Porcine reproductive and respiratory syndrome (PRRS) is endemic in most pig-producing countries worldwide and is reported to be among the diseases with the highest economic impact in the modern pig industry [1
]. The disease is mainly characterized by reproductive failure in pregnant sows, respiratory disorders, and growth retardation in piglets [3
]. The causative agent, PRRS virus (PRRSV), belongs to the Arteriviridae
family in the Nidovirales
order. The viral genome consists of a positive-sense, single-stranded RNA molecule of ~15 kb coding for 10 open reading frames (ORFs) [5
]. Genetically, PRRSV strains are divided into PRRSV-1 (mainly predominant in Europe) and PRRSV-2 (predominant in North America and Asia), sharing only 60% nucleotide sequence identity [7
Since the emergence of PRRS, several modified live PRRSV-1 vaccines (MLV1s), i.e., live attenuated viral strains, have been licensed. They are currently frequently used as a tool to reduce the clinical impact of PRRSV infection and to control the within-herd dynamics of infection [9
]. However, the current MLVs have been associated with certain safety concerns such as reversion to virulence [10
]. Importantly, mutation and recombination events confer remarkable plasticity to the RNA genome of PRRSV, including vaccine strains [11
]. Evidence of recombination between PRRSV-1 or PRRSV-2 field strains has been reported several times, especially in Asia where highly pathogenic PRRSV-2 strains circulate [16
]. Regarding recombination between MLV and PRRSV field strains, most of the cases were documented for PRRSV-2 strains [19
], with in some cases, increased pathogenicity of the recombinant strains compared to their parental strains [24
]. More rarely, recombination phenomena between MLV1 strains and PRRSV-1 field strains have been described. These types of recombinant strains were identified in Great Britain by Frossard et al. [26
] and also more recently in China by Chen et al. [27
], who detected a recombinant strain between Unistrain®
PRRS and a PRRSV-1 field strain in a diseased pig. Nevertheless, no evidence of occurrence of recombination events between two MLVs (MLV1 or MLV2) has been described to date.
On a pig farm in France with documented PRRSV infection, where vaccination was successively implemented a few weeks apart, first with Unistrain®
PRRS and then with Porcilis®
PRRS vaccines, we recently identified a recombinant strain between the two commercial vaccine strains. Full-genome sequencing was performed and phylogenetic analysis displayed three recombination events (all located in ORF1 encoding the viral RNA replicase) with Unistrain®
PRRS vaccine as the major parent and Porcilis®
PRRS as the minor parent [28
The objective of the present study was to evaluate and compare the clinical, virological, and transmission parameters of this field recombinant strain derived from two MLV1s to its two parental MLV1 strains.
Recombination is an important genetic mechanism contributing to the evolution of PRRSV and leading to emergence of novel strains, potentially exhibiting increased virulence [14
In the current study, different parameters such as clinical signs, viremia, nasal excretion, and transmission were measured and compared between a recombinant PRRSV strain isolated from the field and both parental MLV1 strains. Even though no significant clinical signs were observed in our SPF pigs, significant differences in viral parameters were observed.
Animals infected with the recombinant strain showed a viremia level 10- to 100-fold higher in comparison with Unistrain®
PRRS or Porcilis®
PRRS vaccine strains, both in inoculated and contact pigs. Of note, the viremia profile in the Rec group was similar to that in the Uni group, i.e., exhibiting an early viral detection, a quickly reached viremia steady state, and high and persistent viral loads. This profile might be linked to the genetic origin of the recombinant strain with 2/3 of its genome originating from the Unistrain®
vaccine (major parent) and 1/3 originating from Porcilis®
vaccine (minor parent) [28
]. From this point of view, the recombinant strain seems to have overpassed the replication capacities of its major parental vaccine strain in order to be more adapted to its natural host.
To explain the difference in the dynamics of viremia between the recombinant strain and the parental vaccine strains, we cannot exclude an effect of the inoculum dose. Indeed even if the three strains had equivalent titer in MARC-145 cells, the recombinant strain displayed a higher titer in PAMs compared to MARC-145 (104.2
/mL in MARC-145 and 106
/mL in PAMs). This could have resulted in the inoculation of a higher amount of infectious particles for the recombinant strain (compared to the vaccine strains) during the experiment. Nevertheless, even if a higher inoculum dose had been used for the recombinant strain, this should not have substantially impacted on the results of the study. Indeed previous studies have shown that for PRRSV, the viremia level is mostly independent to the inoculum dose. Loving et al., [32
], demonstrated that initial PRRSV dose did not correlate with PRRSV viremia kinetics during the acute stage of infection when using a low (102
) or high (106
) dose of PRRSV inoculum. These results are supported by those of Haiwick et al. [33
], who showed that, for a challenge dose of 103
/mL, pigs had similar post-challenge viremia profiles. Nevertheless, in groups exposed to lower inoculum doses, less than 101.5
/mL, viremia dynamics were altered with a delayed viremia peak.
