To investigate the incidence, viral load and genetic diversity of bovine rotaviruses strains in Tunisia.
To investigate the incidence, viral load and genetic diversity of bovine rotaviruses strains in Tunisia.
A total of 169 faecal specimens, collected from diarrhoeic calves from several farms located in the central eastern regions of Tunisia, between January 2006 and October 2010, were analysed by semi-nested multiplex RT-PCRs for P and G genotypes identification or were genotyped by DNA sequencing. Positive samples were tested by TaqMan real-time RT-PCR to quantify the viral load. Group A bovine rotaviruses were detected in 15·4% (26/169) of the total studied cases of diarrhoea. Overall, G10 was the predominant G type, detected in 12/26 samples (46·2%) and G6 accounted for 42·3% (11/26) while P was the predominant P type, detected in 12/26 samples (46·2%). Two P genotypes (7·7%) were found in the collection. Dual G or P combination and genotype G8 were not found. The most common VP7/VP4 combinations were G6P (30·8%; n = 8) and G10P (11·5%; n = 3). The combination G10P was seen in one sample, and partial typing was assessed in 53·8% (n = 14) of the cases. The viral load determined by real-time RT-PCR showed an average of 1·68 × 109 genome copies/g of faeces.
Knowledge of P and G types could help us understand the relatedness of animal rotaviruses to viruses causing disease in humans.
This is the first time that the viral load and P types of bovine rotaviruses have been determined in Tunisia, and this study contributes to a better understanding of the epidemiology of such viruses circulating in Tunisia. Nevertheless, continuous surveillance is necessary to detect the emergence of new variants.
Group A rotaviruses are a major cause of diarrhoea in young children and animals including cattle (Kapikian and Chanock 1990; Saif et al. 1994). Rotaviruses constitute a genus within the Reoviridae family, characterized by nonenveloped triple-layered viral particles with a viral genome composed of 11 double-stranded RNA segments (dsRNA). The inner capsid is composed of a single protein, VP6, encoded by gene segment 6 and bearing group and subgroup antigenic specificities. The outer capsid is studded with VP7 and VP4 proteins, which elicit neutralizing antibody responses and form the basis of the present dual classification system of G (VP7) and P (VP4) types (Kapikian et al. 2001). So far, 27 G genotypes and 35 P genotypes have been characterized in humans and animals (Matthijnssens et al. 2011). While at least six P genotypes (P, P, P, P, P and P) and eight G genotypes (G1, G3, G5–G8, G10 and G15) have already been described in rotaviruses affecting cattle, only G6, G10 and G8 combined with P, P and P are considered epidemiologically important (Alfieri et al. 2004; Barreiros et al. 2004; Dhama et al. 2009; Martella et al. 2010).
Several genotypes, such as G3, G6 and G8, are shared by both humans and animals (Desselberger et al. 2001), but direct transmissions between different animal species and between humans and animal species have not actually been observed. However, the increasing number of reports of new human rotavirus genotypes that are more commonly found in animals suggests the possibility of interspecies transmission or genetic reassortment of rotavirus strains (Nakagomi et al. 1994; Desselberger et al. 2001).
While bovine rotavirus infections have been reported in Tunisia on previous occasions (Libersou et al. 2004; Fodha et al. 2005), current prevalence, viral load and the circulating G and P genotypes are still unknown. For this purpose, we examined diarrhoeic calves in the central eastern regions of Tunisia for rotavirus infection during the period extending from January 2006 to October 2010. Furthermore, this is the first survey reporting the quantification and the P types of bovine rotaviruses circulating in Tunisian cattle. Regarding the G types, our results add information to previous genotyping.
