Coronado et al. BMC Veterinary Research (2019) 15:247 https://doi.org/10.1186/s12917-019-1982-x RESEARCH ARTICLE Open Access Investigation of chronic and persistent classical swine fever infections under field conditions and their impact on vaccine efficacy Liani Coronado1, Jose Alejandro Bohórquez2, Sara Muñoz-González2, Lester Josue Perez3, Rosa Rosell2,4, Osvaldo Fonseca1, Laiyen Delgado1, Carmen Laura Perera1, Maria Teresa Frías1 and Llilianne Ganges2* Abstract Background: Recent studies have hypothesized that circulation of classical swine fever virus (CSFV) variants when the immunity induced by the vaccine is not sterilizing might favour viral persistence. Likewise, in addition to congenital viral persistence, CSFV has also been proven to generate postnatal viral persistence. Under experimental conditions, postnatal persistently infected pigs were unable to elicit a specific immune response to a CSFV live attenuated vaccine via the mechanism known as superinfection exclusion (SIE). Here, we study whether subclinical forms of classical swine fever (CSF) may be present in a conventional farm in an endemic country and evaluate vaccine efficacy under these types of infections in field conditions. Results: Six litters born from CSF-vaccinated gilts were randomly chosen from a commercial Cuban farm at 33 days of age (weaning). At this time, the piglets were vaccinated with a lapinized live attenuated CSFV C-strain vaccine. Virological and immunological analyses were performed before and after vaccination. The piglets were clinically healthy at weaning; however, 82% were viraemic, and the rectal swabs in most of the remaining 18% were positive. Only five piglets from one litter showed a specific antibody response. The tonsils and rectal swabs of five sows were CSFV positive, and only one of the sows showed an antibody response. After vaccination, 98% of the piglets were unable to clear the virus and to seroconvert, and some of the piglets showed polyarthritis and wasting after 36 days post vaccination. The CSFV E2 glycoprotein sequences recovered from one pig per litter were the same. The amino acid positions 72(R), 20(L) and 195(N) of E2 were identified in silico as positions associated with adaptive advantage. Conclusions: Circulation of chronic and persistent CSF infections was demonstrated in field conditions under a vaccination programme. Persistent infection was predominant. Here, we provide evidence that, in field conditions, subclinical infections are not detected by clinical diagnosis and, despite being infected with CSFV, the animals are vaccinated, rather than diagnosed and eliminated. These animals are refractory to vaccination, likely due to the SIE phenomenon. Improvement of vaccination strategies and diagnosis of subclinical forms of CSF is imperative for CSF eradication. Keywords: CSFV, Chronic infection, Persistent infection, Vaccination failures, Viral evolution * Correspondence: llilianne.ganges@irta.cat 2 OIE Reference Laboratory for Classical Swine Fever, IRTA-CReSA, Campus de la Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain Full list of author information is available at the end of the article © The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. 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. Coronado et al. BMC Veterinary Research (2019) 15:247 Page 2 of 13 Background [11, 15, 17, 18]. Long-term endemism of CSF is charac- Classical swine fever (CSF) is one of the most devastat- terized by the circulation of moderate- and low- ing diseases in the pig industry worldwide, affecting virulence strains [1, 11]. Therefore, laboratory support domestic pigs and wild boars. This disease is endemic in for accurate CSFV diagnosis is fundamental for efficient Asia, Central and South America, and some Eastern Euro- disease control. Unfortunately, not all endemic countries pean countries [1]. CSF virus (CSFV), the aetiological have the necessary resources to achieve this goal, and agent, belongs to the Pestivirus genus of the Flaviviridae only clinical diagnosis is performed. Hence, the lack of family [2]. Three degrees of virulence (low, medium and laboratory diagnostics may also explain the complex high) have been described for CSFV strains, which have epidemiological scenario in some endemic countries. also been linked to different clinical manifestations of the In addition to the persistent viral infection generated disease [3]. Pigs infected with low-virulence strains may after trans-placental transmission, previous works have develop chronic CSF and can shed the virus continuously reported that CSFV is also able to generate persistent or intermittently for months, representing a constant postnatal infection in both domestic pigs and wild boars source of reinfection [3]. Additionally, the role of low- [19, 20]. Subsequently, it was shown that pigs with per- virulence strains in “pregnant carrier sow syndrome”, sistent postnatal infection were unable to elicit a specific which can lead to congenital infection of the foetus by immune response to a CSFV live attenuated vaccine, trans-placental transmission, has been known for more with viral vaccine RNA undetectable in these pigs after than 50 years [4–7]. vaccination via the superinfection exclusion (SIE) mech- Modified live attenuated vaccines based on the C- anism [21, 22]. These studies were performed under ex- strain have played a relevant role in the implementation perimental conditions, with one of the few scientific reports of eradication programmes for CSF [8–12]. These showing the possible negative impact that persistently in- vaccines present several advantages that facilitate their fected animals may have on vaccination programmes, use in developing countries, inducing high titres of neu- which could complicate CSF control in endemic countries. tralizing antibodies that provide protection against Given this background, the present study aimed to highly virulent CSFV strains [9, 13]. Moreover, this type determine whether subclinical forms of CSF, which may of vaccine confers protection against CSFV even in the be misdiagnosed in the field, are present at the time of absence of neutralizing antibodies a