IR, and DNA-PKcsSD/SD MEFs even display a mild resistance to IR (Fig. 6A). Discussion To determine whether S2053 phosphorylation has cell type-dependent In this study, we developed a phosphorylation-specific Ab and roles, which might explain the lack of lymphocyte development showed that murine DNA-PKcs can be phosphorylated at the defects in DNA-PKcsPQR/PQR mice despite the mild IR sensitivity corresponding S2053 site upon radiation. Notably, in contrast to the 8 LYMPHOCYTES DEVELOP WITHOUT DNA-PKcs S2056 PHOSPHORYLATION

FIGURE 5. Normal CSR junction and DNA-PKcs kinase activity in DNA-PKcsPQR/PQR mice. (A) HTGTS analyses of CSR junctions show the MH and insertion (INS) usage among all IgH junctions recovered from DNA-PKcsPQR/PQR (n = 1559), DNA-PKcs+/+ (n = 10,465), and control DNA-PKcs2/2 (n = 717) mice. (B) The distribution of the Sm-Sm joins within the 10-Kb region include the core Sm region. The percentage shows the frequency of junctions that fall down- stream of the core Sm region. The red (in the to centromere orientation) and the Downloaded from blue (in the centromere to telomere orientation) numbers on the right show the percentage and ratio of junctions that fall in given orientations, respectively. Because the IgH locus resides in the telomere to centromere orientation, the normal deletional joints fall in the telomere to centromere orientation (red). (C and D) Western blotting of http://www.jimmunol.org/ ATM and DNA-PKcs substrate phosphorylation upon IR (10 Gy) with or without treatment with ATM inhibitor (KU55933, 10 mM) and DNA-PK kinase inhibitor (NU7441, 10 mM). Primary Abs were used at the following dilutions: anti-H2AX (1:1000; Millipore), anti-gH2AX (1:1000; Millipore), anti-KAP1 (1:1000; Cell Signaling Technology), anti–phospho-KAP1

(S824) (1:1000; Bethyl Laboratories), and anti- by guest on October 1, 2021 actin (1:5000; Sigma Aldrich).

phosphorylation of S2056 in human DNA-PKcs, radiation-induced Nevertheless, careful analyses of V(D)J and CSR junctions show phosphorylation of murine DNA-PKcs at S2053 is not abolished by that S2056 phosphorylation does not affect the quality or the quantity either DNA-PKcs or ATM-specific inhibitors, suggesting potential of the junctions. Collectively, these findings are consistent with the redundancy. Nevertheless, analyses of the DNA-PKcsPQR/PQR, phosphorylation of DNA-PKcs at the S2056 cluster being largely DNA-PKcsPQR/- (presumably ∼50% PQR protein), and DNA-PKcsSD/SD dispensable for physiological DSB repair during lymphocyte de- mice confirmed the role of S2053 phosphorylation in radiation velopment and suggest a potential substrate-specific role of S2056 resistance of both fibroblasts and B cells. Yet unexpectedly, the phosphorylation during DNA repair. in vivo analyses of lymphocyte development clearly indicate This apparent discrepancy between the strong phenotype in that S2053 phosphorylation is dispensable for end-ligation and the kinase-dead model (18) versus the lack of detectable pheno- the hairpin opening required during both V(D)J recombination and type in the S2056 autophosphorylation site mutant is not unique Ig CSR. In this context, DNA-PKcsPQR/PQR MEFs are much less to DNA-PKcs. Similar findings were also noted for ATM kinase sensitive to IR than DNA-PKcs2/2 MEFs, suggesting the S2053 (46–49), forcing us to consider other phosphorylation sites or phosphorylation is only required for a subset of DNA-PKcs– phosphorylation-independent effects of catalysis on DNA-PKcs dependent repair. In contrast to the relatively clean DNA ends and ATM. In this context, the mouse model with alanine substi- generated during lymphocyte development, radiation generates tutions at another DNA-PKcs phosphorylation cluster (T2609) complex DNA lesions with potential base modifications, which (DNA-PKcs3A/3A) succumbed to bone marrow failure shortly after might require additional DNA-PKcs function. Alternatively, the birth (36). Although DNA-PKcs3A/3A B cells are proficient at acute activation of DNA-PKcs upon radiation might require more chromosomal V(D)J recombination, cNHEJ in DNA-PKcs3A/3A robust DNA-PKcs activity than physiological gene rearrangement. B cells are hypersensitive to ATM inhibition (45), suggesting The Journal of Immunology 9

