Recombination occurs in most RNA viruses and has a major impact on their diversification and evolution. More significantly, recombination has also been associated with the emergence of new viruses, increases in virulence and pathogenesis, and the evasion of host immune responses [1Perez-Losada M, Arenas M, Galan JC, et al. Recombination in viruses: Mechanisms, methods of study, and evolutionary consequences. Infection, genetics and evolution journal of molecular epidemiology and evolutionary genetics in infectious diseases 30C 2015; 296-307., 2Simon-Loriere E, Holmes EC. Why do RNA viruses recombine? Nat Rev Microbiol 2011; 9: 617-26.]. Porcine reproductive and respiratory syndrome virus (PRRSV) is the most costly swine virus worldwide, which is an enveloped, positive-sense, single-stranded RNA virus [3Lunney JK, Fang Y, Ladinig A, et al. Porcine Reproductive and Respiratory Syndrome Virus (PRRSV): Pathogenesis and Interaction with the Immune System. Annu Rev Anim Biosci 2016; 4: 129-54.]. Recombination in RNA viruses is a process to form chimeric offspring from parental genomes of mixed origin, which requires co-infection or super-infection of a cell with at least two viruses. RNA recombination is based on RNA-dependent RNA polymerase (RdRp) template switches that may occur during either genomic RNA replication or subgenomic mRNA synthesis [4Pasternak AO, Spaan WJ, Snijder EJ. Regulation of relative abundance of arterivirus subgenomic mRNAs. J Virol 2004; 78: 8102-13.]. The factors that influence template switching include RNA secondary structures and sequence similarity between donor and acceptor templates [5Galetto R, Giacomoni V, Veron M, Negroni M. Dissection of a circumscribed recombination hot spot in HIV-1 after a single infectious cycle. J Biol Chem 2006; 281: 2711-20., 6Zhang J, Temin HM. Retrovirus recombination depends on the length of sequence identity and is not error prone. J Virol 1994; 68: 2409-14.].

Recombination is a common phenomenon between PRRSV isolates in the field [7Martin-Valls GE, Kvisgaard LK, Tello M, et al. Analysis of ORF5 and full-length genome sequences of porcine reproductive and respiratory syndrome virus isolates of genotypes 1 and 2 retrieved worldwide provides evidence that recombination is a common phenomenon and may produce mosaic isolates. J Virol 2014; 88: 3170-81.]. The occurrence of PRRSV recombination was first suggested by the phylogenetic analysis of field isolates [8Kapur V, Elam MR, Pawlovich TM, Murtaugh MP. Genetic variation in porcine reproductive and respiratory syndrome virus isolates in the midwestern United States. J Gen Virol 1996; 77(Pt 6): 1271-6.]. Among PRRSV isolates, intra-genotype recombination is frequent but no inter-genotype recombination has been reported [9Chen N, Yu X, Wang L, et al. Two natural recombinant highly pathogenic porcine reproductive and respiratory syndrome viruses with different pathogenicities. Virus Genes 2013; 46: 473-8.-16Lu WH, Tun HM, Sun BL, et al. Re-emerging of porcine respiratory and reproductive syndrome virus (lineage 3) and increased pathogenicity after genomic recombination with vaccine variant. Vet Microbiol 2015; 175: 332-40.]. PRRSV could undergo homologous recombination with the frequency from <2% up to 10% in vitro and ~38% (133/352) in vivo [12Yuan S, Nelsen CJ, Murtaugh MP, Schmitt BJ, Faaberg KS. Recombination between North American strains of porcine reproductive and respiratory syndrome virus. Virus Res 1999; 61: 87-98., 14Liu D, Zhou R, Zhang J, et al. Recombination analyses between two strains of porcine reproductive and respiratory syndrome virus in vivo. Virus Res 2011; 155: 473-86.]. The most widely accepted model of PRRSV recombination is copy-choice model [2Simon-Loriere E, Holmes EC. Why do RNA viruses recombine? Nat Rev Microbiol 2011; 9: 617-26., 17Lai MM. RNA recombination in animal and plant viruses. Microbiol Rev 1992; 56: 61-79.].

