Epidemiological and Public Health Significance of Toxoplasma gondii Infection in Wild Rabbits and Hares: 2010–2020

Toxoplasmosis is a zoonosis of global distribution, and Toxoplasma gondii infections are common in humans and animals worldwide. Hares and rabbits are important small game species, and their meat is consumed by humans in many countries. Demand for rabbit meat for human consumption is increasing; therefore, toxoplasmosis in rabbits and hares is of epidemiological significance. Viable T. gondii has been isolated from rabbits. The present review summarizes worldwide information on the seroprevalence, parasitological investigations, clinical cases, isolation, and genetic diversity of T. gondii in wild rabbits, free domestic rabbits, hares, and other rabbits from 2010 to 2020. Differences in prevalence, susceptibility, genetic variants, and clinical implications of T. gondii infection in rabbits and hares are discussed. This review will be of interest to biologists, parasitologists, veterinarians, and public health workers. Additional studies are needed to increase our knowledge of genetic variants and the population structure of T. gondii in rabbits and hares and to understand the differences in susceptibility to T. gondii in hares in different areas.


Introduction
Toxoplasmosis is a zoonosis of global distribution caused by the protozoan parasite Toxoplasma gondii, the only species in the Toxoplasma genus. Toxoplasma can infect virtually every warm-blooded animal, including humans and livestock, as intermediate hosts [1]. Felids (domestic and wild) are the definitive hosts, and the only hosts able to excrete oocysts to the environment. Toxoplasma gondii is considered one of the most important foodborne and waterborne parasites of veterinary and medical importance worldwide [2,3].
Consumption of undercooked or raw meat containing tissue cysts is the primary way of infection of T. gondii for humans [1,4]. Humans can also become infected by consuming food or drinking water contaminated with oocysts or by accidentally ingesting oocysts from the environment. In addition, humans, especially hunters, may also acquire infection through contact with the parasite while field dressing game [1,5,6].
After being consumed in uncooked or raw meat containing tissue cysts (carnivores) or in feed or drink contaminated with oocysts (all warm-blooded animals), T. gondii initiates intestinal replication. Bradyzoites and sporozoites, respectively, are released and infect the intestinal epithelium [1], and then tachyzoites are disseminated to other tissues via the bloodstream and lymph throughout the body and replicate intracellularly until the cells break and cause tissue necrosis. Young and immunocompromised humans and animals may domestic and hunted rabbits. The literature cited is from 2010-2020; information before 2009 was reviewed previously [1,24].

Serological Investigation in Rabbits
Detection of antibodies is important for epizootiological studies in rabbits because most infections are subclinical [1]. Serum antibodies to T. gondii have been found in rabbits, mainly in the European rabbit (O. cuniculus), in several surveys worldwide. Data from 2010 to 2020 are summarized in Table 1. Antibodies to T. gondii in rabbits were reported ranging from 0.9% to 37.5%, with most studies showing seroprevalence levels between 10-15% (Table 1). The highest prevalence in domestic rabbits (O. cuniculus) was reported in Egypt: 26.7% [29] and 37.5% [30], respectively. IgM antibodies were observed in nine slaughtered rabbits (6.0%) as an indication of active infection [30]. Additionally, in Egypt, antibodies against T. gondii were detected in 11.3% of farmed rabbits [31] (Table 1). High seroprevalence (23.4%) was also observed in free domestic rabbits in a rural area in China, and viable T. gondii was isolated [28] (Table 1).
In free wild rabbits, the most comprehensive survey was from Australia and involved 2114 wild rabbits (O. cuniculus) sampled over a 12-year span [20]. Mean seroprevalence was 9.9%. Lower seroprevalences were found in wild rabbit studies in Europe: 3.3% of 548 wild rabbits in Scotland [36] and 2.8% of 36 wild rabbits in Portugal [35] (Table 1). Similarly, in another rabbit species, the wild cottontail rabbit (S. floridanus), low seroprevalence of T. gondii antibodies (2.1%) was observed in Italy [13]. Wild rabbits and domestic rabbits that have access to the outdoors are likely more susceptible to T. gondii infections than pet rabbits confined indoors due to the possibility of direct contact with the contaminated environment [17].
Animal management (living and rearing outdoors or in a pet shop), the possibility of consuming contaminated vegetables, and cohabitation with cats are the main risk factors that could account for some of these observed differences. Additionally, different serological tests used, types of samples analyzed, cut-off titers, and sample sizes could contribute to these differences as well.
The Indirect fluorescent antibody test (IFAT) is a well-established technique for detecting anti-T. gondii antibodies in different animal species, including rabbits, but requires conjugate and is not automated. ELISA has the advantage of being automated and allows the testing of many samples, but also needs species-specific conjugates. A commercial ELISA is available for rabbits (see legend Table 1). Therefore, many studies in rabbits and hares rely on agglutination tests that do not need species-specific secondary antibodies. The specificity, sensitivity, and cut-off value of these serological tests have been validated in a few studies (see Section 5). Development and standardization of diagnostic tests for T. gondii infection are still necessary [3].

