Diagnosis and variability in coconut lethal yellowing phytoplasmas in Africa

In several countries of Africa, coconut palms are affected by lethal yellowing syndromes (LYS). Phytoplasmas are associated with Cape Saint Paul Wilt (CSPW) in Ghana, Kaincopé disease in Togo, Lethal Disease (LD) in Tanzania, and Lethal Yellowing (LYM) in Mozambique. No constant difference can be found between all these syndromes. They are all ascendant yellowing, starting on the lower leaves, associated with an aberrant drop in mature and immature nuts. Using some 'phytoplasma universal primers' - such as P1/P6 or P1/P7 we characterized different phytoplasma isolates involved in these LYS. "Specific primers", such as G813/AKSR claimed to be specific to CSPW, were used for Ghanaian isolates from Western and Central Region. In parallel, we tested these primers on a range of other mollicutes or bacteria. Cloning and sequencing of PCR products led us to some interesting conclusions or queries. It was shown in our study that 'phytoplasma universal primers' enabled amplification of a wide range of mollicutes such as Mesoplasma spp., Entomoplasma spp, Acholeplasma spp. These mollicutes are known to occur in or on plant surfaces and flowers. 'Universal primers' could even amplify bacteria (Leifsonia xylii for instance). Consequently, it meant they could not be reliable tools for diagnosing these LYS. Primers G813/AKSR, claimed to be specific to CSPW in Ghana, did not diagnose all cases of CSPW. We were able to find the origin of those false negative. Studies of the 16Sr RNA sequences showed significant variability. Some variability was detected inside the sequence of the AKSR primer site. A new primer was designed that made it possible to improve the diagnosis with almost no false negatives. A heteroduplex mobility assay (HMA) was used to confirm the genomic diversity existing among African LYS. They were also compared to the phytoplasmas associated with a similar lethal yellowing syndrome known as "Coconut Lethal Yellowing" (LY). A DNA fragment was amplified with primers P1 and P7 and subsequently subjected to HMA analysis. A PCR product amplified from our reference Ghanaian isolate GH5D, was combined with each PCR product and electrophoresed on polyacrylamide gel. Three groups of LYS were identified by HMA. The samples from Mozambique, and Ghana formed one group. The second group contained only the phytoplasmas from Tanzania. These two groups were different from the Caribbean isolates. The difference between the phytoplasmas associated with LYS on each side of a border in East Africa -Tanzania and Mozambique- has no explanation so far. (Texte intégral