animals Article A Methodology to Quantify Resilience in Growing Pigs Houda Laghouaouta 1 , Ramona N. Pena 1 , Roger Ros-Freixedes 1 , Josep Reixach 2 , Marta Díaz 2 , Joan Estany 1 , Ramon Armengol 1 , Anna Bassols 3 and Lorenzo Fraile 1, * 1 Agrotecnio-CERCA Center, Department of Animal Science, University of Lleida, 25198 Lleida, Spain; houda.laghouaouta@udl.cat (H.L.); romi.pena@udl.cat (R.N.P.); roger.ros@udl.cat (R.R.-F.); joan.estany@udl.cat (J.E.); ramon.armengol@udl.cat (R.A.) 2 Selección Batallé SA, 17421 Riudarenes, Spain; jreixach@batalle.com (J.R.); mdiaz@batalle.com (M.D.) 3 Departament de Bioquímica, Universitat Autónoma de Barcelona, 08193 Barcelona, Spain; anna.bassols@uab.cat * Correspondence: lorenzo.fraile@udl.cat; Tel.: +34-973-70-28-14 Simple Summary: The concept of resilience can be defined as the ability of animals to maintain their productivity despite the stressors that might occur during the rearing period. Despite the growing interest in the genetic determinism of resilience and its possible inclusion in selection objectives, there is no straightforward method to measure this trait. Several quantifying methods estimate resilience based on productivity such as body weight or milk production, or non-productivity related traits such as immunity or stress responses. The objective of this study was to elaborate novel resilience indicators in growing pigs based on both productivity (body weight) and non-productivity (acute-phase proteins) related traits. We propose the deviation from the expected growth curve and the increment of the acute-phase protein haptoglobin, after applying a common vaccine, as resilience indicators in growing pigs under standard rearing conditions. We showed that the suggested  resilience indicators are under a genetic control, show a substantial variability in the population, and  may be improved through selection. Citation: Laghouaouta, H.; Pena, R.N.; Ros-Freixedes, R.; Reixach, J.; Abstract: There is a growing concern about the genetic determinism of resilience and its possible Díaz, M.; Estany, J.; Armengol, R.; implementation in breeding programs. The objective of our study was to elaborate novel resilience Bassols, A.; Fraile, L. A Methodology indicators in growing pigs based on the deviation from the expected growth curve and the increment to Quantify Resilience in Growing of the acute-phase protein haptoglobin (HP) after applying a common vaccine. A total of 445 pigs Pigs. Animals 2021, 11, 2970. were vaccinated with an attenuated Aujeszky vaccine at 12 weeks of age. Deviation from the expected https://doi.org/10.3390/ ani11102970 body weight (∆BW) given the growth curve of unvaccinated pigs at 28 days post-vaccination (DPV) and the increment of HP at 4 DPV (∆HP) were suggested as resilience indicators. Challenged pigs Received: 6 September 2021 that maintained their productivity and had a minor activation of HP were deemed resilient, whereas Accepted: 13 October 2021 pigs that had low ∆BW values and a high activation of HP were deemed susceptible. Pigs were also Published: 15 October 2021 classified based on ∆BW and ∆HP relative to the expected BW at 28 DPV and to the basal level of HP, respectively. The concordance was high between both methods, indicating that ∆BW and ∆HP are Publisher’s Note: MDPI stays neutral not sensitive to the animal’s expected BW nor the basal level of HP. The heritability estimates were with regard to jurisdictional claims in moderate for ∆BW (0.33) and low-to-moderate for ∆HP (0.16). Our study suggests ∆BW and ∆HP as published maps and institutional affil- novel resilience indicators in pigs. The suggested indicators capture different aspects of resilience, iations. are easy to measure, and are genetically controlled. Thus, they may be improved through selective breeding. Further analyses are needed to validate our findings. Keywords: body weight; haptoglobin; pigs; resilience indicators; vaccine challenge Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and 1. Introduction conditions of the Creative Commons Attribution (CC BY) license (https:// Pork is one of the most important sources of animal protein for humans and a sustained creativecommons.org/licenses/by/ increase in pig production will be necessary to cope