3. Rabbit haemorrhagic disease (RHD)
3.3 Evolution of RHDV in the Iberian Peninsula: A brief review of recent findings
findings.
Adapted from: II Seminario Internacional sobre el Conejo Silvestre. Córdoba 28-30
Abril 2010 (in press)
Muller, A. and Thompson, G.
Departamento de Clínicas Veterinárias, Instituto de Ciências Biomédicas Abel Salazar (ICBAS), Universidade do Porto
Adress: R. Padre Armando Quintas, P- 4485-661, Vairão-Portugal
99 Abstract
Since the early 1990s, rabbit haemorrhagic disease (RHD) has caused high mortality rates in wild rabbit populations in the Iberian Pensinsula. It is now considered endemic and continues to have considerable impact on species conservation and hunting industry. Here we review current knowledge on the evolution of rabbit haemorrhagic disease virus (RHDV) in the Iberian Peninsula. Our previous work (Muller et al. 2009, Vet. Microbiol. 135: 368-373) as well as recent work on the origin and phylodynamics of RHD (Kerr et al., 2009, J. Virol. 83, 12129-12138; Kinnear and Linde, 2010, J. Gen. Virol. 91, 174-181) is presented and discussed. Currently available sequence data of RHDV of the Iberian Peninsula derive from the beginning of RHD outbreaks (AST89, MC89) and from different geographic locations in Portugal (40 samples) and Spain (3 samples), as well as Southern France (3 samples) obtained between 1994 and 2007. Phylogenetic analyses based on partial VP60 gene sequences allowed a grouping of these RHDVs into three groups, termed “Iberian” Groups IB1, IB2 and IB3. Interestingly, these three Iberian groups clustered separately, though not far from earlier RHDVs of Genogroup 1 (containing e.g. strain “AST89”), but clearly distinct from globally described RHDV strains. This result gave rise to the hypothesis that the virus evolved independently on the Iberian Peninsula, with the Pyrenees acting as a natural barrier to rabbit and virus dispersal (Muller et al 2009). No differences were observed in RHDV sequences obtained from geographic regions where the rabbit subspecies O. c. algirus prevails compared with those obtained from O. c. cuniculus. The distinct clustering of these same Iberian RHDV sequences was confirmed by more sophisticated phylogenetic analyses (Kerr et al., 2009). Times to most common recent ancestor (TMRCA) for the different RHDV branches were estimated. These gave a mean date of 1948 for the origin of “Iberian” strains. Both studies also estimated that virulent RHDV probably emerged in the early 1900s, and that multiple virus lineages were already circulating before the first reports of disease in 1984 (Kerr et al., 2009; Kinnear and Linde, 2010). Both studies relate the emergence of virulent RHDV to the intensification of rabbit production. We discuss these findings and conclude that if virulent RHDV already existed in the Iberian Peninsula in the 1950s, their impact could have been masked by mortalities due to the concomitant recent introduction of myxomatosis. Despite evidence of genetic differences between RHDV field strains currently circulating in the Iberian Peninsula and those circulating elsewhere; there is no evidence that these affect protection induced by current commercially available vaccines. Antigenic variants (RHDVa) have been found in a commercial rabbit farm in Portugal, but not yet in wild rabbits. To our knowledge, no avirulent forms of RHDV or RCV-like viruses have yet been found in Iberian wild rabbit populations, but further investigations are
100
warranted to address these issues. Also, more strains, especially from different locations in Spain, should be analysed to enhance current understanding of RHDV evolution on the Iberian Peninsula. Work was funded by FCT: SFRH/BD/31093/2006.
101 Introduction
Similar to other European countries, myxomatosis and rabbit haemorrhagic disease (RHD) have been introduced into the Iberian Peninsula in the 1950s and early 1990s, respectively (Anonymous, 1989; Monteiro, 1999; Muñoz, 1960; Villafuerte et al., 1995). Within a few years of their introduction in wild rabbit populations, a severe decline in rabbit abundance was recorded in the Iberian Peninsula to the extent that the wild rabbit is currently considered “near threatened” by the World Conservation Union in 2008 (Smith and Boyer, 2008). RHD is now considered endemic in Spain and Portugal, and despite many efforts, rabbit numbers have not fully recovered (Delibes-Mateos et al., 2008b, 2009; Dias-Pereira et al., 2004; Moreno et al., 2007; Muller et al., 2004; Santos et al., 2006; Villafuerte et al., 1995; Ward, 2005). The objective of the present manuscript is to review current knowledge on the evolution of rabbit haemorrhagic disease virus (RHDV) in the Iberian Peninsula. We will briefly present our previous work (Muller et al., 2009) as well as recent work on the origin and phylodynamics of RHD (Kerr et al., 2009; Kinnear and Linde, 2010) and discuss the practical implications of these findings.
