FOXO3a Mediates Homologous Recombination Repair (HRR) via Transcriptional Activation of MRE11, BRCA1, BRIP1, and RAD50

To test whether homologous recombination repair (HRR) depends on FOXO3a, a cellular aging model of human dermal fibroblast (HDF) and tet-on flag-h-FOXO3a transgenic mice were studied. HDF cells transfected with over-expression of wt-h-FOXO3a increased the protein levels of MRE11, BRCA1, BRIP1, and RAD50, while knock-down with siFOXO3a decreased them. The protein levels of MRE11, BRCA1, BRIP1, RAD50, and RAD51 decreased during cellular aging. Chromatin immunoprecipitation (ChIP) assay was performed on FOXO3a binding accessibility to FOXO consensus sites in human MRE11, BRCA1, BRIP1, and RAD50 promoters; the results showed FOXO3a binding decreased during cellular aging. When the tet-on flag-h-FOXO3a mice were administered doxycycline orally, the protein and mRNA levels of flag-h-FOXO3a, MRE11, BRCA1, BRIP1, and RAD50 increased in a doxycycline-dose-dependent manner. In vitro HRR assays were performed by transfection with an HR vector and I-SceI vector. The mRNA levels of the recombined GFP increased after doxycycline treatment in MEF but not in wt-MEF, and increased in young HDF comparing to old HDF, indicating that FOXO3a activates HRR. Overall, these results demonstrate that MRE11, BRCA1, BRIP1, and RAD50 are transcriptional target genes for FOXO3a, and HRR activity is increased via transcriptional activation of MRE11, BRCA1, BRIP1, and RAD50 by FOXO3a.


Introduction
One of the most deleterious types of DNA damage is the double-strand break (DSB), which can lead to growth arrest and cell death if not repaired. DSB is repaired by nonhomologous end joining (NHEJ) and homologous recombination repair (HRR). NHEJ repairs double-strand breaks by direct ligation of DNA break-ends throughout the entire cell cycle [1], while HRR repairs double-strand breaks by homologous recombination mostly during the S and G2 phases [2,3]. NHEJ is an abundant repair activity for DSB in mammalian cells, but it is error-prone as end processing can lead to small deletions and translocations [4]. By comparison, HRR takes advantage of a DNA template with high sequence homology and provides high-fidelity repair for complex DNA damage including , and β-actin were measured by Western blotting in HDF cells transfected with increasing amounts of wt-h-FOXO3a plasmid. HDF cells of PD24 (young) were transfected with 0, 300, 600, and 1200 ng of wt-h-FOXO3a using Lipofectamine 3000. The HDF cells were then cultured for 24 h. Transfection efficiency was determined by pCMV-beta-Gal-transfected HDF. Relative band intensities of MRE11, BRCA1, BRIP1, RAD50, RAD51, NBS1, BARD1, PALB2, and FOXO3a versus β-actin were measured by densitometry and the data were plotted as histograms. (B) The protein levels of MRE11, BRCA1, BRIP1, and RAD50 were measured in siFOXO3a-transfected HDF PD24 (young) cells by Western blotting. Standard deviations as error bars were obtained from three different experiments. Statistical significance is indicated as * p < 0.05, ** p < 0.01, and *** p < 0.001.