FIGURE 6. DNA-PKcsPQR/PQR cells are moderately sensitive to radiation. (A) The sensitivity to IR (1, 2, 5 Gy) of DNA-PKcs+/+, DNA-PKcs2/2, DNA-PKcsPQR/PQR,and DNA-PKcsSD/SD primary MEFs. The relative survival represents the relative cellularity at the given dose as a proportion of the unirradiated sample of the same ge- notype. The data represent the average and SE of at least five independent wells per genotype per dose. The radiation sensitivity of DNA-PKcsPQR/PQR MEFs is sta- tistically significantly different from both DNA-PKcs+/+ and DNA-PKcs2/2 MEFs at all three doses tested. p , 0.001. (B) Relative fold of growth (by cell counts) for DNA-PKcs+/+ and DNA-PKcsPQR/PQR splenic B cells during CSR (4 d). The bars mark the means and SEs. Downloaded from p = 0.13 by unpaired Student t test. (C) The sensitivity to IR (1, 2, 5 Gy) of DNA-PKcs+/+ and DNA-PKcsPQR/PQR B cells activated for CSR. Two independent pairs of DNA-PKcs+/+ and DNA-PKcsPQR/PQR B cells were assayed in two independent experiments (another is shown in Supplemental Fig. 2E). DNA-PKcsPQR/PQR

B cells are significantly more sensitive to radiation http://www.jimmunol.org/ than the DNA-PKcs+/+ controls at all three radiation doses. p , 0.05. by guest on October 1, 2021 that ATM kinase might phosphorylate other targets or other cells might be vulnerable to other disturbances, including DNA-PKcs– sites of DNA-PKcs to promote cNHEJ. Yet, ATM inhibition or mediated phosphorylation of other sites on DNA-PKcs and the ATM deletion does not further impair the end-ligation required loss of other nonessential cNHEJ factors (e.g., XLF, PAXX, or for lymphocyte development or CSR in DNA-PKcsPQR/PQR and MRI). This possibility should be tested in future experiments. DNA-PKcsSD/SD lymphocytes, suggesting that at least the cNHEJ PQR/PQR function in DNA-PKcs cells cannot be explained by Acknowledgments compensatory phosphorylation via ATM. Although we did not We thank Dr. Eric Mcintush and the Bethyl Laboratories Inc. for helping formally exclude the contribution of ATR kinase, ATR kinase develop the phosphorylation-specific Ab against murine DNA-PKcs. We activity is essential for cell and embryonic viability, and ATR is thank Dr. Chyuan-Sheng (Victor) Lin and Li Chen for technical assistance usually activated by RPA-coated ssDNA, not DSBs (50). In ad- and advice on the generation of the new mouse models. We thank other dition, the catalysis itself might trigger conformational changes members of the Zha laboratory for helpful discussions and technical advice. of DNA-PKcs independent of any specific autophosphorylation We apologize to colleagues whose work could not be cited because of space sites. Early biochemical studies suggested that binding of puri- limitations and was covered by reviews instead. fied DNA-PKcs to DNA is affected by ATP hydrolysis but not the resulted phosphorylation (51). Such catalysis-dependent Disclosures allosteric changes have been reported for PARP1, which is also ac- The authors have no financial conflicts of interest. tivated upon DNA binding (52). The published structure of purified DNA-PKcs adopted a few different conformations (32, 53) but how and which conformations reflect the in vivo structure of References 1. Shiloh, Y. 2003. ATM and related protein kinases: safeguarding genome integ- DNA-PKcs remain elusive. rity. Nat. Rev. Cancer 3: 155–168. In summary, our findings unequivocally demonstrate an im- 2. Gurley, K. E., and C. J. Kemp. 2001. Synthetic lethality between mutation in portant distinction between autophosphorylation and catalysis Atm and DNA-PK(cs) during murine embryogenesis. Curr. Biol. 11: 191–194. 3. Sekiguchi, J., D. O. Ferguson, H. T. Chen, E. M. Yang, J. Earle, K. Frank, itself, indicating that autophosphorylation at S2056, although a S. Whitlow, Y. Gu, Y. Xu, A. Nussenzweig, and F. W. Alt. 2001. Genetic in- good marker for DNA-PK activation, is largely dispensable for teractions between ATM and the nonhomologous end-joining factors in genomic cNHEJ and DNA damage responses required for lymphocyte stability and development. Proc. Natl. Acad. Sci. USA 98: 3243–3248. PQR/PQR 4.Gapud,E.J.,Y.Dorsett,B.Yin,E.Callen,A.Bredemeyer,G.K.Mahowald, development. Although end-ligation in DNA-PKcs cells K. Q. Omi, L. M. Walker, J. J. Bednarski, P. J. McKinnon, et al. 2011. Ataxia tel- and mice are not hypersensitive to ATM loss, the moderate radi- angiectasia mutated (Atm) and DNA-PKcs kinases have overlapping activities during PQR/PQR chromosomal signal joint formation. Proc. Natl. Acad. Sci. USA 108: 2022–2027. ation sensitivity of DNA-PKcs cells suggest the possibility 5. Zha, S., W. Jiang, Y. Fujiwara, H. Patel, P. H. Goff, J. W. Brush, R. L. Dubois, PQR/PQR that the apparently efficient end-ligation in DNA-PKcs and F. W. Alt. 2011. Ataxia telangiectasia-mutated protein and DNA-dependent 10 LYMPHOCYTES DEVELOP WITHOUT DNA-PKcs S2056 PHOSPHORYLATION

