Abstract: Serological diversity of 178 RYMV isolates was determined by phylogenetic analysis of Serological Differentiation Indices (SDI) data generated from antigen coated-plate enzyme-linked immunosorbent assay (ACP-ELISA) using 26 RYMV Polyclonal antisera. These RYMV isolates were obtained from northern, southern, eastern and western Cote d`Ivoire. All the RYMV isolates was classified into three main serogroups (Sg1, Sg2 and Sg3) and six subgroups (Sg1a, Sg1b, Sg2a, Sg2b, Sg3a and Sg3a). This indicates the existence and levels of serodiversity among RYMV isolates in Cote d`Ivoire. These results provide evidence of a possible relationship between serological property, host plant and ecological origin of RYMV isolates. Phylogenetic classification of each RYMV isolate defined by SDI data in ACP-ELISA is potentially useful in epidemiological studies to assess isolate identity and interaction as well as to assist breeding programs aiming at the development of cultivars with durable resistant to RYMV in Cote d`Ivoire.
INTRODUCTION
Rice yellow mottle virus (RYMV), genus Sobemovirus (Hull, 1988) is the most rapidly spreading disease of rice (Oryza sativa L.) in Africa (Abo et al., 1998). First reported in Kenya, East Africa in 1966. It was reported in West Africa in 1975. The disease is now found virtually in all the countries of Africa where rice is grown (Abo et al., 1998). The virus causes yellowing, mottling, necrosis and stunting of rice plants leading to incomplete emergence of panicles with sterile grains (John and Thottappilly, 1987). In severity of infected plants death do occur. The means of transmission of the virus is by mechanical contacts and insects (Abo et al., 1998). Initially, RYMV was not of great threat to rice production, but with the intensification of rice production under rainfed and continuous irrigation, together with the introduction of many exotic varieties, epidemic out-break occurred. Several yield losses (96-97%) caused by RYMV infection has been reported (Fomba, 1986).
The existence of different RYMV strains in the field (Konate et al., 1997) that cause different kinds of disease is often a matter of considerable practical importance. For this reason reliable criteria are needed for distinguishing and identifying these strains. Knowledge of the serological relationships between RYMV isolates is valuable in diagnostic work and may prove to be important in epidemiological studies and disease control.
In previous studies, African RYMV isolates were serotyped using double immunodiffusion gel assay (Pinner et al., 1988; Mansour and Baillis, 1994; Sere et al., 2005), which were shown to be less sensitive than enzyme linked immunosorbent assay (ELISA). Phylogenetic analysis of some RYMV isolates based on sequences of the coat protein gene has been reported (Ngon et al., 1994; N’Guessan et al., 2000; Pinel et al., 2000; Fargette et al., 2001). However, because of relative expensiveness of sequencing many isolates of RYMV, few sequence data are available for phylogenetic analysis. Besides, few data are available on RYMV serodiversity, disease ecology and interactions between different RYMV strains in rice. However, Serological Differentiation Index (SDI) data has been reportedly used for the phylogenetic analysis of plant viruses serological classification (Pinner and Markharm, 1990; Rybicki, 1991; Pinner et al., 1992), but never been used for RYMV serodiversity. Construction of phylogeny for plant viruses using SDI data generated from antigen coated-plate enzyme linked immunosorbent assay (ACP-ELISA) allows rapid evaluation of serological diversity of many isolates and strains (Pinner and Markharm, 1990). The technique, which can be carried out in any moderately equipped research laboratory, is simple, rapid, cheap and very accurate.
The aim of this study was to investigate the use of serological differentiation indices for phylogenetic analysis of RYMV isolates serological relationships in Cote d’Ivoire. Such information is useful in epidemiological study and for developing rice varieties with durable resistant to RYMV in Cote d’Ivoire.
MATERIALS AND METHODS
Sample collection: In 2001 an intensive RYMV survey and sampling was performed in northern, southern, eastern and western parts of Cote d’Ivoire where rice was produced in upland, lowland and irrigated conditions. Leaf samples were collected based on (i) different host plant species, (ii) typical RYMV symptoms of yellow, mottle and stunting and (iii) different ecosystems. During the surveys, leaf samples were stored in an icebox and there after transferred to the laboratory and were stored in a freezer.
