Sequence of cowpea chlorotic mottle virus RNAs 2 and 3 and evidence of a recombination event during bromovirus evolution

VIROLOGY 172,32 l-330 (1989)

Sequence of Cowpea Chlorotic Mottle Virus RNAs 2 and 3 and Evidence of a Recombination Event during Bromovirus Evolution

RICHARD F. ALLISON, MICHAEL JANDA, AND PAUL AHLQUIST’

Institute for Molecular Virology and Department of Plant Pathology, University of Wisconsin-Madison, Madison, Wisconsin 53706

Received April 13, 1989; accepted May 22, 7 989

The genomic sequence of cowpea chlorotic mottle virus (CCMV) was completed by sequencing biologically active cDNA clones of CCMV RNA2 (2774 bases) and RNA3 (2173 bases). While only the central core of the encoded 94-kDa CCMV 2a protein contains features conserved among known and putative RNA replication proteins from many viruses, both flanking regions of CCMV 2a show substantial similarity to the corresponding protein of the related brome mosaic virus (BMV). The 3a proteins of CCMV and BMV, implicated as contributors to the distinct host specificities of the two viruses, show lower levels of conservation but are still discernibly related throughout. Major differences occur in the organization of noncoding sequences in CCMV and BMV RNA3. With respect to an otherwise similar region preceding the BMV 3a gene, the CCMV RNA3 5’ noncoding sequence contains a clearly bounded 111 -base insertion that must reflect a sequence rearrangement in evolution of at least one of the two viruses. The presence of a subgenomic promoter-like sequence near the end of the novel CCMV sequence makes the organization of genes in CCMV RNA3 reminiscent of the 3’end of tobacco mosaic virus RNA, suggesting that CCMV or its 3a gene might have been derived from an ancestor with fewer genomic RNAs. Sequence similarities between the CCMV and BMV RNA3 intercistronic regions include the subgenomic mRNA promoter and an oligo(A), but not an intercistronic segment required for BMV RNA3 amplification, implying that replication signals on the two RNA3s may be organized quite differently. o 198s Academic Press, Inc.

INTRODUCTION

The bromoviruses are a group of positive strand RNA viruses with isometric virions and tripartite genomes (Lane, 1981). Monocistronic RNA1 (ca. 3.2 kb) and RNA2 (ca. 2.8 kb) encode the nonstructural la (104 kDa) and 2a (94 kDa) proteins, respectively. RNA3 (ca. 2.2 kb) encodes the nonstructural 3a protein (32 kDa) and the 20 kDa coat protein, which is translated from subgenomic RNA4. The best studied member of the group is brome mosaic virus (BMV). The complete BMV genome has been sequenced (Ahlquist et a/., 1981 b; 1984b) and can be manipulated through biologically active cDNA clones from which infectious transcripts can be synthesized (Ahlquist and Janda, 1984; Ahlquist et al., 1984c; Janda et a/., 1987).

Among other uses, BMV has been intensively studied as a model system for RNA-dependent RNA replication in eukatyotic celis. The la and 2a proteins, which are transacting factors in viral RNA replication (Kiber- stis et a/., 1981; French et a/,, 1986; P. Kroner and P. Traynor, unpublished results), share extensive sequence similarity with noncapsid proteins of a wide range of other positive strand RNA viruses of animals and plants, suggesting broad conservation of funda- mental aspects of RNA replication (Haseloff et al., 1984; Cornelissen and Bol, 1984; Ahlquist et a/.,

’ To whom requests for reprints should be addressed,

1985). BMV RNA signals required in cis for (-) strand RNA initiation in vitro (Dreher and Hall, 1988), for complete viral RNA amplification in viva (French and Ahl- quist, 1987) and for subgenomic RNA synthesis in vitro (Miller et a/., 1985; Marsh et a/., 1988) and in viva (French and Ahlquist, 1988) have been defined and characterized. The presence of tRNA-related sequence elements in several genome regions associ- ated with replication suggests that a host factor(s) may also participate in viral RNA synthesis (Rezaian et a/., 1985; French and Ahlquist, 1987; Marsh and Hall, 1987).

We recently constructed complete, biologically active cDNA clones for a second bromovirus, cowpea chlorotic mottle virus (CCMV), and exchanged individual genomic RNAs between BMV and CCMV to dem- onstrate virus-specific interactions at several important steps of infection (Allison et a/., 1988). Heterologous combinations of BMV and CCMV RNAs 1 and 2 failed to direct RNA replication, showing that RNA2 and/or its encoded 2a protein encodes a specific determinant(s) for compatible interaction with the homologous RNA1 RNA3 exchanges displayed highly asymmetric levels of RNA3 amplification, which must reflect differences in &-acting features of RNA3, frans-acting replicase characteristics encoded by RNAs 1 and/or 2, and their interaction to determine replicative specificity. Deter- minants of host specificity in systemic infection are contained in RNA3 and elsewhere in the genome (Ban-

321 0042-6822/89$3.00 CopyrIght C 1989 by kademlc Press. Inc

322 ALLISON. JANDA, AND AHLQUIST

croft, 1972; Allison et al., 1988). Further studies, including designed sequence exchanges between CCMV and BMV, should thus help to define the specific interactions among the viral and cellular components which control these steps in infection.

The 3’ terminal CCMV RNA sequences, the sequence of subgenomic RNA4, and the extreme 5’ends of RNAs l-3 have been reported previously (Ahlquist eta/., 1981 a; Dasgupta and Kaesberg, 1982; Allison et a/., 1988). The complete sequence of CCMV RNA1 has also recently been determined (Jozef Bujarski, personal communication). We report here the full sequences of CCMV RNAs 2 and 3, determined from biologically active cDNA clones. Among-other features, otherwise similar regions preceding the CCMV and BMV 3a genes contain two blocks of sequence which are contiguous in BMV but separated by 111 bases in CCMV. This novel CCMV segment positions a subgenomic promoter-like sequence just 5’ to the 3a gene, which has possible implications for bromovirus evolution. The CCMV 2a and 3a proteins show extensive sequence similarity throughout their length with the analo- gous proteins of both BMV and cucumber mosaic virus (CMV).

METHODS

The complete CCMV RNA2 and RNA3 cDNA inserts in plasmids pCC2TP2 and pCC3TP4 (Allison et al., 1988) were sequenced by subcloning fragments in Ml 3mpl8 or Ml 3mpl9 (Yanisch-Perron et al., 1985) and applying the dideoxynucleotide method (Sanger et a/., 1977; Biggin et al., 1983). All sequences were determined on both cDNA strands. pCC2TP2 subclone libraries were constructed by linearizing the plasmid with either Sacl (5’ cDNA end) or Xbal (3’ cDNA end), treating for varying lengths of time with Ba131 exonu- clease followed by DNA polymerase I large fragment (Maniatis et a/., 1982) cleaving adjacent to the oppo- site end of the cDNA insert, using Pstl for the Sacl library or EcoRl for the Xbal library, and subcloning into Smal-, Pstl-cut Ml 3mpl9 or Smal-, EcoRI-cut M 13mpl8 as appropriate. Subclones were selected

for sequencing by dot-blot hybridization of phage cul- ture supernatants to pCC2TP2 and ssDNA sizing on 0.8% agarose gels (Messing, 1983). pCC3TP4 subclones were made by cloning various restriction fragments. The contiguity of adjacent restriction fragments was verified by sequencing across the relevant restriction site on one or more overlapping clones. RNA sequencing to establish the 5’ and 3’ ends of the CCMV RNAs has been described previously (Ahlquist et a/., 1981 a; Allison et al., 1988). Sequences were assem- bled and analyzed with the assistance of programs from the University of Wisconsin Genetics Computer Group (Devereux et al., 1984).

