To study the transcription of ac73 in AcMNPV infected Sf9 cells, we first predicted the promoter of ac73. A TTAAG motif which is the typical feature of baculovirus late promoter (Morris and Miller 1994), was found from nt 61–57 upstream of ATG of ac73 (Fig. 1A), indicating that ac73 may be a late gene. This was consistent with a report that in AcMNPV infected T. ni cells, ac73 was mainly transcribed during late infection at nt 57 upstream of ATG of ac73 (Chen et al. 2013). To further determine whether ac73 is really a late gene, the transcripts of ac73 at different time points of infection were detected through PCR amplification of an inner fragment (~ 270 bp) within ac73. Result showed that the transcripts of ac73 could be detected from 12 to 72 h p.i. (Fig. 1B), indicating that ac73 was expressed at the late stage of infection. In addition, Western blot analysis was performed to detect AC73 protein levels in infected cells. The AC73 protein was under the detectable level before 18 h p.i., but was clearly detected since 24 h p.i. (Fig. 1C). For reference, VP39, the well-known viral late protein (Thiem and Miller 1989), could also be clearly detected since 18 h p.i. (Fig. 1C). With the above results, we can conclude that ac73 is a late gene of AcMNPV.
Figure 1. Ac73 is a late viral gene. A The promoter prediction of ac73. The predicted late transcription motif TTAAG is underlined and indicated in red. B Transcription detection of ac73. The ac73 transcripts from different time points of infection were detected through PCR amplification of a fragment of the ORF ac73 (~ 270 bp) using cDNA as templates. bp, base pair. C Expression detection of ac73. The AC73 protein at different time points of infection was detected by Western blot using anti-AC73 antibody. VP39 which is expressed during late infection served as a control together with cellular GAPDH.
Next, the subcellular localization of AC73 in infected cells was determined by immunofluorescence microscopy. In Ac-egfp infected cells, AC73 could be clearly detected in the nucleus of infected cells since 18 h p.i., and it was mainly localized in the ring zone region peripheral to the nuclear membrane during virus infection (Fig. 2A). To further determine whether AC73 can enter the nucleus independently, EGFP fused AC73 or the control EGFP alone was transiently expressed by transfection of the corresponding plasmid into Sf9 cells. The result showed that both EGFP and EGFP fused AC73 were evenly distributed in the cytoplasm and nucleus (Fig. 2B), suggesting that AC73 alone could not enter the nucleus completely. This is consistent with the fact that no nuclear localization signal could be predicted in AC73 (data not shown). To exclude the possibility of effect of EGFP on the localization of AC73, Sf9 cells were first transfected with plasmid encoding EGFP or EGFP fused AC73 and then infected with WT AcMNPV. Compared to the result of transient expression, EGFP fused AC73 but not EGFP in the superinfected cells showed clear nuclear localization and was embedded into OBs (Fig. 2B). Thus, the results suggested that AC73 could enter the nucleus in an infectiondependent way and it seemed to be assembled into OBs.
Figure 2. The subcellular localization of AC73. A Immunofluorescence assay of AC73 localization during Acegfp infection. Sf9 cells were infected with Ac-egfp which does not contain ph, thus no OB formed. Cells were fixed at indicated time points of infection, and the anti-AC73 antibody and Alexa 647-conjugated goat anti-rabbit antibody were used as the primary and secondary antibody for detection of AC73, respectively. EGFP fluorescence indicated the cells were successfully infected; the nuclei of cells were stained with Hoechst 33258 dye (blue). B The localization of EGFP fused AC73 in transfected and infected cells. V - the cells were transfected with plasmid encoding EGFP or EGFPAC73. V + transfected cells (V -) were further infected with WT AcMNPV which can express polyhedrin for the formation of OBs and observed at 48 h p.i.. DIC differential interference contrast. Bars, 10 μm.
Previous proteomics data revealed that AC73 was associated with BV (Wang et al. 2010), but not with ODV (Braunagel et al. 2003). However, our result found that EGFP fused AC73 could be assembled into OBs, suggesting that AC73 may also be ODV-associated. To test this possibility, BVs and ODVs were prepared from the supernatant of WT AcMNPV infected cells and larvae, respectively, and then analyzed by Western blot with antiAC73 antibody. As shown in Fig. 3A, AC73 could be probed in both BV and ODV samples. To further confirm this result and to determine the localization of AC73 in virion more accurately, the BVs and ODVs were fractionated into envelope and nucleocapsid components. As the Western blot result showed, AC73 could be detected in nucleocapsid samples of BV and ODV, but not in the envelope samples (Fig. 3B). Thus, AC73 is a nucleocapsid component of both BV and ODV.
