The potential role of scavenging flies as mechanical vectors of Lagovirus europaeus/GI.2

Rabbit haemorrhagic disease virus (RHDV) is a calicivirus belonging to the genus Lagovirus that causes rabbit haemorrhagic disease (RHD) in adult European rabbits (Oryctolagus cuniculus). The virus has a hepatic tropism and causes cell loss as a result of virally-induced apoptosis leading to a necrotizing hepatitis (reviewed in [1]). Originally detected in China in 1984, RHDV rapidly disseminated worldwide and reached almost all continents, becoming enzootic in several countries (reviewed in [1]). In addition to the significant ecological losses associated with the crashes in the wild rabbit populations, RHD has also led to serious negative economic impacts in the rabbit-associated industries (e.g. [2, 3]). In 2010, French rabbit populations experienced atypical RHD outbreaks [4]. These were shown to be caused by a novel RHDV genotype later named Lagovirus europaeus/GI.2 [5], which is antigenically dissimilar and fatally infects kittens as young as 11 days old [6]. GI.2 replaced older RHDV circulating strains (genotype GI.1; [7‐10]) and quickly spread worldwide [11, 12]. Recombination is an important mechanism in GI.2 evolution with all known strains resulting from recombination events with either pathogenic (GI.1b and GII.1) or non-pathogenic strains (GI.3 and GI.4), including those associated with the first outbreaks [13‐16].

Transmission of RHDV occurs via contact of a susceptible rabbit with an infected animal or carcass, through contaminated surfaces, food, burrows, cages, etc., or by vectors such as insects, scavenging birds and mammals (reviewed in [1]). Calliphorid and muscid flies were first implicated in GI.1 RHDV dissemination following the escape of the virus from a quarantine compound on the off-shore Australian Wardang Island, subsequently spreading to mainland Australia [17, 18]. Similarly, spread of RHDV to the United Kingdom was also suggested to have been mediated by insect vectors [19]. Later investigations, both in the field and under laboratory conditions, suggested that flies (Calliphora, Chrysomya, Hydrotaea, Lucilia, Musca, Oxysarcodextia and Sarcophaga genera), fleas (Spilopsyllus cuniculi and Xenopsylla cunicularis) and mosquitoes (Aedes notoscriptus, Ae. postspiraculosus and Culex annulinostris) could mechanically transmit RHDV [20‐29]. Rabbits are infected by ingestion or contact with contaminated flyspots deposited on vegetation or at burrow entrances. Absorption of virus particles from flyspots deposited on mucous membranes such as rabbit conjunctiva was also put forward as a pathway of disease transmission [20]. Flyspots were shown to contain enough viral particles to cause RHD in susceptible rabbits [4]. Yet, insects do not support RHDV replication.

A recent study by Calvete and colleagues [30] showed the inability of GI.2 mechanical transmission by the mosquito Aedes albopictus (Culicidae) and a limited ability of the sandfly Phlebotomus papatasi (Psychodidae). These insects were selected for study based on their availability in laboratory colonies and their feeding habits that render them efficient mechanical vectors of viruses.

While flies were shown to be the main vectors of RHDV in Oceania [4, 15, 17, 25], their epidemiological role in Europe is unknown. In mainland Portugal, insects from orders Coleoptera and Diptera are the most commonly found near rabbit enclosures, with Psycodidae, Scarabaeidae and Staphylinidae families and Culicoides genus being the most abundantly trapped [31]. RHDV GI.2 has been detected in insects from Mycetophilidae, Staphylinidae and Simuliidae families, Forcipomyiinae subfamily and Culicoides genus [31, 32]. However, other fly species that may constitute vectors of RHDV have not been investigated. The main goals of this study were to search for evidence of mechanical transmission of GI.2 by insects collected in Portugal in the scope of a longitudinal epidemiological study of wild rabbits and to determine the potential of these insects to act as sentinels of GI.2 circulation.

Overall, 3027 fly specimens were collected between June 2018 and February 2019 and morphologically identified. Morphological identification of each specimen was performed to the species (n = 2301), genus (n = 20), or family level (n = 706) (Table 1). The average daily number of flies collected peaked in October, with a second smaller peak in January and February (Fig. 1). The most abundant Diptera collected in every session were Muscidae and Calliphoridae specimens, as was also shown by other authors in studies carried out in the Iberian Peninsula with pig [36, 45, 46], dog [47] and cat (unpublished observations), except in January, when Drosophilidae was the second most collected family. In Portugal, Calliphora vicina and C. vomitoria can be found all year round, especially during the winter and spring seasons, while other species inside the Calliphoridae family are captured from late spring until autumn (e.g., Chrysomya albiceps, Lucilia sericata, L. ampullacea and L. caesar) [47‐49]. Flies from the genus Fannia are also present all year round, being especially abundant during all spring season [46].

Molecular identification by sequencing of a fragment of the COI gene (GenBank accession numbers OQ860783-OQ860806) and comparison by BLAST analysis with publicly available sequences in GenBank allowed the assignment of the insects into four families, split into 8 genera, in a total of 11 different species (Table 2). In general, there was a good agreement between morphological and molecular identification, with some instances with further assignment to the genus/species level based on the sequences obtained; yet, species identification was not possible for insects of the families Phoridae and Drosophilidae. All the detected species have been previously reported as occurring in Portugal, associated with carrion and organic matter decomposition [e.g. 36, 46, 47].

