|Home | About | Journals | Submit | Contact Us | Français|
The replication of plus-strand RNA viruses depends on subcellular membranes. Recent genome-wide screens have revealed that the sterol biosynthesis genes ERG25 and ERG4 affected the replication of Tomato bushy stunt virus (TBSV) in a yeast model host. To further our understanding of the role of sterols in TBSV replication, we demonstrate that the downregulation of ERG25 or the inhibition of the activity of Erg25p with an inhibitor (6-amino-2-n-pentylthiobenzothiazole; APB) leads to a 3- to 5-fold reduction in TBSV replication in yeast. In addition, the sterol biosynthesis inhibitor lovastatin reduced TBSV replication by 4-fold, confirming the importance of sterols in viral replication. We also show reduced stability for the p92pol viral replication protein as well as a decrease in the in vitro activity of the tombusvirus replicase when isolated from APB-treated yeast. Moreover, APB treatment inhibits TBSV RNA accumulation in plant protoplasts and in Nicotiana benthamiana leaves. The inhibitory effect of APB on TBSV replication can be complemented by exogenous stigmasterol, the main plant sterol, suggesting that sterols are required for TBSV replication. The silencing of SMO1 and SMO2 genes, which are orthologs of ERG25, in N. benthamiana reduced TBSV RNA accumulation but had a lesser inhibitory effect on the unrelated Tobacco mosaic virus, suggesting that various viruses show different levels of dependence on sterol biosynthesis for their replication.
Plus-stranded RNA [(+)RNA] viruses usurp various intracellular/organellar membranes for their replication. These cellular membranes are thought to facilitate the building of viral factories, promote a high concentration of membrane-bound viral proteins, and provide protection against cellular nucleases and proteases (1, 12, 35, 44). The membrane lipids and proteins may serve as scaffolds for targeting the viral replication proteins or for the assembly of the viral replicase complex. The subcellular membrane also may provide critical lipid or protein cofactors to activate/modulate the function of the viral replicase. Indeed, the formation of spherules, consisting of lipid membranes bended inward and viral replication proteins as well as recruited host proteins, has been demonstrated for several (+)RNA viruses (20, 30, 48). These virus-induced spherules serve as sites of viral replication. Importantly, (+)RNA viruses also induce membrane proliferation that requires new lipid biosynthesis. Therefore, it is not surprising that several genome-wide screens for the identification of host factors affecting (+)RNA virus replication unraveled lipid biosynthesis/metabolism genes (8, 23, 38, 50). However, in spite of these intensive efforts, understanding the roles of various lipids and lipid biosynthesis enzymes and pathways in (+)RNA virus replication is limited.
Tomato bushy stunt virus (TBSV) is among the most advanced model systems regarding the identification of host factors affecting (+)RNA virus replication (32). Among the five proteins encoded by the TBSV genome, only the p33 replication cofactor and the p92pol RNA-dependent RNA polymerase (RdRp) are essential for TBSV RNA replication (55). p33 and p92pol are integral membrane proteins, and they are present on the cytosolic surface of the peroxisomes, the site of replicase complex formation and viral RNA replication (30, 42). Electron microscopic images of cells actively replicating tombusviruses have revealed the extensive remodeling of membranes and indicated active lipid biosynthesis (30, 34).
Additional support for the critical roles of various lipids in TBSV replication comes from a list of 14 host genes involved in lipid biosynthesis/metabolism, which affected tombusvirus replication and recombination based on systematic genome-wide screens in yeast, a model host. These screens covered 95% of the host genes (16, 38, 50, 51). The 14 identified host genes involved in lipid biosynthesis/metabolism included 8 genes affecting phospholipid biosynthesis, 4 genes affecting fatty acid biosynthesis/metabolism, and 2 genes affecting ergosterol synthesis. These findings suggest that these lipids likely are involved, directly or indirectly, in TBSV replication in yeast.
To further understand the roles of cellular membranes, lipids, and host factors in viral (+)RNA replication, we analyzed the importance of sterol biosynthesis in tombusvirus replication. Sterols are ubiquitous and essential membrane components in all eukaryotes, affecting many membrane functions. Sterols regulate membrane rigidity, fluidity, and permeability by interacting with other lipids and proteins within the membranes (4, 5). They also are important for the organization of detergent-resistant microdomains, called lipid rafts (45). The sterol biosynthesis differs in several steps in animals, fungi, and plants, but the removal of two methyl groups at the C-4 position is critical and rate limiting. The C-4 demethylation steps are performed by SMO1 (sterol4α-methyl-oxidase) and SMO2 in plants and by the orthologous ERG25 gene in yeast (10). Accordingly, erg25 mutant yeast accumulates 4,4-dimethylzymosterol, an intermediate in the sterol biosynthesis pathway (3). However, sterol molecules become functional structural components of membranes only after the removal of the two methyl groups at C-4. Therefore, ERG25 is an essential gene for yeast growth.
