Surveillance of the rabies-related lyssavirus, Mokola in non-volant small mammals in South Africa

The reservoir host of Mokola virus (MOKV), a rabies-related lyssavirus species endemic to Africa, remains unknown. Only sporadic cases of MOKV have been reported since its first discovery in the late 1960s, which subsequently gave rise to various reservoir host hypotheses. One particular hypothesis focusing on non-volant small mammals (e.g. shrews, sengis and rodents) is buttressed by previous MOKV isolations from shrews (Crocidura sp.) and a single rodent (Lophuromys sikapusi). Although these cases were only once-off detections, it provided evidence of the first known lyssavirus species has an association with non-volant small mammals. To investigate further, retrospective surveillance was conducted in 575 small mammals collected from South Africa. Nucleic acid surveillance using a pan-lyssavirus quantitative real-time reverse transcription polymerase chain reaction (qRT-PCR) assay of 329 brain samples did not detect any lyssavirus ribonucleic acid (RNA). Serological surveillance using a micro-neutralisation test of 246 serum samples identified 36 serum samples that were positive for the presence of MOKV neutralising antibodies (VNAs). These serum samples were all collected from Gerbilliscus leucogaster (Bushveld gerbils) rodents from Meletse in Limpopo province (South Africa). Mokola virus infections in Limpopo province have never been reported before, and the high MOKV seropositivity of 87.80% in these gerbils may indicate a potential rodent reservoir.

http://www.ojvr.org Open Access marked with a unique tattoo number near the base of their tail, and released back to their respective capture sites. Voucher specimens were euthanised with an overdose of Isofor, after which their organs were harvested (i.e. brain, tongue, salivary glands, heart, kidney, lungs, pectoral muscle, spleen, intestines, rectum and bladder) in 2.0 mL Sarstedt tubes for a broader pathogen surveillance study and immediately stored in liquid nitrogen until storage at -80 °C. Carcasses were placed in a 3 L PathoPak (Intelsius Solutions, United Kingdom [UK]) containing 80% ethanol and were submitted to Ditsong National Museum of Natural History and the Natural History Collection for Public Health and Economics for voucher-based morphological identification, and museum archiving.
All of the brain samples were negative for the presence of viral RNA with the pan-lyssavirus qRT-PCR assay (Appendix Table 1-A1). Negative results were expected as these animals were apparently healthy individuals and did not exhibit any visible signs of disease. An overall MOKV seropositivity of 87.80% (36 out of 41) was observed for the gerbils (Gerbilliscus leucogaster) tested from Meletse at the cut-off 1:25 serum dilution (Figure 2; Appendix Tables 1-A1, 2-A1, 3-A1 &  4-A1). The titre ranges for this rodent species were high when compared to another serological surveillance study conducted in Zimbabwe (Foggin 1988). Foggin identified MOKV VNAs in 5.63% (18 out of 320) of all rodents that were tested. An overall MOKV seropositivity of 17.57%     (1) PCR assay refers to the molecular barcoding assay that was used to determine the genetic identity of the rodent -Cytochrome B (CytB); (2) DNA source refers to the material that was used to extract DNA from for the PCR assay; (3) Query cover refers to how much the submitted sequence (i.e. the query sequence) is covered by the target sequence; (4) Per cent identity refers to the similarity of the query sequence to the target sequence; (5) GenBank accession number refers to GenBank's reference for the target sequence; (6) BLAST results refer to the genetic identity (i.e. genus and species name) of the target sequence's organism; §,  (Aghomo et al. 1990;Kemp et al. 1972;Nottidge, Omobowale & Oladiran 2007;Ogunkoya et al. 1990). Even though MOKV has been shown to cross-react in serological assays with other closely-related lyssaviruses (Kuzmin et al. 2008), cross-reactivity with other phylogroup II lyssaviruses was not investigated in this study.
Of the 36 gerbils showing MOKV seropositivity, only 28 were genetically identified with the CytB barcoding PCR assay ( Table 2). The same identification was obtained from morphological examination of 24 voucher specimens ( Table  2). Eight gerbils could not be identified to species level as they were released and no additional sample material was available. The Highveld gerbil, Gerbilliscus brantsii, is sympatric with G. leucogaster, however, based on known museum records, no G. brantsii has been caught at Meletse before (Rautenbach 1982) and these were, therefore, allocated to G. cf. leucogaster. The variability observed in the per cent identity (i.e. 83.78% -100.00%) between the individual gerbils is expected since previous molecular characterisation assays performed on the Gerbilliscus genus have recorded intraspecies genetic variation that range from 1% to 20% (Aghová et al. 2017;Colangelo et al. 2007).
Members of the Gerbilliscus genus are nocturnal and terrestrial, exhibit no sexual dimorphism (Skinner & Chimimba 2005) and occupy simple to complex, deep burrows (i.e. warrens) (De Graaff 1981; Granjon & Dempster 2013). They are physiologically, morphologically and behaviourally adapted to live in arid climates (Granjon & Dempster 2013;Monadjem et al. 2015). Gerbilliscus leucogaster, however, is less arid adapted and can be found along rivers and drainage lines in open grasslands and wooded savannas (Dempster 2013;Monadjem et al. 2015). The breeding pattern and social organisation of G. leucogaster rodents are not well-understood, however, studies have reported a communal nature (De Graaff 1981;Smithers 1971) with burrows being occupied by a pair (Skinner & Chimimba 2005) and some warrens housing families or several adults (Choate 1972). The ecological nature of Bushveld gerbils may potentially be the reason why this specific rodent species are more likely to be MOKV seropositive compared to solitary rodent species belonging to the Steatomys and Rhabdomys genera occurring at Meletse.
More nucleic acid and serological surveillance studies in non-volant small mammal populations are required to obtain a better understanding of MOKV distribution, prevalence and its potential reservoir species. Brain and serum samples in this study were collected from seemingly healthy small mammals in areas that do not coincide with areas where previous MOKV cases have been reported in South Africa. Surveillance should be expanded to areas where MOKV spillover infections in cats and dogs have previously been reported. Furthermore, because lyssavirus distribution and dynamics might be influenced by seasonality, surveillance efforts should also include samples that were collected in different seasons and over multiple years. This expansion, together with representative sample sizes of certain nonvolant small mammal species, will collectively increase the possibility of identifying more of these animals that are infected or that have previously been exposed to MOKV.

