BmNPV Orf 65 (Bm65) Is Identified as an Endonuclease Directly Facilitating UV-Induced DNA Damage Repair

ABSTRACT

Baculoviruses have been used as biopesticides for the control of Lepidoptera larvae. However, solar UV radiation reduces the activity of baculovirus. In this study, an UV endonuclease, Bm65, was found encoded in the genome of Bombyx mori nuclear polyhedrosis virus (BmNPV). Bm65 (the ortholog of AcMNPV orf79) was guided by a key nuclear localization signal to enter the nucleus and accumulated at UV-induced DNA damage sites. Subsequent results further showed that Bm65-mediated DNA damage repair was not the only UV damage repair pathway of BmNPV. BmNPV also used host DNA repair proteins to repair UV-induced DNA damage. In summary, these results revealed that Bm65 was very important in UV-induced DNA damage repair of BmNPV, and BmNPV repaired UV-damaged DNA through a variety of ways.

IMPORTANCE Baculovirus biopesticides are environmentally friendly insecticides and specifically infect invertebrates. UV radiation from the sunlight greatly reduces the activity of baculovirus biopesticides. However, the molecular mechanisms of most baculoviruses to repair UV-induced DNA damage remain unclear. Nucleotide excision repair (NER) is a major DNA repair pathway that removes UV-induced DNA lesions. At present, there are few reports about the nucleotide excision repair pathway in viruses. Here, we showed for the first time that the baculovirus Bm65 endonuclease actually cleaved UV-damaged DNA. Meanwhile, we found that BmNPV used both viral-encoded enzymes and host DNA damage repair proteins to reverse UV-induced DNA damage. These results will provide a reference for the research of UV damage repair of other viruses.

INTRODUCTION

Baculoviruses are enveloped double-stranded DNA viruses, specifically infecting insects, especially Lepidoptera, Hymenoptera, and Diptera insects (1). Baculoviruses are environmentally friendly insecticides and used in the control of insect pests, but UV radiation from the sunlight greatly reduces the activity of baculovirus biopesticides (2). An important target molecule of UV radiation is DNA. Cyclobutane pyrimidine dimers (CPDs) and pyrimidine 6‐4 pyrimidone photoproducts accounted for 75% and 25% of the total UV-induced DNA damage products, respectively (3). There are three pathways to repair UV-induced DNA damage, including photoreactivation, nucleotide excision repair (NER), and base excision repair (4). NER is the most important pathway to remove different DNA lesions, widely existing in prokaryotes and eukaryotes. NER generally includes damage recognition, incision, excision, repair synthesis, and DNA ligation (5). At present, photolyases that are responsible for photoreactivation, and directly reverse UV-induced DNA damage are found in a few baculoviruses (6–9). However, the DNA damage repair mechanisms of other sequenced baculoviruses are not clear.

BmNPV belongs to the family Baculoviridae and specifically infects silkworm. BmNPV ORF65 (Bm65) is a highly conserved gene in the genera Alphabaculovirus and Betabaculovirus, but not a core gene existing in all baculoviruses (10). In addition, orthologs of Bm65 are found in ascoviruses, iridoviruses, and several bacteria (11). Bm65 is a member of the GIY-YIG nuclease superfamily which is related to DNA repair, DNA recombination, foreign DNA restriction, and genome stability maintenance (11–17). The GIY-YIG module includes four or five conserved sequence motifs, two highly conserved residues and four invariant residues (13). Autographa californica multiple nucleopolyhedrovirus (AcMNPV) ORF79 shows 99% amino acid sequence identity to Bm65 and shows structural similarity to bacterial UvrC which has been found to participate in the NER pathway (11, 18). In our previous study, Bm65 was found to improve the survival rate of UV-irradiated Escherichia coli. The recombinant BmNPV budded viruses (BVs) overexpressing Bm65 also showed stronger resistance to UV radiation (15). Subsequent research showed that Bm65 significantly increased the expression level of UV-damaged mCherry reporter gene and enhanced survival of UV-damaged silkworm cells (19). Research on Ac79 showed that Ac79 encoded an early gene product with structural similarities to UvrC and intron-encoded endonucleases, and was required for efficient budded virus production (11).

In the current study, Bm65 was confirmed to be a very important UV damage repair protein and the absence of Bm65 resulted in a virus phenotype more sensitive to UV radiation. However, Bm65 was not a necessary protein in UV-induced DNA damage repair. BmNPV not only uses the UV endonuclease Bm65, but also uses host DNA damage repair proteins to repair UV-induced damage.

RESULTS

Bm65 is an UV endonuclease and the GIY-YIG domain is the key functional domain for cleaving UV-damaged DNA.

In order to preliminarily identify key functional domains of Bm65, we constructed recombinant plasmids separately expressing full-length Bm65, nuclear localization signal-missing Bm65 (Bm65-T5,1 to 36 aa:amino acid), and GIY-YIG domain-missing Bm65 (Bm65-T3,37 to 104 aa) (Fig. 1A, panel a).

FIG 1.

