Generation of a Mouse Model Lacking the Non-Homologous End-Joining Factor Mri/Cyren

Classical non-homologous end joining (NHEJ) is a molecular pathway that detects, processes, and ligates DNA double-strand breaks (DSBs) throughout the cell cycle. Mutations in several NHEJ genes result in neurological abnormalities and immunodeficiency both in humans and mice. The NHEJ pathway is required for V(D)J recombination in developing B and T lymphocytes, and for class switch recombination in mature B cells. The Ku heterodimer formed by Ku70 and Ku80 recognizes DSBs and facilitates the recruitment of accessory factors (e.g., DNA-PKcs, Artemis, Paxx and Mri/Cyren) and downstream core factor subunits X-ray repair cross-complementing group 4 (XRCC4), XRCC4-like factor (XLF), and DNA ligase 4 (Lig4). Accessory factors might be dispensable for the process, depending on the genetic background and DNA lesion type. To determine the physiological role of Mri in DNA repair and development, we introduced a frame-shift mutation in the Mri gene in mice. We then analyzed the development of Mri-deficient mice as well as wild type and immunodeficient controls. Mice lacking Mri possessed reduced levels of class switch recombination in B lymphocytes and slow proliferation of neuronal progenitors when compared to wild type littermates. Human cell lines lacking Mri were as sensitive to DSBs as the wild type controls. Overall, we concluded that Mri/Cyren is largely dispensable for DNA repair and mouse development.


Mouse Genotyping
Two polymerase chain reactions (PCRs) were designed to determine the mouse genotypes. The first PCR was performed using TCAGGTCTGCCCTACACTGA and GTGGTGGTGCTTCTCTGTGA primers, detecting both wild type (428 bp) and null (414 bp) alleles ( Figure 1B). The second PCR performed with TCAGGTCTGCCCTACACTGA and AGAGGGGAGGACCC primers was used to validate the presence of the WT allele (234 bp, Figure 1B). The PCRs were performed using 50 ng of genomic DNA extracted from murine tissues (e.g., ears, tails), in a final reaction volume of 25 µL, using the Taq 2x Master Mix Kit (New England Biolabs ® Inc., Ipswich, MA, USA; #M0270L). A 2.5% agarose gel was used to separate 428 bp and 414 bp PCR products during 18 h at 4 • C, 90 V; and 0.7% agarose gel was used to detect the 234 bp PCR product (75 min, room temperature, 124 V). Genomic DNA isolated from the Mri +/+ and Mri −/− cells as well as samples with no genomic DNA were used as the PCR controls ( Figure 1B).

Fluorescence-Activated Cell Sorting, Splenocyte, and Thymocyte Count
Fluorescence-activated cell sorting (FACS) analysis was performed as previously described [11,32]. Briefly, spleens and thymi were isolated from 2-month-old mice, and splenocytes and thymocytes were counted using Countess™ Automated Cell Counter (Invitrogen, Carlsbad, CA, United States); the cell suspension was spun down and diluted with PBS to obtain a final cell concentration of 2.5 × 10 7 /mL. The samples of 2.5 × 10 6 splenocytes or thymocytes were blocked for 15 min at room temperature with Mouse BD fragment crystallizable (Fc) Block™ (1:50 dilution) (BD Biosciences, Franklin Lakes, NJ, USA; #553142). The cells were then incubated with fluorochrome-conjugated antibodies (see below) and sorted.

