Porcine Reproductive and Respiratory Syndrome (PRRS) is a disease of major economic impact worldwide. The etiologic agent of this disease is the PRRS virus (PRRSV). Increasing evidence suggest that microevolution within a coexisting quasispecies population can give rise to high sequence heterogeneity in PRRSV.
We developed a pipeline based on the ultra-deep next generation sequencing approach to first construct the complete genome of a European PRRSV, strain Olot/9, cultured on macrophages and then capture the rare variants representative of the mixed quasispecies population. Olot/91 differs from the reference Lelystad strain by about 5% and a total of 88 variants, with frequencies as low as 1%, were detected in the mixed population. These variants included 16 non-synonymous variants concentrated in the genes encoding structural and nonstructural proteins; including Glycoprotein 2a and 5.
Using an ultra-deep sequencing methodology, the complete genome of Olot/91 was constructed without any prior knowledge of the sequence. Rare variants that constitute minor fractions of the heterogeneous PRRSV population could successfully be detected to allow further exploration of microevolutionary events.
PRRSVMicroevolutionVariant spectraUltra-deep next generation sequencing
Porcine Reproductive and Respiratory Syndrome virus (PRRSV) is the causative agent of a significant disease of the domestic pig (Sus scrofa) with global consequences. The severity of PRRSV infection ranges from subclinical to lethal and it affects pigs in both growing and reproductive stages. The virus has a positive-sense 15 kb RNA genome and its genetic diversity has been well characterised within and between European and North American strains . Extensive viral genetic heterogeneity may have contributed towards the observed variations between PRRSV isolates and clones in term of virulence, interactions with the immune system, and antigenic properties of viral proteins. Such a broad diversity indeed poses serious challenges to diagnostics and control measures.
Most previous studies of PRRSV genetic diversity have been restricted to the ORF5 and ORF7 sequences of type 2 “North American-like” viruses that also include the Asian variants. Only 14 of the 303 completed PRRSV genomes in Genbank belong to genotype 1. Furthermore, studies have shown that PRRSV mutates rapidly and multiple intra-strain variants can coexist in individually infected pigs . The extensive genetic diversity displayed by PRRSV and other RNA viruses such as HIV and influenza reflects the error prone nature of RNA polymerases, which lack a proofreading function [3, 4].
To identify PRRSV quasispecies, previous studies have employed conventional methodologies including reverse-transcription, PCR, cloning and Sanger sequencing of a subset of PRRSV structural and non-structural proteins [2, 5]. More recently next-generation sequencing (NGS) of fragments generated by long range RT-PCR has been used to characterise multiple PRRSV genomes . However, this approach relies upon prior knowledge of the target sequence and the assumption that the PCR primer binding sites are non-variable. Here we describe an approach which requires no prior knowledge of the target sequences and which should enable the detection of low-frequency nucleotide variants and hence provides a snapshot of the microevolution in the entire viral population.
We analysed the intra-strain sequence diversity of low passage PRRSV Olot/91 strain, passaged exclusively on primary porcine alveolar macrophages (PAM). First reported in Spain in 1991, Olot/91 is the parent strain of the commercial Suvaxyn PRRSV inactivated vaccine which is used in Spain and Portugal. Only a partial sequence of 3,383 nt [GenBank:X92942], that covers ORFs 2-7 and the 3'-UTR of this strain has previously been published .
To reduce potential complications with sub-genomic RNA, the virus was grown in porcine alveolar macrophages for 5 days and viral particles were pelleted from cell culture supernatant through a 30% sucrose cushion. Viral RNA was then prepared using BioSprint 96 DNA Blood kit (Qiagen) and sequencing libraries produced using the Illumina TruSeq RNA sequencing library preparation kit, with a modification eliminating the initial mRNA isolation steps. A total of 7.56 million 225 bp reads were generated on a Illumina MiSeq machine. They were analysed following a 2-step strategy whereby the consensus sequence of the major Olot/91 strain was first obtained and then used to call rare variants present in the evolving PRRSV viral population (Figure 1). To account for potential contamination with host DNA, the raw reads were initially filtered against the pig genome (Sscrofa10.2 ) using the program Best Match Tagger (v1.1.0)  which removed about 88% of the reads. The remaining reads then underwent a stringent quality filtering whereby the first 5 bp of the 5' ends were trimmed to remove any remnant adapter sequences; reads with ambiguous bases and Phred scores lower than 20 were discarded using Sickle (v1.2)  and the 3' ends were further scanned for run-through adapter sequences using Scythe (v0.991b) . The final high quality reads (4.3%) were next either mapped against the reference Lelystad (LV) genome [Genbank:M96262]  using BWA (v0.6.2) /Samtools (v0.1.18)  or assembled de novo with Velvet (v1.2.9) . Variants used to call the 15,111 nt consensus sequence in the mapped major Olot/91 strain were determined as those having allele frequencies higher than 0.7 by the program Lofreq (v0.6.0) . A total of 742 single nucleotide variants were identified that differed from the reference LV strain (Figure 2, Additional file 1: Table S1). Of these, 180 resulted in non-synonymous changes in coding sequences. No insertions or deletions were found. For the de novo assembly, contigs from two kmer values of 161 and 195 were merged with Lasergene's SeqMan Pro (v10.1) to yield a single contig of 15,082 nt. Both the mapped consensus sequence and the de novo assembled contig were aligned against each other to ensure the accuracy of the genome. These sequences were exactly identical to each other and the full length Olot/91 genome has now been deposited in Genbank with the accession number KF203132. This full length Olot/91 contains 8 SNPs and a 3 nt insertion when compared with the published partial Olot/91 sequence. However, this PAM-passaged virus only shares 96.8% similarity with the MARC-145 cell-adapted PRRSV Olot/91 strain (Genbank:KC862570) . The differences may represent the molecular basis for the adaption to growth on MARC-145 cells which are of African Green Monkey origin .