UV endonuclease activity assay of Bm65. (A) Effects of truncated Bm65 proteins on the UV sensitivity of E. coli. (a) Schematic diagram of truncated Bm65. (b) Survival rate analysis of recombinant E. coli expressing Bm65 or truncated Bm65 after UV treatment. Error bars indicate the mean ± standard deviation (SD) from three independent experiments. ** indicates a statistically significant difference (P < 0.01). (B) Expression and purification of Bm65 and truncated Bm65 protein. (a) SDS-PAGE and Western blotting of Bm65 and truncated Bm65 in E. coli. Lane 1, 3, Expression of GST in E. coli; lane 2, Expression of Bm65-GST in E. coli; lane 4, Expression of Bm65(T5)-GST in E. coli; lane 5, Expression of Bm65(T3)-GST in E. coli. (b) SDS-PAGE analysis of Bm65 and truncated Bm65 after purification. Lane 1, Expression of GST in E. coli; lane 2, Expression of Bm65-GST; lane 3 to 8, Bm65-GST proteins bounded to glutathione-agarose beads were eluted with elution buffer from the first time to the sixth time; lane 9, Expression of Bm65(T5)-GST; lane 10, The effluent of total proteins after passing through the ProteinIso GST Resin affinity column; lane 11 to 15, Bm65(T5)-GST proteins bounded to glutathione-agarose beads were eluted from the first time to the fifth time; lane 16, Expression of Bm65(T3)-GST; lane 17, The effluent of total proteins after passing through the ProteinIso GST Resin affinity column; lane 18 to 22, Bm65(T3)-GST proteins bounded to glutathione-agarose beads were eluted from the first time to the fifth time. (C) The UV endonuclease activity of Bm65 on UV-treated pIB-egfp plasmids. (a) DNA dot blot analysis of UV-induced damage (CPD) in UV-treated plasmid DNA. The plasmid pIB-egfp was exposed to UVC radiation (500 J/m² UV). The plasmid pIB-egfp without UV treatment was used as a negative control. Mouse monoclonal antibodies against CPD was used to detect UV-induced DNA damage. (b) The assay of UV endonuclease activity of Bm65. Nontreated and UV-treated pIB-egfp plasmids were, respectively, incubated with BSA, T4 endonuclease V, or a purified GST fusion of Bm65. T4 endonuclease V was used as a positive control, and BSA was used as a negative control. Endonuclease activity on UV-treated pIB-egfp would nick the plasmid at the pyrmidine dimmers to a slower migrating nicked circular form. The reaction products were separated by electrophoresis in 1% agarose gels.(c) The assay of UV endonuclease activity of truncated Bm65. Nontreated and UV-treated pIB-egfp plasmids were, respectively, incubated with BSA, T4 endonuclease V, Bm65 (T5), or Bm65 (T3).

E. coli BL21 cells were separately transformed with these recombinant plasmids and exposed to increasing doses of UV radiation. The effect of Bm65 and truncated Bm65 on the UV sensitivity of these recombinant E. coli was subsequently analyzed by calculating the survival rate. The results showed that after UV treatment, survival rate of BL21 cells expressing Bm65 or Bm65(T5) was significantly higher than that of BL21 cells containing the empty vector or expressing Bm65(T3) (Fig. 1A, panel b). Bm65(T5) had a similar capacity as full-length Bm65 protein to improve survival of E. coli after UV radiation. However, Bm65 mutant (T3) losing GIY-YIG conserved domain lost its ability to repair UV-damaged DNA.

To further investigate the DNA-cleaving ability of Bm65 in vitro, the UV endonuclease analysis was performed as described under “Materials and Methods.” First, full-length Bm65 protein and truncated Bm65 protein with GST label were expressed and purified (Fig. 1B).

T4 endonuclease V selectively cleaves UV-induced dimers and the unirradiated DNA was insensitive to T4 endonuclease V. Here, T4 endonuclease V was used as a positive control and BSA was used as a negative control. The plasmid DNA was induced by UV irradiation to produce CPD (the most prevalent form of UV-induced DNA damage), which was verified by DNA dot blot analysis (Fig. 1C, panel a). These plasmids with CPDs were then used as substrates for UV endonuclease activity analysis. The plasmids with pyrimidine dimer are converted to the nicked circular, slower mobility form by T4 endonuclease V treatment. By utilizing agarose gels, the size change of plasmid DNA can be visualized. Incubation of UV-irradiated plasmids with T4, Bm65, and Bm65(T5) resulted in conversion of circular plasmid DNA to nicked circular DNA and to slower mobility bands (Fig. 1C, panels b and c). The significant difference between the electrophoretic migration of UV-treated and untreated plasmid DNA, indicated that Bm65 and truncated Bm65 (T5) selectively cleaved UV-induced DNA, just like T4 endonuclease V. However, truncated Bm65 (T3) without GIY-YIG domain did not have UV endonuclease activity, suggesting that the GIY-YIG domain was the key functional domain of Bm65.

Full-length Bm65 protein, but not truncated Bm65 and mutant Bm65 proteins accumulate at UV-induced DNA damage sites of host cells.

BmN cells were separately transfected with pHTB-Pie1-Bm65-gfp, pHTB-Pie1-Bm65(T3)-gfp, pHTB-Pie1-Bm65(T5)-gfp, and pHTB-Pie1-Bm65(M2)-gfp to express Bm65-GFP, truncated Bm65-GFP, and mutant Bm65-GFP. To investigate whether Bm65 was recruited to sites of UV-induced DNA damage, we utilized a technique involving irradiation through a UV-opaque polycarbonate filter membrane with 5-μm pores. This method produced areas of local UV irradiation, such as a single damage site. The proteins recruited to the sites of UV-induced DNA damage could be readily visualized (Fig. 2A). An antibody specific for the most prevalent form of UV-induced DNA damage (CPD) was used to immunofluorescently visualize induced dimers. In this study, Bm65-GFP was found to be colocalizated with CPD. However, truncated Bm65 (T5-GFP or T3-GFP) and mutant Bm65 (M2-GFP) (the mutation of nuclear localization signal in ⁷⁶KRKCSK motifs) were all not recruited to UV-induced DNA damage sites (Fig. 2B). The results suggested that both the GIY-YIG domain and the nuclear localization signal were indispensable for the colocalization of Bm65 with UV-induced DNA damage.

