To obtain Mrad1-disrupted ES cells, we used a promoterless gene targeting strategy to delete the gene 20 . Mrad1 was disrupted by homologous recombination in ES cells using the targeting vector illustrated in Fig. 1A. This targeting construct contains a selectable neo gene. This gene, which lacked the start codon (ATG), was inserted between the third exon and third intron of Mrad1, deleting parts of the third exon and third intron. Homologous recombination of the targeting construct into mouse genomic Mrad1 was predicted to generate mutant cells that express a fusion protein containing the 77AA of the Mrad1 N-terminus and a full length neo protein, but lacking the rest of the Mrad1 protein (part of the third exon and the complete 4, 5 and 6 exons). We then targeted a 129Sv/Ev ES cell line obtained from Dr. Victor Lin's laboratory (Columbia University) and Mrad1 ^+/- ES cell clones were obtained after transfection and challenge with 300 μg/ml G418. Multiple heterozygous Mrad1-deleted clones were identified by Southern blot analysis (Fig. 1B). A few were selected for PCR genotyping and confirmed to be Mrad1 heterozygous (Fig. 1C).

Mrad1 ^+/- ES cells were used to generate Mrad1 targeted mice (see Methods for details). Genotypes of the mice were analyzed by Southern blot hybridization (Fig. 1D) and PCR (Fig. 1E), and indicated that Mrad1 ^+/- animals were successfully generated. Mating between Mrad1 ^+/- mice only produced Mrad1 ^+/+ and Mrad1 ^+/- offspring, providing evidence that Mrad1 ^-/- causes embryonic lethality. Mating between Mrad1 ^+/- and Mrad1 ^+/+ mice generated almost equal numbers of Mrad1 ^+/- and Mrad1 ^+/+ pups (111:109), suggesting no effect of Mrad1 heterozygous mutation on embryonic survival. Mrad1 ^+/- and wild type mice were maintained and monitored up to 1.5 years of age, and heterozygotes had no overt defects compared to wild type mice.

Mouse embryos at 7.5 dpc with either the Mhus1 ^-/- or Mrad9 ^-/- genotype are smaller than wild-type littermates, look morphologically abnormal, and eventually die during embryonic development 21 22 . To determine the embryonic morphology and the stage of embryonic lethality of Mrad1 homozygous mutants, embryos resulting from Mrad1 ^+/- X Mrad1 ^+/- crosses were retrieved at different stages of gestation. We examined embryos at 6.5d, 7.5d, 8.5d, 10.5d and 11.5d. Table 1 shows the numbers and genotypes of the embryos obtained. PCR was used to genotype DNA isolated from yolk sacs (Fig. 2A). Gross morphology of embryos at the different stages of development is presented in Fig. 2B. Mrad1 ^-/-, Mrad1 ^+/+ and Mrad1 ^+/- embryos were identified at all stages analyzed. At E6.5, 23 of the 27 embryos are close in size and morphology except 2 Mrad1 ^-/-, 2 Mrad1 ^+/+ and 1 Mrad1 ^+/-embryos were relatively smaller but with normal morphology. Overall, deletion of Mrad1 does not impact on the size and gross morphology of the embryos by E6.5. At E7.5, however, there were many Mrad1 ^-/- embryos significantly smaller than those with the Mrad1 ^+/- or Mrad1 ^+/+ genotype. We retrieved three liters at E7.5, and the widths of the embryos were measured. The average width among the embryos with the same genotype was calculated, and then normalized to the average width of wild type embryos. Afterwards, the average widths of the embryos from the three liters were subjected to statistical analysis (T-test), and we found that the width of Mrad1 ^-/- embryos was statistically significantly smaller than that of Mrad1 ^+/+ embryos (P = 0.016); the widths of Mrad1 ^+/- and Mrad1 ^+/+ embryos were statistically equal (P = 0.11). At E8.5, Mrad1 ^+/- and Mrad1 ^+/+ embryos were further developed while Mrad1 ^-/- embryos still looked like the Mrad1 ^-/- embryos at E7.5 (data not shown). At E10.5 only amorphous material was found as Mrad1 ^-/- embryos (Fig. 2B), and at E11.5 Mrad1 ^-/- embryos were completely resorbed. Taken together, as Mrad9 and Mhus1, Mrad1 is critical for embryonic development at or before E7.5.