Our previous genome-wide screens for factors affecting tombusvirus replication have identified two sterol synthesis genes, ERG25 and ERG4, that participate in different steps in the sterol biosynthesis pathway (11). In this work, we further characterized the importance of ERG25 in TBSV replication in yeast. The downregulation or pharmacological inhibition of ERG25 in yeast led to a 4- to 5-fold decreased TBSV RNA accumulation. The in vitro activity of the tombusvirus replicase was reduced when isolated from the yeast cells described above. We also found that the stability of p92pol viral replication protein decreased by 3-fold in yeast treated with a chemical inhibitor of ERG25. The inhibition of sterol biosynthesis in plant protoplasts or in plant leaves with a chemical inhibitor or the silencing of SMO1 and SMO2 genes also resulted in a reduction in TBSV RNA accumulation, supporting the roles of sterols in tombusvirus replication in plants as well.
Saccharomyces cerevisiae strain BY4741 (MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0) and the single-gene deletion strain Δpex19, as well as the ERG25/THC (BY4741; URA3::CMV-tTA) strain with the regulatable TET promoter from the Hughes collection, were obtained from Open Biosystems (Huntville, AL). The following yeast expression plasmids have been generated before: pGAD-His92 (41), pGBK-His33/DI-72 (16), pGBK-His33/CUP1 (15), pGAD-His92CUP1 (25), and pYC/DI72 (37).
To generate pCM189-tetDI72, the DI-72 sequence was PCR amplified using the primers 1803 (GGCGAGATCTGGAAATTCTCCAGGATTTCTC) and 3176 (CGGTCAAGCTTTACCAGGTAATATACCACAACGTGTGT), with pYC/DI72 as a template. The obtained PCR product was purified and digested with HindIII and BglII, followed by ligation into the HindIII- and BamHI-digested vector pCM189.
The ERG25/THC strain transformed with pGAD-His92 and pGBKHis33/DI-72 plasmids was pregrown in a synthetic complete dropout medium lacking leucine and histidine (SC-LH− medium) containing 2% glucose and then cultured for 24 h at 29°C until an optical density at 600 nm (OD600) of ~0.8 to 1.0 in SC-LH− medium containing 2% galactose. For the maximum level of ERG25 gene expression, yeast was grown in the absence of doxycycline, whereas to reduce the expression level of the ERG25, yeast was grown in the same medium in the presence of 10 mg/liter doxycycline (16).
To study the effect of 6-amino-2-n-pentylthiobenzothiazole (APB) on TBSV repRNA accumulation, we used different APB concentrations to treat both BY4741 (wild type) and ERG25/THC strains. Briefly, the above-described strains were transformed with pGAD-His92 and pGBK-His33/DI-72 plasmids and pregrown overnight at 29°C with or without APB in SC-LH− medium containing 2% glucose. The replication of TBSV repRNA was induced by transferring the culture to SC-LH− medium containing 2% galactose (37) followed by being grown at 29°C. After 24 h, cells were used to obtain total RNA extracts.
To do the time course study with APB, the ERG25/THC strain carrying pGAD-His92 and pGBK-His33/DI-72 plasmids was grown with 30 μM APB added at different time points. Briefly, yeast was pregrown overnight at 29°C in 6 different batches (treatments 1 to 6) in SC-LH− medium containing 2% glucose. TBSV repRNA was expressed by transferring the culture to SC-LH− medium containing 2% galactose at 29°C. The APB inhibitor was removed, and the pellet was washed with the growth medium at the beginning of repRNA induction or after a 12-h treatment (treatments 4 and 5, respectively), followed by culturing as described above.
To study the effect of lovastatin on TBSV DI-72 repRNA accumulation in yeast, we prepared a stock solution of lovastatin (Sigma) according to reference 27. In our studies, we used 20 μg/ml (final concentration) of the inhibitor or 0.1% of the solvent in mock-treated samples. BY4741 yeast was cotransformed with plasmids pGBK-His33, pGAD-His92, and pYC/DI72 and pregrown overnight at 29°C with 20 μg/ml of lovastatin or 0.1% of solvent in synthetic complete dropout medium lacking uracil, leucine, and histidine (SC-ULH− medium) containing 2% glucose. TBSV repRNA replication was started by transferring the yeast culture to SC-ULH− medium containing 2% galactose along with inhibitor. After 24 h of culturing at 29°C, cells were collected and processed for RNA isolation to analyze repRNA accumulation as described below.