FIG 2.

Recruitment of Bm65 to UV-induced DNA damage sites. (A) Schematic diagram of the local UV damage. (B) The localization analysis of Bm65-GFP with UV-induced DNA damage (CPD). From top to bottom: the localization of GFP with CPD; the localization of Bm65-GFP with CPD; the localization of Bm65(T5)-GFP with CPD; the localization of Bm65(T3)-GFP with CPD; the localization of Bm65(M2)-GFP with CPD.

Full-length Bm65 protein, but not truncated Bm65 and mutant Bm65 proteins enhance the expression of UV-damaged reporter gene.

Host cell reactivation assays were used to evaluate the effect of truncated Bm65 and mutant Bm65 proteins on the expression level of a UV-treated reporter gene in BmN cells. UV-irradiated plasmids containing a red fluorescent reporter gene (pHTB-Pie-1-mCherry) were cotransfected with plasmid pHTB-Pie-1-gfp, pHTB-Pie-1-Bm65-gfp, pHTB-Pie-1-Bm65(T5)-gfp, pHTB-Pie-1-Bm65(T3)-gfp, or pHTB-Pie-1-Bm65(M2)-gfp into BmN cells. At 72-h posttransfection, the repair extent of UV-induced DNA damage was quantitatively determined by monitoring the expression of mCherry. The results showed that there was no difference in the expression yield of unirradiated mCherry in these experimental groups and the mCherry expression decreased markedly with increasing UVC dose (Fig. 3A). Because UV radiation reduces mCherry expression, the increased expression level of mCherry in experimental groups than that in control groups indirectly reflects the more effective repair of DNA damage. The cells expressing GFP, truncated Bm65, or mutant Bm65 proteins (T5, T3, or M2) showed less red fluorescence than the cells expressing Bm65, suggesting that truncated Bm65 or mutant Bm65 did not help repair UV-damaged mCherry reporter gene like full-length Bm65 (Fig. 3A).

FIG 3.

Host cell reactivation assay. (A) Fluorescence images of BmN cells. The plasmid pHTB-Pie-1-mCherry (UV-treated or untreated) was cotransfected into BmN cells with plasmid pHTB-Pie-1-gfp, pHTB-Pie-1-Bm65-gfp, pHTB-Pie-1-Bm65(T5)-gfp, pHTB-Pie-1-Bm65(T3)-gfp, or pHTB-Pie-1-Bm65(M2)-gfp. At 72-hpt, cells were observed under a fluorescence microscope. (B) mCherry expression analysis. At 72-hpt, the expression level of mCherry was expressed as a percentage of unirradiated mCherry. Error bars indicate the mean ± SD from three independent experiments. ** indicates a statistically significant difference (P < 0.01).

Furthermore, the ratio of expression of UV-treated mCherry to untreated mCherry was analyzed at 72 hour post infection (hpt). The expression level of UV-treated mCherry in cells expressing full-length Bm65 was higher than that in cells expressing GFP, truncated Bm65, or mutant Bm65 (T5, T3, or M2), suggesting that none of the truncated Bm65 or mutant Bm65 (T5, T3, or M2) had the same capacity as Bm65 to repair a UV-damaged reporter gene (Fig. 3B).

Bm65-deleted virus is more sensitive to UV radiation than wild-type virus.

In order to investigate the effect of Bm65 deletion on the UV sensitivity of BmNPV, vBm^WT-GFP, Bm65-deleted virus (vBm^Bm65KO-GFP) and repair-type virus (vBm^Bm65-Rep-GFP) were exposed to UVC radiation and then infected BmN cells. The expression of green fluorescent protein was observed and the virus titer was determined using a TCID50 assay. After 500 J/m² and 1,000 J/m² UV treatment, the fluorescence of wild-type viruses, repair-type viruses, and Bm65-deleted viruses increased with time (Fig. 4A), suggesting that Bm65-deleted viruses still had the ability of replication and proliferation. However, the results further showed that the fluorescence of Bm65-deleted viruses decreased more significantly than that of control viruses after UV radiation (Fig. 4A), indicating that Bm65-deleted BmNPV was much more UV sensitive than wild-type virus and repair-type viruses. The virus titer of vBm^Bm65KO-GFP was significantly lower than that of vBm^WT-GFP and vBm^Bm65-Rep-GFP after UV radiation (Fig. 4B). These results indicated that Bm65-deleted BmNPV was more sensitive to UV radiation than wild-type virus, but Bm65-deleted virus might also repair UV-induced DNA damage in a Bm65-independent manner.

FIG 4.

Effect of Bm65 deletion on UV sensitivity of BmNPV. (A) Fluorescence microscopy analysis of virus-infected cells. BmN cells were infected with UV-irradiated BmNPV BVs (vBm^WT-GFP, vBm^Bm65KO-GFP, or vBm^Bm65-Rep-GFP) and observed by fluorescence microscope. (B) Proliferation analysis of virus after UV radiation. The titer of UV-irradiated BmNPV BVs (vBm^WT-GFP, vBm^Bm65KO-GFP, or vBm^Bm65-Rep-GFP) was determined by the method of TCID50. Error bars indicate the mean ± SD from three independent experiments. ** indicates a statistically significant difference (P < 0.01).

