Coronaviruses in farm animals: Epidemiology and public health implications

3.1.1 Cattle

Bovine coronavirus (BCoV) was first associated with neonatal calves diarrhoea by Mebus et al. in the United States (1973). To date, 38 BCoV genomes were published in GenBank (He et al., 2019). The BCoV is the causative agent of winter dysentery (WD) in adult cattle (Saif, Redman, Brock, Kohler, & Heckert, 1988) and causes respiratory tract infections in calves, cattle and small ruminants (Storz et al., 2000). Recent reports identified differences in antigenic, genomic and culture characteristics between respiratory and enteric BCoV strains (Vilček, Jacková, Kolesárová, & Vlasáková, 2017). Several Bovine-like coronaviruses were detected in domestic animals and were reviewed by Amer (2019).

The seroprevalence of anti-BCoV antibodies among newborn diarrheic calves ranged from 20% (17/82) in Algeria (Ammar et al., 2014) to 93.9% (172/183) in Turkey (Yavru et al., 2018). Molecular detection of the virus by qRT-PCR was reported from several countries leading to variable prevalences (Figure 2, Table 2). In Vietnam, 16 out of 232 (6.9%) diarrheic calves were tested positive and the genetic characterization of S and HE proteins reveals that Vietnam BCoV sequences might share a common ancestor with Cuban and Chinese BCoVs (Shin et al., 2019). Recently, the occurrence of BCoV in neonatal calf diarrhoea was estimated to 7.2% (14/194) in Iran using antigen-capture ELISA (Lotfollahzadeh et al., 2020).

Calves and cattle with either diarrhoea or respiratory infections are the main source of the virus (Saif, 2010) and infection by BCoV occurs either by faecal-oral or aerosol contamination. The BCoV is shed both through respiratory and enteric secretions in high amounts (1 billion virus particles per ml of faeces) for up to 14 days (Kapil et al., 1990). Some studies showed that the virus persists in sub-clinically infected adult cattle (Park et al., 2007; Toftaker, Holmøy, Nødtvedt, Østerås, & Stokstad, 2017). Furthermore, high animal density seems to be the main BCoV risk factors (Boileau & Kapil, 2010).

The BCoV is one of the main causative agents of neonatal calf diarrhoea during the first month of life (Ammar et al., 2014; Brandão et al., 2006) and the most common deathly disease in calves (Gomez & Weese, 2017). Other studies showed that BCoV may be involved in 10 to 30% of neonatal diarrhoea cases (Alfieri et al., 2018).

In adult cattle, the WD occurs as epidemics during winter and is characterized by a contagious watery diarrhoea, fever, depression and reduced milk yield (Toftaker et al., 2017). Singasa, Songserm, Lertwatcharasarakul, and Arunvipas (2017) reported that milk production might decrease up to 10% for 2 weeks during the BCoV infection.

Vaccines are available for the prevention of BCoV infections in neonatal calves and in adult. An inactivated virus vaccine was developed for pregnant cows to enhance passive maternal immunization of calves via colostrum and prevent neonatal diarrhoea (Decaro et al., 2009). Another inactivated hemagglutinating antigens-enriched vaccine showed their efficacy to protect cattle against WD (Takamura, Matsumoto, & Shimizu, 2002).

The BCoV was also associated, alone or with other pathogens, to respiratory infections such as enzootic pneumonia in calves and shipping fever in young cattle (Saif, 2010). The prevention from respiratory CoV in cattle is based on multivalent killed or attenuated-live intra-nasally administrated vaccines (Hay et al., 2016; Richeson et al., 2019).

3.1.2 Water buffaloes

A bovine-like CoV in water buffalo (Bubalus bubalis) was first described in 1985 in Bulgaria, using serological methods (Muniiappa, Mitov, & Kharalambiev, 1985). The sequencing of the isolated buffalo CoV from Italy, showed sequence homology with the BCoV, with some differences justifying its classification as a new variant of BCoV and was named as Bubaline CoV (BuCoV) (Decaro, Martella, et al., 2008). In Italy, gastroenteritis disease among water buffalo calves is caused by both BuCoV and BCoV (Decaro et al., 2010). Closely related strains to the BCoV (displaying 98% sequence homology) were also detected among faecal samples of domestic buffaloes in Bangladesh (Lau et al., 2016). Infection of water buffalo calves was also reported from three districts in Egypt, with an overall molecular prevalence of 26.8% (22/82) in faecal samples (El-Sadek et al., 2019) (Table 2).