Promoter Activities of Human MRE11, BRCA1, BRIP1, and RAD50 were FOXO3a-Dependent and ChIP Assay on FOXO Consensus Sites of Human MRE11, BRCA1, BRIP1, and RAD50 Promoters Showed Cellular-Aging-Dependent Decrease in FOXO3a Binding
To test whether promoter activities of human MRE11, BRCA1, BRIP1, and RAD50 are FOXO3a-dependent, HDF cells (young, PD24) were transfected with promoter-luciferase constructs and increasing amounts of wt-h-FOXO3a, and promoter activities of these genes were measured. Promoter activities increased FOXO3a-dependently ( Figure  3B,E,H,K), indicating that the transcription of these genes are FOXO3a-dependent. To test the hypothesis, FOXO3a binding to the endogenous promoters of the FOXO3a-de-pendentHRR genes was measured by ChIP assay. FOXO3a binding to endogenous promoters decreased during cellular aging of HDF; ChIP assays with anti-FOXO3a antibody and PCR primers were performed. The results showed that FOXO3a binding on FOXO consensus sites in human endogenous MRE11 (FOXO site #1) ( Figure (Figure 3). These results showed that the protein levels of MRE11, BRCA1, BRIP1, and RAD50 decreased during cellular aging of HDF due to decreased FOXO3a binding on FOXO consensus sites of human endogenous Figure 2. The protein levels of MRE11, BRCA1, BRIP1, RAD50, RAD51, BRCA2, FOXO3a, and β-actin were measured by Western blotting. Cell extracts were prepared from HDF cells at PD24 ("young"), PD30, PD36 ("middle"), PD46, and PD56 ("old"). Relative band intensities of MRE11, BRCA1, BRIP1, RAD50, RAD51, BRCA2, and FOXO3a versus β-actin were measured by densitometry and the data were plotted as histograms. Standard deviations as error bars were obtained from three different experiments. Statistical significance is indicated as * p < 0.05, ** p < 0.01, and *** p < 0.001. To test whether promoter activities of human MRE11, BRCA1, BRIP1, and RAD50 are FOXO3a-dependent, HDF cells (young, PD24) were transfected with promoter-luciferase constructs and increasing amounts of wt-h-FOXO3a, and promoter activities of these genes were measured. Promoter activities increased FOXO3a-dependently ( Figure 3B,E,H,K), indicating that the transcription of these genes are FOXO3a-dependent. To test the hypothesis, FOXO3a binding to the endogenous promoters of the FOXO3a-dependentHRR genes was measured by ChIP assay. FOXO3a binding to endogenous promoters decreased during cellular aging of HDF; ChIP assays with anti-FOXO3a antibody and PCR primers were performed. The results showed that FOXO3a binding on FOXO consensus sites in human endogenous MRE11 (FOXO site #1) ( Figure 3A-C), BRCA1 (FOXO site #1, #2, #3) ( Figure 3D-F), BRIP1 (FOXO site #1, #2, #3, #4) ( Figure 3G-I), and RAD50 promoters (FOXO site #1, #2) ( Figure 3J-L) gradually decreased during cellular aging, namely at PD26 (young) and PD44 (old) (Figure 3). These results showed that the protein levels of MRE11, BRCA1, BRIP1, and RAD50 decreased during cellular aging of HDF due to decreased FOXO3a binding on FOXO consensus sites of human endogenous MRE11, BRCA1, BRIP1, and RAD50 promoters. The decreased FOXO3a binding is because of the decreased protein levels of FOXO3a during cellular aging of HDF, as shown in Figure 2.   Cell lysates were prepared from HDF cells of PD26 ("young") and PD44 ("old"). Cell lysates were sonicated on ice to obtain sheared average DNA fragments of 300 bp to 1000 bp. Immunoprecipitation was carried out as described in the Materials and Methods section. Isolated genomic DNA from immunoprecipitation was used in the ChIP-PCR reaction. ChIP-PCR amplification of FOXO3abinding consensus sites in (C) the human MRE11 promoter, (F) the human BRCA1 promoter, (I) the human BRIP1 promoter, and (L) the human RAD50 promoter was performed as described in the Materials and Methods section. The PCR products were separated by electrophoresis in 1.5% agarose gel. Standard deviations as error bars were obtained from three different experiments. Statistical significance is indicated as * p < 0.05, ** p < 0.01, and *** p < 0.001. Tet-on h-FOXO3a transgenic mice with inducible wt-h-FOXO3a were prepared from crossing of flag-h-FOXO3a-tetO transgenic mice with tet-R transgenic mice and subsequent selection of double-positive mice offspring. FOXO3a target genes to be studied were induced in the tet-on flag-h-FOXO3a transgenic mice by oral administration of doxycycline in 1% sucrose. The mRNA levels of flag-h-FOXO3a, MRE11, BRCA1, BRIP1, and RAD50 increased in a doxycycline-dose-dependent manner in mice tail-tip samples, and those for mouse FOXO3a and β-actin were not changed ( Figure 4A). The protein levels of flag-h-FOXO3a, MRE11, BRCA1, BRIP1, and RAD50 increased in a doxycycline-dose-dependent manner in mice tail-tips, and those for mouse FOXO3a and β-actin were not changed ( Figure 4B). The result indicates that the mRNA and protein levels of flag-h-FOXO3a, MRE11, BRCA1, BRIP1, and RAD50 were also regulated FOXO3a-dependently in vivo.

