Molecular monitoring of African swine fever virus using surveys targeted at adult Ornithodoros ticks : a re-evaluation of Mkuze Game reserve , South Africa

ArNoT, L.F., DU ToIT, j.T. & BASToS, A.D.S. 2009. Molecular monitoring of African swine fever virus using surveys targeted at adult Ornithodoros ticks: a re­evaluation of Mkuze Game reserve, South Africa. Onderstepoort Journal of Veterinary Research, 76:385–392 The Mkuze Game reserve (MGr), in north­eastern KwaZulu­Natal Province, South Africa is an African swine fever virus (ASF) controlled area. In a survey conducted in 1978, ASF prevalence in warthogs and Ornithodoros ticks in MGr was determined to be 2 % and 0.06 %, respectively. These values, acknowledged as being unusually low compared to other East and southern African ASF­ positive sylvatic­cycle host populations, have not been assessed since. The availability of a sensitive PCr­based virus detection method, developed specifically for the sylvatic tampan host, prompted a re­evaluation of ASF virus (ASFV) prevalence in MGr ticks. of the 98 warthog burrows inspected for Ornithodoros presence, 59 (60.2 %) were found to contain tampans and tick sampling was signifi­ cantly male­biased. Whilst gender sampling­bias is not unusual, the 27 % increase in infestation rate of warthog burrows since the 1978 survey is noteworthy as it anticipates a concomitant increase in ASFV prevalence, particularly in light of the high proportion (75 %) of adult ticks sampled. However, despite DNA integrity being confirmed by internal control amplification of the host 16S gene, PCr screening failed to detect ASFV. These results suggest that ASFV has either disappeared from MGr or if present, is localized, occurring at exceptionally low levels. Further extensive surveys are required to establish the ASFV status of sylvatic hosts in this controlled area.

Ornithodoros (Kleiboeker & Scoles 2001), is the most important maintenance host and vector of the ASF virus (ASFV) in Africa (Plowright, Parker & Peirce 1969;Thomson 1985).The distribution of ASF in southern Africa is determined largely by the distribution of this sylvatic cycle soft tick and its pre ferred vertebrate host, the common warthog Phaco choerus africanus (Thomson 1985).In the inverte brate tick host, the virus can be transmitted sexually, transovarially and transstadially (reviewed by Klei boeker & Scoles 2001).Infection rates increase with increasing size and age of the tick, and females dis play higher infection rates than males, due most likely to directional sexual transmission of the virus from males to females (Plowright 1977).In the ab sence of an infectious vertebrate blood meal, the virus can be maintained for periods of up to 15 months in the sylvatic Ornithodoros tick (Plowright, Perry, Pierce & Parker 1970) and possibly even in definitely (Plowright 1977).Whilst the virus can per sist for prolonged periods in infected ticks, which are considered to be the hosts of primary impor tance in ASF epidemiology, warthogs play an impor tant role in amplifying the virus in tick colonies.This ostensibly occurs at those times when virus levels in viraemic neonate warthogs surpass the threshold level required to infect ticks (Thomson 1985;Plow right et al. 1994).As warthogs are seasonal breed ers in southern Africa, the proposed viral transmis sion between the vertebrate and invertebrate hosts is cyclical, coinciding with the warthog farrowing sea son that runs from November to December (Cum ming 2005).ASFV seroprevalence in large ASFinfected warthog populations is generally high, ranging from 50-100 %, while the presence of virus in extensively surveyed East African tick popula tions is substantially lower, ranging from 0.26 to 5.03 % in adult ticks (Plowright 1977).
A study was undertaken in 1978 to assess the ASF status in warthog and Ornithodoros sylvatic cycle hosts in Mkuze Game reserve (MGr), South Africa (Thomson, Gainaru, Lewis, Biggs, Nevill, Van der Pypekamp, Gerber, Esterhuysen, Bengis, Bezuiden hout & Condy 1983;Thomson 1985).In this study, the ASFV infection rate for 5018 ticks was deter mined by inoculating pig leucocyte cultures with pooled ground tick suspensions (3-5 adult females, 6-8 adult males and > 10 nymphs; Thomson et al. 1983).Following blind passage, three haemadsorp tion(HAd)negative isolates, of reduced virulence for domestic pigs, were obtained, corresponding to a 0.06 % infection rate.This is 23fold lower than the ASFV infection rate in Ornithodoros in the Kruger National Park (KNP), South Africa (Thomson et al. 1983).Similarly, ASFV seroprevalence in warthogs in MGr was unusually low (2 %) on immunoelectro osmophoresis (IEoP) compared to the 94 % on IEoP for KNP, with ELISA seroprevalence esti mates being 6 % and 85 %, respectively.The MGr prevalence levels were the lowest documented for the six ASFpositive southern African localities eval uated (Thomson et al. 1983), and remain the lowest reported thus far for an ASFpositive region in which substantial numbers of both sylvatic hosts cooccur.The exceptionally low ASFV prevalence in MGr, availability of a sensitive PCrbased method for de termining infection rates in ticks (Bastos, Arnot, jacq uier & Maree, 2009), lack of largescale surveys since 1978 and the recent incorporation of MGr into the Greater St Lucia Wetland Park, prompted a reevaluation of the ASFV status in the sylvatic host that is of primary importance in ASF epidemiology, the warthogburrowdwelling Ornitho doros tampan.

