【病毒外文文献】2001 Enhanced Accumulation of Coronavirus Defective Interfering RNA from Expressed Negative-Strand Transcripts by Coexpr

Enhanced Accumulation of Coronavirus Defective Interfering RNA from Expressed Negative Strand Transcripts by Coexpressed Positive Strand RNA Transcripts Sangeeta Banerjee John F Repass 1 and Shinji Makino 2 Department of Microbiology and Institute for Cellular and Molecular Biology The University of Texas at Austin Austin Texas 78712 and Department of Microbiology and Immunology University of Texas Medical Branch at Galveston Galveston Texas 77555 1019 Received April 6 2001 returned to author for revision May 10 2001 accepted June 14 2001 published online August 3 2001 Expression of negative strand murine coronavirus mouse hepatitis virus MHV defective interfering DI RNA transcripts in MHV infected cells results in the accumulation of positive strand DI RNAs M Joo et al 1996 J Virol 70 5769 5776 However the expressed negative strand DI RNA transcripts are poor templates for positive strand DI RNA synthesis The present study demonstrated that DI RNA accumulation from the expressed negative strand DI RNA transcripts in MHV infected cells was enhanced by the coexpression of complementary RNA transcripts that correspond to the 59 region of positive strand DI RNA The positive strand RNA transcripts corresponding to the 59 end most 0 7 2 0 kb DI RNA had a similar enhancement effect The coexpressed positive strand RNA transcripts lacking the leader sequence or those containing only the leader sequence failed to demonstrate this enhancement effect demonstrating that the presence of the leader sequence in the coexpressed positive strand RNA transcripts was necessary but not sufficient for the enhancement of DI RNA accumulation from the coexpressed negative strand DI RNA transcripts Negative strand DI RNA transcripts that were coexpressed with the partial length positive strand RNA transcripts were no more stable than those expressed alone suggesting that a higher stability of the expressed negative strand RNA transcripts was an unlikely reason for the higher DI RNA accumulation in cells coexpressing two complementary DI RNA transcripts Sequence analyses unexpectedly demon strated that the leader sequence of the majority of accumulated DI RNAs switched to helper virus derived leader sequence suggesting that enhancement of DI RNA accumulation was mediated by the efficient utilization of helper virus derived leader sequence for DI RNA synthesis Furthermore our data suggested that this leader switching a type of homologous RNA RNA recombination occurred during positive strand DI RNA synthesis and that MHV positive strand RNA synthesis mechanism may have a preference toward recognizing double stranded RNA structures over single stranded negative strand RNA to produce positive strand DI RNAs 2001 Academic Press Key Words coronaviruses mouse hepatitis virus negative strand RNA defective interfering RNA leader sequence leader switching RNA replication RNA recombination double stranded RNAs RNA expression INTRODUCTION Murine coronavirus mouse hepatitis virus MHV is a single stranded positive sense RNA virus approxi mately 31 kb in length Bonilla et al 1994 Lai and Stohlman 1978 Lee et al 1991 Pachuk et al 1989 MHV infected cells generate seven to eight species of virus specific mRNAs whose sequences comprise a 39 coterminal nested set structure Lai et al 1981 Leibowitz et al 1981 The mRNAs are numbered 1 to 7 in decreas ing order of size Lai et al 1981 Leibowitz et al 1981 MHV particles carry only mRNA 1 and only mRNA 1 contains a packaging signal Fosmire et al 1992 van der Most et al 1991 At their 59 ends all the mRNAs are fused to a 72 to 77 nucleotide long leader sequence Lai et al 1984a 1983 Spaan et al 1983 MHV mRNA body sequences begin from a transcription consensus se quence in the intergenic region that is located upstream of each gene Lai et al 1984a 1983 Makino et al 1988b Spaan et al 1983 Genomic size and subgenomic size negative