Application of a Simplified Molecular Protocol to Reveal Mixed Infections with Begomoviruses in Cassava

Samples of cassava leaves exhibiting severe symptoms of cassava mosaic disease (CMD) were collected with the PhytoPASS kit in fields surrounding the city of Bujumbura (Burundi). These materials were then sent to Belgium for polymerase chain reaction determination of the CMD begomoviruses inducing the observed symptoms. Different pairs of specific primers were used to amplify DNA sequences specific to African cassava mosaic virus (ACMV), East African cassava mosaic virus (EACMV), East African cassava mosaic Cameroon virus (EACMCV), East African cassava mosaic Malawi virus (EACMMV), East African cassava mosaic Zanzibar virus (EACMZV), the Uganda variant of East African cassava mosaic virus (EACMV-UG) and South African cassava mosaic virus (SACMV). It was revealed that mixed infections were prevailing in the analyzed materials. Most of the samples submitted to this analysis were found to be co-infected by three different begomoviruses (ACMV + EACMV + EACMV-UG). The so revealed mixed infections could explain the high severity of CMD symptoms noticed on cassava in the region of Bujumbura while the diversity within the CMD causal agents illustrates the importance to take this parameter into consideration for a successful use of plant genetic resistance to control the disease.     

The effect of cassava mosaic geminiviruses on symptom severity, growth and root yield of a cassava mosaic virus disease-susceptible cultivar in Uganda

A study was carried out to assess the effect of different cassava mosaic geminiviruses (CMGs) occurring in Uganda on the growth and yield of the susceptible local cultivar ‘Ebwanateraka’. Plants infected with African cassava mosaic virus (ACMV), ‘mild’ and ‘severe’ strains of East African cassava mosaic virus‐Uganda (EACMV‐UG2) and both ACMV and EACMV‐UG2 were grown in two experiments in Kabula, Lyantonde in western Uganda. The most severe disease developed in plants co‐infected with ACMV and EACMV‐UG2 and in those infected with the ‘severe’ form of EACMV‐UG2 alone; disease was least severe in plants infected with the ‘mild’ strain of EACMV‐UG2. ACMV‐infected plants and those infected with the ‘mild’ strain of EACMV‐UG2 were tallest in the 1999–2000 and 2000–2001 trials, respectively; plants dually infected with ACMV and EACMV‐UG2 were shortest in both trials. Plants infected with ‘mild’ EACMV‐UG2 yielded the largest number and the heaviest tuberous roots followed by ACMV and EACMV‐UG2 ‘severe’, respectively, whilst plants dually infected with ACMV and EACMV‐UG2 yielded the least considering the two trials together. Reduction in tuberous root weight was greatest in plants dually infected with ACMV and EACMV‐UG2, averaging 82%. Losses attributed to ACMV alone, EACMV‐UG2 ‘mild’ and EACMV‐UG2 ‘severe’ were 42%, 12% and 68%, respectively. Fifty percent and 48% of the plants infected with both ACMV and EACMV‐UG2 gave no root yield in 1999–2000 and 2000–2001, respectively. These results indicate that CMGs, whether in single or mixed infections, reduce root yield and numbers of tuberous roots produced and that losses are substantially increased following mixed infection.     

Role of a novel type of double infection in the geminivirus-induced epidemic of severe cassava mosaic in Uganda