The mRNA levels of host DNA repair genes increases more significantly in UV-treated control virus-infected cells than that in UV-treated vBm^Bm65KO-infected cells.

Viruses can interact with host cells to directly activate and deactivate host DNA-damage response pathways (20). To investigate the effect of Bm65 deletion on the mRNA levels of host DNA-repair genes in BmNPV-infected cells, RT-qPCR assay was performed. The results showed that when viruses were exposed to UV radiation, the relative mRNA levels of host DNA damage repair genes (Bombyx mori Rad4, Bombyx mori Rad23, and Bombyx mori fen1) in vBm^Bm65KO-infected cells were significantly lower than that in vBm^Bm65-Rep-infected cells (Fig. 5A to C). However, the mRNA levels of the control protein Bombyxin did not differ between UV-treated control virus-infected cells and UV-treated Bm65 deletion virus-infected cells (Fig. 5D).

FIG 5.

RT-qPCR analysis of relative mRNA levels of host DNA damage repair genes in BmN cells. (A) The relative mRNA levels of BmRad4 in BmNPV-infected cells. (B) The relative mRNA levels of BmRad23 in BmNPV-infected cells. (C) The relative mRNA levels of Bmfen1 in BmNPV-infected cells. (D) The relative mRNA levels of Bombyxin in BmNPV-infected cells. The experiments were divided into the following groups: (a) BmN cells were infected with control viruses (vBm^Bm65-Rep) or Bm65-deleted viruses (vBm^Bm65KO); (b) 60 J/m² UVC treated cells were infected with viruses (vBm^Bm65-Rep or vBm^Bm65KO); (c) Cells were infected with 500 J/m² UVC treated viruses (vBm^Bm65-Rep or vBm^Bm65KO). Error bars indicate the mean ± SD from three independent experiments. * indicates a statistically significant difference (P < 0.05). ** indicates a statistically significant difference (P < 0.01).

mRNA levels of host DNA damage repair genes were still upregulated after UV-treated Bm65-deleted viruses’ infection.

The RT-qPCR results further showed that the relative mRNA levels of host DNA damage repair genes (BmRad4, BmRad23, and Bmfen1) were still increased when vBm^Bm65KO or BmN cells were exposed to UVC light, and was significantly higher than that in vBm^Bm65KO-infected cells without UV treatment (Fig. 6). When cells were exposed to UV radiation, the expression of DNA repair related genes increases to help repair UV-induced damage. The expressions of host DNA repair proteins were still increased after UV-treated Bm65-deleted virus infection, indicating that host DNA damage repair proteins were also associated with UV damage repair of Bm65-deleted viruses.

FIG 6.

RT-qPCR analysis of the relative mRNA levels of host DNA damage repair genes in vBm^Bm65KO-infected cells. (A) The relative mRNA levels of BmRad4 in vBm^Bm65KO-infected cells. (B) The relative mRNA levels of BmRad23 in vBm^Bm65KO-infected cells. (C) The relative mRNA levels of Bmfen1 in vBm^Bm65KO-infected cells. (D) The relative mRNA levels of Bombyxin in vBm^Bm65KO-infected cells. The experiments were divided into the following groups: BmN cells were infected with Bm65-deleted viruses (vBm^Bm65KO); 60 J/m² UVC treated cells were infected with vBm^Bm65KO; cells were infected with 500 J/m² UVC treated vBm^Bm65KO; 60 J/m² UVC treated cells were infected with 500 J/m² UVC treated vBm^Bm65KO. Error bars indicate the mean ± SD from three independent experiments. * indicates a statistically significant difference (P < 0.05). ** indicates a statistically significant difference (P < 0.01).

BmNPV uses host DNA damage repair protein (BmRad23) to repair UV-induced DNA damage.

Immunofluorescence assay was used to observe whether DNA damage repair proteins still colocalized with UV-induced DNA damage of Bm65-deleted viruses. The results showed that Bm65 and the crucial DNA damage recognition protein BmRad23 were accumulated at the sites of UV-induced DNA damage (CPD) of Bm65-deleted viruses (Fig. 7A, panel b). From this we speculate that Bm65-deleted viruses might use host DNA damage repair proteins to repair UV-induced damages.

FIG 7.

