Detection and Molecular Characterization of Porcine Teschoviruses in India: Identification of New Genotypes

Abstract

Porcine Teschoviruses (PTVs) are ubiquitous enteric viral pathogens that infect pigs and wild boars worldwide. PTVs have been responsible for causing the severe clinical disease (Teschen disease) to asymptomatic infections. However, to date, limited information is available on large-scale epidemiological data and molecular characterization of PTVs in several countries. In this study, we report epidemiological data on PTVs based on screening of 534 porcine fecal samples from different states of India and a RT-PCR based detection of PTVs shows a percent positivity of 8.24% (44/534). The PTV prevalence varied among different regions of the country with the highest detection rates observed in the state of Karnataka (38.1%). Phylogenetic analysis based on VP1 gene reveals the presence of PTV genotype 6 and 13 along with some unassigned novel genotypes which did not cluster with any of the established PTV genotypes (PTV 1–PTV 13). Indian PTV 6 strains are genetically closest to the Spanish strains (85.7–94.4%) whereas PTV 13 and novel genotype strains were found to be more similar to the Chinese strains (88.1–99.1%). Using recombination detection software, no Indian PTVs found to be recombinant on VP1 gene and selection pressure analysis revealed the purifying selection in the several sites of the VP1 gene of PTVs. The Bayesian analysis of Indian PTVs shows 1.16 × 10^–4 substitution/site/year as the mean evolutionary rate. Further, isolation of the novel PTV strains from India and more detailed investigation much needed to know the evolutionary history of PTV strains circulating in porcine populations in India.

Introduction

Porcine teschoviruses (PTVs) are classified under the genus Teschovirus within the family Picornaviridae which includes 68 Genus and 158 species of viruses (www.picornaviridae.com June 2023). Originally, Teschovirus was classified under porcine enteroviruses (PEVs) [1]. Further, genomic analysis PEV 1–7, and PEV 11–13 classified them under Teschovirus, PEV 9 and 10 as PEV B (recently known as Enterovirus G) and PEV 8 as Sapelovirus [2–4].

After the first detection in the Czech Republic in 1929, PTVs were found to be responsible for causing Teschen disease in pigs during the 1940s and 1950s, incurring substantial economic losses [5, 6]. Afterward, the severity of polio-encephalomyelitis diminished and the highly virulent Teschen strains were supplanted by less virulent Talfan strains [1, 7]. PTVs are ubiquitous and detected in pig populations worldwide [8]. Pig and wild boars are the only known natural host for PTV infections [9]. Though PTV infections mostly have been associated with subclinical but a variety of clinical conditions, including polio-encephalomyelitis, reproductive disease, enteric disease, pneumonia, and skin lesions also have been reported [9]. Many studies from different countries like America, Russia, Spain, Italy, China, Germany, Japan, and India have reported prevalence and genetic diversity among PTVs [10–14].

PTV has a positive-sense single-stranded RNA genome of 7.1 kb, containing a single long open reading frame (ORF) that is processed to form at least 12 individual mature proteins, comprising a leader (L) protein followed by the viral capsid proteins (VP1–4), and seven non-structural proteins (2A–C, 3A–D) [8]. Among all the structural genes VP1 is the most immunodominant and responsible for genetic variation [15]. To date, 13 distinct genotypes have already been well characterized based on VP1 [8]. Several new genotypes have been proposed from recent genotyping studies from China and Japan [8] but they have not yet been formally recognized by the ICTV. The molecular mechanism involved in the variation and evolution of picornavirus results from genomic rearrangements, point mutation, and particularly recombination [16], and plays a significant role in the evolution by reducing mutational load, producing genetic variations, and by developing viruses with new properties [17]. These studies mentioned the diverse genetic nature of PTV and evolutionary forces responsible for the mutation and generation of new strains under PTV.

Information on PTV is limited from India on PTV detection and genotyping [10, 18, 19], the present study was designed to get long-term comprehensive data about PTV epidemiology in different parts of India. The study further aimed to investigate the genetic diversity and evolutionary rate in PTV isolates from different regions of India.

Materials and Methods

Sample Collection

A total of 534 porcine fecal samples from both healthy (390) and diarrhoeic (144) animals, were collected during the year 2014 to 2019 from different pig farms, located in different regions of India (southern-80, northern-312, and north-eastern-142). Samples were collected from piglets with diverse age groups, sex, and health status.