In Vitro HRR Activity Increased in a Doxycycline-Dose-Dependent Manner in MEF Obtained from Embryos of Tet-On h-FOXO3a Transgenic Mice and Decreased during Cellular Aging of HDF
Mouse embryonic fibroblast (MEF) cells, established from embryos of tet-on h-FOXO3a transgenic mice, and young (PD24) and old (PD46) HDF cells were used for an in vitro HRR assay. wt-MEF cells were prepared from embryos of wild-type mice (C57BL/6N) to use as a control. Before transfection of recombination plasmid, MEF cells were treated with three concentrations of doxycycline (0, 0.5, 1.5, and 4.5 µg/mL) in DMEM-10% FBS. The MEF cells and HDF cells were transfected with DR-GFP (Addgene, #26475) plasmid and I-SceI (Addgene, #26477) plasmid for the in vitro HRR assay. DR-GFP is composed of two differentially mutated GFP regions. GFP* and iGFP have mutations in the sequence of GFP and do not make functional GFP protein. Homologous recombination between GFP* and iGFP after formation of a double-strand break in the GFP* site by I-SceI produces a functional GFP. Following transfection of the MEFs with the HRR vector and I-Scel vector, the MEF cells were again treated with three concentrations of doxycycline or control (0, 0.5, 1.5, and 4.5 µg/mL) and wt-MEF cells were treated with 4.5 µg/mL of doxycycline in DMEM-10% FBS for 48 h. HDFs were also transfected with the HRR vector and I-Scel vector and incubated for 48 h. Total RNA was isolated and a real-time qPCR (quantitative PCR) was performed to assay the HRR. The results showed that the mRNA levels of the recombined GFP increased in a doxycycline-dose-dependent manner in MEF cells (from transgenic mice) but not in the wt-MEF (from wild-type mouse), and the mRNA levels of the unrecombined GFP level did not change in a doxycycline dose-dependent manner ( Figure 5B); our results indicate that FOXO3a activates HRR. The mRNA levels of the recombined GFP increased doxycycline-dependently but the mRNA levels of the unrecombined GFP were unchanged ( Figure 5B). The mRNA levels of both the recombined GFP and the unrecombined GFP were unchanged even though doxycycline was treated ( Figure 5C). The results indicate that in vitro HRR activity is FOXO3a-dependent. The mRNA levels of the recombined GFP largely decreased (by 68%) in old (PD46) HDF cells when compared to young (PD46) HDF cells ( Figure 5B). The result suggests that cellularaging-dependent decrease of FOXO3a levels cause a decrease in HRR.  Statistical significance is indicated as * p < 0.05, ** p < 0.01, and *** p < 0.001. the recombined GFP and the unrecombined GFP were unchanged even though doxycycline was treated ( Figure 5C). The results indicate that in vitro HRR activity is FOXO3adependent. The mRNA levels of the recombined GFP largely decreased (by 68%) in old (PD46) HDF cells when compared to young (PD46) HDF cells ( Figure 5B). The result suggests that cellular-aging-dependent decrease of FOXO3a levels cause a decrease in HRR.  HDFs were also transfected with the HRR vector and I-Scel vector and incubated for 48 h. Old HDF cells were transfected with wt-h-FOXO3a for 24 h to induce FOXO3a before HRR assay to test the restoration of HRR activity. Total RNAs were isolated and RT-qPCRs for recombined and unrecombined GFP were performed to measure HRR activity. (A) Schematic diagram depicting the in vitro HRR assay. (B) RT-qPCR for total RNA was performed to detect recombined and unrecombined GFP in MEF (from tet-on h-FOXO3a transgenic mice) and (C) wt-MEF (from wild-type mice). Statistical significance is indicated as *** p < 0.001, ns: not significant.