The effect of BmRad23 on BmNPV to repair UV-induced DNA damage. (A) The localization analysis of BmRad23-GFP with UV-induced DNA damage (CPD) of Bm65-deleted viruses. (a) Schematic diagram of UV damage of Bm65-deleted viruses; (b) recruitment of DNA damage repair proteins to UV-induced DNA damage sites of Bm65-deleted viruses. BmN cells were, respectively, transfected with plasmids pHTB-Pie-1-BmRad23-gfp and pHTB-Pie-1-gfp to express BmRad23-GFP and GFP. Bm65-deleted viruses (vBm^Bm65KO) were exposed to 500 J/m² UV radiation and then infected Bm65-GFP-expressing cells or BmRad23-GFP-expressing cells, respectively. The colocalization of DNA damage repair proteins (Bm65 or BmRad23) with UV-induced DNA damages (CPD) of Bm65-deleted viruses were analyzed by immunofluorescence at 3 h after infection. (B) The effect of BmRad23 on UV-induced DNA damage repair of BmNPV. (a) Knockdown efficiencies of BmRad23 siRNA in BmN cells. BmN cells were transfected with BmRad23 siRNA. After incubation for 48 h, the BmRad23 mRNA levels were determined by RT-qPCR. (b) The effect of BmRad23-knockdown on UV-induced DNA damage repair of control virus and Bm65-deleted virus. The experiments were divided into the following groups: BmN cells were infected with UV-treated control viruses (vBm^Bm65-Rep) or UV-treated Bm65-deleted viruses (vBm^Bm65KO); BmRad23-knockdown cells were infected with UV-treated viruses (vBm^Bm65-Rep or vBm^Bm65KO). (c) The effect of BmRad23 on UV-induced DNA damage repair of Bm65-deleted virus. The experiments were divided into the following groups: BmN cells were infected with UV-treated Bm65-deleted viruses (vBm^Bm65KO); BmRad23-knockdown cells were infected with UV-treated vBm^Bm65KO; BmRad23-overexpression cells were infected with UV-treated vBm^Bm65KO. The UV-induced DNA damage of BmNPV was analyzed by CPD ELISA Kit at 3-, 6-, and 12-h postinfection, as shown, with the indicated viruses. Error bars indicate the mean ± SD from three independent experiments. * indicates a statistically significant difference (P < 0.05), ** indicates a statistically significant difference (P < 0.01).

CPD was the major UV-induced DNA lesion and the UV-induced DNA damage of BmNPV was further analyzed by using a CPD ELISA Kit. At different time points after infection, the changes of CPD in these UV-treated control virus-infected cells or UV-treated Bm65-deleted virus-infected cells were analyzed. The results showed that UV-induced DNA damage in control and Bm65-deleted virus-infected cells both decreased with time. Meanwhile, the UV-induced DNA damage in Bm65-deleted virus-infected cells decreased more slowly (Fig. 7B, panel b). These results indicated that Bm65 was very important in UV-induced DNA damage repair. However, the UV-induced DNA damage of Bm65-deleted viruses was still repaired by a Bm65-independent repair pathway.

To further analyze whether BmNPV uses host DNA damage repair proteins to repair UV-induced DNA damage, BmN cells were transfected with siRNA to knock down the expression of BmRad23. The results showed that BmRad23 mRNA expression was effectively downregulated in siRNA-transfected cells (Fig. 7B, panel a). The DNA ELISA results further showed that CPDs of control viruses and Bm65-deleted viruses both decreased more slowly in siRNA-transfected cells than that in control cells (Fig. 7B, panel b), suggesting that BmRad23 was also important for BmNPV to repair UV-induced DNA damage. Furthermore, BmN cells were transfected with pHTB-Pie-1-BmRad23-gfp to overexpress BmRad23. It was further confirmed that overexpression of BmRad23 contributed to faster repair of UV-damaged DNA of Bm65-deleted viruses (Fig. 7B, panel c). Based on the above results, we speculated that BmNPV not only used its own DNA repair protein Bm65, but also used host DNA damage repair proteins to repair UV-induced damage.

DISCUSSION

Irradiation with UV is an effective means of virus inactivation. Viruses usually use host-encoded enzymes or viral-encoded enzymes to reverse UV-induced DNA damage and survive in varied environments (21, 22). Baculovirus is also very sensitive to UV rays in sunlight. Pyrimidine dimers induced by UV radiation were reported as one factor responsible for inactivation of baculovirus (23).

The NER is critically important in repair of UV-induced DNA lesions and is found in most organisms. Herpes simplex virus 1 has been reported to encode a functional thymidine kinase. The repair of UV-irradiated herpes simplex virus 1 (HSV-1) influenced by thymidine kinase activity is by excision repair or a process dependent on excision repair (24). Autographa californica multicapsid nucleopolyhedrovirus (AcMNPV) ORF79 (Ac79), the protein homologue of Bm65, was reported to have predicted structural similarities to UvrC, bacteriophage T4 endonuclease II, and the I-TevI intron-encoded endonucleases with the GIY-YIG domain (11). In this study, BmNPV Bm65 protein was shown to be an endonuclease and cleaved UV-damaged DNA, indicating that some baculoviruses use their own nucleic acid excision repair pathway to repair UV-induced DNA damages. We further found that GIY-YIG domain was the key functional domain of Bm65 endonuclease. It was interesting that Bm65-T5 (1 to 36 aa which retains the key GIY-YIG domain and losing the nuclear location signal) still had endonuclease activity for UV-irradiated DNA, and also enhanced the UV tolerance of E. coli. However, none of the truncated Bm65 or mutant Bm65 (Bm65-T5; Bm65-T3, 37 to 104 aa, losing the key GIY-YIG domain and retaining the nuclear location signal; Bm65-M2, retaining the key GIY-YIG domain and mutating the nuclear location signal) played a role in UV damage repair in silkworm cells. In our previous studies, the ⁷⁶KRKCSK motif was found to be critical for nuclear localization of Bm65. The mutation in ⁷⁶KRKCSK abrogated the import ability of Bm65 from cytoplasm to nucleus (10). In this study, it was further found that none of Bm65(T5), Bm65(T3), and Bm65(M2) was colocalizated with UV-induced DNA damage in infected silkworm cell. It was speculated that because prokaryotes had no nuclear membrane, Bm65(T5) could act directly on UV-damaged DNA of E. coli. In eukaryotes, the nuclear import of proteins is generally selective and signal-dependent (25). It is generally considered that small proteins can passively diffuse between cytoplasm and nucleus, and nuclear location signals are required for nuclear import of larger proteins (26, 27). However, Bm65(T5) that only retained the functional domain of endonuclease and located in cytoplasm and nucleus, still did not accumulate at sites of UV-induced DNA damages in host nucleus and did not help repair UV-induced DNA damage. Therefore, it is speculated that Bm65 needs a more precise regulatory mechanism of localization signal to help play the role of UV damage repair in host cells.

