Martín‑Valls et al. Porcine Health Management (2023) 9:1 https://doi.org/10.1186/s40813-022-00298-3 RESEARCH Open Access Introduction of a PRRSV‑1 strain of increased virulence in a pig production structure in Spain: virus evolution and impact on production Gerard E. Martín‑Valls1†, Martí Cortey1†, Alberto Allepuz1, Francesc Illas2, Montserrat Tello1 and Enric Mateu1* Abstract Background: A strain of Porcine reproductive and respiratory syndrome virus (PRRSV), showing characteristics of enhanced virulence, affected a pyramidal production system from Spain with 7600 sows in 4 genetic nuclei and 13,000 sows in multipliers. Different PRRSV strains circulating in this production system from December 2020 to Octo‑ ber 2021 were detected and sequenced. The spread of each isolate was examined and their impact on health and production in three of the affected farms was evaluated. Results: The newly emerged PRRSV isolate with enhanced virulence entered the system before the onset of the study (January 2020) and afterwards four significantly different clades were detected during the study period in differ‑ ent farms, probably because of independent introduction events. The diversification of the enhanced virulence strain was higher for those clades (substitution rates up to 1.1% nucleotides/year) compared to other PRRSV strains present in the production system (up to 0.17%), suggesting a faster spread and adaptation. The impact of the infection in the first affected farm was dramatic, with an average abortion rate above 27% during 17 weeks before returning to the baseline production. Fertile sow mortality reached 6.5% for 39 weeks. In two farms infected later by other clades of this enhanced virulence strain, the impact was less acute; despite the fact that for parameters such as the proportion of stillbirths or mummies, more than 10 months were needed to recover pre-outbreak values. In the examined nurser‑ ies, mortalities reached peaks between 28 and 50% and several months were needed to return to normality. Conclusion: Introduction of a PRRSV strain of enhanced virulence in a production system where several farms were previously positive for other PRRSV strains, resulted in a fast spread such as would be observed in naïve farms. The productive and health impact was very high taking several months to return to normality. Background industry, of which weaners and growers accounted for Porcine reproductive and respiratory syndrome (PRRS) 55% of the total cost. In Europe, Nieuwenhuis et al. [20] is one of the most economically important diseases of estimated an average loss of € 126 per sow (€ 59–379) pig industry. According to Holtkamp et al. [10] the dis- during an outbreak period. More recently, Nathues et al. ease cost about U$ 664 million annually to the American [19] estimated a cost of 126.79 €/sow/year in a farm with slight reproductive problems, and 3.77 €/fattener in a farm with slight respiratory disease. † Gerard E. Martín-Valls and Martí Cortey contributed equally to this work Losses caused by PRRS result from reproductive disor- *Correspondence: enric.mateu@uab.cat ders, such as abortions and stillbirths, but also from an 1 Dept. Sanitat i Anatomia Animals, Facultat de Veterinària, Universitat increased proportion of weak-born piglets and increased Autònoma de Barcelona, Travessera dels Turons s/n, Campus UAB, mortality in new-borns [2, 6, 25]. Besides, PRRSV is one 08193 Cerdanyola del Vallès, Spain Full list of author information is available at the end of the article major component of the porcine respiratory disease © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. The Creative Commons Public Domain Dedication waiver (http://​creat​iveco​ mmons.​org/​publi​cdoma​in/​zero/1.​0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Martín‑Valls et al. Porcine Health Management (2023) 9:1 Page 2 of 11 complex, and the presence of PRRSV circulating in the clade (c4) appeared in May 2021 in the nucleus Nu4 and nurseries commonly leads to increased mortality when then spread to M7 in June that year. The source of infec- combined with other bacterial or viral agents [4, 9, 28]. In tion for the original outbreak of each clade could not be many affected nurseries, the presence of PRRSV results identified. Subsequent outbreaks in other farms were in increased medication costs as well [27]. Indirect costs, likely the result of the movement of animals to nurseries. arising from the distortion caused by the disease on the All other farms remained with its previous PRRSV status normal production flow together with control costs, are as of December 2020. Additional file 1 shows the chro- also important. nology of positive PRRSV detections in the system. PRRS is caused by Betaarterivirus suid 1 and 2 viruses, namely PRRS virus (PRRSV) 1 and 2. One of the impor- tant features of PRRSV is its high genetic and antigenic Evolution of the newly emerged PRRSV isolate diversity, which is constantly evolving [18]. PRRSV-1, the Next, we analyzed the diversification of the newly predominant species in Europe, is subclassified in 4 sub- emerged highly virulent PRRSV isolate in comparison types, each one of which include several clades [22]. In to the isolates previously present in the system (Fig. 1). Western Europe, subtype 1 (PRRSV-1.1) is predominant. The substitution rates (expressed as percentage of yearly This genetic diversity also translates to a variable viru- changes in the whole genome) for the newly emerged lence. In general, PRRSV-1 was considered less virulent strain ranged from 0.33 to 1.11%, compared to 0.03%- than PRRSV-2 [17], a