Conventional methods to detect PRRSV recombination are based on PCR and sequencing followed by data analysis using recombination detection programs, such as RDP and SimPlot [18Lole KS, Bollinger RC, Paranjape RS, et al. Full-length human immunodeficiency virus type 1 genomes from subtype C-infected seroconverters in India, with evidence of intersubtype recombination. J Virol 1999; 73: 152-60., 19Martin DP, Lemey P, Lott M, Moulton V, Posada D, Lefeuvre P. RDP3: A flexible and fast computer program for analyzing recombination. Bioinformatics 2010; 26: 2462-3.]. A major limitation of these PCR based methods is that all genomic, subgenomic and defective RNAs could serve as templates for PCR amplification. Therefore, the viability of the recombinants identified by them is unknown. To address this limitation, we developed a new in vitro system targeting at recombination events that present in viable offspring [20Chand R. Study of Recombination in PRRSV using a novel in-vitro system. 2013. PhD Thesis. Kansas State University, Manhattan, KS.]. This reverse genetics system uses green fluorescence protein (GFP)- and enhanced GFP (EGFP)-expressing PRRSV infectious clones to study recombination in the chimeric genes that are nonessential for virus replication. Successful recombination is evidenced by generating a viable fluorescent virus from the co-transfection of a non-fluorescent viable virus with a mutation in EGFP (P129-EGFP-97C) and a fluorescent defective virus (P129-GFPm-d (2-6)) (Fig. 1). However, what types of cross-over events occurred to produce viable viruses in this system are not clarified yet. Here we took advantage of high throughput sequencing to assess the locations of all cross-over events between EGFP and GFPm genes and explore the cross-over patterns that are potentially utilized to produce viable recombined viruses.

Fig. (1) Recombination between two PRRSV infectious clones P129-EGFP-97C and P129-GFPm-d (2-6). P129-EGFP-97C is a fully functional non-fluorescent virus and P129-GFPm-d (2-6) is a fluorescent defective virus lacking ORF2-6. Successful recombination between the two parental viruses is evidenced by producing viable fluorescent progeny viruses.

By combined utilization of the in vitro reverse genetics system and high throughput sequencing, we inferred recombination events between inserted fluorescent genes in PRRSV infectious clones. Based on thousands of sequences from viable progeny viruses, we found that the nucleotide identities changed between the variable and conserved regions of EGFP/GFPm genes, indicating that the cross-over events occurred in the conserved regions. The advantages of this new method in measuring recombination include: 1) This reporter system targets at detecting viable recombinant viruses. 2) The system tests recombination events in nonessential gene without affecting the virus replication. 3) The co-utilization of in vitro reverse genetics system and deep sequencing reveals all types of cross-over events which occurred in a target gene to produce viable recombinants. Similar reporter systems are generally used to measure the virus recombination [27Levy DN, Aldrovandi GM, Kutsch O, Shaw GM. Dynamics of HIV-1 recombination in its natural target cells. Proc Natl Acad Sci USA 2004; 101: 4204-9., 28Runckel C, Westesson O, Andino R, DeRisi JL. Identification and manipulation of the molecular determinants influencing poliovirus recombination. PLoS Pathog 2013; 9: e1003164.]. The disadvantage of using 454 pyrosequencing in our system is the introduction of errors (error rates ranged from 0.04-0.66%) [29Di Giallonardo F, Zagordi O, Duport Y, et al. Next-generation sequencing of HIV-1 RNA genomes: determination of error rates and minimizing artificial recombination. PLoS One 2013; 8: e74249.], which resulted in the variations of the identities in a same conserved region (Table 2); however, it did not interrupt the evaluation of the obviously changes between conserved regions. In addition, this system was developed to analyze the requirements for generating viable recombinants but not for detecting all recombination events between these two infectious clones, therefore, it could not detect the recombination events in the PRRSV genomes.

The first and second conserved regions (166-288nt and 290-455nt) are only separated by a C₂₈₉T mutation. And the variation frequencies in two flanked variable regions (21-165nt and 456-651nt) are nearly identical (~37% and ~38%), which arouse suspicion that the lower variation frequency of C₂₈₉T mutation (20%) might have resulted from mutation or inaccurate sequencing rather than from recombination. However, our results provide strong support that the changes in sequence identity are due to cross-over events. First, the 20% mutation rate of C₂₈₉T results from 15531 reads (Table 2). Second, the sequencing result of the sample (80% nucleotides at position 289 are cytosine, which is identical to the fluorescent gene GFPm) is consistent with the cell sorting result (about 80% sorted cells express fluorescence). Third, the C₂₈₉T mutation is highly stable during serial passage in Marc-145 cells [20Chand R. Study of Recombination in PRRSV using a novel in-vitro system. 2013. PhD Thesis. Kansas State University, Manhattan, KS.].