3.1.3 Small ruminants

Although the BCoV is initially a cattle virus, it is occurring also in small ruminants, in which it was neglected for a long time due to its low prevalence and insignificant clinical manifestation (Amer et al., 2019). In Australia, sheep were reported in 1978 to excrete CoV-like particles that were visualized by electron microscopy examination during a diarrhoea episode (Tzipori, Smith, Makin, & McCaughan, 1978). The infection of small ruminants by the BCoV is rather to occur through close contact with cattle in mixed flocks than as a natural infection (Tråvén, Carlsson, Lundén, & Larsson, 1999). The most recent study in Ghana reported anti-BCoV antibodies in 26 out of 102 (25.5%) and 29 out of 66 (43.9%) sheep and goats, respectively, associated to lambs’ newborn diarrhoea and respiratory infections (Burimuah et al., 2020) (Table 2).

3.2.1 Dromedaries

Nowadays, it is well established that dromedaries are the natural host of the MERS-CoV and a source for human and other domestic animals infections (Kandeil et al., 2019). Since the first case of human infected by the MERS-CoV was identified in September 2012 in Saudi Arabia (World Health Organization, 2019), interest to dromedaries as sources of the virus increased and the isolated strains were shown to be genetically very similar to those isolated from humans (Omrani, Al-Tawfiq, & Memish, 2015).

The close contacts between infected dromedaries and humans enhance the continuing zoonotic transmission and may explain why the MERS-CoV continues to occur in humans (De Wit et al., 2016). Moreover, the MERS-CoV was also detected in sheep, goats and donkeys reared close to camels (Kandeil et al., 2019).

Transmission from dromedaries to humans occurs by direct contact with mucous and nasal discharge or by consumption of meat or raw milk (Gossner et al., 2016; Mirkena et al., 2018). Indeed, MERS-CoV RNA was detected using qRT-PCR in whole and skimmed milk collected from milking animals that were kept with camels that have frequent contact with multiple origin camels participating at racing events (Reusken, Farag, et al., 2014).

The MERS-CoV detection in dromedaries is difficult as the infection is asymptomatic. However, experimental MERS-CoV infection of dromedaries caused mild to moderate rhinitis with nasal discharge, tracheitis and bronchitis in addition to the shedding of a large amount of virus from the respiratory tract (Haverkamp et al., 2018).

Several studies showed that the wild strains of the MERS-CoV circulate in dromedaries in more than 24 countries across Africa and the Middle East (Figure 2), with high seroprevalences rates recorded from Egypt (71%) (Ali et al., 2017), Nigeria (94%) (Reusken, Messadi, et al., 2014) and Saudi Arabia 84% (Harrath & Abu Duhier, 2018). It was shown that the seroprevalence among dromedaries increases with age, reaching 80%–100% in adults (Harrath & Abu Duhier, 2018).

In experimentally challenged dromedaries, a modified orthopox-based vaccine expressing the MERS-CoV spike protein conferred mucosal immunity. This vaccine induced a significant reduction of virus shedding, and the conferred protection was associated with neutralizing antibodies production (Haagmans et al., 2016). As a strategy to prevent human contaminations, mass vaccination of camels is discussed (Dighe et al., 2019).

3.2.2 Alpacas and llamas

Following a severe diarrhoea outbreak that induced high morbidity and mortality among neonatal alpacas (Vicugna pacos) in Silly (Southern Peru), 40% (20/50) of samples collected from dead alpacas were positive to CoVs using RT-PCR (Rojas et al., 2016). In Oregon (USA), 39 out of 45 diarrheic crias (young llamas and alpacas) were positive to CoVs, based on morphological characterization of the virus after potassium phosphotungstate staining (Cebra et al., 2003). Coronavirus was also suspected to cause respiratory disease among alpacas in California, ranging from mild to fatal syndromes (Crossley et al., 2010). Reusken et al. (2016), found that 15 healthy alpacas in Al-Shahaniya (Qatar) were seropositive to MERS-CoV whereas both faecal and nasal swabs of the tested animals were negative by molecular assays. Phylogenetic analysis showed that the alpaca CoV genome displayed high identity (> 99.5%) with those of BCoV strains (Jin et al., 2007).

