Mutation of host DnaJ homolog inhibits brome mosaic virus negative-strand RNA synthesis

Abstract

The replication of positive-strand RNA viruses involves not only viral proteins but also multiple cellular proteins and intracellular membranes. In both plant cells and the yeast Saccharomyces cerevisiae, brome mosaic virus (BMV), a member of the alphavirus-like superfamily, replicates its RNA in endoplasmic reticulum (ER)-associated complexes containing viral 1a and 2a proteins. Prior to negative-strand RNA synthesis, 1a localizes to ER membranes and recruits both positive-strand BMV RNA templates and the polymerase-like 2a protein to ER membranes. Here, we show that BMV RNA replication in S. cerevisiae is markedly inhibited by a mutation in the host YDJ1 gene, which encodes a chaperone Ydj1p related to Escherichia coli DnaJ. In the ydj1 mutant, negative-strand RNA accumulation was inhibited even though 1a protein associated with membranes and the positive-strand RNA3 replication template and 2a protein were recruited to membranes as in wild-type cells. In addition, we found that in ydj1 mutant cells but not wild-type cells, a fraction of 2a protein accumulated in a membrane-free but insoluble, rapidly sedimenting form. These and other results show that Ydj1p is involved in forming BMV replication complexes active in negative-strand RNA synthesis and suggest that a chaperone system involving Ydj1p participates in 2a protein folding or assembly into the active replication complex.

FIG. 1.

Schematic diagram of DNA-based expression of BMV RNA3 replicons and subsequent BMV-directed, RNA-dependent RNA replication and subgenomic mRNA synthesis. The cDNA-based RNA3 launching cassette includes BMV NCRs, the 3a gene, an intergenic replication enhancer (RE), a 5′-end-flanking GAL1 promoter, and 3′-end-flanking ribozymes (Rz) from hepatitis delta virus or hammerhead ribozyme. X represents the BMV coat protein gene or its replacements, the coding sequences for GUS or Ura3p. Upon galactose induction, cellular RNA polymerase II-dependent transcription synthesizes positive-strand RNA3 transcripts that serve as the templates for 1a- and 2a-dependent RNA3 replication and synthesis of subgenomic mRNA (RNA4) required for expression of the coat gene or its replacements.

FIG. 2.

Cloning of the MAB3/YDJ1 gene. A region of S. cerevisiae chromosome XIV corresponding to base coordinates 511000 to 495000 is shown. Arrows indicate the positions of genes and the directions of transcription. The regions carried by yeast genomic clones p1024 and p1020, which complemented the temperature-sensitive growth and low-level BMV-directed GUS gene expression in the mab3 mutant, are shown by thick lines. The square bracket shows the minimal complementing region corresponding to the 2.1-kb AflII-XhoI fragment containing the complete YDJ1 open reading frame.

FIG. 3.

The phenotype of the ydj1i yeast. The 1a and 2a proteins were expressed from pB1CT19 and pB2CT15, respectively, in wt (YPH500) and ydj1i yeast with (+) and without (−) pYDJ1. (A) BMV-directed GUS expression from a DNA platform. B3GUS was expressed in vivo from a GAL1 promoter. Cell cultures were grown in a defined galactose liquid medium, and GUS activity was determined as described in Materials and Methods. The histogram shows averages and standard deviations of relative GUS activity (the average for wt = 1) from three or four transformants. (B) BMV-directed CAT expression from directly transfected B3CAT RNA. Cells were cotransfected with in vitro-synthesized B3CAT and luciferase RNA, and CAT and luciferase activity was determined as described in Materials and Methods. The histogram shows averages and standard deviations of relative ratios of CAT activity to luciferase activity (the average for wt = 1) from four transformants. (C and D) Accumulation of BMV RNA3-related RNAs. BMV RNA3 was expressed in vivo from a GAL1 promoter.Cell cultures were grown in a defined galactose liquid medium, and RNA was analyzed as described in Materials and Methods. Northern blots were probed with ³²P-labeled RNA hybridizing with positive-strand (left panels) and negative-strand (right panels) BMV RNA3 and -4 and autoradiographed. A representative result of Northern hybridization is shown in panel C. Histograms in panel D show averages and standard deviations of relative RNA accumulation (the average for wt = 1) from five or three transformants. The accumulation of each RNA species was quantified and normalized to that of actin mRNA. (E) Accumulation of the 1a and 2a proteins. Two-day cultures of yeast in a defined galactose liquid medium were inoculated in a fresh galactose medium to give an initial A₆₀₀ = 0.05 and grown at 30°C as described in Materials and Methods. At each time point, total protein was extracted and subjected to Western analyses using antibodies against 1a, 2a, or Ydj1p. Parts of Coomassie brilliant blue-stained gels (from 0.1% SDS-9% polyacrylamide gel electrophoresis ) are shown to indicate the amount of loaded protein. Positions of molecular mass markers are shown at the right. Growth curves are also shown. Similar results were obtained in two other independent experiments.