notion reinforced by the emergence 0.17% for the previous PRRSV strains present in the of an extremely virulent strain in China some years ago system. In contrast, the mean nucleotide diversity was [26]. However, the discovery of strains like Lena in Bela- similar between the newly emerged and the previous rus (PRRSV-1.3) [11], or PR40 (PRRSV-1.1) in Italy [5], strains (0.21–0.93% versus 0.37–0.63% for the pre-exist- demonstrated the existence of highly virulent PRRSV-1 ing ones). In general, the highest rates within the genome strains. were observed for segments nsp2, ORF3, ORF4, ORF5, Starting in 2020, severe PRRSV outbreaks charac- with nsp9 and nsp10 showing the lowest. Of note, the terized by high abortion rates and increased mortality pattern was not different when newly emerged or pre- rates in weaners, were increasingly reported in North- vious strains were compared. The highly virulent iso- eastern Spain (NE Spain) [16]. Isolation and sequencing late was identified as a mosaic, with a backbone derived of the virus in several outbreaks [15] showed that they from a strain related to the isolate PR40 [5] that incor- were caused by a mosaic PRRSV strain, derived from the porated four recombinant segments. During the study, PR40 strain reported by Canelli et al. [5], after drifting an additional recombination event involving nsp9 was and undergoing several recombination events with other detected in all isolates of c2. A new recombinant isolate PRRSV-1.1 isolates. was detected in nursery N5 in July 2021. In this case, the TRAIN®), with a recombinant segment in ORF1b (region The aim of the present study was to assess the evolu- virus contained the backbone of a vaccine strain (UNIS- tion of this PRRSV strain with enhanced virulence after its introduction in a production system. The productive encoding nsp9) derived from the highly virulent PRRSV impact of the strain in selected breeding sites and nurser- isolate. Additional file 2 shows the recombination events ies was also examined. found for the newly emerged strain. Results Impact of the infection by the newly emerged PRRSV virus Emergence of the highly virulent PRRSV isolate on PRRSV vaccinated farms in the production system To analyze the impact of the infection in different farms, Four different clades of the newly emerged highly viru- productive data of farms M2 (c1), M7 (c3 and c4) and lent PRRSV were detected (c1-c4) (Fig. 1). The first Nu4 (c4) were examined. These farms were infected at detection (c1) of the new highly virulent strain occurred different moments of the dissemination of the new highly before the start of the present study, in farm M2 in Janu- virulent strain (January 2020 to April 2021). M2 had been ary 2020; no other outbreaks occurred until the begin- detected as positive stable for a Spanish clade of PRRSV-1 ning of 2021. In March 2021, two new outbreaks of the before January 2020, when reproductive problems started highly virulent strain were reported. The first occurred again; in that case caused by the highly virulent PRRSV in the farrow-to-finish farm M8 (c2). This same virus strain. M7 was PRRSV positive stable until December was later found in nursery N4b (June 2021). The second 2020, when a reproductive outbreak characterized by detection (c3), appeared in nursery N7 in March 2021, abortions started. Nu4, that was positive for a Spanish and shortly thereafter (April 2021 onwards) in nurseries clade of PRRSV1, suffered a reproductive outbreak start- N5 and N6a, and in June 2021 in farm M7. The fourth ing April 2021 (confirmed in May 2021). Martín‑Valls et al. Porcine Health Management (2023) 9:1 Page 3 of 11 GQ461593 R Recombinant isolate 0.02 AMERVAC-PRRS-GU067771 N7/Jun-2021 N6b/May-2021 MSR = Mean substuon rate First detecon date N7/Jul-2021 MND = Mean nucleode diversity 100 N5/May-2021 for a given cluster 100 CReSA3267-JF276435 M3/Jun-2021 95 N5/Jul-2021 95 Lelystad virus-NC 043487 100 Porcilis PRRS-KJ127878 Suvaxyn PRRS MLV-LQ787782 Ingelvac PRRSFLEX EU-KT988004 M3/May-2021 81 M3-2/May-2021 M6/Dec-2020 6-month period 100 N6a/Mar-2021 MSR= 0.09% [0.01-0.36] M6/Feb-2021 M6/Dec-2020 MND= 0.37% ± 0.03 M6-2/Dec-2020 95 Nu1/May-2021 6-month period Nu1/Apr-2021 100 Nu1/Dec-2020 MSR= 0.17% [0.01-0.52] Nu1/Feb-2021 MND= 0.63% ± 0.05 Nu4/Feb-2021 Nu4/Mar-2021 N4c/Mar-2021 100 M4/May-2021 N4a/Jun-2021 N4a/Apr-2021 7-month period M4/Dec-2020 N4a/Jan-2021 MSR= 0.03% [0.01-0.09] N4a/Feb-2021 MND= 0.55% ± 0.04 N4a/Mar-2021 M4-2/Jun-2021 M4-2/May-2021 M4/Jun-2021 N4b/Jun-2021-OM893850 M8/Mar-2021-OM893851 5-month period c2 92 M8/Jun-2021-OM893854 100 M8/Apr-2021-OM893852 MSR= 0.63% [0.26-1.22] M8/May-2021-OM893853 M8/Jul-2021-OM893855 MND= 0.53% ± 0.04 M2/Feb-2021-OM893828 M2/Feb-2021-OM893829 5-month period 100 N2/Jun-2021-OM893835 N2/May-2021-OM893834 c1 MSR= 1.1% [0.82-1.37] 100 N2/Jul-2021-OM893836 N2/Mar-2021-OM893833 MND= 0.61% ± 0.04 N6a/May-2021-OM893839 M7/Jun-2021-OM893830 4-month period 86 N5/Apr-2021-OM893837 N7/Mar-2021-OM893845 100 N6a/Jun-2021-OM893840 c3 MSR= 0.84% [0.62-1.09] N6a/Jul-2021-OM893841 MND= 0.93% ± 0.05 N6a/Apr-2021-OM893838 85 M7/Oct-2021-OM893832 Nu4/Jul-2021-OM893849 6-Month period 100 Nu4/Jun-2021-OM893848 M7/Jun-2021-OM893831 c4 MSR= 0.33% [0.01-0.98] Nu4/May-2021-OM893847 Nu4/May-2021-OM893846 MND=0.21% ± 0.03 Lena-JF802085 Fig. 1 Maximum likelihood phylogenetic tree build up using whole genome sequences, a general time reversible model with 1,000 iterations and pairwise deletion. Numbers along the branches represent the confidence probability of the corresponding clade Table 1 and Figs. 2 and 3 summarize the impact