Viral RNA Extraction and cDNA Synthesis

Total RNA from suspended (10% PBS) stool samples was isolated using Qiazol reagent (Qiagen GmbH, Hilden, Germany) following the manufacturer’s protocol and quantified on a Nanodrop-spectrometer. Extracted RNA (500 ng) was used to prepare a pool of the first-strand cDNA by reverse transcription (RT) using recombinant MMLV-RT (Promega Corporation, Madison, WI, USA) and random hexamer (Qiagen GmbH, Hilden, Germany) at 37 °C [20].

Diagnostic RT-PCR Assay

PTV in stool samples was screened using the duplex RT-PCR assay (for both porcine Tescho and Sapelo virus) targeting the 5’-UTR regions [21] using the SapphireAmp Fast PCR master mix (Takara Bio Inc., Shiga, Japan). These samples were also screened for other gastroenteritis causing viruses viz. Rotavirus, Calicivirus [22], Astrovirus [23], and Picobirnaviruses [24] employing the primers in RT-PCR assays as outlined in the published reports.

Amplification, Cloning, and Sequencing of the PTV VP1 Gene

Major capsid gene VP1 has been amplified (533 bp) using published primers [25] and amplicons were visualized under a UV trans-illuminator after resolving in 2% agarose gel electrophoresis. Desired amplified products were cloned into the pDRIVE (Qiagen GmbH, Hilden, Germany) cloning vector and transformed in Escherichia coli DH5α competent cells. The recombinant plasmids were obtained using the GeneJET plasmid Miniprep kit (Thermo Fisher Scientific, Waltham, MA, Vilnius, Lithuania). Positive recombinant clones were sequenced using the Bigdye terminator Sanger sequencing method in an ABI 3730 × l sequencer (Eurofins Genomic India Ltd., Bangalore, India). The VP1 sequences (18) were deposited in GenBank with Accession Numbers MK139557-74.

Statistical analysis

The statistical analysis was performed in SPSS version 20 on the Windows platform (IBM Corp, Armonk, NY, USA). Information from the questionnaire was entered into a Microsoft Excel spreadsheet (Microsoft Corporation) and piglet PTV results were coded as negative = 0 and positive = 1. The data collected about weaning health status, sex, mixed species farm/ presence of other animals in the farm, location of the farm altitude, ventilation, water source, feed type, use of heater/cooler for temperature control in the sheds, sheds in different elevation, use of disinfectant were used as independent variables and PTV positivity was used as the outcome variable. Cross-tabulation or Fisher’s exact test was used to test the significance of the univariate associations between the explanatory variables and the outcome variable. Fisher’s exact test was used when expected cell frequencies were < 5. Analysis of multiple predictors of PTV positivity was performed using stepwise forward logistic regression analysis considering only those factors significant p ≤ 0.1 in univariate analysis and retaining only factors significant at p ≤ 0.05 in the final model [26].

HeatMap Analysis of PTV Genotypes All over the World

Heat maps of PTV genotype distribution all over the world were generated using GraphPad Prism version 7. All over the world at least 13 different genotypes of PTVs have been reported based on the VP1 structural gene. Though recently several studies described isolates beyond genotype 13, they are not yet formally recognized by ICTV and not designated as assigned genotypes. Also, the sequences for which country, year, and genotypes have not been mentioned were excluded for the heat map analysis.

Phylogenetic Analyses

For the phylogenetic analysis, all the PTV sequences (> 533 bp, VP1) available in the NCBI GenBank database (strains without year, country, and genotype excluded) were retrieved. A total of 400 sequences including our 18 VP1 Indian PTV sequences were aligned using clustalW algorithm and were used for creating a phylogenetic tree based on maximum likelihood and using the IQ-TREE webserver [27]. The generated tree from the IQ webserver was further annotated in the iTOL (version 4.4.2) [28]. Further 400 sequences were reduced to 124 sequences to obtain a focussed, enlarged, and well-annotated phylogenetic tree for the reader presented in the manuscript (Fig. 3).

Fig. 3.

Phylogenetic analysis of PTV isolates from all over the world based on the VP1 nucleotide sequence (533 bp). Strains are designated as genotype, accession no, country, and isolation year. Strains belonging to different genotypes and Indian strains are indicated with different color codes

Recombination Analysis

The potential recombination events within the PTV VP1 gene sequence were analyzed by Recombination Detection Program (RDP) 4.84 [29]. Seven methods were used in the RDP software which include RDP, GeneConv, BootScan, MaxChi, Chimaera, SiScan, and 3Seq.