FIG. 4.

Inhibition of accumulation of negative-strand B3(5′GAL, CP^fs) in ydj1i yeast. (A) Schematic representation of BMV-directed RNA synthesis pathway for a BMV RNA3 derivative, B3(5′GAL, CP^fs). B3(5′GAL, CP^fs) was constructed by replacing the viral 5′ NCR of B3CP^fs (38) with the 5′ NCR of yeast GAL1 mRNA and was deficient in the initiation of positive-strand RNA synthesis. (B and C) Accumulation of positive- and negative-strand B3(5′GAL, CP^fs). B3(5′GAL, CP^fs) was expressed in vivo from a GAL1 promoter. Cell cultures were grown in a defined galactose liquid medium and RNA was extracted from the cells and analyzed by Northern hybridization as described in Materials and Methods. Northern blots were probed with ³²P-labeled RNA hybridizing with positive-strand (left panels) and negative-strand (right panels) BMV RNA3 and -4 and autoradiographed. A representative result of Northern hybridization is shown in panel B. Histograms in panel C show averages and standard deviations of relative RNA accumulation (the average for wt = 1) from three transformants. The accumulation of each RNA species was quantified and normalized to that of actin mRNA.

FIG. 5.

Effects of ydj1 mutation on 1a-dependent recruitment of RNA3 to membranes. Yeasts expressing BMV RNA3 with (+1a) or without (−1a) 1a protein were grown in defined galactose medium, spheroplasted, and lysed osmotically. The lysate was then centrifuged at 10,000 × g to yield pellet and supernatant (sup) fractions as described by Chen and Ahlquist (11). RNA was purified by phenol-chloroform extraction and ethanol precipitation, and an equal percentage of each fraction was analyzed by Northern blot hybridization to detect positive-strand RNA3 or cellular actin mRNA. In the lanes labeled “total,” RNA from the unfractionated lysate of the same volume as the supernatant fraction was analyzed.

FIG. 6.

Effect of ydj1 mutation on 1a-dependent recruitment of 2a protein to membranes. Yeast cells expressing 2a in the presence (+1a) or absence (−1a) of 1a were grown, spheroplasted, and lysed osmotically as described by Chen and Ahlquist (11). (A) The lysate was then centrifuged as indicated in the legend to Fig. 5 to yield pellet and supernatant (sup). An equal percentage of each fraction was separated by 0.1% SDS-9% PAGE, and Western analysis was performed with antibodies against 1a or 2a proteins or phosphoglycerate kinase (Pgkp), a cytosolic protein marker. (B) The lysate was subjected to a flotation analysis in a discontinuous sucrose gradient as described by Ahola and Ahlquist (2), with slight modifications. Briefly, 1 ml of the lysate was mixed with 5 ml of 67% (wt/wt) sucrose in HN buffer containing protease inhibitors (2) in an SW40Ti ultracentrifuge tube (Beckman) and overlaid by 5 ml of 50% (wt/wt) sucrose and 1 ml of 10% (wt/wt) sucrose in NH buffer containing protease inhibitors (2). After centrifugation at an average speed of 150,000 × g for ca. 17 h, a membrane fraction (top 2.5 ml; 10 to 50% interface layer), an intermediate fraction (next 2.5 ml; 50% layer), and the sample loading layer fraction (bottom 7 ml; 60% loading layer) were recovered. The same percentage of each fraction was analyzed by SDS-PAGE and Western blotting with the antibodies against 1a or 2a proteins or phosphoglycerate kinase.

FIG. 7.

Effects of addition and depletion of Ydj1p on in vitro BMV RdRp activity. BMV RdRp extract was prepared from protease-deficient BJ5465 yeast expressing 1a and 2a proteins and a replicating BMV RNA3 derivative, B3URA3 RNA, and purified with DEAE-Bio gel A column chromatography. (A) Purified BMV RdRp (50 μl) was treated with protein A-Sepharose-conjugated anti-Ydj1p antibody (+) or protein A-Sepharose without antibodies (−), followed by addition of 500 ng of E. coli-expressed, purified Ydj1p or phosphate-buffered saline. The RdRp reaction procedure was performed with exogenously added BMV virion RNA as a template in the presence of [α-³²P]UTP as described by Quadt et al. (42). The product of the RdRp reaction was treated with S1 nuclease and analyzed by electrophoresis on a 1% agarose gel and by autoradiography. Positions of double-stranded BMV RNAs 1 to 4 are indicated in the figure. (B) Confirmation of depletion of Ydj1p. Purified BMV RdRp, treated with protein A-Sepharose-conjugated anti-Ydj1p antibody (+) or just protein A-Sepharose (−), was subjected to SDS-PAGE followed by Western analysis with anti-Ydj1p antibodies. The position of the Ydj1p band is indicated. Arrowhead shows a background signal presumably derived from leaked protein A.