of the and mummies some weeks the count of piglets born infection on productive parameters of the examined alive was close to zero. farms. Additional file 3 shows the impact of the dis- In M7, infected by the strain of increased virulence ease on the actual number of weaned piglets versus the 11 months later (December 2020), the impact was more expected values. Additional file 4 shows the excess on persistent. Thus, abortion rates reached on average only mortality in nurseries. In M2, probably one of the first 11%, but 42 weeks were required to return to the base- farms infected by the emerging highly virulent strain line. Interestingly, in M7 the proportion of mummies was in Spain, the impact of the infection was devastating. very high (6.8%) for 40 weeks. In this case, mortality of For 13 consecutive weeks the abortion rates surpassed sows was lower and non-significantly different from the 25%, reaching 71% some weeks; the return to baseline rates recorded before the outbreak. Mortality of suckling production required 17 weeks. In this farm, the death piglets was not restored to normal values before the 4­ 3rd rate of fertile sows was affected for 39 weeks after the week after the onset of the outbreak. onset of the outbreak, with a weekly average of 6.5%. In Nu4, the abortion rate increased less than in the During the outbreak, mortality of suckling piglets was other two farms (5.9% on average) and the return little affected because summing up abortions, stillbirths to baseline only required 12 weeks. In contrast, the Martín‑Valls et al. Porcine Health Management (2023) 9:1 Page 4 of 11 Table 1 Summary of several productive parametres before and during PRRS outbreaks caused by the strain of enhanced virulence Before the outbreak During the outbreak Weeks to baseline FARM M2 Abortion rate Average: 1.0% Average: 28.7% (+ 27.7)**** 17 Median: 0.7% Median: 29.2% (+ 28.5) Peak: 71.0% Stillbirths Average: 6.8% Average: 9.4% (+ 2.6)* 17 Median: 5.7% Median: 8.4% (+ 2.7) Peak: 27.0% Mummies Average: 0.6% Average: 1.8% (+ 1.2) (n.s) 16 Median: 0.7% Median: 1.7% (+ 1.0) Peak: 9.6% Fertile sow mortality rate Average: 1.0% Average: 6.5% ( +)*** 39 Median: 0.8% Median: 4.8% ( +) Peak: 25.0% Suckling piglet mortality Average: 10.9% Average: 7.3% (-2.6) (n.s.) 0 Median: 7.9% Median: 6.3% (-1.6) Peak: 10.6% FARM M7 Abortion rate Average: 1.5% Average: 11.1% (+ 9.6)**** 42 Median: 1.4% Median: 9.5% (+ 8.1) Peak: 35.4% Stillbirths Average: 8.3% Average: 13.0% (+ 4.7)* 37 Median: 8.0% Median: 13.2% (+ 5.2) Peak: 16.7% Mummies Average: 2.9% Average: 6.8% (+ 3.9)** 40 Median: 2.8% Median: 4.6% (+ 1.8) Peak: 28.5% Fertile sow mortality rate Average: 1.2% Average: 3.7% (+ 2.5) p = 0.1 11 Median: 1.3% Median: 3.2% (+ 1.9) Peak: 7.9% Suckling piglet mortality Average: 13.5% Average:15.0% (+ 1.5)*** Not yet reached after 42 weeks Median: 12.8% Median: 16.7% (+ 3.9) Peak: 31.3% FARM Nu4 Abortion rate Average: 0.7% Average: 5.9% (+ 5.2)**** 12 Median: 0.7% Median: 5.6% (+ 4.9) Peak: 10.4% Stillbirths Average: 1.6% Average: 10.7% (+ 4.7)**** Not reached yet after 39 weeks Median: 1.2% Median: 10.8% (+ 5.2) Peak: 37.1% Mummies Average: 1.7% Average: 2.6% (+ 0.9)** Not reached yet after 39 weeks Median: 1.7% Median: 2.4% (+ 0.7) Peak: 11.5% Fertile sow mortality rate Average: 0.7% Average: 2.5% (+ 1.8)** 39 Median: 0.7% Median: 2.1% (+ 1.4) Peak: 5.2% Suckling piglet mortality Average: 12.2% Average: 16.1% (+ 3.9)**** Not reached yet after 39 weeks Median: 12.6% Median: 15.6% (+ 3.0) Peak: 38.3% The table shows the main indicators for the farrowing units before and during the outbreak and the time needed to recover the baseline values. Statistical significance (Mann–Whitney): *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001; n.s. = non significant proportion of stillbirths and mummies remained ele- For comparison purposes, for breeding farm M4, that vated for more than 39 weeks. The sow death rate had been infected endemically with PRRSV since before although lower (2.5%) when compared to M2, remained 2019, the average figures for the period 2019–2020–2021 elevated for 39 weeks in comparison to the pre-out- were: abortion rate: 2.3% ± 1.03%; mortality rate of fer- break values. Suckling piglet mortality remained ele- tile sows: 1.3 ± 1.2%; stillbirths: 9.2% ± 2.0%; mummies: vated for more than 39 weeks. 0.4 ± 0.09/litter, pre-weaning mortality: 14.3% ± 2.6%. Martín‑Valls et al. Porcine Health Management (2023) 9:1 Page 5 of 11 Fig. 2 Impact of an increased virulence PRRSV isolate compared to a strain of moderate virulence. The graphs depict the temporal evolution of the abortion rate, fertile sow mortality and suckling piglet mortality starting on the 31st week of 2019 until the beginning of 2022. A–C farms (M2, Nu4 and M7) that suffered outbreaks caused by the increased virulence PRRSV strain. D A farm (M4) that was endemic instable for a Spanish PRRSV isolate of moderate virulence. The red arrows indicate the starting of the outbreak for each A, B and C farms, farm D had been recurrently positive since before 2019 Nurseries N2 and Nu4-nur that were filled with ani- not known but, in our opinion, this is a likely hypothesis, mals from M7 and Nu4 were also followed (Fig. 4). In since all full genome sequences of this highly virulent iso- both cases, the impact on mortality from weaning to the late known to us harbor these recombining segments. end of the nursery phase was evident, reaching peaks of It is difficult to determine the precise source of intro- almost 60% in N2. Return to normality required several duction of the new PRRSV strain in