with the challenge of providing 4.0/). worldwide food security [1]. This challenge can be tackled by increasing the number of Animals 2021, 11, 2970. https://doi.org/10.3390/ani11102970 https://www.mdpi.com/journal/animals Animals 2021, 11, 2970 2 of 12 animals and by improving the efficiency of the pork production sector. Each animal is exposed to a diversity of stimuli that arise from its internal and external environments and it is desirable for the animal’s welfare and its commercial productivity, that it, has the capacity to cope with these challenges. This defines the concept of resilience, that is, the ability of the animal to be minimally affected by all the perturbations that might occur during its productive life and quickly return to the physiological, behavioral, cognitive, health, affective, and production states that pertained before exposure to a disturbance [2]. In recent years, there has been a growing interest in the genetic determinism of resilience and its possible inclusion in selection objectives. Selective breeding for improved resilience would provide animals with more robust phenotypes. However, there is no straightforward quantification method for resilience. Indirect indicators that reflect the disturbance caused by stressors such as residual variance or variation coefficients have been used as proxies to phenotype this complex trait [3]. Although there is a general agreement that internal and external challenges nega- tively impact productivity through the dysregulation of basic biological mechanisms [4], there is no consensus on how to measure resilience. Several quantification methods es- timate resilience based on productivity-related traits such as body weight (BW) in layer chickens [5], feed intake in pigs [6], litter size in rabbits [7], and milk yield in cattle [8,9]. All these methods assume that resilient animals are able to maintain their productivity despite stressors. The relationship between resilience and immunity has also been well documented in the literature. Stressors modulate the immune system [10,11] and increase disease sus- ceptibility [12]. Resilient and susceptible animals respond differently toward stressors and express different immune phenotypes [10]. In this sense, several resilience indicators have been assessed through non-productivity related traits such as medication records [13] or antibody levels [14]. In this context, acute-phase proteins are frequently measured in relation to the innate immune response. These are a group of blood proteins that change their concentration in animals subjected to infection, inflammation, surgical trauma, or stress [15]. During the first days after the challenge, there is an increase in the plasma concentration of positive acute-phase proteins such as haptoglobin (HP) and a decrease in negative acute-phase proteins. Serum acute-phase protein levels have been used as biomarkers for the presence and extent of disease processes [16]. Given the strong relation- ship between immune state, welfare, and productivity, lower serum acute-phase proteins have been correlated with better production parameters in pigs [17,18] and therefore could be used as indicators of resilience. The objective of the present study was to evaluate novel resilience indicators in grow- ing pigs based on both the deviation from the expected growth curve and the increment of the acute-phase protein HP after applying a common vaccine under standard rearing conditions, with the goal of evaluating the capacity of these indicators for classifying animals as resilient or susceptible. 2. Materials and Methods 2.1. Animals and Experimental Design Animals belonged to a commercial Duroc line reared in high-health status farms from Selección Batallé SA (Riudarenes, Northeast of Spain). A total of 540 barrows at approximately 10 weeks of age (71.4 ± 2.4 days) were identified, ear-tagged, and randomly distributed into five fattening batches, ensuring a stock density of 1 m2 per animal. Pigs were reared under the same conditions and fed ad libitum with commercial diets. After 14 days (85.6 ± 2.4 days of age), 445 pigs (experimental group, E) were intramuscularly vaccinated with an attenuated Aujeszky vaccine (Auskipra, batch 9R83, Laboratorios