To obtain the membrane-enriched (ME) fraction, containing the assembled replicase complex with the repRNA template, from the ERG25/THC strain, yeast transformed with pGAD-His92 and pGBK-His33/DI-72 plasmids was grown for 24 h at 29°C in SC-LH− medium containing 2% galactose with or without doxycycline (10 mg/liter) until reaching an OD600 of ~0.8 to 1.0. To collect the ME fraction from the APB-treated yeast, strain BY4741 was transformed with plasmids pGBK-His33/CUP1 and pGAD-His92/CUP1 expressing 6×His-tagged cucumber necrosis virus (CNV) p33 and 6×His-tagged p92, respectively, from the inducible CUP1 promoter, as well as pCM189tetDI72 expressing DI-72 repRNA from the doxycycline-repressible TET promoter. Briefly, yeast cells were grown overnight with or without APB in SC-ULH− medium containing 2% glucose at 29°C. repRNA replication then was induced with 50 μM copper sulfate for 24 h at 23°C with shaking at 250 rpm. After 24 h of growth, yeast samples were collected at an OD600 of ~0.8 by centrifugation at 1,100 × g for 5 min. The procedure used to obtain the ME fractions from the above-described strains was described earlier (15, 40, 41). Briefly, the yeast pellet was washed with 20 mM Tris-HCl, pH 8.0, and resuspended in 1 ml of 20 mM Tris-HCl, pH 8.0, followed by centrifugation at 21,000 × g for 1 min. Yeast cells were broken by glass beads in a Genogrinder (Glen Mills Inc., Clifton, NJ) for 2 min at 1,500 rpm. After being mixed with 600 μl chilled extraction buffer (200 mM sorbitol, 50 mM Tris-HCl [pH 7.5], 15 mM MgCl2, 10 mM KCl, 10 mM β-mercaptoethanol, yeast protease inhibitor mix; Sigma), the samples were centrifuged at 100 × g for 5 min at 4°C. The supernatant was moved to a new microcentrifuge tube and centrifuged at 21,000 × g for 10 min at 4°C. The pellet was resuspended in 0.7 ml extraction buffer, resulting in the ME fraction. The replicase assay with the ME fraction was performed in a 100-μl volume containing RdRp buffer [40 mM Tris, pH 8.0, 10 mM MgCl2, 10 mM dithiothreitol, 100 mM potassium glutamate, 0.2 μl RNase inhibitor, 1 mM ATP, CTP, GTP, 0.3 μl radioactive [32P]UTP (800 mCi/mmol; ICN)] and a 50-μl ME fraction. Samples were incubated at 25°C for 2 h. The reaction was terminated by adding 70 μl SDS-EDTA (1% SDS, 50 mM EDTA, pH 8.0) and 100 μl phenol-chloroform (1:1). After the isopropanol precipitation of the RNA products, the RNA samples were electrophoresed under denaturing conditions (5% PAGE containing 8 M urea) and analyzed by phosphorimaging using a Typhoon (GE) instrument as described previously (15, 40, 41).
Total RNA isolation from yeast and Northern blot analyses of the accumulation of TBSV repRNA were performed as described previously (37, 41). Western blotting for measuring p33/p92 levels was performed using anti-His antibody, whereas the secondary antibody was alkaline phosphatase-conjugated anti-mouse immunoglobulin G (Sigma), as described previously (49).
To test the level of ERG25 mRNA expression, the ERG25/THC yeast strain was grown for 12 h in YPD medium at 29°C with shaking at 250 rpm. Doxycycline was added, and samples were collected at 0-, 5-, 11-, 24-h time points. After the cells were pelleted, total RNA was extracted by using a modified hot-phenol method (as described above). For the Northern blot analysis, the total RNA samples were diluted 100 times (except for the detection of ERG25 mRNA, for which undiluted samples were used) before electrophoresis, followed by the transfer of RNA to membranes. The 32P-labeled RNA probes were prepared by in vitro transcription with T7 RNA polymerase from appropriate PCR products. The PCR product used to transcribe the labeled 18S ribosomal RNA (rRNA) probe was amplified from the yeast genome with primers 1251 (GGTGGAGTGATTTGTCTGCTT) and 1252 (TAATACGACTCACTATAGGTTTGTCCAAATTCTCCGCTCT). The template for the probe to detect ERG25 mRNA was obtained by PCR with primers 2793 (GCCGGATCCATGTCTGCCGTTTTCAACAAC) and 2794 (GTAATACGAGTCACTATAGGGAGATAGAAGAACGGATTTCAAAC) using yeast genomic DNA as the template.
Yeast strain BY4741 was transformed with either pGBK-His33/CUP1 or pGAD-His92/CUP1 expressing 6×His-tagged CNV p33 and 6×His-tagged p92, respectively, from the inducible CUP1 promoter. Yeast transformants were cultured overnight in SC H− medium (or SC L− medium for p92 expression) containing 2% glucose with or without 30 μM APB at 29°C. To induce the expression of p33 or p92, 50 μM CuSO4 was added to the yeast cultures for 30 min at 29°C, followed by the addition of cycloheximide to a final concentration of 100 μg/ml to inhibit protein synthesis. Equal amounts of yeast cells were collected at given time points as indicated in the figure legends after cycloheximide treatment (26), and cell lysates were prepared by the NaOH method as described previously (37). The total protein samples were analyzed by SDS-PAGE and Western blotting with anti-His antibody using ECL (Amersham) as described previously (15).
The in vitro translation reaction mixtures were based on a wheat germ extract containing 1 mM amino acid mixture without methionine (Promega) and [35S]methionine (10 mCi/ml). We added 0.3 pmol of p33 or p92 mRNAs to program the assay (modified from reference 26), while APB was added at a 30 μM final concentration. After 1 h of incubation at room temperature, samples were mixed with SDS-PAGE loading dye and incubated at 100°C for 2 min and electrophoresed in an SDS-PAGE gel followed by phosphorimaging as described previously (26).