The same set of cleaned reads was then remapped as previously onto the newly generated Olot/91 genome to an average depth of approximately 530×. Eighty-eight low-frequency single-nucleotide variants representing the genotypic makeup of the heterogeneous PRRSV population were identified, of which 16 represented non-synonymous changes in coding sequences (Figure 2, Additional file 1: Table S1). Among the 22 low frequency variants that overlap with the major Olot/91 versus reference LV strain variants, 17 have evolved into genotypes identical to the reference strain. The frequency of intra-strain variants detectable in this study ranged from 1 to 20% with each supported by a minimum of 3 reads. A subset of these low-frequency variants was validated using two alternative methods; Sanger sequencing and real-time PCR. Reverse transcription was performed as described previously . Briefly, one microgram of the Olot/91 RNA was reverse transcribed using a TaqMan kit (Applied Biosystems, Foster City, CA). For Sanger sequencing, PCR products encompassing the variants were sequenced in both directions using BigDye terminator v3.1 Cycle Sequencing kit (Applied Biosystems). Variants with frequencies as low as 3% could be successfully identified by the overlapping peaks (Figure 3A and Additional file 2: Figure S1). Major and minor allele specific real-time PCRs employing Platinum SYBR Green PCR SuperMix UDG (Invitrogen, Paisley, UK) were run on a Stratagene MX3000P (Stratagene). Samples were run in triplicate, beta-actin was utilized as the housekeeping gene and results were calculated as described previously . Again, the main Olot/91 sequence and the variants present in the related mixed population could both be successfully detected (Figure 3B).
While variants in the major Olot/91 strain do congregate on known highly hypervariable B- and T-cell epitopes of type I PRRSV (Figure 2 and Additional file 3: Figure S2) [19–23], most of the variants representative of the probable quasispecies population map to the nsp12 and gp2a genes of ORF1b and ORF2 respectively (Figure 2); suggesting a higher microevolutionary rate at this region under the investigated culture conditions. In addition, the predicted short signal peptide of GP2a was found to harbour a high concentration of non-synonymous variants and a potential N-glycosylation site was created from a residue change (S37N) on GP5. Further analysis, like molecular modelling, may be necessary to decipher the potential impacts of these evolving variants have on such functions as interactions between these proteins and other viral or host cellular proteins.
Using ultra-deep NGS we have constructed the complete sequence of the PRRSV Olot/91 genome cultured in their natural host - porcine alveolar macrophages - and identified single-nucleotide variants present in the associated viral population. Further investigation using this methodology will help to establish if a link exists between the microevolutionary dynamics and pathogenesis of PRRSV viral strains. However, it is also important to note that the exact nature of the pathogenesis cannot be truly identified until the viral haplotypes within the quasispecies population can be reconstructed with confidence .
We are indebted to Lise K. Kvisgaarda and Lars E. Larsen for kindly providing the sequence of the Marc 145-adapted Olot/91 strain. TAA, ALA and ZHL were supported by UK Biotechnology and Biological Sciences Research Council (BBRSC) Institute Strategic Programme Grants (BB/J004235/1). ADW was supported by EC FP7 PoRRScon Grant (Agreement number 245141). Edinburgh Genomics is funded by BBSRC National Capacity Grants (BB/J004243/1). We are also grateful for support from the COST Action FA0902.