protein kinase have complementary V(D)J recombination functions. Proc. Natl. 29. Chen, B. P., N. Uematsu, J. Kobayashi, Y. Lerenthal, A. Krempler, H. Yajima, Acad. Sci. USA 108: 2028–2033. M. Lo¨brich, Y. Shiloh, and D. J. Chen. 2007. Ataxia telangiectasia mutated 6. Calle´n, E., M. Jankovic, N. Wong, S. Zha, H. T. Chen, S. Difilippantonio, M. Di (ATM) is essential for DNA-PKcs phosphorylations at the Thr-2609 cluster upon Virgilio, G. Heidkamp, F. W. Alt, A. Nussenzweig, and M. Nussenzweig. 2009. DNA double strand break. J. Biol. Chem. 282: 6582–6587. Essential role for DNA-PKcs in DNA double-strand break repair and apoptosis 30. Douglas, P., G. P. Sapkota, N. Morrice, Y. Yu, A. A. Goodarzi, D. Merkle, in ATM-deficient lymphocytes. Mol. Cell 34: 285–297. K. Meek, D. R. Alessi, and S. P. Lees-Miller. 2002. Identification of in vitro and 7. Falck, J., J. Coates, and S. P. Jackson. 2005. Conserved modes of recruitment of in vivo phosphorylation sites in the catalytic subunit of the DNA-dependent ATM, ATR and DNA-PKcs to sites of DNA damage. Nature 434: 605–611. protein kinase. Biochem. J. 368: 243–251. 8. Ma, Y., U. Pannicke, K. Schwarz, and M. R. Lieber. 2002. Hairpin opening and 31. Yajima, H., K. J. Lee, and B. P. Chen. 2006. ATR-dependent phosphorylation of overhang processing by an Artemis/DNA-dependent protein kinase complex in DNA-dependent protein kinase catalytic subunit in response to UV-induced nonhomologous end joining and V(D)J recombination. Cell 108: 781–794. replication stress. Mol. Cell. Biol. 26: 7520–7528. 9. Hung, P. J., B. Johnson, B. R. Chen, A. K. Byrum, A. L. Bredemeyer, W. T. 32. Ding, Q., Y. V. Reddy, W. Wang, T. Woods, P. Douglas, D. A. Ramsden, Yewdell, T. E. Johnson, B. J. Lee, S. Deivasigamani, I. Hindi, et al. 2018. MRI is S. P. Lees-Miller, and K. Meek. 2003. Autophosphorylation of the catalytic a DNA damage response adaptor during classical non-homologous end joining. subunit of the DNA-dependent protein kinase is required for efficient end pro- Mol. Cell 71: 332–342.e8 10.1016/j.molcel.2018.06.018. cessing during DNA double-strand break repair. Mol. Cell. Biol. 23: 5836–5848. 10. Ochi, T., A. N. Blackford, J. Coates, S. Jhujh, S. Mehmood, N. Tamura, 33. Cui, X., Y. Yu, S. Gupta, Y. M. Cho, S. P. Lees-Miller, and K. Meek. 2005. J. Travers, Q. Wu, V. M. Draviam, C. V. Robinson, et al. 2015. DNA repair. Autophosphorylation of DNA-dependent protein kinase regulates DNA end PAXX, a paralog of XRCC4 and XLF, interacts with Ku to promote DNA processing and may also alter double-strand break repair pathway choice. Mol. double-strand break repair. Science 347: 185–188. Cell. Biol. 25: 10842–10852. 11. Xing, M., M. Yang, W. Huo, F. Feng, L. Wei, W. Jiang, S. Ning, Z. Yan, W. Li, 34. Meek, K., P. Douglas, X. Cui, Q. Ding, and S. P. Lees-Miller. 2007. Trans Q. Wang, et al. 2015. Interactome analysis identifies a new paralogue of XRCC4 Autophosphorylation at DNA-dependent protein kinase’s two major autophos- in non-homologous end joining DNA repair pathway. Nat. Commun. 6: 6233. phorylation site clusters facilitates end processing but not end joining. Mol. Cell. 12. Kumar, V., F. W. Alt, and R. L. Frock. 2016. PAXX and XLF DNA repair factors Biol. 27: 3881–3890. are functionally redundant in joining DNA breaks in a G1-arrested progenitor 35. Uematsu, N., E. Weterings, K. Yano, K. Morotomi-Yano, B. Jakob, G. Taucher-Scholz, B-cell line. Proc. Natl. Acad. Sci. USA 113: 10619–10624. P. O. Mari, D. C. van Gent, B. P. Chen, and D. J. Chen. 2007. Autophosphorylation of 13. Liu, X., Z. Shao, W. Jiang, B. J. Lee, and S. Zha. 2017. PAXX promotes KU DNA-PKCS regulates its dynamics at DNA double-strand breaks. J. Cell Biol. accumulation at DNA breaks and is essential for end-joining in XLF-deficient 177: 219–229. Downloaded from mice. Nat. Commun. 8: 13816. 36. Zhang, S., H. Yajima, H. Huynh, J. Zheng, E. Callen, H. T. Chen, N. Wong, S. Bunting, 14. Lieber, M. R. 2010. The mechanism of double-strand DNA break repair by the Y. F. Lin, M. Li, et al. 2011. Congenital bone marrow failure in DNA-PKcs mutant nonhomologous DNA end-joining pathway. Annu. Rev. Biochem. 79: 181–211. mice associated with deficiencies in DNA repair. J. Cell Biol. 193: 295–305. 15. Gao, Y., J. Chaudhuri, C. Zhu, L. Davidson, D. T. Weaver, and F. W. Alt. 1998. A 37. Zhang, S., S. Matsunaga, Y. F. Lin, B. Sishc, Z. Shang, J. Sui, H. Y. Shih, targeted DNA-PKcs-null mutation reveals DNA-PK-independent functions for Y. Zhao, O. Foreman, M. D. Story, et al. 2016. Spontaneous tumor development KU in V(D)J recombination. Immunity 9: 367–376. in bone marrow-rescued DNA-PKcs(3A/3A) mice due to dysfunction of telo- 16.Taccioli,G.E.,A.G.Amatucci,H.J.Beamish,D.Gell,X.H.Xiang, mere leading strand deprotection. Oncogene 35: 3909–3918.