The preparation of Nicotiana benthamiana protoplasts, electroporation with TBSV and CNV RNA, as well as viral RNA analysis were performed as described previously (39). APB was dissolved in dimethyl sulfoxide (DMSO) and was added at a concentration of 60 μM (or as indicated in the figure legends) before or after electroporation to the N. benthamiana protoplasts. Campesterol (4 mg/ml) and stigmasterol (4 mg/ml) (from Steraloids Inc., Newport, RI) were prepared using a 1:1 mixture of DMSO and 95% ethanol and then added to protoplast preparations at 10, 20, or 40 μg/ml before or after electroporation.
APB (1,500 μM) or DMSO (1.5% as a control) was used for infiltration into N. benthamiana leaves, followed by the inoculation of the same leaves with inoculum containing TBSV, CNV, or TMV. Total RNA extraction from the inoculated leaves was done at 4 dpi (7).
pTRV1 and pTRV2 plasmids to launch Tobacco rattle virus (TRV) infection through agroinfiltration were kindly provided by S. P. Dinesh-Kumar (13). To amplify partial SMO1 and SMO2 cDNA fragments from N. benthamiana, primers were designed according to consensus regions of aligned SMO1 and SMO2 sequences (10). To generate plasmid pTRV2-SMO1, a 387-bp fragment of N. benthamiana SMO1 was amplified by reverse transcription-PCR (RT-PCR) with primers 2901 (GGCGGAATTCACAAGTTTGCCCCTGCCGTC) and 2902 (GGCGCTCGAGAACATAGTGATGGTAGTCATGGTAATC) (10) using total RNA extract obtained from uninfected N. benthamiana leaves. The obtained RT-PCR product was treated with EcoRI and XhoI and then was cloned at the EcoRI and XhoI sites of pTRV2 vector. Construct pTRV2-SMO2 carrying 450 bp of N. benthamiana SMO2 sequence was generated as described for pTRV2-SMO1 except using primers 2903 (GGCGGAATTCATGGCTTCCATGATCGAATCTGCTTGG) and 2904 (GGCGCTCGAGGACAAGAAAAAGAATTTCAGCAGGGTGAGC) (10).
The VIGS assay to silence SMO1 and SMO2 in N. benthamiana was performed as described previously (53). Eleven days after the agroinfiltration of pTRV2-SMO1 and pTRV2-SMO2, the accumulation levels for N. benthamiana SMO1 and SMO2 mRNAs in N. benthamiana were determined by semiquantitative RT-PCR with primers 3273 (CCTAATCTTCTCTTGTGTCCCTC) and 2902 and the primer pair of 3274 (CAATTGACTTGTCTTGGTGGGTTT) and 3275 (CAAGGACTCTAAGTGAGACCC), respectively (10). The accumulation level of the control tubulin mRNA was measured by semiquantitative RT-PCR using primers 2859 (TAATACGACTCACTATAGGAACCAAATCATTCATGTTGCTCTC) and 2860 (TAGTGTATGTGATATCCCACCAA). The obtained RT-PCR products were sequenced to confirm their identities. The VIGS-silenced leaves were sap inoculated with TBSV or TMV as described previously (53). Total RNA extracts from the infected leaves were prepared at 3 dpi (days postinoculation) and from systemically infected leaves at 5 dpi to analyze TBSV and TRV accumulation.
To confirm the role of ERG25 in TBSV replication, we used the ERG25/THC (TET::ERG25) strain from the yTHC collection (Open Biosystems). The expression of the essential ERG25 gene is under the control of a doxycycline-titratable promoter in the yeast genome in the ERG25/THC strain. Therefore, the expression of the ERG25 gene can be downregulated/turned off by the addition of doxycycline to the yeast growth medium (31). This approach allowed us to test tombusvirus RNA replication in the presence of high levels of Erg25p (when yeast was grown without added doxycycline after the induction of tombusvirus repRNA replication) or various reduced levels of Erg25p (when yeast was grown in the presence of doxycycline) (31).
To test the replication of the TBSV repRNA, which is an efficiently replicating surrogate RNA template derived from the TBSV genomic RNA (gRNA) (37, 54), we expressed the p33 and p92pol replication proteins and DI-72 repRNA from plasmids in the ERG25/THC yeast strain. These experiments revealed the efficient replication of DI-72 repRNA in the ERG25/THC strain grown under standard growth conditions without added doxycycline (see Materials and Methods) (Fig. (Fig.1A,1A, lanes 1 and 3). The addition of doxycycline to the growth medium led to a rapid decrease in ERG25 mRNA levels at the beginning of repRNA replication (5-h time point) (Fig. (Fig.1B,1B, lane 5). The level of ERG25 mRNA was undetectable by Northern blotting 11 h after the addition of doxycycline (Fig. (Fig.1B,1B, lane 6) as well as at the 24-h time point (Fig. (Fig.1B,1B, lane 7), when the samples for TBSV repRNA analysis were collected (Fig. (Fig.1A).1A). The accumulation level of DI-72 repRNA was ~3-fold lower in ERG25/THC yeast lacking detectable levels of ERG25 mRNA (Fig. (Fig.1A,1A, lane 4), suggesting that the high expression level of ERG25 promotes TBSV repRNA replication in yeast.