To study the cause of the current epidemic of severe mosaic in Ugandan cassava, PCR analysis was used to detect and identify African cassava mosaic virus (ACMV), East African cassava mosaic virus (EACMV) and the recently reported recombinant geminivirus (UgV), which is derived from ACMV and EACMV, in leaf extracts from cassava plants grown from cuttings in the glasshouse at Dundee. The cuttings were collected from plants showing symptoms of different severities and growing at different sites in Uganda inside, at the periphery of, and outside, the area affected by the epidemic. ACMV occurred throughout the nine districts sampled but UgV was detected only in the area affected by the epidemic. EACMV was not found in Uganda. Most plants containing ACMV alone expressed mild or moderate mosaic, whereas very severe mosaic developed in most plants containing UgV plus ACMV and a few of those containing UgV only. Very severe mosaic in cassava from southern Sudan was likewise associated with co-infection by UgV and ACMV. The very severe disease was reproduced by graft-inoculating geminivirus-free cassava with UgV plus ACMV; plants inoculated with either UgV or ACMV developed severe or moderate symptoms, respectively. Unlike ACMV, Malawian EACMV did not enhance the severity of symptoms induced by UgV. However, a very severely affected plant from Ukerewe Island, Tanzania, contained ACMV and EACMV but not UgV. UgV attained a much greater concentration in cassava than did ACMV but the opposite occurred in Nicotiana benthamiana. In neither host was total virus antigen concentration affected by co-infection. Factors affecting the genesis, selection and spread of UgV are discussed. The evidence indicates that UgV is probably of relatively recent origin, that such variants do not appear often, and that the current epidemic has resulted from the rapid spread of UgV to infect plants and to invade regions in which ACMV already occurred. The novel type of virus complex so produced, consisting of an interspecific recombinant virus (UgV) and one of its parents (ACMV), typically has even more severe effects than UgV alone.     

Occurrence and distribution of cassava begomoviruses in Kenya

A survey for cassava mosaic disease (CMD) was conducted in Kenya, to investigate the factors contributing to the generally increased incidence and severity of CMD in the cassava growing regions and to study the distribution of the disease's causal begomoviruses, African cassava mosaic virus (ACMV) and East African cassava mosaic virus (EACMV) and their strains. Special emphasis was given to the occurrence of the destructive recombinant Uganda variant strain of EACMV (EACMV-UG2). Samples from 91 farmers' fields in the main cassava-growing areas of coastal and western Kenya were collected and subjected to ELISA and PCR for detection and typing of the begomoviruses present. CMD incidence was highest in western Kenya (80–100%) and lowest in the Coast province (25–50%). In Western and Nyanza provinces, 52% of the samples tested contained EACMV-UG2, 22% ACMV and 17% contained both ACMV and EACMV-UG2. EACMV was found in four cases at different sites. In cassava samples from the coast province, only EACMV with DNA-A sequences similar to EACMV strains present in Kenya and Tanzania was found. East African cassava mosaic Zanzibar virus (EACMZV) was present in several farms in the Kilifi district. In 15% of all cassava samples with CMD symptoms, flexuous, filamentous virus-like particles were also found, providing evidence for a more complex virus situation in cassava grown at the Kenyan coast. In western Kenya, where intense cassava cultivation takes place, CMD is rampant and EACMV-UG2 was found in mixed virus infections with ACMV driving the epidemics. In coastal areas, where farms are scattered and in isolation, EACMV is endemic, however, with a lower disease incidence and with a limited impact to cassava production.     

Viruses Associated with Cassava Mosaic Disease in Senegal and Guinea Conakry

A survey in Senegal and Guinea Conakry established the presence and incidence of cassava mosaic virus disease (CMD) in both countries. CMD occurred in all the fields surveyed, although its incidence was higher in Senegal (83%) than in Guinea (64%). Populations of the whitefly vector, Bemisia tabaci, were low in both countries averaging 1.7 adults per shoot in Guinea and 3.2 in Senegal. Most infections were attributed to the use of infected cuttings, 86 and 83% in Senegal and Guinea, respectively, and there was no evidence of rapid current‐season, whitefly‐borne infection at any of the sampled locations. Disease severity was generally low in the two countries and averaged 2.5 in Guinea and 2.3 in Senegal. No plants with unusually severe CMD symptoms characteristic of the CMD pandemic in East and Central Africa were observed. Restriction fragment length polymorphism (RFLP)‐based diagnostics revealed that African cassava mosaic virus (ACMV) is exclusively associated with CMD in both the countries. Neither East African cassava mosaic virus (EACMV), nor the recombinant Uganda variant (EACMV‐UG2) was detected in any sample. These survey data indicate that CMD could be effectively controlled in both countries by phytosanitation, involving the use of CMD‐free planting material and the removal of diseased plants.     