Between January 2006 and October 2010, 169 faecal samples from mixed dairy-beef calves with diarrhoea were collected from 17 cattle herds (designated as herds A-Q) as part of an ongoing surveillance study of potential zoonotic micro-organisms associated with gastroenteritis in humans. These cattle herds were located in the central eastern regions of Tunisia: Mahdia, Monastir, Ouerdanine, Moknine and Kairouan. In these regions, 80% of calves belong to the pure Holstein breed and are essentially oriented towards milk production. Farm sizes ranged from 5 to 25 animals. All calves with diarrhoea were sampled. Sixty-six originated from Mahdia (16, 9, 14, 12, 5 and 10 from herds A, B, C, D, E and F, respectively), 34 from Monastir (12, 7 and 15 from herds G, H and I, respectively), 13 from Ouerdanine (13 from herd J), 49 from Moknine (10, 5, 3,15, 9 and 7 from herds K, L, M, N, O and P, respectively) and 7 from Kairouan (7 from herd Q). The ages of the calves under experiment ranged from 3 to 90 days (mean age, 55 days; median age, 30 days). All of the specimens were transported to the laboratory on ice and stored at −20°C until analysis. The same faecal samples were previously tested for the detection of bovine caliciviruses (Hassine-Zaafrane et al. 2012).
The viral RNA was extracted as previously described (Hassine-Zaafrane et al. 2011) and then was analysed using one-step RT-PCR kit (QIAGEN, Hilden, Germany) and primers amplifying partial VP6 gene (Iturriza-Gomara et al. 2002). The PCR conditions involved an initial reverse transcription step of 30 min at 50°C, followed by PCR activation at 95°C for 15 min, 35 cycles of amplification (1 min at 94°C, 1 min at 55°C and 1 min at 72°C), with a final extension of 10 min at 72°C. All rotavirus-positive samples were quantified by VP2 TaqMan RT-PCR assay using primers and probe as previously described (Gutiérrez-Aguirre et al. 2008). The number of genome copies present in each positive sample that could be evaluated was estimated by comparing the sample Ct value to standard curves. The detection limit of this real-time RT-PCR assay was 100 copies of viral RNA, indicating a good sensitivity of the assay. To obtain the standard curves, a 531-bp fragment was amplified by RT-PCR from the VP2 gene then cloned into the pGEM-T Easy vector (Promega Corporation, Madison, WI). After transformation in E. coli, and production of the clones, the plasmid DNA was purified and quantified; then, serial dilutions were prepared and used as standard curves. The final concentration in the samples was adjusted based on the volume of nucleic acids analysed and was expressed per gram of faeces.
Bovine rotavirus G and P genotyping was performed using semi-nested type-specific multiplex RT-PCRs that could detect five G types and five P types. For G typing, the first-round PCR used VP7-F and VP7-R primers to amplify an 881-bp region of the VP7 gene (Iturriza-Gomara et al. 2001). The nested multiplex PCR was performed using a pool of internal primers (Gouvea et al. 1994b) specific for G5, G6, G8, G10 and G11 bovine rotavirus genotypes in combination with the appropriate forward consensus primer (VP7-F).
For P typing, primers VP4-F and VP4-R were used in the first-round PCR to amplify a 663-bp fragment of the VP4 gene (Simmonds et al. 2008). The second-round PCR amplification was carried out with primer VP4-R (antisense) and a pool of primers specific to P genotypes P, P, P, P and P (sense) (Gentsch et al. 1993; Gouvea et al. 1994a). All PCR products were examined by gel electrophoresis in 2% agarose gels containing 0·4 μg ml−1 ethidium bromide and then visualized under UV light. P and G genotypes were determined by the size of the amplicons.
All VP6 RT-PCR-positive faecal specimens that were negative in semi-nested multiplex RT-PCR were typed by sequencing part of the VP7 (20 specimen typed) and/or VP4 (4 specimen typed) gene. The samples were amplified by one-step RT-PCR for VP7 and VP4 genes as previously described (Iturriza-Gomara et al. 2001; Simmonds et al. 2008).
The amplicons of the VP7 and VP4 genes were purified using Amicon® Ultra 0.5 30K Centrifugal Filters (Millipore™ Corporation, Billerica, MA) according to the manufacturer's protocol. The purified PCR products were used as a template for sequencing using an ABI PRISM® Big Dye® Terminator Cycle Sequencing Ready Reaction Kit on an automated sequencer (model 3130XL DNA Genetic Analyzer), (Applera Corporation, Foster City, CA) and were sequenced from both directions.