RESULTS AND DISCUSSION

Plasmids pCC2TP2 and pCC3TP4 contain complete cDNA copies of CCMV RNAs 2 and 3, respectively, fused to a T7 RNA polymerase promoter. The mixture of in vitro transcripts from these clones and from pCClTP1, a similar CCMV RNA1 cDNA clone, is infectious-to both protoplasts and whole cowpea plants. This set of clones has been selected to define a stan- dardgenome, designated the “M 1” strain of CCMV, for further studies (Allison’& al., 1988). The complete sequences of CCMV RNAs 2 and 3 were determined from pCC2TP2 and pCC3TP4 as described under Methods, and are presented in Figs. 1A and 1 B. These sequences, together with the previously determined sequence of CCMV RNA1 (J. Bujarski, personal communication), complete the sequence of the CCMV genome.

Relation of CCMV 2a to other proteins

The CCMV 2a protein shows strong sequence similarity with the 2a proteins of both BMV and CMV (Fig. 2A). As previously described for BMV and CMV, certain regions of these three 2a proteins also show significant sequence similarity with RNA-dependent RNA polymerases from poliovirus, phage Q@, and infectious bursa disease virus, a dsRNA birnavirus, as well as with proteins encoded by the animal alphaviruses and coro- naviruses and diverse plant viruses (Haseloff et a/.,

FIG. 1. Sequences of CCMV RNAs 2 and 3. (A) Sequence of CCMV RNA2 as determined from the biologically active cDNA insert in pCC2TP2. Initiation and termination codons for the 2a gene are marked by double underlining and the 2a protein sequence is given in Fig. 2A. U residues at positions 2561 and 2600 in the cDNA sequence differ from previous RNA sequencing results, which found Gs at the corresponding positions (Ahlquist ef a/., 1981 a). The first change occurs 5’ to the 1 go-base 3’ terminal sequence (underlined) whose secondary structure has been described previously (Ahlquist eta/., 1981 a). The second change allows replacement of an internal loop by an A. U pair in the stem of a conserved hairpin, enhancing the structural similarity of this RNA2 region to other bromoviral and CMV RNAs. (6) Sequence of CCMV RNA3 as determined from the biologically active cDNA insert in pCC3TP4. Initiation and termination codons for the 3a and coat protein genes are shown by double underlining, as well as the adjacent initiation and terminations preceding the 3a gene (bases 38-43; see text). The 3a protein sequence is given in Fig. 2B and the coat protein sequence in Dasgupta and Kaesberg (1982). The 1 1 1 -base insertion with respect to the BMV RNA3 5’noncoding sequence (bases 1 13-223; see Fig. 4A), the intercistronic oligo(A) (bases 1292-l 331) and the 190 b of 3’terminal sequence whose secondary structure has been described previously (Ahlquist et a/., 1981 a) are underlined. The G residue at position 1350, marked by an inverted triangle (v). corresponds to the capped 5’ end of subgenomic RNA4. The A residue at position 1810 in the cDNA contrasts with a G found in the corresponding position of RNA4 by Dasgupta and Kaesberg (1982) altering amino acid 150 of CCMV coat protein from alanine to threonine.

A m'G~~~GUAAUCCAC~AGAGCGAGG"UCAAUCCCUUGUCGACUCACGGGUCUCCAUCAG"UGAkAACAGUUUAUACAUUUUC"UCUUGAUAUUUUUCIJUCUUUACU

101 UCCAUUAA"APGUCUAAGU;CAUUCCAG~~UGAGACU;ACCACGUUCCCUCAUUCChAUGGAUGUUUtAUCAGACUC;CG~UCUGACUCACACCAUG -2a-B

201 AUGAGGCGAUAUUCGUFIAC~GAnUCGA"U~UGAAAGU~AGUUGAUAC;UCUGUUGkAAU~CCGCAG~UGGCACGCUAGCAAGUUAUAUGCAUGCCG~

301 AAAGCCCCUAGUGGAGGAUGGUCUUCUGPAUCCCCCUUU;UAGAUGGGGUCUUUGCUGCAAGAACGUCGUUGACGUUUAUGACGGGCUGCUC

401 GGUUAUAGACUCAUACCAAUGGCUGAAGCCGCUAGAAUGUUGUACUUGGACCGG

501 UAGAUACCU~UGAUGGUUUCACCGAAGCAAUGUUUGAUG;GAUGAAUGA;;AUUCCUGGCGAC APAAAAUACAUGCGCUUUAAGUCU~JGAAGCUGA

601 AUCAAGGC~GCUCCAGAAACUUCCGAUA;GGUGCCGUCUGAAUAUACGUUGGCAGAUAGGUACGUUACCACCAGAGAGGAGUUCGCGUCUGUUGACUCG

701 GAUUAUGAC;IUAUCCUUPCCUGGUGAGCCCUGUGGAG;UCAGGGUGGGAGUGUGUGAAGACACAUAC~GUCAUUCGG~GCUGAUGA~C~:UACGAUGE

801

901

1001

1101

1201

1301

1401

1501

i6Ol

1701

:a01

CUCM"AUCACGAUAGGAUCAGUUUAAAA;CGCUGGAGGCGUACCGACUCAUGCCUAUUUUGACGACACUUACUACCAGGC"U;

GGAAGAGCUAGGCGAUUAU;UGUCGAUA;UAG"~GUU~UCUGUCCGGCAGAG"GAUG;UGAUUGGUA;CGUGACCCUGAAAAGUACUAUGAGCCUGAG

GUCCAGAGGACUUGAGUACCAUAAGAAAUGGAAAGACCACCAC~GACCUGACUGGUGUGACAGUUUUGUCUGAGAUU~UUUGCAGA~UAUCAGCACAUG

AUAAAGUCUGAUAUUAAACCAGUUGUCUCGGAUACGUUACACCUCGAACAUUUCAUGGUAhAGGAGUUACUAGCUGC;

UCUCACCAUAUUUUACGGC;UGUUUCGAGAAGUUUUC~GCUUU~UCAAGGUUUGUGG"CCCCA;AGGGAAGAUCUCCUCCCUGGRACUGAAAAA

UGUUCCCCU~UCGAAUAAA;GGUUUCUUG;iGGCGGAUUU~AGUUUUCUCAGGGUGAGCUUCAUCUUGAGUUCCAIWGAGAGAUAUUGUU;;

UCAUUGGGUUUUCCAGCCCCUUUGACUAAUUGGUGGUGUGAUUUCCAUAGGGAAUCUAUGCUAUCGGAUCCUCAUGCUGGAGUUAACAUGCCAGUUUCC;

UUCAGCGUCtUACUGGUGA;GCUUUUAC";AUUUUGGGMUACUU"GGUGACUAUGGCCAUGAUGGCCUAUUGUUGCGAUAUGAACACCGUGGACUGUGC

UAUCUUUUCCW;UGAUGAU;CUCUGUUAA;UUGUAAAAG;FIAACCACAUCUGGAUGCUAAUGUUUUUCAAUCUCUGUUUAAUAUGGAAAUUAAAGU"AUG

1901 GACCCAAGU;UGCCAUACG;UUGUAGUAAtUUUCU"U"A~~CUGAAAUG~UAACUUGGUG"CUGUGCCUGAUC;UAUGAGAGAGA"ASAGAGACUG;;

ZOO1 CU~9AGCGAAAGAUCAUCAAAUCGCCUGAGUUGUUAAGAGCCCACUUUGAGUCCUUUUGUGAUAGGAUGFlRAUUCCU~C~UUGGAUG 'AAAAAAUGA;

