Comparison of capsid-associated RNA-dependent RNA polymerase activity of double-stranded RNA virus-like particles from different yeast genera

Abstract

Killer yeasts secrete proteinaceous toxins that kill sensitive yeasts I 11. In Saccharornyces cerevisiae, killer determinants are cytoplasmically inherited double-stranded RNA (dsRNA) genomes, L and M, encapsidated in virus-like particles (VLPs), ScV-L and ScV-M. ScV particles possess capsid-associated RNA-dependent RNA polymerase activity [2]. With L or M dsRNA as template, RNA polymerase activity results in full-length single-stranded RNA (ssRNA) (+) strand transcripts that the same RNA polymerase replicates to produce dsRNAs identical to the parental species [2]. From sequence analysis, the RNA polymerase has been attributed to a 180kDa protein [3] associated with ScV particles that provides both transcriptase and RNA replicase activities. Capsid-associated RNA polymerases have been reported for several fungal dsRNAs viruses, including those of Aspergillus foetidus [4], Pmicillium species 151 and Ustilago maydis [ 6 ] . Amongst these, even in he most thoroughly studied cases, the RNA polymerase has not been identified but is again inseparable from the intact capsid with enclosed dsRNA 171. This work describes assays of capsid-associated RNA polymerase of SCV-LIA and SCV-LBC particles from S. cerevisiae K7 and S. cerevisiae I IA, respectively. As these two particles are capable of independent existence, presumably each has its own capsidassociated RNA polymerase. Assays were also performed with ScVL3/ScV-M3 particles from group 3 killer S. cerevisiae NCYC761 and those isolated from two other yeasts, Yarrowia lipolyfica CBS6124-2, YIV-Ly, and Pichia membranaefaciens NCYC333, PmV-Lp. VLPs were extracted from 72 hour cultures as described earlier [8]. RNA polymerase was assayed as follows. For each VLP species, 2.5 A254 absorbance units were added to Iml reaction mixture containing 50mM Tris-HCI, pH7.4, 5mM MgCI2, l0mM PME, 1OOp M EDTA pH8.0, 20mM NaCI, 5mM KCI, 2mg Bentonite, 500pM each ATP. CTP, GTP, 75pCi ~ ' [ c ' ~ ~ P ] UTP and incubated at 30OC. Samples (IOpI) were removed at 30 to 60 minute intervals, and TCA precipitated onto nitrocellulose filters. Radioactive counts were measured and a time-course assay of [ c c ~ ~ P ] UTP incorporation plotted. When the time-course y indicated that the reaction had stopped they were phenol extracted and RNA precipitated. The results indicate RNA polymerase activities associated with each VLP species assayed. However, reactions did not proceed at the same rate, with corresponding differences in total label incorporated. ScV-L I A panicles consistently yielded the highest levels of radio-labelled product and SCV-LBC particles consistently the lowest. ScV-LI A particles showed the highest initial rate of reaction, incorporating more label in the first twenty minutes than any other VLP species. The relative order of the apparent rates of reaction for the VLP species remained constant when the assays were repeated. This suggests that differences in reaction rates observed between VLP species are true differences and not due to random (e.g. sampling) errors. Ss and ds RNA were separated by fractionation on a 5ml CFI I(Whatman) column [9] and alcohol precipitated. Yield of ss and ds RNA varied. SCV-LIA and S C V L ~ A particles yielded predominantly dsRNA. about 6S% of total product. SCV-LBC, YIV-Ly and PmV-Lp particles yielded 85% ssRNA. Aliquots of ssRNA were co-electrophoresed with Ipg rRNA (23s and 16S, E. coli) on non-aqueous formamide, 4% polyacrylamide tube gels, and distance migrated by the cold rRNA markers determined by UV-absorbance scanning (Joyce-Loebl) at 265nm. Gels were cut into 2.5mm slices perpendicular to the direction of migration, and radioactive counts determined for each slice. Radioactivity profiles for ssRNA product was plotted together with a similar profile of A265 For each ssRNA product the peak of radioactivity was calculated to be half the molecular weight of the relevant dsRNA species. It was concluded that ssRNA products are full-length transcripts of their dsRNA templates. Two aliquots of ssRNA product were each resuspended in 2ml IxSSC (150mM NaCI, 15mM sodium citrate, pH 7.0). 45% formamide and 1 pg unlabelled dsRNA (appropriate to product analysed). Samples were either denatured at 85OC for three minutes, then incubated at 60OC for 30 minutes, or kept at 60OC for 33 minutes. Both were ethanol precipitated, digested with I pgml-] RNase A in 2xSSC, phenol extracted and re-precipitated. Samples were loaded onto 4% polyacrylamide tube gels, in tris-borate buffer, electrophoresed, UV-scanned, sliced and Cerenkov counted, followed by plotting the resulting profiles of radioactivity and A265. When subjected to both denaturing and re-annealing conditions, the peak of the radioactivity profile coincided with that of the ,4265 profile, corresponding to dsRNA. That is, ssRNA product hybridised to template dsRNA. No activity was found on gels loaded with transcript and dsRNA that were incubated at 60OC without prior denaturation. This confirms RNase sensitivity of the transcript and suggests that it does not form RNase resistant dsRNA by a process of self-annealing. SsRNA product from each polymerase assay therefore seems to be a transcript of single, same sense strand of the original dsRNA template. DsRNA product was co-electrophoresed with 1 pg dsRNA appropriate to the product analysed on 4% polyacrylamide tube gels. The radioactivity profile of dsRNA product was plotted together with a profile of ,4265. The peak of radioactivity coincided with the peak in the ,4265 profile, corresponding to cold dsRNA. It was concluded that dsRNA product and dsRNA template shared the same molecular weight. Capsid-associated RNA-dependent RNA polymerase activity has been detected for each of the VLP groups tested. These activities each resulted in the production of full-length ssRNA, transcribed from one strand of dsRNA template, and full-length dsRNA. There are differences between these activities, with of SCV-LIA and ScVL ~ A consistently yielding predominantly dsRNA and the other VLPs yielding predominantly ssRNA. As RNA polymerase performs both transcriptase (ds+ss) and replicase (ss+ds) functions, I t might suggest that for SCV-LIA and S C V L ~ A there is a more active RNA replicase activity, and a more active transcriptase for the other VLPs.