Seroprevalence in Hares
In the last decade, there have been several reports of T. gondii antibodies in hares from Europe as well as in two studies from China ( Table 2).
The European brown hare (L. europaeus) was the most studied hare species. Prevalence of anti-T. gondii antibodies ranged from 0% to 21% among studies in different countries ( Table 2). The low seroprevalence (0% in L. europaeus and 4.2% in L. timidus) observed in Finland, concomitantly with cases of fatal toxoplasmosis, might indicate a natural, inherent susceptibility of these hare species to T. gondii at least in those areas [40]. On the other hand, higher seroprevalence levels have been observed in other countries, e.g., France, Austria, Czech Republic or Slovakia [21,41], suggesting that European brown hares can survive T. gondii infection in the wild in other areas [6]. In another hare species, the Iberian hare (Lepus granatensis) endemic in Spain, no clinical cases were detected, and a moderate to low seroprevalence (11.4%) was observed [5]. In that study, significantly higher seroprevalence was observed in juvenile Iberian hares compared with the adult ones [5], and the authors suggested that possibly T. gondii infection affected the survival of infected hares, and thus infected juveniles reached the adult stage at lower rates than the non-infected hares or that a short-lived humoral immune response against T. gondii existed in this species [5].

Isolation of Viable T. gondii DNA from Rabbits and Hares
Viable T. gondii was isolated by bioassay from rabbits in Argentina [44,45], Brazil [23], and China [28] in the last decade (Table 3). In Argentina, isolation was performed in an aborted rabbit dam [44,45] with disseminated toxoplasmosis (Table 3). Two seropositive rabbits in Brazil were bioassayed in mice and in cats, and one cat excreted T. gondii oocysts after consumption of one of the rabbit's brain. Both strains were pathogenic to mice [23]. In China, viable T. gondii was isolated from one seropositive rabbit. The isolate was also pathogenic to mice [28]. Little is known about the genetic diversity of T. gondii in rabbits and hares, but three of the four isolates characterized by PCR-RFLP markers were different genotypes (Table 3).

Detection of T. gondii DNA in Tissues of Rabbits and Hares
The most frequently used target genes for T. gondii PCR in both humans and animals are the repetitive regions of the 35-copy B1 gene and 300-copy 529 bp repetitive element. Toxoplasma gondii DNA was detected in three (2.4%) of 126 domestic rabbits (O. cuniculus) from Zhengzhou, eight of 248 (3.2%) from Luoyang, and two of 96 (2.1%) rabbit brains and hearts from slaughtered rabbits from food markets in China [7] (Table 4). A very high prevalence of infection was recently reported in European wild rabbits from central Portugal, with 67.9% of 28 hunted rabbits showing T. gondii DNA based on a nested PCR targeting the surface antigen 2 (SAG2) gene [46]. T. gondii DNA was also detected in tissues of three (brain of one, skeletal muscle of two) of 144 wild cotton-tailed rabbits (S. floridanus) [13]. The tissues were collected from animals culled for demographic control of this species in Italy [13]. Shin et al. 2013 [32] found 16.2% of 142 rabbits to be positive for the presence of T. gondii DNA at higher levels than those shown by serological levels by enzyme-linked immunosorbent assay (ELISA) (10.6%). In this study, differences were observed among breeds with higher prevalence in Flemish giant and chinchilla rabbits compared to crossbreeds and New Zealand white rabbits [31]. On the contrary, a nested PCR of blood from pet rabbits in Poland showed all tested blood samples (n = 360) were negative for the T. gondii B1 gene in these pet rabbits [33].
Toxoplasma gondii DNA has also been detected in hares (  [19] detected T. gondii using a qPCR targeting the 529 bp repetitive region in the tissues of two of 36 hares killed or found dead (5.5%). One of the infected hares had high levels of parasites in the spleen and lung, a unique finding. Toxoplasma gondii DNA was also detected in the Tolai hare (Lepus tolai) in three different areas of China [13] (Table 4). On the other hand, T. gondii infection was not found by PCR in the liver of seropositive Iberian hares (L. granatensis) in Spain [5] or in livers of L. europaeus in Greece [40]. In the latter species, Aubert et al. (2010) [41] were not able to detect viable T. gondii from heart tissues using a bioassay in mice; it should be noted that only three seropositive L. europaeus were bioassayed.