In this study, Bm65-deleted viruses were found to be more UV-sensitive than wild-type viruses, but were not completely defective in the repair of UV-induced DNA damages. The results suggested that the repair of UV-irradiated BmNPV in host cells depends, at least in part, on expression of Bm65.

UV-radiation induces DNA damage and activates DNA repair pathway in host cells (28). Viruses usually use host enzymes and related proteins to repair DNA damage, such as herpes simplex virus, which does not encode enzymes suitable for handling UV-damaged DNA and uses host nucleotide excision repair pathways (29, 30). However, not all repair pathways are readily available to the virus, and not all repair proteins are required. Radiation sensitive 4 (Rad4) and radiation sensitive 23 (Rad23) are the critical DNA repair proteins that can form complexes to recognize various types of DNA lesions (31–33). Flap endonuclease1 (Fen1) is also an important protein involved in repair of UV-induced DNA damage (34–36). Bombyxin regulates the utilization of storage carbohydrates (37). So far, there is no evidence that Bombyxin is associated with DNA damage repair, and is used as a negative control in this study. Our previous RT-qPCR results showed that the transcription level of BmRad4, Bmfen-1, and Bm65 were significantly increased in BmNPV-infected cells when BmNPV or BmN cells were treated by UV radiation, suggesting that BmNPV and host might mutually utilize the same DNA repair proteins for repairing UV-induced DNA damage (38). In the current study, we further analyzed whether Bm65-deleted viruses still affected the transcription level of host DNA damage repair genes after UV radiation. The results showed that the mRNA levels of these host DNA repair genes were still increased when Bm65-deleted viruses were treated by UV irradiation. However, the increased transcription level of these DNA damage repair genes observed in UV-treated Bm65-deleted virus-infected cells did not reach the same level as that in UV-treated control virus-infected cells. Compared with Bm65-deleted viruses, higher UV tolerance of control virus might be because Bm65 directly cleaved UV-damaged DNA, and due to higher expression of host repair genes after UV-treated control virus infection.

Immunofluorescence assay further showed that host DNA damage repair proteins (BmRad23) were still accumulated at the UV-damaged DNA sites of Bm65-deleted viruses and the UV-induced DNA damages of Bm65-deleted viruses decreased gradually with time. Furthermore, DNA ELISA analysis confirmed that the UV damage repair ability of BmNPV significantly decreased in BmRad23-knockdown cells. The overexpression of BmRad23 could make up for the decline of the UV damage repair ability of Bm65-deleted viruses in a certain extent, suggesting that Bm65-deleted viruses could still use host proteins to repair UV-induced DNA damages.

In summary, Bm65 is a very important UV endonuclease in UV-induced DNA damage repair of BmNPV, and the absence of Bm65 resulted in a more sensitive virus phenotype to UV radiation. BmNPV uses both viral-encoded enzymes and host DNA damage repair proteins to reverse UV-induced DNA damage (Fig. 8).

FIG 8.

Schematic diagram showing the ways of BmNPV repairing UV-induced DNA damage. 1, Bm65 directly participates in DNA damage repair by cleaving UV-induced dimers; 2, BmNPV uses host DNA repair proteins to repair UV-induced DNA damage.

MATERIALS AND METHODS

Plasmids, viruses, and cell.

The plasmid pGEX-5X-3 was conserved in our laboratory. The plasmid pHTB-Pie-1-mCherry expressing mCherry reporter gene was constructed previously (39). The plasmid pHTB-Pie-1-gfp was constructed by Li et al. (40) in our laboratory. The plasmids pHTB-Pie-1-Bm65-gfp and pHTB-Pie-1-BmRad23-gfp were constructed by Tang et al. (15, 38, 41). The plasmid pIB-egfp was conserved in our laboratory. BmN cells were cultured at 27°C in TC-100 insect medium supplemented with 10% fetal bovine serum (Gibco/Thermo Fisher Scientific, Waltham, MA, USA). The plasmids pHTB-Pie1-Bm65(T3)-gfp, pHTB-Pie1-Bm65(T5)-gfp, and pHTB-Pie1-Bm65(M2)-gfp (the mutations in 76KRKCSK motifs) were constructed by Li et al. (10). Bm65-knockout BmNPV BVs (vBm^Bm65KO-GFP) which were obtained by homologous recombination, and repair-type viruses (vBm^{Bm65-Flag-GFP}) which were obtained by transposition were all conserved in our laboratory (10). Recombinant BmNPV (vBm^WT-GFP) expressing green fluorescent protein (GFP) was constructed in previous work (15). All primers in the study were listed in Table 1.

TABLE 1.