M. I. Torres Arzayus, A. Priestley, S. P. Jackson, A. Marshak Rothstein, 38. Zha, S., J. Sekiguchi, J. W. Brush, C. H. Bassing, and F. W. Alt. 2008. Com- http://www.jimmunol.org/ P. A. Jeggo, and V. L. Herrera. 1998. Targeted disruption of the catalytic plementary functions of ATM and H2AX in development and suppression of subunit of the DNA-PK gene in mice confers severe combined immunodefi- genomic instability. Proc. Natl. Acad. Sci. USA 105: 9302–9306. ciency and radiosensitivity. Immunity 9: 355–366. 39. Dong, J., R. A. Panchakshari, T. Zhang, Y. Zhang, J. Hu, S. A. Volpi, R. M. Meyers, 17. Rooney, S., J. Sekiguchi, C. Zhu, H. L. Cheng, J. Manis, S. Whitlow, J. DeVido, Y. J. Ho, Z. Du, D. F. Robbiani, et al. 2015. Orientation-specific joining of AID- D. Foy, J. Chaudhuri, D. Lombard, and F. W. Alt. 2002. Leaky Scid phenotype initiated DNA breaks promotes antibody class switching. Nature 525: 134–139. associated with defective V(D)J coding end processing in Artemis-deficient 40. Hu, J., R. M. Meyers, J. Dong, R. A. Panchakshari, F. W. Alt, and R. L. Frock. mice. Mol. Cell 10: 1379–1390. 2016. Detecting DNA double-stranded breaks in mammalian genomes by linear 18. Jiang, W., J. L. Crowe, X. Liu, S. Nakajima, Y. Wang, C. Li, B. J. Lee, amplification-mediated high-throughput genome-wide translocation sequencing. R. L. Dubois, C. Liu, X. Yu, et al. 2015. Differential phosphorylation of DNA- Nat. Protoc. 11: 853–871. PKcs regulates the interplay between end-processing and end-ligation during 41. Faust, G. G., and I. M. Hall. 2012. YAHA: fast and flexible long-read alignment nonhomologous end-joining. Mol. Cell 58: 172–185. with optimal breakpoint detection. Bioinformatics 28: 2417–2424.

