MATERIALS AND METHODS

Field data and leaf samples collecting: Surveys were conducted in the South-Kivu province, Eastern D.R. of Congo in different cassava growing villages. From Bukavu town, the administrative centre of the province, five directions (Southern, Western, Northern, North-Western and the costal lake-Kivu). Twenty two villages and a total of 150 fields were covered during the survey (data not shown).

Within the selected fields, all the plants on the diagonals were systematically observed independently of the field square. For each plant examined, CMD incidence, CMD severity, the whitefly number (counted from the five top youngest leaves) and the type of infection (whitefly infection or cutting-borne infection) were recorded as proposed by Hahn et al. (1980) and adapted by Okao-Okuja et al. (2004) and Sseruwagi et al. (2004).

Leaf samples were systematically collected in the survey area on diseased plants. Samples were kept in alcohol prior to DNA extraction, Polymerase Chain Reaction (PCR) diagnosis and sequencing realised in the plant pathology laboratory at the Universite catholique de Louvain in Belgium.

Total DNA extraction, PCR amplification and sequencing: Total DNA was extracted from collected leaf samples using the protocol described by Dellaporta (Dellaporta et al., 1983) and the FastDNA^R Kit with FastPrep^R instruments (Qbiogene, inc., CA).

DNA-A AC2 and AC4 regions specific primers described by Monde et al. (2010) were used.

The mix PCR was prepared in a final volume of 50 μL using H₂O depc 26.25 μL, MgCl₂ 25 mM 5 μL, GoTaq^® 5x flexi buffer 10 μL, Dntp 100 mM 1.5 μL, each upstream and downstream primer 1 μL, GoTaq^®DNA polymerase (Promega) 5 units μL^-1 0.25 μL, extract DNA samples 5 μL. The PCR was cycling by a thermocycler ICycler Biorad^® version 4.006 at 94°C for 2 min for denaturizing, followed of 38 cycles of amplification at 94°C for 30 sec for denaturizing, hybridization at 58°C for 30 sec; 72°C for 1 min for elongation. The final elongation has been done at 72°C for 7 min. The revelation was made in 1.2% agarose gel under UV after electrophoresis at 120 V in ethidium bromide. Direct sequencing using Biosequencer Genetic analyser 3100 was made with PCR product diluted tenfold in distilled water. Each amplicon was sequenced many times forward and reverse. Partial sequences produced were translated in amino acids by a molecular toolkit (www.vivo.colostate.edu) and submitted to EMBL EBI server. NCBI accessions numbers were attributed with accessions FN433646 to FN433697. Concatenated amino acids of AC2 and AC4 from the NCBI Genbank and those collected in DR Congo were used to carry out phylogenetic trees using MEGA4.0 software (Tamura et al., 2007). The evolutionary history was determined using the Neighbour-Joining method with the complete deletion option and 1000 replicates for bootstrap values. The expressed mutations in amino acids sequences are shown in front of respective phylogenetic trees by alignment (Fig. 1 and 2).

Concatenated amino acids sequences of AC2 and AC4 of the following NCBI accessions were used: Cameroon [CM] [AF366902]¹; Nigeria [NG] [AJ427910]¹; Uganda [UG] [AF126800, AF126802]¹; [AJ717533, AJ717534, AJ618958, AM502326, AM502327, AM502329, AM502334]²; Ivory Coast [IC] [AF259894]¹ and Tanzania [TZ] [AY795988]².

Molecular diversity of the analysed sequences were reported in the geographic map of Sud-Kivu (Fig. 3) established in basis of general map of soil occupation of DR Congo provided by UCL-Geomatics, Louvain la Neuve, Belgium, 2007 where only elavation is highlightened.

RESULTS AND DISCUSSIONS

CMD incidence, severity, infection type and whitefly population: CMD incidence, its severity, whitefly population and virus diagnostics are presented in Table 1. The field data collected showed a clear geographic and ecological distribution of CMD incidence, its severity and the whitefly population in the surveyed area. CMD mean incidence values were moderate to high, ranging from 9.48 to 100% with the highest values recorded in the low altitude of Uvira (high temperature values, grassland and intensive agriculture environment) and Bunyakiri (high temperature values, secondary forest) where more fields were recorded with 85 to 100% of diseases plants than anywhere else (Table 1).

