Innovative molecular diagnosis of T richinella species based on β‐carbonic anhydrase genomic sequence

Summary Trichinellosis is a helminthic infection where different species of T richinella nematodes are the causative agents. Several molecular assays have been designed to aid diagnostics of trichinellosis. These assays are mostly complex and expensive. The genomes of T richinella species contain certain parasite‐specific genes, which can be detected by polymerase chain reaction (PCR) methods. We selected β‐carbonic anhydrase (β‐ CA) gene as a target, because it is present in many parasites genomes but absent in vertebrates. We developed a novel β‐CA gene‐based method for detection of T richinella larvae in biological samples. We first identified a β‐CA protein sequence from T richinella spiralis by bioinformatic tools using β‐CAs from C aenorhabditis elegans and D rosophila melanogaster. Thereafter, 16 sets of designed primers were tested to detect β‐CA genomic sequences from three species of T richinella, including T . spiralis, T richinella  pseudospiralis and T richinella  nativa. Among all 16 sets of designed primers, the primer set No. 2 efficiently amplified β‐CA genomic sequences from T . spiralis, T . pseudospiralis and T . nativa without any false‐positive amplicons from other parasite samples including T oxoplasma gondii, T oxocara cati and P arascaris equorum. This robust and straightforward method could be useful for meat inspection in slaughterhouses, quality control by food authorities and medical laboratories.


Introduction
Trichinella spp. are the causative agents of a zoonotic helminthic disease, which is called trichinellosis (Gottstein et al., 2009). Within the main sources of Trichinella spp.
[including pigs, horses, wild boars, dogs, walruses, foxes, bears and birds (Alban et al., 2011)], the most important source of human infection worldwide is the domestic pig, e.g. in Europe, whereas horse and wild boar meat have played a significant role during outbreaks within the past three decades (Gottstein et al., 2009). Infection in humans occurs with the ingestion of insufficiently cooked meat contaminated with Trichinella larvae, which are encysted in muscle tissues of the host animals (Clausen et al., 1996;Pozio, 2007;Gottstein et al., 2009). Due to the re-emerging problem also in Europe, the European Union and some associated non-European Union member countries still run Trichinella monitoring programmes (Gottstein et al., 2009). Following delivery by the gravid female worm, which lives within the intestinal mucosa of the host, newborn larvae (NBL) predominantly migrate into lymphatic and blood vessels of the host and thereafter to the ultimate target organ, i.e. highly oxygenated muscle. NBL enter the striated muscle cells by the aid of their stylet (Gottstein et al., 2009). NBL survive in muscle nurse cells for years (up to 40 years in humans and over 20 years, e.g. in polar bears) (Froscher et al., 1988;Kumar et al., 1990). Among 12 different genetically detected Trichinella spp., Trichinella spiralis is the most common species in domestic and wild swine and is also the most important aetiological agent that causes wide and global distribution of trichinellosis in humans (Pozio and Darwin Murrell, 2006). Other Trichinella spp. include Trichinella nativa (La Rosa et al., 2001;Nelson et al., 2003), Trichinella pseudospiralis (La Rosa et al., 2001;Pozio and Darwin Murrell, 2006), Trichinella britovi (Gottstein et al., 2009), Trichinella murelli (Ancelle, 1998), Trichinella nelsoni (Marucci et al., 2009), Trichinella papuae (Pozio et al., 2005), Trichinella zimbabwensis (Pozio et al., 2002), Trichinella genotype T8 (Marucci et al., 2009), Trichinella genotype T9 (Zarlenga et al., 2006) and Trichinella genotype T12 (Krivokapich et al., 2012). Seven percent of infected animals carry mixed infections (Airas et al., 2010).
In this study, we designed and piloted a PCR method for rapid detection of three species of Trichinella larvae, including T. spiralis, T. pseudospiralis and T. nativa in meat samples.

Multiple sequence alignment (MSA)
MSA of the β-CA protein sequence from T. spiralis (UniProt ID: E5SH53) against the sequences from Drosophila melanogaster (UniProt ID: Q9VHJ5) and Caenorhabditis elegans (UniProt ID: Q22460) revealed that they all contain the first (CxDxR; C: cystein, D: aspartic acid, R: arginine, and x: any residue) and second (HxxC; H: histidine, C: cystein, and x: any residue) highly conserved residues ( Fig. 1). Hence, the β-CA protein sequence from T. spiralis (UniProt ID: E5SH53) extracted correctly through the Basic Local Alignment Search Tool (BLAST) homology analysis and contained identical highly conserved residues to the predefined β-CA protein sequences.

