The study presented here has greatly extended our knowledge of the diversity of betasatellites associated with begomovirus disease of okra across Indian subcontinent. Prior to this study only a single OLCuB and six BYVB isolates had been characterized. Our analysis shows BYVB to be the major betasatellite associated with okra and that this occurs across the whole of southern Asia. OLCuB appears to have a more limited geographic distribution across northern Pakistan and northern India.
The malvaceous plants in the New World are affected by bipartite begomoviruses [27–29] betasatellites not apparently occurring in the New World . In the Old World begomoviruses infecting okra are invariably associated with betasatellites and are, for the few which have been characterized, monopartite--lacking the DNA B component. Across Africa okra and other malveaceous hosts are affected by is affected by leaf curl disease, which is associated with begomoviruses and a single betasatellite, Cotton leaf curl Gezira betasatellite (CLCuGB) [30–32]. Although first identified in cotton , CLCuGB is also widespread in hollyhock, Sida spp. and tomato [7, 34, 35]. This situation in Africa, apparently a single betasatellite affecting okra, contrasts markedly with the different betasatellites affecting situation we have shown in India. Okra, as we have shown is affected by distinct betasatellites. There is at this time some debate as to the geographic origins of okra, with the majority favoring a southern Asian origin over a North African origin . The evidence presented here, a greater diversity of betasatellites of okra in India than in Africa suggesting a longer association of these on the sub-continent, might add weight to this argument. This is in agreement with the conclusions of Nawaz-ul-Rehman and Fauquet  who showed the center of diversity, and thus likely the center of origin, of begomoviruses and betasatellites to reside in Southeast Asia.
The presence of CroYVMB has been shown in different host plants belonging to non- malvaceous plants such as, Croton bonplandianus (EF597245), Croton sp. (AM410551), Crotalaria juncea (GQ183865, GQ183866, EU557375), radish (FJ593630), Jatropha gossypifolia (EU604296) and papaya (HM143903, HM143908), which are supported by different helper begomoviruses [23, 38]. However, the identification of CroYVMB in okra is something of a conundrum. CroYVMB is not a "malvaceous betasatellite" and has not previously been identified in any malvaceous plant species. The possible explanation for this may just be phenomenon known as pathogen "reassortment" due to insect transmission by B. tabaci, vector of a pathogen from another host plant that is maintained in trans by a helper begomovirus.
Betasatellites, on the whole, fall into two distinct groups, those isolated from species of the family Malvaceae and those isolated from non-malvaceous species [5, 7]. Although the malvaceous betasatellites are frequently identified in non-malvaceous plants, the converse has not been reported so frequently. This has been taken to indicate that the requirements for infection of species in the family Malvaceae differ from those for non-malvaceous species.
With three (possibly four) distinct betasatellites associated with disease in okra, this raises the question as to whether they induce distinct symptoms? Based on the data in Table 1, there is no clear correlation between symptom type (leaf curl/enation or yellow vein) and a particular betasatellite species. The leaf curl/enation and yellow vein phenotypes are associated with infections of both OLCuB and BYVB. Of course, since only single clones were characterized from each sample (multiple clones were characterized in only few isolates (data not shown)), it is possible that the major betasatellite (of a mixed infection) was not characterized, leading to deceptive results. However, for CLCuMB it has been shown that βC1 can phenocopy all symptoms of the disease (CLCuD) when introduced into a plant using a Potato virus X vector , showing that the major symptom determinant, at least for the CLCuD begomovirus-betasatellite, not the virus. It would thus be somewhat surprising to find three distinct betasatellite species that share a high level of sequence identity with respect to βC1 gene could induce distinct symptoms. The possible reasons for this may be attributed to helper viruses, other betasatellites in mixed infections, or okra varieties. Again, only experimental inoculation with defined clones will provide a definitive answer to this question.
The finding that the leaf curl/enation phenotype for infections of okra is restricted to Pakistan and northern Indian is puzzling. The possible reason reasons for this may be again, the helper virus, other betasatellites in mixed infections, okra varieties or that some other factor, such as for example co-infection with another (as yet unidentified) virus or satellite (-like) components determine the symptom differences. Recently two studies have shown that alphasatellites, the third partner in many begomovirus-betasatellite infections, can have a significant effect on symptoms [21, 40], although this did not lead to a change in symptom phenotype, merely an amelioration in symptom severity. It is also interesting to note that the geographical occurrence of the leaf curl/enation phenotype in okra overlaps the range of CLCuD and one of the viruses (Cotton leaf curl Multan virus), as well as the betasatellite know to cause CLCuD (CLCuMB), has recently been implicated in okra leaf curl disease for the first time reported from China .
The identification of a relatively high level of diversity of betasatellites in okra has implications for the development of resistance, by both conventional breeding and non-conventional (transgenic) approaches. Any resistant varieties produced will need to be able to counter begomoviruses supporting all possible betasatellites to have any chance of being durable. The finding that three of the betasatellites identified in okra, OLCuB, BYVB and BYVIB, have a βC1 gene with high sequence identity provides a possible means of achieving a broad spectrum resistance in okra to begomovirus-betasatellite diseases. Targeting the shared βC1 gene (by for example RNAi -mediated down regulation of transcript levels) or the product of the βC1 gene (using, for example, peptide aptamers [42, 43]) might yield a resistance active against three of the four betasatellites identified here--"one stone killing three birds".
Our efforts are now centered on analyzing the diversity of begomoviruses that occur in okra in India. Only a single virus, BYVV, has thus far been identified in okra in association with BYVB . The results indicated association of multiple betasatellites which are supported by multiple distinct helper begomoviruses. We have identified at least three distinct begomoviruses associated with okra in India (unpublished).