19. Boboila, C., C. Yan, D. R. Wesemann, M. Jankovic, J. H. Wang, J. Manis, 42. Borghesani, P. R., F. W. Alt, A. Bottaro, L. Davidson, S. Aksoy, G. A. Rathbun, by guest on October 1, 2021 A. Nussenzweig, M. Nussenzweig, and F. W. Alt. 2010. Alternative end-joining T. M. Roberts, W. Swat, R. A. Segal, and Y. Gu. 2000. Abnormal development of catalyzes class switch recombination in the absence of both Ku70 and DNA Purkinje cells and lymphocytes in Atm mutant mice. Proc. Natl. Acad. Sci. USA ligase 4. [Published erratum appears in 2013 J. Exp. Med. 210: 641.] J. Exp. 97: 3336–3341. Med. 207: 417–427. 43. Difilippantonio, S., E. Gapud, N. Wong, C. Y. Huang, G. Mahowald, H. T. Chen, 20. Yan, C. T., C. Boboila, E. K. Souza, S. Franco, T. R. Hickernell, M. Murphy, M. J. Kruhlak, E. Callen, F. Livak, M. C. Nussenzweig, et al. 2008. 53BP1 facilitates S. Gumaste, M. Geyer, A. A. Zarrin, J. P. Manis, et al. 2007. IgH class switching long-range DNA end-joining during V(D)J recombination. Nature 456: 529–533. and translocations use a robust non-classical end-joining pathway. Nature 449: 44. Li, G., F. W. Alt, H. L. Cheng, J. W. Brush, P. H. Goff, M. M. Murphy, S. Franco, 478–482. Y. Zhang, and S. Zha. 2008. Lymphocyte-specific compensation for XLF/ 21. Bosma, G. C., J. Kim, T. Urich, D. M. Fath, M. G. Cotticelli, N. R. Ruetsch, cernunnos end-joining functions in V(D)J recombination. Mol. Cell 31: 631–640. M. Z. Radic, and M. J. Bosma. 2002. DNA-dependent protein kinase activity 45. Lee, B. S., E. J. Gapud, S. Zhang, Y. Dorsett, A. Bredemeyer, R. George, E. Callen, is not required for immunoglobulin class switching. J. Exp. Med. 196: J. A. Daniel, O. Osipovich, E. M. Oltz, et al. 2013. Functional intersection of ATM 1483–1495. and DNA-dependent protein kinase catalytic subunit in coding end joining during 22. Manis, J. P., D. Dudley, L. Kaylor, and F. W. Alt. 2002. IgH class switch re- V(D)J recombination. Mol. Cell. Biol. 33: 3568–3579. combination to IgG1 in DNA-PKcs-deficient B cells. Immunity 16: 607–617. 46. Yamamoto, K., Y. Wang, W. Jiang, X. Liu, R. L. Dubois, C. S. Lin, T. Ludwig, 23. Crowe, J. L., Z. Shao, X. S. Wang, P. C. Wei, W. Jiang, B. J. Lee, V. M. Estes, C. J. Bakkenist, and S. Zha. 2012. Kinase-dead ATM protein causes genomic F. W. Alt, and S. Zha. 2018. Kinase-dependent structural role of DNA-PKcs instability and early embryonic lethality in mice. J. Cell Biol. 198: 305–313. during immunoglobulin class switch recombination. Proc. Natl. Acad. Sci. 47. Pellegrini, M., A. Celeste, S. Difilippantonio, R. Guo, W. Wang, L. Feigenbaum, USA 115: 8615–8620. and A. Nussenzweig. 2006. Autophosphorylation at serine 1987 is dispensable 24. Block, W. D., Y. Yu, D. Merkle, J. L. Gifford, Q. Ding, K. Meek, and S. P. Lees- for murine Atm activation in vivo. Nature 443: 222–225. Miller. 2004. Autophosphorylation-dependent remodeling of the DNA- 48. Daniel, J. A., M. Pellegrini, J. H. Lee, T. T. Paull, L. Feigenbaum, and dependent protein kinase catalytic subunit regulates ligation of DNA ends. A. Nussenzweig. 2008. Multiple autophosphorylation sites are dispensable for Nucleic Acids Res. 32: 4351–4357. murine ATM activation in vivo. J. Cell Biol. 183: 777–783. 25. Chan, D. W., B. P. Chen, S. Prithivirajsingh, A. Kurimasa, M. D. Story, J. Qin, 49. Daniel, J. A., M. Pellegrini, B. S. Lee, Z. Guo, D. Filsuf, N. V. Belkina, Z. You, and D. J. Chen. 2002. Autophosphorylation of the DNA-dependent protein ki- T. T. Paull, B. P. Sleckman, L. Feigenbaum, and A. Nussenzweig. 2012. Loss of nase catalytic subunit is required for rejoining of DNA double-strand breaks. ATM kinase activity leads to embryonic lethality in mice. J. Cell Biol. 198: 295–304. Genes Dev. 16: 2333–2338. 50. Menolfi, D., W. Jiang, B. J. Lee, T. Moiseeva, Z. Shao, V. Estes, M. G. Frattini, 26. Chen, B. P., D. W. Chan, J. Kobayashi, S. Burma, A. Asaithamby, K. Morotomi- C. J. Bakkenist, and S. Zha. 2018. Kinase-dead ATR differs from ATR loss by Yano, E. Botvinick, J. Qin, and D. J. Chen. 2005. Cell cycle dependence of limiting the dynamic exchange of ATR and RPA. Nat. Commun. 9: 5351. DNA-dependent protein kinase phosphorylation in response to DNA double 51. Jovanovic, M., and W. S. Dynan. 2006. Terminal DNA structure and ATP in- strand breaks. J. Biol. Chem. 280: 14709–14715. fluence binding parameters of the DNA-dependent protein kinase at an early step 27. Douglas, P., X. Cui, W. D. Block, Y. Yu, S. Gupta, Q. Ding, R. Ye, N. Morrice, prior to DNA synapsis. Nucleic Acids Res. 34: 1112–1120. S. P. Lees-Miller, and K. Meek. 2007. The DNA-dependent protein kinase cat- 52. Langelier, M. F., L. Zandarashvili, P. M. Aguiar, B. E. Black, and J. M. Pascal. alytic subunit is phosphorylated in vivo on threonine 3950, a highly conserved 2018. NAD+ analog reveals PARP-1 substrate-blocking mechanism and allo- amino acid in the protein kinase domain. Mol. Cell. Biol. 27: 1581–1591. steric communication from catalytic center to DNA-binding domains. Nat. 28. Merkle, D., P. Douglas, G. B. Moorhead, Z. Leonenko, Y. Yu, D. Cramb, Commun. 9: 844. D. P. Bazett-Jones, and S. P. Lees-Miller. 2002. The DNA-dependent protein 53. Dobbs, T. A., J. A. Tainer, and S. P. Lees-Miller. 2010. A structural model for kinase interacts with DNA to form a protein-DNA complex that is disrupted by regulation of NHEJ by DNA-PKcs autophosphorylation. DNA Repair (Amst.) 9: phosphorylation. Biochemistry 41: 12706–12714. 1307–1314. A. SupFigure 1 CTA and TTA both encode Leucine. PQR allele +/PQR