Today, interest to llamas increased as their antibodies could be engineered to block the attachment of the SARS-CoV-2 to the angiotensin-converting enzyme (ACE2) receptor present on the human cell membranes of type 2 pneumocytes and intestinal epithelial cells (Dong et al., 2020).

3.4.1 Transmissible gastroenteritis virus (TGEV)

The TGE is the sole mammalian disease caused by a CoV in the World Organisation of Animal Health (OIE) list (www.oie.int, Chapter 3.8.10.). TGE was reported since 1946 in the United States (Doyle & Hutchings, 1946). Then, it was reported in several regions including European countries, North, Central and South America, Southeast Asia: China, Japan, Korea, Nepal, Myanmar (Burma) and South and West Africa (Chen et al., 2019).

The virus could contaminate a non-infected herd through the introduction of asymptomatically infected animal, then TGEV is transmitted by faecal–oral route (MacLachlan & Dubovi, 2017). Large amounts of TGEV are present in the faeces of infected pigs, lasting up to 18 months, beside virus milk shedding (Piñeyro et al., 2018). Domestic animals such as cats and dogs may play a role as host for the TGEV (Sestak & Saif, 2002).

As for most of mammalian CoVs, the TGE occurs during winter after an incubation period varying between 18 hr and 3 days (Pensaert, 1976). The course of the disease is characterized by vomiting and profuse diarrhoea and is marked by high morbidity and mortality rates mainly among piglets (Animal Health Australia, 2016). Although pigs of all ages are susceptible to the TGEV infection, animals older than 5 weeks display milder clinical symptoms than piglets (Piñeyro et al., 2018).

As the TGEV shares common epitopes for neutralizing antibodies with PRCV, a significant decrease of the TGE incidence occurred as a consequence of the cross protection conferred by the PRCV (Wesley & Woods, 1996).

Bivalent or trivalent live-attenuated vaccines, combined with rotavirus, PEDV and/or Escherichia coli are available in Europe and North America. They are recommended for sows during gestation in order to provide lactogenic immunity to newborn piglets. Inactivated vaccines are mostly used in Asia (Gerdts & Zakhartchouk, 2017).

3.4.2 Porcine hemagglutinating encephalomyelitis virus (PHEV)

The emergence of the PHE was traced back to 1957 in Canada, with the occurrence of high several episodes of vomiting, wasting and anorexia followed by neurological signs in pig nurseries (Roe & Alexander, 1958). In 1962, the virus was isolated from baby pigs suffering from encephalomyelitis (Greig et al., 1962), then several outbreaks with similar clinical signs were reported in Canada and Europe and the disease was named vomiting and wasting disease (VWD) (Cartwright et al., 1969). In 1971, the PHEV was classified as a CoV (Greig et al., 1971) and subsequently the disease was reported from numerous countries, like China, South Korea, Japan, Belgium, Canada, the United States and Argentina (Mora-Díaz, Piñeyro, Houston, Zimmerman, & Giménez-Lirola, 2019). Recently, the PHEV was detected and characterized for the first time in Czech Republic in 7.9% of pigs’ nasal swabs from different age categories (Moutelikova & Prodelalova, 2019). The PHEV circulates in swine populations in silent way infecting animals after replacement or weaning and the transmission occurs via nose-to-nose contact, by the inhalation of infected nasal secretions (Saif et al., 2019).

The PHEV infects all naive pigs at any age, but the clinical manifestations are most severe in piglets under 4 weeks of age, mainly born from naive dams, with mortality rate reaching 100% (Mora-Díaz et al., 2019). Currently the PHEV infection is most likely to remain subclinical in affected herds, because the good protection conferred by the colostral immunity transferred to newborn piglets, which is considered as the best way to prevent from the infection (Rho et al., 2011).