Discussion
Regulation of homologous recombination repair (HRR) by FOXO transcription factors has not been reported. Here, we showed that FOXO3a activates HRR through transcriptional activation of its target genes, MRE11, BRIP1, RAD50, and BRCA1, using a cellular aging model of HDF and a model of tet-on flag-h-FOXO3a inducible transgenic mice ( Figure 6). As a DSB is one of the most deleterious DNA lesions, the cell repairs DSBs by two pathways, NHEJ and HRR. When HRR is deficient, DSB repair leads to enhanced dependence on alternative pathways including NHEJ, alternative end joining, and singlestrand annealing [32,33]. However, these pathways repair DSBs without a homologous DNA template, and are error prone, often producing small deletions and translocations around the DSB [4]. As such, HRR deficiency has been closely related to increases in cellular DNA mutation and increased incidence of cancers, particularly for breast and ovarian cancers [34][35][36][37]. As homologous recombination (HR) is also required for chromosome segregation during meiotic division, defects in the HR process also lead to genomic instability and enhanced cancer predisposition. The details of regulation of HRR by FOXM1 and FOXO3a remain to be solved. FOXO3a has been reported to interact with ATM to promote phosphorylation of ATM [45]. FOXM1 has also been shown to be upregulated following exposure to radiation by an undescribed mechanism [39]. Further characterization of the detailed regulation of FOXM1/FOXO3a by a DNA damage signal (ATM or γ-H2AX) and the detailed regulation of HRR by FOXM1/FOXO3a will provide a better understanding of the regulation of HRR.
Cellular aging in HDF cells decreased the levels of FOXO3a, MRE11, BRCA1, BRIP1, RAD50, and RAD51 ( Figure 2). HRR activity also declines distinctly in many tissues of old mice [10], and senescent fibroblasts display a dramatic decrease in HRR when compared to early growing fibroblasts [12,13]. FOXO3a levels have been shown to decrease in senescent fibroblasts [46] with a decreased binding to FOXO consensus sites of target genes in aged Drosophila [47]. HRR decreases in aged cells appear to be due to decreased levels and binding of FOXO3a to the target promoters involved in mediating HRR. We also observed a decrease in the levels of FOXM1 in aged cells (data not shown). The reduced levels of FOXM1 in aged cells also appears to contribute to decreased HRR in aged cells. On the separate or joint contribution of FOXO3a and FOXM1 to the reduced levels of HRR in aged cells, a detailed study is necessary for a better understanding. Additional contributions to reduced HRR in senescent fibroblasts may be via changes in the levels of SIRT6, as HRR was stimulated by overexpression of SIRT6, possibly through the activation of PARP1 [12,13]. Thus, in addition to FOXO3a/FOXM1, contributions of SIRT6 and PARP1 to HRR also merit a detailed study. We only showed aging-related regulation of HRR by FOXO3a in cellular aging of HDF cells, but not in body aging. Because cellular aging does not exactly represent body aging, regulation of HRR by FOXO3a should be further verified in body aging. In summary, in this study, key transcriptional targets of FOXO3a for HRR in the context of the aging models and the tet-on flag-h-FOXO3a inducible transgenic mouse were identified, pointing to a level of regulation by FOXO3a in modulating a key DNA repair mechanism.