By RT-qPCR, GI.2 was detected in seven pools of insects, with Cq values ranging between 23 and 30 (Table 3). The RT-qPCR system employed in this study was previously shown to detect as few as nine copies of viral RNA, which corresponded to a mean Cq value of 37.56 [41]. Since the Cq values observed in the present study are below that value, these flies were considered as positives for the presence of GI.2. Positive insects were identified as belonging to four families, Muscidae, Calliphoridae, Fanniidae and Drosophilidae, and were collected in January and February 2019. These samples were further screened by conventional RT-PCR. Only one insect pool collected in January 2019 and identified as Hydrotaea armipes was positive (sample 12; Table 2). From this, we were able to sequence a 107 bp fragment within the capsid gene (GenBank accession number: OQ859633). BLAST analysis with sequences available in the GenBank database revealed the highest score with the GI.2 strain AUS/VIC/MLD-6/2017 (98.6% nucleotide identity; accession number MW460206), confirming its identity as GI.2.

The apparent seroprevalence of RHDV in the rabbit population was 30.0% (CI₉₅ 20.5–41.5%; data not shown). No seroconversions were detected among the 50 individual rabbits trapped between May and December 2018 (117 captures), and no dead rabbits were found, supporting that RHDV was not circulating in the local population during that time. No rabbits were captured during the January-February 2019 trapping session, but the fresh carcass of an adult female, initially live-captured in August 2018 and seronegative for RHDV, was found. At necropsy, lesions suggestive of RHD were found (data not shown) and RHDV RNA was amplified from liver and lung samples. Two fragments of 762 bp and 503 bp, covering p16 and part of p23, and a partial fragment of VP60, were sequenced (GenBank accession numbers OQ859632 and OQ859631). Standard nucleotide BLAST searches revealed 97.3% and 97.6% identity with GI.2 sequences (GenBank accession numbers KM115683 and KM115716) for the p16/p23 and VP60 fragments, respectively. The GI.2 sequence obtained from the fly sample 12 (also collected in January 2019) has 100% nucleotide identity with the sequence obtained from the rabbit samples, confirming it originated from the recorded RHDV GI.2 outbreak. While identical, the samples were handled independently at different periods of time, with no possibility of cross-contamination. Moreover, the sequences obtained were not identical to any of the sequences previously obtained in our laboratory, further discarding the possibility of contamination from other tested samples.

The RT-qPCR method used in this study is a highly sensitive method that detects minimum amounts of GI.2 viral RNA [41]. Previously, Gehrmann and Kretzschmar [22] found that 10–100 virus particles were the minimum dose required to induce disease in rabbits; a more recent study showed that the minimum infective dose for GI.1 is ≤ 10⁴ gRNA copies [50]. However, this has not been fully assessed for GI.2. Thus, despite the detection of GI.2 RNA in these flies, showing their ability to carry the virus, it remains to be determined if the amount of viral particles present was sufficient to induce disease in susceptible rabbits. The detection of RHDV in the abdominal content of the flies after dissection makes it unlikely that it was the result of cross-specimen contamination in the traps or in the subsequent insect pools [24]. Thus, the virus detected would likely later be excreted in flyspots. A single flyspot (faecal and regurgitation spots) has been shown to contain enough RHDV particles to infect susceptible rabbits [20, 27].

Previous observational and experimental studies showed that muscid and calliphorid flies, including Calliphora, Hydrotaea, and Musca, can transmit RHD between susceptible rabbits and might have an essential role in the epidemiology [20‐29]. Our results are in line with these studies as RHDV positive flies found in this study also belong to these families. We further detected GI.2 RNA in Drosophilidae and Fannidae flies, which had never been reported. The role of these flies in RHDV transmission remains to be determined; however, their life cycles and feeding habits lend support to their role as mechanical vectors for GI.2. Indeed, several RNA viruses belonging to families Rhabdoviridae, Dicistroviridae, Birnaviridae, Reoviridae, and Errantiviridae have been reported in Drosophila melanogaster [51], while Fannia canicularis is a known vector of Newcastle disease virus and Aleutian mink disease virus [52]. Furthermore, Fanniidae flies have a necrophagous life cycle [53], with females being attracted to decaying material, carrion and feces, as well as sweat and mucus from animals. Their larvae can also be found in vertebrate carrion and burrows, causing myasis in animals with unhealed hounds [54]. Drosophilid flies usually feed on substrates rich in bacteria, yeasts, and other fungi, but some feed on animal tissues or secretions, especially those of Amiota, Apsiphortica and Phortica genera [reviewed by 55]. Previous results reported the presence of GI.2 RNA in insects belonging to families Ceratopogonidae (genus Culicoides and subfamiily Forcipomyiinae), Staphylinidae, Simuliidae [31] and Mycetophilidae [32]. While their feeding habits and dispersal behaviour are compatible with a role as mechanical vectors for GI.2, their life cycles reduce the likelihood of acting as virus reservoirs [56]. However, since these insects were not detected in the present study, we could not further assess their role.

In conclusion, our results appear to indicate that scavenging flies C. vomitoria, C. vicina, Hydrotea armipes, Muscina levida, Fannia lepida and those from the Drosophilidae family may have a role as mechanical vectors of RHDV in the native range of the southwestern Iberian subspecies of European rabbit (Oryctolagus cuniculus algirus), as shown in Oceania [24, 27]. Furthermore, our observations suggest that the detection of GI.2 in scavenging flies might be used as a tool to monitor viral circulation [24]. Indeed, we were able to detect GI.2 in scavenging flies during a GI.2 outbreak, but not when there was no  evidence of viral circulation in the local rabbit population. Future studies are warranted to fully determine the potential of scavenging flies in the epidemiology of RHD and as monitoring tools for surveillance of RHDV outbreaks in the field.