Serological and biological variations of African cassava mosaic virus, in Nigeria

To determine the occurrence of variants of African cassava mosaic virus, 316 cassava leaf samples were collected from mosaic‐affected cassava plants in 254 farmers. fields in 1997 and 1998, covering the humid forest, coastal/derived, southern Guinea and northern Guinea savannas and arid and semi‐arid agroecologies of Nigeria. The samples were tested in triple antibody sandwich enzyme‐linked immunosorbent assay using a panel of 10 monoclonal antibodies (MAbs) against the virus in which 29 reaction patterns were observed. In cluster analysis, nine serotypes were obtained at 0.80 Jaccard similarity coefficient index in which at least 50% of isolates of each serotype reacted alike. The serotypes ranged between two extremes: serotype 1 with 90% isolates reacting with the 10 MAbs and serotype 8 in which 90% of its isolates failed to react with the antibodies. Isolates of serotypes 1, 2, 4 and 8 were widely distributed while those of the other serotypes were estricted to certain agroecologies. Four representative isolates 227 (serotype 1), 231 (serotype 2), 235 and 283 (serotype 8) elicited different responses in Nicotiana, benthamiana, with isolate 283 not able to infect this and other test plants used. The serological variations did not necessarily reflect the biological variations. In polymerase chain reaction tests, one out of the five pairs of ACMV primers tested distinguished only isolate 283. The humid forest, derived/coastal and southern Guinea savannas where most of the crop is grown in Nigeria had a high number of variants, which makes the agroecologies suitable for the selection of resistant cassava clones against ACMV.     

Field Studies of Cross Protection with Cassava Mosaic Geminiviruses in Uganda

The effect of cassava mosaic virus disease (CMD) was compared on plants grown from cuttings that were initially virus‐free at planting and those infected with a mild strain of East African cassava mosaic virus‐Uganda (EACMV‐UG). All initially healthy plants developed CMD symptoms within 5 months of planting (MAP) at both trial sites in Uganda, although spread was more rapid at Kamuli than at Serere. Significantly (P < 0.001) higher symptom severity scores were recorded in initially healthy plants, which had average scores of 3.6 and 3.5 at Kamuli and Serere, respectively, compared with 2.8 for mildly diseased plants at each location. Severity scores of 4 and 5 were more frequent in initially healthy plants, accounting for 77 and 39% of the total infections recorded in comparison with 47 and 11% in mildly diseased plants at Kamuli and Serere, respectively. Mildly diseased plants were significantly taller than initially healthy plants 8 and 12 MAP at both locations. However, the converse was true 4 MAP although differences were significant at Serere but not at Kamuli. Mildly diseased plants yielded significantly more tuberous roots than initially healthy plants at Kamuli but not at Serere. Average total weights of tuberous roots per plant were 2.48 and 1.63 kg for mildly diseased and initially healthy plants at Kamuli and 4.46 and 4.61 kg at Serere, respectively. These results may help to explain the increased prevalence in recent years of mildly diseased plants of local CMD‐susceptible cultivars in eastern Uganda, from where these varieties virtually disappeared following the severe CMD epidemic in the 1990s. The results also provide the first field evidence of a cross protective effect of mild strains of a cassava mosaic geminivirus.     