the studied farms. In months. The feed conversion index and the average daily the case of farm M2, the first case in the studied produc- weight gain were severely affected as well. tion system and one of the first reported cases in Spain, no animals were introduced from external sources and Discussion introduction by people or fomites was unlikely given PRRS is one of the costliest diseases of pigs. For many the biosecurity characteristics of the herd. However, M2 years, it was considered that the species PRRSV-2 was located in a flat area about 800 m from a fattening included strains of higher virulence compared with unit belonging to a different company that frequently the species PRRSV-1. However, in the last years, the imported fatteners from other locations and that suffered description of isolates such as Lena (PRRSV-1.3) or PR40 a sudden increase of mortality just before the outbreak (PRRSV-1.1) [5, 11] suggest that PRRSV-1 also include in M2, supposedly caused by PRRSV. In this case, air- strains of elevated virulence. borne transmission from the fattening unit to M2 cannot During the first half of 2020, a new highly virulent be ruled out. In Nu4, the proximity to a main highway PRRSV strain emerged in Spain [15]. That strain was (450 m westwards) can be also considered a risk factor shown to derive from a previously reported strain cir- for airborne transmission. The land surrounding Nu4 is culating in Italy [5]. However, the strain circulating in flat land with cultivated fields and no farms are in the Spain have undergone several recombination events with vicinity (> 1.5 km). other PRRSV-1 isolates. Some of these recombination The study of this production system allowed to see events probably took place locally in Spain, as identified the evolution of the infection by this strain of enhanced in the analysis of the available sequences. Whether these virulence. The first case took place in January 2020, recombination events resulted in enhanced virulence is before the onset of the present study and resulted in a Martín‑Valls et al. Porcine Health Management (2023) 9:1 Page 6 of 11 Fig. 3 Temporal evolution of stillbirths and mummies in farms suffering PRRSV outbreaks of increased virulence compared to a farm infected by a strain of moderate virulence. The graphs show the percentages of stillbirths and mummies during 2020 and 2021 in three farms:M2 (A), M7 (B) and Nu4 (C) affected by the PRRSV strain of increased virulence compared to a PRRSV endemic farm (M4, D) where a moderate virulence isolate was circulating since before 2019. The red arrow shows the onset of the reproductive PRRS outbreak in each farm devastating impact on the affected farm. The productive long before the outbreak with the new strain. In compari- data suggested that the introduction of the virus resulted son, data published for PRRSV-2 in the USA indicated a in an explosive dissemination of the infection within the median time to baseline production of 16.5 weeks (range farm, with almost all sows probably infected in a short 0–29 weeks) [14]. These figures may somewhat fit the period of time. Although the farm was positive stable and case of M2, except for sow mortality, but are well below vaccination against PRRSV was implemented, the behav- of the figures for M7 and Nu4, suggesting that the impact ior of the infection resembled a scenario of a fully naïve of the PRRSV-1 strain affecting the present produc- population stricken by a new agent. It is worth noting tion system was probably above the average of PRRSV-2 that the infection in M2 was associated with an impor- infected farms in the USA. Time to stability was not tant mortality in sows. In contrast, when farm M7 was established since, at the end of the study, weaned animals infected, some 11 months later, the impact was differ- of the farms infected by the increased virulence isolate ent: lower abortion rates, lower mortality of sows, but a were still testing positive. higher persistence of the disease and more time needed Looking at the mean substitution rates of each clade of to return to baseline normality (10–11 months for most the virus, it can be observed that while the clade affect- of the examined parameters). Finally, for Nu4, the infec- ing M2 (the first farm infected) showed a substitution tion produced less abortion, less mortality in sows, but rate of 1.1%, the mean substitution rate for M7 was 0.84 even more time was needed to return to normality, with and it was 0.33 for Nu4. In other words, later variants of most parameters still altered 39 weeks after the onset of the virus showed less acute impact in the farm, but more the outbreak. This farm was the only one that was not persistence, together with lower substitution rates. It is implementing vaccination before the outbreak, although tempting to hypothesize that the virus reduced its viru- it had been classified in the PRRSV-positive category lence as gained capability to persist in the herd, in the Martín‑Valls et al. Porcine Health Management (2023) 9:1 Page 7 of 11 Fig. 4 Evolution of mortality, feed conversion index and average daily weight gain in nurseries N2 and Nu4-nurs, after being infected with a highly virulent PRRSV-1 isolate, compared to nursery N4 that received piglets from a farm endemically infected with a PRRSV-1 strain of moderate virulence. The dotted read line indicates the median of each parametre before the outbreak. Data are shown per trimesters (Q1–Q4 each year) framework of a selection process. The more adapted the positive and were implementing a vaccination program. virus and the more fit for