Cacao swollen shoot virus (CSSV), a member of the family Caulimovi-ridae, genus Badnavirus occurs in all the main cacao-growing areas of West Africa. We amplified, cloned and sequenced complete genomes of five new isolates, two originating from Togo and three originating from Ghana. The genome of these five newly sequenced isolates all contain the five putative open reading frames I, II, III, X and Y described for the first sequenced CSSV isolate, Agou1 originating from Togo. Their genomes have been aligned with the genome of Agou1. The nucleotide and amino acid sequence identities between isolates have been calculated and a phylogenetic analysis has been made including other pararetroviruses. Maximum nucleotide sequence variability between complete genomes of CSSV isolates was 29.4%. Geographical differentiation between isolates appears more important than differentiation between mild and severe isolates. ORF X differs greatly in size and sequence between the Togolese isolates Nyongbo2 and Agou1, and the four other isolates, its functional role is therefore clearly questionable.

Oligonucleotides derived from the flanking regions of the putative coat protein gene of the cacao swollen shoot badnavirus isolate 1 A (CSSV-l A) were able to prime the synthesis of specific products directly from extracts from CSSV-l A-infected leaves by immunocapture polymerase chain reaction (IC-PCR), following trapping of virions with polyclonal antibodies to CSSV-l A. CSSV isolates serologically distinct from CSSV-l A were not detected by IC-PCR when the CSSV-l A-derived primers were used following trapping with homologous antisera. IC-PCR was at least 100-fold more sensitive than double antibody sandwich (DAS)-ELISA in comparative tests on samples from greenhouse-grown cacao plants. The superior sensitivity of IC-PCR over DAS-ELISA was confirmed in attempts to detect and identify CSSV-l A isolates in field samples and permitted detection of CSSV-l A isolates even in symptomless leaves from plants showing stem swelling only. The IC-PCR products obtained from four randomly selected field samples were sequenced and shown to contain a region of the CSSV-1 A genome where ORF X overlaps ORF 3. Analysis of the partial amino acid sequences deduced from ORF 3 and ORF X of the four field isolates revealed a considerable variation in these CSSV-l A gene products.

No abstract

Objective: To determine the true prevalence of HIV dual infections in a previously characterised HIV seropositive patient group due to inconsistencies between different diagnostic methods. Design: A cross-sectional study of an HIV seropositive group with different diagnostic methods. Setting: Three hospitals in the Northern, Ashanti and Greater Accra Regions of Ghana. Subjects: One hundred and forty five HIV infected patients/individuals sampled from June to September 2002. Main outcome measures: Using serological and molecular methods, the seropositive status of HIV-infected patients, previously determined by a preliminary screening process, was confirmed and discrepancies noted. The data was used to propose a more accurate laboratory diagnosis of HIV dual infections involving HIV-1 and HIV-2. Results: HIV-1 infections were mostly accurately detected, but difficulties were encountered in diagnosing HIV-2 infections. To achieve a positive detection on confirmatory immunoblots, antibody concentration in some samples tested was enhanced by using larger volumes. In other cases, diagnosis of HIV infections by PCR, especially HIV-2, was possible only after increasing the DNA template or MgCl 2 concentrations. Such samples would otherwise have been inaccurately scored for HIV infections. Conclusion: Based on the results of this study, we propose that the accurate diagnosis of HIV dual infections, especially HIV-2 component, must use an algorithm that involves PCR. Our results however underscore conclusions of a previous study that most dually seroreactive samples are predominantly HIV-1 infections with crossreactivity to HIV-2 antigens.