Validation and Comparison of Serological Results by Different Techniques in Rabbits
Little is known of the validity of different serological tests for the detection of T. gondii antibodies in rabbits. In a recent study, validation of the modified agglutination test (MAT) for detection of T. gondiispecific antibodies was performed using wild rabbit samples and was compared to an indirect fluorescent antibody test (IFAT) as a reference test [20]. The two assays showed 94.3% agreement, with the MAT performing with a sensitivity of 84.1% (95% CI: 69.9-93.4%) and specificity of 96.7% (95% CI: 93.0-98.8%) relative to the IFAT.
In a study on pet rabbits in Japan, 337 serum samples were tested by different serological methods; initially, IgG and IgM ELISAs were performed, and those samples seropositive by ELISA were further analyzed using a latex agglutination test (LAT), Western blotting (WB), and an indirect immunofluorescence assay (IFAT) [17]. The rates of seropositivity for T. gondii were 0.89% (3/337) and 0.29% (1/337) for the IgG and IgM ELISAs, respectively. SAG1 and SAG2 were detected as major antigens by the positive rabbit sera in WB analyses and were associated with strong staining observed by IFAT in T. gondii tachyzoites [17]. In pet rabbits from breeding farms in Korea [32], whole blood was analyzed by serology, nested PCR, and immunoblotting (three samples negative and six samples positive based on ELISA). Interestingly, there was a higher T. gondii prevalence as detected using nested PCR (16.2%) relative to ELISA (10.6%). The authors suggested that the nested PCRpositive/ELISA-negative rabbits might have harbored prepatent or latent infections with T. gondii. All PCR-positive samples were sequenced and corresponded (99% homology) to partial T. gondii B1 gene sequences.

Cinical Cases in Hares and Rabbits in the Last Decade
Infections in rabbits are documented to be mainly subclinical [36]. On the other hand, since the 1950s, high susceptibility and fatal systemic toxoplasmosis cases have been reported in European Brown hare (L. europaeus), mainly from Scandinavia, where harsh winter conditions are common [1,6,40]. To our knowledge, these epizootics have not been seen in other countries. Reports of clinical toxoplasmosis in rabbits and hares, preceding the year 2010, were reviewed earlier [1]. Only two studies reported acute toxoplasmosis in a few rabbits in the USA [1]. Since then, a retrospective report of fatal acute disease caused by T. gondii was diagnosed in a domestic rabbit in Brazil in which T. gondii was identified by immunohistochemistry (IHC) in histological sections of spleen and liver tissue and genotyped [11] (Table 5). The animal was from a farm where there were 400 young rabbits (approximately 40 to 45 days old) that developed diarrhea and severe dehydration after weaning. A total of 150 animals died during a 2-month period, 24 h after the onset of clinical signs [11].
As indicated above, an outbreak of abortion occurred on a small rabbit farm in Argentina [44]. A rabbit that had aborted was necropsied; there was no information on the aborted fetus. The rabbit had disseminated toxoplasmosis, and the diagnosis was confirmed by IHC testing, PCR, and isolation of viable T. gondii [45] (Table 3).
In hares, in a retrospective study of L. europaeus and L. timidus that died of visceral toxoplasmosis, the results were confirmed by positive serology and IHC. The proportional mortality rates and the T. gondii antibody prevalence differed significantly between the two host species. Limited genotyping revealed ToxoDB genotype #1 (type II) in those animals [40]. In another study, two hares were found dead in the Czech Republic and macroscopic observations indicated acute toxoplasmosis [19]. One hare had a very high number of T. gondii parasites detected in the lungs and spleen, exceeding 1,000,000 parasites/gram. Both studies on L. europaeus hares, in Finland and in the Czech Republic, revealed the genotype to be T. gondii type II clonal lineage [19,40]. Retrospective study of wild brown hares that died. Genotyped (see Table 4). Yes Yes [40] 2 found dead, Czech Republic Acute toxoplasmosis, one animal with more than 1 × 10 6 parasites/g of tissue. Genotyped (see Table 4).