persistence, the lower the sub- However, the impact changed over time, transitioning stitution rate. Compared to the evolution of the Spanish from a relatively short (3 months), but devasting impact clades already existing in the production system (mean on the reproductive performance, to a milder but sus- substitution rates from 0.03 to 0.17%), it is evident that tained impact that required almost one year to return to the evolution of the new strain and the already existing baseline production. These results highlight the impor- ones was taking place at very different speeds, prob- tance of implementing surveillance and reporting sys- ably reflecting the high dissemination of the new strain. tems that may alert of the arising of new PRRSV variants To note, in our hands, the in vitro replication of the of concern, as well as of the need to extreme the biosecu- highly virulent strain produced titers 1–2 l­og10 higher rity of pig farms in Europe. than those of the classical Spanish PRRSV-1 strains (not shown). Material and methods It is also worth to comment that up to four significantly Description of the production system different clades of the emergent PRRSV were detected, The production structure consisted of four pyramids that suggesting that these were the result of 4 different intro- accounted for 7,600 sows in 4 nuclei and > 13,000 sows in duction events from different sources. The sources of the multipliers (Fig. 5). An additional farrow-to-finish farm virus could not be established. (1,600 sows) belonging to the same company sometimes received surplus animals (weaners) from other farms of Conclusions the system. A new highly virulent PRRSV1 strain resulting from dif- All nuclei and breeding farms applied strict biosecurity ferent recombination events emerged in Spain in 2020. measures including, but not limited to, double perimetral This strain had a considerable impact in the affected fence, parking area, limitation of access to farms to non- farms, regardless of whether they were previously PRRSV authorised people, compulsory shower before entering Martín‑Valls et al. Porcine Health Management (2023) 9:1 Page 8 of 11 Fig. 5 Structure of the production pyramids included in the study the animal facilities, workwear of exclusive use for each with the others through personnel, trucks or by any other farm, and ≥ 24 h downtime. All trucks used for the trans- mean. PRRS-positive farms were vaccinating gilts and portation of pigs were washed and disinfected after each sows against PRRSV with modified-live vaccines (twice four months with UNISTRAIN PRRS®) except for Nu4 transport and disinfected again before the transporta- before first service and recall blanket vaccination every tion of a new batch of animals the next day of operation. Dead animals were disposed in tightly closed containers that did not apply vaccination against PRRSV. located outside the perimetral fence. The containers were Regarding the location of the farms, all were picked up by trucks designed for this purpose and that located ≥ 2.5 km form other pig farms except for M2, never entered the perimeter of the farm. that was located 800 m from a fattening unit belonging Farms on top of each pyramid and the farrow-to-fin- to a different company. Nu4 was located at < 1 km from a ish farm only used self-replacements. In all cases, gilts highway with high density of pig transports. sent from the nucleus farms (Nu1-Nu4) to the multipli- Pyramid 1 (P1) included one nucleus (Nu1, 2,800 ers were tested for PRRSV by RT-qPCR at weaning and sows), three multipliers (M1, M2, M3, 6,430 sows in only negative gilts were sent to the destination farm. Each total) and several nurseries that then feed other produc- pyramid was operated independently and had no contact tion farms (within or outside the company) with breeders Martín‑Valls et al. Porcine Health Management (2023) 9:1 Page 9 of 11 or fatteners. Nu1 was PRRSV-positive at the beginning were already PRRSV-positive because of previous infec- of the study (December 2020), as revealed by the occa- tions and that in two cases (M2 and M7 were vaccinat- sional detection of positive sows or piglets. M1 was sta- ing). Data collected included the numbers of served sows, ble to PRRSV (no RT-PCR positive pigs at weaning), but number of abortions, mortality of productive sows, total in M2 and M3 sows or piglets were occasionally found number of piglets born per sow (broke down as born to be positive for PRRSV by RT-PCR. Pyramid 2 (P2) alive, mummified, stillbirths), mortality from birth to consisted of a nucleus (Nu2, 400 sows) that sent sows to weaning, and fertility rates. Two nurseries (N2 and N7) two breeding farms (M4, M5; 4,700 sows in total), plus were also included in the analysis (mortality rates, feed several nurseries that fed several other farms as before. conversion index from weaning to 10 weeks of age, and Nu2 and M5 were historically free from PRRSV, while average daily weight gain). Analysed data comprised M4 was known to be endemically infected with viral cir- 2019, 2020 and 2021. Average data for each farrow- culation in the farrowing units. Pyramid 3 (P3) included ing batch (weekly) were rolling averaged (three weeks). one nucleus (Nu3, 2,800 sows), one multiplier (M6; Return to baseline for a given parameter was considered 1,200 sows) and several nurseries. M6 was known to be to happen when, for at least 4 consecutive weeks, the val- endemically infected by PRRSV at the beginning of the ues for that parameter were in or above the median of the study. Pyramid 4 (P4) included one nucleus (Nu4, 1,600 farm calculated