Estimation of Substitution Rates and TMRCA (Time to the Most Recent Common Ancestor)

The evolutionary rate of the PTV strains from all over the world was inferred using the Bayesian Markov Chain Monte Carlo (MCMC) statistical framework implemented in the BEAUti/BEAST package v1.10.4. The TempEst program (v1.5.1) was used to determine that sequences used in this work showed eligible temporal structure to proceed with molecular clock analyses. Different combinations of demographic models and clock models were compared as suggested, like constant size or Bayesian skyline plot models, strict or relaxed (lognormal or exponential) clocks [30]. At least 30 million MCMC iterations were run for each gene and each demographic model/clock model combination. The estimation of parameters and divergence time was carried out using Tracer1.7.1 (http://beast.bio.ed.ac.uk/Tracer). A correct mixing of MCMC was verified by effective sampling size (ESS) calculations available in Tracer.

Selection pressure analysis

All available PTVs were analysed for estimation of the nature of selection pressure on the virus using the ORF of PTVs with the Datamonkey web-server (www.datamonkey.org). Four different codon-based ML methods were used to estimate the ratio of non-synonymous to synonymous substitutions (dN/dS, also known as Ka/Ks or x) at every codon in the alignment: the single-likelihood ancestor counting (SLAC), fixed effects likelihood (FEL), random effects likelihood (REL) and fast unconstrained Bayesian approximation for inferring selection (FUBAR). In each method, the confidence level was set to a Bayes factor of 50 or a p-value of 0.1. Site-specific selection pressure was measured on the VP1 gene using the HyPhy software implemented in the Datamonkey webserver [31]. In this study, a total of 400 VP1 coding sequences of PTV (382 sequences retrieved from NCBI and 18 sequences of the present study) were chosen and this server accepted 326 sequences by declining the similar/duplicate sequences (Supplementary data 1). The sites under selection pressure were evaluated using classic maximum likelihood methods i.e., Single Likelihood Ancestor Counting (SLAC) model and Fixed Effect Likelihood (FEL) model [32] and other more recently developed methods; Mixed Effects Model of Evolution (MEME) capable of identifying both episodic and pervasive positive selection [33], and the Fast Unbiased Bayesian AppRoximation (FUBAR) method that can detect positive selection. All these methods were employed using the best nucleotide substitution model, 010010 (HKY85 model). The strength of selection pressure was determined based on the ratio of nonsynonymous (dN) to synonymous (dS) substitutions per site (x = dN/dS). To avoid a high false-positive rate, due to the reduced number of sequences, sites with P values < 0.1 for SLAC, FEL, and MEME models, and a posterior probability > 0.90 for FUBAR were accepted as candidates for selection.

Results

Prevalence and Risk Factor Analysis

Porcine teschovirus was detected in 44 out of 534 samples (8.24%). All the fecal samples were found negative for rotavirus, calicivirus, astrovirus, and picobirnaviruses. The highest prevalence was detected in the Indian state of Karnataka (8/21–38.1%), followed by Tamil Nadu (3/11–27%), West Bengal (12/80–15%), Uttar Pradesh (19/312, 6.1%), and Kerala (2/48, 4.1%) (Table 1, Fig. 1). The risk factor analysis revealed that mixed-species farm (OR- 0.52, 95% CI 0.27–0.99) and diarrhoeic piglets (OR- 0.43, 95% CI 0.22–0.85) were associated with reduced risk for PTV infection while the location of the farm in non-hilly/plain regions (OR- 9.08, 95% CI 1.14–72.18), and pre-wean piglets (OR- 2.02, 95% CI 1.02–4.04) was associated with a higher risk of PTV infection. The risk factor associated with climate and sex of the animal, no significant relationship has been found.

Table 1.

PTV prevalence in different states of India

State	Sample screened	PTV positive
Uttar Pradesh	312	19 (6.1%)
Meghalaya	17	0
Nagaland	35	0
WB	80	12 (15%)
Assam	10	0
Karnataka	21	8 (38.1%)
Tamil Nadu	11	3 (27.27%)
Kerala	48	2 (4.16%)
Total	534	44 (8.24%)

Fig. 1.

PTV prevalence in different states of India

Heat Map Analysis of PTV Genotypes from All over the World

The heat map analysis shows that China, Italy, Spain, and Germany have been found with reports of most of the established PTV genotypes (Fig. 2). In the heat map, the rows represent the country and columns report the genotypes found in that specific PTV isolate. China, Germany, Japan, India, and the USA have reported some novel diverse genotypes assigned as NA (Not assigned). The Dominican Republic, Haiti, and Ukraine have reported only a single PTV genotype (PTV 1). In India PTV genotype 5, 6, 8, and 13 have been detected among which PTV 5 and 8 reported earlier studies whereas our finding reveals the presence of PTV genotype 6 and 13. In India also some novel genotypes we have found in our present study.

