The broad bean strain of Tobacco mosaicirus (TMV-B) infects Nicotiana tabacum White Burley systemically whereas the tomato strain of Tomato mosaicirus (ToMV-S1) induces necrotic local lesions and is restricted to inoculated leaves. To examine the possible role of the viral movement protein (MP) in these symptom differences, a chimaeric virus (T/OMP) was produced in which the TMV-B MP gene was replaced by the ToMV-S1 MP gene. T/OMP induced the same symptoms as TMV-B in N. tabacum White Burley. However, in N. tabacum Samsun NN and other plants containing the N resistance gene, T/OMP caused necrotic lesions that were smaller than those produced by TMV-B but similar in size to those of ToMV-S1. We conclude that ToMV MP gene can substitute functionally for the TMV-B MP gene, and that the MP gene influences the size of necrotic local lesions on N-containing hosts.

Summary. Virus isolate Y23V, obtained from squash showing leaf curl symptoms inYunnan, China, was readily differentiated from four studied Chinese begomovirus isolates in reactions with a set of monoclonal antibodies raised against begomoviruses. The complete nucleotide sequence (2714 nts) of the DNA-Alike molecule of Y23V was determined. The DNA-A of Y23V is most closely related to that of tomato yellow leaf curl Thailand virus-[1] (TYLCTHV-[1]) (84% sequence identity). However, the AC1 and AC4 gene of Y23V DNA-A resembled to Pepper leaf curl virus from Bangladesh (PepLCBDV). The DNA-A ofY23V has three distinct regions: the region from 74-2071 nts is 95% identical to TYLCTHV-[1] excluding a 27 nt deletion; the following 386 nts are 91% identical to PepLCBDV and the rest of the DNA-A is not closely related to any reported begomovirus.Y23V, therefore, is considered to have arisen by recombination. The 84%sequence identity ofY23VwithTYLCTHV-[1] allowsY23Vto be considered as a distinct begomovirus species, for which the name squash leaf curl Yunnan virus (SLCYNV) is proposed.

DNAβ is a type of single-stranded (ss) circular satellite DNA found in association with monopartite-genome begomoviruses, such as Tomato yellow leaf curl China virus isolate Y10 (TYLCCNV-Y10). Y10 DNAβ is required for symptom expression in plants but depends on TYLCCNV-Y10 genornic DNA (DNA-A) for replication and encapsidation. When we converted DNAβ into a gene-silencing vector (modified DNAβ (DNAmβ)) by replacing its C1 open-reading frame (ORF) with a multiple cloning site (MCS), it was replicated but no longer induced symptoms in association with TYLCCNV-Y10 DNA-A, so allowing theeffects of gene inserts to be recognized easily. Insertion into DNAmβ of sequences from any of the three host genes (proliferating cell nuclear antigen (PCNA), phytoene desaturase (PDS), and sulfur (Su), or from a transgene (green fluorescent protein (GFP), resulted in silencing of the cognate gene in Nicotiana benthamiana. The silencing persisted for more than a month and was associated with decreased levels of mRNA of the gene targeted. Although DNAmβ probably does not enter meristematic tissue, the PCNAgene could be silenced there. DNAmβ was an effective silencing vector in tested N. glutinosa, N. tabacum Samsun (NN or nn), and Lycopersicon esculentum plants, and was able to silence two genes simultaneously. This satellite DNA vector-based form of virus-induced gene silencing (VIGS) promises to be applicable to other begornovirus/DNAβ systems, which are recently reported to occur in several dicotyledonous crop species, thereby providing a powerful approach to gene discovery and the analysis of gene function in these crops.

Sugarcane mosaic virus (SCMV)was detected in all 62 maize samples collected from eight maize-growing provinces in China showing dwarf mosaic symptoms by immunocapture reverse-transcription polymerase chain reaction (RT-PCR). Maize dwarf mosaic virus (MDMV), Sorghum mosaic virus (SrMV) and Johnsongrass mosaic virus (JGMV), however, were not detected in any of the samples by RT-PCR. Eleven cDNA fragments of approximately 0.8 kilobases covering most of the coat protein (CP) gene of SCMV were sequenced and sequence analysis indicates that these eleven isolates share 98.1 to 100% identity at the amino acid level. Sequence comparison and phylogenetic analysis of the CP genes from the eleven Chinese isolates as well as 21SCMVsubgroup virus isolates indicate that the eleven Chinese virus isolates were closely related to SCMV with 97.0 to 98.1% sequence identity at the amino acid level, while relatively lower sequence identity was found with MDWV, SrMV or JGMV. The results indicate that the Chinese isolates are members of the SCMV species, and thus, SCMV can be considered as the most common and important potyvirus infecting maize in China.

Complete nucleotide sequences of the DNA-A-like molecules of three East African cassava mosaic virus (EACMV) isolates from Kenya (-K, 2801 nt) and Malawi (-MH and-MK, both 2804 nt) were determined. These sequences were compared with that published for a Tanzanian isolate (-T, 2801 nt) and the partial sequence of a third Malawian isolate. Intergenic region sequences of all isolates, and deduced amino acid sequences of their AC1 (Rep) proteins, each formed a tightly related cluster thatwas distinct from the comparable components of other begomoviruses. Other complementary-sense genes (AC2, AC3, AC4) differed between EACMV isolates in a way consistent with the accumulation of point mutations. In contrast, virus-sense genes (CP, AV2) of isolates-MH and-MK differed (substantially for AV2) from those of other EACMV isolates but somewhat resembled those of tomato yellow leaf curl virus-Israel, suggesting they had been acquired by recombination with an unidentified begomovirus.

