1. Introduction
In mammalian cells, DNA double-strand breaks (DSBs) activate the DNA damage response signaling (DDR). During DDR, Ataxia telangiectasia mutated (ATM) protein kinase phosphorylates multiple substrates, including histone H2AX and the scaffold proteins, mediator of DNA damage checkpoint protein 1 (MDC1) and p53-binding protein 1 (53BP1) [
1]. The E3 ubiquitin ligases, really interesting new gene (RING) finger (RNF) 8 and RNF168, function downstream of the ATM to enhance 53BP1 binding, which, in turn, facilitates the recruitment of DDR effectors, Pax transactivation domain-interacting protein (PTIP), and Rap1-interacting factor 1 (RIF1) [
1]. Moreover, methylated [
2,
3,
4] and acetylated [
5] histones may facilitate the DDR. In particular, histone H4 lysine 20 di-methylation (H4K20me2) [
3] and histone H3 lysine 79 mono- and di-methylation (H3K79me1/2) [
4] were thought to facilitate recruitment of 53BP1 to the sites of damaged DNA. Homologous recombination (HR), classical non-homologous end joining (NHEJ), and alternative end joining (A-EJ) are cellular pathways that recognize and repair DSBs. NHEJ is initiated by the recruitment of the core Ku70/Ku80 (Ku) sensor to the DSB sites. Ku facilitates the recruitment of downstream factors, including the DNA-dependent protein kinase, catalytic subunit (DNA-PKcs), and the NHEJ core factors DNA ligase 4 (Lig4) and X-ray repair cross-complementing protein 4 (XRCC4). A number of NHEJ proteins, including accessory factors, stabilize the DNA repair complex and process DNA overhangs to facilitate ligation [
1]. Among them, nuclease Artemis [
6], XRCC4-like factor (XLF, or Cernunnos) [
7,
8], a paralogue of XRCC4 and XLF (PAXX) [
9,
10,
11], and modulator of retrovirus infection (Mri) [
12,
13].
During the B and T lymphocyte development, both DDR and NHEJ pathways function in response to the recombination activating gene (RAG)-induced DSBs in the process known as the variable (V), diversity (D) and joining (J) gene segments recombination (V(D)J recombination). RAG is the nuclease that generates DSBs adjacent to the
V,
D, and
J gene segments of immunoglobulin and T cell receptor genes. NHEJ is the only known process to recognize and efficiently repair RAG-induced DSBs [
1,
14]. V(D)J recombination is ablated in mice lacking core NHEJ factors, Ku70 [
15] and Ku80 [
16]. Inactivation of
XRCC4 or
Lig4 resulted in embryonic lethality in mice, while conditional inactivation or knocking down of
XRCC4 or
Lig4 in lymphocytes blocked the V(D)J recombination and NHEJ [
1,
17,
18]. Accessory NHEJ factors DNA-dependent protein kinase, catalytic subunit (DNA-PKcs) and Artemis are required for the V(D)J recombination-associated DNA repair. Artemis is a nuclease that processes RAG-induced hairpin-sealed DNA ends, and DNA-PKcs is required to both structurally stabilize and phosphorylate Artemis [
6,
19,
20,
21,
22,
23]. On the contrary, germline inactivation of
XLF [
24,
25],
PAXX [
26,
27,
28,
29], or
Mri [
12,
13] had no or modest impact on the DNA repair and lymphocyte development in general, and the V(D)J recombination in particular. Combined inactivation of XLF and PAXX resulted in the V(D)J recombination defect in cells [
30,
31,
32] and synthetic lethality in mice [
26,
28,
29,
33]. Moreover, XLF is functionally redundant with DNA-PKcs [
33,
34,
35], Mri [
12,
13], and RAG2 [
36].