strand RNAs are present in coronavirus in fected cells in amounts that are significantly lower than the amounts of corresponding positive strand RNAs Sethna et al 1989 Two independent studies showed that nascent leader sequence containing MHV sub genomic mRNAs are elongating on a genomic length replicative intermediate RNA containing a genomic length negative strand RNA late in infection Baric et al 1983 Mizutani et al 2000 The genomic length nega tive strand RNA also serves as a template for sub genomic mRNA synthesis early in infection An et al 1998 These studies established that the negative strand genomic RNA is a template for MHV subgenomic mRNA synthesis and that leader sequence joins to the body of subgenomic RNA during subgenomic mRNA synthesis It has been proposed that negative strand subgenomic RNAs also serve as templates for sub 1 Present address M D Anderson Cancer Center Science Park Research Division Smithville TX 78957 2 To whom correspondence and reprint requests should be ad dressed at Department of Microbiology and Immunology The Univer sity of Texas Medical Branch at Galveston Galveston Texas 77555 1019 Fax 409 772 5065 E mail shmakino utmb edu Virology 287 286 300 2001 doi 10 1006 viro 2001 1047 available online at on 0042 6822 01 35 00 Copyright 2001 by Academic Press All rights of reproduction in any form reserved 286genomic mRNA synthesis in coronavirus Baric and Y ount 2000 Sawicki et al 2001 Sawicki and Sawicki 1990 Sethna et al 1989 and related arterivirus van Marle et al 1999 Cloned defective interfering DI RNAs of coronavi ruses have been used to study the mechanism of coro navirus RNA replication Chang et al 1994 Dalton et al 2001 de Groot et al 1992 Izeta et al 1999 Kim et al 1993a b Kim and Makino 1995 Liao and Lai 1995 Lin and Lai 1993 Makino and Lai 1989b Repass and Makino 1998 van der Most et al 1994 Transfection of in vitro synthesized DI RNA transcripts into helper virus infected cells results in DI RNA replication Makino and Lai 1989b Expression of DI RNAs in helper virus in fected cells using a recombinant T7 vaccinia virus ex pression system also results in DI RNA replication Lin and Lai 1993 Three discontiguous regions are required for replication of DI RNAs derived from the JHM strain of MHV MHV JHM these regions are derived from the 59 end 0 47 kb of DI RNA an internal 58 nt long region internal cis acting replication signal corresponding to 0 9 kb from the 59 end of DI RNA and the 39 end 0 46 kb of DI RNA Kim and Makino 1995 Lin and Lai 1993 Repass and Makino 1998 Among these three regions only the 39 end 55 nucleotides plus a poly A tail are necessary for negative strand RNA synthesis Lin et al 1994 In MHV JHM DI RNAs the secondary structure of the internal cis acting replication signal of positive strand RNA is important for positive strand RNA synthe sis Repass and Makino 1998 Coronavirus DI RNAs have two unique biological prop erties that are not described in other DI RNAs of positive strand RNA viruses One is leader switching in which the leader sequence of the DI RNA switches to the leader sequence of helper virus with a high efficiency during DI RNA replication Makino and Lai 1989b The 39 region of the leader sequence of MHV genomic RNA contains two to four repeats of an UCUAA sequence Makino and Lai 1989a Makino et al 1988b Fig 1A Immediately down stream of this pentanucleotide repeat is a nine nucle otide sequence of UUUAUAAAC which is found in most MHVs Makino and Lai 1989a Makino et al 1988b Fig 1A All naturally occurring MHV DI RNAs characterized so far contain three to four repeats of UCUAA and lack the nine nucleotide sequence Makino et al 1985 1988a When MHV DI RNAs containing the nine nucle otide sequence are constructed and transfected into MHV infected cells the leader sequence of DI RNA switches to that of helper virus with a very high effi ciency while this leader switching does not occur in DI RNAs lacking the nine nucleotide sequence Makino and Lai 1989b Leader switching is also found in