Study area and warthog census data
The MGr is a 37 000 ha conservation area situated in northeastern KwaZuluNatal (KZN) Province, South Africa and constitutes the northwestern spur of the Greater St Lucia Wetland Park.The park is flanked to the north and south by the highpreva lence ASF and zeroprevalence warthog populations of the KNP and HluhluweiMfolozi Park (HiP), re spectively.In 1984, the number of warthog estimated to be present in the northern (main) part of the re serve (Fig. 1) was 630, while during the 2002 game census conducted by the KZN Nature Conservation Service in the same part of the reserve, the warthog population was estimated to be in the vicinity of 1 000 individuals (Craig Mulqueeny, personal com munication 2002).

tick collection
A total of 98 warthog burrows throughout the reserve were sampled for tampans (Fig. 1).The exact posi tion of each burrow was recorded using a hand held GPS unit.Loose soil was removed from the burrows with a spade and sieved over a white plastic sheet, using a stainlesssteel Labotec test sieve, with aper tures of 3 350 µm.The plastic sheet facilitated the de tection of small ticks that passed through the sieve.

Mkuze Game reserve
surfaces of all ticks were photographed and they were visually sexed using the guidelines of Walton (1962), and classed as male, female or nymphs.A two sample ttest was used to test for any difference in the overall occurrence of male and female ticks across all samples collected in MGr, using GENSTAT version 4.2 (Lane & Payne 1996).

Molecular analyses
DNA was extracted using a silicaguanidine thiocy anate technique from 10 % W/V ground tick suspen sions, prepared as described previously (Bastos et al. 2009).ASFV genome presence was evaluated using a duplex PCr approach that, in addition to be ing confirmatory for the virus, is also informative with respect to host taxonomy and ASFV genotype (Bastos et al. 2009).False negative results, due either to a failed PCr or a failed DNA extraction, are precluded through amplification of the tick mitochon drial 16S gene, which acts as an internal control.Positive and negative controls were always includ ed to ensure reaction efficiency and freedom from contamination of the reagents used.
To determine the genetic relatedness of the three MGr 1978 viruses to each other and to previously characterised, regional viruses, nucleotide sequenc es corresponding to the Cterminal end of the p72 gene were generated as previously described (Bas tos, Penrith, Cruciere, Edrich, Hutchings, roger, Cou acyHymann & Thomson 2003).In addition, three South African tampan isolates sampled between 1979 and 1987, were characterised.Sequences generated by a manual sequencing approach were submitted to Genbank (www.ncbi.nlm.nih.gov)un der accession numbers Fj455835 to Fj455840, and were complemented with homologous reference sequence data available for each of the 14 southern African genotypes described to date (Bastos et al. 2003;Boshoff, Bastos, Gerber & Vosloo 2007).A molecular phylogeny was inferred in MEGA4 (Ta mura, Dudley, Nei & Kumar 2007) using the neigh bourjoining algorithm with nodal support being as sessed by 10 000 bootstrap replications.

Burrow infestation rate and sex ratios of adult ticks
Tampans were collected from 59 of the 98 burrows inspected, which corresponds to a 60.2 % tick infes tation rate. of the 348 ticks collected, 88 (25 %) were nymphs and 161 of the 260 adult ticks (62 %) were males.This 62:38 male:female bias was sta tistically significant (P < 0.01; twosample ttest).

Screening for ASFV genome presence
All 348 ticks were screened for the presence of ASFV using the sylvatic tick duplex PCr (Bastos et al. 2009).None of these 348 ticks amplified the ex pected 478 bp ASFV p72 gene target.However, with the exception of the negative controls, all reac tions amplified the expected 313 bp 16S host mito chondrial gene target of the internal control, thereby confirming template DNA and reaction integrity and precluding the possibility of false negative results.

retrospective genetic analysis of the 1979 Mkuze tick isolates
Nucleotide sequencing of the Cterminal end of the p72 gene revealed that the three MGr tick viruses isolated in 1978 were identical to each other and also to two viruses that caused outbreaks of ASF in domestic pigs in 1973 and 1975 in South Africa (Fig. 2).Phylogenetic analysis of the homologous 401 nucleotide (nt) dataset identified the MGr viruses as belonging to genotype XX, whilst three additional South African tampan viruses clustered within south ern African, domestic pig genotypes III and XXII.None of the tick viruses were unique across the gene region sequenced, with each being identical to a domestic pig virus (Fig. 2).