C. B. S2053D mutation WT-TATTCATACAGTTCTCAAGAT Y S Y S S Q D Marker PQR/PQR PQR/PQR +/PQR +/PQR H2O ctrl SD-TATTCATACTCTGATCAAGAT 500bp- Y S Y S D Q D 400bp- -384bp(PQR) TATTCATACTCTGATCAAGAT 300bp- -280bp(WT) 200bp-

MEFs Hela E. +/+ IR -+++-++ ATMi--+---+ PKi ---+--- D. pS2056(h) - DNA-PKcs Genotype n= Weight(g) -250KDa +/+ 9 22.3±3.0 DNA-PKcs -250KDa PQR/PQR 9 25.3±3.9 pT2609(h) - PQR/- 2 22.9±1.4 -250KDa -150KDa SD/SD 9 21.5±3.4 Vinculin -100KDa +/SD 3 22.2±1.9

Supplementary Figure 1: The alignment of genomic and predicted protein sequence of the DNA- PKcsPQR (A) and DNA-PKcsSD (B) allele from homozygous DNA-PKcsPQR/PQR, DNA-PKcsSD/SD mice. The synonymous substitution (CTA->TTA) generates a new restriction site (CTTAAG) for AfiII, which was used for screening. (C) A set of representative genotyping results from crosses between DNA- PKcsPQR/+ parents. The PCR products corresponding to the WT allele is 280bp, to PQR is 387bp. (D) The body weight of adult DNA-PKcsPQR/PQR, DNA-PKcsPQR/- (compound heterozygous with null), DNA-PKcs+/SD DNA-PKcsSD/SD mice (averager standard error). Based on student t test, there is no significant difference between any two genotype groups. (E) Western blotting using commercial antibodies developed against pS2056 of human DNA-PKcs (Bethyl Laboratories Inc., BL2423, 1:500) SupFigure2

A. At 8-10 week Weight (mg) Count (106) Thymus Spleen Thymus Spleen WT (n=3) 38.90±11.24 95.93±3.56 68.5±33.2 64.8±18.8 PQR/PQR(n=3) 34.33r9.72 80.07r13.15 58.57r11.9 52.2r14.9 PQR/- (n=3) 41.25±6.45 92.50±35.6 61.5±11.3 64.5±35.6

B. At 4 week Count (106) Thymus Spleen WT (n=4) 173.4±34.3 65.9±9.8 SD/SD(n=11) 127.8r46.9 59.8r19.7 +/SD (n=3) 160.8±35.4 56.9±19.9

C. 45% E. 40% 100% 35% +/+(2) 30% 25% 80% PQR/PQR(2) 20% 15% 60% 10% IgG1+ among B cells IgG1+ 5% 0% 40% WT PQR WT PQR WT PQR D. ATMiATMi Pki Pki 20% Relative Cell number to no-IR Relative Cell number to ATM -/- +/- +/+ p 0% SD/SD 5116 012345 PKcs +/+ 26 57 32 >0.9 Irradiation dose (Gy)