DDR factors were thought to be dispensable for the V(D)J recombination, because germline inactivation of
ATM [
37],
H2AX [
38,
39],
MDC1 [
40], or
53BP1 [
41] resulted in modest or no effect on early stages of B and T lymphocyte development. Strikingly, combined inactivation of
XLF and
ATM [
42], or
XLF and
53BP1 [
43,
44], resulted in live-born mice with nearly no mature B and T lymphocytes due to the impaired V(D)J recombination. Additional ATM-dependent DDR factors, including MDC1, may be involved in the V(D)J recombination, and their functions might be revealed in the
XLF-deficient background [
1,
42,
43,
44].
XLF is the NHEJ factor. Mutations in the
XLF gene in humans result in combined immunodeficiency [
8,
45], and inactivation of the
XLF gene in mice results in a modest reduction of B and T lymphocytes count [
24,
25]. XLF shares a structure with XRCC4, and binds XRCC4 to stimulate the Lig4 activity [
7]. XLF has a yeast homolog Nej1 that also stimulates the DNA repair in yeast [
46]. Moreover, the lack of XLF results in increased levels of medulloblastoma in
Trp53-deficient mice [
24]. Together, these observations place XLF to the group of “core” NHEJ factors. MDC1 is a DNA damage response protein acting downstream of ATM and upstream of 53BP1 [
47]. Like XLF, the MDC1 has no enzymatic activity and likely stabilizes the DNA repair complex and facilitates the recruitment of other DNA repair factors. Both MDC1 and XLF can be phosphorylated by ATM and likely by DNA-PKcs to regulate their functions in DNA repair [
1]. Moreover, both XLF and MDC1 were proposed to tether the DNA at the DSB sites before the DNA ligation [
1,
48].
Here, we generated MDC1−/−XLF−/− double-knockout cell lines and demonstrated that MDC1 is stimulating the V(D)J recombination in cells lacking XLF. Moreover, we demonstrated that combined inactivation of MDC1 and XLF resulted in synthetic lethality in mice.
2. Materials and Methods
2.1. Generation of Abelson Murine Leukemia Virus-Transformed (vAbl) Cell Lines
Eμ-Bcl2+ and
XLF−/−Eμ-Bcl2+ vAbl cells were published earlier [
34,
42,
43]. Five independent clones of
MDC1−/−Eμ-Bcl2+ were generated using two three-week-old mice following the procedure described previously [
34,
42,
43,
49,
50]. Additionally, the
XLF gene was inactivated in
Eμ-Bcl2+ vAbl cells to obtain
XLF−/−Eμ-Bcl2+ cell lines, and in
MDC1−/−Eμ-Bcl2+ to generate
MDC1−/−XLF−/−Eμ-Bcl2+ vAbl lines, using the clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein 9 (Cas9) gene-editing approach as described earlier [
51]. Briefly, oligonucleotides corresponding to single guide RNAs (sgRNAs) were cloned into the plasmid vector LentiCRISPR v2 (Addgene plasmid #52961, Addgene, Watertown, MA, USA) [
52]. The following sgRNAs were used to target
exon 3 of the
XLF gene: sgRNA1_FWD: 5′-CTTAGCATACACCAACTTC-3′; sgRNA1_REV: 5′-GAAGTTGGTGTATGCTAAG-3′; sgRNA2_FWD: 5′-CCACCAACAGGTACTCATA-3′; sgRNA2_REV: 5′-TATGAGTACCTGTTGGTGG-3′. Parental vAbl cells were transduced with lentiviral vectors containing corresponding sgRNA sequences, and up to 200 clones were screened by western blot. The cells lacking the XLF signal were used to validate the deletion of the
exon 3 by DNA sequencing (available upon request). Two
XLF−/− clones and four
MDC1−/−XLF−/− clones were used for experiments. Mock-treated and parental vAbl cells were used as DNA repair-proficient controls.
