【病毒外文文献】1998 Characterizations of Coronaviruscis-Acting RNA Elements and the Transcription Step Affecting Its Transcription Effi

Characterizations of Coronavirus cis Acting RNA Elements and the Transcription Step Affecting Its Transcription Efficiency Sungwhan An and Shinji Makino 1 Department of Microbiology and Institute for Cellular and Molecular Biology The University of Texas at Austin Austin Texas 78712 1095 Received November 14 1997 returned to author for revision December 19 1997 accepted January 26 1998 Seven to eight species of viral subgenomic mRNAs are produced in coronavirus infected cells These mRNAs are produced in different quantities and their molar ratios remain constant during viral replication We studied RNA elements that affect coronavirus transcription efficiency by characterizing a series of cloned coronavirus mouse hepatitis virus MHV defective interfering DI RNAs containing an inserted intergenic sequence from which subgenomic DI RNA is transcribed in MHV infected cells Certain combinations of upstream and downstream flanking sequences of the intergenic sequence suppressed subgenomic DI RNA transcription yet changing one of the flanking sequences to a different sequence eliminated transcription suppression The suppressive effect of certain combinations of flanking sequences but not all combinations could be counteracted by altering the intergenic sequence Thus the combination of intergenic sequence and flanking sequence affected transcription efficiency We also characterized another set of DI RNAs designed to clarify which transcription step determines the relative molar ratios of coronavirus mRNAs Our study indicated that if subgenomic mRNAs were exclusively synthesized from negative strand genomic RNA then the relative molar ratios of coronavirus mRNAs were most likely determined after synthesis of the genomic sized template RNA If negative strand subgenomic RNAs were templates for subgenomic mRNAs then the relative molar ratios of coronavirus mRNAs probably were determined after synthesis of the genomic sized template RNA used for subgenomic sized RNA transcription but prior to the completion of the synthesis of subgenomic sized RNAs containing the leader sequence The relative molar ratios of coronavirus mRNAs therefore seem to have been established prior to a putative replicon type amplification of subgenomic mRNAs 1998 Academic Press INTRODUCTION Many eukaryotic RNA viruses express their gene s by producing subgenomic mRNA s in infected cells Coro navirus an enveloped virus containing a large positive sense single strand RNA belongs to this group Cells infected with mouse hepatitis virus MHV a prototypic coronavirus produce seven to eight species of virus specific mRNAs that make up a 39 coterminal nested set structure Lai et al 1981 Leibowitz et al 1981 These mRNAs which are named mRNAs 1 through 7 in de creasing order of size Lai et al 1981 Leibowitz et al 1981 are produced in different quantities and their molar ratios remain constant during MHV replication The 59 end of the MHV genomic RNA and the sub genomic mRNAs start with a leader sequence that is approximately 72 to 77 nucleotides nt long Lai et al 1983 1984 Spaan et al 1983 the presence of the leader sequence in subgenomic mRNAs is a unique character istic in coronavirus and arterivirus de Vries et al 1990 which is closely related to coronavirus Curiously on the genome the leader sequence is not found any place besides the 59 end yet the subgenomic mRNAs have the leader sequences fused with the mRNA body se quences The mRNA body sequences begin from a UC UAAAC transcription consensus sequence or a very sim ilar sequence of intergenic sequences which is located upstream of each MHV gene Coronavirus transcription undergoes a discontinuous transcription step in which the leader sequence somehow joins to the body of the subgenomic RNA Jeong and Makino 1994 Zhang et al 1994 Genomic sized and subgenomic sized negative strand RNAs each of which corresponds to one of the subgenomic mRNA species are also present in corona virus infected cells Sethna et al 1989 These negative strand RNAs contain an antileader sequence at the 39 end and a poly U sequence at the 59 end Sethna et al 1991 Several models have been proposed to explain coro navirus subgenomic RNA synthesis One model pro poses that negative strand RNA synthesis starts on genomic RNA and terminates at the intergenic sequence Sawicki and Sawicki 1990 In this model the leader sequence joins to the body of subgenomic RNA either by relocalization of negative strand subgenomic RNA to the leader sequence of the genomic RNA antileader se quence joins to negative strand subgenomic RNA or during subgenomic mRNA synthesis on negative strand subgenomic RNA template Sawicki and Sawicki 1990 1 To whom correspondence and reprint requests should be ad dressed Fax 512 471 7088 E mail makino mail utexas edu VIROLOGY 243 198 207 1998 ARTICLE NO VY989059 0042 6822 98 25 00 Copyright 1998 by Academic Press All rights of