The cassava mosaic geminiviruses occurring in Uganda following the 1990s epidemic of severe cassava mosaic disease

The cassava mosaic geminiviruses (CMGs) isolated from cassava plants expressing mild and severe symptoms of cassava mosaic disease (CMD) in 2002 in Uganda were investigated using the polymerase chain reaction and restriction fragment length polymorphism (PCR-RFLP) molecular techniques and DNA sequencing. Two previously described cassava mosaic geminiviruses: African cassava mosaic virus (ACMV) said East African cassava mosaic virus - Uganda variant (EACMV-UG2) were detected in Uganda. The RFLP technique distinguished two polymorphic variants of ACMV (ACMV-UG1 and ACMV-UG2) and three of EACMV-UG2 (EACMV-UG2[1], EACMV-UG2[2] and EACMV-UG2[3]). ACMV-UG1 produced the fragments predicted for the published sequences of ACMV-[KE]/UGMld/ UGSvr, while ACMV-UG2, which produced the RFLP fragments predicted for the West African ACMV isolates ACMV-[NG], ACMV-[CM], ACMV-[CM/DO2] and ACMV-[CI], was shown to be ACMV-UGMld/UGSvr after DNA sequencing. EACMV-UG2[1] produced the RFLP fragments predicted for the published sequences of EACMV-UG2/UG2Mld/UG2Svr. However, both EACMV-UG2[2] and EACMV-UG2[3], which produced East African cassava mosaic vzras-[Tanzania]-like polymorphic fragments with RFLP analysis, were confirmed to be isolates of EACMV-UG2 after DNA sequencing. Thus, this study emphasises the importance of DNA sequence analysis for the identification of CMG isolates. EACMV-UG2 was the predominant virus and occurred in all the surveyed regions. It was detected in 73% of the severely and 53% of the mildly diseased plants, while ACMV was less widespread and occurred most frequently in the mildly diseased plants (in 27% of these plants). Mixed infections of ACMV and EACMV-UG2 were detected in only 18% of the field samples. Unlike previously reported results the mixed infection occurred almost equally in plants exhibiting mild or severe disease symptoms (21% and 16%, respectively). The increasing frequency of mild forms of EACMV-UG2 together with the continued occurrence of severe forms in the field warrants further studies of virus-virus and virus-host interactions.     

Particle purification and properties of cassava Ivorian bacilliform virus

A virus found in cassava from the north-west of the Ivory Coast was transmitted by inoculation with sap extracts to herbaceous species in six plant families. Chenopodium quinoa was used as a propagation host and C. murale was used for local lesion assays. The virus particles are bacilliform, c. 18 nm in diameter, with predominant lengths of 42,49 and 76 nm and a structure apparently similar to that found in alfalfa mosaic virus. Purified preparations of virus particles had A₂₆₀/A₂₈₀ of 1.7 ±0.05, contained one protein of M_rc. 22 000, and yielded three species of RNA with M_r (× 10^-6) of c. 0.7, 0.8 and 1.2. Although the virus particles were poorly immunogenic, an antiserum was produced and the virus was detected by enzyme-linked immunosorbent assay (DAS-ELISA) in leaf extracts at concentrations down to c. 6 ng/ml. Four other field isolates were also detected, including a strain which caused only mild systemic symptoms in C. quinoa instead of necrosis. The naturally infected cassava source plants were also infected with African cassava mosaic virus (ACMV) but when the new virus was cultured in Nicotiana benthamiana, either separately or together with ACMV, its concentration was the same. The new virus did not react with antisera to several plant viruses with small bacilliform or quasi-bacilliform particles, and alfalfa mosaic virus reacted only weakly and inconsistently with antiserum to the cassava virus. The new virus, for which the name cassava Ivorian bacilliform virus is proposed, is tentatively classified as the second member of the alfalfa mosaic virus group.     