Primer sets used in the study

Primer name	Primer sequence (5′→3′)	Enzyme digestion sites
Bm65-F	ATGGATCCATGGCGACGACTCTGTACACCA a	BamH I
Bm65-R	ATCTCGAGCAACTTATTTGCTAACAGAAATTT	Xho I
Bm65-GST-F	ATGGATCCCCATGGCGACGACTCTGTA	BamH I
Bm65-GST-R	ATCTCGAGTTACAACTTATTTGCTAACAGAAATTTAT	Xho I
Bm65-T5(1 to 36aa)-F	ATGGATCCATGGCGACGACTCTGTACACCA	BamH I
Bm65-T5(1 to 36aa)-R	ATCTCGAGTTTTATGCGTCTGTTAAGGTTGCTC	Xho I
Bm65-T3(37 to 104aa)-F	ATGGATCCATGCAGCATTCGAACAAACAAG	BamH I
Bm65-T3(37 to 104aa)-R	ATCTCGAGCAACTTATTTGCTAACAGAAATTT	Xho I
qBmRad4-F	GGGTCAGAAGCGCACATTTC
qBmRad4-R	GCGAACGGTTTCTTCGATGG
qBmRad23-F	TTCCATACCCGAAGCAGG
qBmRad23-R	GGTGTTGGGAGTCTCGGTAA
qBmfen-1-F	ACCTGAAGTCGCTGACCCTA
qBmfen-1-R	TAGCTCCATTCCTCACTCGTTC
qBombyxin-F	CAATGCGGAAAGGGTGTA
qBombyxin-R	CGTTGCTTGGTCGTGTCT
qα-tubulin-F	CTCCCTCCTCCATACCCT
qα-tubulin-R	ATCAACTACCAGCCACCC
BmRad23 siRNA	CGCGCCUCAUUUAAUAAUATT UAUUAUUAAAUGAGGCGCGTT
siRNA (negative control)	UUCUCCGAACGUGUCACGUTT ACGUGACACGUUCGGAGAATT

^aUnderlined letters indicate restriction enzyme digestion sites.

Effects of truncated Bm65 proteins on the UV sensitivity of E. coli.

Bm65 and truncated Bm65 was amplified by PCR and ligated with pET30a. The E. coli BL21 containing pET30a-Bm65, pET30a-Bm65(T5), pET30a-Bm65(T3), and pET30a were, respectively, spread onto LB plates and exposed to 0 J/m², 0.15 J/m², or 0.3 J/m² UVC. After an 18-h incubation in a dark, the colonies were counted. The ratio of the bacteria colonies of the experimental group (UV-treatment) to that of the control group (untreatment) was used as the survival rate.

Expression and purification of recombinant protein Bm65-GST, Bm65(T5)-GST, and Bm65(T3)-GST.

Primer pairs were designed to amplify Bm65, Bm65(T5), or Bm65(T3) from Bm-Bacmid. Then, target DNA was purified and ligated with pGEX-5X-3. These recombinant plasmids were separately transformed into E. coli to express Bm65-GST, Bm65(T3)-GST or Bm65(T5)-GST. Isopropyl-beta-d-thiogalactopyranoside (IPTG) was added to induce the expression of Bm65-GST, Bm65(T3)-GST, or Bm65(T5)-GST. The cell pellets were harvested by centrifugation and resuspended in buffer A (140 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄, 1.8 mM KH₂PO₄, pH 7.3). Then, the cell suspensions were sonicated at 2 W on ice and clarified by centrifugation. The supernatants were loaded onto a ProteinIso GST Resin affinity column (TRAN, China). After washing with PBS, the fusion proteins were eluted with elution buffer (50 mM Tris-HCl, 10 mM reduced glutathione, pH 8.0). Eluted fractions were subjected to 15% SDS-PAGE analysis. Using the mouse monoclonal antibody against GST (Abcam, UK) as the first antibody and the goat anti-mouse IgG H&L (HRP) (Abcam, UK) as the second antibody, the expression of Bm65 and truncated Bm65 were analyzed by Western blot.

DNA dot blot assay.

To generate CPD (the most prevalent form of UV-induced DNA damage), the plasmids were exposed to 500 J/m² UVC light generated by a hand-held wand (Thermo, USA). The UV doses were determined by an UV radiometer (Lutron, China). To verify whether CPDs were produced in the plasmids DNA after UV treatment, a DNA dot blot assay was performed. For negative control, plasmid DNA without UV treatment was used. UV-treated (500 J/m² UVC) and untreated plasmid DNA were added onto nitrocellulose membrane. The membrane was blocked with blocking solution (1X PBS-T buffer with 5% nonfat milk powder). Then, the membrane was incubated with mouse monoclonal antibodies against CPD (Cosmo Bio, Japan). The secondary antibody used in the study was goat anti-mouse IgG H&L (HRP) (Abcam, UK). These plasmids with CPDs were then used as substrates for UV endonuclease activity analysis.

Assay of UV endonuclease activity.

UV endonuclease activity analysis was performed according to previous methods (42, 43) with some modifications. UV-treated plasmids containing CPDs and untreated plasmids (5 μg) were, respectively, incubated with bovine serum albumin (BSA, 20 μM), T4 endonuclease V (10 U), Bm65 (20 μM), Bm65-T5 (20 μM), and Bm65-T3 (20 μM) for 30 min at 37°C according to the T4 endonuclease V manufacturer’s instructions (Biolab, China). T4 endonuclease V specifically cleaved UV-induced pyrimidine dimers and was used as a positive control. BSA was used as a negative control. All reaction products were resolved by electrophoresis in 1% agarose gels.

Colocalization analysis of Bm65 with UV-induced DNA damage (CPD) of host cells.

