Molecular prevalence of Toxoplasma gondii DNA in goats’ milk and seroprevalence in Northwest Tunisia

Abstract Toxoplasmosis is a worldwide zoonosis with high impact on human and animal health. Consumption of unpasteurized milk is a risk factor of human toxoplasmosis. The aim of this study was to estimate the seroprevalence and molecular prevalence of T. gondii in goats’ milk in Northwest of Tunisia (Jendouba Governorate). A total number of 77 blood samples were collected from six herds were screened with a commercial ELISA kit for T. gondii antibodies. For the same goats’ samples, a nested PCR was performed to detect T. gondii DNA in milk. The seroprevalence of T. gondii infection was 31.2% (±0.05) while the molecular prevalence of this parasite in milk was estimated to 7.8% (±0.03). A very low value of kappa showed that there is not agreement between seroprevalence and parasite prevalence in milk. These results suggest that the consumption of raw milk from naturally infected goats is a potential source of human infection. An extension programme should be implemented to decrease related to goats’ raw milk consumption.


Introduction
Toxoplasmosis is a zoonotic infection caused by Toxoplasma gondii, a protozoan parasite belonging to the order Coccidia and the Apicomplexa phylum. T. gondii infects warm-blooded animals including humans with felines (mainly cats) as definitive hosts (Dubey 2010). Toxoplasmosis is a serious risk for seronegative pregnant women and immunocompromised persons (Montoya & Liesenfeld 2004;Tenter 2009). Infection with T. gondii is frequently asymptomatic, but remains serious even fatal for specific groups including congenitally infected fetuses and newborns, immunocompromised individuals (AIDS patients), and transplanted persons (Saadatnia & Golkar 2012). It is estimated that T. gondii infects up to one-third of the human population in the world (Dubey & Jones 2008). In Tunisia, human seropreva-lence is high; it has been estimated to 70% in female Tunisian aged of 30 years in the northwest of the country (Bouratbine et al. 2001).
Among the food-borne diseases, toxoplasmosis is a real burden in several countries (EFSA, 2007;Havelaar et al. 2010;Scallan et al. 2011). For example, in USA, toxoplasmosis is the second cause of mortality after salmonellosis, and the fourth cause of hospitalization after salmonellosis, campylobacteriosis and norovirus infections (Scallan et al. 2011). Many foods are a reservoir of T. gondii. The main concerned foods are raw and undercooked meat infected with T. gondii, unpasteurized milk from infected animals, infected vegetables and water contaminated with oocysts (Dubey 1994(Dubey , 1996Baril et al. 1995). Among food animals, sheep and goats are the main sources of human infection (Dubey 2010). Infection with T. gondii in goats is ranging between 3.7 and 81.8% (Dubey & Adams 1990;Costa et al. 2012). In North Africa, studies about T. gondii infection in goats are scarce. In Egypt, positive ELISA and PCR results were found in 41.7 and 25% of goats, respectively (Ghoneim et al. 2010). In Tunisia, a single study was performed by Ben (1986), the seroprevalence in goats was estimated to 16%. Recently, a serological survey carried out in Morocco, reported that 8.5% (9/106) of goats were positive to toxoplasmosis (Benkirane et al. 2015). Several studies showed that drinking unpasteurized goats' milk could cause clinical and even fatal toxoplasmosis infections in humans (Riemann et al. 1975;Patton et al. 1990;Skinner et al. 1990). When excreted in milk, T. gondii tachyzoites are infective (Skinner et al. 1990). Recently, few studies indicated the presence of T. gondii in goat's milk and reported low prevalence ranging between 6 and 9.4% (Bezerra et al. 2013;Dehkordi et al. 2013).
Among the reliable methods, polymerase chain reaction (PCR) has the highest accuracy, sensitivity and specificity compared with conventional diagnostic methods (Held et al. 2000;Kompalic-Cristo et al. 2007). This study aimed to estimate the seroprevalence of T. gondii infection and the molecular prevalence of T. gondii DNA in milk samples to provide a preliminary assess of contamination risk by T. gondii in Tunisian goats' milk.

Study area
This study was carried out in goat farms located in Northwest Tunisia (localities of Tabarka and Hammam Bourguiba, Jendouba Governorate (Fig. 1)). These localities are situated between 36°45 0 and 36°57 0 North latitudes and between 8°35 0 and 8°45 0 East longitudes with an altitude ranging between 20 and 150 m. The governorate of Jendouba has the most temperate climate in Tunisia with a mean annual rainfall of 1000 mm. The mean minimal and maximal temperature are 5 and 30°C in winter and summer, respectively (National Institute of Meteorology, Tunisia).

Samples collection
This study was carried out from February to May 2014 in six herds totalling 77 lactating goats of different ages and breeds (Table 1). Blood samples were collected in dry tubes from the jugular vein; the sera were separated and stored at À20°C until used. Simultaneously, 50 mL of milk was collected from each goat and kept frozen in sterile tubes for molecular study.

Serology
Samples of 10 lL of sera were examined for IgG antibodies against T. gondii using an ELISA kit (ID Screen â Toxoplasmosis Indirect, Montpellier, France) in accordance to the manufacturer's instructions.