for the six months before the outbreak. sows), one multiplier (M7, 1,200 sows) and two nurseries Farm M4 and nursery N4 that were endemically infected that fed other farms outside the pyramid or sent animals by a Spanish strain of moderate virulence since before for fattening. Nu4 was PRRSV-positive with occasional 2019 were used for comparative purposes. M4 applied a detection of infected animals at weaning; M7 was PRRSV vaccination plan to gilts and sows as mentioned above. positive but stable, and circulation of the virus was not detected in weaners. The farrow to-finish farm M8 was Virus isolation, sequencing, and phylogenetic analyses endemically infected by PRRSV. After March 2021, M6 Virus isolation in alveolar macrophages was attempted sent weaned pigs to nursery N7 that was previously used using the samples showing the lowest Cts for PRRSV. for weaners coming from M7 (M7 growers were sent to To avoid biasing the results, only a single passage was nursery N6a afterwards). performed. According to previous reports [8], with this method the discrepancy between the PRRSV sequence Sampling and RT‑qPCR analysis obtained from the original sample, or from the iso- In this study (December 2020–October 2021) samples late was 1–3 nucleotides per 1 ­ 04 positions. Viral RNA taken for routine PRRSV monitoring or diagnosis were was extracted from the isolates using the Trizol reagent used. All nucleus and breeding farms were monitored (Thermo Fisher), with an elution volume of 20 µl. monthly for PRRSV. Results included in the present study The presence of PRRSV RNA in the sample was encompassed the period between December 2020 and assessed by using a commercial RT-PCR kit (LSI VET- October 2021. In breeding herds, 30 random suckling MAX PRRSVEUNA, ThermoFisher), according to the piglets at 3 weeks of age were bled monthly for moni- instructions provided with the kit. If the Ct of the sam- toring. When a reproductive outbreak compatible with ple was lower than 20, the viral RNA was used for NGS PRRSV appeared, sampling was aimed at affected lit- using Illumina Miseq without performing any previous ters and sows. In nurseries, monitoring was performed amplification. The protocol developed in five steps. First, every month by a cross-sectional sampling at 6 and the genomic library was constructed using a commer- mRNA or rRNA Depleted RNA and NEBNext® Ultra™ II 9 weeks of age (10 animals each). Samples were submit- cial protocol and reagents (Protocol for use with Purified RNA Library Prep Kit for Illumina®, New England Bio- ted to the Veterinary Laboratory for the Diagnosis of Infectious Diseases of UAB for diagnosis. Upon recep- tion, serum was separated, and RNA was extracted using labs). After the NGS run, sequences of low quality were the MagMax Core RNA Extraction kit in a KingFisher trimmed (QC > 20) using Trimmomatic [3]. Then, reads Flex System (ThemoFisher). Samples were analyzed by were mapped against a reference sequence (Burrows- RT-qPCR using the LSI VETMAX PRRSVEUNA 2.0 kit Wheeler Aligner applying the BWA-MEM algorithm for (Thermofisher). long reads) [13]. The reference sequence was produced form the earliest available isolate obtained during the Impact of the infection by a new highly virulent PRRSV outbreak by de novo assembly using SPADES [1]. In the strain in previously PRRSV‑infected farms fourth step, variant calling to determine the frequency To assess the impact of the infection caused by the new of each nucleotide at each position of the reference highly virulent isolate, productive and mortality data genome was performed with SnpSift [7]. Finally, the viral were collected from 3 farms (M2, Nu4 and M7) that quasi-species was constructed in fasta format and the Martín‑Valls et al. Porcine Health Management (2023) 9:1 Page 10 of 11 consensus sequence was obtained using Consensus soft- Additional file 3. Variation in the number of weaned piglets with regards ware (http://​www.​hiv.​lanl.​gov/​conte​nt/​seque​nce/​CONSE​ to the pre-established production objectives of each farm. NSUS/​conse​nsus.​html). Additional file 4. Excess mortality in nurseries associated to the PRRS A phylogenetic tree was built up for the whole genome outbreak. sequences using MEGA X [24], applying the maximum likelihood method, a general time reversible model Acknowledgements with 1,000 iterations and pairwise deletion. The data- Our grateful thanks to the veterinary technical services from Grup Batallé for sample collection and Laura Valls (Diagnos Hipra) for helping in the selection set contained, apart from the 55 consensus sequences of samples and the analysis. obtained in this work, two reference PRRSV1 genomes (NC_043487-Lelystad and JF276435-CReSA3267), plus 4 Author contributions GEMV, MC, AA, FI and EM designed the study. FI collected and quality-con‑ vaccine strains available in GenBank. The tree was rooted trolled data of the farms. GEMV, MC and MT analysed the samples, performed with a PRRSV1 subtype 3 strain (JF802085-Lena). the viral isolations, and sequenced the isolates. GEMV, MC and EM analysed the sequences and prepared the phylogenetic trees. GEMV, EM and FI, ana‑ lysed the productive data. All authors read and approved the final manuscript. Recombination analyses Funding The existence of potential recombination points within This study was partially funded by a Boehringer Ingelheim European PRRS the viral genome was initially assessed using GARD Research Award. Funds from the Veterinary Diagnostic Laboratory for Infec‑ Kosakovsky Pond [12] and RDP v5 (Darren et al. 2021). tious Diseases of the UAB partially supported