To determine if Mrad1 is important for tumorigenesis, the skin of mice with Mrad1 ^+/+ and Mrad1 ^+/- genotypes was treated with DMBA plus TPA. A total of 38 mice were divided into 2 groups, each with the Mrad1 ^+/+ or Mrad1 ^+/- genotype. The 2 groups were from 19 litters, each litter consisting of 2 mice with Mrad1 ^+/+ and Mrad1 ^+/- genotypes, and identical sex, either female (14 liters) or male (5 liters). Under this setting, Log-Rank Test in the Kaplan-Meier PL method can be used for statistical analysis on the significance of differences of tumor development between two groups of animals 23 . DMBA was used to initiate skin tumorigenesis, and TPA was used to promote skin tumor growth. Mice were first painted with 15 μg DMBA in 100 μl acetone. One week after DMBA treatment, mice were painted with 2 μg TPA in 100 μl acetone twice a week (Monday and Thursday) for 17 weeks. Only skin tumors larger than 1 mm were recorded. After 7 weeks of TPA treatment, Mrad1 ^+/- mice began to develop skin tumors in the treated area. It took 13 weeks of TPA treatment for the first Mrad1 ^+/+ mouse to develop skin tumors (Fig. 3A and 3B). At the 17^th week of TPA treatment, when the experiment ended, 14 Mrad1 ^+/- mice had skin tumors while only 7 Mrad1 ^+/+ mice, treated in the same fashion, developed skin tumors (Fig. 3B). Kaplan-Meier PL method 23 was used for comparison of the relative risk of tumor development induced by DMBA plus TPA between mice with different genotypes. The rate of tumor development in Mrad1 ^+/- mice was significantly higher than in Mrad1 ^+/+ mice (P = 0.003, Log-Rank Test). Additionally, the average number of tumors in each of Mrad1 ^+/- mice that developed tumors was significantly higher (P = 0.010) than that in tumor-bearing Mrad1 ^+/+ mice (Fig. 3C). We also measured the size of tumors on mouse skin after 17 weeks of TPA treatment. Tumors larger than 6 mm in diameter only appeared in Mrad1 ^+/- mouse skin (Fig. 3D; data not shown). We made skew analysis on the tumor size distribution and found that G1 values were 0.40528 and 1.84488 for the wild type and Mrad1 heterozygous mice, respectively, suggesting that Mrad1 heterozygous mice bore significantly larger tumors. Staining skin specimens with H&E or anti-keratin 14 indicated that the tumors were derived from keratinocytes and possessed characteristics of papillomas (Fig. 3E and 3F). All these data suggest that Mrad1 plays an important role in the prevention of skin papilloma development.

DNA repair and cell cycle control are two processes important for maintaining genomic integrity and preventing carcinogenesis 1 2 , thus we investigated these processes in keratinocytes bearing an Mrad1 deletion. First, we examined whether there is a significant change in the proliferation of the isolated keratinocytes resulting from the deletion of an allele of Mrad1. Cells were grown in defined Keratinocyte-SFM medium. Five hundred thousand cells, either Mrad1 ^+/+ or Mrad1 ^+/-, were seeded into each well of 6-well plates. The number of cells in each population dropped to nearly 40% of the seeded, original number (2 × 10⁵ per well) after the first 2 days of incubation. Subsequently, the total number of keratinocytes increased until day 12 when the experiment was terminated (Fig. 4A). There is a slight but significant difference in cell proliferation between Mrad1 ^+/+ and Mrad1 ^+/- keratinocytes (at Day 12, P = 0.03). This result indicates that Mrad1 deletion affects the normal proliferation of cells although the influence is not dramatic.