reproduction in any form reserved 198Another way to envision negative strand subgenomic RNA synthesis is provided by the idea that processing of the full length negative strand genomic RNA occurs after transcription from the genome Subgenomic mRNA syn thesis would follow processing of subgenomic negative strand RNAs the leader sequence could join the body sequence during processing of negative strand genomic RNA or during subgenomic mRNA synthesis on negative strand subgenomic RNA These two models propose that negative strand subgenomic RNAs are synthesized prior to the synthesis of subgenomic mRNAs All other tran scription models propose that subgenomic mRNAs are synthesized prior to the synthesis of negative strand sub genomic RNA Subgenomic mRNAs may be synthesized by the processing of the positive strand genomic RNA Baric et al 1983 or by transcription from negative strand genomic RNA by a unique leader primed tran scription mechanism in which short free leader RNAs are used as primers for subgenomic mRNA synthesis Baric et al 1983 or by polymerase jumping from leader sequence to intergenic sequence during subgenomic mRNA synthesis on the negative strand genomic RNA template Spaan et al 1983 In these models once subgenomic mRNAs are synthesized negative strand subgenomic RNAs are copied on subgenomic mRNAs Negative strand subgenomic RNAs may have several different effects on transcription These negative strand subgenomic RNAs may be templates for subgenomic mRNA synthesis Sawicki and Sawicki 1990 Schaad and Baric 1994 or may become replicons which undergo replication to amplify more subgenomic mRNA and neg ative strand subgenomic RNAs Sethna et al 1989 Al ternatively these negative strand subgenomic RNAs may not participate in subsequent transcription they may be dead end transcription products Jeong and Makino 1992 All of the coronavirus transcription models postu late several transcription steps for the production of mature mRNAs one of these unidentified transcriptions steps is that which defines the relative molar ratios of subgenomic mRNAs Due to the large size of coronavirus genomic RNA a coronavirus infectious cDNA clone has not yet been constructed Instead for studying transcription we use coronavirus defective interfering DI RNAs that contain an inserted intergenic sequence from which transcrip tion in DI RNA replicating coronavirus infected cells yields a subgenomic DI RNA with a leader sequence Makino et al 1991 Studies using MHV DI RNAs iden tified RNA elements that affect transcription Jeong et al 1996 Joo and Makino 1992 1995 Makino and Joo 1993 Makino et al 1991 van der Most et al 1994 van Marle et al 1995 One study showed that a series of DI RNAs each of which contains a 0 4 kb long sequence derived from various regions of the MHV sequence with an intergenic sequence in the middle transcribe different amounts of subgenomic DI RNA Jeong et al 1996 demonstrating that sequences flanking the intergenic sequence affect transcription Other studies showed that differences in the sequence of the inserted intergenic sequence also affect transcription Jeong et al 1996 Joo and Makino 1992 van der Most et al 1994 A DI RNA containing a 12 nt long intergenic sequence 12 nt se quence UCUAAUCUAAAC that is flanked by 0 2 kb long sequences upstream and downstream of the genes 6 7 junction transcribes only a small amount of sub genomic DI RNA while a DI RNA containing an 18 nt long naturally occurring intergenic sequence at genes 6 7 18 nt sequence AAUCUAAUCUAAACUUUA in the place of the 12 nt sequence transcribes a significantly higher amount of subgenomic DI RNA Jeong et al 1996 demonstrating that the 18 nt sequence overcomes the flanking sequence mediated transcription suppression Among the many unanswered questions about coro navirus transcription regulation we addressed the fol lowing three questions I Does replacement of the 12 nt sequence with the 18 nt sequence in the intergenic se quence always overcome flanking sequence mediated transcription suppression II Does the presence of one of the flanking sequences or the presence of certain combinations of upstream and downstream flanking se quences suppress transcription III During which step of transcription are the relative ratios of MHV mRNAs mainly determined RESULTS In an attempt to test for their susceptibility to transcrip tion suppression by flanking sequences the 18 nt or the 12 nt sequences were independently placed in the mid dle of different 0 4 kb sequences that had been inserted in a complete MHV DIss EcDNA clone DE5 w3 Makino and Lai 1989b between the AflII site and the SacII site Fig 1 this placement of the intergenic sequence re sulted in its being flanked by two 0 2 kb sequences We called the 0 2 kb long regions that were located up stream and downstream of the intergenic sequence the 0 2 kb upstream flanking sequence and the 0 2 kb down stream flanking sequence respectively Effect of 18 nt sequence on flanking sequence mediated transcription suppression The 18 nt