few days post vac- vaccination in a conventional farm from an endemic cination (dpv) [8, 9, 13, 14]. In addition to horizontal country. Likewise, this study aimed to evaluate vaccine protection, modified live attenuated vaccines based on efficacy for these types of infections in field conditions, the C-strain also induce protection in pregnant sows as an important aspect that is poorly understood to date. early as 5 dpv, preventing viral trans-placental transmis- sion [8, 9]. Although the immune response induced by Results this type of vaccine does not allow differentiation of in- CSFV detection and antibody response in sows fected animals from vaccinated animals (the DIVA con- Despite being vaccinated, the rectal swabs of five of the cept) [8, 9, 12], this vaccine continues to be used in sows were CSFV RNA positive (from B to F), with Ct endemic countries (a majority of which are developing values corresponding to high (litter C) to moderate countries). The reasons underlying the worldwide distri- (litters B and D) and low (litters E and F) viral RNA bution of this vaccine are its high efficacy and safety as loads. In contrast, the sow from litter A was negative. well as its relatively low cost of production [10]. The results obtained from tonsil samples corresponded It has been previously discussed that CSF has not been with those from rectal swabs, i.e., there were five positive controlled despite long-term vaccination programmes sows (from B to F) and one negative sow (sow A). The with live attenuated C-strain vaccines conducted in sow from litter F showed a Ct value less than 28, and endemic areas [11, 15, 16]. Factors associated with vari- the others showed values greater than 29. Notably, ability, quality, and stability of the vaccine batches pro- although the rectal swabs of five sows were positive, the duced; loss of the cold chain; and lack of vaccination sera of only two of these sows were positive, showing coverage, mainly in rural areas with poor accessibility low viral RNA loads (sows B and C had Ct values of 30 and backyard farms, among others, have been used to and 32, respectively). explain this discrepancy [9, 11, 15, 16]. In terms of humoral response, notably, only the sow Recent studies have demonstrated evolutionary cap- from litter E showed an antibody response (55.3% block- acity of CSFV in endemic conditions under inefficient ing) in the ELISA test. With the exception of the animals vaccination programmes. In this regard, it has been from the six litters under study, all the animals from the hypothesized that circulation of new viral variants when farm were slaughtered once the samples collected at the immunity induced by the vaccine is not sterilizing weaning were found to be CSFV RNA positive, and might favour viral persistence in the swine population strict biosecurity measures were applied in the farm. Coronado et al. BMC Veterinary Research (2019) 15:247 Page 3 of 13 CSFV detection and antibody response in piglets at bladder and stomach; marginal infarcts of the spleen; weaning haemorrhages; and necrosis in the tonsils (Table 2). In At the day of weaning (33 days after birth), the rectal the remaining litters, there were no obvious clinical swabs of all piglets in four of the six litters were CSFV signs; two cases of sudden death occurred in litters A RNA positive (litters B, C, D and F) and showed Ct and D, while cases of animals with polyarthritis were values corresponding to high (Ct value less than 23) and recorded in litters B and C. A mortality rate of 8% was moderate (Ct values between 23 and 28) viral RNA loads obtained between 21 and 96 dpv, with a 14% mortality (Table 1). In the other two litters, only two out of nine observed during the entire study. Additionally, the mor- animals were CSFV RNA positive with high RNA viral tality for each litter was also determined (Table 2). Inter- loads in rectal swabs in litter A, whereas in litter E, eight estingly, in litters B, C and F, no anatomopathological out of eleven were CSFV RNA positive with high and lesions in internal organs were observed after necropsy. moderate RNA viral load in rectal swabs (Table 1). Like- Finally, the three litters that remained until 96 dpv wise, the results obtained for sera were similar to those showed growth retardation from approximately 36 dpv obtained for rectal swabs. All serum samples from litters until the end of the study (litters A, D and F), with clear B, C, D and F were CSFV RNA positive, showing Ct signs of apathy, anorexia, diarrhoea, dyspnoea and con- values corresponding with high viral RNA loads (Ct junctivitis observed at that time in some animals. The value less than 23) in 38 out of 42 piglets (Table 1). largest accumulation of anatomopathological lesions of Different profiles were observed in the two remaining haemorrhages in tonsils and kidneys, enteritis and even litters. In litter E, five out of eleven piglets were CSFV development of interstitial pneumonia were observed in RNA positive, with two animals exhibiting low viral litters A and D, which remained in the study until 96 RNA loads, one with a moderate RNA load and the dpv; however, the mortality percentage in both litters other two with high viral RNA loads (Table 1). In litter was 22.2% (Table 2). A, the serum samples of four out of nine animals were positive (44.4%), three with Ct values corresponding to CSFV detection in piglets after vaccination low RNA loads and one with a moderate viral RNA load. The presence of CSFV RNA in sera, rectal swabs and In general, the subclinical infection detected in the tonsils was monitored by qRT-PCR at different times piglets from the present study at weaning included 82% post vaccination. After 21 dpv, 98% of the piglets from animals that were viraemic. Among the remaining 18%, the six litters were unable to clear the virus, with a ma- most of the piglets, although not viraemic, were positive jority of the piglets exhibiting high and moderate viral in rectal swabs. RNA loads in sera and rectal swabs (Tables 3 and 4). At In terms of humoral response, only five out of eleven 21 dpv, serum samples from the six litters that were piglets from litter B showed a specific