Fig. 1:

Phylogenetic and molecular analysis based on amino acids in concatenated sequences of AC2 and AC4 for ACMV inferred from the Neighbour-joining method (Saitou and Nei, 1987) using MEGA software version 4.0 with complete deletion option. Bootstrap values (1000 replicates) are shown as percentages in the branches (Tamura et al., 2007). The phylogenetic tree was constructed with AC2 and AC4 sequences from Cameroon [CM] AF366902; Ivory Coast [IC] AF259894; Nigeria [Ng] AJ427910; Uganda [UG] AF126800, AF126802 and Democratic Republic of Congo FN433646, FN433647, FN433648, FN433649, FN433650, FN433651, FN433652, FN433653, FN433654, FN433655, FN433656, FN433657, FN433658, FN433659, FN433660, FN433661, FN433662, FN433663, FN433664, FN433665, FN433666, FN433667, FN433668, FN433669, FN433670, FN433671, FN433672, FN433673, FN433674, FN433675, FN433676, FN433677

Fig. 2:

Phylogenetic and molecular analysis based on amino acids in concatenated sequences of AC2 and AC4 for EACMV-UG inferred from the Neighbour-joining method (Saitou and Nei, 1987) using MEGA software version 4.0 with complete deletion option. Bootstrap values (1000 replicates) are shown as percentages in the branches (Tamura et al., 2007). The phylogenetic tree was constructed with AC2 and AC4 sequences from Uganda [Ug]AM502327, AM502329, AM502326, AM502334, AJ717534, AJ717533, AJ618958; Tanzania [Tz]AY795988 and Democratic Republic of Congo FN 433678, FN433679, FN433680, FN433681, FN433682, FN433683, FN433684, FN433685, FN433686, FN433687, FN433688, FN433689, FN433690, FN433691, FN433692, FN433693, FN433694, FN433695, FN433696, FN433697

Fig. 3:	Cassava mosaic Begomovirus distribution on the Sud-Kivu map. The location of sampling were made in the basis of the general map of DR Congo. Five different isolates of ACMV (1 to 5) and two isolates of EACMV-UG (1 and 2) were identified by molecular analysis based on AC2 and AC4 genes. Fractional numbers (X/Y) represent the dual infections by both ACMV and EACMV-UG

In those villages more than 50% of infected plants were recorded with high severity score 3-5 and cassava asymptomatic plants in the fields were rare. The high incidence and severity values were positively correlated to whitefly population (7-17.9 insects per leaf) depending on the local environment temperature.

Table 1:	CMD incidence, severity and whitefly population and CMD diagnostic in the South-kivu province. Number with different succeeding alphabets are significantly different at p = 0.05 level, mean incidence was estimated as a percentage of CMD diseased plants in a field and in a village. Mean severity was estimated by using a 1-5 scale proposed by Sseruwagi et al. (2004) where score 1 is attributed to a cassava plant CMD symptom free and 5 to a cassava plant with CMD severe symptoms. Whitefly population was estimated as a mean whitefly number per leaf on the 5 top cassava leaves
Mean values with same letters are not signifficantly LSD different LSD at p = 0.05

Xie et al. (2011) have mentioned that whitefly fecundity and development were significantly increased at 24°C comparing to 15, 18 and 21°C.

However, great values of CMD incidence, severity and whitefly population were recorded at Kalehe, despite its location in altitude, moderate temperature and intensive agriculture environment. The lowest values were observed in the high altitude (Lwiro, Kavumu, Katana, Luhihi) where 9.5 to 37.8% of infected plants were recorded. The lowest value was observed in Lwiro (9.5%) where most of the cassava plants on the fields were asymptomatic, CMD severity score moderate (2-3) and whitefly population low (2 insects per leaf).

The high incidence and CMD severity score observed in the low altitude area (Uvira, Kalehe and Bunyakiri) was recorded in a positive correlation with the EACMV presence, alone or in mixed infections, in cassava fields with 57.2; 64.2 and 63.7% EACMV-positive samples respectively. These zones can be considered with the high epidemic index 3 (Legg et al., 2006). All sampled fields were predominantly grown with local and susceptible genotypes, among which Nambiyombiyo, M’Bailo and Cizinduka are most cultivated in altitude zones whereas Nakarasi and Naunde are most cultivated in Uvira zones.

The CMD severity score all around the covered area was ranged from 2.3 at Kavumu to 5 at Bunyakiri and Uvira with a mean score of 3.5 for the three most affected zones (Uvira, Kalehe and Bunyakiri). There was no genotypes variability on the incidence and severity, homogenous results have been recorded for all the local genotypes observed.

The whitefly number counted on upper cassava leaves varied from the different villages with low mean value of 1-3 insects per leaf in the high altitude (Katana, Luhihi, Lwiro, Kavumu, Nyangezi, Walungu and Mudaka) to 17.9 insects per leaf, the highest value observed in low altitude in Uvira-Kiliba. The positive correlation of white fly population, CMD incidence and severity and EACMV presence reaffirms the hypothesis EACMV to be easily transmitted by whitefly than ACMV (Colvin et al., 2004). The type of infection was homogenously distributed in each ecological zone, roughly 75 to 80% for cutting-borne infection and 20 to 25% for whitefly infection, except in Kamanyola where whitefly infection is 32.6 versus 67.4% for cutting-borne infection. These results suggest an ecological impact, mainly the temperature, on CMD vector development and its impact on the disease transmission and spread.