Double Strand Break Sensitivity Assay
The DSBs sensitivity assay was performed as previously described [10,32,33]. Human nearly-haploid HAP1 cells were generated by the Horizon Discovery Group (Waterbeach, Cambridge, UK, #HZGHC005061c001 and #HZGHC005061c004) and are commercially available. HAP1 cells were cultured according to the manufacturer's instructions. Doxorubicin (Selleckchem, Houston, TX, USA; #S1208), bleomycin (Selleckchem; #S1214), and etoposide (Sigma-Aldrich, St. Louis, MS, USA; #E1383) were used to induce DSBs, and PrestoBlue™ Cell Viability Reagent (Thermo Fisher, Waltham, MA, USA; #A13262) was used to estimate cellular metabolism levels. Briefly, 2000 cells per well were seeded into 96-well plates in 100 µL of Iscove Modified Dulbecco Media (IMDM) medium (day 0). On day 1, 50 µL of the medium was replaced with 50 µL of fresh medium containing doxorubicin, bleomycin, or etoposide, when indicated. Each experimental condition was performed in triplicates. On day 4, 11 µL of 10× PrestoBlue reagent was added to the wells and incubated for 30 min at 37 • C. The cellular viability was estimated according to manufacturer's instructions, using the excitation/emission wavelengths set at 544/590 nm.

Neural Stem Progenitor Cell Proliferation and Self-Renewal Assays
Early passage NSPCs (P3-P10) were used throughout all of the NSPC experiments. A PrestoBlue™ Cell Viability Assay was used to investigate the neurosphere proliferation rates, following the manufacturer's instructions during each incubation on days 1 to 7. The capacity of neural stem cells to maintain their multipotency ex vivo was assessed by determining the number and two-dimensional size of neurospheres [34]. Single NSPCs were plated onto 6-well suspension plates in the proliferation medium on day 0. During days 8 and 10 in culture, images of the entire wells were captured using an EVOS microscope. Only areas between 50 and 1500 pixels were included in the analyses.

Generation of Mri −/− Mice
To investigate the impact of Mri on mouse development, we generated a mouse model with 14 bp frame-shift deletion in Mri exon 2 on a C57BL/6 background ( Figure 1A). Purified sgRNA and Cas9 RNA were introduced to fertilized oocytes, resulting in complete inactivation of the Mri gene. Mri status (WT, wild type, +/+; heterozygous, +/−; and null, −/−) was confirmed for every experiment by PCR screening ( Figure 1B). Mri +/+ , Mri +/− , and Mri −/− mice were born from Mri +/− parents at ratios close to 1:2:1 ( Figure 1C). Thirty-day old Mri −/− mice possessed an average body weight of 15.0 g, which was slightly lower, but not significantly different from the Mri +/+ controls, with a bodyweight of 17.5 g, on average ( Figure 1D). The lifespan of Mri −/− and Mri +/− mice was monitored for up to 12 months, according to the local regulations. During this time frame, both Mri −/− and Mri +/− mice were fertile and had no cancer incidence, similar to the Mri +/+ controls.

Class Switch Recombination to IgG1 Is Reduced in Mri −/− Mice
To determine whether Mri deficiency affects CSR, we isolated B cells from the spleens of Mri +/+ and Mri −/− mice and stimulated the cells with LPS and IL-4. After 96 h, we detected that average IgG1 levels were 33% in Mri −/− mice, which was significantly lower (p = 0.0031) than in the Mri +/+ controls (average 39%; Figure 2D). B lymphocytes isolated from Ung −/− mice were used as the negative control and possessed on average only 2% of IgG1 at the end of the experiment (96 h), which was lower than in Mri +/+ or Mri −/− mice (p < 0.0001).

Lack of Mri Results in the Reduced Proliferation Rate of Neuronal Stem Progenitor Cells
Previous studies have shown that single knockout of NHEJ DNA repair genes (e.g., Xrcc4, Lig4, Ku70) results in impaired nervous system development in mice [1,13,14]. To determine the impact of Mri on the developing nervous system, we used NSPC isolated from Mri +/+ and Mri −/− mice at postnatal day 1. We performed four independent experiments using two cell lines from two mice of each genotype. The average proliferation rate of Mri −/− neurospheres was approximately 35% lower than that in the WT controls, p = 0.0043 ( Figure 3B).