The isolated keratinocytes were cultured in an incubator for 3 days before analysis of DNA damage. The neutral comet assay was used to detect DNA double-strand breaks (DSBs). There were significantly more DSBs in the incubated Mrad1 ^+/- keratinocytes than inMrad1 ^+/+ cells (P = 0.004) (Fig. 4B). The phosphorylation of histone H2AX (γ-H2AX), a marker for the presence of DSBs, was also used to evaluate DSBs in Mrad1 ^+/-and Mrad1 ^+/+ keratinocytes in this study. Significantly more Mrad1 ^+/- keratinocytes contained γ-H2AX-positive foci than Mrad1 ^+/+ cells, while significantly less Mrad1 ^+/- keratinocytes contained no γ-H2AX foci compared to Mrad1 ^+/+ cells (Fig. 4C). Therefore, Mrad1 is critical for maintaining genomic integrity. We also studied the effects of Mrad1 deletion on apoptosis. The percentage of apoptotic cells among Mrad1 ^+/- keratinocytes is slightly higher than within theMrad1 ^+/+ cell population (Fig. 4D), but the difference is not statistically significant. We also examined the apoptosis levels induced by DMBA (0.15 μg/ml, 24 h), and found that the apoptosis levels in both Mrad1 ^+/- and Mrad1 ^+/+ keratinocytes were enhanced, but there were no statistically significant differences between the DMBA-induced apoptosis levels of Mrad1 ^+/- and Mrad1 ^+/+ cells, and between the mock-treated and DMBA-treated cells with either genotypes (Fig. 4D). The above results suggest that Mrad1 deletion in keratinocytes does not alter DMBA-induced apoptotic response in this experimental setting.

Flow cytometric analyses of PI-stained Mrad1 keratinocytes indicated that Mrad1 ^+/- cells contained a slightly smaller G1 subpopulation and a slightly larger S subpopulation in the cell cycle than Mrad1 ^+/+ cells (Fig. 4E). After incubation for 24 h in medium containing 0.15 μg/ml DMBA, more cells with either Mrad1 phenotype were accumulated in G1 phase (Fig. 4E), indicating a functional G1 phase checkpoint control in both cell types. Measurement of BrdUrd uptake by replicative S phase cells in combination with DNA content via PI staining in individual cells can reveal more information on cell cycle distribution. Therefore, we investigated cell cycle profiles in more detail by pulse labeling with BrdUrd and staining cells after 4 days of incubation. There is again no major difference in cell cycle distribution between Mrad1 ^+/+ and Mrad1 ^+/- keratinocytes. The number of BrdUrd-positive Mrad1 ^+/+ cells in S phase is nearly the same as the number of Mrad1 ^+/- cells (Fig 4F). Based on the above data, we conclude that Mrad1 deletion has only a slight effect on the distribution of cell cycle phase during in vitro incubation in both mock-treated and DMBA-treated conditions. This result is consistent with the finding that Mrad1 deletion only leads to a slight delay in proliferation of mouse skin keratinocytes.

In a previous study, we reported that deletion of either one or two alleles of Mrad9 induced expression of cell cycle checkpoint genes and dramatically reduced cell proliferation of keratinocytes in culture. The results herein show that Mrad1 ^+/- keratinocytes have only a slightly lower proliferation rate than the Mrad1 ^+/+ controlcells (Fig. 4A). We therefore examined whether expression levels of p53 and p21 in Mrad1 ^+/- keratinocytes are higher than in the related Mrad1 ^+/+ cells. Indeed, heterozygous deletion of Mrad1 did not by itself alter expression of p21or p53 in skin keratinocytes in culture (Fig. 5A). These results can explain why Mrad1 ^+/- keratinocytes do not have significantly higher levels of apoptosis and only a slightly lower proliferation rate than those of the wild type keratinocyte control. We further investigated p21 and p53 expression levels in keratinocytes treated with DMBA (0.15 μg/ml, 24 h) and found that the treatment induced expression most dramatically of p53 but also somewhat p21 in both Mrad1 ^+/+ and Mrad1 ^+/- keratinocytes (Fig. 5A). The expression levels of Mrad9 and Mhus1 in keratinocytes withMrad1 ^+/+ and Mrad1 ^+/- genotypes were also examined by real-time PCR. The results show that Mrad9 and Mhus1 expression levels are not altered by one allele Mrad1 deletion (Fig. 5B), suggesting that there is no regulatory effect of Mrad1 on either Mrad9 or Mhus1 expression in mouse keratinocytes. Interestingly, Mrad1 expression levels in Mrad1 ^+/- mouse skin tumors were higher than those in Mrad1+/⁺mouse skin tumors, but not statistically significantly (Fig. 5C).

