Case Report: First Report and Phylogenetic Analysis of Porcine Astroviruses in Chile

Porcine Astrovirus (PoAstV) causes mild diarrhea in young pigs and is considered an emerging virus in the swine industry worldwide. PoAstV has high genetic diversity and has been classified into five genetic lineages, PoAstV1–5. In Chile, only human astroviruses have been reported. This study aimed to determine the presence and genetic diversity of PoAstV circulating in intensive pig farms in Chile. Seventeen Chilean intensive swine farms from Valparaíso, Metropolitana, O'Higgins, Ñuble and Araucanía regions were sampled. A selection of oral fluid and fecal material samples from 1–80 days-old pigs were collected and analyzed using next-generation sequencing. The circulation of PoAstV was confirmed in all studied farms. We obtained complete or partial sequences of PoAstV-2 (n = 3), PoAstV-4 (n = 2), and PoAstV-5 (n = 7). In 15 out of 17 farms, we detected more than one lineage co-circulating. Phylogenetic analyses grouped the seven PoAstV-5 strains in a monophyletic cluster, closely related to the United States PoAstV-5 strains. The three PoAstV-2 were located into two separate sub-clusters. PoAstV-4 sequences are also grouped in two different clusters, all related to Japanese strains. Thus, our results indicate that PoAstV circulates in Chile with high frequency and diversity. However, the lack of reference sequences impairs local evolution patterns establishment and regional comparisons. This is the first contribution of PoAstV genomes in Latin America; more studies are needed to understand the diversity and impact of PoAstV on swine health.

In humans, AstV is the third most common cause of viral diarrhea in young children worldwide, with high seroprevalences as 94% in children of 6-9 years old (13,14). Despite this high prevalence, due to the lack of cell culture systems and animal models, AstV are among the least studied enteric RNA viruses (12). However, advances in sequencing technologies have increased the availability of genome sequences and the identification of new strains. Typically, human AstV (HAstV) infections cause acute self-limiting mild diarrhea (12). Nevertheless, immunocompromised patients occasionally exhibit systemic spread, resulting in neurologic disease (15). In addition, a zoonotic potential of AstV is suspected but remains unclear (16)(17)(18). Genetic and evolutionary studies support the idea that both cross-species transmission and recombination events among AstV of human, porcine, and other species origin, may have occurred (11, 16).
Porcine astrovirus (PoAstV) has high genetic diversity, it is worldwide distributed, and it is commonly detected and shed by healthy and diarrheic swine (19). Five lineages of porcine AstV (PoAstV1-5) have been described by Laurin,Dastor (20), but only PoAstV-1 has been officially classified under the species Mamastrovirus 3 (21). In the US, pigs are commonly (13.9%) coinfected with multiple astrovirus strains (19). High prevalence and co-infections may create appropriate conditions for viral recombination and the potential emergence of viral variants that pose a higher risk of clinical disease. Recently, PoAstV has been linked to extraintestinal infections suggesting more complex pathogenesis and serious outcomes than previously thought (3,4,22).
In Chile, a recent study reported human AstV infections (14%) as a predominant cause of viral gastroenteritis in rural zones, in addition to norovirus (15%) and rotavirus (14%) (23). However, animal AstV has not been described in Chile. Considering the high prevalence and worldwide distribution of the AstV and PoAstV, the suspected zoonotic potential, and the lack of information regarding these viruses in a regional context, this study aims to determine the presence and genetic diversity of PoAstV circulating in Chilean intensive pig farms. These samples were taken in the context of other viral disease surveillance programs .

Sample Collection
During influenza virus and rotavirus surveillance and diagnosis programs, we collected oral fluids and fecal samples from 1-80 days-old pigs in 17 intensive pig farms from mainland Chile in 2015 and 2017. The sampled farms are located in an area that concentrates 95% of the national intensive pig production (Valparaíso, Metropolitana, Libertador General Bernardo O'Higgins, Maule, Ñuble, and Araucanía Regions), which represents approximately 50% of the pig inventory in Chile (24,25). Each fecal sample corresponds to a pool of 5 diarrheic feces, which were collected using nylon gloves, deposited in sterile 50 mL tubes with 20 mL of viral transport media (Minimum Essential Medium, 1X Trypsin TPCK, 2% bovine serum albumin, and 1% antifungal antibiotic solution), and then centrifuged at 7,000 rpm for 5 min. Oral fluids were collected by groups of 20-30 healthy pigs kept in pens. Briefly, a 16 mm braided cotton rope was hung in each pen for about 30 min. The ropes were deposited inside plastic ziplock bags and squeezed to obtain the oral fluid and deposited into 50 mL tubes. All samples were kept at −20 • C until processing. One sample per farm (n = 17) was selected for next-generation sequencing. The criteria to select the samples included the location, geographic distance between farms and detection of other pathogens such as rotavirus and influenza.

Viral RNA Extraction and Whole-Genome Sequencing
The RNA extraction was carried out using the Chomczynskiphenol solution (Winkler, BM-1,755, Chile) following the manufacturer's recommendations. The Next-generation sequencing (NGS) was performed at the Molecular Diagnostic Development Laboratory at the Veterinary Diagnostic Laboratory of the University of Minnesota (MVDL, UMN), USA, using the Illumina MiSeq platform. Library pre-paration was performed using the SMARTer Stranded Total RNA-Seq Kit v2-Pico Input Mammalian (Takara bio, USA). De novo assembly of the reads was carried out using an automated pipeline that identifies viral reads using DIAMOND protein alignment and the Swissprot Uniref90 database. The viral reads are then grouped by the lowest common ancestor and assembled using SPAdes and subsequently, the contigs are joined using an Advanced Genome Aligner (http://www.genomedetective.com/ app/typingtool/virus/). Complementary, the assembly using PoAstV reference sequences was performed using Geneious Prime R 2021.2.2.