antibody response positive by qRT-PCR with Ct values below 28 were veri- to E2 glycoprotein, as detected by ELISA, whereas the fied by a viral isolation test in PK-15 cells. remaining piglets from the study were negative. Regard- Notably, piglets from litters B, C and F showed the high- ing the ELISA-positive samples, the antibody response est viral RNA loads during the study, with strong viral levels measured as blocking percentage values were be- RNA signals observed in sera, rectal swabs and tonsils tween 42 and 47% (Table 1). These results were con- (Tables 3, 4 and Fig. 1), and showed constant and similar firmed by the NPLA, resulting in low neutralizing Ct values during the study. In the remaining litters (A, D antibody titres of 1:25 in the five sera. and E), the viral RNA load ranged from low to moderate and was undetected in some cases. In these litters, rectal Clinical manifestations registered at weaning and after swab samples allowed the detection of a greater number vaccination of positive animals than serum samples (Tables 3, 4 and Notably, at weaning, all the animals were apparently Fig. 1). In particular, a majority of the pigs (seven out of clinically healthy (Table 2), exhibiting a weight gain simi- nine) in litter A tested negative for viral RNA in serum lar to the normal weight gain observed in Cuba in this samples at 21 dpv. After this time, at 33, 66 and 96 dpv, type of commercial farm (data not shown); likewise, the positive samples were detected intermittently, although absence of clinical symptoms was correlated with the with low viral RNA loads. Nevertheless, during the period absence of mortality at weaning. At 21 dpv, clinical signs from 21 to 96 dpv, the rectal swabs from all animals were were registered in only litter E, such as anorexia, positive at some point (Tables 3 and 4). Finally, in pigs conjunctivitis, weakness of hindquarters, dyspnoea, diar- from litter D, two different profiles were detected; three of rhoea, incoordination of movements and nervous disor- these pigs developed some antibody response to the virus, ders; three animals died suddenly, and the remaining as two animals were able to clear the virus in sera and rec- animals were euthanized at 21 dpv. Lesions found after tal swabs from 33 to 96 dpv. In the remainder of the pigs, necropsy included petechial haemorrhages in the kidney, constant viraemia was observed during the study until 96 Coronado et al. BMC Veterinary Research (2019) 15:247 Page 4 of 13 Table 1 Ct values from rectal swabs and sera obtained by CSFV-specific qRT-PCR and CSFV-specific antibodies at weaning b The antibody test result against the E2 glycoprotein was considered positive when the blocking % was more than 40% (in bold). Values between 30 and 40% were considered doubtful (in black), and values less than 30% were considered negative (in grey) dpv with a high viral RNA load. Furthermore, the tonsils positive, although with low viral RNA loads. Finally, the of the animals from this litter were also positive, with a tonsils of all the animals under study were positive at low viral RNA load observed in the two animals that either 21 dpv or 96 dpv (Fig. 1). cleared the virus in sera. In contrast, animals that showed persistent viremia exhibited moderate and high viral RNA Antibody response profile detected after vaccination loads (Tables 3, 4 and Fig. 1). Additionally, despite the Notably, none of the animals in litters B, C and F intermittent viral RNA detection during the study in litter showed an antibody response detected by ELISA at 21 A (Tables 3 and 4), the tonsils of all the animals were dpv (litters B and C) or 96 dpv (litter F). Among the Coronado et al. BMC Veterinary Research (2019) 15:247 Page 5 of 13 Table 2 CSFV-positive animals at weaning and clinical signs, mortality rates and pathological findings in the litters under study remaining three litters, in litter A, five out of nine piglets E2 gene of Pinar del Rio/2016 (LT985811), as well as all were able to seroconvert at 21 dpv (Table 4 and Fig. 2). the available full-length E2 gene sequences from Cuban Furthermore, between 33 and 96 dpv, intermittent detec- CSFV strains, were analysed as described in Methods sec- tion of antibodies in animals from this litter was regis- tion. By in silico analysis, the amino acid at position 72 (R) tered. In terms of neutralizing antibody titres, only four on the E2 glycoprotein was identified under positive selec- pigs were capable of clearing the virus in sera, as tion pressure for the strain LT985811 with a 0.01 level of determined by the NPLA at 96 dpv, with neutralizing significance (2LnL = 12.6 M2vsM1 and 2LnL = 36.8 antibody titres from 1:100 to 1:400 (Fig. 3). M8vsM7). Likewise, sites 20 (L) and 195 (N) were also In the case of litter D, only three pigs showed CSFV- identified in silico as positions with adaptive advantage in specific antibodies to E2 glycoprotein and neutralizing the above-mentioned virus with a 0.01 significance level antibody titres at 21 dpv (Table 4 and Fig. 3). One of (θLTR = 28.85684). Other mutations without statistical these pigs died (pig number 32) despite being the animal significance were detected in the CSFV strain from the with the highest level of neutralizing antibodies gener- present study, namely, 3AxV and 7DxG in the external ated against infection. The other two antibody-positive loop of the B/C domain, 91IxV in the A/D domain and pigs showed neutralizing antibody titres ranging from 1: 210DxN and 228AxV in the transmembrane domain 200 to 1:400 at 96 dpv (Fig. 3). (Additional file 1: Figure S1). Finally, in litter E, five pigs showed CSFV-specific anti- body responses, as measured by ELISA and the NPLA Discussion (Figs. 2 and 3). This group generated the highest neutral- In the present work, the coexistence of two forms of izing antibody response at 21 dpv, with titres ranging subclinical CSF was demonstrated: (i) chronic CSF and from 1:400 to 1:800 (Fig. 3). (ii) persistent CSF, likely transmitted congenitally from sows to their litters. Nevertheless, a postnatal persistent E2 sequence determination for each litter and sequence infection in some of these animals cannot be ruled out