:101 AAAUUUAUUkJGCAAGU"U&lGGCUCUCA;IGUAU& CCUGACGUUGAAAACGAUG;CAGAGUAGC;‘AUUGCUGCU;UCGGCUACU;ICUCAGAAAA;

::01 UUCUUGAGAUUUUGCGhAUGUUAUGCGACUGAhGGGGUCAAUAUAUAUAAGGUACAUCCCAUCAC'~CAGGAGUGGUUCGAGGCCUCUA~GGAUCGAG

:301 ACGGUGACUGGUUCCAUGACUGGCGUAAUCCGAAGUUUCCCACUGCCUUAGAUAAGGUUUGG""UGGAGAU"CUUUG~UACGC'~AGAGAUGAUCCUAUGAA

2401 GCACAUAG~GAGAGAGAUAGGAGACAUAGACAUAGGCUU~UCGAGCCAUGhAUUCUUCCUU~AAAZUUGCCUAUGAUCGUAGGAAIUCUUAGU~GGAA

1591 ACCGUUGCG;I;GGUGCGU~GACCCUUUCIi~~UGU;GGUCACAUU;iylGACUWGU;UAGUCCACA;UAGGACUGG;UCUAACAGU;UCUUUAAAC;

2601 G"AAUCGUCG"UGCGACGUUGGUUUGCUUACA;IGCAAUC.4AGCUGCCUUUGAGUUUUACUCCUUGAACUCUUCAGAAGMUUCUUCGG;IAUUCGUACCAG

2'01 UAUCUCACA;;IGUGAGGU~"~~GACUG~~~GCAGC~~C~UAGUC;~~AC"AGGUGA;CUCUAAGGA~ACCA :--4

201 A"A"CAUAA;UCCUCGUUC;"U~"GUUA;AGC"CCCG~~C"~CAC~AC"UUUAGA~CU"U"ACU~"UCCUCCA~ACCGUGGUC~AGGGAG~C~ 7-w

301 AGCCGGCGCCCAGGAUGAUAUG"CGUUGUUACAG"CACUU""""CCGAC~UCCAGGGI;GGAGUU"GC;AAGGAGUGU~GU"~GUA~GUAUACC~~

401 U"AUCCUC"~"AACCGGCU"AA"UAUA"AGA"C"AGUCCCCAAGAACACUGGUAG"AGAGCUCUGAAC"UAUUUAAG"5AGAGUAUGAAAAAGGUCACA

501 U"CCCUCCAGCGGUGUGCUUAGUA"ACCUAGAGUGC"GG"UUU"CUUGUGAGGACGAC~CAGUGAC"G~"C"GGGAGUGUCACCAUUAGA""~UU~

601 CUUGA"AAG;:GCUUCG"C~UUGAGAU"U;AGAACCUGU;;GAUGG"ACG~~GA~C"A~UAUUCC"AU;UCUAGUCUU~CG~UAUCG;""GUU"U"C;

701 CC"AGUUAU~ACUGUCCCA~GCAGAUGAU~~G~"AGA~ACAGAUGUU~CGG"U"~U~CUCAACU~A"GGUGUCA~A"CCUCAG~UC"ACCG"C~

801 UUAUGAG"CT;UGCGUAU"G~UCUGCGAAC;UUCGUAG"~CCU~"~~UACAAGCAG;ACGCACCUA;G"AUAAG"A;GUGGAACCC;U"GACAGG";

901 GAAACGUUU~AGCCGUAAA~AAUUG~UUAUGUUA~GGCAU~CG~UC~UCUGU~UCAUGGUUAUCUAUU~GU~CCAC~ACUG~GAC~

1001 GACGAGCAAGAUCCAGAGA;GAUUGUGU"~GAGGAGGAAUACCGACAGCAACGGAGUCGGCAAGGAUAAAAUCGCCGUAACCGCU~~

1101 CCG""GCGG~GCUWCCGAC~GCCUCGUUG;CUAUCAAUC~UAGA~U~GA"CUA"UG~UAACAGACU;AGUGUCA"~GAGCAAAAC;CAUAUGAGG; a.%? -

1201

1301

1401

1501

1601

1701

1801

13Ul

:001

‘?ll)l

324 ALLISON, JANDA, AND AHLQUM

1984; Cornelissen and 601, 1984; Kamer and Argos, 1984; Rezaian et al., 1984; Gorbalenya and Koonin, 1988). The most heavily conserved core of these similarities includes an Asp-Asp pair that is found in similar sequence context in retroviral reverse transcriptases and may be functionally related to an Asp-Thr-Asp motif in many DNA-dependent DNA polymerases (Kamer and Argos, 1984; Argos, 1988).

While similarity with other viral proteins is confined to a central core region (see brackets in Fig. 2A) whose limits correspond to the physical boundaries of some other viral proteins (Haseloff et al., 1984) sequence similarity among bromovirus and CMV 2a proteins extends over nearly the entire 2a protein length. While many of the conserved residues of the 2a protein core might contribute to catalytic functions in RNA polymer- ization, portions of the ends and some interior regions may contribute to other functions, such as proper interaction of the 2a protein with other factors in RNA replication. The specificity of such interactions might un- derlie the inability of CCMV and BMV RNA2 to direct RNA replication with RNA1 from the converse virus (Al- lison et a/., 1988). The extent and distribution of differences between the CCMV and BMV 2a proteins (Fig. 2A) are thus of particular interest because of their possible role in virus-specific interactions, and the possible use of such differences to reveal specific interactions in the RNA replication process.

Relation of CCMV 3a to other proteins

CCMV RNA3 encodes the 3a and coat protein genes. The relation between sequences of the RNA3- encoded CCMV and BMV coat proteins has been discussed previously (Dasgupta and Kaesberg, 1982) as well as their ability to encapsidate either CCMV or BMV RNAs in vivo (Allison et al., 1988). The 3a gene, whose biochemical function is unknown, is dispensable for RNA synthesis (French and Ahlquist, 1987). However, expression of a functional 3a gene is required for systemic infection by CCMV and BMV (R. Allison and R. Sacher, unpublished results), and the bromovirus 3a gene is circumstantially related to the 30-kDa gene of tobacco mosaic virus (TMV) (Haseloff et al., 1984) which is required for cell-to-cell movement of TMV (Zimmern and Hunter, 1983; Meshi et a/., 1987; Deom et a/., 1987). Bromovirus 3a genes thus might contribute significantly to the host specificity shown by RNA3 for systemic infection (Bancroft, 1972; Allison et al., 1988).

With 52% identity, the 3a proteins are the least conserved proteins encoded by BMV and CCMV. Never- theless, BMV and CCMV 3a are similar to each other and, to a lesser degree, to the CMV 3a protein (Davies and Symons, 1988) throughout their entire length (Fig.

2B). Previously noted similarity between the 35-kDa protein of red clover necrotic mottle virus (RCNMV) and amino acids 181-234 of BMV 3a (Lommel et a/., 1988) also extends to CCMV 3a. Moreover, the preceding 70-75 residues of the bromovirus 3a sequences also show similarity with the corresponding portion of the RCNMV 35kDa protein, with stronger similarity between RCNMV and CCMV (Fig. 2B). Dot matrix comparisons failed to reveal any extended similarity between the bromovirus 3a proteins and the 30-kDa proteins of alfalfa mosaic virus M and S strains (Barker et al., 1983; Ravelonandro et al., 1984) and tobacco streak virus (Cornelissen et al., 1984) or the 30-kDa proteins of the TMV common, L, OM, and Cc strains (Goelet et al., 1982; Takamatsu et a/., 1983; Meshi et al., 1982a,b). The relationship between the CCMV 3a protein and TMV Cc strain 30-kDa protein was also ex- amined by direct sequence comparisons, since these viruses share a common systemic host, cowpea. Com- puter-optimized alignments displayed weak matching throughout the two proteins. However, while the level of similarity found exceeded the average simila’rity of the two sequences after repeated shuffling, it did not satisfy the usual criteria for statistical significance (Doolittle, 1981).