A targeting vector was made to produce a deletion in Mrad1. We used the promoterless selection strategy to obtain a high efficiency of homologous gene targeting 20 . The targeting vector was constructed in three steps starting with pBluescript SK(+) vector. First, the 5'end fragment, a 1523 bp Mrad1 sequence between exon 2 and exon 3, was generated by PCR from 129 SvEv mouse genomic DNA with primers:

5'-GTCTCAGGTTTTCACACATCTTCC-3' and 5'-CTACGCGTCGACCTTCCTGAATGACAAATTCCTG-3' (Fig. 1A). The PCR product was cut with Kpn1 and Sal1, and subcloned into pBluescript SK(+). Second, the neo gene was amplified from pRc/CMV2 vector without the promoter and ATG using primers:

5'-CTACGCGTCGACATTGAACAAGATGGATTGCACGC-3' and 5'-AAGGAAAAAAGCGGCCGCAGACATGATAAGATACATTGATGAG-3'. Then, the PCR product was cut with Sal1 and Not1, and inserted in frame with Mrad1 into the plasmid constructed in the first step. Third, the 3'end fragment, 5591 bp long between intron 3 and intron 6, was generated by PCR from 129 SvEv mouse genomic DNA with primers:

5'-AAGGAAAAAAGCGGCCGCCTACTACAACTACTGCTACTAC-3' and 5'-TCCCCGCGGCACAGGACAGTACAGTAAGTCG-3'. The product was cut with SalI and SacII, and inserted into the vector constructed in the second step. This yielded the final targeting construct with the selectable neo gene, which was linearized with Kpn1 prior to transfection into ES cells.

ES cells derived from 129 SvEv mice were cultured by established methods 38 . ES cells used to make gene-targeted mice were grown on feeder cells, electroporated with targeting vector linearized by Kpn1, and then grown in the presence of G418 at 300 μg/ml. The G418-resistant clones were picked, expanded and subjected to Southern blot hybridization and PCR analyses to identify Mrad1 ^+/- targeted clones. Positive clones were injected into C57BL/6 blastocysts. Chimeric offspring were born and mated to C57BL/6 mice to confirm successful germ line transmission of the targeted Mrad1 allele. Genomic DNA from tails was analyzed by Southern blot hybridization and PCR analyses. Mrad1 heterozygous mutant mice were intercrossed and maintained.

For Southern blotting, genomic DNA was isolated from ES cells and tails of mice using published methods 22 . DNA was digested with HindIII, separated on a 0.7% agarose gel, then transferred to a nylon membrane, and hybridized to a ³²P-labeled probe, which was generated by PCR using primers:

5'-GTGGCCTAGGTGGTTGCGTATCTGAAC-3' and 5'-GTCGGCTCCGAGAAGAAGGATGCTCC-3' in conjunction with mouse genomic DNA as template.

To genotype ES cells and mice by PCR, the reaction was performed using genomic DNA templates and the following primer pair:

5'-GTCTCAGGTTTTCACACATCTTCC-3' and 5'-GCTTATATTCTAGAAACCTTCCTGTATG-3'. PCR conditions were 95°C for 5 min, followed by 35 cycles of 95°C for 30 s, 59°C for 30 s, and 72°C for 3 min, with a final extension at 72°C for 10 min.

Mouse embryos were obtained at several stages of gestation, including E6.5, E7.5, E8.5, E10.5, and E11.5. All dissections were performed in 1 × PBS. Whole embryos were rinsed with 1 × PBS. Pictures of whole embryos were taken while viewed by a Wild Heerbrugg dissecting microscope.