analysis considering the ability of CSFV to generate this form of The complete E2 sequences from tonsil samples of one infection soon after birth [20]. Although both forms of animal from each litter studied were determined. Notably, infection can be considered as chronic CSF manifesta- the six consensus sequences obtained were identical and tions, since both exceed 28 days of infection [1, 24], the were deposited in the European Nucleotide Archive under knowledge and distinction of the pathogenesis of both the accession number LT985811. This viral strain was iso- types of infections is relevant for accurate diagnosis and lated and designated Pinar del Rio/2016. The full-length control in endemic areas. One of the main differences Coronado et al. BMC Veterinary Research (2019) 15:247 Page 6 of 13 Table 3 Ct values in rectal swabs obtained by specific qRT-PCR at different times post vaccination Coronado et al. BMC Veterinary Research (2019) 15:247 Page 7 of 13 Table 4 CSFV RNA detection in sera and CSFV-specific antibodies after vaccination a Death or euthanasia, there is no sample. bThe antibody test result against the E2 glycoprotein was considered positive when the blocking % was more than 40% (in bold). Values between 30 and 40% were considered inconclusive (in black), and values less than 30% were considered negative (in grey) Coronado et al. BMC Veterinary Research (2019) 15:247 Page 8 of 13 Fig. 1 CSFV RNA detection in tonsil samples from piglets. CSFV viral RNA was detected using qRT-PCR [23] in litters either at 21 dpv (grey dots) or 96 dpv (white dots). Viral RNA load is expressed as the Ct value; the mean Ct value for each litter is also indicated (black triangle). Ct values equal to or less than 42 were considered to indicate positive results between pigs with persistent CSFV infection and pigs tonsils were considered to be chronically infected, suffering the chronic form is that, in chronic infec- such as piglets from litters A, D and E. tion, a specific immune response against CSFV can be In contrast, persistently infected animals have been generated, although inefficiently, and in some cases, previously defined as CSF-infected pigs that show per- the antibody response can mask the detection of the manent viremia with high viral titres over long periods virus in sera [20, 25]. Considering this fact, even of time and are unable to generate a specific immune re- though antibodies were intermittently detected in sponse to the virus, mainly in terms of the humoral im- some animals, the non-viraemic animals detected in mune response [20, 25]. Recently, a type I and II this study that exhibited positive rectal swabs and interferon response blockade has been demonstrated in Fig. 2 CSFV RNA detection in serum samples at weaning and 21 dpv versus antibody response after vaccination. CSFV RNA loads in sera expressed as Ct values (left Y axis) at weaning (white dots) and 21 dpv (white rhombuses). Ct values equal to or less than 42 were considered to indicate positive results (black line). Antibody response against the E2 glycoprotein detected by a commercial ELISA test (IDEXX) expressed as blocking % (right Y axis); samples with blocking % values more than 40 were considered positive (dotted line) Coronado et al. BMC Veterinary Research (2019) 15:247 Page 9 of 13 Fig. 3 Neutralizing antibody response in piglets. Neutralizing antibody titres were detected by NPLA at 21 dpv (grey bars) and 96 dpv (white bars). Only the litters in which neutralizing antibody titres were detected are shown. * indicates animals that died before 21 dpv. † indicates animals that died between 21 and 96 dpv animals with persistent CSFV infection [20, 22]. This animals showed permanent viremia until the end of the phenomenon explains the immunological anergy in pigs study (96 dpv; more than 4 months of infection), exhibit- with persistent CSFV infection, blocking the immuno- ing the inability to seroconvert. Likewise, given that the logical mechanisms of the host to fight the infection animals in our study belonged to a commercial farm, dif- [20, 26]. Notably, CSFV persistent form was the pre- ferent aetiological agents could be associated with the dominant infection detected in the animals from the clinical signs registered in some of these animals, which present study, including those from litters B, C and F supports the importance of identification of concomitant and some piglets from litter D. infections that may be related, as described above; these Although other porcine pathogens were not diagnosed infections will be identified in future studies. in this study, previous work has shown that, in Cuba, Previous work has shown detection of a low RNA load CSFV is circulating in concomitance with other viral of the lapinized C-strain vaccine by specific qRT-PCR porcine pathogens, such as porcine circovirus 2 (PCV-2) [10] in serum and rectal swab samples soon after vaccin- and porcine parvovirus (PPV) [27]. However, consider- ation in Pestivirus-free vaccinated pigs [21]. Based on ing the high relevance, future studies will be carried out the levels of CSFV RNA loads detected in the rectal to understand the implications of other porcine patho- swabs and tonsils from vaccinated sows, as well as the gens involved in coinfection with CSFV circulating in inconsistent humoral response observed after vaccin- endemic scenarios. Unfortunately, currently, the re- ation in these animals, we determined that five of these sources have been used to detect the presence of sub- animals were asymptomatic CSFV carriers (sows from B clinical forms of CSF, given the high impact of CSFV on to F). Only one vaccinated sow (sow E) was antibody the porcine immune system and given that CSF is con- positive at weaning, although at very low levels. The re- sidered a notifiable disease by the World Organisation sults of the present study indicate that antibody response for Animal Health (OIE). detection and CSFV diagnosis in serum samples from Previous results showed that chronic forms of CSF can subclinically infected CSFV animals may be hindered by be fatal, with clinical manifestations similar to those the coexistence of some antibodies and CSFV [20]. described in the present study, as was the case for pig- Therefore, in addition to detecting the virus in tonsils or lets from litter E [1, 24] (Table 2). This effect is probably tonsil scrapings, these results indicate the effectiveness due to the immunosuppression caused by persistent of using samples such as rectal swabs for the detection CSFV infections [20]. Previous work performed under of subclinical forms of CSF. laboratory conditions has indicated that pigs that exhibit The failure in the response to vaccination in the sows CSFV permanent viremia for more than 30 days post in- was suggested by the lack of maternal immune response fection may develop a similar clinical condition [20, 21]. to protect piglets from either prenatal or postnatal infec- Our data also show the clinical evolution over time of tions. Factors associated with poor handling and animals with persistent CSFV infection. In these ani- malpractice in vaccination have been widely discussed to mals, signs of growth retardation between 21 and 36 dpv explain the constant failures in vaccination in endemic were observed. It was not until after 66 dpv (approxi- areas [11, 15, 16]. Thus, these results prove how vaccine mately 3 months) that, in a sporadic manner, some of failure in the field may allow viral persistence in an en- the animals developed severe symptoms, as previously demic scenario. Improvement in vaccine quality, as well described for the chronic form of CSF [24]. However, as optimization of vaccination dose and schedule, in gilts Coronado et al. BMC Veterinary Research (2019) 15:247 Page 10 of 13 and sows may aid the prevention of CSFV carrier gener- infections in animals are not detected by clinical diagno- ation [28]. Nevertheless, the reason underlying the lack sis and, despite being infected with CSFV, the animals of effective maternal immunity in sows after vaccination are vaccinated, rather than diagnosed and eliminated. must be studied in depth. Animals with persistent and chronic CSFV infections Recently, SIE or homologous interference has been de- are refractory to vaccination, likely due to the SIE scribed as a mechanism that can interfere with CSFV phenomenon. Hence, CSFV eradication in some en- vaccine efficacy in pigs with persistent postnatal CSFV demic areas will require the elimination of persistently infection under experimental conditions [21, 22]. Here, and chronically infected animals by relying heavily on we show that this mechanism may also occur in animals laboratory tools for effective diagnosis. Likewise, CSF suffering from subclinical CSF, either chronic or persist- control programmes may include a vaccination strategy ent, remaining unnoticed at the time of vaccination in that confers sterilizing immunity to avoid the generation field conditions. Pigs with persistent infection from the of subclinical infections in the field. Therefore, govern- present study were unable to seroconvert at any time ment support and financial aid will be key factors to during more than 4 months under study. Likewise, vac- ensure the institution of policies for effective control cination was also ineffective in litters suffering from a and eradication of this disease globally. chronic form of CSF. Therefore, vaccination against sub- clinical forms of CSFV is counterproductive and gener- Methods ates additional expenses for control programmes. Thus, Cells and viruses along with improvement in the vaccination strategy, ef- The porcine kidney cell line PK-15 (ATCC CCL 33, forts must also be made in the diagnosis of these CSF U.S.A.) was cultured in pestivirus-free Dulbecco’s modi- disease forms. fied Eagle medium (DMEM; Lonza, Switzerland) supple- With respect to the nature of the circulating viral mented with 5% foetal bovine serum (FBS) (Euroclone, strain, notably, the consensus sequence identified in the Italy) at 37 °C in 5% CO2. The cell monolayer was incu- full-length E2 gene of the virus recovered from the six bated with 100 μl of serum. Following 1 h of adsorption, litters was the same regardless of the different CSF the serum was removed, and the cells were cultured in forms recorded. Therefore, our results corroborate, in a DMEM supplemented with 2% FBS. After 48 h of incu- field infection, previous results showing CSFV stability bation, a peroxidase-linked assay (PLA) [31] was used [11, 29]. The CSFV Pinar del Rio strain isolated in 2010 for viral identification. (Pinar del Rio/2010) was phylogenetically the closest A lapinized live attenuated C-strain-based vaccine relative to the Pinar del Rio/2016 strain recovered in the (Labiofam strain) belonging to subgenotype 1.1 (batch present study. Previous studies have determined that the number 16001) and used in Cuba since 1965 for vaccin- full-length E2 sequence is the most reliable sequence for ation against CSF [15, 32] was employed in this study. phylogenetic analysis of CSFV [17, 30]. Position 72 (R) This vaccine batch had been previously tested in 33-day- on the E2 protein, previously described in Pinar del Rio/ old naïve pigs, and all the pigs showed a specific CSFV 2010 as a positive selection site for Cuban strains [15], antibody response at 21 dpv, as measured by ELISA was also identified in the Pinar del Rio/2016 isolate. This (IDEXX, Netherlands). finding may suggest the relevance of this codon in the evolution of the CSFV 1.4 genotype in the Cuban Experimental design endemic scenario. Likewise, codon 20 (L), previously This study was performed in a recently repopulated described as a position associated with adaptive advan- Cuban commercial pig farm located in the western tage in the CSFV Pinar del Rio/2010 strain [17], was also region. Six randomly chosen sows and their respective found in the strain from the present study. Moreover, litters (designated A to F) were included. The first-parity codon 195 (N), present in only the CSFV Pinar del Rio/ sows had been vaccinated with one dose of the lapinized 2016 strain, was also identified in silico as a position CSFV C-strain vaccine (Labiofam strain) at weaning and associated with adaptive advantage. Further studies will again at 6 months before delivery. The experiment was clarify the implications