The third conserved region (652-671nt) shows to be a cross-over hotspot and has higher cross-over rate than the above two conserved regions (Fig. 2). The result is consistent with Sanger sequencing results that all six individual clones of the whole fluorescent gene from the same co-infection sample had cross-over at the 20bp-conserved region [20Chand R. Study of Recombination in PRRSV using a novel in-vitro system. 2013. PhD Thesis. Kansas State University, Manhattan, KS.]. Although both RNA secondary structure and the length of sequence identity are crucial molecular determinants for recombination [5Galetto R, Giacomoni V, Veron M, Negroni M. Dissection of a circumscribed recombination hot spot in HIV-1 after a single infectious cycle. J Biol Chem 2006; 281: 2711-20., 6Zhang J, Temin HM. Retrovirus recombination depends on the length of sequence identity and is not error prone. J Virol 1994; 68: 2409-14., 28Runckel C, Westesson O, Andino R, DeRisi JL. Identification and manipulation of the molecular determinants influencing poliovirus recombination. PLoS Pathog 2013; 9: e1003164.], RNA secondary structure seems to play a more important role in this case considering that the length of the cross-over hotspot (20bp) is shorter than the other two cross-over regions (123bp and 166bp). In addition, the 20bp-conserved region encoded “117-HMVLLE-222” is located at the junction region of β-sheet strand 11 linking to the loop. Glu222 at the junction point has alternative conformations [30Arpino JA, Rizkallah PJ, Jones DD. Crystal structure of enhanced green fluorescent protein to 1.35 A resolution reveals alternative conformations for Glu222. PLoS One 2012; 7: e47132.]. Previous studies implied that recombination events occurred more frequency at the transcriptional pausing sites or polymerase-binding motifs [5Galetto R, Giacomoni V, Veron M, Negroni M. Dissection of a circumscribed recombination hot spot in HIV-1 after a single infectious cycle. J Biol Chem 2006; 281: 2711-20., 17Lai MM. RNA recombination in animal and plant viruses. Microbiol Rev 1992; 56: 61-79.]. Therefore, the mechanisms responsible for the hotspot of the 20bp-conserved region are probably associated with RNA secondary structure. Notably, recombination events were not detected in the 20bp region when using infectious clones containing EGFP-97C and GFP, sharing 83% nucleotide identity [20Chand R. Study of Recombination in PRRSV using a novel in-vitro system. 2013. PhD Thesis. Kansas State University, Manhattan, KS.]. Recombination events occurred between EGFP-97C and GFPm, which share 91% nucleotide identity, suggesting that the overall nucleotide identity may be also essential to produce recombinants. And another explanation is that the relative positions of the EGFP sequence inserted in GFPm gene may be important [6Zhang J, Temin HM. Retrovirus recombination depends on the length of sequence identity and is not error prone. J Virol 1994; 68: 2409-14.].

Four recombination patterns might be used to produce viable recombinants during co-infection (Fig. 3). The exact rate of each pattern could not be determined due to the limitation of the read length (~400bp) of the 454 sequencing method used in this study, but two cross-over events were confirmed by sequence alignments (Fig. 4). The results indicated that different cross-over patterns were utilized and distinct recombination events occurred simultaneously in the viral quasispecies.

This study presented the detailed cross-over map of the chimeric genes in two PRRSV infectious clones by high throughput sequencing. Pattern analyses and sequence alignments indicated that different patterns may be utilized and distinct cross-over events may occur simultaneously. Our results also showed that the 20bp-conserved region is likely a cross-over hotspot, suggesting that RNA secondary structure may play a more important role for recombination in the case. This new in vitro reverse genetics system accompanied with deep sequencing creates a viable platform to study different types of recombination events and contributes to understanding of the requirements for recombination.