Isolates propagation: Sixty four, 68, 28 and 18 isolates respectively from northern, southern, eastern and western parts of Cote d’Ivoire (Table 1-4) were propagated in the method of mechanical transmission was according to Fauquet and Thouvenel (1977).
| Table 1: | Identity of isolates of Rice yellow mottle virus collected from Southern Cote d’Ivoire |
| Table 2: | Identity of isolates of Rice yellow mottle virus collected from Northern Cote d’Ivoire |
| Table 3: | Identity of isolates of Rice yellow mottle virus collected from Eastern Cote d’Ivoire |
Infected leaf samples were ground with 0.01 M phosphate buffer pH 7.0 at the ratio of 1:10 (w/v) and the resulting homogenate filtered through cheesecloth. Carborundum powder (600 mesh) was added to the inoculum to aid the penetration of the virus into leaf tissues. Four weeks after inoculation, leaves from each RYMV isolate bearing typical yellow mottle symptoms were harvested and used for RYMV serodiversity study.
Polyclonal antiserum production: Twenty-six polyclonal antibodies used in this study were obtained from Plant Pathology Unit, WARDA, raised against different RYMV isolates previously collected from different parts of West Africa (Table 5). These antisera were produced as follows:
| Table 4: | Identity of isolates of Rice yellow mottle virus collected from Western Cote d’Ivoire |
| Table 5: | List of antisera code and country of origin |
Different RYMV isolates, after propagation on susceptible rice variety Bouake189, were purified using the method of Thottappilly and Rossel (1993) and a modified version of Hull (1988). Two weeks after inoculation, rice seedlings showing typical symptoms of RYMV were harvested for purification. Rabbits were immunized with purified RYMV isolates by intramuscular injections at one-week intervals (Pinner and Markham, 1990). A purified RYMV suspension (0.5 mL, 1 mg mL-1) was emulsified with 0.5 mL of Freund’s complete adjuvant and used to immunize a rabbit followed by three more injections prepared with Freund incomplete adjuvant. One week after the fourth injection, the rabbit was bled and serum collected. All the twenty-six antisera produced were preabsorped with healthy plant materials (to remove any possible antibody produced against the host rice plant), as described by Pinner and Markham (1990).
ACP-ELISA: Indirect-antigen coated-plate ACP-ELISA was performed as described by Jaegle and Van Regenmortel (1985). Briefly, the procedure was as follows. Virus saps extracts were directly adsorbed to microtitre plates, followed by blocking with 1% bovine serum albumin. Twofold serial dilutions of antisera were made and bound antibody was detected with goat anti-rabbit serum conjugated to alkaline phosphatase (Sigma). The bound conjugate was detected using p-nitrophenyl phosphate solution and the plates were read at 405 nm.
Serological Differentiation Index (SDI) values between isolates: These values (Jaegle and van Regenmortel, 1985) were determined as described by Pinner and Markham (1990). Each virus was tested against a two fold serial dilution of each antiserum and the process was repeated three times. The SDI represents the average number of twofold dilution steps between homologous and heterologous viruses at a standard absorbance value of 0.5. The SDI values were read directly from the graph and the final value was expressed as a mean of the replicates.
Phylogenetic analysis: Four composite relationships dendrograms were generated, each for 64, 68, 28 and 18 RYMV isolates from northern, southern, eastern and western parts of Cote d’Ivoire from SDI data obtained from ACP ELISA results only (Pinner et al., 1992; Rybicki, 1991; Dekker, 1988) using numerical taxonomy and multivariate analysis system (NTSYS-PC), version 2.1 (Rohlf, 2000). SDI data were first converted to pairwise distance matrices, using the Jaccard coefficient of similarity (Jaccard, 1908) present in NTSYS-PC 2.1 and dendrogram was created by UPGMA cluster analysis (Sneath and Sokal, 1973).