of these findings in the evolution initiated at weaning (33 days after birth), at which time, and pathogenesis of CSFV. vaccination of piglets from the farm was carried out. At this time, piglets were identified individually (1 to 62), Conclusion and the distribution among litters was as follows: litter The presence of chronic and persistent CSF infections A (piglets 1 to 9); B (10 to 20); C (21 to 30); D (31 to was demonstrated in an endemic scenario under vaccin- 39); E (40 to 50); F (51 to 62). After identification, sera ation. The CSF persistent infection was the predominant and rectal swabs were collected from the litters and the form in the animals studied. Here, we provide one of the corresponding sows. The sows were euthanized, and few reports showing that, in field conditions, subclinical tonsil samples were collected. Subsequently, the piglets Coronado et al. BMC Veterinary Research (2019) 15:247 Page 11 of 13 were vaccinated using a pig dose, equivalent to 100 pro- sera from piglets were also evaluated at 21, 36, 66 and tective doses (PDs), of a lapinized live attenuated C- 96 dpv. Samples with blocking percentage values below strain-based vaccine (Labiofam strain) by intramuscular 30% were considered negative, those with values be- injection in the neck. At 21 dpv, sera and rectal swabs tween 30 and 40% were considered doubtful and those were obtained from the 62 piglets. The piglets from with values ≥40% were considered positive. litters B, C and E were then euthanized, and tonsils were collected. Sera and rectal swabs were obtained from the Full-length E2 encoding sequence determination animals from the remaining litters (A, D and F) at 36, 66 The sequence of the full-length E2 region was deter- and 96 dpv (end of experiment), and tonsil samples were mined from tonsils obtained from one piglet of each obtained. Additionally, clinical signs were registered by litter as previously described [30]. The one-step RT-PCR the veterinarian from the farm. protocol was performed using the commercially available All pigs were handled according to the rules described One-Step RT-PCR Kit (Qiagen, Netherlands). The se- in the manual of good practices for porcine production quencing reactions were conducted under the BigDye™ from the Cuban Republic [33]. International standards terminator-cycling conditions (Thermo Fisher Scientific, for animal welfare were used following the regulations of U.S.A.) using an ABI 3130 × L instrument. Forward and the Animal Health Directory (DSA). The Ethics Com- reverse sequences obtained from each amplicon were mittee of the Cuban Ministry of Agriculture (MINAGRI) assembled using the Contig Express application in approved the study, and all efforts were made to Vector NTI software, version 11 (Invitrogen, U.S.A.). minimize animal suffering. Animals were euthanized by bleeding by inserting an appropriately sized knife into Sequence analysis major artery following stunning, according to the Multiple sequence alignment was conducted for all avail- current protocol described in the OIE Terrestrial Animal able full-length E2 gene sequences from Cuban strains Health Code [34]. (CSFV/Margarita/1958 [JX028201, AJ704817], CSFV/ According to the rules for the CSF control programme Santiago de Cuba/2011 [JX028203], CSFV/Holguin/ in the country, in case of positive CSFV detection, only 2009[JX028202], CSFV/Pinar del Rio/2010 [JX028204, the animals under study were to be maintained following KX576461], and the new sequences obtained in the strict biosecurity measures, and the remaining animals present study). Multiple sequence alignment was per- from the farm were to be eliminated. formed using the algorithm Clustal W (EMBL-EBI, United Kingdom) included in the program BioEdit Se- Detection of CSFV RNA and virus isolation quence Alignment Editor [37]. Positive selection analysis RNA was extracted from all of the samples using the from the complete E2 gene was carried out using several RNeasy Mini Kit (Qiagen, Netherlands) according to the models (M0, M1, M2, M7 and M8) available in the manufacturer’s instructions. The presence of CSFV RNA CODEML module of the PAML 4.7 software package in sera, rectal swabs and tonsil samples was analysed by [38]. Finally, functional divergence among the Cuban se- qRT-PCR [23]. Threshold cycle values (Ct) equal to or quences was assessed by estimation of the coefficients of less than 42 were considered to indicate positive results. type-I functional divergence (θI) using DIVERGE 3.0 Samples in which fluorescence was undetectable were software [39]. The 3D structure and surface of the E2 considered negative. Ct values below 23 were considered protein was constructed as previously described [17] for high, values from 23 to 28 were considered moderate, the Cuban ancestral strain “Margarita/1958” [AJ704817] and values from 29 to 42 were considered low RNA viral [11], the previously described “Pinar del Rio/2010” loads [35]. Serum samples that were positive by qRT- [KX576461] strain [11] and the new strain obtained in PCR at 21 dpv were tested by virus isolation in PK15 the present study. cells, as described above. E2-specific and neutralizing antibody detection Additional file Piglet serum samples were tested with a neutralization PLA (NPLA) at weaning (33 days old, 0 dpv) and at 21 Additional file 1: Figure S1. Representation of the mutation pattern for Cuban CSFV strains on the 3D structure of the predicted model of the E2 and 96 dpv [36]. The titres were expressed as the recip- protein of CSFV. The 3D structure and surface of the Cuban ancestral rocal dilution of serum that neutralized 100 TCID50 of strain “Margarita/1958” [AJ704817], the previously described “Pinar del the Margarita strain in 50% of the culture replicates. Rio/2010” [KX576461] strain [11] and the new strain “Pinar del Rio/ 2016”[LT985811] are presented. The in silico mutations associated with Detection of E2-specific antibodies was performed using positive selection (red), associated with adaptation (blue) and not linked a commercial ELISA kit (IDEXX, Netherlands) following previously with any evolutionary advantage (yellow) are denoted. The the manufacturer’s recommendations in serum samples antigenic region B/C (gold), antigenic region A/D (ruby red) and transmembrane domain (grey) are located on monomer B. (JPG 94 kb) collected at weaning from sows and piglets. In addition, Coronado et al. BMC Veterinary Research (2019) 15:247 Page 12 of 13 Abbreviations 3. Weesendorp E, Backer J, Stegeman A, Loeffen W. Transmission of classical CSF: Classical swine fever; CSFV: Classical swine fever virus; Ct: Threshold swine fever virus depends on the clinical course of infection which is cycle; DIVA: Differentiating infected from vaccinated animals; associated with high and low levels of virus excretion. Vet Microbiol. 