Comparative Analysis of Tissue Tropism of Bipartite Geminiviruses

Abutilon mosaic virus (AbMV), a bipartite geminivirus of the genus Begomovirus, has been vegetatively propagated for many years in Abutilon sellovianum in which it is strictly phloem-restricted. Using in situ hybridization and immunological analyses, the tissue tropism of AbMV in the laboratory host Nicotiana benthamiana was compared with that of two other bipartite begomoviruses, African cassava mosaic virus (ACMV) and tomato golden mosaic virus (TGMV). Analysis of the first systemically infected leaves and longitudinal sections of axillary and flower buds revealed that all three viruses are initially confined to the vascular traces, although both ACMV and TGMV are later detectable in nearly all tissue types. In contrast, AbMV remained strictly phloem-limited in this host throughout the course of infection. The ability of ACMV and TGMV to move out of N. benthamiana phloem tissues is correlated with the development of severe symptoms in comparison with the mild symptoms associated with AbMV infection. It was also demonstrated that Sida micrantha mosaic virus, a virus that is closely related to AbMV, is phloem-limited in Malva parviflora even though it induces severe leaf curl, stunting and necrosis in this host. The present data demonstrate that bipartite begomoviruses can exhibit strikingly different patterns of tissue tropism.     

Efficacy of chemotherapy and thermotherapy in elimination of East African cassava mosaic virus from Tanzanian cassava landrace

Cassava mosaic disease is caused by cassava mosaic begomoviruses (CMBs) and can result in crop losses up to 100% in cassava (Manihot esculenta) in Tanzania. We investigated the efficacy of chemotherapy and thermotherapy for elimination of East African cassava mosaic virus (EACMV) of Tanzanian cassava. In vitro plantlets from EACMV‐infected plants obtained from coastal Tanzania were established in the greenhouse. Leaves were sampled from the plants and tested to confirm the presence of EACMV. Plantlets of plants positive for EACMV were initiated in Murashige and Skoog (MS) medium. On the second subculture, they were subjected into chemical treatment in the medium containing salicylic acid (0, 10, 20, 30 and 40 mg/L) and ribavirin (0, 5, 10, 15 and 20 mg/L). In the second experiment, EACMV‐infected plantlets were subjected to temperatures between 35 and 40°C with 28°C as the control. After 42 days of growth, DNA was extracted from plant leaves and PCR amplification was performed using EACMV specific primers. It was found that plant survival decreased with increasing levels of both salicylic acid and ribavirin concentrations. In general, plants treated with salicylic acid exhibited a lower plant survival % than those treated with ribavirin. However, the percentage of virus‐free plants increased with an increase in the concentration of both ribavirin and salicylic acid. The most effective concentrations were 20 mg/L of ribavirin and 30 mg/L of salicylic acid; these resulted in 85.0% and 88.9% virus‐free plantlets, respectively. With regard to thermotherapy, 35°C resulted in 79.5% virus‐free plantlets compared to 69.5% at 40°C. Based on virus elimination, ribavirin at 20 mg/L, salicylic acid 30 mg/L and thermotherapy at 35°C are recommended for production of EACMV free cassava plantlets from infected cassava landraces.     

Restriction of Virus Movement into Axillary Buds is an Important Aspect of Resistance in Cassava to African cassava mosaic virus

Axillary buds and bark samples of resistant, moderately resistant and susceptible (control) cassava genotypes either naturally infected under field conditions or experimentally inoculated by grafting were indexed for African cassava mosaic virus (ACMV). Virus detection was carried out using enzyme‐linked immunosorbent assay and polymerase chain reactions to determine the distribution of the virus within the plant and elucidate the genotypes response to virus movement. Significantly more bud and bark samples were positive for virus on the susceptible genotype TME 117 than resistant genotypes TMS 30001 and TMS 91/02319, or the moderately resistant genotype TMS 30572. Detectable virus concentration was significantly lower in the buds of moderately resistant and resistant genotypes than the susceptible control. Under field conditions, it was significant that more primary stem buds were infected than the buds of secondary and tertiary stems but such a gradient was not obvious with bark samples. Shoots that had asymptomic new leaves after the initial symptomatic leaves had no virus in their buds, but some of the bark samples from the same plants tested positive. A significant interaction was observed between year and stem type, and among year, genotype and stem type with respect to virus detection in bud and bark samples. Restriction of virus movement into axillary buds occurred in all the resistant and moderately resistant genotypes. This may explain ACMV‐infected stem cuttings of resistant genotypes producing healthy plants in subsequent generation.     

