PMC:4504098 / 9233-15429
Annnotations
{"target":"https://pubannotation.org/docs/sourcedb/PMC/sourceid/4504098","sourcedb":"PMC","sourceid":"4504098","source_url":"https://www.ncbi.nlm.nih.gov/pmc/4504098","text":"Discussion\nWe detected both serologic and molecular evidence of RESTV infection in Philippine bats. RESTV RNA in the oropharyngeal swab of three Miniopterus schreibersii clustered phylogenetically with the 2008 pig-derived sequences and the historic 1989 Philippine primate-derived sequence. Sequence from all three bats was identical, and aligned most closely with the 2008 pig isolate from a farm (Farm A) in Bulacan Province [14], less than 40 km from the bat sampling location. All sequenced products from bats had the single nucleotide change; all positive control and related material held at AAHL did not have the change. Limited variation is not surprising with an assay targeting a conserved region of the NP gene following recent introduction of infection a population. While the high Ct values from the qPCR indicate the assay is approaching the limits of detection with these samples, a number of factors support the veracity of the findings. At the laboratory level, the repeatability of positive findings using qPCR in both pooled and individual specimens, the repeatability of positive findings in archived duplicate specimens, the corroboration by conventional PCR, and the direct sequencing results. We detected RNA in archived duplicate samples of two of the three positive M. schreibersii; these duplicates were made in the Philippines, stored at the PAHC, and forwarded to AAHL and tested in a separate run 12 months after the first test. At the epidemiological level, the clustering of the positive samples in one species, in one cave, and in one sampling event is consistent with a current infection dynamic. The inability of the PAHC duplicate sample testing to fully corroborate the first results can be plausibly explained by reduced volume of the duplicate samples (less than half the original sample volume), meaning less extracted RNA in the duplicate assay. In addition, there was potential nucleic acid degradation as the duplicate samples had been stored for a considerably longer period than the original samples. The qPCR findings also suggest the possible presence of the virus in additional sympatric taxa (M. australis, Cynopterus brachyotis and Chaerephon plicata) and in additional locations (Puning Cave), but limits of detection issues and small sample volumes again precluded corroboration and sequencing. Limited blood volumes also constrained the use of additional assays such as cell culture and next generation sequencing. For context, around 85 % of the bats we screened weighed less than 100 g. In combination with our ethical decision not to destructively sample the bats, this meant that individual blood volumes were frequently much less than 100 μl.\nThe serologic findings in flying-foxes, in conjunction with the molecular findings in insectivorous bats, suggest that ebolavirus infection is taxonomically widespread in Philippine bats. Also, while ebolaviruses have previously been detected in other Pteropodidae, this is the first reported detection in flying-foxes. The stronger serologic response of one sample to EBOV than RESTV antigen in the Western blot is intriguing, and parallels recent findings from Rousettus fruit bats in Asia [10]. While acknowledging the potential for non-specific binding in the recombinant N protein-based Western blot, and for cross-reactivity with heterologous antigens [16], the findings could suggest that more than one strain of ebolavirus is circulating in the source population. All three Western blot corroborated seropositives were A. jubatus, and all were captured at the same roost, which is periodically shared with P. vampyrus. The uncorroborated ELISA-positive bat was a captive P. vampyrus from a different location. This scenario supports the veracity of the serologic findings. Additional samples are needed to further interpret the findings. The absence of positive serology in M. schreibersii given the positive PCR findings warrants discussion. In an endemic infection scenario, positive serology would expected in the source population from which viral RNA was detected. However, in a scenario of recent introduction of infection to a population, limited seroconversion in the presence of infected individuals would not be unexpected. The lack of sequence variation in all three PCR-positive M. schreibersii is consistent with the latter.\nOur findings of RESTV infection in Philippine bats are supported by those of Taniguchi et al. [17]. They reported antibodies to RESTV in Rousettus amplexicaudatus from two locations in Luzon. As they sampled different bat populations, and one to two years prior to our study, our negative findings in R. amplexicaudatus in this study, while frustrating, are not overly surprising given the cryptic nature of filovirus infection and detection in bats [5]. Indeed, Tanaguchi et al. [17] screened 141 bats in total from 17 species, only confirmed RESTV-specific antibodies in 3 of 16 R. amplexicaudatus, and failed to detect any RESTV-specific amplicons by RT-PCR.\nThe decision to pool samples in the initial screening PCR reflected logistical constraints, however any saving in cost and time is countered by a loss of diagnostic sensitivity, which becomes particularly problematic when modest amounts of genetic material are present in the samples. In addition, the low level Ebola viral RNA detected from non-invasive swabs has prompted some studies to use tissue samples to maximise the probability of detection in infected bats (e.g., Amman et al. [8]). However, in this study we were constrained from destructively sampling bats, and thus our scope for viral detection may have been reduced. The aim of the study was to identify presence or absence of infection in bat taxa, and an optimistic target sample size was set to allow robust epidemiological interpretation of negative findings. This sample size was not met for any species or genus, and accordingly we refrain from making any interpretation on the lack of detection in any taxa. Conversely, our detection of infection in the modest sample of M. schreibersii indicates that, at the time of the study, infection prevalence was substantially higher than our conservative design prevalence.","divisions":[{"label":"title","span":{"begin":0,"end":10}},{"label":"p","span":{"begin":11,"end":2700}},{"label":"p","span":{"begin":2701,"end":4346}},{"label":"p","span":{"begin":4347,"end":5008}}],"tracks":[{"project":"Zoonoses_partialAnnotation","denotations":[{"id":"T2","span":{"begin":2156,"end":2177},"obj":"Species"},{"id":"T4","span":{"begin":2182,"end":2200},"obj":"Species"}],"attributes":[{"id":"A2","pred":"tao:has_database_id","subj":"T2","obj":"Tax:58060"},{"id":"A4","pred":"tao:has_database_id","subj":"T4","obj":"Tax:478698"},{"subj":"T2","pred":"source","obj":"Zoonoses_partialAnnotation"},{"subj":"T4","pred":"source","obj":"Zoonoses_partialAnnotation"}]},{"project":"2_test","denotations":[{"id":"26184657-19590002-143716470","span":{"begin":429,"end":431},"obj":"19590002"},{"id":"26184657-23343532-143716471","span":{"begin":3194,"end":3196},"obj":"23343532"},{"id":"26184657-21666792-143716472","span":{"begin":3360,"end":3362},"obj":"21666792"},{"id":"26184657-24747773-143716473","span":{"begin":4798,"end":4799},"obj":"24747773"},{"id":"26184657-23055920-143716474","span":{"begin":5497,"end":5498},"obj":"23055920"}],"attributes":[{"subj":"26184657-19590002-143716470","pred":"source","obj":"2_test"},{"subj":"26184657-23343532-143716471","pred":"source","obj":"2_test"},{"subj":"26184657-21666792-143716472","pred":"source","obj":"2_test"},{"subj":"26184657-24747773-143716473","pred":"source","obj":"2_test"},{"subj":"26184657-23055920-143716474","pred":"source","obj":"2_test"}]}],"config":{"attribute types":[{"pred":"source","value type":"selection","values":[{"id":"Zoonoses_partialAnnotation","color":"#ece193","default":true},{"id":"2_test","color":"#c793ec"}]}]}}