Primer design and PCR
Among the four different pairs of primers (Nos. 1-4) only the primer set No. 2 (PCR product size 191 bp) showed a clear positive DNA band on agarose gel electrophoresis for detection of T. spiralis β-CA genomic sequence in larvae and meat samples (data not shown). It also produced positive bands on larvae samples from T. pseudospiralis and T. nativa (Fig. 2). The next rounds of PCR analyses were performed to find out whether it was possible to design primers specific for T. pseudospiralis and T. nativa. We used the same reverse primer (primer set No. 2) which worked well for the first PCR round. The PCR products representing partial sequences of T. pseudospiralis and T. nativa β-CA gene were sequenced, and the forward primers were designed based on the obtained sequences. These primers were included in the sets Nos. 5-7 for T. pseudospiralis and 8-11 for T. nativa, and the PCR results are shown in Fig. 3. Most primers produced multiple bands. When the most promising primers (No. 6 for T. pseudospiralis and No. 10 for T. nativa) were tested against all three species of Trichinella, they completely cross-amplified because of the high sequence similarity (data not shown). Due to the interspecies crossamplification we further designed four new reverse primers which were used together with the previous forward primers resulting in primer set Nos. 12-16.
The results revealed that the primer set No. 13 from T. nativa and primer set No. 12 from T. pseudospiralis produced the strongest DNA bands on the agarose gel electrophoresis (Fig. 4). For evaluation of crossamplification these primers were tested on all three Trichinella spp. Figure 5 shows that there was again a complete cross-amplification between different Trichinella spp. by these primers. This finding led us to conclude that finding of species-specific primers for Trichinella β-CA genes is probably impossible due to the high sequence similarity. Based on our results the primer set No. 2 was considered the most potential tool for diagnostic purposes, even though it was not able to discriminate between different Trichinella spp. Therefore, its specificity was further tested on prepared samples from Toxoplasma gondii, Toxocara cati, and Parascaris equorum. No falsepositive reactions were detectable (Fig. 6).

Discussion
Many protozoan, arthropod, and nematode species (invertebrate metazoans) contain β-CA gene and protein sequences in their genomes and proteomes respectively. BLAST homology analysis of predefined β-CA protein sequences from D. melanogaster (Uniprot ID: Q9VHJ5) and C. elegans (Uniprot ID: Q22460) resulted in identification of more than 75 β-CA protein sequences from protozoans and invertebrate metazoans (Zolfaghari Emameh et al., 2014). One of them was β-CA protein sequence from T. spiralis (UniProt ID: E5SH53). Our previous results also revealed that β-CA proteins show high   (Zolfaghari Emameh et al., 2014). This led us to hypothesize that the high similarity is extensible to other species of one defined genus, such as Trichinella spp. including the clinically most relevant species: T. spiralis, T. pseudospiralis, and T. nativa. In principle, the high sequence homology at gene level could provide a straightforward avenue to develop a genusspecific PCR assay for diagnostics of Trichinella infection (trichinellosis). On the other hand, the high sequence similarity could be a difficult challenge for development of a species-specific assay based on the β-CA gene.
A number of PCR-based diagnostic methods have been developed for detection of infectious diseases, and some of them have been applied for parasitic diseases, such as trichinellosis (Li et al., 2010). In our study, we designed and tested 16 different pairs of primers for detection of β-CA genomic sequences of Trichinella spp. At the beginning of assay development, it is necessary to test the efficiency of designed primers. Our method using the primer set No. 2 seemed to detect minute amounts of DNA (a single larva/reaction). Thereafter, the set No. 2 was considered the most promising primer pair for the development of genus-specific tool for Trichinella diagnostics. The results also revealed that interspecies crossamplification is a recurrent challenge in this assay system when the DNA sequences are highly similar. None of the tested primers were able to discriminate between different Trichinella spp. As already pointed out, the described method was no able to discriminate between various species of Trichinella. This same problem hampers most detection methods reported to date. All species and genotypes of the genus Trichinella are morphologically indistinguishable at all developmental stages. Consequently, only biochemical or molecular methods can be used to identify the genotype of the parasite (Gajadhar et al., 2009).
Our results showed that the developed assay is robust enough for screening of Trichinella infection in routine meat samples. The positive signals were genus-specific since no false-positive bands were detectable when samples from other parasites, a protozoan T. gondii and two nematodes T. cati and P. equorum, were used as templates. Importantly, no β-CA genes are present in mammalian genomes (Syrjanen et al., 2010;Zolfaghari Emameh et al., 2014). Therefore, host-derived falsepositive reactions would be unlikely. Other advantages of the present method include both ease of use and rapidity of detection. The meat samples are lysed overnight, and thereafter, the sample is ready for PCR amplification. No separate DNA isolation procedure is needed for the preparation of the template. The PCR reaction and gel electrophoresis take about 3 h after which the positive bands are clearly visible for documentation.   As an overall conclusion, even though there are many diagnostic methods available for trichinellosis, most of them are either expensive, time-consuming or require special expertise. Therefore, new simple tools are urgently needed for screening of biological samples in livestock slaughterhouses and quality controlling laboratories of meat processing companies. Our developed method is based on detection of β-CA genomic sequences from Trichinella spp. The greatest advantage of this method is that mammalian hosts do not contain any β-CA gene sequences in their genomes. The results indicated that the PCR method is Trichinella genus-specific and robust enough for routine screening purposes.