PCR diagnostic showed that only two species are occurring in the region, ACMV and EACMV species. They were diagnosed alone or in mixed infections in all cassava fields in the survey area except in the villages all along and near the lake Kivu (Luhihi, Mushweshwe and Birava) where only ACMV has been found (87%). EACMV-UG was found alone or in mixed infections with ACMV. Great rates of mixed infections were frequently associated to severe CMD symptoms observed at Bunyakiri, Kalehe and Uvira as it has been observed in Uganda, Kenya, Tanzania and Rwanda, suggesting a synergistic interaction (Harrison et al., 1997; Legg, 1999; Pita et al., 2001b; Fondong et al., 2000).

The blast of EACMV sequences showed high similarity to EACMV-UG (more than 98% of nucleotide and amino acids homology) (Fig. 2). This study is a report on EACMV-UG spread to Sud-Kivu province where it hasn’t been reported yet. The high frequency of EACMV-UG isolates in the north and south means the spread was realized from two different movements, at the north probably from Uganda when at the south, the spread is probably from Burundi, where it has been reported earlier (Bigirimana et al., 2004).

The concatenated AC2 and AC4 amino acids sequences alignment with Ugandan isolates showed that AC2 region of both ACMV and EACMV-UG is less diversified with less than 6% of amino acids substitution. The same observation was made for AC4 of EACMV-UG with less than 4% of amino acids substitution while the same region of ACMV was the most diversified with 6 to 11% of amino acids substitutions (Fig. 1 and 2). The AC4 of ACMV variability percentage is closed to what have been mentioned by Ndunguru et al. (2005) and Duffy and Holmes, (2009) for cassava begomovirus diversity and evolution and confirms ACMV occurring in the region for a long time, while the EACMV-UG less diversity is related to its recent introduction in the region.

The phylogenetic trees of different isolates established based on concatenated amino acids of AC2 and AC4 (Fig. 1 and 2) showed a single ACMV group of truncated AC4 gene presenting a premature codon stop and segregated in five different clusters while all EACMV-UG isolates were segregated in two different clusters.

The three first groups of ACMV are widely distributed in the region and linked to moderate CMD symptoms, while the groups 4 and 5 are restricted to Uvira zone and linked to severe CMD symptoms. The cluster 5 contains one isolate which is completely different from the other isolates and contains a specific fragment length of twenty five amino acids (22.9% of amino acids substitution) at the beginning of the AC2 region. The blast of the differential fragment didn’t reveal any recombination from a known begomovirus, nucleotides and amino acids substitution was the only one way of the virus diversity and evolution identified in this study.

The EACMV-UG isolates are closed to Uganda isolates and segregated in two clusters. The first cluster contains isolates widely distributed in the surveyed area which have homogenous AC2 region and weak AC4 amino acids substitution (3.8%) comparing to Ugandan isolates. The second cluster contains isolates restricted to kalehe and Uvira zones which present a deletion of twenty seven amino acids on the beginning of AC2 region, while AC4 region contains a weak amino acids substitution (3.8%). This particular isolates, both ACMV and EACMV-UG, restricted to Kalehe and Uvira is one of the probable causes of severe CMD symptoms observed in this area beside a great rate of ACMV/EACMV dual infections and the intensive cultivation of local and susceptible cassava genotypes.

The spread to Great Lake region (DRC, Rwanda and Burundi) of EACMV-UG from East African countries was expected to introduce the cassava begomovirus diversity reported by Ndunguru et al. (2005) in its origin region, in contrast no cassava begomoviruses species diversity has been reported by this study similar to the Rwandan and Burundian cases where both ACMV and EACMV-Ug only were reported (Bigirimana et al., 2004; Legg et al., 2001). In the East-Northern DRC, only those two species were reported by Monde et al. (2010) study. In relation to these observations, it can be assume that the Ugandan CMD spread to Great Lake countries was realized before recombination processes. However, recombinations result from the coexistence of several begomovirus in the same environment and no additional favourable conditions are reported to influence this phenomenon. In this way, frequent begomovirus recombinations reported in East Africa can easily occur in the Great lake region and there is no reason to not fear such threat in the Great Lake region where the two species coexistence is reported since ten years. Curiously, the first phenomenon of begomovirus recombination in Uganda was reported related to severe CMD disease (Deng et al., 1997; Zhou et al., 1997) only four years after Hong et al. (1993) report on the coexistence of ACMV and EACMV in East Africa region. Thus, there is no evident explanation of this and more care is needed to permanently monitor the disease epidemiology and aetiology in this region.

ACKNOWLEDGMENTS

This study was achieved under a Belgian technical cooperation (BTC) scholarship associated to technical support of the Earth and Life Institute Applied Microbiology-Phytopathology of the University catholic of Louvain in Belgium and the Catholic University of Bukavu in DR Congo. Authors are grateful to UCL-ENGE-Geomatics for providing the map.