Hipra, Amer, Girona) and 95 pigs (control group, C) were inoculated with phosphate-buffered saline. Live attenuated Aujeszky vaccines are widely used in swine medicine preventive programs across Europe due to the risk of transmission of wild-type Aujeszky virus from wild boar to domestic pigs. Thus, many animals are routinely vaccinated across Europe Animals 2021, 11, x 3 of 13 Animals 2021, 11, 2970 Live attenuated Aujeszky vaccines are widely used in swine medicine preventive 3pro- of 12 grams across Europe due to the risk of transmission of wild-type Aujeszky virus from wild boar to domestic pigs. Thus, many animals are routinely vaccinated across Europe with this vaccine and the measurement of haptoglobin, following this vaccination, could with this vaccine and the measurement of haptoglobin, following this vaccination, could be easily carried out without adding any additional vaccine to the pig population. Each be easily carried out without adding any additional vaccine to the pig population. Each rearing batch contained E and C pigs, and half-sibs were evenly distributed between the rearing batch contained E and C pigs, and half-sibs were evenly distributed between the E and C groups. There is no risk of transmissibility of the Aujeszky virus between E and E and C groups. There is no risk of transmissibility of the Aujeszky virus between E and C pigs due to the well-known safety profile of the vaccine [19]. Experimental individuals C pigs due to the well-known safety profile of the vaccine [19]. Experimental individuals were offspring from 49 sires and 198 dams (around 14 sires and 40 dams per batch). were offspring from 49 sires and 198 dams (around 14 sires and 40 dams per batch). Body Body weight weight was was registered registered at at −14, −14,0,0,and and2828days dayspost-vaccination post-vaccination (DPV) (DPV) and and blood blood samples sampleswerewerecollected collectedfrom fromthe thejugular jugularvein veinatat 4 DPV 4 DPV(Figure 1). 1). (Figure In addition, 81 pigs In addition, (40 81 pigs from the C group and 41 from the E group) were bled at 0 DPV to establish the basal (40 from the C group and 41 from the E group) were bled at 0 DPV to establish the basal level level of of the the acute-phase acute-phase protein protein HPHP in in each batch. Average each batch. daily gain Average daily gain (ADG) (ADG) between between − −14 14 and and 28 28 DPV was estimated. At the end of the fattening period, animals were slaughtered at DPV was estimated. At the end of the fattening period, animals were slaughtered at approximately approximately 30 30 weeks weeks ofof age age (217.8 (217.8 ±±3.7 3.7days) days)following followingthethestandard standardprocedures procedures ofof the the production production company. company. Figure 1. Figure Timeline and 1. Timeline and experimental experimental design. design. 2.2. Determination of the Acute-Phase Protein Haptoglobin 2.2. Determination of the Acute-Phase Protein Haptoglobin Serum HP was quantified at 0 and 4 DPV by using a spectrophotometric method Serum HP was quantified at 0 and 4 DPV by using a spectrophotometric method (hemoglobin binding assay) with the Tridelta PHASE Haptoglobin Assay (Tridelta Devel- (hemoglobin binding assay) with the Tridelta PHASE Haptoglobin Assay (Tridelta Devel- opment Ltd., County Kildare, Ireland) and performed on an automated analyzer (Olympus opment Ltd., County Kildare, Ireland) and performed on an automated analyzer (Olym- AU400, Hamburg, Germany). Intra-assay and inter-assay coefficients of variation for this pus AU400, Hamburg, Germany). Intra-assay and inter-assay coefficients of variation for technique have been previously reported [20]. this technique have been previously reported [20]. 2.3. Proposal of Resilience Indicators 2.3. Proposal of Resilience Indicators The experiment was carried out in commercial farms under standard rearing condi- The experiment tions. Both BW and HP wasdata carried wereout in commercial analyzed farms to quantify theunder standard rearing condi- pigs’ resilience. tions. Both BW and HP data were analyzed to quantify the pigs’ resilience. 