The limited effect of the inoculum dose is also sustained by the results from the contact pigs of our study. Indeed contact pigs from the Rec group had the same level of viremia as Rec inoculated pigs while these contact pigs probably get infected by the nasal route with a lower dose of PRRSV than those received by inoculated pigs. Altogether these results support the hypothesis that the viremia level may be strongly linked to the fitness of the PRRSV strain but not to the inoculum dose.
The increased level of viral particles excreted via the respiratory tract is consistent with the observed highest viremia level of the recombinant strain compared with the other tested strains. This is also in line with the higher PRRSV viral load detected in tonsils from pigs in the Rec and Uni groups compared to the Porci group. Thus, the enhanced nasal excretion capacities of the recombinant strain in inoculated pigs can explain the stronger transmission of the recombinant PRRSV strain to contact pigs in the early stages of infection, compared to the vaccine strains. Using the same methodology, we previously determined the transmission parameters for a PRRSV field strain from France [31
]. Interestingly, the transmission parameters of the recombinant strain were very close to those determined for this PRRSV-1 wild-type strain. Regarding the transmission of Unistrain®
PRRS and Porcilis®
PRRS vaccine strains to contact pigs, we found some unexpected results. Indeed, while viremia and nasal shedding were lower for the Porci group compared to the Uni group, contact pigs from the Porci group became infected earlier. Our results are therefore not consistent with those of Martinez-Lobo et al., [34
], who found faster transmission of the Unistrain®
PRRS vaccine strain compared to Porcilis®
PRRS vaccine strain. To explain this difference in terms of viral transmission between both studies, it has been suggested that in vivo replication of attenuated vaccine strains may lead to mutant selection [35
]. In our study, during the adaptation process of vaccine strains in the swine host, mutations might have conferred increased transmission capacities to Porcilis®
PRRS but not to Unistrain®
High transmission capacities might confer long-term persistence to the recombinant strain in the pig herd. Indeed, two years after the first isolation of the recombinant strain, we could confirm that the recombinant PRRSV strain was still circulating on the farm. The 2016 recombinant strain sequence, with 1.4% genetic divergence (due to additional mutations over the years) compared to the recombinant strain from 2014, is in line with previous data (from field experience and experimental infections) that described a mutation rate of about 0.5% to 1% per year [36
Even though three recombination events were clearly identified in the field strain under study, recombination may not be the only or main reason resulting in increased transmission characteristics. Indeed, the recombinant strain was isolated from the field in December 2014 more than one year after two successive vaccinations in a limited period of time, which are most probably the source of the recombination events between the two MLV1 vaccines. The recombinant strain might have increased its replication capacities thanks to selection of more adapted virus variants following multiple pig passages in the farm. This phenomenon has already been described for MLV2 strains [38
]. Mengeling et al. [39
] suggest that vaccine strains could evolve during in vivo passages, enabling them to replicate more efficiently in their natural target cells. They also found that genetic reversion to virulence is a gradual process, occurring with a prolonged passage time of the strain in the field. Given our results and data from the literature, one of the major assumptions is that the adaptation process of attenuated strains after multiple passages in the pig might have changed the tropism of the virus and increased its replicative capacity. In our laboratory (data not shown), Porcilis®
PRRS and Unistrain®
PRRS vaccine strains were not able to replicate in PAMs, suggesting that they might have lost their ability to replicate in their natural target cells during the attenuation process in cell lines such as MARC-145. However, the recombinant field strain from 2014 replicated both in PAMs and MARC-145 cells, while the recombinant strain isolated in 2016 could only replicate in PAMs and lost its ability to multiply in MARC-145 cells. In fact, the replication capacities of this PRRSV strain of vaccine origin seem to have evolved after multiple in vivo passages, recovering the ability to replicate in vitro in PAMs, the natural target cells.
In this study, we characterized in vivo a field recombinant PRRSV-1 strain derived from two MLV1 strains. Although no significant clinical signs were observed, this recombinant strain demonstrated increased excretion and transmission capacities compared to parental vaccine strains. The viral fitness of this recombinant strain might reveal its safety level since propagation and persistence in the field might increase the risk of reversion to virulence due to genetic evolution over several years. Because vaccine safety in the field is an issue of particular concern, measures should be implemented to avoid any recombination phenomenon of PRRSV vaccine strains under field conditions, i.e., a more careful use of MLVs for PRRSV prevention and control.