Molecular biodiversity of cassava begomoviruses in Tanzania: evolution of cassava geminiviruses in Africa and evidence for East Africa being a center of diversity of cassava geminiviruses

Cassava is infected by numerous geminiviruses in Africa and India that cause devastating losses to poor farmers. We here describe the molecular diversity of seven representative cassava mosaic geminiviruses (CMGs) infecting cassava from multiple locations in Tanzania. We report for the first time the presence of two isolates in East Africa: (EACMCV-[TZ1] and EACMCV-[TZ7]) of the species East African cassava mosaic Cameroon virus, originally described in West Africa. The complete nucleotide sequence of EACMCV-[TZ1] DNA-A and DNA-B components shared a high overall sequence identity to EACMCV-[CM] components (92% and 84%). The EACMCV-[TZ1] and -[TZ7] genomic components have recombinations in the same genome regions reported in EACMCV-[CM], but they also have additional recombinations in both components. Evidence from sequence analysis suggests that the two strains have the same ancient origin and are not recent introductions. EACMCV-[TZ1] occurred widely in the southern part of the country. Four other CMG isolates were identified: two were close to the EACMV-Kenya strain (named EACMV-[KE/TZT] and EACMV-[KE/TZM] with 96% sequence identity); one isolate, TZ10, had 98% homology to EACMV-UG2Svr and was named EACMV-UG2 [TZ10]; and finally one isolate was 95% identical to EACMV-[TZ] and named EACMV-[TZ/YV]. One isolate of African cassava mosaic virus with 97% sequence identity with other isolates of ACMV was named ACMV-[TZ]. It represents the first ACMV isolate from Tanzania to be sequenced. The molecular variability of CMGs was also evaluated using partial B component nucleotide sequences of 13 EACMV isolates from Tanzania. Using the sequences of all CMGs currently available, we have shown the presence of a number of putative recombination fragments that are more prominent in all components of EACMV than in ACMV. This new knowledge about the molecular CMG diversity in East Africa, and in Tanzania in particular, has led us to hypothesize about the probable importance of this part of Africa as a source of diversity and evolutionary change both during the early stages of the relationship between CMGs and cassava and in more recent times. The existence of multiple CMG isolates with high DNA genome diversity in Tanzania and the molecular forces behind this diversity pose a threat to cassava production throughout the African continent.     

Transgenic cassava resistance to <Emphasis Type="Italic">African cassava mosaic virus</Emphasis> is enhanced by viral DNA-A bidirectional promoter-derived siRNAs

Expression of double-stranded RNA (dsRNA) homologous to virus sequences can effectively interfere with RNA virus infection in plant cells by triggering RNA silencing. Here we applied this approach against a DNA virus, African cassava mosaic virus (ACMV), in its natural host cassava. Transgenic cassava plants were developed to express small interfering RNAs (siRNA) from a CaMV 35S promoter-controlled, intron-containing dsRNA cognate to the common region-containing bidirectional promoter of ACMV DNA-A. In two of three independent transgenic lines, accelerated plant recovery from ACMV-NOg infection was observed, which correlates with the presence of transgene-derived siRNAs 21–24 nt in length. Overall, cassava mosaic disease symptoms were dramatically attenuated in these two lines and less viral DNA accumulation was detected in their leaves than in those of wild-type plants. In a transient replication assay using leaf disks from the two transgenic lines, strongly reduced accumulation of viral single-stranded DNA was observed. Our study suggests that a natural RNA silencing mechanism targeting DNA viruses through production of virus-derived siRNAs is turned on earlier and more efficiently in transgenic plants expressing dsRNA cognate to the viral promoter and common region.     