Multiple alignments were carried out using Clustal W (Thompson et al. 1994). Phylogenetic trees were designed by imputing the aligned sequences into the MEGA program (version 4.1) (Tamura et al. 2007) and constructed with the neighbour-joining algorithm (Saitou and Nei 1987). Genetic distances were calculated with the Kimura-2 parameter model (Kimura 1980) with a transition/transverse ratio of 2·0, and the reliability of the trees was determined by bootstrap analysis with 100 pseudo-replicates data sets.
The sequences obtained in this study have been submitted to GenBank under the following Accession Numbers:
VP7: B21/28-05-06/TUN [GenBank: KF724035]; B31/27-07-06/TUN [GenBank: KF724036]; B38/04-09-06/TUN [GenBank: KF724037]; B52/23-10-06/TUN [GenBank: KF724038]; B55/03-11-06/TUN [GenBank: KF724039]; B70/16-12-06/TUN [GenBank: KF724040]; B72/23-12-06/TUN [GenBank: KF724041]; B85/02-02-07/TUN [GenBank: KF724042]; B89/03-03-07/TUN [GenBank: KF724043]; B91/05-03-07/TUN [GenBank: KF724044]; B95/05-04-07/TUN [GenBank: KF724045]; B109/27-06-07/TUN [GenBank: KF724046]; B132/21-02-09/TUN [GenBank: KF724047]; B135/23-02-09/TUN [GenBank: KF724048]; B149/24-03-10/TUN [GenBank: KF724049]; B156/08-04-10/TUN [GenBank: KF724050]; B157/16-04-10/TUN [GenBank: KF724051]; B158/16-04-10/TUN [GenBank: KF724052]; B159/02-10-10/TUN [GenBank: KF724053]; and B165/15-10-10/TUN [GenBank: KF724054].
Statistical analyses were performed with SPSS® software, version 19 as previously described (Hassine-Zaafrane et al. 2011). P values ≤ 0·05 were considered significant.
Twenty-six of 169 (15·4%) diarrhoeic animals were tested positive to group A rotaviruses by one-step VP6 RT-PCR assay. Five (3%) noroviruses of genogroup III were detected as mixed infections with rotavirus.
In this study, the mean age of calves found positive to rotaviruses was 32·4 ± 24·3 days and the median was 30 days.
Regarding the seasonal distribution of bovine rotaviruses, despite the small number of positive samples, a significant relationship was found between rotavirus infection and seasonal distribution (P < 0·05). Indeed, the prevalence of bovine rotaviruses was 25% (12/48) in spring, 7·3% (3/41) in summer, 12·5% (5/40) in autumn and 15% (6/40) in winter.
G and P types of bovine rotavirus strains detected in the 26 positive cases are summarized in Table 1. G and P genotypes were successfully determined for 23 (88·5%) and 15 (57·7%) samples, respectively. The VP7 gene of 3 samples and the VP4 gene of 11 samples could not be amplified.
|Strains||Year||Season||G typing result by||P typing result by|
|Multiplex PCR||Sequencing||Multiplex PCR||Sequencing|
In the current study, G10 (46·2%) was the most predominant strain followed by G6 (42·3%). Among these rotavirus strains, 12 were characterized as P (46·2%), 2 as P (7·7%) and 1 as P (3·8%).
Both G and P types could be assigned to 12 (46·2%) of 26 rotavirus-positive samples. Overall, G6 in combination with P was the most prevalent strain (30·8%) followed by G10P (11·5%), both considered as the most common G/P associations found in bovines. Altogether, these data may suggest that there is a predominance in Tunisia of bovine strains with P VP4 and with either G6 or G10 VP7. One partially typed strain G10P[?] detected in one sample was further characterized by sequence analysis as G10P. Dual G or P types were not found.
When analysing the distribution of G/P type combination through time, it was observed that G6P was the prevalent strain in 2006, 2007 and 2009, but G10P predominated in 2010. The one strain G10P was detected in February 2009.
For strains genotyped by DNA sequencing, phylogenetic analyses were performed by comparing the nucleotide sequences obtained with strain sequences available in the GenBank database.
Phylogenetic trees of nucleotide sequences of bovine rotavirus isolates with representative VP7 (a) and VP4 (b) genotypes were constructed (Fig. 1).