2.3.1. Body Weight Deviation 2.3.1. BW Body Weight data from Deviation the C group were analyzed using the following model: BW data from the C group were analyzed using the following model: BW28 = B + αBW−14 + βBW0 + γA + e BW 28 = B + α BW −14 + βBW 0 + γA + e where BW28 is the BW at 28 DPV; B is the batch fixed effect with five levels; BW−14 , BW0 , where and A BWare 28 theis covariates the BW at of28BW DPV;at B is the −14 DPV,batch BW fixed effectand at 0 DPV, with agefive levels; at 28 DPV,BW −14,and α, β, BW0γ, and A are are the the covariates regression of BWand coefficients; at −14 e isDPV, BW at 0term. the residual DPV, and age at 28 DPV, α, β, and γ are the regression For each pigcoefficients; from the Eand e is the group, BWresidual term. 28 was estimated using the previous model (R2 =For each 85%). pigwas This from the E group, considered BW28 wasBW the expected estimated 28 given using the the previous growth curve of model control(R 2 = pigs. 85%). This was(∆BW) A differential considered the expected was calculated BW as the 28 given the difference growth between thecurve of control observed pigs. A and expected BW28 (∆BW(ΔBW) differential = observed BW28 − expected was calculated BW28 ). Besides, as the difference between%BW was alsoand the observed estimated expected as the ratio between ∆BW and the expected BW28 to have a relative measure of ∆BW. We defined resilient pigs as those minimally affected by the vaccination process and all the uncontrolled events that could have also occurred during the experiment. Resilient pigs were those that had high values of ∆BW and %BW. Animals 2021, 11, 2970 4 of 12 Besides, for each pig from the E group, BW28 was also estimated based on the pig’s ADG before the vaccine challenge as: BW0 − BW−14 BW28 = BW0 + d0,28 d−14,0 BW −BW 0 −14 where d−14,0 is the pig’s ADG before the vaccine challenge (between −14 and 0 DPV) and d0,28 is the number of days between 0 and 28 DPV. For each pig from the E group, a differential ∆BWADG was calculated as the difference between its observed and ex- pected BW28 . 2.3.2. Increment of Haptoglobin after the Vaccine Challenge Data of the acute-phase protein HP were analyzed within each batch. A previous work reported that among the five acute-phase proteins (C-reactive protein, serum amyloid A, HP, pig-MAP, and albumin), HP is the most sensitive acute-phase protein for Aujeszky disease [21] and that the peak of its response occurs four days after the infection [22]. Therefore, the increment of HP at 4 DPV (∆HP) was calculated as the difference between the pig’s HP concentration at 4 DPV and the mean of the HP concentration of the cor- responding batch at 0 DPV. In addition, %HP was calculated as the ratio between ∆HP and the basal level of HP within the batch. As the plasma concentration of HP increases after the challenge [15], resilient pigs should show lower values of ∆HP and %HP than susceptible pigs. 2.4. Classification of Pigs into Resilient and Susceptible Groups Pigs from the E group were classified as resilient (R), average (A), or susceptible (S) based on BW and HP data. Four different classifications were performed. 1. Pigs were classified based on the combination of resilience indicators ∆BW and ∆HP as suggested by Bai et al. [23]. Pigs with ∆BW above the third quartile (Q3 ) and ∆HP below the first quartile (Q1 ) were classified as R. In contrast, pigs with ∆BW below Q1 and ∆HP above Q3 were classified as S. The rest of the animals were considered as A. 2. Pigs were classified based on %BW and %HP following the same previous quartile distribution to examine whether ∆BW and ∆HP are sensitive to the animal’s expected BW and the basal level of HP. 3. Pigs were classified based on ∆BWADG and ∆HP following the same previous quar- tile distribution. 4. Pigs were classified based on the observed BW28 to examine whether ∆BW and ∆HP captured more information than the differences in the observed BW28 . The agreement between the different classifications was assessed using a concordance test with the kappa statistic [24]. Correlation between the suggested indicators was assessed using the R software. 