Nucleotide sequence evidence for the occurrence of three distinct whitefly-transmitted geminiviruses in cassava

The complete nucleotide sequence of the DNA of Indian cassava mosaic virus (ICMV) and a key part of that of a group B isolate of African cassava mosaic virus from Malawi (ACMV-M) were determined and compared at the nucleotide and encoded amino acid levels with the published sequences of an ACMV group A isolate (ACMV-K) and other whitefly-transmitted gemini-viruses (WTGs). The DNA of ICMV consists of two circular single-stranded molecules, DNA-A [2815 nucleotides (nt)] and DNA-B (2645 nt), which differ substantially in sequence from the genome components of ACMV-K (DNA-A 70%, DNA-B 47% sequence identity) and other WTGs. ICMV DNA-A contains eight open reading frames (ORFs) encoding proteins of > 100 amino acid residues, of which four ORFs (one genome sense, three complementary sense) are comparable to those of other WTGs. DNA-B contains one ORF in each sense, as in other WTGs. None of the putative viral proteins are more similar in amino acid sequence to the proteins of ACMV-K than to those of another WTG. The coat protein of ACMV-M is more like that of tomato yellow leaf curl virus from Sardinia (86% sequence identity) than those of ICMV or ACMV-K. The intergenic regions of ACMV-K, ACMV-M and ICMV DNAs differ in size, and largely in sequence, except for two 30 to 40 nt sequences which are also conserved in other WTGs and can form stem-loop structures. The intergenic region of ICMV DNA contains three copies of a 41 nt sequence, and that of ACMV-M DNA contains an imperfect repeat of a 34 nt sequence which resembles the repeated sequence in ICMV DNA. The differences between ACMV-K, ACMV-M and ICMV are considered great enough to justify their separation as isolates of three distinct WTGs: African cassava mosaic virus, East African cassava mosaic virus and Indian cassava mosaic virus.