Within the coding regions, nucleotide sequence similarities between BMV and CCMV RNAs 2 and 3 largely follow the above relationships of the encoded proteins, The more complex sequence relationships among noncoding regions in BMV and CCMV RNAs are described below.

5’terminal sequences of the CCMV genomic RNAs

The 5’ ends of CCMV RNAs 1, 2, and 3 display significant similarities with the genomic RNAs of BMV and CMV (Fig. 3). In interviral as well as intraviral comparisons, the strongest conservation is seen among RNAs 1 and 2. Within each of the three viruses, the 5’ 40-45 bases of RNAs 1 and 2 are nearly identical, while the RNA3 sequences are more distinct, showing greater similarity in inter-viral than intraviral comparisons. As with BMV and CMV (Rezaian et a/., 1985; French and Ahlquist, 1987) the conserved motifs at the 5’ ends of CCMV RNAs 1 and 2 include the ‘element GGUU- CAAYCCCU (Y = pyrimidine), which corresponds to the consensus “box B” recognition sequence of RNA polymerase III promoters (Marsh and Hall, 1987) and thus also to the conserved residues of the T*C loop of tRNAs. Partial agreement with this consensus is present within the various RNA3 5’ends, especially for BMV (Fig. 3). Moreover, strong matches to the consensus motif occur in the intercistronic regions of BMV and CMV RNA3 (French and Ahlquist, 1987; Davies and Sy- mons, 1988) but not CCMV RNA3 (see below). The

SEQUENCE OF CCMV GENOMIC RNAS 2 AND 3 325

A 1 72

. . . . . . . . . . . . . MSKfipEgetyhVPSFQYhnfDQtLEsDshhdeaIfvtEsinesgMtsVEltA..........................DGtLASYmhAVItPLvedGL

. . . . . . . . . . . . IdsSKtWdDDfvrqVPSFQWiiDQsLEdE........ .VE.aaSLqVqEPaDgvAI.........................DGsLASFkWaPLeIGGV mlsppptfsfanLLngsYgvDtpeeVervrReqredaEaalrnykpLpa~vseSVprDEPIvsqtVtaapvtsvddafvsfgaedylemspselLsaF~~~LrVG@V

73 178 LnPPFDqaR~lcCknVvdV..ydgLlgyrLIPmAE~~IdGS~dEsKcDDWrP~TSDGFte~fd~eip~Etknt~lsleaEsRqap...ETsdmvPsE FdPPFDRvRWGSiCdtVqqIdV..qqFtdRpLIPq ~~IpGS~DKiDDWyPeDTSD~GvsFaadedHaSdLklasdssnUIeKvRvtg...DT....PkE LcssFDRslFiSsvamartUlapltstRtLkrfeDLvaaI~k..tdFfLEDDgpq.tdVsqSDvpGy~Fepgq.HsSGFEpppiCAkWL~lyqcpcfdfnalrescaE

507 *t 617 NWWWFHRESmLSDPHAGVnMPVSFQRRTGDAFTYFGNTLVTI.MbA YCWMntVDCAIFSGDDSLLIcKsKPhLDanVFqSLFNtdKIKVMDPSLPYVCSKFLLKTFMnNL NWWsDFRRDSnSDPHAkVBVSFQ~T~~~~~AYasDLsdcD~IFSGDDSLIIS~VLDtdmFTSLF~I~PSVPYVCS~L~T~g~ kWWWPHRfSYIkDkr~GMDISFQRRTQAFTYlrGNTIVTMAefAWCYDtdqfDrLLFSGDDSLaFSKLPPVGDpskFTTLFNMEaKVMEPaVDYICSlCFyslmsLvt~

618 VSVPDPMRKIQRLAKRK

=I 721 II..KspEURAHFeSFW~~EI(MInlLCkFVaLlCY.....klCPdVenDVRvAI~gyYSE~LRcECYaTtCVNIYkVkhPItq

VSVPDPLRKIQRLAKRK IL..RdeqblLRAWvSFWMlKJ'INqLDElQdIttLChFVyLKY...GkEKPwIFe~ s1YSENFLRZSDCYcTEGIrVYqmSDPVck fqsP.tIREIQRLgt~IpysdnnDFLfAHhnSFvDRL~RldsqscIdqLsiFfeLRlkksGnEaalVLgaf~ytAnFnaYkElY..YSDrqqcDl~tFcISE.fr~

11 722 808 .eWfeasRdRDGD~Hd~~Pta~~fGkYarDdp~IeeR~rh~nr~.sSLICLAyDRRs. ..LsKdkKtVAwvRktLsk................... fRRTTeeRKtDGD~Hn~~PgvtDICVYRtiG~YssDcStKeLpvK.RigRLheALeRESLKWLnDRttqrLltKKvDdyAtgRggLtsvdallvkshcetfkpsdlr ~RRTTvkKRKnG........cvdssgVDRrppLsqfaggEtSktkVsrqkpaseglqksqRESaiysetfpdvtipRsRsrgLvs.........................,

B 1 109 MsNt .tfrPFtGSSRTwS.GeQAG*qdDmsLLqS~SD~reKfUcKCeCICUimYtnLsS~r~IDLVPl[nTgSraL~~e~~hIPSSGvLSXPR~WL~~tV MSN. ivsPFsGSSRTtsDW;k~SDKkLIeSLISI~~jKC~cYn~epr~IDLWl[shvSawLsWatSk'IDM;UPS RGPMWPRIVCFLVRTTds . . . . . .maFqGpSRT...LtqQssAaSsDdLqkiLISpdAI~tkmAtdCd~rhh~~aIsvrpLWqvTsnnlLpFFl[SgPDaG~~rS~ qVLCaVtRTVst . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..Sva~~h~sLapTIgV

110 215 tKSGSVTIrLVDLIsAssveILKPVDGtQKATIPISsLPAIVcFSPSYDCPMqwIG.. .~~GLVTPLdGVIssGs~~~ms~~~~.~~kyV ~SGSITVsLcDsG~ra~aIDn.Q~TIqLSaLP~IaLtPS~~~~dsg~~FGIa~~~TtG~A~~~~~~S.gPatiMV daeGSLkIyLaDLG . . ..DkeLsPIDG.QcvTLhnheLPALIsFqPtYDCPMKLVG . ..NRRRCFaWVerhGyIGyGGTtAsvcsnWPlqlsSL[nMIY.RaAagktLV AipGhVT.. .VELInpnvlGpFqvmsGqtlswsPgagkPcLmiFSvhhqlnsD....... .RepFrVrItntG.IpTkkSyArcRAYWgfdvgtRhryYKsepArLieLe

216 302 ePFDRLKrLSRK.qLKNYVRG ItNQSVnBOYLLG~LLktDeqdp~I~KesltptdsnGVg~kiaVtAk~LPtaSlsInrR ~FDRLRqLdKKS.LKWYIRGIsWQSVdHGYLLG RPLqsvDqvaqEdLLVtESesPsALGrGV.KDskSVsAsSVAGLPvsSptLriK LPYnRLaehS~~Varl~sqlNnvsssr~LpnvaLnqnasghGseILkKS.pPiAIGspsasrnnSfrsqvVnGL.......... VgYqRtllsSiK.AVeaW.....................................................................