bovine coronavirus BCV DI RNAs Chang et al 1996 and infectious bronchitis virus Stirrups et al 2000 Leader switching occurs during the rescue of defective RNAs by heterologous strains of the coronavirus infectious bron chitis virus The mechanism of leader switching is not known We have speculated that leader switching occurs during positive strand DI RNA synthesis Makino and Lai 1989b while others have speculated that it may occur during negative strand DI RNA synthesis Chang et al 1996 Another unique property of coronavirus DI RNA is that DI RNA replication occurs from negative strand DI RNA transcripts that are transfected or expressed in helper virus infected cells Joo et al 1996 However the effi ciency of DI RNA accumulation from transfected or ex pressed negative strand DI RNA transcripts is poor MHV DI RNA replicates extremely efficiently after transfection of positive strand DI RNAs into MHV infected cells Makino and Lai 1989b whereas several passages of virus sample are necessary to demonstrate DI RNA rep lication after transfection of large amounts of negative strand DI RNA transcripts Joo et al 1996 We do not know why expressed or transfected negative strand DI RNAs are poor templates Most of the expressed nega tive strand DI RNAs probably exist as single stranded RNAs in MHV infected cells while it has been suggested that MHV negative strand RNAs do not exist as single stranded RNAs but associate with positive strand RNAs Lin et al 1994 Sawicki and Sawicki 1986 Therefore we speculated that negative strand DI RNAs may be poor templates for positive strand DI RNA synthesis since they are probably present as single stranded RNA species in MHV infected cells Joo et al 1996 In the present study we examined whether coexpres sion of negative strand DI RNA transcripts with their complementary positive strand DI RNA fragments corre sponding to the 59 region of positive strand DI RNA enhances DI RNA accumulation We hoped that the ex pression of negative strand DI RNA transcripts with par tial length positive strand RNA would form double strand ds RNA structures in MHV infected cells and that such ds RNAs may be better template RNAs for positive strand DI RNA synthesis Our study demonstrated that DI RNA accumulated more efficiently when partial length positive strand DI RNA fragments were coexpressed with negative strand DI RNA transcripts Unexpectedly we also found that the enhancement of DI RNA accumu lation was mediated by the leader switching mechanism The present data provided further information about the mechanism of leader switching and positive strand MHV RNA synthesis from negative strand template RNA RESULTS Enhancement of DI RNA accumulation from negative strand DI RNA transcripts by coexpressed partial length positive strand DI RNA fragments To test the possibility that positive strand DI RNA syn thesis occurs efficiently from negative strand DI RNA template that exists within a ds RNA structure DI RNA 287 MHV DI RNA REPLICATIONFIG 1 Schematic representation of all the clones used in this study A Structure of MHV A59 helper virus RNA with details of its 59 end sequences The hatched box represents the wt 30 nt region with the actual sequence shown The open rectangle represents the leader sequence Specific nucleotides appear within the rectangle with its position identified above 2R represents two repeats of the UCUAA sequence The black box represents the nine nucleotide UUUAUAAAC sequence located just downstream of the leader sequence B The structure of the full length positive strand DI RNA DIU and its derivatives The shaded box represents the unique 30 nt region The actual nucleotide sequence is shown above the box with the mutated nucleotides written in bold and underlined Probe 3 binds to positive sense DI RNA containing the unique 30 nt sequence The open rectangles represent the leader sequence The solid bold lines define deletions of the nine nucleotide sequence for most of the DI RNAs and deletion of the entire leader sequence in case