Supplementary Figure 2 (A) Thymus and Spleen weight (g) and totally cellularity in DNA-PKcsPQR/PQR and control mice. B) Total cell count of thymus and spleen from DNA-PKcsSD/SD and DNA-PKcs+/+ littermates. The data represents the average and standard deviation of >3 mice of each genotype. (C) The percentage of IgG1+ cells among all B220+ B cells at day 4 of class switch. The data represents the average and standard deviation of 3 independent experiments. ATM inhibitor KU55933 was used at final concentration of 7.5μM and DNA-PKcs inhibitor NU7441 was used at a final concentration of 2.5 μM. (D) The distribution of ATM genotyping from breeding between DNA-PKcsSD/SDATM+/- parents. The breeding between ATM+/- mice were used as control. P was calculated using the Fisher exact test. (E) The independent repeats of radiation sensitivity of B cells activated for class switch recombination. The activated B cells were irradiated on the day of purification and cytokine stimulation (anti-CD40 and IL4) and the cell counts were determined 4 days later. SupTable 1 CODINGJOINTs_ǦLEFT P N P RIGHT N+P= Del= GL AAGCGGAGCACCACAGTGCTAACTGGGACC ACTATGCTATGGACTACTGGGGTCAAGGAA 1 AAGCGGAGCACCACAGTGCTAACTGG CTATGGACTACTGGGGTCAAGGAA 010 2 AAGCGGAGCACCACAGTGCTAACTGGGA TGCTATGGACTACTGGGGTCAAGGAA 06 3 AAGCGGAGCACCACAGTGCTAACTG TGCTATGGACTACTGGGGTCAAGGAA 09 4 AAGCGGAGCACCACAGTGCTAACTGGG T ACTATGCTATGGACTACTGGGGTCAAGGAA 13

+/+ 5 AAGCGGAGCACCACAGTGCTAACTGGGA CG TGCTATGGACTACTGGGGTCAAGGAA 26 6 AAGCGGAGCACCACAGTGCTAACTGGGACC GG ATGCTATGGACTACTGGGGTCAAGGAA 23 7 AAGCGGAGCACCACAGTGCTAACTGGGA GGG ACTATGCTATGGACTACTGGGGTCAAGGAA 32 8 AAGCGGAGCACCACAGTGC CCA ACTATGCTATGGACTACTGGGGTCAAGGAA 311 9 AAGCGGAGCACCACAGTGCT CCG TATGGACTACTGGGGTCAAGGAA 317 10 AAGCGGAGCACCACAGTGCTAACTGGGACC CC T ACTATGCTATGGACTACTGGGGTCAAGGAA 30 11 AAGCGGAGCACCACAGTGCTAACTGGGAC GGG TGCTATGGACTACTGGGGTCAAGGAA 35 12 AAGCGGAGCACCACAGTGCTAACTG AGGG ATGCTATGGACTACTGGGGTCAAGGAA 48 13 AAGCGGAGCACCACAGTGCTAACTGGGA AACG ATGCTATGGACTACTGGGGTCAAGGAA 45 DNA-PKcs 14 AAGCGGAGCACCACAGTGCTAACTG AGGG ATGCTATGGACTACTGGGGTCAAGGAA 48 15 AAGCGGAGCACCACAGTG TCTT TATGGACTACTGGGGTCAAGGAA 419 16 AAGCGGAGCACCACAGTGCTAACTGGGA GTTTT TATGCTATGGACTACTGGGGTCAAGGAA 54 17 AAGCGGAGCACCACAGTGCTAACTGGGAC GT GGT CTATGCTATGGACTACTGGGGTCAAGGAA 52 18 AAGCGGAGCACCACAGTGCTAACTG ACCCT TCC GGACTACTGGGGTCAAGGAA 815 19 AAGCGGAGCACCA GACGGAGA CTATGGACTACTGGGGTCAAGGAA 823 1 AAGCGGAGCACCACAGTGCTAACTGGGAC T TGCTATGGACTACTGGGGTCAAGGAA 15 2 AAGCGGAGCACCACAGTGCTAACTG TT CTATGCTATGGACTACTGGGGTCAAGGAA 26 3 AAGCGGAGCACCACAGTGCTAACTGGGA AGG TGCTATGGACTACTGGGGTCAAGGAA 36 4 AAGCGGAGCACCACAGTGCTAACTGGG TC T ACTATGCTATGGACTACTGGGGTCAAGGAA 33 5 AAGCGGAGCACCACAGTGCTAACTGGGAC GGG GGACTACTGGGGTCAAGGAA 311 6 AAGCGGAGCACCACAGTGCTAACTGGGAC GT A ACTATGCTATGGACTACTGGGGTCAAGGAA 31 PQR/PQR 7 AAGCGGAGCACCACAGTGCTAACTGGGAC GAATACTATGCTATGGACTACTGGGGTCAAGGAA 41 8 AAGCGGAGCACCACAGTGCTAACTGGGAC GT AG CTATGCTATGGACTACTGGGGTCAAGGAA 42 9 AAGCGGAGCACCACAGTGCTAACTGGGAC G GGA CTATGCTATGGACTACTGGGGTCAAGGAA 42 10 AAGCGGAGCACCACAGTGCTAACTGGGAC G GGG GCTATGGACTACTGGGGTCAAGGAA 46 11 AAGCGGAGCACCACAGTGCTAACT CTATT ACTATGCTATGGACTACTGGGGTCAAGGAA 56 12 AAGCGGAGCACCACAGTGCTAACTGGGAC GGGGTACTATGCTATGGACTACTGGGGTCAAGGAA 51