FIG. 2. Comparison of the 2a and 3a proteins of CCMV, BMV, and CMV. Amino acrds are printed in bold caprtals whenever the srmilarity among two or more residues In a gtven column equals or exceeds 0.8 (the similarity between Land V) usrng the amino acrd substitution table of Schwartz and Dayhoff (1979) as normalized by Gribskov and Burgess (1986). (A) Comparison of the 2a proteins of CCMV (first line), BMV (second line), and CMV Q strain (third line). Percentage identity in the pairwise alignments (includrng gaps) are CCMV vs BMV, 60%; CCMV vs CMV, 33%; and BMV vs CMV. 36%. Numbering shows position in the CCMV 2a sequence; BMV and CMV 2a are 822 and 839 residues long. The central region displaying similarity to Sindbis virus nsP4 (see Fig. 2 of Haseloff er al., 1984) is delimited by brackets. The highly conserved Asp- Asp pair found in similar contexts in many other viral proteins, including known polymerases (Kamer and Argos, 1984; Argos, 1988), is marked by asterisks. (B) Comparison of the 3a proteins of CCMV (first line), BMV (second line), and CMV Q strain (third line) and amino acids 60-l 94 of the 35.kDa protein of red clover necrotic mottle virus (fourth line; Lommel et a/., 1988). Percentage identity (including gaps) in the pair-wise alignments of the 3a proteins are CCMV vs BMV, 52%; CCMV vs CMV, 33%; and BMV vs CMV, 33%. Numbering shows posrtion In the CCMV 3a sequence; BMV and CMV 3a are 303 and 279 residues long.

BMV intercistronic copy of this consensus motif occurs within a region required for efficient amplification of BMV RNA3 in infected cells.

While the 5’ noncoding regions of BMV and CCMV RNA2 have similar overall lengths of 103 and 109 bases, the more divergent 5’ noncoding sequences of

CCMV and BMV RNA3 contain 238 bases and 91 bases, respectively. Most of this length difference is accounted for by a clearly bounded 111 -base segment that occurs just 5’to the CCMV 3a gene and separates two sequence blocks matching contiguous portions of BMV RNA3 (Fig. 4A; see also Fig. 1 B). Comparison of

326 ALLISON, JANDA, AND AHLQUIST

(PolIII Box B/T'?C Consensus) GGwcAAnncc ******* **

CMV 1 m’GpppGVuuVaWu ACaAGAgCGuacGGWCAAcCCCV GcCucCvCugua CMV 2 m’GpppGVuuauV CuCaAGAgCGuauGGWCAAcCCCV GcCucCVCuguG

CCMV 1 m’GpppGVAAV CCACGAGAACGA GGWCAAUCCCWGVCGACVCACGG CCMV 2 m’GpppGVAAV CCACGAGAgCGA GGWCMUCCCWGVCGACUCACGGgVCVcC

BMV 1 m’GpppGVAg ACCACG GAACGA GGWCAAVCCCWGVCGAC CACGGWCVGC BMV 2 m’GpppGVAA ACCACG GAACGA GGWCAAVCCCWGVCGACcCACGGW VGC

BMV 3 m’GpppGVAAaaV ACCAacu AAuucuc GWCgAVuCC GgCGA ACa WCVau --- - --- -- ccw 3 m’GpppGVAAVcWuACCAaac AACu --

m’GpppGVAAVcW ACCAC WCAAaCuuV atJa+lJa uGuaguVGCuguq -

CMV 3 uuucu WCAcgs gVGVCGcgI$A GuccacGCuguq -- -

FIG. 3. Alignment of sequence similarities in the 5’sequences of genomic RNAs 1, 2, and 3 of CCMV, BMV, and CMV. Residues belonging to a consensus of more than half of the bases within a given vertical line are shown in bold capitals. Similarities among the RNA3 sequences of the different viruses are emphasized by underlining. The consensus sequence of the polymerase III box B sequence and tRNA T@C loop is shown above the viral sequences, with similarities to the viral sequences indicated by asterisks. The 5’sequence of CCMV RNA1 was derived from direct RNA sequencing and clone pCClTP1 as described previously (Allison et al., 1988). For all of the RNAs shown, U-rich sequences follow the regions displayed.

this CCMV-specific segment to the total sequence of BMV RNA3 shows that its 3’ end has 70% identity with the functional core element of the BMV subgenomic mRNA promoter (Fig. 4A), which is discussed further below. No significant relationship between the remainder of the CCMV block and BMV RNA3 was found.

A 5' Noncoding Sequences

Buv3 &pppGUML..(48 I I I I

CCMV3 rn'Gpp~GVAA...(83

Intercistronic Sequences

gUAaAUCcggU CVAACAagCUcgGU CCAU...(167

II III I IIIIII II II III cUAgAUCuauUgCUlUCAgaCUuaGUguCAU...(113

Although it is not possible to tell from this information whether the BMV sequence was derived from a CCMV-like predecessor by deletion, or the CCMV sequence from BMV via insertion, the observed differences make it clear that a sequence rearrangement at this position has contributed to bromovirus evolution.

* *

b.)...ACAUCGGWUW ucAGLGuAgu GAUA CUGWuUuGuUCCCG /II/IIIIIIlI II I III/ IIIII I I IIIII

b.)...ACAUCGGWWUgaAGcAucgGAUA CUGWaUaGcUCCCG

' 179 b.) . . .MuAGuUCgAUaUCaUMWCcuC GUucWUg '

II II II II II IIIIII I II II:

I

' core Subgan&c Promoter

B

cCMv3

BMv3

m7Gppp GVAA UCuuuaCcaAACAA CU Uu~acuuUAuM;WuAuguAGuUGcuGuGuGAuuCCCGug~ag~a~WAcuG III/ II I IIIII II I II II III/ I II II I II III1 III IIIII ill:1 I

GUAAaUCcgguCu AACMgCUcggUcCAuuucgUAgAGWaAgcaAGcUGggGaGAccCCCGacAGC CGUWGGAUcAgcGcucgcgucu IIIIIII I:

/CGuuuGGgwca.. .

FIG. 4. Comparison of selected regions from BMV and CCMV RNA3. (A) Comparisons between and among the sequences preceding the 3a gene (upper alignment) and the central intercistronic regions between the 3a and coat protein genes (lower alignment). The asterisks above the upper alignment mark the ends of two largely conserved blocks that are contiguous in BMV, but in CCMV are separated by the 11 1 bases denoted here in italics and underlined in Fig. 2. The 70% identity of the 3’ end of this additional CCMV segment with the core subgenomic promoter from the intercistronic region is marked by the double line of vertical bars. The arrow inserted in the intercistronic sequences (lower alignment) marks the transcription initiation site of subgenomic RNA4, and the core subgenomic promoter is bracketed. The intercistronic sequences shown represent bases 997-l 259 of BMV RNA3 (Ahlquist et a/., 1981 b) and bases of 1141-l 368 of CCMV RNA3 (Fig. 1 B). (B) Comparison of the 5’end of CCMV RNA3 (bases l-79) with sequences following the BMV 3a gene (bases 998-l 105).