Cell Culture and Transfections
Human dermal fibroblast (HDF) cells were obtained from the Dermatology Laboratory of Seoul National University Medical School (Seoul, Republic of Korea). HDF cells were transfected by Lipofectamine 3000 (Invitrogen, Waltham, MA, USA) according to the manufacturer's instructions. The reaction was performed with 30 μL of Lipofectamine On the regulation of HRR, FOXM1 has been implicated in the activation of HRR. FOXM1 has been reported to upregulate BRIP1 [38] and NBS1 [39], and FOXM1 activation upregulates RAD51 and BRCA1 in idiopathic pulmonary fibroblasts to activate HRR [40], with FOXM1 being a transcription factor that plays an important role in proliferation, cell cycle control, DNA repair, tumorigenesis, cancer progression, and tumor growth [41,42]. FOXO3a is reported to transcriptionally regulate GADD45 to mediate DNA repair [30], and is known to suppress DNA double-strand-break-induced mutation [43]. FOXO3a has also been reported to be upregulated at the transcriptional and translational level to activate DSB repair in bleomycin-treated MEF and HDF [31]. Therefore, FOXO3a may be involved in regulation of HRR activity. As a transcription factor, FOXM1 appears to activate the expression of target genes involved in HRR such as BRCA1/2, RAD51, BRIP1, and NBS1. The forkhead transcription factors, FOXM1 and FOXO3a, however, have been reported to suppress each other. Overexpression of FOXM1 resulted in downregulation of FOXO3a, and overexpression of FOXO3a induced downregulation of FOXM1 [44,45]. In this context, FOXO3a overexpression is expected to suppress the levels of RAD51, BRIP1, and BRCA1/2 those upregulated by FOXM1 overexpression. However, our results show that this is not true in the system of our study. FOXO3a overexpression was seen to induce transcriptional activation of MRE11, BRCA1, BRIP1, and RAD50 via increased interaction of FOXO3a with FOXO consensus sites of promoters of these genes (Figures 2-5), and FOXO3a overexpression induced activation of HRR ( Figure 6). In our system, FOXM1 overexpression caused upregulation of MRE11, BRIP1, BRCA1, RAD50, and RAD51 and downregulation of FOXO3a (data not shown). FOXM1 overexpression and FOXO3a overexpression were similar in upregulating MRE11, BRCA1, BRIP1, and RAD50. The difference was that FOXM1 overexpression induced RAD51 and BRCA2, but FOXO3a overexpression did not. Therefore, as both FOXM1 and FOXO3a are decreased during cellular aging, overexpression of either FOXM1 or FOXO3a activates HRR.
The details of regulation of HRR by FOXM1 and FOXO3a remain to be solved. FOXO3a has been reported to interact with ATM to promote phosphorylation of ATM [45]. FOXM1 has also been shown to be upregulated following exposure to radiation by an undescribed mechanism [39]. Further characterization of the detailed regulation of FOXM1/FOXO3a by a DNA damage signal (ATM or γ-H2AX) and the detailed regulation of HRR by FOXM1/FOXO3a will provide a better understanding of the regulation of HRR.