2.5. Heritability Estimates Heritability estimates of the suggested resilience indicators (∆BW, %BW, ∆HP, and %HP) were estimated using the following univariate animal model: y = Xb + Za + e where y stands for ∆BW, %BW, ∆HP, and %HP; b is the vector of the batch fixed effect and the covariate of age at 28 DPV; a is the vector of additive effects; X and Z are incidence matrices; and e is the vector of residual terms. Additive genetic effects (a) and residuals (e) were assumed to be normally distributed as a∼N (0, Aσa 2 ) and e∼N (0, Iσe 2 ), where A is the relationship matrix calculated with a five-generation pedigree of 3467 pigs, σa 2 is the genetic additive variance, I is an identity matrix, and σe 2 is the residual variance. All effects were assumed to be independent between them. Marginal posterior distributions were Animals 2021, 11, 2970 5 of 12 estimated using Monte Carlo Markov chains of 1,000,000 samples with a burn-in period of 200,000, of which one in every 100 samples was retained. The heritability was estimated as the mean of the marginal posterior distribution. The probability (P0.10 ) of h2 being greater than 0.10 and the highest posterior density interval at 95% of probability (HPD95% ) were also estimated. Analyses were performed using the TM software [25]. 3. Results Descriptive statistics for the ADG between −14 and 28 DPV are displayed in Table 1. Overall, pigs from the C group showed a slightly greater ADG than challenged pigs except for batch 1. The ADG (coefficient of variation; %) ranged from 594.3 (24.9) to 756.2 (10.6) g/day for the E group and from 556.2 (21.5) to 794.7 (4.9) g/day for the C group. The lowest ADG were observed in batches 1, 2, and 5. The HP concentration at 4 DPV was variable between and within batches ranging from 0.70 (70.5) to 1.95 (42) mg/mL for the E group and from 1.09 (54.7) to 2.07 (33.1) mg/mL for the C group (Table 1). The highest HP concentrations at 4 DPV were reported in batches 2, 3, and 5. The coefficients of variation for ADG and HP were high, indicating a considerable variability between animals, and in general, higher in the E group than in the C group. Table 1. Mean (coefficient of variation; %) for the average daily gain (ADG; g/day) between −14 and 28 days post-vaccination (DPV) and the serum haptoglobin concentration (HP; mg/mL) at 4 DPV. Group Experimental Group Control Group Batch N ADG HP N ADG HP 1 84 594.3 (24.9) 0.70 (70.5) 20 556.2 (21.5) 1.09 (54.7) 2 95 624.8 (17.0) 1.48 (58.0) 16 707.7 (14.1) 1.59 (31.9) 3 86 756.2 (10.6) 1.95 (42.0) 21 768.4 (10.7) 2.07 (33.1) 4 86 756.0 (11.9) 1.08 (55.3) 23 794.7 (4.9) 1.07 (53.5) 5 94 662.4 (16.7) 1.40 (37.3) 15 749.3 (16.2) 1.30 (39.3) All batches 445 677.7 (18.8) 1.33 (59.4) 95 716.9 (17.8) 1.42 (49.1) 3.1. Descriptive Statistics for the Novel Resilience Indicators Descriptive statistics of the suggested resilience indicators are given in Table 2. Aver- age ∆BW and %BW were −0.68 kg and −1.42%, respectively, indicating that on average, the observed BW of challenged pigs at 28 DPV was lower than the expected BW given the theoretical growth curve. Average ∆HP and %HP were +0.03 mg/mL and +5.40%, respec- tively, showing an increment of HP concentration in plasma at 4 DPV. All the resilience indicators had high SD values. Table 2. Descriptive statistics for the resilience indicators in pigs from the experimental group. Trait Mean SD 1 Min Max ∆BW (kg)2 −0.68 3.64 −13.2 +10.1 %BW 3 (%) −1.42 7.26 −24.4 +19.3 ∆HP 4 (mg/mL) +0.03 0.70 −1.41 +2.65 %HP 5 (%) +5.40 60.4 −89.2 +292 1Standard deviation; 2 Body weight deviation from the expected growth curve of control pigs at 28 days post- vaccination (DPV); 3 Ratio between ∆BW and the expected body weight at 28 DPV given the growth curve of control pigs; 4 Haptoglobin increment at 4 DPV; 5 Ratio between ∆HP and the basal level of haptoglobin. Phenotypic correlations between the resilience indicators are reported in Table 3. A negative and low correlation was reported between ∆BW and ∆HP (r = −0.09, p < 0.05), suggesting that they capture different aspects of resilience. ∆BW and ∆BWADG were moderately correlated (r = 0.52, p < 0.001). basal level of haptoglobin. Phenotypic correlations between the resilience indicators are reported in Table 3. A negative and low correlation was reported between ∆BW and ∆HP (r = −0.09, p < 0.05), suggesting that they capture different aspects of resilience. ∆BW and ∆BWADG were mod- Animals 2021, 11, 2970 6 of 12 erately correlated (r = 0.52, p < 0.001). Table 3. Correlations between