1International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru (PO), India 502 324; 2Scottish Crops Research Institute, Dundee DD2 5DA, UK; 3Natural Resources Institute, Central Avenue, Chatham Maritime, Kent ME4 4TB, UK

Background and objectives
Rosette is a serious disease of groundnut (Arachis hypogaea) endemic to sub-Saharan Africa. The disease is caused by a complex of three agents, groundnut rosette umbravirus (GRV) and its satellite RNA (sat RNA), and groundnut rosette assistor luteovirus (GRAV). GRV and its sat RNA must be packaged in the GRAV coat protein for them to be transmitted by the aphid Aphis craccivora. Sat RNA is largely responsible for different forms of the rosette disease, whereas either GRAV or GRV alone cause no obvious symptoms. GRAV can be detected in plants and aphids by TAS-ELISA using monoclonal antibodies (Mabs) raised to either potato leafroll virus or GRAV [1]. GRV and its sat RNA can be detected in plants by nucleic acid hybridization [2]. Due to their cross reactions with different luteoviruses, a panel of Mabs has to be used in TAS-ELISA to verify that a luteovirus detected in groundnut is indeed GRAV [1]. In addition, TAS-ELISA cannot provide information on whether the aphids carry particles containing either GRAV RNA and/or GRV and its sat RNA. Therefore, we have developed RT-PCR assays for reliable and sensitive detection of the three agents of rosette disease in aphid vectors.

Materials and methods
Cultures of chlorotic and green rosette disease were maintained in groundnut cv. Malimba. A non-viruliferous colony of A. craccivora was maintained on groundnut cv. Malimba and used for transmission. Adult aphids were given a 48 h acquisition access period on either chlorotic or green rosette-infected groundnut plants and a 72 h inoculation feeding period on healthy groundnut seedlings prior to extraction of total RNA. Aphids collected from farmers' fields were stored in 70% ethanol for different periods of time and subsequently used for extraction of total RNA. Groundnut leaf samples were tested for GRAV by TAS-ELISA. Three methods were evaluated for extracting RNA from groundnut leaves and aphids. Primers for specific amplification of each the three agents of rosette were designed and used in RT-PCR.

Results and conclusions
Two of the three extraction procedures were found to be useful in giving RNA of good quality for RT-PCR. Of these two, the total RNA extraction kit supplied by Qiagen was found to be the most versatile. GRAV, GRV and sat RNA-specific fragments could be amplified from diseased plants tested positive for GRAV in TAS-ELISA, whereas only GRV and sat RNA could be amplified from diseased plants tested negative for GRAV. Both GRAV and GRV specific fragments could be amplified from total RNA extracted from single aphids that were exposed either to green or chlorotic rosette-infected groundnut plants. However, sat RNA could be amplified only when total RNA extracted from more than two aphids was pooled and used in RT-PCR. All three agents could be detected in aphids stored in 70% ethanol up to 30 days at room temperature. Analysis by TAS-ELISA of the field collected groundnut plants infected with either chlorotic or green rosette revealed that many symptom-showing plants did not contain GRAV. This was further substantiated by RT-PCR. Fragments specific to the three agents of rosette disease were amplified in A. craccivora exposed to rosette diseased plants that tested positive for GRAV but not from aphids exposed to diseased plants that tested negative for GRAV. The ability to detect rosette disease agents by RT-PCR in aphids will enable samples of aphids collected at different times of the season and in different sites, to be tested at a central location. Rosette symptom showing plants without GRAV can not serve as sources of inoculum and thus remain a 'dead end' of the disease.

1. Scott KP, Farmer M-J, Robinson DJ, Torrance L, Murant, AF, 1996. Annals of Applied Biology 128, 77-83.
2. Blok VC, Ziegler A, Scott K, Dangora DB, Robinson DJ, Murant AF, 1995. Annals of Applied Biology 127, 321-328.