sequence overcomes the effect of transcrip tion suppression that the 0 2 kb upstream and down stream flanking sequences exert over the gene 6 7 in tergenic sequence Jeong et al 1996 We attempted to confirm this previous observation through a comparison of the transcripts of in vitro synthesized DI RNA trans fected into MHV infected cells from mutants FDI 6 7 M and FDI 6 7 wt which both have 0 2 kb insertions up stream and downstream of the gene 6 7 intergenic se quence and respectively carry the 12 and 18 nt se quence Northern blot analysis of total cytoplasmic RNAs 199 CORONAVIRUS TRANSCRIPTIONusing probe 1 which corresponds to nucleotides 1488 1610 from the 59 end DE5 w3 showed that FDI 6 7wt produced a significantly larger amount of subgenomic DI RNA than FDI 6 7 M Fig 2 These data were consistent with our previous result Jeong et al 1996 As shown in Fig 2 and subsequent figures sometimes Northern blot analyses did not show a distinct MHV A59 genomic RNA band this was probably due to minor degradation of RNAs To examine whether replacement of the intergenic sequence from the 12 nt sequence to the 18 nt sequence always overcomes the flanking sequence mediated tran scription suppression we examined two additional sets of DI RNAs one set was FDI EagIwt and FDI EagI and the other was FDI NruIwt and FDI NruI FDI EagIwt and FDI NruIwt had an insertion of the 0 4 kb region sur rounding the EagI site and NruI site of DE5 w3 respec tively and both contained the 18 nt sequence Fig 1 FDI EagI and FDI NruI contain the 12 nt sequence in place of the 18 nt sequence Jeong et al 1996 Northern blot analysis showed no significant differences in the molar ratio of subgenomic DI RNA to genomic DI RNA among these four DI RNAs Fig 2 The replacement of the 12 nt sequence with the 18 nt sequence did not always overcome transcription suppression by the flank ing sequences Characterization of transcription suppression mediated by flanking sequences To know whether the presence of one of the flanking sequences or the presence of certain combinations of upstream and downstream flanking sequences suppress transcription we constructed a series of DI RNAs all of which contained the 12 nt sequence and had a different combination of the 0 2 kb upstream and downstream flanking sequences Fig 3 In these clones one of the flanking sequences of FDI EagI FDI NruI and FDI 6 7 M was replaced with one of the 0 2 kb sequences that flank the naturally occurring intergenic sequences at genes 6 7 2 3 and 1 2 Northern blot analysis using a 32 P labeled probe 1 showed that all of the newly constructed FIG 1 Schematic diagram of the structure of MHV genomic RNA DE5 w3 and DE5 w3 derived insertion mutants with flanking sequences The 0 2 kb flanking sequences that were inserted into DE5 w3 derived insertion mutants are shown as 0 2 kb fragments Each mutant has a 0 4 kb long insertion between the AflII site and SacII site of DE5 w3 Molar ratio sG G represents the average molar ratio of subgenomic DI RNA to genomic DI RNA from at least three independent experiments see Fig 2 The location of probe 1 used for Northern blot analysis see Fig 2 is also shown 200 AN AND MAKINODI RNAs efficiently transcribed subgenomic DI RNA Fig 4 FDI 1 EagI M which had a naturally occurring 0 2 kb upstream flanking sequence at gene 1 2 plus a 0 2 kb downstream flanking sequence of the EagI site effi ciently transcribed subgenomic DI RNA while FDI EagI produced only a very low level of subgenomic DI RNA These data may be interpreted as meaning that the 0 2 kb upstream flanking sequence of FDI EagI con tained a transcription suppressive element s However FDI EagI 2 M which had the 0 2 kb upstream flanking sequence of FDI EagI and a 0 2 kb long downstream flanking sequence at gene 1 2 efficiently transcribed subgenomic DI RNA putative transcription suppressive element s at the 0 2 kb upstream flanking sequence did not suppress transcription Comparison of FDI EagI and FDI EagI 2 M may indicate the presence of the transcrip tion suppressive element s at the 0 2 kb downstream flanking sequence of the EagI site Again this putative transcription suppressive element s did not work at all in FDI 1 EagI These data indicated that the presence of either the 0 2 kb upstream or downstream flanking se quences of the EagI site did not suppress transcription Transcription suppression only occurred when the 0 2 kb upstream and downstream flanking sequences of FDI EagI were combined Studies of the remaining two sets of DI RNAs resulted in the same conclusion transcrip tion suppression only occurred when the DI RNAs con tained both upstream and downstream 0 2 kb fragments around the NruI site and when DI RNA had both up stream and downstream 0 2 kb fragments at the gene 6 7 junction In addition we did not see any relation ships between the transcription efficiency and the pres ence of a large open reading frame in subgenomic DI RNA among DI RNAs Transcription step that determined the molar ratio of subgenomic DI RNA to genomic DI