2.2. Antibodies
The following antibodies were used for western blot: rabbit polyclonal anti-XLF (Bethyl, Montgomery, TX, USA; A300-730A, dilution 1:2000), swine polyclonal anti-rabbit immunoglobulin-horseradish peroxidase-conjugated (Ig-HRP; Dako antibodies, Dako, Glostrup, Denmark; #P0399, dilution 1:5000), mouse monoclonal anti-β-actin (Abcam, Cambridge, UK; ab8226, dilution 1:2000), rabbit polyclonal anti-mouse Ig-HRP (Dako antibodies, Dako, Glostrup, Denmark; #P0260, dilution 1:5000), and goat polyclonal anti-mouse Ig-HRP (Dako antibodies, Dako, Glostrup, Denmark; #P0447, dilution 1:5000).
2.3. Variable (V), Diversity (D) and Joining (J) Gene Segments Recombination (V(D)J Recombination) Assays Based on Chromosomally Integrated pMX Cassettes
V(D)J recombination assays were performed using chromosomally-integrated
pMX inversion (
pMX-INV) and
pMX deletion (
pMX-DEL) substrates, as previously described [
34,
42,
43,
49,
50]. In the
pMX-INV cassette, the
green fluorescent protein (
GFP) gene is placed in the reversed orientation and the GFP protein is not expressed. Upon the RAG-induced V(D)J recombination, the
GFP gene is placed in the sense orientation leading to the GFP protein expression. The GFP protein is then detected by flow cytometry to estimate the V(D)J recombination efficiency in indicated vAbl cells [
42,
49,
50]. For the Southern blot-based experiments, we used chromosomally-integrated
pMX-DEL cassettes. During the V(D)J recombination, the
pMX-DELCJ cassette results in an intermediate product with hairpin-sealed coding ends that require Artemis nuclease activity to open the hairpins prior DNA ligase 4-dependent DNA ligation, leading to coding joints (CJ). On the contrary, the
pMX-DELSJ cassette results in the RAG-dependent generation of blunt signal ends (SE) that can be directly ligated by DNA ligase 4 and do not require Artemis nuclease activity, leading to signal joints (SJ) [
34,
42,
43,
49,
50].
2.4. Mice
All experiments involving mice were performed according to the protocols approved by the Norges teknisk-naturvitenskapelige universitet (NTNU), FOTS#8319.
MDC1+/− [
40],
XLF+/Δ [
24], and
Eμ-Bcl2+ [
53] mice were described previously. The
Eμ-Bcl2+ transgenic mice were used to generate vAbl pre-B cells and increase cell survival during the experimental procedures [
49].
2.5. Proliferation Assay
Fifty thousand vAbl cells were plated in 2 mL of Roswell Park Memorial Institute (RPMI) medium in triplicates into 6-well plates. Similarly, fifty thousand human haploid 1 (HAP1) cells were plated in Iscove’s Modified Dulbecco’s Medium (IMDM; Thermo Fisher, Waltham, MA, USA; 21980065) and supplemented with 10% fetal bovine serum, FBS (Sigma, St. Louis, MO, USA; F7524), and 1% penicillin-streptomycin (Thermo Fisher, Waltham, MA, USA; 15140122) at 37 °C with 5% CO
2, according to the manufacturer’s instructions.
MDC1∆ HAP1 cells are nearly haploid human cells that were custom-generated by request and provided by Horizon Discovery (Waterbeach, Cambridge, UK; HZGHC005077c003). The HAP1 cells are human, nearly haploid cell lines derived from the chronic myelogenous leukemia (CML) cell line (KMB-7). The HAP1 model has been recently used to develop knockout human cells (e.g., References [
13,
33,
51,
54]).
Both vAbl and HAP1 cells were counted every 24 h using a Countess™ Automated Cell Counter (Invitrogen, Carlsbad, CA, USA) with Trypan blue staining (Invitrogen, Carlsbad, CA, USA) and bright-field detection. Statistical analyses were performed using GraphPad Prism 8 (La Jolla, CA, USA), one-way analysis of variance (ANOVA), and t-test.