Molecular study
DNA was extracted from milk as described by Mancianti et al. (2013). Briefly, the milk samples were centrifuged at 2200 g for 5 min. To avoid interference with casein, 1 mL of the pellet was treated with 200 lL of TE (1 mM EDTA, 10 mM Tris-HCl (pH = 7.6)) and 300 lL 0.5 M EDTA (pH 8). The solution was resuspended then centrifuged at 3000 g for 10 min and finally diluted in 200 lL of PBS (Psifidi et al. 2010). DNA was isolated from the resuspended pellet using Wizard Genomics DNA (Promega, Madisson, Wisconsin, USA) extraction kit according to manufacturer's instructions; the DNA was stored at À20°C until used.
To evaluate the efficiency of DNA extraction protocol, a PCR assay targeting the hypervariable regions V1-V3 coding for 18S rRNA was performed using two forward primers AACCTGGTTGATC CTGCCAGT (1 A) and reverse primers GGCAC-CAGACTTGCCCTC (564 R). The PCR reaction was carried out in a mixture of 19 PCR buffer, 2 mM MgCl 2 , 10 lM of each primer, 0.2 mM of each dNTP, 2 U Taq polymerase, 1.5 lL of DNA template and distilled water to a total volume of 25 lL (Wang et al., 2014). The PCR conditions were 5 min at 94°C followed by 25 cycles of denaturation at 94°C for 50 s, annealing at 58°C for 50 s and a final extension at 72°C for 10 min. A nested PCR was performed to amplify a T. gondii DNA fragment of 227 bp belonging to ITS1 gene and coding for the 18S -5.8S rRNA according to a modified protocol of Hurtado et al. (2001). Two external primers were used, namely, NN1 (5 0 -CCTTTGAATCCCAAGCA AAACATGAG-3 0 ) and NN2 (5 0 -GCGAGCCAA-GACATCCATTGCTGA-3 0 ). After the first PCR, nested PCR was performed using 2 lL of the first PCR product as template and T. gondii-specific primers Tg-NP1 (5 0 -GTGATAGTATCGAAAGG-TAT-3 0 ) and Tg-NP2 (5 0 -ACTCTCTCTCAAATG TTCCT-3 0 ). Each amplification was performed in 25 lL of total reaction volume consisting of 0.1 lM of each primer, Taq polymerase buffer (1x) supplemented with MgCl 2 (2 mM), 0.2 mM dNTPs, 0.5 U Taq DNA polymerase (R GoTaq, Promega). Thermal amplification programme was 3 min at 94°C

Statistical analysis
A chi-square Mantel-Haenszel test was performed for comparison of infection prevalence with Epi Info 6 at a threshold of 5% (Schwartz 1993). The concordance between PCR and ELISA was estimated with the Kappa test (Toma et al. 2007).
The kappa coefficient between n-PCR and ELISA was very low (Ƙ = 0.01).

Discussion and conclusion
In this study, we used an ELISA and a nPCR in milk for detection of T. gondii from lactating goats in North-western Tunisia. Anti-T. gondii antibodies were found in 31.2% of serum samples, only two of them were also positive for milk-PCR. Our results were comparable with those reported in China (29.54% by ELISA), Thailand and Pakistan (27.9 and 25.5% by Latex Agglutination Test, respectively) (Jittapalapong et al. 2005;Ramzan et al. 2009;Liu et al. 2015). In Europe, T. gondii is widespread with high infection rates. In Italy, 60.6% of goats' serum samples were positive by micro-agglutination test (Mancianti et al. 2013). Similar findings were reported in Bulgaria, where 59.8% of the samples were positive by inhibition of haemagglutination test (Prelozov et al. 2008). In Romania, the seroprevalences were 33.1 and 52.8% in kids and dairy goats, respectively (Lovu et al. 2012;Pas ßtiu et al. 2015). In USA, Korea and Nigeria lower rates were reported (6.8; 5.1 and 4.6%, respectively) (Kamani et al. 2010;Jung et al. 2014;Yaglom et al. 2014). Consuming raw goats' milk represents a real risk factor for human toxoplasmosis (Skinner et al. 1990). Detection of toxoplasmosis in different lactating species was recently studied by Dehkordi et al. (2013), with different techniques. They showed that raw milk from all domestic species (goats, sheep, buffalos, cattle and camel) could be contaminated. The prevalence of T. gondii in the milk of naturally infected goats is relatively low. We found that 7.8% of milk samples were shedding Toxoplasma DNA. Similar results were reported by Mancianti et al. (2013) in Italy where 7.9% of the samples were positive. A slightly lower rate was reported in Brazil where the parasite was detected in 6% of the milk samples (15/248) (Bezerra et al. 2013). The environmental conditions, diagnostic methods and the presence of cats in the herds could explain this discrepancy. In this study, the molecular prevalence of T. gondii was significantly higher in Hammam Bourguiba than in Tabarka (P = 0.02). This might be attributed to its climate which is characterized by a high rainfall enhancing oocysts dispersion.
Our results showed also that the infection rate was higher for goats with abortion history (P = 0.01). Since congenital toxoplasmosis is one of the main causes of abortion in goats (Duncanson et al. 2001), the consumption of raw milk from infected goats may present a real risk to public health.
There was no concordance between the two techniques (4/77; 5.2%). Positive PCR and negative ELISA samples could be due to the occurrence of early infections (Bezerra et al. 2013). Further studies are needed to quantify and estimate the viability of T. gondii excreted in the milk. Our results provided a molecular evidence of T. gondii DNA presence in raw goats' milk. Considering these results, authorities must integrate health education guidance to encourage milk consumers to avoid drinking raw milk. Toxoplasmosis infection must therefore be added to the list of pathogens transmitted by raw milk with Brucella melitensis (Gupta et al. 2006).

Source of funding
The work was funded by the laboratory of "Laboratoire d' epid emiologie des infections enzootiques des herbivores en Tunisie: application a la lutte" (Minist ere de l'enseignement sup erieur et de la recherche scientifique, Tunisia).

Conflicts of interest
The authors declare no conflict of interests in relation to this work.

Contributions
SA, MG, MRR and MM conceived and designed the experiments. SA and MR performed the experiments. HN involved in the collection of samples. LS contributed to the analysis of tools. SA and MG wrote the manuscript.