The three strains B70/16-12-06/TUN, B85/02-02-07/TUN and B158/16-04-10/TUN shared nucleotide identity ranging from 99·3% to 99·8%, and they showed high nucleotide identity with the reference strain BO/B223 (Accession Number D13394) ranging from 95·9% to 96·4%.
The strain B137/22-02-09/TUN showed high identity with bovine strain BO/86 (Accession Number GU984756) and human strain Hu/PR/1300/04 (Accession Number EU835944) with 88·4% and 88·2% nucleotide identity, respectively.
Therefore, B70/16-12-06/TUN, B85/02-02-07/TUN and B158/16-04-10/TUN isolates were classified as isolates with P genotypes while B137/22-02-09/TUN as isolate with P genotype.
The G10 strains displayed nucleotide identities between them ranging from 98·9% to 100% and clustered with Bo/61A (Accession Number X53403).
Phylogenetic analysis revealed that 8 G6 strains were homologous to each other (86·5–100% nucleotide identity). All these showed high nucleotide identities with the Buff/10733 (Accession Number AY281360), Hu/Hun4 (Accession Number AJ487833), ROBVP7G (Accession number M63266) and BO/CIT39A/02 (Accession Number AY629556) ranging from 93·8% to 100%.
The two strains B159/02-10-10/TUN and B109/27-06-07/TUN formed a branch separate from all the 8 established G6 types.
In this study, the quantification of rotaviruses in faecal specimens by TaqMan RT-PCR demonstrated a mean viral load of 1·68 × 109 genome copies/g of faeces, and a G/P combination can be determined only in samples with viral load higher than 1·90 × 105 genome copies/g of faeces (Table 2).
|G/P combinations||Number of G/P combinations||Mean of viral load (genome copies/g of faeces)||Mean of Ct|
|G6P||8||1·09 × 107||17·9|
|G10P||3||1·14 × 1010||13·1|
|G10P||1||3·61 × 108||15·4|
|NDP||2||1·21 × 104||27·1|
|G10ND||8||1·90 × 105||25·6|
|G6ND||3||3·81 × 103||29·4|
|NDP||1||2·48 × 102||32·8|
Bovine group A rotaviruses play an important role in causing gastroenteritis in young calves, and the best hope for prevention is the development of an effective vaccine. A rotavirus surveillance is an essential step in designing vaccines and vaccine strategies to identify regional strain patterns and potential emerging strains.
A previous study of the aetiological agents of calf diarrhoea in Tunisia has been published, which confirmed that rotavirus was the major cause of diarrhoea in the country (Zrelli et al. 1988). Out of this finding, we conducted this study to achieve a better understanding of the epidemiology of such viruses circulating in Tunisia.
The molecular prevalence of rotaviruses in this study is consistent with studies conducted in Western India (14·3%) (Chitambar et al. 2011) and France (15%) (Midgley et al. 2012), but is lower than the prevalence of rotavirus recorded in Denmark (46%) (Midgley et al. 2012) and in Argentina (62·5%) (Garaicoechea et al. 2006). A previous Tunisian study conducted between December 2001 and April 2002 reported that bovine rotaviruses were detected in 30% of dairy calves with diarrhoea (Fodha et al. 2005). Another study conducted in six European countries demonstrated that lower rates were reported when asymptomatic animals were tested (2–16%) and higher rates when diarrhoeic animals were assessed (12–98%) (Dhama et al. 2009).
The mean age of calves found positive to rotaviruses in this study was higher than that found by Reynolds et al. (1986) in Southern Britain (9·8 days) and by Garcia et al. (2000) in Spain (12·9 days).
Regarding the seasonal distribution of bovine rotaviruses, despite the small number of positive samples, a clear detection peak was observed in spring. This distribution was different from the seasonality of human rotaviruses, which peaked in winter (Hassine-Zaafrane et al. 2011). A study conducted in Japan on healthy calves showed that the highest detection rate of rotavirus genes was in January followed by December (Abe et al. 2009). Little is known about the seasonal distribution of bovine rotavirus infection because some countries apply calving programmes and because of the fact that these viruses infect mostly younger calves, most faecal sampling is performed at the same time.