dIScuSSIon
The two notable differences between the 1978 and 2002 surveys in the MGr with respect to the syl vatic hosts of ASFV, relate to changes in burrow infestation as well as warthog numbers.O. porcinus were found in 59 out of the 98 burrows surveyed (60 % infestation rate) in 2002, whilst Thomson (1985) reported that ticks were retrieved from 13 out of the 40 burrows (33 %) surveyed in 1978 in MGr.These results indicate that tick populations were present in 27 % more burrows in 2002 than in 1978, probably due to the estimated 59 % increase in warthog numbers from 1984 to 2002.The marked increase in warthog numbers, together with the ob servation that 10.8 % and 47 % of warthogs sam pled outside their burrows in Namibia and in South Africa, respectively, were infested with Ornithodoros nymphs (Horak, Biggs, Hanssen & Hanssen, 1983;Horak, Boomker, De Vos & Potgieter, 1988), lends credence to the proposed role played by warthogs in ASF epidemiology, as disseminators of the inver tebrate tick host (Thomson 1985).An increase in ASFV prevalence could thus be anticipated in MGr because of the increase in warthog numbers and concomitant increase in burrow infestation rates and resulting tickwarthog contacts.This was not, however, the case.Instead, none of the 348 ticks screened using a molecular approach developed specifically for ASFV screening of Ornithodoros por cinus (Bastos et al. 2009)

KAB/62
Genotype VIII ticks sampled in 1978 were identical to each other and to two historical, genotype XX domestic pig strains, recovered from ASF outbreaks in South Africa, in 1973 and 1975 (Fig. 2).These MGr vi ruses were, however, distinct from the three other South African tampan isolates, with the Kruger Na tional Park (KNP) tampan virus clustering within genotype XXII and the Thabazimbi and Warmbaths (Bela Bela) tick viruses being assigned to genotype III.The two domestic pig viruses, Lillie and 24823, shown to be identical to the MGr tick viruses, were also HAdnegative, with isolate 24823 having re duced virulence in domestic pigs (Pini 1977).These results are consistent with the observations of Thom son and coworkers (1983) that the three Mkuze vi ruses were unusual for their nonhaemadsorbing characteristic and reduced virulence in domestic pigs.other atypical features of the ASFstatus of the MGr sylvatic hosts are the unusually low sero prevalence of the vertebrate (warthog) host (2 % and 6 % on IEoP and ELISA, respectively).In par ticular, it was noted that only one of three warthog subyearlings from the same sounder, was found to be seropositive to ASF (Thomson et al. 1983).This unexpected result indicates that the specificity/ accuracy of ASF serological tests in warthogs is possibly suboptimal and underscores the need to investigate and validate tests specifically for sylvatic vertebrate hosts (jori & Bastos 2009).

MOZA-98
Transformation of the ASFV incidence by sex and size class of Ornithodoros ticks sampled at three southern African localities (Table 4 of Thomson et al. 1983) revealed that at all localities, more adult males than females were recovered, with the aver age male:female ratio being 61:39.The 62:38 ratio recorded in the 2002 MGr survey is therefore con sistent with these previous observations of male bias.However, differences in the size classes were pronounced with substantially more nymphs than adults being collected in earlier surveys (Pini 1977;Thomson et al. 1983).With the exception of KNP, 6.7 more nymphs than adults (13 % adults: 87 % nymphs) on average, were collected from tick pop ulations sampled in the 1970s by Thomson and co workers (1983) FIG. 1 Map of Mkuze Game reserve depicting the distribution of the 98 warthog burrows (M1-M98) investigated for the presence of Ornithodoros

FIG. 2
FIG. 2 Genetic relatedness of the three Mkuze Game reserve tampan viruses isolated in 1978 (indicated in bold with grey shading) and three additional South African tampan viruses (denoted by grey shading alone), based on Cterminal p72 gene sequence analysis Nodal support ≥ 50 % from 10 000 bootstrap replicates is indicated next to each node and the previously defined genotypes recovered following neighborjoining analysis of the homologous 401 nt dataset, are shown to the right of the relevant lineage harboured the virus.Analysis of the Cterminal p72 gene region revealed that the three MGr viruses isolated from pooled This then begs the question as to whether ASFV still exists within the MGr.If the virus still persists then it must be extremely localized and restricted to burrows that were not surveyed in our study.As ASF is a controlled disease of economic signifi cance, it has major implications for the development of pig industries in northern KZN and also impacts on trade in indigenous swine.Any change in ASFV infection of sylvatic hosts in MGr would require re assessment of the continued inclusion of MGr with in the ASF control area.This PCrbased pilot study has highlighted the possible disappearance of the virus from MGr and the value of performing adult Ornithodoros tick targeted molecular surveys.It has also indentified a need for further, more extensive studies to establish beyond doubt whether ASFV still persists within the greater MGr area.concluSIon Ninetyeight burrows examined for the presence of O. porcinus in MGr yielded 348 O. porcinus ticks in 59 burrows.Despite a 27 % increase in burrow in festation and 59 % increase in warthog density since 1978, a nearly sixfold increase in proportion al sampling of adult ticks versus nymphs, and em ployment of a sensitive PCr screening method, no evidence of ASFV genome presence could be found.It is therefore possible that ASFV no longer exists within MGr and that if it does, that the ASF infection rate is extremely low and restricted to a small number of warthog burrows.We propose that molecular screening of adult ticks is reliable and costeffective for evaluating ASFV control area sta tus and for conducting ASFV surveillance in areas of particular concern.