3.4.3 Porcine epidemic diarrhoea virus (PEDV)

The porcine epidemic diarrhoea (PED) was reported for the first time in England in 1971, but the viral aetiology was only proved in 1978 (Pensaert & de Bouck, 1978). The PEDV spread to other European countries (Belgium, France, Hungary and Czech Republic…) (Pensaert & Martelli, 2016) and to Asia where it became enzootic particularly in most prosper pork industry countries like China, South Korea and Philippines (Song & Park, 2012). The economic losses induced by the occurrence of the PED are dramatic since, high mortality rate is associated to the infection, mainly among < 10 days old piglets (Antas & Woźniakowski, 2019). Mortality reached 100% in newborn and suckling piglets in Japan during 1993 and 1994 (Sueyoshi et al., 1995) and in Thailand during 2007 and 2008 (Puranaveja et al., 2009). In the United States, the PEDV emerged in 2013, where almost 10% of the pig population died in less than 1 year after several outbreaks (Chen et al., 2014; Stevenson et al., 2013) and spread to Mexico and Canada in 2014 (Kochhar, 2014). Based on the study of Zhou et al. (2018), the PEDV was ranked as the most important causative agent of porcine diarrhoea, with an infection rate of 78.25%. The transmission of PEDV is faecal–oral and both contaminated vehicles and food was incriminated in PEDV spreading in the United States (Bowman, Krogwold, Price, Davis, & Moeller, 2015; Lowe et al., 2014; Pillatzki et al., 2015).

Due to the severity of the PED and its huge economic impact, several Asian countries are immunizing pigs with killed or live virus vaccines. Killed virus vaccine induced the higher antibody (IgA and IgG) levels than live virus vaccine particularly in sow's colostrum, sow's sera and piglets sera after suckling (Paudel et al., 2014). In Japan, live attenuated virus is used since 1997 (Usami, Yamaguchi, & Kumanomido, 1998) and oral vaccination started in South Korea and Philippines, since 2004 and 2011 respectively (Garcia et al., 2018; Park et al., 2018). As new virulent strains are circulating, trials for more effective vaccine are undergoing using swine-poxvirus-based vaccines designed to express the A epitope of the spike protein (Yuan, Lin, Li, He, & Fan, 2017).

3.4.4 Porcine respiratory coronavirus (PRCV)

The porcine respiratory coronavirus (PRCV) is a mutant of the TGEV, due to a deletion in spike gene (Laude et al., 1993). It was identified first in Belgium in 1984 (Pensaert, Callebaut, & Vergote, 1986) and spread to almost all European countries including France (Madec et al., 2004), Denmark (Have, 1990) and the United Kingdom (Brown & Cartwright, 1986). The PRCV was soon reported from the United States (Halbur et al., 2003) and Japan (Usami et al., 1998).

Most of the PRCV infections are subclinical or inducing mild symptoms (Caswell & Williams, 2016), this could be due to the low proinflammatory cytokines synthesis activation (Van Reeth, Labarque, Nauwynck, & Pensaert, 1999). The PRCV is still detected upto the 21st day in the lungs of experimentally infected pigs and high titres reached 10^8.3 median tissue culture infectious doses (TCDI50) per gram of lung tissue (Jung et al., 2007). Faecal shedding of PRCV was also reported from 37% (21/57) of sentinel pigs introduced in PRCV-infected herds (Costantini et al., 2004).

Because of the insignificant clinical course of the PRCV infection, neither treatment nor vaccination, is applied. However, countries exporting live pigs need negative status to PRCV, which could be obtained only if piglets are pre-weaned early at the 7th day, together with strict all-in, all-out managed barns, rigorous disinfection and cleaning of pig barns and regular seronegative tests in sows (Burlatschenko & Arsenault, 2015).

