2.3. Full-Length Genome Sequencing by Next Generation Sequencing (NGS), Sequence Analysis and Phylogenetic Reconstruction

RNA obtained from concentrated viral particles present in swine stool samples was subjected to NGS. For this, double-stranded cDNA (dscDNA) was generated with random primers using Maxima H Minus Double-Stranded cDNA Synthesis Kit (ThermoFisher Scientific, Waltham, MA, USA). dscDNA was amplified via Multiple Displacement Amplification (MDA) technology using a REPLI-g Mini Kit (Qiagen, Hilden, Germany) followed by purification and quantification using AMPure XP (Beckman Coulter, Brea, CA, USA) and a Qubit fluorometer (Qubit™ DNA-HS Assay kit, Thermo Scientific, Waltham, MA, USA), respectively.

Libraries were constructed from 50 ng of dscDNA using the Nextera DNA Flex Library Preparation kit (Illumina, San Diego, CA, USA) with dual indexing. Control quality libraries were performed on a Fragment Analyzer 5200 system (Agilent Technologies, Santa Clara, CA, USA) using the Standard Sensitivity NGS Analysis Kit (Agilent Technologies, Santa Clara, CA, USA). The library was sequenced on an Illumina MiniSeq Genomic Platform at Facultad de Ciencias (Universidad de la República, Montevideo, Uruguay) using a Mid Output Reagent Cartridge (300-cycles, 150 base-pair paired-end reads) and following standard Illumina protocols.

Raw reads were demultiplexed automatically on the MiniSeq platform. Adapter/quality trimming and filtering were performed with the BBDuk plugin and clean reads were mapped to a previous Uruguayan hepatitis E genome ({"type":"entrez-nucleotide","attrs":{"text":"MW596896","term_id":"2023166707","term_text":"MW596896"}}MW596896) using Geneious mapper (medium-low sensitivity) available in the Geneious Prime 2020.2.1 software (https://www.geneious.com accessed on 3 February 2022).

Sequences were assembled and annotated with SeqMan NGen^® Version 12.0 (DNASTAR. Madison, WI, USA).

Nested-PCR and subsequent sequencing of partial ORF1 and ORF2 [15] were performed to corroborate the NGS data.

The phylogenetic tree was reconstructed based on HEV3 full-length genomes using the neighbor-joining method with Tamura Nei as the substitution method, using Molecular Evolutionary Genetics Analysis (MEGA) v7 software [35]. Reference sequences of subtypes 3a to 3n according to Smith et al. [6] were retrieved from GenBank and included in the analysis. The substitution model that best fitted the data was obtained with MEGA v7 [35]. The robustness of the tree was determined via bootstrap v10 analysis (1000 replicates).

To further characterize the sequences obtained from water sources (wastewater and/or surface water), an additional analysis of HEV-3 subtypes was performed with a 768 bp region within the ORF2. A subset of previously reported human HEV partial-genome sequences from Uruguay were included [15]. The phylogenetic tree for the partial ORF2 region was reconstructed as described for the whole genome dataset.

Nucleotide p-distance matrices between Uruguayan strains and reference sequences of each HEV-3 subtype were constructed with MEGA v7 software for both the partial ORF2 region and full-length genomes.