of DIU SpeDleader 4R represents four repeats of the UCUAA sequence The black arrowheads represent the T7 promoter sequence C The structure of full length negative strand DI RNA pDER pDER4 and pDER derived pDER19 that contains the nine nucleotide sequence Probe 1 hybridizes with positive sense DI RNA containing the wt 30 nt sequence and probe 2 binds to pDER transcripts The open rectangles represent the antileader sequence The solid bold lines define deletions of the nine nucleotide sequence The black arrowheads represent the T7 promoter sequence All nucleotide sequences are shown in positive polarity D Schematic representation of pS5A plasmid The black arrowhead and black circle represent the T7 promoter sequence and T7 terminator sequence respectively 288 BANERJEE REPASS AND MAKINOaccumulation in MHV infected cells coexpressing com plete negative strand DI RNA transcripts and its comple mentary positive strand DI RNA transcripts was com pared with that in infected cells expressing complete negative strand DI RNA transcripts alone We expected a certain population of the coexpressed RNA transcripts to form ds RNA structures using complementary se quences If negative strand DI RNA in such a ds RNA structure is a better template for positive strand DI RNA synthesis then the amount of DI RNA in coexpressing cells might be higher than that in cells expressing neg ative strand DI RNA transcripts alone If full length pos itive strand DI RNA and full length negative strand DI RNA are coexpressed accumulated DI RNAs should be derived from both expressed template RNAs since DI RNA synthesis starts from both expressed positive strand transcripts Lin and Lai 1993 and expressed negative strand DI RNA transcripts Joo et al 1996 To identify DI RNAs that were initially synthesized from the expressed negative strand DI RNA transcripts nucleo tide sequences at a specific region of positive strand DI RNA transcripts were mutated whereas the correspond ing region in the negative strand DI RNA had no such mutation DI RNAs that were initially synthesized from the expressed negative strand DI RNA transcripts should be detected using an oligonucleotide probe that specif ically hybridized with the sequence specific for the ex pressed negative strand DI RNA As a parental plasmid encoding positive strand DI RNA transcripts we constructed DIU in which a full length positive strand DI RNA sequence was placed between a T7 promoter and a T7 terminator in a plasmid Fig 1B DIU had 11 nucleotides substituted within the unique 30 nt region from nucleotide 487 to 516 Fig 1B of the naturally occurring MHV JHM DI RNA DIssE Makino et al 1988a First we tested the feasibility of using DIU for coexpression studies DIU was transfected into cells infected with recombinant vaccinia virus vTF7 3 which expresses the T7 polymerase Fuerst et al 1986 Four hours after DIU transfection cells were infected with MHV A59 and intracellular RNA was ex tracted 10 h postinfection pi of MHV To test if the accumulated DI RNA maintained the unique 30 nt region we performed Northern blot analysis of intracellular RNA using probe 3 Fig 1B Table 1 which specifically hy bridizes with the unique 30 nt region of DIU and probe 1 Fig 1C Table 1 which specifically hybridizes with the corresponding region of wild type wt sequence wt 30 nt region Stringent conditions were set up for the oligonucleotide probe binding to the specific DI RNAs such that probe 1 did not hybridize with in vitro synthe sized DIU at all and probe 3 did not hybridize with in vitro synthesized positive sense RNA containing the wt 30 nt region We found that approximately half the accumu lated DI RNAs contained the wt 30 nt region and the rest contained the unique 30 nt region data not shown RNA recombination between helper virus and the replicating DI RNAs most probably caused the accumulation of DI RNA containing wt 30 nt region Generation and accu mulation of DI RNA containing wt 30 nt region after expression of DIU indicated that we could not