SEQUENCE OF CCMV GENOMlC RNAS 2 AND 3 327

Similar rearrangements have been suggested to ex- plain the nearly identical 5’ ends on BMV and CCMV RNAs 1 and 2, the highly conserved 3’ ends of RNAs l-3, several imperfect direct repeats in BMV, and other bromovirus genome features (Ahlquist et al., 1984b; 1987; French and Ahlquist, 1988). Such recombination events, which are very similar to those frequently seen in short-term evolution of defective interfering RNAs, apparently can also make a significant contribution to the evolution of independent viruses. Similar observations have been made with tobacco rattle virus strains (Angenent et al., 1986; Robinson el al., 1987) and various alphaviruses (Hahn et al., 1988) supporting the concept that recombination has been a mechanism of central importance in the evolution of RNA viruses (Ha- seloff et al., 1984; Ahlquist et al., 1987). A wide variety of observations suggest that such recombination occurs at the level of viral RNA at significant frequencies (Kirkegaard and Baltimore, 1986; Bujarski and Kaes- berg, 1986; King, 1988).

The presence of a subgenomic promoter-like sequence 5’ to the CCMV 3a gene suggests that this gene might have been derived from a predecessor located in the interior of a larger genomic RNA, where its expression required production of a suitable subgenomic mRNA. The organization of the 3a and coat protein genes in such a precursor would be similar to that of the 3a-like 30-kDa gene and coat protein gene at the 3’ end of TMV genomic RNA, where both genes are expressed by subgenomic mRNAs. After division of this hypothetical precursor RNA into separate genomic RNA segments made the 3a gene directly translatable, lack of selection would allow loss of subgenomic promoter function by sequence drift for CCMV or deletion for BMV. In keeping with this, no “3a” subgenomic RNA species has yet been observed during CCMV infection. The above evolutionary path has been consid- ered previously as one possible explanation of the extensive similarities among various multicomponent viruses such as BMV and single component viruses such as TMV and alphaviruses (Haseloff et a/., 1984). As one alternative to fission of the genomic RNA of a single virus to create an ancestral bromovirus, the 3a gene alone might have been recruited from a TMV-like setting by recombination into a previously viable di- vided genome.

Unlike other sequenced bromoviral RNAs, the first AUG codon in CCMV RNA3 (bases 38-40) is followed immediately by a UAG terminator, and so does not ap- pear to direct translation of a major viral protein (Fig. 1 B). In vitro translation of CCMV RNA3 transcripts from pCC3TP4 and other full-length CCMV RNA3 CDNA clones produces an approximately 32-kDa protein (results not shown), as expected for translation of the 3a open reading frame initiated by the second AUG codon

in CCMV RNA3 (bases 239-241). AUG codons 5’to the first known expressed gene are present in a number of other plant viral mRNAs with capped 5’ ends (Cornelis- sen et al., 1983; Rezaian et al., 1984).

lntercistronic region of CCMV RNA3

The over-200.base intercistronic noncoding region separating the 3a and coat protein genes of CCMV RNA3 is likely to be functionally important, as the corresponding segment of BMV RNA3 contains c&acting elements required for efficient amplification of RNA3 it- self and sequences directing subgenomrc RNA4 synthesis (French and Ahlquist, 1987, 1988). The first 30 bases following the CCMV and BMV 3a genes show moderate similarity (Fig. 4A). However, only a few of these bases are even potentially implicated in BMV RNA3 accumulation (French and Ahlquist, 1987). The intercistronic sequences known to be required for efficient BMV RNA3 accumulation in viva are within a 1 OO- to 150-base region 3’ to the displayed similarity, and no statistically significant similarity between these BMV sequences and the CCMV RNA3 intercistronic region was found. However, when these same c&active BMV intercistronic sequences were compared with the entire CCMV RNA3 sequence under a range of comparison parameters, the highest similarity was consis- tently found between a portion of the BMV c&acting element and the 5’ end of CCMV RNA3 (Fig. 48). Simi- larity extends to the extreme 5’ end of CCMV RNA3, mapping this 5’end precisely against the sequence immediately following the BMV 3a gene termination codon. Whether the equivalent position of such boundaries in the two similar sequences has any biological or evolutionary significance is not clear.

Sequences controlling subgenomic RNA4 synthesis are contained within the first 120 bases 5’ to the BMV coat protein gene (French and Ahlquist, 1988). These BMV subgenomic promoter sequences must have functionally equivalent counterparts in CCMV RNA3, since BMV and CCMV RNA3 will direct RNA4 synthesis when coinoculated with RNAs 1 and 2 from either virus (Allison et al., 1988). The 28-base region immediately upstream of the CCMV coat protein gene is quite similar to the equivalently positioned BMV “core promoter” for subgenomic RNA synthesis (Fig. 4A). In both viruses this element is preceeded by an oligo(A) sequence, whose variable length in virion RNA popula- tions averages near 20 for BMV and near 40 for CCMV (Allison et a/., 1988). For BMV it was shown that re- moval of the oligo(A) dramatically reduces RNA4 synthesis (French and Ahlquist, 1988). Normal levels of BMV RNA4 synthesis also require the presence of 35- 50 bases 5’ to the oligo(A). These sequences contain partial direct repeats of core promoter sequences, in-

328 ALLISON, JANDA, AND AHLQUIST

uu -- C 5 ii c -c.c

U.A V*A AOV G*C UbA

RNAl . . . UCU UGA VCCVGAGUG GWCUA s *

AU c V A A -C.G

C-G V-A G-C

2534 A-V I U*G

RNA2 . . . AA4 UAA VGVVGGVCACAVVVAAGACVVGW GWCUA -

K l

uu

C- G

il A -V.A

COG COG G-C

1930 G*V I U-A

RNA3 - .** UAW WAG VGCCCGCVGAAGAGCGVVACACVAGUG GWAUA - Y *

FIG. 5. Potential for an additional hairpin stem and loop within the 3’ terminal noncoding regions of CCMV RNAs 1, 2, and 3. The sequences shown begin with the last codons of the la, 2a, and coat protein genes, respectively (italics), and end just 5’to stem and loop “h,” which bounds the 190.base 3’-terminal region (underlined in Figs. 1 A and 1 B) whose extensive secondary structure is shown in Fig. 2 of Ahlquist et a/.(1 981a). Bold letters highlight motifs present in two or more of the RNAs. The underlined residues 5’ to the base of the stem and in the loop of each sequence have the potential to base pair to form a “pseudoknot,” using the remainder of the loop and the intervening unpaired nucleotide at the base of the stem to bridge between the resultant two helices (Pleij et a/., 1987). Alterna- tively, the intervening nucleotide in RNA2 could participate in an additional G.U pair at the base of the stem shown. The potential for con- siderable base pairing within the displayed sequences 5’to the hairpin is also apparent, but currently lacks support from either direct structural data or apparent conservation between the various RNAs.

&ding close approximations of the consensus element AUCUAUGUU (French and Ahlquist, 1988). Searching upstream of the CCMV RNA3 oligo(A) re- vealed the presence of the related sequences AUucAUGU, located 24 bases 5’to the oligo(A) (bases 1261-l 268 of Fig. 1 B), but no other extensive similarities.

3’ noncoding regions

An extensive pattern of secondary structure is conserved over approximately 190 bases at the 3’ ends of the genomic RNAs from several bromoviruses including CCMV and BMV (Ahlquist et a/., 198 1 a). These regions direct (-) strand RNA synthesis, aminoacylation, and other virion RNA interactions (Ahlquist et a/., 1984a; Bujarski et a/., 1986; Dreher and Hall, 1988).