3.4.5 Porcine deltacoronavirus (PDCoV)

The porcine deltacoronavirus (PDCoV) was first detected in asymptomatic pigs in 2009 in Hong Kong (China) during a molecular survey (Woo et al., 2012). However, another retrospective molecular study showed that the virus might circulate since 2004 in China (Dong et al., 2015). It was only in 2014 that the clinical signification associated to the PDCoV infection was addressed after multiple outbreaks occurred in Ohio State (USA) (Marthaler, Jiang, Collins, & Rossow, 2014; Wang, Byrum, & Zhang, 2014a). Then, the PDCoV spread to other American States (Iowa, Illinois, Minnesota and Nebraska), Canada (Wang, Byrum, & Zhang, 2014b), South Korea (Lee & Lee, 2014), mainland China (Dong et al., 2015), Thailand, Vietnam and Lao Popular Democratic Republic (Saeng-Chuto et al., 2017).

According to Sun et al. (2020) the PDCoV infection became enzootic in certain region of China. Molecular prevalences ranged from 10.06% (17/169) to 16.9% (29/172), in Hong Kong and Taiwan respectively (Woo et al., 2012; Hsu et al., 2018). The co-infection PDCoV-PEDV is frequently reported in pigs’ farms where gastroenteric problems occur (Jang et al., 2017; Lee et al., 2016; Zhang, Liu, et al., 2019). According to Zhang, Liu, et al. (2019) suckling pigs were most infected than finishing pigs and the PDCoV was recovered from milk samples at the rate of 7.92% (8/101).

Epidemiological patterns of the PDCoV infection are similar to those of PEDV, but the reported prevalences remain lower in diarrheic pigs compared with those reported for PEDV and TGEV infections (Zhang, 2016). Despite mortality rates are lower than that usually observed for PEDV infection, the emergence of the PDCoV disease in Thailand in 2014, caused heavy economic loses to pork farms with a mortality rate reaching 64.27% (2,892/4,500) in piglets (Janetanakit et al., 2016).

Mixing different animal species in swine farms may play an important role in the epidemiology of the PDCoV infection, as chicks and poultry are susceptible to the PDCoV (Liang et al., 2019), thus it is not excluded that they may act as intermediate hosts for the virus. Moreover, in orally PDCoV-inoculated calves, a persisting faecal viral RNA shedding associated to PDCoV-specific serum IgG antibody responses were observed (Jung et al., 2017). There is no commercial available vaccine for the prevention from the PDCoV infection (Zhang, Liu, et al., 2019). Active surveillance programme effectively led to the decrease of the PDCoV herd-level prevalence below 0.5% in two years in Canada (Ajayi et al., 2018).

3.4.6 Swine acute diarrhoea syndrome virus

The swine acute diarrhoea syndrome virus (SADS-CoV) has emerged in pigs in China since August 2016 (Zhou, Sun, et al., 2019). This virus, also qualified as Porcine entreric alphacoronavirus (PEAV) (Xu et al., 2019) or Swine enteric coronavirus (SeCoV) (Pan et al., 2017) was associated with the occurrence of severe diarrhoea of suckling piglets in Guangdong province in 2017 (Gong et al., 2017; Pan et al., 2017) and 2019 (Zhou, Li, et al., 2019), and in Fujian province in 2018 (Li et al., 2018). The economic loss induced by the SADS-CoV was dramatically high, as 24,693 piglets in four farms, died from the infection in Guangdong province, where the first human SARS epidemic started in 2003 (Zhou et al., 2018). This virus is antigenically distinct from the PEDV, the TGEV and the PDCoV (Yang, Yu, & Huang, 2020). Nevertheless, the complete genome sequencing of the N and the S genes showed a high nucleotides homology (96%–98%) with those of four bat-CoV HKU2 (Gong et al., 2017). During the first outbreak in China, the SADS-CoV infection was controlled by immunizing pregnant sows using inactivated filtrated virus obtained from infected piglets intestines (Zhou, Sun, et al., 2019b). As there is an urgent need to develop efficient vaccines to control the SADS-CoV in pigs, trials on attenuation of a virulent strain via cell culture passage are now on-going (Sun et al., 2019).