2011; DMEM: Dulbecco’s modified Eagle medium; dpv: Days post vaccination; 147:262–73. https://doi.org/10.1016/j.vetmic.2010.06.032. DSA: Animal Health Directory; FBS: Foetal bovine serum; MINAGRI: Ministry of 4. Aynaud JM, Corthier G, Vannier P, Tillon PJ. Swine fever: in vitro and in vivo Agriculture; NPLA: Neutralization peroxidase-linked assay; PCV-2: Porcine properties of low virulent strains isolated in breeding farms having circovirus 2; PD: Protective dose; PK-15 cells: Porcine kidney-15 cell line; reproductive failures. In: Liess B, editor. Proceedings of the agricultural PLA: Peroxidase-linked assay; PPV: Porcine parvovirus; qRT-PCR: Quantitative research seminar on hog cholera. hog cholera. classical swine fever and reverse transcription polymerase chain reaction; RNA: Ribonucleic acid; african swine fever. Brussels: Commission of the European Communities, TCID50: 50% tissue culture infective dose Publication EUR 5904 EN; 1977. p. 273–7. 5. Carbrey EA, Stewart WC, Kresse JI, Snyder ML. Inapparent hog cholera Acknowledgements infection following the inoculation of field isolates. In: Lies B, editor. The authors thank Dr. Félix Dominguez and all the field veterinarians who Proceedings of the agricultural research seminar on hog cholera. Hog helped in the development of the present work. cholera. Classical swine fever and African swine fever. Brussels: Commission of the European Communities, Publication EUR 5904 EN; Authors’ contributions 1977. p. 214–30. LG and MTF conceived, designed and coordinated this study. LC, LJP, LD, CLP 6. van Oirschot JT. Experimental production of congenital persistent swine and MTF performed the experiments. LC, JAB, SM, LJP, CLP, OF, MTF, RR and LG fever infections. I Clinical, pathological and virological observations. Vet analysed the data. LG, LJP, CLP contributed reagents/materials/analysis tools. LC, Microbiol. 1979;4:117–32. JAB, MTF and LG wrote the paper. SM, CLP, and LG critically read the 7. van Oirschot JT. Experimental production of congenital persistent swine manuscript. All authors have read and approved the final manuscript. fever infections: II Effect on functions of the immune system. Vet Microbiol. 1979;4:133-47. Funding 8. van Oirschot JT. Vaccinology of classical swine fever: from lab to field. Vet This research was supported by Program I.1: Research grants for and about Microbiol. 2003;96:367–84. development from the MAEC–AECID programme of the Spanish government 9. Ganges L, Nuñez JI, Sobrino F, Borrego B, Fernández-Borges N, Frías- and by grants AGL2015–66907 and RTI2018-100887-B-100 from the Spanish Lepoureau MT, et al. Recent advances in the development of recombinant government; the CERCA Programme, Generalitat de Catalunya, is also acknowl- vaccines against classical swine fever virus: cellular responses also play a edged for funding this research. The funding bodies did not play a role in the role in protection. Vet J. 2008;177:169–77. design of the study; collection, analysis, and interpretation of the data; or writ- 10. Liu L, Xia H, Everett H, Sosan O, Crooke H, Meindl-Böhmer A, et al. A generic ing of the manuscript. real-time TaqMan assay for specific detection of lapinized Chinese vaccines against classical swine fever. J Virol Methods. 2011;175:170–4. Availability of data and materials 11. Coronado L, Liniger M, Muñoz-Gonzalez S, Postel A, Perez LJ, Perez- The datasets used and/or analysed in the present study are available from Simó M, et al. Novel poly-uridine insertion in the 3’UTR and E2 amino the corresponding author on reasonable request. acid substitutions in a low virulent classical swine fever virus. Vet Microbiol. 2017; In press. Ethics approval and consent to participate 12. Blome S, Moß C, Reimann I, König P, Beer M. Classical swine fever The animals in this study were from a recently repopulated Cuban vaccines—state-of-the-art. Vet Microbiol. 2017;206:10–20. commercial farm located in the western region. The authorities responsible 13. Graham SP, Haines FJ, Johns HL, Sosan O, Anna S, Rocca L, et al. for the farm provided written consent allowing the animals to be used in Characterisation of vaccine-induced, broadly cross-reactive IFN-γ secreting T this study. International standards for animal welfare were used following the cell responses that correlate with rapid protection against classical swine regulations of the Animal Health Directory (DSA). The Ethics Committee of fever virus. Vaccine. 2012;30:2742–8. the Cuban Ministry of Agriculture (MINAGRI) of the Cuban government 14. Suradhat S, Intrakamhaeng M, Damrongwatanapokin S. The correlation of approved the study, and all efforts were made to minimize animal suffering. virus-specific interferon-gamma production and protection against classical swine fever virus infection. Vet Immunol Immunopathol. 2001;83:177–89. Consent for publication https://doi.org/10.1016/S0165-2427(01)00389-0. Not applicable. 15. Pérez LJ, Díaz de Arce H, Perera CL, Rosell R, Frías MT, Percedo MI, et al. Positive selection pressure on the B/C domains of the E2-gene of classical Competing interests swine fever virus in endemic areas under C-strain vaccination. Infect Genet The authors declare that they have no competing interests. Evol. 2012;12:1405–12. 16. Ji W, Niu D-DD, Si H-LL, Ding N-ZZ, He C-QQ. Vaccination influences the Author details evolution of classical swine fever virus. Infect Genet Evol. 2014;25:69–77. 