Complete DNA-A sequences of nine Pakistani geminivirus isolates from leaf curl-affected cotton (CLCuV-PK) or from okra, and the partial sequences of several additional isolates were determined. Sequences of isolates from cotton were of four types. Isolates from leaf curl-affected okra had virtually the same sequences as those from cotton. Isolates from yellow vein mosaic-affected okra were of two types (OYVMV types 201 and 301), both distinct from but closely related to the virus isolates from cotton. Of these six types, two types of CLCuVPK are the most closely related but another (CLCuVPK type 72b) is the most distinct. Of the encoded proteins, coat protein (CP) is the most strongly conserved (92-100%amino acid sequence identity), and AC4 protein the most variable (41-87%). The 5' and 3' halves of the intergenic region of some isolates had different affinities and occurred in seven combinations, suggesting that recombination had occurred and that the origin of replication was a favoured recombination site. Similarly, the first 1520 nt of CLCuV-PK type 804a DNA resembled those of OYVMV type 301 DNA but the remaining 1224 nt were very different. The AC1 (Rep) gene and 5

Geminivirus isolates associated with the epidemic of severe cassava mosaic disease in Uganda were studied and compared with virus isolates from the part of Uganda outside the epidemic area, and with African cassava mosaic virus (ACMV) and East African cassava mosaic virus (EACMV). Isolates of a novel type [the Uganda variant (UgV)] were detected in severely affected plants from the epidemic area, whereas those from plants outside the epidemic area were typical of ACMV. The complete nucleotide sequences of DNA-A of UgV (2799 nt) and of a Tanzanian isolate of EACMV (2801 nt) were determined and are extremely similar, except for the coat protein (CP) gene. The CP gene of UgV has three distinct regions: the 5'219 nt are 99% identical to EACMV (only 79% to ACMV); the following 459 nt are 99% identical to ACMV (75% to EACMV); and the 3'93 nt are 98% identical to EACMV (76% to ACMV). UgV DNA-A therefore is considered to have arisen by interspecific recombination of EACMV and ACMV. Despite the hybrid nature of their CP, UgV isolates were indistinguishable from ACMV in tests with 20 monoclonal antibodies (MAbs), including seven which reacted with ACMV but not EACMV. The discontinuous epitopes detected by these seven MAbs must involve amino acids which lie in the central part of the CP (residues 74-226) and which differ in ACMV and EACMV. UgV isolates were detected in severely mosaic-affected plants from all 11 widely separated locations sampled. The probable role of recombination in geminivirus evolution in the short to medium term is discussed.

Eighteen samples of begomoviruses isolated from tobacco, tomato and weed species in Yunnan, China were found to be associated with DNAb molecules, for which the complete nucleotide sequences were found to contain 1333-1355 nt. The 18 DNAb molecules identified consist of three main types, each associated with a different begomovirus species: 72-99 % nucleotide identity was found within one type, but only 39-57 % identity was found between types. All the DNAb molecules reported here and elsewhere contain a 115 nt conserved region that has 93-100 % identity with a consensus sequence, and have a common ORF encoding 118 amino acids on the complementary strand (designated C1). Coagroinoculation of the DNA-A component of Tomato yellow leaf curl China virus tobacco isolate Y10, with its associated DNAb (Y10b), shows this DNAb to be involved in symptom induction in tobacco and tomato. The in-frame ATG mutation of C1 of Y10b caused much milder symptoms as compared with wild Y10b, indicating a functional role for this ORF. Pairwise nucleotide sequence identity comparisons of DNAb molecules and their cognate viral DNA-A molecules indicate that DNAb molecules have co-evolved with their cognate helper viruses. Recombination between DNAb molecules is documented and a DNAb species concept is proposed and discussed.

The VP37 protein encoded by the RNA2 of Broad bean wilt virus 2 (BBWV2) was overexpressed in Escherichia coli. The protein was purified and a polyclonal antibody specific for the protein was produced. Time course studies byWestern blot assays in BBWV2-infected Chenopodium quinoa leaves showed that the VP37 protein was present in cells of the inoculated leaves by 12 h post inoculation and in cells of systemically-infected leaves by 2 days post inoculation. The protein was able to accumulate to a high level in infected leaves at the late infection stage. Gel retardation and UV cross-linking assays demonstrated that the VP37 protein bound preferentially single-stranded (ss) RNA and DNA in a non-sequence-specific manner. The VP37 protein-RNA complex was stable in solutions containing less than 400mM NaCl, but became fully dissociated in the solutions containing 800mM NaCl. Sequence analysis of the VP37 protein and its ability to bind ssRNA and ssDNA suggest that the protein may play a role similar to the movement proteins reported for other plant viruses.

The nucleotide sequence of the RNA1 of broad bean wilt virus 2 (BBWV2) isolate B935 has been determined from overlapping cDNA clones. It contains 5956 nucleotides in length excluding the 30 terminal poly(A) tail and contains a single long open reading frame (ORF) of 5613 nucleotides extending from nucleotide 234 to 5846. A repeated motif has been found in the 50 non-coding region. The predicted polyprotein encoded by the long ORF is 1870 amino acid in length with a molecular weight of 210 K. Amino acid sequence comparisons between portions of the BBWV2 RNA1-encoded polyprotein and proteins encoded by several species in Comoviridae revealed the putative functions of BBWV2 RNA1-encoded proteins and the same general genetic organization as that of comoviruses and nepoviruses. Based on the determined sequence, full-length cDNA clone of RNA1 designated as pU1FL was constructed. Together with transcripts from full-length cDNA clone of RNA2 (pU2FL), transcripts from pU1FL infected Chenopodium quinoa successfully.