DNA-PKcs 13 AAGCGGAGCACCACAGTGCTAACTGGG GCCCT CTATGCTATGGACTACTGGGGTCAAGGAA 54 14 AAGCGGAGCACCACAGTGCTAACTGGGAC GT AGGG GCTATGGACTACTGGGGTCAAGGAA 66 15 AAGCGGAGCACCACAGTGCTAACTGGGAC GATATTACTATGCTATGGACTACTGGGGTCAAGGAA 61 16 AAGCGGAGCACCACAGTGCTA T CGGGAG GCTATGGACTACTGGGGTCAAGGAA 714 SupTable 1 Endogenous DQ52-JH4 junctions from peripheral B cells in DNA-PKcs+/+ (n=19) and DNA-PKcsPQR/PQR (n=16) B cells mice. GL= Germline line sequence; P=palindromic (p) nucleotides; N=nontemplate-encoded nucleotides. Del=deletion, from the predicated GL seq. SupTable 2

CODING JOINTs _Ǧ LEFT P N P RIGHT N+P= Del= GL AAGCGGAGCACCACAGTGCTAACTGGGACC ACTATGCTATGGACTACTGGGGTCAAGGAA

1 AAGCGGAGCACCACAGTGCTAACTGGGACC GG G GCTATGGACTACTGGGGTCAAGGAA 35 2 AAGCGGAGCACCACAGTGCTAACTGGGAC TGCTATGGACTACTGGGGTCAAGGAA 05 3 AAGCGGAGCACCACAGTGCTAACTGGGAC TATGCTATGGACTACTGGGGTCAAGGAA 03 4 AAGCGGAGCACCACAGTGCTAACTGGGAC GGT TATGCTATGGACTACTGGGGTCAAGGAA 33 5 AAGCGGAGCACCACAGTGCTAACTGGGAC GA GCTATGGACTACTGGGGTCAAGGAA 26 6 AAGCGGAGCACCACAGTGCTAACTGGGAC GGCGCTATGGACTACTGGGGTCAAGGAA 36 7 AAGCGGAGCACCACAGTGCTAACTGGGAC GAGG GCTATGGACTACTGGGGTCAAGGAA 46 8 AAGCGGAGCACCACAGTGCTAACTGGGAC GAC CTATGCTATGGACTACTGGGGTCAAGGAA 32 9 AAGCGGAGCACCACAGTGCTAACTGGGAC G GGACTTA ATGGACTACTGGGGTCAAGGAA 89 10 AAGCGGAGCACCACAGTGCTAACTGGGAC GGGATTACTATGCTATGGACTACTGGGGTCAAGGAA 61 11 AAGCGGAGCACCACAGTGCTAACTGGGA GGGAGAGT G CTATGCTATGGACTACTGGGGTCAAGGAA 93 12 AAGCGGAGCACCACAGTGCTAACTGGG GGGGTGTC TGGACTACTGGGGTCAAGGAA 812 13 AAGCGGAGCACCACAGTGCTAACTGGG CC TC CTATGCTATGGACTACTGGGGTCAAGGAA 44 14 AAGCGGAGCACCACAGTGCTAACTGGG TC ATGCTATGGACTACTGGGGTCAAGGAA 26 15 AAGCGGAGCACCACAGTGCTAACTGGG ACTGGGGTCAAGGAA 018 16 AAGCGGAGCACCACAGTGCTAACTGGG T ACTATGCTATGGACTACTGGGGTCAAGGAA 13 17 AAGCGGAGCACCACAGTGCTAACTGG TATGCTATGGACTACTGGGGTCAAGGAA 06 18 AAGCGGAGCACCACAGTGCTAACTG TTACTATGCTATGGACTACTGGGGTCAAGGAA 25 19 AAGCGGAGCACCACAGTGCTAACT CCCT T ACTATGCTATGGACTACTGGGGTCAAGGAA 56 20 AAGCGGAGCACCACAGT A TGAGGAT T ACTATGCTATGGACTACTGGGGTCAAGGAA 913

Supplementary Table 2 Endogenous DQ52-JH4 junctions (n=20) from peripheral DNA-PKcsSD/SD B cells. GL= Germline line sequence; P=palindromic (p) nucleotides; N=nontemplate-encoded nucleotides. Del=deletion, indicating deletion from the predicated GL seq.