Figure 3. AC73 is a nucleocapsid protein of BV and ODV. A AC73 is the component of BV and ODV. The BVs and ODVs were purified and the intact virions were subjected to SDS-PAGE for Western blot assay by using anti-AC73. B AC73 is a nucleocapsid protein. The purified BVs and ODVs were fractionated into envelope (E) and nucleocapsid (NC) fractions, and then analyzed together with intact BV and ODV samples. The healthy cells (HC) and infected cells (IC) were served as controls. for A, B GP64 serves as positive control for the envelope of BV; PIF5 and ODV-E66 served as positive controls for the envelope of ODV; VP39 was the control for nucleocapsid of BV and ODV.
To determine the function of AC73 in virus infection, ac73 knockout and repaired bacmids were constructed. A 168 bp of ac73 in AcMNPV bacmid bMON14272 was replaced with egfp and Cmr genes to generate AcΔ73 bacmid, and then ph alone or both ph and HAtag-fused ac73 genes were inserted into the AcΔ73 bacmid to produce AcΔ73-ph or AcΔ73-ac73R-ph bacmid (Fig. 4A). Then, AcΔ73-ph or AcΔ73-ac73R-ph bacmid was transfected into Sf9 cells. The result showed that the number of fluorescent cells increased obviously from 48 to 96 h p.t. in both AcΔ73-ph and AcΔ73-ac73R-ph transfected cells (Fig. 4B, left two panels), indicating that infectious BVs could be produced when ac73 was deleted. To further confirm this, the supernatants from transfected cells were collected and then used to infect healthy Sf9 cells. Cells were successfully infected with AcΔ73-ph or AcΔ73-ac73R-ph virus as indicated by the occurrence of EGFP fluorescence (Fig. 4B, right panel). To confirm the correctness of recombinant viruses, AcBac-egfp-ph, AcΔ73-ph, or AcΔ73-ac73R-ph virus infected Sf9 cells were analyzed by Western blot. The result showed that the WT AC73 and HAtag-AC73 protein could be detected in AcBac-egfp-ph and AcΔ73-ac73R-ph infected cells, respectively, but no signal could be detected in AcΔ73-ph infected cells (Fig. 4C), suggesting AcΔ73-ph and AcΔ73-ac73R-ph were correctly constructed and produced. Taking together, these results indicated that ac73 is non-essential for BV propagation in cultured cells.
Figure 4. Construction and identification of ac73 knockout and repaired viruses. A Construction of AC73 knockout and repaired bacmids. A fragment of ac73 in AcMNPV bacmid was replaced by egfp and Cmr, and was then inserted with ph or both ph and ac73 to construct AC73 knockout (AcΔ73-ph) or repaired (AcΔ73-ac73R-ph) bacmid. B Transfection and infection assay. Sf9 cells were transfected with the bacmid of AcΔ73-ph or AcΔ73-ac73R-ph and observed at 48 and 96 h p.t. (the left two panels). At 120 h p.t., the supernatants of infected cells were collected and used to infect healthy Sf9 cells, and the infections were detected at 48 h p.i. based on fluorescence (the right panel). Bars, 40 μm. C Detection of AC73 in the recombinant viruses infected cells. Sf9 cells were infected with AcBac-egfp-ph, AcΔ73-ph, or AcΔ73-ac73R-ph virus at an MOI of 10 for 36 h and then collected for Western blot assay. VP39 and GAPDH were used as controls.
To quantify whether ac73 contributes to BV production, one-step growth curve analysis of AcBac-egfp-ph, AcΔ73-ph, or AcΔ73-ac73R-ph was performed. Sf9 cells were infected with these viruses at an MOI of 10, and BV titers at 0, 24, 48, 72, and 96 h p.i. were determined. In contrast to the result of bm59 deletion which did not affect BV production (Hu et al. 2016), one-step growth curve assay revealed the BV titers of AcΔ73-ph decreased by approximately 8- and 5-fold compared to that of AcBac-egfp-ph virus at 72 and 96 h p.i. respectively (P < 0.05) (Fig. 5). By comparison, at all the time points of the infection, the BV titers of AcΔ73-ac73R-ph showed no significant difference with those of AcBac-egfp-ph (P> 0.05) (Fig. 5). Thus, though ac73 is non-essential for BV production, it does play a role in optimal production of infectious BVs.