easily identify the origin of the accumulated DI RNA containing wt 30 nt region after coexpression of DIU and negative strand DI RNA transcripts containing wt 30 nt region Thus DIU was not suitable for cotransfection studies Next we examined whether DI RNA synthesis from negative strand DI RNA transcripts is enhanced by the coexpression of a positive strand RNA fragment that contains only the 59 region of the DI RNA We hoped that the partial length positive strand DI RNA transcripts would hybridize with the 39 region of the expressed negative strand DI RNA transcripts to create a ds RNA region which would promote efficient positive strand DI RNA synthesis Four DIU derived clones DIU Nru DIU TABLE 1 Synthetic Oligonucleotides Used in This Study Oligonucleotide Sequence Binding site Polarity 1024 59 ATCTGATGCATTAAAGTC 39 DIssE 856 873 Negative 2326 59 CACCGCATATGGTGCA 39 pT7 4 319 334 Positive 10066 59 TATAAGAGTGATTGGCGTCCG 39 DIssE 1 21 Positive 10080 59 GGCAACGCCGTCCTCTTCTTGGGTATCGGC 39 DIssE 931 960 Negative 10120 59 CTTTAGACAACGCCAGTT 39 DIssE 1594 1611 Negative 10134 59 AAGACATCCTCATAGGTCTTGTCC 39 DIssE 1236 1259 Negative 10239 59 CCCCAGAAGGTGGAGGCCTCGACGATGATGGCGCTACAATTT GGCTCAGCGGTCTTGGTCAAGCCATCC 39 DIssE 466 534 Positive 10258 59 CTGGCGCCGAATGGACACGTC 39 DIssE 168 188 Negative 10285 59 CGTCCGTACGTACCTAATCTACTC 39 DIssE 16 39 Positive 10682 59 CCCCCTCTAGAGTTTAGATTAGATTAGATTAGATTTAAAC 39 DIssE 53 81 Negative 10683 59 CCCCCTCTAGATTTAAACTACAAGAG 39 DIssE 45 59 Negative Probe 1 59 AGCACTACCGAACTGCAATGCCATCATAGT 39 DIssE 487 516 Negative Probe 2 59 TTGGTTAATCACGTGAGGGTGGATTGTAGC 39 DIssE 370 399 Positive Probe 3 59 CGCTGAGCCAAATTGTAGCGCCATCATCGT 39 DIssE 487 516 Negative 289 MHV DI RNA REPLICATIONSpe DIU Sph and DIU Eag were constructed which had a 2 0 1 5 1 0 and 0 7 kb long 59 end region of DIU respectively Fig 1B All these clones contained the unique 30 nt region and their RNA transcripts should not replicate since they all lack the 39 cis acting replication signal Kim et al 1993a Lin et al 1994 Northern blot analysis using probe 1 showed that DI RNA containing the wt 30 nt region did not accumulate after expression of any of these DIU derived clones in MHV infected cells see Fig 2 lanes 10 and 11 for DIU Spe data not shown for other clones demonstrating that the expressed tran scripts did not undergo RNA recombination to produce full length DI RNA containing the wt 30 nt region These clones were suitable for subsequent studies Plasmid pDER was used to express negative strand DI RNA transcripts Joo et al 1996 In this plasmid the DI specific sequence was placed between the T7 pro moter and T7 terminator such that T7 RNA polymerase mediated transcription produced negative strand DI RNA transcripts containing the wt 30 nt region Fig 1C Equal amounts of pDER plasmid and each of the DIU derived plasmids encoding the partial length positive strand DI RNA transcripts were mixed and transfected into vTF7 3 infected cells In a control group pDER was mixed with equal amounts of plasmid pS5A Woo et al 1997 which encodes the chloramphenicol acetyl trans ferase CAT gene and no MHV specific sequences Fig 1D and then transfected Total amount of DNA used for all the transfections was the same DNA transfected cells were infected with MHV and intracellular RNAs were extracted 10 h post MHV infection Northern blot analysis using probe 1 showed that in all cases DI RNAs containing the wt 30 nt region accumulated significantly higher in cells coexpressing pDER and positive strand RNA fragments than in cells expressing pDER alone Representative data from these experiments using pDER and DIU Spe are shown in Fig 2 Since these DI RNAs contained the wt 30 nt region they were most likely synthesized initially from the pDER transcripts We per formed these experiments at least five times for all DIU derived clones and obtained consistent results data not shown Although the enhancement effect differed slightly from experiment to experiment phosphorimag ing analysis of the membra