Fig. 2.

Distribution of PTV genotype in different countries. The heat map shows the PTV genotypes (red & blue) and absence (cross) profiles. The color gradient strip depicts the number of PTV genotypes associated with a particular country

Phylogenetic Analysis of Partial VP1

Based on the phylogenetic analysis all the 18 PTV partial VP1 sequences (533 bp) from India were grouped into three different clusters according to the genotype (Fig. 3). The PTV field isolates were grouped as follows: eight strains in genotype 6 (Karnataka, West Bengal, Kerala, Uttar Pradesh), six strains in genotype 13 (Tamil Nadu, West Bengal, Uttar Pradesh), and four strains (Karnataka, Kerala, Tamil Nadu) did not cluster with any established genotype and were represented as unassigned genotype (as already different study proposed genotype beyond 13 but not yet formally recognized by ICTV). In the phylogenetic tree Indian PTV 6 strains were grouped with strains from China, USA, Germany, and Spain whereas Indian PTV 13 were clustered with Chinese, Japanese, and Hungarian PTV strains. All the not assigned PTV genotypes from India clustered with Chinese PTV strains.

Sequence Distance Analysis

The Indian PTV strains show a wide range of genetic variations based on VP1 gene nucleic acid (NA) and amino acid (AA) sequence analysis (63.1%-99.3%). Based on both NA and AA level sequence comparison all the Indian strains were closely related to Spain (85.7–94.4%) and Germany (86.4–94.4%) isolates (Fig. 4). Indian PTV13 isolates have shown the most similarity with Chinese strains (88.1–96.6%) (Fig. 4). PTV strains (MK139559, MK139561, MK139572, MK139573) from India (Kerala, Tamil Nadu, and Karnataka) have not clustered with any of the 13 genotypes of PTV. These strains are described as unassigned (UA) PTVs. These Indian UA PTVs show the most similarity with Chinese strains (98.3–99.1%) (Fig. 4).

Fig. 4.

Sequence distance based on nucleotide and amino acid of Indian PTV strains along the PTV strains from all over the world

Recombination Analysis

None of the VP1 sequences from India showed any recombination events within the selected gene (partial VP1) portion.

TMRCA Analysis

The temporal signal was estimated with TempEst software and sequences with dissimilar data/variability were discarded. As in our study, we have detected the Indian PTVs belong to the genotype 6, 13, and some not assigned so separate files according to the genotype were prepared for evolutionary analysis in Beast 1.10.4. Among all the combinations only 32 non-recombinant PTV 6 isolates were shown good temporal signal and we got the Beast analysis results. Both the strict and relaxed molecular clock models were implemented with Bayesian Skyline Plot for population growth. Among them, the strict molecular clock converged with good ESS values. So, we have mentioned here only PTV 6 strict molecular clock analysis. Other model log files are attached in supplementary. All of the Indian strains of PTV6 were clustered in one lineage in the MCCT, supported by posterior probabilities ranges from 0.32–0.99 (Fig. 5). The estimated evolutionary rate was 1.16 × 10^–4 subs/site/year (95% HPD 3.91 × 10^–4–2.13 × 10^–3).

Fig. 5.

Time-scaled Bayesian MCC phylogeny of PTV6 genotype. PTV6 VP1 sequences from India, UK, Spain, and China were used for phylogeny reconstruction. Branch containing Indian sequences indicated in red

The demographic reconstruction of PTV 6 strains has been observed in the skyline plot (Fig. 6). The viral population has been divided into two phases. The first phase shows a constant population size until around 2005. Then there is a sudden decrease in population during a period of 10 years and again become constant and after 2015 it shows a mild increase in population, suggesting the population is growing.

Fig. 6.

Population dynamics of the 32 PTV of genotype6 strains. The demographic history of PTVs shown since its origin in 1958 until the most recent sequence of the dataset, 2018. The Bayesian Skyline plot shows the evolution of population size. Median (dark line) and upper and lower 95% HPD (blue region) estimates of effective population size (y-axis) through time in years (x-axis) are shown

Selection Pressure Analysis

To increase the robustness of the selected sites, we applied four methods (SLAC, FEL, MEME, FUBER) on all sequences of PTV. MEME result shows 5 sites with evidence of episodic diversifying selection at a significance level of 0.1. Whereas, only SLAC analysis showed the overall dN/dS ratio for all non-recombinant PTVs as 0.0658, suggesting that PTV is subjected to strong purifying (negative) selection.