Full-thickness skin removed from newborn (1-2 days old) mice was treated with 0.25% trypsin overnight at 4°C. The epidermis was peeled off from the dermis and minced into pieces smaller than 1 mm. They were placed into a sterile flask, then dispersed by stirring into single cells for 30-60 min, then suspended in Keratinocyte-SFM medium with supplements (Invitrogen). Cells were first incubated in dishes coated with collagen type I at 34°C in 5% CO₂ for 12 h to allow cells to attach to the bottom. Afterwards, unattached cells were removed by washing with PBS. Attached cells were further cultured in fresh medium, which was replaced every 2 days.

For preparing protein from epidermis, full-thickness skin removed from newborn mice was treated with 0.25% trypsin overnight at 4°C. The epidermis was peeled off from the dermis and dispersed in lysis buffer. To prepare cell lysate, keratinocytes incubated for 3 days were either left untreated or treated for 24 h with 0.15 μg/ml DMBA (Sigma). Then, the cell lysate was prepared in 1× SDS-sample buffer, to a final concentration of 10⁴ cells/μL. Fifty μg of protein were resolved on a 10% SDS-PAGE gel, and proteins were transferred to a polyvinylidene difluoride membrane. The membrane was probed consecutively with primary and peroxidase-conjugated secondary antibodies. Primary and secondary antibodies used in this study are mouse anti-GAPDH (KangChen, China), mouse anti-p21 (Santa Cruz), mouse anti-p53 (Oncogene), peroxidase-conjugated anti-rabbit IgG (A9169, Sigma) and peroxidase-conjugated anti-mouse IgG (A9044, Sigma).

Mice (7-8 weeks old) were shaved on their backs 2 days before tumor induction. To induce tumors, the shaved dorsal skin of mice was treated topically with 15 μg of DMBA (Sigma) in 100 μL acetone once. After 1 week, each animal received subsequent topical treatments of 2 μg of TPA (Sigma) in 100 μL acetone twice weekly for 17 weeks. Treated areas were examined weekly for the presence of tumors, which were scored positive if they reached at least 1 mm in diameter.

Dorsal skin samples and tumors were fixed in 4% paraformaldehyde at 4°C overnight, embedded in paraffin, and sectioned as 8-μm slices. The sectioned tissues on slides were stained with H&E 39 40 . Immunohistochemical staining was carried out using a kit (ImmunoCruz Staining Systems, Beijing Zhongshan Golden Bridge Biotechnology). The endogenous peroxidase activity in the specimens was blocked by treatment with 0.3% H₂O₂ and samples were then rinsed with PBS. The specimens were probed consecutively with primary antibodies against Keratin 14 (BAbCo), secondary antibody biotin-conjugated goat anti-rabbit IgG, and horseradish peroxidase-streptavidin complex, and then visualized by diaminobenzidine. Afterwards, sections were counterstained with hematoxylin.

Keratinocytes were isolated as described above and seeded into 6-well plates (5 × 10⁵ cells per well) containing Keratinocyte-SFM medium with supplements. Cell numbers were determined every 2 days.

The cell cycle profiles of cells in different phases were determined using previously established methods 41 . Briefely, 1 × 10⁷ keratinocytes were plated in each 10-cm dish. After incubation for 3 days the cells were mock-treated or treated with 0.15 μg/ml DMBA (Sigma) for 24 h, then processed and stained with propidium iodide (PI), and analyzed by a FACSCalibur cytometer (Becton Dickinson). To assess DNA synthesis, 10 μM BrdUrd was added to medium and cells were pulse labeled for 40 min. Cells were then processed, probed with FITC-conjugated anti-BrdUrd antibody (Becton Dickinson) and stained with PI. Flow cytometric analyses were performed on a FACSCalibur.