To test if Erg25p affected the tombusvirus replicase activity, we generated enriched tombusvirus replicase preparations containing the copurified repRNAs from ERG25/THC yeast grown in the absence or presence of doxycycline. Using comparable amounts of p33 replication protein during the assay (Fig. (Fig.1C,1C, lower), we found that the tombusvirus replicase preparation obtained from yeast with an undetectable level of ERG25 mRNA was ~5-fold less active than the comparable replicase preparation obtained from yeast grown in the absence of doxycycline (Fig. (Fig.1C,1C, compare lanes 5 to 8 to lanes 1 to 4). These experiments demonstrated that the expression of Erg25p in yeast is important for the activity of the tombusvirus replicase.
To further test the role of ERG25 in TBSV replication, we took advantage of a potent chemical inhibitor, namely, 6-amino-2-n-pentylthiobenzothiazole (APB). APB has been shown to bind to Erg25p and inhibits its catalytic function in a competitive manner (9, 21). The treatment of yeast with APB inhibited the biosynthesis of ergosterol and led to the accumulation of the methylated sterol precursors, such as ergosta-5,7-dienol and squalene, but had no significant effect on the composition and the rate of biosynthesis of fatty acids (22). We found that applying increasing amounts of APB to the growth medium affected both yeast growth and TBSV accumulation (Fig. (Fig.2A).2A). The normalization of TBSV repRNA accumulation based on yeast rRNA levels indicated that APB, when applied between at concentrations of 20 and 40 μM, inhibited TBSV repRNA accumulation four to five times more than it inhibited yeast growth (Fig. (Fig.2A).2A). This finding was confirmed in both ERG25/THC (TET::ERG25) (Fig. (Fig.2A)2A) and wild-type BY4741 yeast strains (Fig. (Fig.2B2B).
TBSV replicates on the peroxisomal membrane surface in yeast and in plant cells (17, 30, 36, 42). However, TBSV replication can shift to the endoplasmic reticulum (ER) membrane in the absence of peroxisome, which results in replication as efficient as that occurring on the peroxisomal membrane surface (17, 42). To test if the APB-driven inhibition of sterol biosynthesis also could reduce TBSV replication occurring on the ER membrane, we used the pex19Δ yeast strain, which lacks peroxisomes (17, 42). The accumulation of the TBSV repRNA decreased to ~20% in the 30 μM APB-treated pex19Δ yeast strain, a level of inhibition comparable to that in the wild-type BY4741 strain (Fig. (Fig.2B,2B, compare lanes 10 to 12 to lanes 4 to 6). Therefore, TBSV replication seems to be dependent on sterol biosynthesis equally when it takes place on the peroxisomal membrane surface and on the ER membrane.
Since the downregulation of the ERG25 mRNA level (Fig. (Fig.1)1) or the use of the APB inhibitor of Erg25p is known to lead to the accumulation of methylated sterol precursors, it is possible that these compounds directly inhibit TBSV repRNA replication instead of the lack of sterols in the above-mentioned yeast cells, resulting in the reduction of viral replication. To test this possibility, we also applied lovastatin, which is a potent inhibitor of hydroxymethylglutaryl-coenzyme A (HMG-CoA) reductase (coded by the HMG1 and HMG2 genes in yeast), a rate-limiting enzyme in the mevalonate pathway that regulates cholesterol synthesis. Lovastatin acts as a competitive inhibitor of HMG-CoA reductase, effectively lowering sterol levels in yeast (11, 27). The application of lovastatin reduced TBSV repRNA accumulation in yeast by 4-fold (Fig. (Fig.2C,2C, lanes 5 to 8) without affecting the p33 level. The strong inhibitory effect of lovastatin on TBSV replication is not compatible with the model that the accumulation of methylated sterol precursors in yeast with downregulated ERG25 mRNA levels is responsible for the direct inhibition of TBSV repRNA replication.
Previous electron microscopic images indicated that tombusvirus replication likely utilizes the preexisting peroxisomal membrane surfaces in the infected cells at the beginning of infection, followed by the induction of new membranes and possibly utilizing other membranes than the peroxisomal membrane at late time points of infection (34). To test when TBSV is the most sensitive to the inhibition of sterol biosynthesis, we devised a scheme for the time-restricted inhibition of sterol biosynthesis via treatment with APB, followed by removing the inhibitor at given time points (Fig. (Fig.3A3A).