the resilience indicators. Table 3. Correlations between the resilience 2indicators. Trait ∆BW1 ∆HP %BW 3 %HP 4 ∆BWADG 5 ∆BW Trait ∆BW 1 1 ∆HP 2* −0.09 0.99 %BW*** 3 −0.14 %HP ** 4 ∆BW *** 5 0.52ADG ∆HP ∆BW 1 1 * −0.09 −0.09 ns 0.99 *** 0.94 − 0.14*** ** −0.02 ns 0.52 *** ∆HP %BW 1 1 −0.09 ns −0.14 ** 0.94 *** 0.51 − ns 0.02*** %BW %HP 1 −0.14 1 ** 0.51 ***ns −0.05 %HP 1 pigs at 28 −0.05 ns 1 Body weight deviation from the expected growth curve of non-vaccinated days post- 1 Body weight deviation from the expected growth vaccination (DPV); 2 Haptoglobin increment at curve 4 DPV; of 3non-vaccinated Ratio betweenpigs ∆BW at 28 anddays thepost-vaccination expected body (DPV); 2 Haptoglobin increment at 4 DPV; 3 Ratio between ∆BW and the expected body weight at 28 DPV weight at 28 DPV given the growth4 curve of control pigs. 4 Ratio between ∆HP and the basal level given the growth curve of control pigs. Ratio between ∆HP and the basal level of haptoglobin; 5 Body weight of haptoglobin; 5 Body weight deviation from the expected BW at 28 DPV estimated based on each deviation from the expected BW at 28 DPV estimated based on each pig’s average daily before challenge. * p < 0.05; pig’s ** average p < 0.01; *** p daily before < 0.001; challenge. * p < 0.05; ** p < 0.01; *** p < 0.001; ns non-significant. ns non-significant. 3.2. Classification of Pigs’ Resilience Pigs were classified based on the resilience indicated by both the deviation from the expected growth curve and the increment of HP at 4 DPV. DPV. First, individuals were grouped into R R or or SSgroups groupsbased basedon ∆BW on∆BW and and ∆HP ∆HP (Figure 2). 2). (Figure On Onaverage, the resilient average, pigs pigs the resilient (N = 25) =showed (N positive 25) showed values positive values of ∆BW of ∆BW (+3.54 kg) kg) (+3.54 andand %BW %BW (+6.60%) andand (+6.60%) negative values negative of values ∆HP of ∆HP (−0.71 (−0.71 mg/mL) mg/mL) and and %HP %HP (−61.2%). (−61.2%). In contrast, the susceptible In contrast, group the susceptible (N =(N group 33)=had 33) low and had low negative values and negative of ∆BW values of ∆BW (−6.00(− kg) and 6.00 kg)%BW %BW (−11.7%) and (−11.7%) and positive values of and positive ∆HP values of ∆HPmg/mL) (+1.17 (+1.17 mg/mL) and %HPand %HP (Table (+108%) (+108%) 4).(Table 4). Figure 2. Classification Figure 2. Classification of of pigs pigs as as resilient, resilient, average, average, and and susceptible susceptible based based on on the the first first (Q (Q11)) and and the the ∆BW and third (Q33) quartiles of ∆BW ∆HP. ∆BW: and ∆HP. ∆BW: body body weight weight deviation from the expected growth curve curve of control pigs at 28 days post-vaccination, ∆HP: haptoglobin increment at 4 days post-vaccination. post-vaccination. Pigs were also grouped into R and S groups based on the resilience indicated by %BW and %HP (Figure 3). Individuals were colored according to their group classification using ∆BW and ∆HP in order to visualize the concordance between the first and the second classifications. The concordance was high with a kappa value of 0.8 and an overall agreement of 95%. Pigs were also classified into R, A, and S based on the observed BW28 (Figure S1) and the combination of ∆BWADG and ∆HP (Figure S2). The concordance was low (kappa = 0.1) between the classification obtained by the observed BW28 and the combination of “∆BW and ∆HP” and moderate (kappa = 0.5) between the classification indicated by “∆BWADG and ∆HP” and “∆BW and ∆HP”. ∆BW 1 (kg) S −6.00 2.44 R +6.60 2.81 %BW 2 (%) S −11.7 5.13 R −0.71 0.18 Animals 2021, 11, 2970 ∆HP 3 (mg/mL) 7 of 12 S +1.17 0.61 R −61.2 18.9 %HP 4 (%) S +108.4 68.2 1 Body weight deviation from the expected growth curve of control pigs at 28 days post-vaccination Table 4. Mean and standard deviation (SD) of the resilience indicators of pigs from the resilient (DPV); 2 Ratio between ∆BW and the expected body weight at 28 DPV given the growth curve of (R, N = 25) and susceptible (S, N = 33) groups. control pigs; 3 Haptoglobin increment at 4 DPV; 4 Ratio between ∆HP and the basal level of hapto- globin. Trait Group Mean SD R +3.54 1.42 ∆BW Pigs 1 were (kg) also grouped into RS and S groups based−on the resilience indicated 6.00 2.44 by %BW and %HP (Figure 3). Individuals were R