Cellular aging in HDF cells decreased the levels of FOXO3a, MRE11, BRCA1, BRIP1, RAD50, and RAD51 ( Figure 2). HRR activity also declines distinctly in many tissues of old mice [10], and senescent fibroblasts display a dramatic decrease in HRR when compared to early growing fibroblasts [12,13]. FOXO3a levels have been shown to decrease in senescent fibroblasts [46] with a decreased binding to FOXO consensus sites of target genes in aged Drosophila [47]. HRR decreases in aged cells appear to be due to decreased levels and binding of FOXO3a to the target promoters involved in mediating HRR. We also observed a decrease in the levels of FOXM1 in aged cells (data not shown). The reduced levels of FOXM1 in aged cells also appears to contribute to decreased HRR in aged cells. On the separate or joint contribution of FOXO3a and FOXM1 to the reduced levels of HRR in aged cells, a detailed study is necessary for a better understanding. Additional contributions to reduced HRR in senescent fibroblasts may be via changes in the levels of SIRT6, as HRR was stimulated by overexpression of SIRT6, possibly through the activation of PARP1 [12,13]. Thus, in addition to FOXO3a/FOXM1, contributions of SIRT6 and PARP1 to HRR also merit a detailed study. We only showed aging-related regulation of HRR by FOXO3a in cellular aging of HDF cells, but not in body aging. Because cellular aging does not exactly represent body aging, regulation of HRR by FOXO3a should be further verified in body aging. In summary, in this study, key transcriptional targets of FOXO3a for HRR in the context of the aging models and the tet-on flag-h-FOXO3a inducible transgenic mouse were identified, pointing to a level of regulation by FOXO3a in modulating a key DNA repair mechanism.

Cell Culture and Transfections
Human dermal fibroblast (HDF) cells were obtained from the Dermatology Laboratory of Seoul National University Medical School (Seoul, Republic of Korea). HDF cells were transfected by Lipofectamine 3000 (Invitrogen, Waltham, MA, USA) according to the manufacturer's instructions. The reaction was performed with 30 µL of Lipofectamine 3000 added to 0.5 mL of DMEM (Dulbecco's modified Eagle's medium, Thermo Fisher Scientific, Waltham, MA, USA), and 300 ng, 600 ng, or 1200 ng of wt-h-FOXO3a plasmid (Addgene, #8360) (Watertown, MA, USA) or HDF PD24 (young) cells were transfected with siFOXO3a (100 nM or 300 nM) (Cell Signaling Technology, Danvers, MA, USA) dissolved in 2 µL of Lipofectamine 3000 and 0.5 mL DMEM, respectively. The two solutions were mixed and added to 10 mL culture of HDF in DMEM in a 10 cm dish and incubated at 37 • C for 6 h. The HDF cells were then incubated at 37 • C for overnight in DMEM and supplemented with 10% fetal bovine serum (FBS) (Thermo Fisher) also containing an antibiotics mix of penicillin and streptomycin (Thermo Fisher). The HDF cells were subsequently used in Western blotting, ChIP and luciferase reporter assays. Transfection efficiency was determined by pCMV-beta-Gal-transfected HDF cells, according to the manufacturer's instructions (Takara Bio, Kusatsu, Shiga, Japan).

Primary Mouse Embryonic Fibroblasts
All animal protocols were approved by the Hallym University Institutional Animal Care and Use Committee (IACUC) (approval H20180119). Primary mouse embryonic fibroblasts (MEFs) were obtained from the embryos of pregnant tet-on flag-h-FOXO3a transgenic mice, and wt-MEFs were obtained from the embryos of pregnant wild-type mice (C57BL/6N). The embryos (embryonic day 12.5-14.5) of pregnant transgenic mice were taken out after incision surgery and rinsed in sterile cold PBS on ice to remove blood. These were then decapitated and their visible internal organs were removed, followed by mincing with sterile blade into small fragments as fine as possible and cultured in 10 cm dishes containing DMEM (Thermo Fisher), 10% FBS (Thermo Fisher), and penicillin/streptomycin (Thermo Fisher) at 37 • C in a humidified atmosphere containing 5% CO 2 . MEF cells were treated with control and three concentrations of doxycycline (0, 0.5, 1.5, and 4.5 µg/mL) and wt-MEF cells were treated with doxycycline (0, 4.5 µg/mL) DMEM-10% FBS for 24 h before transfection. Lipofectamine 3000 (L3000) (Invitrogen) was used for transfection of MEF cells according to the manufacturer's instructions.