Surprisingly, the inhibition of sterol biosynthesis only during the period of TBSV replication was as effective as the inhibition of sterol biosynthesis prior to and during TBSV replication (treatment 2 versus treatment 3) (Fig. 3B and C, compare lanes 5 to 8 to lanes 9 to 12). Also, the treatment of yeast cells with APB inhibitor only prior to the beginning of TBSV replication had a limited inhibitory effect on TBSV accumulation (treatment 4) (Fig. 3B and C, lanes 13 to 16). These data suggest that the preexisting sterol level is not as critical as the newly synthesized sterols during TBSV replication.
To further define when sterol biosynthesis is most critical during TBSV replication, we used shorter treatments with APB. These experiments revealed that sterol biosynthesis is most critical during the beginning of TBSV replication (between 1 and 12 h; treatment 5) (Fig. 3B and C, lanes 17 to 20) and less effective at the late time point (between 12 and 24 h; treatment 6) (Fig. 3B and C, lanes 21 to 24). These data indicate that TBSV replication requires new sterol biosynthesis the most at an early stage of infection.
In the experiments shown in Fig. Fig.1,1, we expressed the p33 and p92pol replication proteins from the constitutive ADH promoter, which could allow some limited assembly of the replicase complex prior to the downregulation of ERG25 levels. Therefore, we have retested the replicase activity when the p33 and p92pol replication proteins were expressed from the inducible CUP1 promoter to allow the replicase assembly to take place only when ergosterol biosynthesis was inhibited (Fig. (Fig.4).4). Testing the activity of tombusvirus replicase preparations obtained from APB-treated ERG25/THC yeast grown in the absence of doxycycline revealed that a 40 μM concentration of APB inhibited the replicase activity by ~90% at a late time point (24 h) when we adjusted the preparations to contain comparable amounts of p33 replication protein (Fig. (Fig.4A,4A, top, compare lanes 4 to 6 to lanes 1 to 3). Interestingly, the amount of p92pol was lower in the replicase samples obtained from the APB-treated than from the untreated yeast (Fig. (Fig.4A,4A, middle, compare lanes 4 to 6 to lanes 1 to 3). These experiments demonstrated that the APB treatment of yeast could inhibit the in vitro activity of the tombusvirus replicase and decrease the level of p92pol.
To test the stability of p33 and p92pol replication proteins when sterol biosynthesis is inhibited, we treated yeast cells first with APB and then with cycloheximide (to inhibit new protein synthesis), followed by measuring protein levels (Fig. 4B and C). We found that the half-life of p33 did not change significantly (Fig. (Fig.4C),4C), while that of p92pol was reduced by ~3-fold in APB-treated yeast compared to that of the DMSO-treated control yeast (based on the reduction of the half-life of p92 from ~100 to ~30 min) (Fig. (Fig.4B).4B). Interestingly, APB treatment did not affect the translation of p33 or p92pol in vitro (Fig. (Fig.4D),4D), suggesting that the degradation of p92pol is accelerated in APB-treated yeast, leading to reduced stability selectively for p92pol but not for p33.
The above-described experiments demonstrated that the sterol level is important for TBSV repRNA accumulation in yeast cells and in vitro. To test if sterol biosynthesis also is important for TBSV RNA accumulation in plant cells, we inhibited sterol biosynthesis by APB treatment based on the conserved function of Erg25p protein in yeast and Smo1 and Smo2 proteins in plants (9, 10). We found that 40 μM APB applied either before (Fig. (Fig.5A,5A, lanes 2 and 5) or after the electroporation (not shown) of TBSV gRNA into N. benthamiana protoplasts inhibited TBSV gRNA as well as subgenomic RNA1 (sgRNA1) and sgRNA2 accumulation by ~3-fold. Time course experiments showed that treatment with APB that started at the zero time point (Fig. (Fig.5B,5B, lanes 6 and 7) was more effective than treatments starting from 3 or 6 h postelectroporation. These experiments indicated that sterol biosynthesis also is important in plant cells for TBSV RNA accumulation, especially at the early time point.
To test if the negative effect on TBSV RNA accumulation by the APB-mediated inhibition of sterol biosynthesis could be complemented by the addition of phytosterols to the growth medium of plant protoplasts, we used stigmasterol (the major phytosterol in plant cell membranes) and campesterol in various concentrations prior to or after the electroporation of the TBSV gRNA. Interestingly, we found that 20 μg/ml stigmasterol (Fig. (Fig.6A,6A, lane 9) or 10 μg/ml stigmasterol (not shown) applied in combination with 60 μM APB before the electroporation of TBSV gRNA into N. benthamiana protoplasts increased TBSV gRNA accumulation from 27% (APB treatment) (Fig. (Fig.6A,6A, lanes 5 to 6) to 91% (20 μg/ml stigmasterol treatment) (Fig. (Fig.6A,6A, lane 9) and 75% (10 μg/ml stigmasterol treatment; not shown). The complementation was less pronounced with 40 μg/ml campesterol (Fig. (Fig.6B,6B, lane 10), which resulted in 52% TBSV gRNA accumulation, up from 25% (APB treatment) (Fig. (Fig.6B,6B, lanes 5 to 6) in N. benthamiana protoplasts. Overall, these experiments demonstrated that phytosterols, especially stigmasterol, could complement the inhibitory effect of APB treatment on TBSV gRNA replication in plant protoplasts.