Ethical considerations
This study formed part of a larger surveillance programme of the

Data availability
The authors confirm that the data supporting the findings of this study are available within the article and its supplementary materials.

Disclaimer
The contents of this research communication are solely the responsibility of the authors. The opinions, findings and conclusions expressed do not necessarily reflect the official view of the National Research Foundation.

Non-volant small mammal type
Non-volant small mammal species   Small non-volant mammal individuals that were collected as voucher specimens and whose brains were negative for the presence of MOKV RNA. Animal ethics clearance was obtained from the University of Pretoria's Animal Ethics Committee (Reference Numbers: EC071-15 & H008-18); ‡, Results for the 1:10, 1:25, 1:50, 1:250 & 1:1250 serum dilutions are recorded as a number that represents the number of fields (out of a total of 10) that contain MOKV 12/458 infected cells for both initial (i) and duplicate (f) rounds of the micro-neutralisation test; §, The log10 50% end-point (ED) neutralisation titre for each serum sample as calculated by Reed and Muench (1938); ¶,

Brain samples
The average log10 50% ED neutralisation titre, together with the standard deviation (s.d.) for each serum sample as calculated by Reed and Muench (1938); † †, Follow-up screening could not be completed as the serum sample was depleted during initial screening.    Sp., species. †, Results for the 1:10 and 1:25 serum dilutions are recorded as a number that represents the number of fields (out of a total of 10) that contain MOKV 12/458 infected cells for both initial (i) and duplicate (f) rounds of the micro-neutralisation test.