1. Brief history
RHD was first reported in China in 1984. The disease spread throughout Europe, also reaching the Iberian Peninsula, between 1987 and 1989. In the 1990s, the etiological agent of RHD was characterised as a calicivirus (Ohlinger et al., 1990), and the virus was introduced into Australia and subsequently New Zealand as bio-control agent against the wild European rabbit population (Forrester et al., 2003). In an attempt to understand the molecular epidemiology and ultimately the origins of RHD, phylogenetic analyses based on partial sequences of the viral capsid protein VP60 of many viral isolates from different countries were conducted (Asgari et al., 1999; Forrester et al., 2006; Le Gall-Recule et al., 2003; Matiz et al., 2006; McIntosh et al., 2007; Moss et al., 2002; Nowotny et al., 1997). These showed that there were up to six or more viral groups or subgroups and that virus strains may or may not cluster according to their geographical origin and/or the year of isolation, but that they did not always do so. Also, evidence was gathered on the current and previous existence of RHDV-like viruses, collectively denominated rabbit caliciviruses (RCV), which may have been circulating more or less harmlessly in Europe for many years before the first epidemic appearance in China in 1984 (Forrester et al., 2006; Moss et al., 2002; Rodak et al., 1990). RHDV-specific signals were detected by RT-PCR in rabbit sera taken in the 1950s, concomitantly with the detection of RHD antibodies, but infectiousness of viral RNA could not be proven (Moss et al., 2002). So, until recently, a
102
lot of information on the molecular epidemiology of RHDV in Europe and other countries was gathered. However, detailed information on RHDV in the Iberian Peninsula was still lacking, and further questions on the origin of RHD remained.
2. Evolution of rabbit haemorrhagic disease virus (RHDV) in the European rabbit (Oryctolagus cuniculus) from the Iberian Peninsula
RHDV strains obtained from wild rabbits in different years and geographic locations in the Iberian Peninsula were recently characterised (Muller et al., 2009). Briefly, 47 partial RHDV sequences were analysed, that were obtained from wild rabbits that were found dead during known epidemics of RHD in Portugal, Spain and France between 1994 and 2007. The majority, 40 viral sequences, were obtained from different geographic locations of Portugal. Three RHDV sequences were obtained from wild rabbits from Spain: two from Alicante and Albacete collected in 2004 (Dr. Ramon Soriguer, CSIC), and one from Toledo collected in 1994 (CIBIO). Three of the four recent RDHV sequences from France were collected between 2000 and 2002 in the Department “Pyrénées-Orientales”, and the fourth in 2005 in the Department “Manche” (Dr. Le Gall-Recoulé, AFSSA). Nested PCR was used to amplify partially the RHDV VP60 capsid protein gene (Moss et al. 2002). The amplicons of 547bp were directly sequenced in both directions. The phylogenetic tree was estimated by the neighbor-joining method as considered adequate for comparing relatively short sequences (Takahashi and Nei, 2000). The obtained sequences were compared with published homologues from GenBank database (NCBI). Eighteen sequences representing previously described RHDV genogroups were selected and included. These sequences were grouped into Genogroup 1 to 6, adapted to the classification used by Le Gall-Reculé et al. (2003). The results of this phylogenetic analysis allowed a grouping of these RHDV strains into three groups, termed “Iberian” Groups IB1, IB2 and IB3. Interestingly, these three Iberian groups clustered separately, though not far from earlier RHDVs of Genogroup 1 (containing e.g. strain “AST89”), but clearly distinct from globally described RHDV strains. The Groups IB1 and IB2 contained sequences collected between 1994 and 1997, whereas Group IB3 included those collected between 2000 and 2007. Each group contained sequences from different geographic locations, except Group IB1, that contained five sequences obtained in Santarém district of central Portugal. Interestingly, Group IB2 also included the virus obtained in 1994 from Toledo, Spain, and Group IB3 also included the three sequences from the French Department “Pyrénée- Orientales”, as well as the Spanish strains “Albacete” and “Alicante”. On the contrary, the sequence obtained in 2005 from the Departement Manche in the North of France, clustered amidst other European RHDV sequences of Genogroups 3-5. These findings
103
led to the hypothesis that virus evolution occurred independently in wild rabbit populations on the Iberian Peninsula, with the Pyrenees acting as a natural barrier to rabbit and hence to virus dispersal
3. The origin and phylodynamics of RHDV
Recently, two detailed and sophisticated phylogenetic analyses on RHDV were published (Kerr et al., 2009; Kinnear and Linde, 2010). Importantly, Bayesian Markov Chain Monte Carlo analyses were applied, allowing the estimation of viral nucleotide substitution rates, as well as times to most common recent ancestors (MCRA), adding a temporal dimension to the evolutionary reconstruction of RHDV. Despite some differences in methodology and results, both studies agree on the identification of four distinct monophyletic groups or clades of RHDV. Although contrasting previous studies (Forrester et al., 2003; Le Gall- Recule et al., 2003; Moss et al., 2002), these findings are contributing substantially to the warranted harmonisation of RHDV classification. Group 1 (Clade D) contained the antigenic variants. Group 2 (Clade C) contained predominantly strains from the Iberian Peninsula and South of France. Group 3 (Clade A) contains the “original strain” dating to China 1984 as well as viruses from central Europe between 1987 and 1993, Korea and Mexico. Group 4 (Clade B) contained strains from central Europe between 1989 and 2004 and Bahrein. The estimated dates for MRCA of the different groups of RHDV indicated that virulent RHDV probably emerged in the early 1900s, and that multiple lineages were likely circulating long before the first reports of disease in 1984. Both studies related the evolution of RHDV with the intensification of rabbit production. Importantly, one of the studies included some of our recent sequence data on Iberian RHDV, and confirmed our findings of distinct clustering of Iberian strains (Kerr et al., 2009). Times to MRCA indicated a mean date of 1948 for the origin of “Iberian” strains. Within this predominantly “Iberian” group, two subgroups were described: viruses isolated between 1989 and 1997 and those isolated between 2000 to 2007, which have a mean TMRCA of around 1962. The rabbit caliciviruses (RCV) displayed considerable divergence between each other and also from all RHDV lineages, and a common ancestor was estimated to fall in 19th or early 20th century (Kerr et al., 2009; Kinnear and Linde, 2010). Thus, virulent RHDV most likely did evolve from an avirulent RCV, but this would have occurred approximately 300 years ago, and therefore would not explain the first observed epidemic in 1984 (Kerr et al., 2009).