the analysis. The dataset was the same used for the phylogenetic Availability of data and materials analysis. For every sequence, the recombinant segments Sequences are available at Genbank under accession numbers OM893828- delimited by the recombination points detected by both OM893841, OM893845-OM893855 and OP822960-OP822989. Productive data can be available from the authors upon reasonable request. programs were submitted to BLAST, to identify the potential parental strains. Declarations Ethics approval and consent to participate Substitution rate estimations Samples used in this work were submitted by field veterinarians for diagnostic The evolutionary rate within every clade of related trans- and monitoring purposes on the basis of clinical testing. Therefore, according to Directive 63/2010 -Article 1, point 5(b) non-experimental clinical veterinary mission -for the whole genome, as well as per every practices- the procedure carried out to obtain data is not within the scope genome segment- were calculated using a Bayesian of the named directive. Written informed consent for participation was not Markov chain Monte Carlo (MCMC) approach, imple- obtained from the owners because the data used in the present paper comes from routine diagnostic of pig farms. Also, the data in the present work is used mented in BEAST v1.10.4 [23]. Firstly, the aligned con- anonymously. sensus sequences of every isolate were used to generate a BEAST input file, using the software BEAUTi. Three Competing interests The authors declare no competing interests. independent runs of MCMC per dataset were performed under a strict molecular clock model, using the Tamura- Author details 1 Nei model of sequence evolution, with a proportion of Dept. Sanitat i Anatomia Animals, Facultat de Veterinària, Universi‑ tat Autònoma de Barcelona, Travessera dels Turons s/n, Campus UAB, invariant sites and gamma distributed rate heterogene- 08193 Cerdanyola del Vallès, Spain. 2 Grup Batallé, Avinguda dels Segadors s/n, ity (TN + I + Γ), with partitions into codon positions, and 17421 Riudarenes, Spain. the remaining default parameters in the prior’s panel. Received: 22 November 2022 Accepted: 15 December 2022 Every MCMC run was 100 million steps long and the posterior probability distribution of the chains was sam- pled every 1000 steps. Convergence was assessed based on an effective sampling size after a 10% burn-in, using References Tracer v1.7.2 [21]. The substitution rate estimations were 1. Bankevich A, Nurk S, Antipov D, Gurevich AA, Dvorkin M, Kulikov AS, Lesin the mean values obtained for the three runs. VM, Nikolenko SI, Pham S, Prjibelski AD, Pyshkin AV, Sirotkin AV, Vyahhi N, Tesler G, Alekseyev MA, Pevzner PA. SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing. J Comput Biol. Supplementary Information 2012;19:455–77. https://​doi.​org/​10.​1089/​cmb.​2012.​0021. The online version contains supplementary material available at https://​doi.​ 2. Benfield DA, Nelson E, Collins JE, Harris L, Goyal SM, Robison D, Christian‑ org/​10.​1186/​s40813-​022-​00298-3. son WT, Morrison RB, Gorcyca D, Chladek D. Characterization of swine infertility and respiratory syndrome (SIRS) virus (isolate ATCC VR-2332). J Vet Diagn Invest. 1992;4:127–33. https://​doi.​org/​10.​1177/​10406​38792​ Additional file 1. Emergence of the different clades of the highly virulent 00400​202. PRRSV isolate or its recombinants. 3. Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illu‑ Additional file 2. Recombination patterns among the recombinant mina sequence data. Bioinformatics. 2014;30:2114–20. https://​doi.​org/​10.​ isolates detected for the new strain of enhanced virulence. 1093/​bioin​forma​tics/​btu170. Martín‑Valls et al. Porcine Health Management (2023) 9:1 Page 11 of 11 4. Brockmeier SL, Palmer MV, Bolin SR. Effects of intranasal inoculation 21. Rambaut A, Drummond AJ, Xie D, Baele G, Suchard MA. Posterior sum‑ of porcine reproductive and respiratory syndrome virus, Bordetella marisation in Bayesian phylogenetics using Tracer 1.7. Syst Biol. 2018; bronchiseptica, or a combination of both organisms in pigs. Am J Vet Res. syy032. doi:https://​doi.​org/​10.​1093/​sysbio/​syy032 2000;61:892–9. https://​doi.​org/​10.​2460/​ajvr.​2000.​61.​892. 22. Stadejek T, Stankevicius A, Murtaugh MP, Oleksiewicz MB. Molecular 5. Canelli E, Catella A, Borghetti P, Ferrari L, Ogno G, De Angelis E, Corradi A, evolution of PRRSV in Europe: current state of play. Vet Microbiol. Passeri B, Bertani V, Sandri G, Bonilauri P, Leung FC, Guazzetti S, Martelli P. 2013;165:21–8. https://​doi.​org/​10.​1016/j.​vetmic.​2013.​02.​029. Phenotypic characterization of a highly pathogenic Italian porcine repro‑ 23. Suchard MA, Lemey P, Baele G, Ayres DL, Drummond AJ, Rambaut A. ductive and respiratory syndrome virus (PRRSV) type 1 subtype 1 isolate Bayesian phylogenetic and phylodynamic data integration using BEAST in experimentally infected pigs. Vet Microbiol. 2017;210:124–33. https://​ 1.10. Virus Evolution. 2018; 4:vey016. https://​doi.​org/​10.​1093/​ve/​vey016 doi.​org/​10.​1016/j.​vetmic.​2017.​09.​002. 24. Tamura K, Stecher G, Kumar S. MEGA11: Molecular Evolutionary Genetics 6. Christianson WT, Collins JE, Benfield DA, Harris L, Gorcyca DE, Chladek Analysis version 11. Mol Biol and Evol. 2021;38:3022–7. DW, Morrison RB, Joo HS. Experimental reproduction of swine infertility 25. Terpstra C, Wensvoort G, Pol JM. Experimental reproduction of porcine and respiratory syndrome in pregnant sows. Am J Vet Res. 1992;53:485–8. epidemic abortion and respiratory syndrome (mystery swine disease) 7. Cingolani P, Platts A, Wang le L, Coon M, Nguyen T, Wang L, Land SJ, Lu by infection with Lelystad virus: Koch’s postulates fulfilled. Vet Q. X, Ruden DM. Fly (Austin). A program for annotating and predicting the 1991;13:131–6. https://​doi.