4. Discussion
Inactivation of
RAG and most of the known NHEJ factor genes in mice leads to immunodeficiency [
12,
56]. Recently, we and others found that single inactivation of
XLF,
PAXX, or
Mri genes results in mice with the nearly normal immune system, due to the overlapping functions between XLF and PAXX [
26,
27,
28,
29,
33], as well as XLF and Mri [
12,
13] (
Table 2). The ATM-dependent DDR pathway was initially thought to be dispensable for the V(D)J recombination, although more recent studies using combined genetic inactivation of
XLF and
ATM [
42], as well as
DNA-PKcs and
ATM [
21,
57], revealed that ATM is indeed involved in the early stages of B and T lymphocyte development and its function is partially compensated by XLF and DNA-PKcs. Later, we and others found that ATM substrates, H2AX and 53BP1, are also required for B and T lymphocyte development due to their functions in V(D)J recombination [
42,
43,
44] (
Table 2). Here, we show that another ATM substrate, MDC1, is involved in the V(D)J recombination and its function is compensated in WT cells by XLF. Combined inactivation of
ATM and
XLF, or
53BP1 and
XLF, resulted in immunodeficient mice of smaller sizes than single knockouts or wild-type controls, with abrogated NHEJ, resembling
Ku70−/− or
Ku80−/− knockouts [
1,
42,
43,
44]. Differently, combined inactivation of
DNA-PKcs and
XLF [
34,
35],
H2AX and
XLF [
42], or
MDC1 and
XLF ([
33]; and this study) resulted in embryonic lethality in mice, challenging genetic interaction studies in vivo (
Table 2). One option to overcome this obstacle is to develop conditional knockouts allowing inactivation of
DNA-PKcs,
XLF, or
MDC1 in developing B and T lymphocytes in adult mice. An alternative option is to develop more complex mouse models using. for example,
p53−/− or
p53+/− backgrounds, allowing for the rescue of embryonic lethality (e.g., References [
33,
35]).
Knocking out genes of interest in cell lines may complement and sometimes substitute in vivo experiments using transgenic mice. In particular, vAbl cell lines can be modified using the CRISPR/Cas9 gene-editing approach and serve as a model system to elucidate the specific roles of a particular gene (e.g., References [
30,
31,
32,
50]). Moreover, human, nearly haploid HAP1 cells derived from the KMB-7 cell lines have been recently used to develop genetically-modified cells (e.g., References [
13,
33,
51,
54]).
It becomes more accepted that the DDR pathway contributes to the V(D)J recombination in developing B and T lymphocytes [
1,
34,
42,
43,
44]. However, the mechanistic aspects underlying the specific roles of the DDR factors in this process remain unclear. One can speculate that DDR factors share the functions with XLF, e.g., by stabilizing the DNA repair complex or supporting timely recruitment and dissociation of the NHEJ factors. The DDR pathway may also contribute to distinct but complementary XLF aspects of the DNA repair, e.g., by recruiting the downstream enzymes, supporting the DNA damage-induced post-translational modifications of DNA repair factors and histones, or protecting the free DNA ends from the nuclease-dependent processing before the DNA ligation step [
1,
34,
42,
43,
44,
55]. In particular, the role of MDC1 during the V(D)J recombination might be to stabilize the DNA repair complex, to protect the free DNA ends, to ensure efficient recruitment of downstream DDR factors, such as 53BP1, PTIP, RIF1, Shieldin, etc. [
1,
42,
43,
44,
47,
55,
58], or to exit from the G1 phase of the cell cycle following the RAG-induced DSB [
59]. Further research is required to identify specific roles of MDC1 and XLF in DNA repair.
The proliferation rate of vAbl cells lacking both XLF and MDC1 was reduced when compared to single-deficient and WT controls (
Figure 1) at 72 h. Moreover, proliferation rates of MDC1-deficient cells were also reduced when compared to WT, although not significant. Furthermore, the lack of MDC1 alone resulted in significantly reduced proliferation rates of human HAP1 cells at 96 and 120 h (
Figure 1). These observations may suggest that, first, the lack of MDC1 is compensated by the presence of XLF in murine cells, and second, that the MDC1 is required for efficient DNA repair and proliferation of human cells, likely by supporting the cell cycle progression and DNA damage tolerance [
47,
59].