The Roslin Institute and Royal (Dick) School of Veterinary Studies, University of Edinburgh, Easter Bush
Schommer SK, Kleiboeker SB: Use of a PRRSV infectious clone to evaluate in vitro quasispecies evolution.Adv in Exp Med and Biol 2006, 581:435–438.View Article
Kvisgaard LK, Hjulsager CK, Fahnøe U, Breum SØ, Ait-Ali T, Larsen LE: A fast and robust method for full genome sequencing of Porcine Reproductive and Respiratory Syndrome Virus (PRRSV) Type 1 and Type 2.J Virol Methods 2013, 193:697–705.PubMedView Article
Urniza A, Climent I, Duran J, Corte´s E, Sarraseca J, Casal JI, Vela C: Baculovirus expression of proteins of porcine reproductive and respiratory syndrome virus strain olot/91. Involvement of ORF3 and ORF5 proteins in protection.Virus Genes 1997, 14:19–29.PubMedView Article
Groenen MA, Archibald AL, Uenishi H, Tuggle CK, Takeuchi Y, Rothschild MF, Rogel-Gaillard C, Park C, Milan D, Megens HJ, Li S, Larkin DM, Kim H, Frantz LA, Caccamo M, Ahn H, Aken BL, Anselmo A, Anthon C, Auvil L, Badaoui B, Beattie CW, Bendixen C, Berman D, Blecha F, Blomberg J, Bolund L, Bosse M, Botti S, Bujie Z, et al.: Analyses of pig genomes provide insight into porcine demography and evolution.Nature 2012, 491:393–398.PubMed CentralPubMedView Article
BMTagger: Best match tagger for removing human reads from metagenomics datasets. ftp://ftp.ncbi.nlm.nih.gov/pub/agarwala/bmtagger/
Meulenberg JJM, Hulst MM, de Meijer EJ, Moonen PLJM, den Besten A, de Kluyver EP, Wensvoort G, Moormann RJM: Lelystad Virus, the Causative Agent of Porcine Epidemic Abortion and Respiratory Syndrome (PEARS), Is Related to LDV and EAV.Virology 1993, 192:62–72.PubMedView Article
Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, Marth G, Abecasis G, Durbin R, Genome Project Data Processing Subgroup: The sequence alignment/Map format and SAMtools.Bioinformatics 2078–2079, 2009:25.
Wilm A, Aw PP, Bertrand D, Yeo GH, Ong SH, Wong CH, Khor CC, Petric R, Hibberd ML, Nagarajan N: LoFreq: a sequence-quality aware, ultra-sensitive variant caller for uncovering cell-population heterogeneity from high-throughput sequencing datasets.Nucleic Acids Res 2012, 40:11189–11201.PubMed CentralPubMedView Article
Delrue I, Van Gorp H, Van Doorsselaere J, Delputte PL, Nauwynck HJ: Susceptible cell lines for the production of porcine reproductive and respiratory syndrome virus by stable transfection of sialoadhesin and CD163.BMC Biotechnol 2010, 10:48.PubMed CentralPubMedView Article
Ait-Ali T, Wilson AD, Carre W, Westcott DG, Frossard JP, Mellencamp MA, Mouzaki D, Matika O, Waddington D, Drew TW, Bishop SC, Archibald AL: Host inhibits replication of European porcine reproductive and respiratory syndrome virus in macrophages by altering differential regulation of type-I interferon transcriptional response.Immunogenetics 2011, 63:437–448.PubMedView Article
Díaz I, Pujols J, Ganges L, Gimeno M, Darwich L, Domingo M, Mateu E: In silicoprediction andex vivoevaluation of potential T-cell epitopes in glycoproteins 4 and 5 and nucleocapsid protein of genotype-I (European) of porcine reproductive and respiratory syndrome virus.Vaccine 2009, 27:5603–5611.PubMedView Article
Oleksiewicz MB, Bøtner A, Normann P: Porcine B-cells recognize epitopes that are conserved between the structural proteins of American- and European-type porcine reproductive and respiratory syndrome virus.J Gen Virol 2002, 83:1407–1418.PubMed
Costers S, Vanhee M, Van Breedam W, Van Doorsselaere J, Geldhof M, Nauwynck HJ: GP4-specific neutralizing antibodies might be a driving force in PRRSV evolution.Virus Res 2010, 154:104–113.PubMedView Article
Parida R, Choi I-S, Peterson DA, Pattnaik AK, Laegreid W, Zuckermann FA, Osorio FA: Location of T-cell epitopes in nonstructural proteins 9 and 10 of type-II porcine reproductive and respiratory syndrome virus.Virus Res 2012, 169:13–21.PubMedView Article
Liao YC, Lin HH, Lin CH, Chung WB: Identification of cytotoxic T lymphocyte epitopes on swine viruses: multi-epitope design for universal T cell vaccine.PLoS One 2013, 8:e84443.PubMed CentralPubMedView Article
Schirmer M, Sloan WT, Quince C: Benchmarking of viral haplotype reconstruction programmes: an overview of the capacities and limitations of currently available programmes.Brief Bioinform 2012. doi:10.1093/bib/bbs081
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.