To demonstrate that the inhibition of sterol biosynthesis is important for TBSV RNA accumulation in plants, we infiltrated leaves of N. benthamiana with 60 to 1,500 μM APB prior to inoculation with infectious TBSV virion preparations. The isolation of total RNA from the inoculated leaves 4 days after inoculation, followed by Northern blotting, revealed that APB treatment, when applied at 1,500 μM, inhibited TBSV RNA accumulation by ~90% (Fig. (Fig.7A,7A, lanes 11 to 17). The similar treatment with APB of N. benthamiana prior to inoculation with infectious CNV (a close relative of TBSV) virion preparations reduced CNV gRNA accumulation to below the detection limit at 4 dpi (Fig. (Fig.7C,7C, lanes 10 to 21). Similarly to TBSV, the CNV-infected and APB-treated plants showed a delay in symptom development in systemically infected leaves compared to that of the DMSO-treated and CNV or TBSV-inoculated plants (Fig. 7B and D). In contrast, APB-treated leaves supported TMV (an unrelated plus-strand RNA virus of the alphavirus supergroup) RNA accumulation almost as efficiently as the DMSO-treated leaves (Fig. (Fig.7E).7E). As expected, there was no delay in symptom appearance in APB- or DMSO-treated plants infected with TMV (Fig. (Fig.7F).7F). The APB treatment had no obvious effect on the leaves or the whole plants (Fig. (Fig.7F).7F). These experiments demonstrated that APB treatment could lead to the significant reduction of TBSV and CNV RNA accumulation in the treated leaves and a delay in symptom development, whereas similar treatments had no significant effect on TMV RNA accumulation and did not delay symptom development in TMV-infected plants. Thus, the requirement for sterol biosynthesis seems to be different for tombusviruses and TMV.
The two orthologs of yeast ERG25 gene in plants are the SMO1 and SMO2 genes, which are 4α-methyl oxidases involved in phytosterol biosynthesis (10). The SMO1 and SMO2 genes are involved in different steps of phytosterol biosynthesis, and they show only ~50% sequence identity, allowing for the separate silencing of these genes (10). Accordingly, we silenced individually or in combination the expression of SMO1 and SMO2 genes by using a VIGS strategy in N. benthamiana. Indeed, RT-PCR analysis has shown decreased levels of SMO1 and SMO2 mRNAs 11 days after the infiltration of Agrobacterium carrying the VIGS constructs (Fig. (Fig.8G,8G, lanes 4 to 6 and 10 to 12). The inoculation of the SMO1- and SMO2-silenced leaves with TBSV led to an ~3-fold reduced TBSV gRNA accumulation in the inoculated leaves compared to that of nonsilenced leaves (plants infiltrated with Agrobacterium carrying the pTRV empty vector) (Fig. (Fig.8A,8A, compare lanes 8 to 14 to lanes 1 to 7, and C, compare lanes 8 to 14 to lanes 1 to 7). The silencing of both SMO1 and SMO2 mRNAs led to even further reductions in TBSV RNA accumulation in the inoculated leaves, reaching only ~10% of the TBSV RNA level in the nonsilenced plants (Fig. (Fig.8E,8E, compare lanes 8 to 14 to lanes 1 to 7). The development of the TBSV symptoms in plants silenced for SMO1, SMO2, and both SMO1 and SMO2 was significantly delayed compared to that of pTRV-treated plants (Fig. 8B, D, and F). However, the silencing of the SMO1, SMO2, and combined SMO1 and SMO2 mRNAs did not lead to the complete protection of the plants from TBSV infections, as shown by the appearance of systemic symptoms (Fig. 8B, D, and F) and the accumulation of TBSV gRNA in systemically infected upper leaves (not shown).
On the contrary to the results described above, the SMO1- and SMO2-silenced leaves accumulated TMV RNAs almost as efficiently in the inoculated leaves as in the nonsilenced leaves (~60 to 74%) (Fig. (Fig.8I).8I). Also, symptom development was comparable in the silenced versus nonsilenced N. benthamiana plants inoculated with TMV (Fig. 8J and L). The systemically infected leaves supported TMV RNA accumulation in the silenced plants as efficiently as in the nonsilenced plants (Fig. (Fig.8K).8K). The cosilencing of both SMO1 and SMO2 genes in N. benthamiana plants also had no detectable effect on TMV accumulation (Fig. 8M, N, and O). These results show that the inhibition of sterol biosynthesis by the silencing of SMO1 and SMO2 genes specifically reduces tombusvirus replication, but the effect on TMV accumulation is weaker in the silenced plants. In addition, the results described above make it unlikely that the SMO1- and SMO2-silenced plants inhibit tombusvirus replication due to nonspecific effects (such as the sick plant phenotype-based general inhibition of virus accumulation), since these plants are compatible for supporting TMV replication. Moreover, it seems that various plant viruses show different levels of dependence on sterol biosynthesis for their replication.