The potential for an additional hairpin stem and loop 27 bases 5’ to this region is conserved in all three BMV

genomic RNAs (Ahlquist et a/., 198413). This hairpin is not required for RNA replication in viva and so may have another function(s) (French and Ahlquist, 1987). Based on partial 3’sequences of the CCMV RNAs (Ahl- quist et al., 1981a), Pleij et a/. (1987) referred to the potential for additional secondary structure in the corresponding regions of at least some of the CCMV RNAs. From the complete sequences of RNAs 2 and 3 and the 3’ noncoding sequence of pCC 1 TPl (Allison et al., 1988), we find the potential for an additional stem and loop, positioned six bases 5’to the previously described common 3’ bromoviral structure, in all three CCMV RNAs (Fig. 5). The size and certain sequence features of these hairpins are similar among the CCMV RNAs but have relatively few features in common with the previously identified loops in BMV (Ahlquist et a/., 1984b). As with RNA3 (Dasgupta and Kaesberg, 1982), the overall 3’ noncoding regions of CCMV RNAs 1 and 2 are 50-60 bases shorter than their BMV counterparts.

The CCMV RNA and protein sequences reported here define a number of specific experimental ques- tions, such as whether the specificity to direct RNA replication with only homologous RNA1 maps to the more variable 5’ and/or 3’ portions of CCMV and BMV RNA2 and their encoded 2a proteins; whether the CCMV RNA3 signals for in vivo amplification are organized differently from those on BMV RNA3 (French and Ahlquist, 1987); and whether specific amino acid changes in the CCMV and BMV 3a proteins are related to host specificity. Studies to address these and other related issues are in progress in our laboratory and elsewhere.

ACKNOWLEDGMENTS

We thank Craig Thompson for excellent technical assistance. This research was supported by the National Institutes of Health under Public Health Service Grant GM35072 and by the National Science Foundation under a Presidential Young Investigator Award to P.A. (Grant DMB-8451884). R.F.A. was supported by NIH Viral Oncology Training Grant CA09075.

REFERENCES

AHLQUISI, P., BUJARSKI, J.. KAESBERG, P., and HALL, T. C. (1984a). Lo- calization of the replicase recognition site within brome mosaic virus RNA by hybrid-arrested RNA synthesis. Plant Mol. Ho/. 3, 37-44.

AHLQUIST. P., DASGUPTA, R., and KAESBERG, P. (1981a). Near identity of 3’ RNA secondary structure in bromoviruses and cucumber mosaic virus. Ce//23, 183-l 89.

AHLQUIST, P., DASGUPTA, R., and KAESBERG. P. (1984b). Nucleotide sequence of the brome mosaic virus genome and its implication for viral replication. 1. Mol. Biol. 172, 369-383.

AHLQUIST, P., FRENCH, R., and BUJARSKI, J. (1987). Molecular studies of brome mosaic virus using infectious transcripts from cloned cDNA. Adv. Virus Res. 32, 2 15-242.

SEQUENCE OF CCMV GENOMIC RNAS 2 AND 3 329

AHLQUIST, P., FRENCH, R., JANDA, M., and LOESCH-FRIES, L. S. (1984c). Multicomponent RNA plant virus infection derived from cloned viral cDNA. Proc. Nat/. Acad. Sci. USA 81,7066-7070.

AHLQUIS~, P., and JANDA, M. (1984). cDNA cloning and in vitro transcription of the complete brome mosaic virus genome. Mol. Cell. Biol. 4, 2876-2882.

AHLOUIST, P., LUCKOW, V., and KAESBERG, P. (1981 b). Complete nucleotlde sequence of brome mosaic virus RNAB. /. Mol. Viol. 153,23- 38.

AHLQUIST, P., STRAUSS, E., RICE, C., STRAUSS, J., HASELOFF, J., and ZIMMERN, D. (1985). Sindbis virus proteins nsP1 and nsP2 contain homology to nonstructural proteins from several RNA plant viruses. J. Viral. 53, 536-542.

ALLISON, R. F., JANDA, M.. and AHLQUIST, P. (1988). Infectious in v&o transcripts from cowpea chlorotic mottle virus cDNA clones and exchange of individual RNA components with brome mosaic virus. J. Virol. 62, 358 l-3588.

ANGENENT, G. C., LINTHORST, H. J. M., VAN BELKUM, A. F., CORNELIS- SEN, B. J. C., and BOL, 1. F. (1986). RNA 2 of tobacco rattle virus strain TCM encodes an unexpected gene. Nucleic Acids Res. 14, 4673-4682.

ARGOS, P. (1988). A sequence motif in many polymerases. Nucleic Acids Res. 16, 9909-9916.

BANCROFT, J. B. (1972). A virus made from parts of the genomes of brome mosaic and cowpea chlorotic mottle viruses. /. Gen. Viral. 14,223-228.

BARKER, R., JARVIS, N., THOMPSON, D., LOESCH-FRIES, L. S., and HALL, T. C. (1983). Complete nucleotide sequence of alfalfa mosaic virus RNA3. NucleicAcids Res. 11,2881-2891.

BIGGIN. M., GIBSON, T., and HONG, G. F. (1983). Buffer gradient gels and % label as an ald to rapid DNA sequence determination. Proc. Nat/. Acad. Sci. USA 80, 3963-3965.

BUJARSKI. J., AHLQUIST, P., HALL, T.. DREHER, T., and KAESBERG, P. (1986). Modulation of replication, aminoacylation and adenylation in vitro and infectivity in vivo of BMV RNAs contatning deletions within the multifunctlonal 3’ end. EMBOI. 5, 1769-l 774.

BUJARSKI, J., and KAESBERG, P. (1986). Genetic recombination between RNA components of a multipartite plant virus. Nature (Lon- don)321,528-531.

CORNELISSEN, B. 1. C., and BOL, 1. F. (1984). Homology between the proteins encoded by tobacco mosaic virus and two tricornavl- ruses. P/ant Mol. Biol. 3, 379-384.

CORNELISSEN, 8. J. C., BREDERODE, F. T., MOORMANN, R. I. M., and BOL, J. F. (1983). Complete nucleotide sequence of alfalfa mosaic virus RNA 1. Nucleic Acids Res. 11, 1253-l 265.

CORNELISSEN, 8. J. C., JANSSEN, H., ZUIDEMA. D., and BOL, J. F. (1984). Complete nucleotide sequence of tobacco streak virus RNA 3. Nucleic Acids Res. 12, 2427-2437.

DASGUPTA, R., and KAESBERG, P. (1982). Complete nucleotide sequences of the coat protein messenger RNAs of brome mosaic virus and cowpea chlorotic mottle virus. Nucleic Acids Res. 10, 703-713.

DAVIES, C., and SYMONS, R. H. (1988). Further implications for the evolutionary relationships between tripartite plant viruses based on cucumber mosaic virus RNA 3. Virology 165, 216-224.

DEOM, C. M., OLIVER, J. J., and BEACHY. R. N. (1987). The 30-kilodalton gene product of tobacco mosaic virus potentiates virus movement. Science 238, 389-394.

DEVEREUX, J., HAEBERLI, P., and SMITHIES, 0. (1984). A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res. 12,387-395.

DOOLI~LE, R F. (1981). Similar amino acid sequences: Chance or common ancestry? Science 214,149-l 59.

DREHER, T. W., and HALL, T. C. (1988). Mutational analysis of the sequence and structural requirements in brome mosaic virus RNA for minus strand promoter activity. J. Mol. Biol. 201, 3 l-40.

FRENCH, R.. and AHLQUIST, P. (1987). lntercistronic as well as terminal sequences are required for efficient amplification of brome mosaic virus RNAB. /. Viral. 61, 1457-l 465.

FRENCH, R., and AHLQUIST, P. (1988). Characterization and englneer- ing of sequences controlling in vitro synthesis of brome mosaic virus subgenomic RNA. /. Viral. 62, 241 l-2420.

FRENCH, R., JANDA. M., and AHLQUIST, P. (1986). Bacterial gene inserted In a engineered RNAvirus: Efficient expression in monocot- yledonous plant cells. Science 231, 1294-l 297.