Preparation of parasites and infected meat samples
Parasite samples and Trichinella spp.-infected mouse muscle specimens were prepared at the Department of Veterinary Biosciences, University of Helsinki, Finland. The Animal Experiment Board in Finland had approved the study protocol. The tested Trichinella spp. involved T. spiralis, T. pseudospiralis and T. nativa. The blinded samples, including both infected and uninfected meat, were packed and shipped to Tissue Biology group, School of Medicine, University of Tampere, Finland, for PCR analysis. After arriving in the laboratory, the samples were transferred to −20°C freezer. To test the specificity of different primers, three more parasite samples were analysed including T. gondii (in culture medium), T. cati and P. equorum (in 70% ethanol).

Lysis of the parasites and meat samples
The meat and larvae samples were handled under the laminar hood. Toxoplasma gondii parasites were separated from the culture medium by centrifugation (Heraeus Biofuge Fresco, Thermo Scientific, Waltham, MA) at 13 000 RPM for 10 min (Leroux et al., 2015). Then the following lysis protocol was performed on all the samples: 300 μl of lysis buffer (DirectPCR Tail Lysis reagent, Viagen Biotech, LA) was added to 5 mg of samples. Then 6 μl of proteinase K (Thermo Scientific) was added to the mixture. The lysis tubes were incubated in rotating incubator (HYBAID, Thermo Scientific) at 55°C overnight.

Primer design
Three sets of primers for genomic sequence of β-CA (exonic and intronic) from T. spiralis (Ensembl Metazoa ID: EFV55868) were designed automatically by NCBI Primer-BLAST tool (http://www.ncbi.nlm.nih.gov/tools/primer-blast/) (Ye et al., 2012), and one additional primer set was designed manually (Table 1). The expected sizes of the PCR products were calculated by Oligo Calc (http:// www.basic.northwestern.edu/biotools/oligocalc.html) (Kibbe, 2007). To obtain genomic sequences for β-CA of T. pseudospiralis and T. nativa we first performed PCR amplification using the designed T. spiralis primers, and T. pseudospiralis and T. nativa larvae were used as templates. The corresponding bands for T. pseudospiralis and T. nativa on the agarose gel were cut and sequenced as described in the sequencing section. The obtained β-CA genomic sequences from T. pseudospiralis and T. nativa were aligned by Clustal Omega algorithm (http:// www.ebi.ac.uk/Tools/msa/clustalo/) (Sievers et al., 2011) within EMBL-EBI database (http://www.ebi.ac.uk/Tools/msa/ clustalo/) against the genomic sequence of β-CA from T. spiralis, and the most non-identical regions were selected for the second round of primer design. We designed 3 and 4 different forward primers manually for T. pseudospiralis (primer set Nos. 5-7) and T. nativa (primer set Nos. 8-11), respectively, and used with the same reverse primer No. 2 which was described for the first round. In the third round of primer design, one new reverse primer was designed for T. pseudospiralis (primer set No. 12) and used with its own forward primer No. 7. Also, four new reverse primers were designed for T. nativa and tested with its own forward primers Nos. 8 and 10 (primer set Nos. 13-16). Three other species, T. gondii, T. cati, and P. equorum were tested for any falsepositive cross-amplifications with other parasites.
Sequencing of β-CA genomic sequences from T. pseudospiralis and T. nativa We cut the PCR products of T. pseudospiralis and T. nativa templates from an agarose gel. Illustra-GFX PCR DNA and Gel Band Purification Kit (GE Healthcare, Buckinghamshire, UK) was applied for extraction and purification of the corresponding DNA bands. At the final step, the overnight-dried samples were suspended in HiDi (Life Technologies Europe, Applied Biosystems, Finland), which was followed by vortexing and spinning. The samples were heated at 95°C for 2 min and transferred to ice for 2 min. The samples were sent to Core Facilities and Research Services, BioMediTech, Tampere, Finland for DNA sequencing. The sequencing was carried out with HITACHI 3130x/Genetic Analyzer (Life Technologies Europe, Applied Biosystems).

Authors' contributions
RZE carried out blast homology analysis and MSA on T. spiralis, T. pseudospiralis, T. nativa, C. elegans, and D. melanogaster β-CA sequences. RZE and MK participated in the primers design and setting up the PCR method. AN and AS prepared the parasite and infected meat samples. All authors participated in the design of the study. RZE and MK drafted the first version of the paper. All authors read and approved the final manuscript.

Conflict of Interest
None declared.   T. nativa 319