Since the replicase complexes of (+)RNA viruses are membrane bound, the lipid composition of membranes influencing membrane fluidity, rigidity, and permeability is expected to affect the activity of the viral replicase. Accordingly, we demonstrate that the downregulation of Erg25p, a critical enzyme in the sterol biosynthesis pathway (3), or the inhibition of the activity of Erg25p by APB (9) reduced TBSV repRNA replication in yeast by 3- to 5-fold. Moreover, the inhibition of sterol biosynthesis by lovastatin also resulted in a 4-fold reduction in TBSV repRNA replication in yeast. In addition, a previous genome-wide screen for the identification of host factors for TBSV revealed reduced TBSV repRNA accumulation in erg4Δ yeast (38), suggesting that the sterol biosynthesis pathway is required for TBSV replication.
The reduced level of sterols might have a direct inhibitory effect on the tombusvirus replicase activity, since the isolated replicase complex with the copurified repRNA showed an ~5-fold reduced activity in vitro when obtained from yeast with downregulated Erg25p or treated with APB inhibitor. Interestingly, similar levels of the inhibition of TBSV RNA accumulation by APB treatment were observed in yeast lacking peroxisomal membranes, in which TBSV replication occurs on the ER membrane (Fig. (Fig.2B).2B). These data suggest that TBSV replication is greatly affected by sterols in yeast regardless of the subcellular location of the replicase complexes.
The pretreatment of yeast cells with APB had only a minor inhibitory effect on TBSV repRNA replication (Fig. 3B and C, treatment 4), while APB treatment after the induction of TBSV replication had a larger inhibitory effect, especially when applied in the first 12 h (Fig. 3B and C, treatment 5). These data suggest that TBSV replication mostly depends on the newly synthesized sterols in yeast, while the inhibition of sterol biosynthesis prior to TBSV replication to reduce the level of preexisting sterols in the cellular membranes had only a minor inhibitory effect. Similarly, we found that APB treatment was most effective in N. benthamiana protoplasts when applied from the beginning of TBSV RNA replication (Fig. (Fig.5).5). Based on these data, it is possible that sterols regulate TBSV replication not only by affecting the structure and features of the subcellular membranes supporting TBSV replication but also by playing additional roles during TBSV replication.
Although the functions of sterols during TBSV replication are not yet known, it seems that sterols are needed for the stability of p92pol in yeast (Fig. (Fig.4).4). It is possible that the bulky p92pol replicase protein is exposed more to cytosolic proteases in sterol-poor microenvironments or the structure of p92pol is different under sterol-depleted conditions, leading to the premature degradation of p92pol. It also is possible that the subcellular localization of p92pol is different if less than the normal level of sterols was available in cells.
Although the sterols synthesized in yeast (ergosterol) and in plants (phytosterols, of which the sterol stigmasterol is the most abundant) are different, they might play comparable roles in TBSV replication. Accordingly, the downregulation of ERG25 expression in yeast or silencing the orthologous SMO1 and SMO2 genes in N. benthamiana or the APB treatment of yeast and plant cells had comparable inhibitory effects on TBSV RNA accumulation. Also, the negative effect of the APB treatment on TBSV RNA accumulation could be complemented in plant protoplasts by exogenous stigmasterol, strongly suggesting that sterols are the active compounds that affect TBSV RNA replication. It is intriguing that TBSV replication can take advantage of different sterols in yeast and in plants, suggesting high flexibility for TBSV in different subcellular environments.
Interestingly, sterols seem to be needed for tombusvirus replication but are less critical for TMV replication based on silencing SMO1 and SMO2 genes in N. benthamiana or the APB treatment of plant leaves. This different effect could be due to (i) different subcellular compartments (tonoplast/vacuole and peroxisome, respectively) where TMV and TBSV replicate, and (ii) different features of the replication proteins or their abilities to bind to sterols. For example, the TMV replication proteins likely are peripheral membrane proteins (14, 19), while TBSV replication proteins are integral membrane proteins (30, 33, 36, 41).
Similarly to tombusviruses, the replication of other viruses, such as Dengue virus, Norwalk virus, and hepatitis C virus (HCV), also depends on sterols (6, 18, 46, 47). It has been shown that the HCV replicase complex is associated with cholesterol-rich lipid rafts (2). Infection with West Nile virus has been demonstrated to lead to the redistribution of cholesterol to the sites of virus replication, possibly from the plasma membrane, and results in reduced antiviral responses (28). Cholesterol also is important for animal virus entry into cells, infections, and the exit of virus particles from cells (24, 29, 43, 52). These findings invite further studies of dissecting the functional and mechanistic roles of sterols during virus infections. This could lead to the development of novel, broad-range antiviral strategies for animals and plants.
We thank Daniel Barajas, Muhammad Shah Nawaz Khan, Zhenghe Li, and Judit Pogany for valuable comments. T. Kuchta's (Bratislava, Slovakia) gift of APB is highly appreciated.
This work was supported by the National Science Foundation (IOB-0517218), NIH-NIAID (5R21AI072170-02), and the Kentucky Tobacco Research and Development Center (P.D.N.).
Published ahead of print on 16 December 2009.