Normal Self-Renewal Capacity of Mri-Deficient Neuronal Stem Progenitor Cells
To analyze the capacity of NSPCs to maintain the features of stem cells throughout cell divisions and numerous propagations (self-renewal capacity), we counted the number of neurospheres formed in cell culture. In four independent experiments, we plated 10,000 single NSPCs and cultured for eight days. In total, we counted 5123 neurospheres that originated from Mri +/+ , and 4608 from Mri −/− mice. On average, there were 256 neurospheres in each of the 20 Mri +/+ samples analyzed, and 230 neurospheres in each of the 20 Mri −/− samples (p = 0.7254, n.s., Figure 3C). In addition, images of the neurospheres were collected and the surface was calculated using ImageJ software. Inactivation of Mri did not affect the average diameter of neurospheres, which was 461 px 2 on average in Mri +/+ and 427 px 2 in Mri −/− neurospheres, p = 0.4915 ( Figure 3D). We concluded that Mri is dispensable for the self-renewal capacity of NSPC.  Previous studies have shown that single knockout of NHEJ DNA repair genes (e.g., Xrcc4, Lig4, Ku70) results in impaired nervous system development in mice [1,13,14]. To determine the impact of Mri on the developing nervous system, we used NSPC isolated from Mri +/+ and Mri −/− mice at postnatal day 1. We performed four independent experiments using two cell lines from two mice of each genotype. The average proliferation rate of Mri −/− neurospheres was approximately 35% lower than that in the WT controls, p = 0.0043 ( Figure 3B).

Normal Self-Renewal Capacity of Mri-Deficient Neuronal Stem Progenitor Cells
To analyze the capacity of NSPCs to maintain the features of stem cells throughout cell divisions and numerous propagations (self-renewal capacity), we counted the number of neurospheres formed in cell culture. In four independent experiments, we plated 10,000 single NSPCs and cultured for eight days. In total, we counted 5123 neurospheres that originated from Mri +/+ , and 4608 from Mri −/− mice. On average, there were 256 neurospheres in each of the 20 Mri +/+ samples analyzed, and 230

Human HAP1 Cells Lacking Mri Possess Normal Levels of Sensitivity to DNA Double-Strand Breaks
Deficiency for XRCC4, LIG4, XLF, or DNA-PKcs results in hypersensitivity to DSBs in human HAP1 cells [10,32,33]. To determine the effect of Mri on DSB sensitivity, we obtained two independent clones of MRI-deficient HAP1 cells as well as WT and XRCC4-deficient controls. We exposed the HAP1 cells to DSB-inducing agents bleomycin (0 to 0.4 µM), doxorubicin (0 to 4 nM), and etoposide (0 to 160 nM), and evaluated the cell viability four days later (Figure 4). We observed no hypersensitivity of HAP1 cells lacking Mri when compared to WT controls. However, cells lacking XRCC4 were hypersensitive to all indicated compounds, bleomycin, doxorubicin, and etoposide (p < 0.0001, Figure 4). HAP1 cells to DSB-inducing agents bleomycin (0 to 0.4 μM), doxorubicin (0 to 4 nM), and etoposide (0 to 160 nM), and evaluated the cell viability four days later (Figure 4). We observed no hypersensitivity of HAP1 cells lacking Mri when compared to WT controls. However, cells lacking XRCC4 were hypersensitive to all indicated compounds, bleomycin, doxorubicin, and etoposide (p < 0.0001, Figure 4).

Discussion
We have generated a new knockout mouse model with 14 bp deletion in exon 2 of the Mri gene, Mri −/− . While we were characterizing our mouse model, another group reported an independentlygenerated Mri-deficient mouse [20], which possessed a similar phenotype. Thus, our observations are confirmatory to the findings observed by the Sleckman group [20].
The CSR to IgG1 was reduced in primary B cells isolated from Mri −/− mice when compared to WT controls (Figure 2), which suggests that Mri is involved in specific DNA repair-mediated event. Furthermore, we isolated neuronal stem progenitor cells from Mri −/− brains and found that these cells proliferate slower when compared to Mri +/+ controls. Reduced proliferation rates of Mri-deficient neuronal stem progenitor cells could be explained, as one option, by lower expression or lack of factors functionally redundant with Mri in these cell types. Future studies would be required to address this option. Moreover, future studies may address questions such as neurological defects and