RESULTS
SDI validity: All reactions that reached or exceeded the standard value of 0.5 and three times the background level was considered to be positive (Pinner et al., 1992). For each antiserum and its homologous, the SDI was defined as 0. The validity of high SDI values was demonstrated by obtaining similar results with virus purified further using CsCl gradient as described by Thottappilly and Rossel (1993). SDI values of the two were considered to be significant (Jaegle and van Regenmortel, 1985).
Diversity among southern isolates: Phylogenetic analysis revealed serological differences among 28 southern isolates (Fig. 1). The similarity ranges from 20 to 70% Jaccard similarity coefficient. At 25% Jaccard similarity level, all the isolates were separated into two main serogroups (Sg1 and Sg2), while at 40% Jaccard similarity level Sg1 and Sg2 were further separated into two subgroups (Sg1a and Sg1b) and (Sg2a and Sg2b) respectively. However, according to the pairwise genetic distances among the isolates analysed at 100% similarity level all isolates were separated, except in Sg1a subgroup in which isolates, SB-U-2 and SB-U-3, AS-L-3 and AS-L-4 were, respectively identical.
Diversity among western isolates: There were considerable diverse serological differences among 18 western isolates (Fig. 1) as indicated by phylogenetic analysis, giving rise to 25 to 75% similarity ranges. All the isolates were separated into two main serogroups (Sg1 and Sg2) at 30% similarity level. Sg1 and Sg2 were further separated into two subgroups (Sg1a and Sg1b) and (Sg2a and Sg2b), respectively at 40% similarity level. However, Sg2a was further divided into Sg2a1 and Sg2a2 higher subgroups at 45% similarity level. At 100% similarity level all isolates were separated, except in Sg2a2 subgroup in which TL-I-18 and SG-U-8 isolates were identical.
Diversity among eastern isolates: Broad serodiversity was obtained among 68 eastern isolates analysed (Fig. 2). The similarity ranges from 20-75% Jaccard similarity coefficient. At 30% Jaccard similarity level, all the isolates were separated into three main serogroups (Sg1, Sg2 and Sg3), while at 40% Jaccard similarity level Sg2 and Sg3 were further separated into two subgroups (Sg2a and Sg2b) and (Sg3a and Sg3b), respectively. However, according to the pairwise genetic distances among the isolates analysed at 100% similarity level, all isolates of subgroups Sg3a and Sg3b were distinct, while identical isolates were found among Sg1, Sg2a and Sg2b serogroups.
Diversity among northern isolates: Phylogenetic analysis revealed wide serological differences among 64 northern isolates analysed (Fig. 2). The similarity ranges from 25% to 75% Jaccard similarity coefficient. At 25% Jaccard similarity level, all the isolates were separated into three main serogroups (Sg1, Sg2 and Sg3), while at 30% Jaccard similarity level Sg1 and Sg2 were further separated into two subgroups (Sg1a and Sg1b) and (Sg2a and Sg2b), respectively. However, according to the pairwise genetic distances among the isolates analysed at 100% dissimilarity level, all isolates of serogroups Sg2a and Sg3 were distinct, while identical isolates were found among Sg1a, Sg1b and Sg2b subgroups.
DISCUSSION
RYMV has been described as a variable virus with many pathological variants (Thottappilly and Rossel, 1993; Sasaya et al., 1997; Konate et al., 1997; N’Guessan et al., 2000) and the unlimited number of pathological and virulence characters of RYMV and lack of standardisation of pathological conditions and virulence tests among different researchers have led to confusion and uncertainty in the characterization of this pathogen from rice (Taylor et al., 1990; Mansour and Baillis, 1994). The classification of all virus isolates into three main serogroups (Sg1, Sg2 and Sg3) and six subgroups (Sg1a, Sg1b, Sg2a, Sg2b, Sg3a and Sg3a) indicates the existence and levels of serodiversity among RYMV isolates in Cote d’Ivoire. This conformed to the earlier study of existence of several serotypes of RYMV isolates in Cote d’Ivoire (N’Guessan et al., 2000). In this study, many isolates emanating from same locality, field and host were observed to be serologically different (Fig. 1 and 2). For example, northern isolates (KG-34-I and KG-24-I), eastern isolates (DK-2-L and DK-3-L), western isolates (BF-I-1 and BF-I-2) and southern isolates (SB-U-1 and SB-U-2) were in each locality obtained from same host plant but were different in serodiversity. This explains the fact that within a set of isolates of related strains in the same host plant, many possibilities of interaction exist (Matthews, 1991). This possible isolate and host plant interaction varies between one locality to another thus account for diverse serological variability that exist among different RYMV isolates in Cote d’Ivoire. The serological similarities observed between isolates within the same and different localities confirm the great cross-infection potential of RYMV transmitted under natural conditions by different insect vectors (Bakker, 1971; Hammond et al., 1999).