Figure 5. Deletion of ac73 decreased infectious BV production. Sf9 cells were infected with AcBac-egfp-ph, AcΔ73-ph, or AcΔ73-ac73R-ph virus at an MOI of 10. The supernatants of infected cells were harvested at the indicated time points of infection, and virus titers were determined by endpoint dilution assay for one-step growth curve analysis. The points represent the average titers from triplicate infections and error bars indicate standard deviations (SD).
Next, we determined whether ac73 is essential for the morphogenesis of ODV and OB. To this end, AcBac-egfpph, AcΔ73-ph, and AcΔ73-ac73R-ph infected cells at 24, 48, and 72 h p.i. were subjected to electron microscopy. At 24 h p.i., the nucleocapsids and ODVs could be detected in the nucleus of infected cells for AcBac-egfp-ph and AcΔ73-ac73R-ph, as well as AcΔ73-ph (Fig. 6, upper panel), suggesting the ac73 was neither essential for the nucleocapsid assembly, nor for the envelopment of nucleocapsids to form ODV. At 48 and 72 h p.i., the OBs that embedded with ODVs were formed in AcBac-egfp-ph and AcΔ73-ac73R-ph, as well as in AcΔ73-ph infected cells (Fig. 6, lower two panels). These results showed that ac73 was not required for ODV or OB formation in infected cells.
Figure 6. EM analysis of ODV and OB formation. Sf9 cells were infected with AcBac-egfp-ph, AcΔ73-ph, or AcΔ73-ac73R-ph virus at an MOI of 10 and fixed at indicated time points of infection for EM analysis. The representative ODVs or OBs are indicated by black arrow. N nucleus, C cytoplasm. Bars, 1 μm.
To further investigate the function of AC73 in vivo, bioassay was performed to determine the effects of ac73 deletion on viral infectivity in host level. The early thirdinstar S. exigua larvae were orally infected with OBs of AcBac-egfp-ph, AcΔ73-ph, or AcΔ73-ac73R-ph at different concentrations using droplet method (Hughes et al. 1986). Liquefaction of the infected larvae after death was observed for AcBac-egfp-ph, AcΔ73-ph, and AcΔ73-ac73R-ph viruses (data not shown), indicating that ac73 was not essential for oral infection and liquefaction of infected larvae. In two independent experiments, the potency ratio test showed that there was no significant difference between the AcBac-egfp-ph and AcΔ73-ac73Rph viruses as evidenced by the including of the value 1.0 for 95% CL (Robertson et al. 2007), but the LC50 of AcΔ73-ph virus was 3–4 fold higher than that of AcBac-egfp-ph virus (the potency ratio didn't include 1.0) (Table 1), suggesting that the deletion of ac73 reduced the viral infectivity of AcMNPV in S. exigua larvae. Thus, ac73 is a virulent gene that contributes to virus infection in vivo.
Virus Test 1 Test 2 LC50(95% CL)
(× 104 OBs/mL)
LC50 (95% CL)
(× 104 OBs/mL)
Potency ratioa(95% CL) AcBac-egfp-ph 6.84 (4.35, 10.53) 5.19 (2.77, 9.26) AcΔ73-ph 25.32 (16.60, 38.30) 3.70 (1.96, 7.56) 15.14 (8.71, 25.57) 2.92 (1.312, 7.20) AcΔ73-ac73R-ph 6.71 (4.24, 10.40) 0.98 (0.53, 1.82) 2.79 (1.43, 5.12) 0.538 (0.22, 1.24) aThe potency ratio was calculated by dividing the LC50 of the AcΔ73-ph and AcΔ73-ac73R-ph viruses by that of AcBac-egfp-ph. A significant difference was based on whether the 95% confidence limits (CL) of the potency ratio included the value 1.0.
Table 1. Bioassay results of recombinant viruses against early third-instar S. exigua larvae.