RNA For identification of the transcription step that mainly determines the relative molar ratios of MHV mRNAs we first characterized constructs FDI 2 7 M and FDI 6 7 M Fig 5 They only differed from each other in their 0 2 kb upstream flanking sequence FDI 2 7 M and FDI 6 7 M had 0 2 kb upstream flanking sequences from the up stream flanking sequences at genes 2 3 and genes 6 7 respectively Fig 5 In repeated experiments the genomic DI RNA from the two DI RNAs Figs 4 and 6 replicated with very similar efficiencies The difference between the subgenomic DI RNAs of FDI 6 7 M and FDI 2 7 M however was striking FDI 6 7 M subgenomic DI RNA was significantly lower than FDI 2 7 M sub genomic DI RNA Figs 4 and 6 The structures of the subgenomic DI RNAs of both DI RNAs were expected to be the same because both DI RNAs had the 12 nt se quence and had the same 0 2 kb downstream flanking sequence Sequence analysis of cloned RT PCR prod ucts of subgenomic DI RNA showed that both sub genomic DI RNAs had the helper MHV derived leader sequence and the same leader body fusion site data not shown the difference between the leader se quences of FDI 6 7 M and FDI 2 7 M and that of helper virus allowed for the identification of the origin of leader sequence in the subgenomic DI RNA These data dem onstrated that the subgenomic DI RNAs of both DI RNAs indeed had the same structure FIG 2 Northern blot analysis of DE5 w3 derived insertion mutants Intracellular RNAs were extracted from DI RNA transfected MHV infected cells lanes 4 5 11 14 or DI RNA transfected mock infected cells lanes 2 3 6 9 Lanes 1 and 10 represent RNA from MHV infected cells Total cytoplasmic RNAs lanes 1 5 or poly A containing RNAs lanes 6 14 were analyzed by Northern blot analysis using probe 1 see Fig 1 Arrowhead and arrow point to subgenomic DI RNAs and genomic DI RNAs respectively The minute amount of genomic DI RNA sized signal in lanes 2 and 3 represents the transfected DI RNA transcripts that were not degraded in MHV uninfected cells 201 CORONAVIRUS TRANSCRIPTIONThese data indicated that the molar ratio of sub genomic DI RNA to genomic DI RNA was determined at a transcription step s that occurred prior to or concom itant with the synthesis of subgenomic sized DI RNA of either polarity containing leader sequence The follow ing are the logical bases for the previous statement I The similar replication efficiencies of FDI 6 7 M and FDI 2 7 M indicated that similar amounts of helper virus derived and host derived factors were available for DI RNA synthesis of both constructs II Efficiency of puta tive RNA synthesis from subgenomic sized DI RNA tem plate containing leader sequence of those identical sub genomic DI RNA probably is the same whether they were positive stranded negative stranded or both If sub genomic DI RNAs accumulate by replicon type amplifi cation Sethna et al 1989 then these subgenomic DI RNAs with identical structures probably amplify with the same efficiency Any difference in the subgenomic to genomic DI RNA molar ratios of the two DI RNAs there fore should occur prior to a possible replicon type am plification One of the transcription steps that possibly deter mines the ratio of subgenomic DI RNA to genomic DI RNA is the step that synthesizes biologically functional genomic sized template RNA genomic sized template RNA may be either positive stranded or negative stranded FDI 6 7 M might have produced significantly less genomic sized transcription template RNA than did FIG 3 Schematic diagram of the structure of MHV genomic RNA DE5 w3 and DE5 w3 derived insertion mutants with flanking sequences The 0 2 kb flanking sequences that were inserted into DE5 w3 derived insertion mutants are shown as 0 2 kb fragments All mutants contained the 12 nt sequence in the middle of the inserted sequence Molar ratio sG G represents the average molar ratio of subgenomic DI RNA to genomic DI RNA from at least three independent experiments see Fig 4 The location of probe 1 used for Northern blot analysis see Fig 4 is also shown 202 AN AND MAKINOFDI 2 7 M the sequence differences at the upstream flanking sequence of these two DI RNAs may be respon sible for the differences in the amounts of biologically active template Quantitative comparison of negative strand or positive strand DI genomic RNA of FDI 6 7 M and FDI 2 3 M may not reveal the difference in the amount of biologically functional template both DI RNAs produced similar amounts of positive strand DI genomic RNA Figs 4 and 6 and may produce similar amounts of negative strand DI genomic RNAs yet the amount of biologically functional genomic sized template RNA that is used for transcription may differ between the two DI RNAs To examine whether FDI 6 7 M and FDI 2 7 M synthe sized different amounts of genomic sized RNA templates that function for transcription we constructed and char acterized FDI 2 7 M dIG and FDI 6 7 M dIG Fig 5 FDI 2 7 M dIG and FDI 6 7 M dIG each had an additional 12 nt sequence located 0 2 kb downstream of the 12 nt se