​org/​10.​1080/​01652​176.​1991.​96942​97. effects of single nucleotide polymorphisms, SnpEff: SNPs in the genome 26. Tian K, Yu X, Zhao T, Feng Y, Cao Z, Wang C, Hu Y, Chen X, Hu D, Tian X, Liu of Drosophila melanogaster strain w1118; iso-2; iso-3. 2012; 6:80–92. D, Zhang S, Deng X, Ding Y, Yang L, Zhang Y, Xiao H, Qiao M, Wang B, Hou 8. Cortey M, Arocena G, Pileri E, Martín-Valls G, Mateu E. Bottlenecks in the L, Wang X, Yang X, Kang L, Sun M, Jin P, Wang S, Kitamura Y, Yan J, Gao transmission of porcine reproductive and respiratory syndrome virus GF. Emergence of fatal PRRSV variants: unparalleled outbreaks of atypical (PRRSV1) to naïve pigs and the quasi-species variation of the virus during PRRS in China and molecular dissection of the unique hallmark. PLoS infection in vaccinated pigs. Vet Res. 2018;49:107. https://​doi.​org/​10.​ ONE. 2007; 13;2(6):e526. doi: https://​doi.​org/​10.​1371/​journ​al.​pone.​00005​ 1186/​s13567-​018-​0603-1. 26. 9. Galina L, Pijoan C, Sitjar M, Christianson WT, Rossow K, Collins JE. Interac‑ 27. Trevisi P, Amatucci L, Ruggeri R, Romanelli C, Sandri G, Luise D, Canali tion between Streptococcus suis serotype 2 and porcine reproductive M, Bosi P. Pattern of Antibiotic Consumption in Two Italian Production and respiratory syndrome virus in specific pathogen-free piglets. Vet Rec. Chains Differing by the Endemic Status for Porcine Reproductive and 1994;134:60–4. https://​doi.​org/​10.​1136/​vr.​134.3.​60. Respiratory Syndrome. Front Vet Sci. 2022; 9:840716. doi: https://​doi.​org/​ 10. Holtkamp DJ, Kliebenstein JB, Neumann EJ, Zimmerman JJ, Rotto HF, 10.​3389/​fvets.​2022.​840716. Yoder TK, Wang C, Yeske PE, Mowrer CL, Haley CA. Assessment of the 28. Yu J, Wu J, Zhang Y, Guo L, Cong X, Du Y, Li J, Sun W, Shi J, Peng J, Yin F, economic impact of porcine reproductive and respiratory syndrome virus Wang D, Zhao P, Wang J. Concurrent highly pathogenic porcine repro‑ on United States pork producers. J Swine Health Prod. 2013;21:72–84. ductive and respiratory syndrome virus infection accelerates Haemophi‑ 11. Karniychuk UU, Geldhof M, Vanhee M, Van Doorsselaere J, Saveleva TA, lus parasuis infection in conventional pigs. Vet Microbiol. 2012;158:316– Nauwynck HJ. Pathogenesis and antigenic characterization of a new East 21. https://​doi.​org/​10.​1016/j.​vetmic.​2012.​03.​001. European subtype 3 porcine reproductive and respiratory syndrome virus isolate. BMC Vet Res. 2010;6:30. https://​doi.​org/​10.​1186/​1746-​6148-6-​30. 12. Kosakovsky Pond SL, Posada D, Gravenor MB, Woelk CH, Frost SD. Publisher’s Note GARD: a genetic algorithm for recombination detection. Bioinformatics. Springer Nature remains neutral with regard to jurisdictional claims in pub‑ 2006;22:3096–8. https://​doi.​org/​10.​1093/​bioin​forma​tics/​btl474. lished maps and institutional affiliations. 13. Li H, Durbin R. Fast and accurate long-read alignment with Burrows- Wheeler transform. Bioinformatics. 2010;26:589–95. https://​doi.​org/​10.​ 1093/​bioin​forma​tics/​btp698. 14. Linhares DC, Cano JP, Torremorell M, Morrison RB. Comparison of time to PRRSv-stability and production losses between two exposure programs to control PRRSv in sow herds. Prev Vet Med. 2014;116:111–9. https://​doi.​ org/​10.​1016/j.​preve​tmed.​2014.​05.​010. 15. Martin DP, Varsani A, Roumagnac P, Botha G, Maslamoney S, Schwab T, Kelz Z, Kumar V, Murrell B. RDP5: a computer program for analyzing recombination in, and removing signals of recombination from, nucleo‑ tide sequence datasets. Virus Evol. 2020;12:veaa087. https://​doi.​org/​10.​ 1093/​ve/​veaa0​87. 16. Martín-Valls GE, Cortey M, Allepuz A, Illas F, Tello M, Mateu E. Description of a New Clade within Subtype 1 of Betaarterivirus suid 1 Causing Severe Outbreaks in Spain. Microbiol Resour Announc. 2022;11:122. https://​doi.​ org/​10.​1128/​mra.​00304-​22. 17. Martínez-Lobo FJ, Díez-Fuertes F, Segalés J, García-Artiga C, Simarro I, Castro JM, Prieto C. Comparative pathogenicity of type 1 and type 2 isolates of porcine reproductive and respiratory syndrome virus (PRRSV) in a young pig infection model. Vet Microbiol. 2011;154:58–68. https://​ Ready to submit your research ? Choose BMC and benefit from: doi.​org/​10.​1016/j.​vetmic.​2011.​06.​025. 18. Murtaugh MP, Stadejek T, Abrahante JE, Lam TT, Leung FC. The ever- • fast, convenient online submission expanding diversity of porcine reproductive and respiratory syndrome virus. Virus Res. 2010;154:18–30. https://​doi.​org/​10.​1016/j.​virus​res.​2010.​08.​ • thorough peer review by experienced researchers in your field 015. • rapid publication on acceptance 19. Nathues H, Alarcon P, Rushton J, Jolie R, Fiebig K, Jimenez M, Geurts V, • support for research data, including large and complex data types Nathues C. Cost of porcine reproductive and respiratory syndrome virus at individual farm level—an economic disease model. Prev Vet Med. • gold Open Access which fosters wider collaboration and increased citations 2017;142:16–29. https://​doi.​org/​10.​1016/j.​preve​tmed.​2017.​04.​006. • maximum visibility for your research: over 100M website views per year 20. Nieuwenhuis N, Duinhof TF, van Nes A. Economic analysis of outbreaks of porcine reproductive and respiratory syndrome virus in nine sow herds. At BMC, research is always in progress. Vet Rec. 2012;170:225. https://​doi.​org/​10.​1136/​vr.​100101. Learn more biomedcentral.com/submissions