Unusually severe symptoms are a characteristic of the current epidemic of mosaic virus disease of cassava in Uganda

Mosaic disease (MD) is more severe in cassava plants infected within the area of the current epidemic in northern and central Uganda than to the south of the affected area. This difference in severity was recorded within a single cultivar as well as amongst the mixtures of cultivars found commonly in farmers' fields. An increase in severity also occurred as the epidemic passed through localities. Varietal or agroecological factors coincident with the area of the epidemic are therefore unlikely to cause the increased severity. The severe disease could also be graft and cutting transmitted and could super-infect mildly diseased plants. Both mildly and severely diseased plants gave positive reactions in ELISA tests to antisera prepared against African cassava mosaic virus (ACMV) and an unusually severe form of ACMV or a closely related geminivirus is likely to be the cause of the severe mosaic disease. The epidemic also involves increased populations of the whitefly vector of ACMV, Bemisia tabaci , and possible hypotheses are presented as to how these phenomena may be related.     

Disease resistance characterisation of improved cassava genotypes to cassava mosaic disease at different ecozones

Twenty-two cassava genotypes and eight controls were evaluated in two cropping seasons for resistance to cassava mosaic disease (CMD) at the International Institute of Tropical Agriculture (IITA) fields, located at different ecozones of Nigeria. Disease incidence (DI) and index of symptom severity data were obtained monthly at each location and genotype. Symptomatic leaves were also collected during evaluation at each location, and virus was indexed by amplification in polymerase chain reaction. Significant differences within and across locations were observed in the reactions of cassava genotypes to CMD. DI across cassava genotypes was significantly (p = 0.05) highest in the Ibadan (22.6%), followed by Onne (19.3%). Generally, plants of clones 96/0860, 96/1439, 96/0160, 96/1089A, 96/1632, 96/1613, 96/1708, 96/0191, 96/0249 and 96/1565 had significantly lower values of DI in each location. African cassava mosaic virus in single infection was the predominant causal agent of CMD in IITA experimental fields under study.     

Occurrence of three distinct begomoviruses in cassava in Madagascar

The presence of East African cassava mosaic virus in association with cassava mosaic disease in Madagascar has previously been reported. We now describe virus isolates from mosaic‐affected Madagascan cassava with epitope profiles typical of African cassava mosaic virus, and an isolate with a nucleotide sequence similar to that of South African cassava mosaic virus. Thus, three distinct begomoviruses occur in cassava in Madagascar.     

Loss of CMD2‐mediated resistance to cassava mosaic disease in plants regenerated through somatic embryogenesis

Cassava mosaic disease (CMD) and cassava brown streak disease (CBSD) are the two most important viral diseases affecting cassava production in Africa. Three sources of resistance are employed to combat CMD: polygenic recessive resistance, termed CMD1, the dominant monogenic type, named CMD2, and the recently characterized CMD3. The farmer‐preferred cultivar TME 204 carries inherent resistance to CMD mediated by CMD2, but is highly susceptible to CBSD. Selected plants of TME 204 produced for RNA interference (RNAi)‐mediated resistance to CBSD were regenerated via somatic embryogenesis and tested in confined field trials in East Africa. Although micropropagated, wild‐type TME 204 plants exhibited the expected levels of resistance, all plants regenerated via somatic embryogenesis were found to be highly susceptible to CMD. Glasshouse studies using infectious clones of East African cassava mosaic virus conclusively demonstrated that the process of somatic embryogenesis used to regenerate cassava caused the resulting plants to become susceptible to CMD. This phenomenon could be replicated in the two additional CMD2‐type varieties TME 3 and TME 7, but the CMD1‐type cultivar TMS 30572 and the CMD3‐type cultivar TMS 98/0505 maintained resistance to CMD after passage through somatic embryogenesis. Data are presented to define the specific tissue culture step at which the loss of CMD resistance occurs and to show that the loss of CMD2‐mediated resistance is maintained across vegetative generations. These findings reveal new aspects of the widely used technique of somatic embryogenesis, and the stability of field‐level resistance in CMD2‐type cultivars presently grown by farmers in East Africa, where CMD pressure is high.