1 17. Rios L, Coronado L, Naranjo-Feliciano D, Martínez-Pérez O, Perera CL, National Centre for Animal and Plant Health (CENSA), OIE Collaborating Centre for Disaster Risk Reduction in Animal Health, San José de las Lajas, Hernandez-Alvarez L, et al. Deciphering the emergence, genetic diversity Mayabeque, Cuba. 2OIE Reference Laboratory for Classical Swine Fever, and evolution of classical swine fever virus. Sci Rep. 2017;7:17887. https:// IRTA-CReSA, Campus de la Universitat Autònoma de Barcelona, 08193 doi.org/10.1038/s41598-017-18196-y. Bellaterra, Spain. 3University of Illinois, College of Veterinary Science, 18. Yoo SJ, Kwon T, Kang K, Kim H, Kang SC, Richt JA, et al. Genetic evolution of Department of Clinical Veterinary Medicine, Urbana, Illinois 61802, United classical swine fever virus under immune environments conditioned by States. 4Departament d’Agricultura Ramaderia i Pesca (DARP), Centre de genotype 1-based modified live virus vaccine. Transbound Emerg Dis. 2018: Recerca en Sanitat Animal (CReSA, IRTA-UAB), Campus de la Universitat 1–11. https://doi.org/10.1111/tbed.12798. Autònoma de Barcelona, 08193 Bellaterra, Spain. 19. Cabezón O, Colom-Cadena A, Muñoz-González S, Pérez-Simó M, Bohórquez JA, Rosell R, et al. Post-Natal persistent infection with classical swine fever Received: 29 June 2018 Accepted: 26 June 2019 virus in wild boar: a strategy for viral maintenance? Transbound Emerg Dis. 2015. https://doi.org/10.1111/tbed.12395. 20. Muñoz-González S, Ruggli N, Rosell R, Pérez LJ, Frías-Leuporeau MT, Fraile L, References et al. Postnatal persistent infection with classical swine fever virus and its 1. Blome S, Staubach C, Henke J, Carlson J, Beer M. Classical swine fever—an immunological implications. PLoS One. 2015;10:e0125692. https://doi.org/1 updated review. Viruses. 2017;9:1–24. https://doi.org/10.3390/v9040086. 0.1371/journal.pone.0125692. 2. Simmonds P, Becher P, Collett M, Gould EA, Heinz FX, Meyers G, et al. 21. Muñoz-González S, Perez-Simó M, Muñoz M, Bohorquez JA, Rosell R, Family Flaviviridae. In: King AMQ, Adams MJ, Carstens EB, Lefkowitz EJ, Summerfield A, et al. Efficacy of a live attenuated vaccine in classical swine editors. Ninth report of the international Commitee on taxonomy of viruses. fever virus postnatally persistently infected pigs. Vet Res. 2015;46:78. https:// Ninth. San Diego: Elsevier Academic Press; 2012. p. 1004–20. doi.org/10.1186/s13567-015-0209-9. Coronado et al. BMC Veterinary Research (2019) 15:247 Page 13 of 13 22. Muñoz-González S, Pérez-Simó M, Colom-Cadena A, Cabezón O, Bohórquez JA, Rosell R, et al. Classical swine fever virus vs. classical swine fever virus: the superinfection exclusion phenomenon in experimentally infected wild boar. PLoS One. 2016;11:e0149469. 23. Hoffmann B, Beer M, Schelp C, Schirrmeier H, Depner K. Validation of a real- time RT-PCR assay for sensitive and specific detection of classical swine fever. J Virol Methods. 2005;130:36–44. 24. Moennig V, Floegel-Niesmann G, Greiser-Wilke I. Clinical signs and epidemiology of classical swine fever: a review of new knowledge. Vet J. 2003;165:11–20. https://doi.org/10.1016/S1090-0233(02)00112-0. 25. Liess B. Persistent infections of hog cholera: a review. Prev Vet Med. 1984;2: 109–13. https://doi.org/10.1016/0167-5877(84)90054-0. 26. Muñoz-González S, Sordo Y, Pérez-Simó M, Suárez M, Canturri A, Rodriguez MP, et al. Efficacy of E2 glycoprotein fused to porcine CD154 as a novel chimeric subunit vaccine to prevent classical swine fever virus vertical transmission in pregnant sows. Vet Microbiol. 2017. doi:https://doi.org/10.1 016/j.vetmic.2017.05.003. 27. Pérez LJ, de AHD, Frias MT, Perera CL, Ganges L, Núñez JI. Molecular detection of torque Teno sus virus in lymphoid tissues in concomitant infections with other porcine viral pathogens. Res Vet Sci. 2011;91:e154–7. https://doi.org/10.1016/J.RVSC.2011.02.012. 28. Anonymous. Chapter 2.8.3 classical swine fever. In: OIE terrestrial manual. Paris: World Organisation for Animal Health; 2014. 29. Jenckel M, Blome S, Beer M, Höper D. Quasispecies composition and diversity do not reveal any predictors for chronic classical swine fever virus infection. Arch Virol. 2017;162:775–86. https://doi.org/10.1007/s00705-016-3161-8. 30. Postel A, Schmeiser S, Bernau J, Meindl-Boehmer A, Pridotkas G, Dirbakova Z, et al. Improved strategy for phylogenetic analysis of classical swine fever virus based on full-length E2 encoding sequences. Vet Res. 2012;43:50. https://doi.org/10.1186/1297-9716-43-50. 31. Wensvoort G, Terpstra C, Boonstra J, Bloemraad M, Van ZD. Production of monoclonal antibodies against swine fever virus and their use in laboratory diagnosis. Vet Microbiol. 1986;12:101–8 Elsevier Sci Publ BV. 32. Díaz de Arce H, Núñez JI, Ganges L, Barreras M, Teresa Frías M, Sobrino F. Molecular epidemiology of classical swine fever in Cuba. Virus Res. 1999;64:61–7. 33. Anonymous. Manual de procedimientos técnicos para la crianza porcina. Instituto de Investigaciones Porcinas; 2008. https://books.google.es/books/ about/Manual_de_procedimientos_técnicos_para.html?id= zyxsMwEACAAJ&redir_esc=y. Accessed 16 Mar 2018. 34. Anonymous. Chapter 7.5. Slaughter of animals. In: Terrestrial animal health code. Twenty-sev. Paris: World Organisation for Animal Health; 2018. p. 369–87. 35. Tarradas J, Monsó M, Fraile L, de la Torre BG, Muñoz M, Rosell R, et al. A T- cell epitope on NS3 non-structural protein enhances the B and T cell responses elicited by dendrimeric constructions against CSFV in domestic pigs. Vet Immunol Immunopathol. 2012;150:36–46. https://doi.org/10.1016/j. vetimm.2012.08.006. 36. Terpstra C, Bloemraad M, Gielkens AL. The neutralizing peroxidase-linked assay for detection of antibody against swine fever virus. Vet Microbiol. 1984;9:113–20. 37. Hall TA. BioEdit: a user-friendly biological sequence alignment editor and analysis program for windows 95/98/NT. Nucleic Acids Symp. 1999;41:95–8. 38. Yang Z. PAML 4: phylogenetic analysis by maximum likelihood. Mol Biol Evol. 2007;24:1586–91. https://doi.org/10.1093/molbev/msm088. 39. Gu X, Zou Y, Su Z, Huang W, Zhou Z, Arendsee Z, et al. An update of DIVERGE software for functional divergence analysis of protein family. Mol Biol Evol. 2013;30:1713–9. https://doi.org/10.1093/molbev/mst069. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.