Fluorescence microscopy was further used to observe the colocalization of Bm65-GFP with UV-induced DNA damage. BmN cells were cultured on coverslips and separately transfected with pHTB-Pie-1-Bm65-gfp, pHTB-Pie-1-Bm65(T5)-gfp, pHTB-Pie-1-Bm65(T3)-gfp, and pHTB-Pie-1-Bm65(M2)-gfp. After culturing for 48 h at 27°C, the cells were washed with PBS. After being covered by a UV-opaque polycarbonate filter membrane (5-μm pore size, Millipore, Billerica, MA, USA), the transfected cells were irradiated with 200 J/m² UVC. After culturing for 3 h, the cells were fixed with 4% paraformaldehyde for 15 min and permeabilized in 0.1% Triton X-100 for 15 min. After incubation with blocking solution (1X PBS-T buffer with 10% nonfat milk powder), the coverslips were incubated with mouse monoclonal antibodies against CPD (Cosmo Bio, Japan). The secondary antibody used in the study was goat anti-mouse IgG H&L (Alexa Fluor 647) (Abcam, UK). After being washed with PBS-T, the cells were counterstained with DAPI (Sigma, UK) to visualize the nuclei and photographed using a fluorescence microscope (Leica, Wetzlar, Germany).

Effect of Bm65 on the UV sensitivity of BmNPV.

To further determine whether deletion of Bm65 affected the UV sensitivity of virus, wild-type viruses (vBm^WT-GFP), Bm65-deleted virus (vBm^Bm65KO-GFP), and repair-type viruses (vBm^Bm65-Rep-GFP) were diluted to an equal titer using PBS and, respectively, exposed to 500 J/m², 1,000 J/m², or 1,500 J/m² UVC radiation. Then, BmN cells were infected with these UV-treated viruses. Viral infections were characterized by observation of green fluorescence in BmN cells using fluorescence microscopy (Leica, Germany). The supernatants of cells infected with wild-type viruses, Bm65-deleted virus, or repair-type viruses were harvested at selected time points. The viral titer was determined by 50% tissue culture infective dose (TCID50) assays (44, 45).

RT-qPCR analysis.

RT-qPCR was used to determine the transcription level of host DNA-repair genes in vBm^Bm65KO-infected BmN cells. Four groups of BmN cells were used in the experiment: BmN cells were infected with vBm^Bm65-Rep or vBm^Bm65KO (MOI = 5); 60 J/m² UVC-treated BmN cells were infected with vBm^Bm65-Rep or vBm^Bm65KO (MOI = 5); BmN cells were infected with 500 J/m² UVC-treated viruses (vBm^Bm65-Rep or vBm^Bm65KO, MOI = 5); and 60 J/m² UVC-treated BmN cells were infected with 500 J/m² UVC-treated viruses (vBm^Bm65-Rep or vBm^Bm65KO, MOI = 5). Total RNA was separately extracted from these cells using TRIzol reagent (Invitrogen, USA) at different time points after infection. Normal cells without any treatment were used as mock controls. Bombyx mori alpha-tubulin (Bmtubulin) was used as an internal control. RT-qPCR analysis was performed using the 2^-ΔΔCt method. The mRNA levels of DNA repair genes from the mock group were set as 1, and the mRNA levels of these genes from the experimental groups relative to the mock group were calculated to be a ratio.

Colocalization analysis of DNA damage repair proteins (Bm65 and BmRad23) with UV-induced DNA damage (CPD) of Bm65-deleted viruses.

To examine whether the DNA damage repair proteins were still recruited to UV-induced DNA damages sites of Bm65-deleted viruses, vBm^Bm65KO were exposed to an incident UVC dose of 500 J/m². First, BmN cells were transfected with pHTB-Pie-1-Bm65-gfp or pHTB-Pie-1-BmRad23-gfp to express Bm65-GFP or BmRad23-GFP. After culturing for 48 h at 27°C, the cells were infected with 500 J/m² UV-treated vBm^Bm65KO. After culturing for 3 h, the cells were fixed with 4% paraformaldehyde and permeabilized in 0.1% Triton X-100. An anti-CPD mouse monoclonal antibody (Cosmo Bio, Japan) was used as primary antibody. The secondary antibody was goat anti-mouse IgG H&L (Alexa Fluor 647) (Abcam, UK). The cells were counterstained with DAPI (Sigma, UK) to visualize the nuclei and photographed using a fluorescence microscope (Leica, Wetzlar, Germany).

Effect of BmRad23 knockdown on UV damage repair of BmNPV.

BmRad23 siRNA was designed and synthesized (GenePharma, China), and transfected into BmN cells using Cellfectin II reagent (Invitrogen, USA). Total RNA was isolated using TRIzol Reagent (Invitrogen, USA) at 48-h posttransfection. The knockdown efficiency was confirmed by RT-qPCR.

Control viruses (vBm^Bm65-Rep) and Bm65-deleted viruses (vBm^Bm65KO) were diluted to an equal titer using PBS and, respectively, exposed to 500 J/m² UVC radiation. At 48 hpt, siRNA-transfected cells were infected with these UV-treated viruses. After culturing for different time, the total DNA of infected cells was extracted. UV-induced DNA damage in extracted DNA was further analyzed using CPD ELISA Kit (Abcam, UK).

Accession numbers.

All sequence data that support the findings of this study are available in GenBank with the following accession numbers: Bm65 (BBA20568.1), BmRad4 (ADC53488.1), BmRad23 (ACK38234.1), Bmfen1 (ACY92094.1), and Bombyxin (BAA00246.1).