Western Blot Analysis
HDF cells or tail-tip samples of tet-on flag-h-FOXO3a transgenic mice were lysed in lysis buffer (50 mM Tris-HCl, pH 7.4; 150 mM NaCl; 1 mM EDTA; 0.25% sodium deoxycholate; 1% NP-40; and supplemented with a protease inhibitor cocktail) (Sigma-Aldrich, St. Louis, MO, USA). A total of 25 micrograms of protein extract from the HDF cells or 40 µg of protein extract from the tail-tip samples were separated on an SDSpolyacrylamide gel and transferred to an ImmunoBlot PVDF membrane. The membranes were incubated with primary antibodies, washed, and then incubated with horseradish peroxidase-conjugated secondary antibodies. After washing, the resulting protein bands were visualized by ECL (Amersham, Little Chalfont, UK). The Western probing antibodies for FOXO3a, MRE11, BRCA1, BRCA2, BRIP1, RAD50, and RAD51 were purchased from Santa Cruz Biotechnology (Dallas, TX, USA); antibodies for PALB2, NBS1, and BARD1 were purchased from Proteintech (Rosemont, USA); Anti-Flag and anti-beta actin antibodies were from Cell Signaling Technology (Danvers, MA, USA).

Promoter Reporter Assay
HDF PD26 cells were transfected with the human promoter reporter plasmids (pGL3-MRE11, pGL3-BRCA1, pGL3-BRIP1, and pGL3-RAD50) and an increasing amount of wt-h-FOXO3a using Lipofectamine 3000 transfection reagent (Invitrogen) according to the manufacturer's instructions. The cell extracts were then prepared by incubating cells in cell lysis buffer (Promega, Madison, WI, USA) and the luciferase activity of the cell extracts was measured with a luminometer (GloMAX 20/20, Promega) using a luciferase assay reagent (Promega).

Development of the Tet-On Flag-h-FOXO3a Transgenic Mice
Ph-FOXO3a-tetO (5.17 kb) was constructed by insertion of flag-human FOXO3a cDNA between SV40 promoter-tet-operator and SV40 poly (A) signal in a vector. Ph-FOXO3a-tetO was then linearized by digestion with restriction enzymes and separated by agarose gel electrophoresis. The DNA was purified using a gel extraction kit (Qiagen). The purified DNA was microinjected at a concentration of 10 µg/mL in PBS into fertilized C57/BL6 mouse eggs (Macrogen, Seoul, Republic of Korea). Transgenic animals were identified by PCR of genomic DNA obtained from tail-tip samples. The sense and antisense primers were 5 -GCGGAGCGAGGAACTGAG-3 and 5 -CCGCCCTGGGA ATGATAG-3 , respectively. pLenti6-tet-R was constructed in a vector to include the CMV promoter/β-globin intron/tet-R/poly(A). Linearized pLenti6-tet-R was microinjected at a concentration of 10 µg/mL in PBS into fertilized C57/BL6 mouse eggs (Macrogen). Transgenic animals were identified by PCR of genomic DNA obtained from the tail-tip sampling. The sense and antisense primers were 5 -GCGGAGCGAGGAACTGAG-3 and 5 -CCGCCCTGGGA ATGATAG-3 , respectively. Two transgenic mice were mated and double-positive mice (flag-h-FOXO3a-tetO and tet-R) were obtained, as confirmed by PCR of tail-tip genomic DNA. All animal experimental protocols were approved by the Laboratory Animal Committee of Hallym University (IACUC) (approval number Hallym 2019-17). Twelve double-positive tet-on flagh-FOXO3a transgenic mice (12 weeks old) were used for testing flag-h-FOXO3a-dependent gene expression in the tail tissue samples. Carrier-only or solution containing doxycycline at 0.2 mL (0, 0.1, 1, 10 mg/mL 1% sucrose) was orally administered to transgenic mice once a day for 2 days. Cell extracts were prepared from the tail-tip samples on day 3, and their proteins were separated and analyzed by Western blotting. Anti-flag, anti-FOXO3a, anti-RAD50, anti-BRCA1, anti-BRIP1, anti-MRE11, anti-flag, and anti-β-actin antibodies were used for protein identification.