Keratinocytes incubated for 4 days were mock-treated or treated for 24 h with 0.15 μg/ml DMBA, trypsinized for 10 min using 0.1% trypsin at 37°C, washed twice with cold PBS, then resuspended in 1× binding buffer [10 mmol/L HEPES (pH 7.4), 140 mmol/L NaCl, and 2.5 mmol/L CaCl₂] at a concentration of 1 × 10⁶ cells/mL. Then cells were stained with Annexin V-FITC (Jingmei Biotech) and PI for 15 min at room temperature before flow cytometric analysis.

Keratinocytes were cultured in standard medium for 4 days. The comet assay was carried out according to the manufacturer's instructions (Trevigen). Briefly, cells at a concentration of 1 × 10⁵/mL were mixed gently with premelted low-temperature-melting agarose at a volume ratio of 1 to 10 (v/v) and spread on glass slides. The slides were then submerged in precooled neutral lysis buffer at 4°C for 30 min. After rinsing, the slides were equilibrated in Tris-borate EDTA solution, electrophoresed at 1.0 V/cm for 20 min, and then stained with PI. Fluorescence images for at least 50 nuclei were captured using a Nikon microscope and analyzed by CASP-1.2.2 software (University of Wroclaw) for tail moment (i.e., the geometric mean of fluorescence on the tail from the nucleus).

All statistical analyses were performed using statistical software package SPSS Version 10.0. The Kaplan-Meier PL method 23 was used for comparison of the relative risks of tumor development induced by DMBA-TPA between the mice with the two different Mrad1 genotypes. We designed the tumor development experiment to meet a set of conditions so the Log-Rank Test in the Kaplan-Meier PL method could be used. The Student's t test was performed to determine statistical significance of the differences for the comet assay. Wilcoxon rank-sum test was used to compare the difference in tumor numbers between the two groups of mice having different Mrad1 genotypes. In all the above analyses, a P value of < 0.05 was considered statistically significant. Skewness was used to compare the difference of tumor size distributions between Mrad1 wild type and heterozygous mice.

Keratinocytes grown on coverslips were fixed with 4% paraformaldehyde in PBS for 15 min at room temperature, washed in PBS twice, incubated in PBS containing 0.5% Triton-X100 for 15 min and in PBS containing 5% BSA and 0.1% Triton-X100 for 1 hr, and washed in PBS once, followed by incubation with anti-phospho-H2AX (Upstate) primary antibody (1:100 dilution) in PBS containing 5% BSA and 0.1% Triton-X100 for 1 hr at room temperature. Afterwards, the coverslips were washed two times for 5 min each in PBS and incubated with Texas Red -conjugated anti-mouse antibody (1:100 dilution in PBS containing 5% BSA and 0.1% Triton-X100) for 1 hr at room temperature. Finally, the coverslips were counterstained with DAPI (10 ng/ml). The images were captured using a fluorescence microscope.

Total RNA was isolated from mouse tumors (3 wild type and 3 Mrad1 heterozygous tumors) or keratinocytes cultured for 4 days using the RNeasy Mini kit, as described by the manufacturer (QIAGEN). Two μg total RNA were reverse transcribed in a 20 μL reaction volume to form cDNA using the SuperScript First-Strand Synthesis System for RT-PCR (Invitrogen). Real-time PCR was performed using the StepOnePlus system(ABI) with SYBR Green I (Takara) to label amplified DNA. A standard curve method of quantification was used to calculate the expression of target genes relative to the housekeeping gene β-actin. Experiments were performed thrice. The following primer pairs were used for the PCR reactions: Mrad9, 5'-GCCTCTTACTATCCACTTCG-3' and 5'-AGCCCTCATTGCCTCC-3'; Mrad1, 5'-GCCCTATTTCAGGTTGT-3' and 5'-TGCCCATCTTCATTTCT-3'; Mhus1, 5'-TCCCTGTCTTACCGTGTC-3' and 5'-CTCCCTTTAGGTTTGCTT-3'; β-actin, 5'-GTAAAGACCTCTATGCCAACA-3' and 5'-GGACTCATCGTACTCCTGCT-3'. We used the following PCR procedure: 94°C for 3 min, then 40 cycles of 94°C for 15 s, 55°C for 20 s, 72°C for 19 s, and a final extension at 72°C for 3 min.