In this study, dual G or P types were not found. However, a study conducted in India between 2007 and 2010 indicated that G3P alone or in combination with G10 or G8 was predominant among bovine populations (Malik et al. 2012). Besides, co-infections by G8 and G6 were registered in Argentina from 2004 to 2010 (Badaracco et al. 2012).
The genotype P was found in combination with G10 genotype in one sample. In India, the genotype P was detected in association with G6 and G10 (Ghosh et al. 2007), while it was found in association with G8 in Japan (Fukai et al. 1999) and Western India (Chitambar et al. 2011).
As in the current study, the most common G types in cattle were G10 and G6 as recorded in Brazil (Alfieri et al. 2004), the Netherlands (van der Heide et al. 2005), Ireland (Cashman et al. 2010), Italy (Monini et al. 2008), Turkey (Alkan et al. 2010), earlier in Sweden (De-Verdier Klingenberg et al. 1999), Japan (Fukai et al. 2004) and India (Rao et al. 2000).
The first Tunisian study has reported the isolation of a single genotype of bovine rotaviruses: genotype G6 (Libersou et al. 2004). However, results obtained by Fodha et al. (2005) demonstrated that genotype G8 was the dominant strain followed by genotype G6 or that faecal specimens contained a mixture of both.
We know that genotype G8 is considered as the third G type of epidemiological importance in cattle. It was previously detected in bovine in Japan (Fukai et al. 2002), Sweden (De-Verdier Klingenberg et al. 1999) and Italy (Falcone et al. 1999), but in this study, genotype G8 was not found.
The frequency of G10P (11·5%) was relatively low, with respect to the values obtained from studies in India (Gulati et al. 1999) (81% between 1994 and 1997) and Italy (Falcone et al. 1999) (31·5% between 1994 and 1998), but close to values obtained in Brazil (Alfieri et al. 2004) (16% between 1996 and 1999).
Rotavirus strains bearing G10P are common pathogens of cattle in various regions (Fukai et al. 2002, 2004; Garaicoechea et al. 2006; Monini et al. 2008). It has also been reported that G10P strains are associated with symptomatic and asymptomatic infections in children in India (Iturriza-Gomara et al. 2004). Indeed, interspecies transmissions of group A rotaviruses have been suggested, especially between humans and cattle (Das et al. 1993). Several publications reported G6 (Gerna et al. 1992; Steyer et al. 2012) and G10 (Urasawa et al. 1992) genotype strains in humans. Also, uncommon genotypes such as P, P and P are increasingly detected in humans in different areas of the world (Gerna et al. 1992; Gentsch et al. 1993; Santos and Hoshino 2005; El Sherif et al. 2011).
Concerning untypeable genotypes, in most of these cases, there was either no amplified product after the nested multiplex RT-PCR reaction, or the VP7 and/or VP4 sequences of these strains could not be determined. The large number of untypeable genotypes can be explained by the fact that, in nature, any combination of G and P types may occur and the untypeable genotypes may represent other existing or new G or P types. These G or P types may escape classification if there are no suitable diagnostic reagents available. These samples may also be classified as untypeable because they were tested only with primers representing the G and P types traditionally associated with the bovine population. Another explanation that should be considered is that the rotavirus samples from India and Africa are more diverse and thus less likely to be amplified with a given set of primers (Simmonds et al. 2008). In these cases, DNA sequencing of a part of VP7 and VP4 genes was shown to be useful as a quick determination of uncommon or novel strains whose genotyping cannot be performed by genotyping PCR. However, the lack of amplification and the absence of sequences could be due to inhibitors in faecal samples, conservation problems or mismatches with the sequence of the primer.
Out of this study, it can be concluded that the usual bovine P and G genotype rotaviruses circulate in Tunisia. Therefore, it is suggested that the study of rotavirus G and P genotyping of human and animal rotaviruses of different species should be carried on and that the methods used for rotavirus typing need to be monitored and updated regularly.
This work was supported by the AUF Project (code 2092 RR823) and the National Reference Center (NRC) for Enteric Viruses, CHU Dijon, France. We thank Nedra Kerkeni for her editorial assistance.
None of the authors have a commercial or other association that might pose a conflict of interest.