Discussion
We have generated a new knockout mouse model with 14 bp deletion in exon 2 of the Mri gene, Mri −/− . While we were characterizing our mouse model, another group reported an independently-generated Mri-deficient mouse [20], which possessed a similar phenotype. Thus, our observations are confirmatory to the findings observed by the Sleckman group [20].
The CSR to IgG1 was reduced in primary B cells isolated from Mri −/− mice when compared to WT controls (Figure 2), which suggests that Mri is involved in specific DNA repair-mediated event. Furthermore, we isolated neuronal stem progenitor cells from Mri −/− brains and found that these cells proliferate slower when compared to Mri +/+ controls. Reduced proliferation rates of Mri-deficient neuronal stem progenitor cells could be explained, as one option, by lower expression or lack of factors functionally redundant with Mri in these cell types. Future studies would be required to address this option. Moreover, future studies may address questions such as neurological defects and cognitive functions in mice lacking Mri as well as whether the Mri-deficient mice are prone to infections.
In addition, we found that human nearly haploid HAP1 cell lines lacking Mri possessed no proliferation defect or hypersensitivity to DNA damaging agents such as etoposide, doxorubicin, and bleomycin ( Figure 4). Previously, it was reported that murine embryonic fibroblasts (MEF) lacking Mri were hypersensitive to ionizing radiation when compared to WT controls, although the sensitivity was less obvious than in XLF-deficient MEFs [20]. The discrepancy between our and previously published data could be due to the usage of different cell types, the difference between species as well as distinct ways to induce DNA damages (e.g., chemicals vs. irradiation). Further studies involving various cell type models originated from different species, and using various DNA damaging strategies would deepen our understanding of the specific functions of Mri in DNA repair in mammalian cells. Overall, we concluded that the lack of Mri has a rather minor effect on murine and human cells.
Combined inactivation of Mri and Xlf [20], however, revealed an important role of Mri in mouse development, which was overlooked due to its functional redundancy with XLF. Synthetic lethality between Mri and Xlf complicated studies of genetic interaction between these factors in vivo. There are several potential ways to overcome this challenge. One option is to use conditional knockouts of Xlf or Mri genes. Moreover, there might be another genetic-based approach. Inactivation of one or two alleles of Trp53 partially rescued synthetic lethality between Xlf and Dna-pkcs [10,18,19] and Xlf and Paxx [10]. In addition, deficiency or haploinsufficiency for Trp53 rescued embryonic lethality of Lig4 −/− and Xrcc4 −/− mice [13,14]. Inactivation of the Atm gene rescued embryonic lethality of Lig4 −/− mice [41]. Finally, inactivation of both alleles of Ku80 rescued embryonic lethality of Lig4 −/− mice [17], and inactivation of Ku70 rescued synthetic lethality between Xlf and Dna-pkcs [19]. Based on these data, one can speculate that inactivation of Trp53, Atm, Ku70, or Ku80 will rescue synthetic lethality between Xlf and Mri. Moreover, given the critical roles of Ku70 and Ku80 in the initiation of classical NHEJ, one could propose that mice lacking all known NHEJ factors (e.g., Ku70 −/− Ku80 −/− Dna-pkcs −/− Artemis −/− Xlf −/− Paxx −/− Mri −/− Xrcc4 −/− Lig4 −/− ) will be viable, indistinguishable from Ku-deficient mice, and serve as a suitable in vivo model to investigate alternative end-joining, A-EJ.

Conclusions
A new Mri-deficient mouse model was generated. Mri-deficient mice possessed normal body size and number of B and T lymphocytes; however, Mri is required for an efficient class switch recombination process in mature B cells. Mri −/− neurospheres showed a reduced proliferation rate, but similar self-renewal capacity when compared to the Mri +/+ controls.  Acknowledgments: We gratefully acknowledge support by the Molecular Imaging Core Facility (CMIC) and Comparative Medicine Core Facility (CoMed) at the Faculty of Medicine and Health Sciences, NTNU (Trondheim).

Conflicts of Interest:
The authors declare no conflicts of interest.