GOELET, P., LOMONOSSOFF, G. P., BUTLER, P. J. G., AKAM, M. E., GAIT, M. J., and KARN, 1. (1982). Nucleotide sequence of tobacco mosaic virus RNA. Proc. Nat/. Acad. Sci. USA 79, 5818-5822.

GORBALENYA, A., and KOONIN, E. (1988). Birnavlrus RNA polymerase IS related to polymerases of positive strand RNA viruses. Nucleic Acids Res. 16,7735.

GRIBSKOV, M., and BURGESS, R. R. (1986). Sigma factors from E. co//, El. subtilis, phage SPOl, and phage T4 are homologous proteins. Nucleic Acids Res. 14, 674556763.

HAHN, C., LUSIIG, S., STRAUSS, E., and STRAUSS, J. (1988). Western equine encephalltls virus IS a recombinant virus Proc. Nat/. Acad. Sci. USA 85, 5997-6001.

HASELOFF, J.. GOELET. P., ZIMMERN, D.. AHLQUIST, P., DASGUPTA, R., and KAESBERG, P. (1984). Striking slmilarltles in amino acid sequence among nonstructural proteins encoded by RNA viruses that have dlssimllar genomic organlratjon. Proc /Vat/. Acad. Sci. USA. 81,4358-4362.

JANDA, M., FRENCH, R., and AHLQUIST, P. (1987). High efflclency T7 polymerase synthesis of infectious RNA from cloned brome mosaic virus cDNA and effects of 5’ extensions on transcript infectivity. Virology 158, 259-262.

KAMER, G., and ARGOS, P. (1984). Primary structural comparison of RNA-dependent polymerases from plant, animal, and bactenal viruses. Nucleic Acids Res. 12, 7269- 7282.

KIBERSTIS, P., LOESCH-FRIES, L. S., and HALL, T. (1981). Ural protein synthesis in barley protoplasts inoculated with native and fraction- ated brome mosaic virus RNA. Virology 1 12, 804-808.

KING, A. M. Q. (1988). Genetic recombination in positive strand RNA viruses, Chap. 7. In “RNA Genetics, Vol. II: Retrovlruses, Viroids, and RNA Recomblnatlon” (E. Domingo, 1. J. Holland, and P. Ahl- quist, Eds.), pp. 149-165. CRC Press, inc., Boca Raton, FL.

KIRKEGAARD, K., and BALTIMORE, D. (1986). The mechanism of RNA recomblnatlon In poliovirus. Cell 47, 433-443.

LANE, L. (1981). Bromoviruses, in “Handbook of Plant Virus Infec- tions and Comparative Dlagnosls” (E. Kurstak, Ed.), pp. 333-376. Elsevier Biomedical Press, Amsterdam.

LOMMEL, S. A , WESTON-FINA, M., XIONG, Z., and LOMONOSSOFF, G. P. (1988). The nucleotide sequence and gene organization of red clover necrotic mosaic virus RNA-2. Nucleic Acids Res. 16, 8587- 8602.

MANIATIS. T., FRITSCH, E. F., and SAMBROOK, 1. (1982). “Molecular Cloning: A Laboratory Manual.” Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.

MARSH, L. E., DREHER, T. W., and HALL, T. C (1988). MutatIonal analysis of the core and modulator sequences of the BMV RNA3 subgenomic promoter. Nucleic Acids Res. 16, 98 l--995.

MARSH, L. E., and HALL, T. C. (1987). Evidence lmpllcating a tRNA heritage for the promoters of positive-strand RNA synthesis In brome mosaic and related viruses. Co/d Spring Harbor Symp. Quant. Bioi. 52, 331-341.

MESHI, T., OHNO. T., and OKADA, Y (1982a). Nucleotlde sequence and its character of cistron coding for the 30 K protein of tobacco mosaic virus (OM strain). J. Biochem. 91, 1441 1444.

MESHI, T., OHNO, T.. and OKADA, Y. (1982b). Nucleotide sequence of the 30K protein clstron of cowpea strain of tobacco mosaic virus. Nucielc Acids Res. 10, 6 1 1 l-6 1 17

330 ALLISON. JANDA. AND AHLQUIST

MESHI, T., WATANABE, Y., SAITO, T., SUGIMOTO, A., MAEDA, T., and OKADA, Y. (1987). Function of the 30 kd protein of tobacco mosaic virus: Involvement in cell-to-cell movement and dispensability for replication. EMBO/. 6, 2557-2563.

MESSING, J. (1983). New M 13 vectors for cloning. ln “Methods in Enzymology” (R. Wu, L. Grossman, and K. Moldave, Eds.), Vol. 101, pp. 20-78. Academic Press, New York.

MILLER, W. A., DREHER, T., and HALL, T. (1985). Synthesis of brome mosaic virus subgenomic RNA in vitro by internal initiation on (-)- sense genomic RNA. Nature (London) 313,68-70.

PLEIJ, C. W. A., ABRAHAM& J. P., VAN BELKUM, A., RIETWLD, K., and

BOSCH, L. (1987). The spatial folding of the 3’ noncoding region of aminoacylatable plant viral RNAs. In “Positive Strand RNA Vi- ruses” (M. A. Brinton and R. R. Rueckert, Eds.), pp. 299-316. A. R. Liss, New York.

RAVELONANDRO, M., PINCK, M., and PINCK, L. (1984). Complete nucleotide sequence of RNA 3 from alfalfa mosaic virus, strain S. Biochi- mie 66,395-402.

REZAIAN, M. A., WILLIAMS, R. H. V., GORDON, K. H. J., GOULD, A. R., and SYMONS, R. H. (1984). Nucleotide sequence of cucumber mosaic virus RNA 2 reveals a translation product significantly homologous to corresponding proteins of other viruses. Eur. J. Biochem. 143,277-284.

REZAIAN, M. A., WILLIAMS, R. H. V., and SYMONS, R. H. (1985). Nucleo- tide sequence of cucumber mosaic virus RNA 1. Eur. J. Biochem. 150,331-339.

ROEINSON, D. J., HAMILTON, W. D. O., HARRISON, B. D., and BAUL-

COMBE, D. C. (1987). Two anomalous tobravirus isolates: Evidence for RNA recombination in nature. J. Gen. Viral. 68, 2551-2561,

SANGER, F., NICKLEN, S., and COULSON, A. (1977). DNA sequencing with chain-terminating inhibitors. Proc. Nat/. Acad. SC;, USA 74, 5463-5467.

SCHWARTZ, R. M., and DAYHOFF, M. 0. (1979). In “Atlas of Protein Sequence and Structure” (M. 0. Dayhoff, Ed.), Vol. 5, pp. 101- 110. Nat. Biomed. Res. Found., Washington, DC.

TAKAMATSU, N., OHNO, T., MESHI, T., and OKADA, Y. (1983). Molecular cloning and nucleotide sequence of the 30K and the coat protein cistron of TMV (tomato strain) genome. Nucleic Acids Res. 11, 3767-3778.

YANISCH-PERRON, C., VIEIRA, J., and MESSING, 1. (1985). Improved Ml 3 phage cloning vectors and host strains: Nucleotide sequences of theM13mp18and pUC19vectors. Gene33,103-119.

ZIMMERN, D., and HUNTER, T. (1983). Point mutation in the 30K open reading frame of TMV implicated in temperature-sensitive assem- bly and local lesion spreading of mutant Ni 2519. EMBO J. 2, 1893-l 900.

Sequence of cowpea chlorotic mottle virus RNAs 2 and 3 and evidence of a recombination event during bromovirus evolution

Documents