Discussion

Teschovirus encephalomyelitis (previously known as Teschen/Talfan disease) caused by at least 13 genotypes of porcine teschovirus (PTV 1–13) is considered to be of socio-economic importance. Many studies on experimentally infected PTV cases have reported non-suppurative encephalomyelitis with neurological signs [7, 34, 35]. In previous studies, PTV prevalence in porcine fecal samples was mentioned between 6.84% (India) to 90% (China) [12, 13]. In our present study, we have detected a 7.14% prevalence which falls between the results from previous studies from India [10, 18]. Moreover, in our study sample size (534) was larger than the previous studies covering several regions of India. Regarding risk factor analysis data, a previous study from India has shown more PTV prevalence in healthy, nursery age group, male, and in the winter season [18]. In our present study, we have detected a significantly higher infection rate of PTV in asymptomatic pigs than in those with diarrhoea, which is consistent with previous studies [36]. In contrast to findings related to the absence of PTV infection in suckling pigs by John et al. [18], we have found more PTV prevalence in pre-weaned or suckling pigs. Other factor includes farm type; we have detected more PTV prevalence in farms located in the plain regions and farms where only porcine species were kept. Phylogenetic analysis based on partial VP1 gene confirms the presence of two established genotypes (PTV 6 and PTV 13) and some novel or unassigned genotypes. PTV genotype 6 strains have been detected for the first time in India with the highest prevalence among all the detected genotypes. Previously, a study from India described PTV genotype 5(KY548743, KY548746), 8 (KY548745), and variant (KY548744) [12]. Chinese and Spanish strains were recognized as closest to Indian strains. Previous studies on Indian PTV molecular characterization also show some Indian PTV genotypes were most similar with Chinese (89%) PTV isolates which correlate with our study [10]. Our Indian PTV VP1 sequences were non-recombinant though several recombination events are mentioned in Chinese studies [8]. In earlier studies, recombination events were not mentioned in Indian PTVs. Also, for the first time, we have estimated the evolutionary analysis of Indian PTV strains (PTV 6) which revealed the evolutionary rate of 1.16 × 10^–4 substitutions per site per year (subs/site/year). Picornaviruses, especially enteroviruses, generally have high nucleotide substitution rates [37]. Previous studies regarding VP1 gene-based mean evolutionary rate analysis of PTV suggested higher (2.46 × 10^–3 subs/site/year) [38]. However, a longer gene sequence is considered a much better target than a short region for estimating the nucleotide substitution rate of PTV [8].

RNA viruses, especially picornaviruses, have high mutation rates than DNA viruses. Our study analysis shows a lesser rate of evolution than several recent studies from China and Spain [8]. As recombination is one of the major factors in evolution so, it may be one of the reasons for less mutation rate of our non-recombinant strains. In the selection pressure analysis of PTVs, we have detected negative selection rather than positive selection. In previous studies also it has been found that purifying or negative selection is predominant among PTVs [8, 39]. The negative selection also affects the population growth of viruses, which we have seen in our Bayesian skyline plot as a reduction of the population (Fig. 5). Other several evolutionary studies of RNA viruses regarding Bayesian skyline plot shows a reduction in the population over time [40, 41].

Conclusions

In conclusion, in our study 18 PTV partial VP1 sequences were obtained which were analysed with a metadata set encompassing cognate PTV sequences from different parts of the world, revealing predominant genotypes and a comparatively lower rate of evolution of PTVs in India. Genotype 5, 6, 8, and 13 along with novel strains has been prevalent in India based on previous and our study. PTV 6 was found to be the most prevalent PTV genotype circulating in the Indian pig population. Also, PTV reported more in asymptomatic healthy animals rather than with overt clinical symptoms which shows the importance of easy diagnostics which can screen rapidly the asymptomatic carrier animals to prevent the transmission of infection. This also keeps the chances of healthy animals serving as a reservoir for disease dissemination. Further in-depth studies are needed to reframe the classification pattern of PTV as new species and genotypes mentioned in several studies. Also, the pathogenicity studies related to different PTV genotypes must be done in the future, especially focusing on the novel strains.

Author Contributions

YSM and SG conceptualized and designed the experiments; SB screened the samples; SBJJK and SS, VOR, and PT performed the experiments and designed the figures. SB, YSM written the manuscript; SG edited the manuscript; YSM critically analyzed and finalized the manuscript. All authors have read and agreed to the published version of the manuscript.

Funding

The present study was funded by ICAR-National Fellow project, Indian Council of Agricultural Research, Ministry of Agriculture and Farmers Welfare, Government of India.

Data Availability

Data are available from the authors upon reasonable request; sequences are available via GenBank.

Declarations

Conflict of interest

The authors have no relevant financial or non-financial interest to disclose.