Histopathology and Immunohistochemistry
Microscopic lesions of T. gondii infection in rabbit tissues (brain, heart, and diaphragm) revealed mainly granuloma, mononuclear cell infiltrates, and degeneration areas; necrosis was observed in brain and heart tissues [26].
Splenomegaly seems to be a common finding for rabbits and hares with fatal toxoplasmosis [11,40]. In a retrospective study of spleen and liver tissue of a domestic rabbit with fatal acute disease caused by T. gondii in Brazil, the spleen was grossly enlarged and had extensive necrosis of parenchyma [11], as well as a marked quantity of transparent fluid in the abdominal and thoracic cavity, likely attributable to hepatomegaly and congestion. In hares with very high parasitic burdens [19], the most typical gross findings in T. gondii positive animals were pulmonary congestion, splenomegaly, and granulomatous and necrotic lesions in the liver. As indicated above, in this study, a high number of T. gondii parasites were detected, especially in the lungs and spleen (more than 1,000,000 parasite counts per gram of tissue). In other hare studies, the pathomorphological findings consisted of enormous enlargement of the spleen, lung edema, enlarged mesenterial lymph nodes and paleness, an uneven color, and/or hemorrhagic areas in the liver [40]. Most of the animals had an otherwise good body condition.
The Toxoplasma gondii load in tissues of asymptomatic animals is low [1], thus the chances of detection of the parasite by histopathology and IHC are also low. To our knowledge, only a few studies [11,40,45] have evidenced the presence of T. gondii in rabbits and/or hares by IHC in the last decade. Jokelainen et al. (2011) [40] performed retrospective IHC on formalin-fixed liver tissue samples from hares with fatal toxoplasmosis as the recorded cause of death, and toxoplasmosis was confirmed by IHC and positive serology in 8.1% (14 of 173) of brown hares (L. europaeus) and 2.7% (4 of 148) of mountain hares (L. timidus) tested. The liver was chosen because it is typically affected in fatal toxoplasmosis in hares, and it has successfully been used as the organ of choice for IHC. Based on IHC detection of the parasite in liver tissue, 17 of 18 (94%) cases originally diagnosed as fatal toxoplasmosis of hares were retrospectively confirmed, and one new case was found. Only one hare with recorded toxoplasmosis was IHC negative.
In the rabbit that had aborted at a farm in Argentina, disseminated toxoplasmosis was confirmed by IHC testing, PCR, and isolation of viable T. gondii [45] (Table 3). T. gondii was also demonstrated retrospectively by IHC in histological sections of spleen and liver tissues of one rabbit suffering fatal acute toxoplasmosis in Brazil [11].

Genotypes Based on Isolation of Viable T. gondii
In the studies that isolated viable T. gondii from domestic rabbits by bioassays (Table 3), the genotypes reported were #48 from Argentina, #11 and #19 from Brazil, and #2 from China [23,27,44].

Genotypes Based on Isolation of T. gondii DNA from Tissues
Based on genotyping in tissues in the last decade, different genotypes were observed in rabbits and hares. In rabbits, interestingly, in a household where raw rabbit meat was eaten and cases of human toxoplasmosis occurred, a virulent strain of the parasite was isolated from the brain of a rabbit fed green fodder from the home garden in Poland, but no details were given by Sroka and Szmańska (2012) [34].
In the case of fatal acute toxoplasmosis in a domestic rabbit in Brazil [11], genotyping of DNA isolated from the frozen spleen revealed nonarchetypal ToxoDB genotype #8 [11], type BrIII genotype, which is a typical clonal Brazilian lineage (Table 4); this was the first description of acute disease associated with this genotype in a non-human naturally infected animal host. The genotype found in Brazil reported the first occurrence of the 291 allele for the typing marker TUB2 in a type BrIII strain, emphasizing that isolates homogenously classified as type BrIII via PCR-RFLP can be genetically diverse as well as the genetic diversity of T. gondii in Brazil [11].
As indicated above, in the European brown hares, ToxoDB genotype #1 (type II clonal lineage) has been reported from clinical samples in Finland [40] and the Czech Republic [19]. Both studies used microsatellites for genotyping; six microsatellites were used in the Finnish study, and 15 microsatellites were used in the Czech report. In the Czech Republic, one T. gondii isolate was archetypal clonal type II and the other a type II variant (W35 = 244). The animal infected with Type II variant W35 had severe disseminated toxoplasmosis, with an estimated 7 million T. gondii per gram of tissue [19].
In China, isolation and genotyping of T. gondii DNA in two studies, one in rabbits and the second in Tolai hares, found genotype #9 [7,14]; this genotype has been previously identified in several hosts, including domestic animals in China [14].

Conclusions
Although in the last decade there have been several studies performed on the epidemiology and clinical features of T. gondii in domestic and wild rabbits and hares, more studies on the prevalence and clinical and immunological implications of T. gondii infection in rabbits and hares are needed. Additional studies would also help to understand the differences in susceptibility to T. gondii in hares in different areas and to increase our knowledge of genetic variants and the population structure of T. gondii in rabbits and hares. More studies of the specific factors that exacerbate disease in rabbits and hares are also needed.
Viable T. gondii could be present in the organs of seropositive rabbits and hares. Therefore, hunters should be aware of T. gondii infection in clinically healthy rabbits and hares and are advised to take precautions while field dressing game and to freeze or cook rabbit and/or hare meat before human consumption [5]. In domestic rabbit farms, it is important to implement good hygienic conditions and, if possible, prevent the presence of cats and dogs on the farms [18].

Conflicts of Interest:
The authors declare no conflict of interest.