3.4.7 Recombinant emerging pigs’ coronavirus

decaA new chimeric virus containing the S protein of the PEDV on a TGEV backbone was discovered in Italy in 2016 and might circulate there from mid-2009 to 2012 according to Boniotti et al. (2016). This recombinant virus was also detected in Germany, Central Eastern Europe and Slovakia precluding for an old circulation in Europe (Akimkin et al., 2016; Belsham et al., 2016; Mandelik et al., 2018).

4.4.1 Agriculture intensification and environmental changes

Intensification of livestock production implies increasing animal density in farms and increasing movement of human and vehicles on and off farms, which promotes pathogen transmission and spreading (Cutler, Fooks, & van der Poel, 2010). The intensively development of swine industry the last decade is considered of high risk for the trigger of severe pandemic viruses for humans (Borkenhagen, Salman, Ma, & Gray, 2019). This was shown particularly with swine influenza viruses, and Nipah virus. Indeed, swine exposed workers displayed significantly higher titers (in hemagglutination inhibition assay) against swine influenza subtypes H1N1 and H1N2 than non-exposed workers (Odds Ratio = 54.9 and 95% confidence interval: [13.0–232.6] (Gray et al., 2007). In late 1998 and 1999, several fatal encephalitis cases (n = 105) caused by a Paramyxovirus were recorded in Malaysia (Chua, 2000) and quickly spread to Singapore (Paton et al., 1999). Respiratory disease and encephalitis were recorded few weeks ago in pigs in the same district and the isolated virus, named Nipah virus, was shown to be transmitted from bats to pigs, then to humans (Chua et al., 2002).

The consequences of anthropogenic activities leading to deforestation, habitat fragmentation and replacement of natural vegetation by crops definitely modify wildlife biology. These anthropogenic activities foster wildlife migration and create new environments that increase host, vector and pathogen contacts (Jones et al., 2013). Indeed, the deforestation was incriminated in the occurrence of the Nipah virus outbreak by pushing fruitbats (the natural reservoir of the Nipah virus) to pig farms high-density areas and fostered the contact between both species (Chua et al., 2002).

4.4.2 Trade of domestic and wild animals

It is well established that animals and animal product trade foster pathogens spreading (Travis, Watson, & Tauer, 2011). A study on livestock trade network assessment in Europe, showed that pigs (for fattening and slaughter) and cattle (for fattening) were ranked second as the most heavily moved animal species within Europe in 2011, which foster the risk for exotic diseases introduction (Hardstaff, Häsler, & Rushton, 2015).

Thousands of wild animals are traded in Asian markets every day, such as raccoon dogs, pangolin, masked palm civets, ferret badgers and hedgehogs (www.animalasia.org). As several domestic and wild animals were shown to harbour CoVs, handling, touching or eating these animals, could increase the risk for humans to be in contact with CoVs.

4.4.3 Changing in the distribution of invasive animal species

Rodents have a worldwide distribution and were shown to harbour more than 60 known pathogens infecting both animals and humans (Meerburg, Singleton, & Kijlstra, 2009), including CoVs (Wang et al., 2015). Bats also are considered to be the most abundant and diverse vertebrate after rodents (Rodhain, 2015), with more than 1,300 known species and worldwide geographic distribution (Teeling, Jones, & Rossiter, 2016). Bats are considered as an important reservoir of highly lethal zoonotic viruses and were shown to harbour more CoVs than any other animal species (Hu et al., 2015). Indeed, beside the SARS-CoV, the SARS-CoV-2 and the MERS-CoV, two porcine CoVs, have a bat-CoV origin: the SADS-CoV and the PEDV (Banerjee, Kulcsar, Misra, Frieman, & Mossman, 2019). The geographical range of some bat species was estimated to expand by near to 400% in the last four decades as a response to climate change (Ancillotto, Santini, Ranc, Maiorano, & Russo, 2016). One could imagine, that this expansion would be favourable to a higher contact likelihood with wild and domestic animals.

4.4.4 Herd management practices

Mixing different animal species would increase CoVs host jumping probability. The BCoV contaminates sheep and goats, when they share the same barns or graze with infected cattle (Tråvén et al., 1999). The MERS-CoV was also detected from sheep, goats and donkeys kept in close contact with infected dromedaries (Reusken et al., 2013).