In Vitro Homologous Recombination Repair (HRR) Assay
In vitro homologous recombination assay was performed as described previously [44]. Briefly, MEF cells were established from embryos of tet-on flag-h-FOXO3a mice, and wt-MEF cells were prepared from C57BL/6N wild mice. MEF (from tet-on flag-h-FOXO3a transgenic mice) cells were treated with three concentrations of doxycycline (0, 0.5, 1.5, and 4.5 µg/mL); wt-MEF cells were treated 4.5 µg/mL doxycycline. Lipofectamine 3000 (L3000) (Invitrogen) was used for transfection of MEF cells as described in the section of cell culture and transfection. DR-GFP (Addgene, #26475) plasmid is composed of two differentially mutated GFP regions. The downstream GFP gene and the upstream GFP sequence contained the inactive I-SceI site within the GFP sequence before the homologous recombination repair. MEF cells were transfected with 2 µg DR-GFP plasmid and 2 µg I-SceI (Addgene, #26477) plasmid as described in the cell culture and transfection section. I-SceI expression vector was used to induce a double-strand break at a genomic I-SceI site. After transfection with the plasmids, MEF cells were treated with three concentrations of doxycycline or control (0, 0.5, 1.5, and 4.5 µg/mL) in DMEM-10% FBS for 48 h. Total RNA was prepared from MEF cells as described in the section for RNA extraction and quantitative real-time RT-qPCR. The cDNAs from unrecombined and recombined GFP mRNAs were synthesized from 1 µg total RNA in 20 µL by quantitative qPCR. Two sets of primers were prepared for use in RT-qPCR. The un-recombination (unrec) primer sequence was localized on the I-SceI site present only upstream of GFP; the second primer set was the recombinant (rec) primer sequence present downstream of GFP. The forward primer sequences for unrec was 5 -GCTAGGGATAACAGGGTAAT-3 ; for rec, it was 5 -GAGGGCGAGGGCGATGCC-3 ; for the reverse primer, it was 5 -TGCACGCTGCCGTCCTCG-3 . The conditions were 95 • C for 15 min, followed by 40 cycles of 95 • C for 10 s, 60 • C for 15 s, and 72 • C for 35 s. The PCR products were resolved and visualized in a 1.5% agarose gel.

Statistical Analysis
Statistical analysis was carried out by Student's t-test using the GraphPad Prism Version 4.0 software for Windows (GraphPad Software, San Diego, CA, USA). p-values less than 0.05 were considered to indicate statistical significance. All values are expressed as mean ± S.E.M.

Conclusions
The regulation mechanism of HRR activity by FOXO3a was addressed in HDF cells during cellular aging and in a FOXO3a-inducible transgenic mouse model system. MRE11, BRCA1, BRIP1, and RAD50 among nine HRR factors turned out to be regulated cellularaging-dependently and transcriptionally by FOXO3a in a cellular aging model of human dermal fibroblast (HDF). Furthermore, FOXO3a upregulated MRE11, BRCA1, BRIP1, and RAD50 at transcriptional and translational levels in tet-on flag-h-FOXO3a inducible transgenic mice. Likewise, HRR activity was shown to be regulated FOXO3a-dependently in an in vitro DR-GFP reporter HRR assay using an MEF (mouse embryonic fibroblast) obtained from the tet-on flag-h-FOXO3a inducible transgenic mice. In conclusion, MRE11, BRCA1, BRIP1, and RAD50 are transcriptional target genes of FOXO3a, and FOXO3a contributes to homologous recombination repair activity in MEF cells and mouse model systems.