Phylogeny
Complete or near to complete Chilean PoAstV genomes, with >77% of coverage, were used for phylogenetic analysis ( Table 1). These sequences were compared with all complete or near complete PoAstV genome available in GenBank database. We used the lineage classification described by Lee et al. (26). The final data set comprised 93 Astrovirus genome sequences that were aligned using MUSCLE (27). The phylogeny was constructed using RAxML with the GTR+G+I substitution model and 1,000 bootstrap replications in Geneious Prime R 2021.2.2. Additionally, the phylogeny was constructed for the ORF2 region with the same methodology, and p-distances at the nucleotide and amino acid level of the ORF2 sequences were estimated using MEGA X (28).

RESULTS AND DISCUSSION
All samples were successfully sequenced by Illumina. The most consistent virus family found in the samples was Astroviridae. Additionally, reads that belong to other families, such as Caliciviridae, Parvoviridae, and Reoviridae, were found at lower rates and were not considered for further analysis as they are beyond the aim of this study. Porcine astrovirus reads were observed in all samples/farms included, confirming its ubiquity in Chilean swine intensive production. Overall results of PoAstV identified 229,496 reads for PoAstV-5, 33,917 for PoAstV-4 and 31,751 for PoAstV-2. Only 10 reads were classified as PoAstV3,   Table 1). These results are in agreement with estimations made in the US and some European and Asian countries, where the most prevalent lineage is PoAstV-4, mainly followed by PoAstV-2 (19,(29)(30)(31)(32)(33). Contrary to the situation in China, where the most widely distributed strain is presumably PoAstV-2 (34).
We identified at least two different PoAstV lineages cocirculating in 15 out of 17 farms (Supplementary Table 1). Thus, the Chilean swine exhibits conditions for PoAstV recombination events. The circulation of multiple PoAstV strains in the same farm has been reported previously in China, Denmark, Slovakia, Thailand, and the USA (31,(35)(36)(37). Co-infection of different lineages in the same individual has also been reported (19,38).
Eleven complete or near-to-complete genomes of PoAstV were obtained, which were used for the phylogenetic analysis ( Table 1). The genomes were obtained from both fecal and oral fluids samples and were recovered from 8 different farms. Figure 1 shows the geographic distribution of the sequences obtained (Figure 1). Interestingly, from farm two, it was possible to obtain the genome of PoAst-2, 4, and 5.
The phylogenetic analysis grouped the Chilean strains into 5 (Figure 2 and Supplementary Figure 1). The seven PoAstV-5 strains are grouped into one monophyletic cluster (97.4% pairwise identity), closely related to strains detected in the USA. Interestingly, the PoAstV-5 were obtained from seven different farms distributed in five geographic regions. The three PoAstV-2 genomes were detected from three different farms in different regions and grouped into two separate sub-clusters. The PoAstV-2/Swine/CHI/FB036/2017 (Metropolitan region) and PoAstV-2/Swine/CHI /CF2673/2017 (Maule region) genomes formed one sub-cluster, while PoAstV-2 /Swine/CHI/FB0148/2017 (Araucania region) grouped with sequences from USA and Japan. Finally, the two PoAstV-4 from two different farms and regions (Metropolitan and Valparaiso region) were phylogenetically distant also related to Japanese strains. The phylogeny demonstrates a genetic relationship between Chilean PoAstV-5 strains, suggesting the same origin for those strains. On the contrary, PoAstV-2 and PoAstV-4 results indicate more diversity even with fewer sequences (For details on genetic distances see Supplementary Tables 2, 3).
Due to the limited number of sequences obtained in this study and the scarce of sequences in GenBank database, conclusions about the origin of the viral strains cannot be elucidated. Indeed, most of the available sequences in GenBank database are from the USA, Japan, and China. Other limitations of the phylogenetic analysis are the incomplete genome coverage in several samples ( Table 1) and the potential errors derived from sequencing methods, as these may alter the phylogenetic tree estimation. This is the first report characterizing the PoAstV sequences circulating in Chile.
This result represents, in turn, the first PoAstV genomes from swine in Latin America. PoAstV studies in Latin America are very scarce, and only two have been published. One study identified PoAstV in healthy pigs from a farm in Brazil (39), and another study conducted in Colombia, which obtained partial PoAstV sequences from diarrheic piglets and humans (40). To date, most of the PoAstV sequences available in GenBank database were obtained in the Northern hemisphere.
Our results support the detection of PoAstV in the Chilean swine population, similar to other observations worldwide. Further studies are needed to understand the relevance of PoAstV to swine health and the evolution and spread of PoAstV locally and globally.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.

ETHICS STATEMENT
The animal study was reviewed and approved by Institutional Animal Care and Use Committees of the Universidad de Chile, protocol number 02-2016. Written informed consent was obtained from the owners for the participation of their animals in this study.

AUTHOR CONTRIBUTIONS
NA, GR-T, and VN: study design and conceptualization. SM and VN: funding and resources. CF and JM: samples collection and processing. CF, JM, and SM: performed the assays. CF, NA, CV, BB, and VN: data analysis. CF, NA, GR-T, and VN: wrote the paper. All authors critically evaluated the paper. All authors contributed to the article and approved the submitted version.

ACKNOWLEDGMENTS
We are grateful to the staff of Animal Virology Lab, Facultad de Ciencias Veterinarias y Pecuarias, Universidad de Chile for all their support in sampling processing, specially to Dr. Felipe Berrios and Valentina Valdes. This study was partially funded by the Animal Virology Laboratory, Faculty of Veterinary and Animal Sciences, Universidad de Chile, and Programa Fondecyt N • 11170877 and N • 1211517 to VN.