| Fig. 1: | Composite relationship dendrogram of RYMV isolates from Southern and Western Cote d’Ivoire derived from SDI data using 26 polyclonal antisera in ACP ELISA |
| Fig. 2: | Composite relationship dendrogram of RYMV isolates from Eastern and Northern Cote d’Ivoire derived from SDI data using 26 polyclonal antisera in ACP ELISA |
In this study there were indications of localised micro variation among northern and eastern isolates with the emergence of Sg3 serogroup which was not found among the southern and western isolates. This is practically important for various RYMV identifications and further strengthens the deployment of durable rice germplasms in the region.
Besides, in the current study, all the isolates were completely separated and distinct. Twenty-two isolates gave homologous reactions with five antisera out of twenty-six antisera used, showing relatively low level of homologous reaction between isolates and the antisera. This might account for complete distinction of all isolates at 80% similarity level (Fig. 1 and 2). These serological distinctions among isolates suggest that they all differed in a specific combination of epitopes (N’Guessan et al., 2000). However, some isolates were serologically identical at 100% similarity (Fig. 1 and 2), this might be due to the similarity in the ecological origin of these isolates (Konate et al., 1997).
However, isolates from the same leaf and same host plant (KG-34-I and KG-24-I) were observed to be serologically different (Fig. 1). This could probably explain the fact that within a set of isolates of related strains, many possibilities of interaction exist (Matthews, 1991). As a result of possible interaction between different strains of the same isolates in the same host plant, diverse serological variability tends to exist between different isolates of RYMV across different ecologies in Cote d’Ivoire. The serological diversity observed between isolates within the same and from different ecologies in West Africa confirms the great cross-infection potential of RYMV transmitted under natural conditions by different insect vectors (Bakker, 1971; Hammond et al., 1999). Such possibilities of interaction within a set of isolates of related strains might lead to frequent occurrence of mutants which might be responsible for the high level of serological variation among the isolates (Boccard and Baulcombe, 1993).
Present results revealed that the use of SDI data generated from ACP-ELISA has great potential for serological identification and classification of RYMV isolates in Cote d’Ivoire. The specific distinction pattern of each isolate, revealed by phylogenetic analysis of SDI data generated from ACP-ELISA, is consistent, repeatable and reliable. The definition of specific distinct pattern for each isolate or strain should be a simple and straightforward task. Obviously, for these distinctions to have a practical meaning for the rice breeder, specific distinct pattern for each isolate must be related to the degree of virulence present. This could be achieved by a systematic comparison of distinct serotyped isolates or strains contrasting to their degree of virulence to rice. A similar approach has been used to determine the serological relationships of geminivirus isolates from gramineae in Australia (Pinner et al., 1992) and to identify maize streak geminivirus strains (Pinner and Markharm, 1990). Phylogenetic classification of each isolate of RYMV defined by SDI data in ACP-ELISA should be useful for the surveillance of RYMV in rice growing regions, in epidemiological studies to assess isolate identity and interaction as well as assist breeding programs aiming at the effective development of cultivars with durable resistant to RYMV in Cote d’Ivoire.
ACKNOWLEDGMENTS
This research was funded by the Department for International Development (DFID), UK and the Government of Japan. The authors would like to acknowledge Mr. Mensah Yao and Mr. Zai Kamelan, for their technical support.