104 4. Discussion and Conclusions
Genetic differences have been found between current RHDV field strains of the Iberian Peninsula and those elsewhere (Kerr et al., 2009; Muller et al., 2009). We hypothesized that virus evolution occurred independently in wild rabbit populations on the Iberian Peninsula (Muller et al., 2009). A mean date of 1948 for the origin of “Iberian” strains has been estimated recently (Kerr et al., 2009). This suggests that virulent RHDV must have been circulating already before the first recorded rabbit mass mortality in the late 1980s and early 1990s in Spain and Portugal. The following question arises: is it possible that RHD outbreaks already occurred in wild rabbit populations well before the 1980s, e.g. in the 1950s and 1960s? We think it is possible and that they would likely have gone unnoticed. Rabbit mortalities from the 1950s onwards coincided with the known introduction of myxomatosis. RHD is an acute infection that induces discrete if any visible signs and lesions, requiring necropsy for diagnosis. Other co-factors may have played a role in incrementing the impact of RHD in wild rabbit populations between the 1950s and 1980s. Maybe the most notorious being major changes in agricultural practices, leading to an increase in rabbit habitat defragmentation. Finally, only with the first description of RHD in 1984, investigations on this new disease exploded and the disease was subsequently actively searched for and thus reported from many countries, including Spain and Portugal (Anonymous, 1989; Monteiro, 1999; Villafuerte et al., 1995). Current sequence data consists mostly of strains obtained in Portugal, and it is unclear whether these are representative for the whole of the Iberian Peninsula. Analyses of more RHDV field strains especially from Spain and South of France are required to strengthen current findings and to gain further insight into viral evolution and hopefully origins of RHD in the Iberian Peninsula.
Only one single serotype of pathogenic RHDV is known, which contains two major subtypes, denominated RHDV and the antigenic variants denominated RHDVa (Capucci et al., 1998a; Schirrmeier et al., 1999), which have by now been described worldwide (Farnos et al., 2007; Kerr et al., 2009; Le Gall-Recule et al., 2003; Matiz et al., 2006; McIntosh et al., 2007). Although antigenic characterisations of current Iberian RHDV has not been carried out, to our knowledge there is no evidence of antigenic variation. Current vaccines are thus expected to be fully protective. Interestingly, although antigenic variants (RHDVa) have been found in a commercial rabbit farm in Portugal (Muller et al., 2009), these have not yet been found in wild rabbits in the Iberian Peninsula. Rabbit caliciviruses (RCV) are a group of predominantly apathogenic viruses that display some antigenic similarities, but are phylogenetically distant to RHDV (Bergin et al., 2009; Capucci et al.,
105
1996; Forrester et al., 2009; Forrester et al., 2007; Kerr et al., 2009; Moss et al., 2002; Strive et al., 2010; Strive et al., 2009). The presence of a nonpathogenic virus closely related to RHDV was already suspected at the time of the first RHD seroepidemiological surveys, when specific RHDV antibodies were found in sera of farm and laboratory rabbits where no disease was reported, and also in rabbit sera collected before the appearance of RHD (Moss et al., 2002; Rodak et al., 1990). The drastic reduction in wild rabbit numbers observed on the Iberian Peninsula has been historically unprecedented, suggesting that RHDV-like apathogenic viruses were either not circulating, or, if they did, they were probably not fully cross-protective. Active investigations on the presence of avirulent RCV-like viruses on the Iberian Peninsula seem limited. We have analysed 68 samples from 2 healthy wild rabbit populations by nested RT-PCR. All samples were negative, however serology was not performed (Muller et al., 2009). Thus to our knowledge, no avirulent forms of RHDV nor RCV-like viruses have been found in Iberian wild rabbit populations, but further investigations are warranted to address these issues as they are expected to significantly add to current understanding of RHDV epidemiological history.
Acknowledgements
107