colored according +6.60 to their group classification 2.81 us- %BW 2 (%) S − 11.7 ing ∆BW and ∆HP in order to visualize the concordance between the first and the second 5.13 classifications. The concordance was R high with a kappa −0.71 0.18 agree- value of 0.8 and an overall ∆HP 3 (mg/mL) S +1.17 0.61 ment of 95%. Pigs were also classified into R, A, and S based on the observed BW28 (Figure 4 R −61.2 18.9 S1) and %HPthe combination (%) of ∆BW S ADG and ∆HP (Figure S2). The concordance +108.4 68.2 was low (kappa = 0.1) between the classification obtained by the observed BW 28 and the combina- 1 Body weight deviation from the expected growth curve of control pigs at 28 days post-vaccination (DPV); tion 2 of between Ratio “∆BW and ∆BW∆HP” and and the moderate expected body (kappa weight at= 28 0.5) between DPV thegrowth given the classification indicated curve of control pigs; 3 Haptoglobin increment at 4 DPV; 4 Ratio between ∆HP and the basal level of haptoglobin. by “∆BWADG and ∆HP” and “∆BW and ∆HP”. Figure 3. Figure Projectionofofthe 3.Projection theresilient, resilient, average, average, andand susceptible susceptible groups groups obtained obtained withwith the first the first (Q1) (Q and1) and third third (Q3 ) quartiles (Q3) quartiles of ∆BW of ∆BW and ∆HP ∆HP and on theonplane the plane defined defined by %HP by %HP and %BW. and %BW. Individuals Individuals were colored according were colored to their according to group classification their group usingusing classification the criterion from from the criterion FigureFigure 2 to visualize con- 2 to visualize cordance between both methods. ∆BW: body weight deviation from the expected concordance between both methods. ∆BW: body weight deviation from the expected growth curve growth curve of control pigs at 28 days post-vaccination (DPV), ∆HP: haptoglobin increment at 4 DPV, of control pigs at 28 days post-vaccination (DPV), ∆HP: haptoglobin increment at 4 DPV, %BW: ratio%BW: ratio between ∆BW and the expected body weight at 28 DPV, %HP: ratio between ∆HP and the basal level of haptoglobin. The growth curves of animals from the R and S groups were similar at the beginning of the experiment (Figure 4). After the challenge at 12 weeks of age, resilient animals were able to withstand the perturbations and showed a faster growth than susceptible ones. At the end of the fattening period (30 weeks of age), the resilient pigs showed a greater carcass weight than susceptible ones (107.7 and 92.1 kg, respectively). level of haptoglobin. The growth curves of animals from the R and S groups were similar at the beginning of the experiment (Figure 4). After the challenge at 12 weeks of age, resilient animals were Animals 2021, 11, 2970 able to withstand the perturbations and showed a faster growth than susceptible ones. 8 ofAt 12 the end of the fattening period (30 weeks of age), the resilient pigs showed a greater car- cass weight than susceptible ones (107.7 and 92.1 kg, respectively). Figure 4. Figure Growth curves 4. Growth curves of of pigs pigs from from the the resilient resilient and and susceptible susceptible groups. The The grey grey band band represents represents the the confidence confidence interval. interval. Individuals Individuals were were colored colored according according to their group classification using the criterion criterion from Figure 2. 3.3. Heritability 3.3. Heritability Estimates Estimates The features The features ofof the the marginal marginal posterior posterior distributions of the distributions of the heritability heritability estimates are estimates are displayed in Table 5. Both ∆BW and %BW had a moderate heritability displayed in Table 5. Both ∆BW and %BW had a moderate heritability of 0.33 and 0.37,of 0.33 and 0.37, with PP0.10 with 0.10ofof0.94 0.94and and 0.93, 0.93, respectively. respectively. Heritability Heritability estimates estimates of ∆HP of ∆HP and and %HP%HPwerewere 0.16 0.16 and 0.13, with P 0.10 of 0.66 and 0.53, respectively. and 0.13, with P0.10 of 0.66 and 0.53, respectively. Table 5. Table Heritability estimates 5. Heritability estimates for for the the resilience resilience indicators. indicators. Trait Mean 1 1 PP0.10 2 HPD 3 Trait Mean 0.10 2 HPD95% 95% 3 ∆BW 44 ∆BW 0.33 0.33 0.94 0.94 0.02-0.65 0.02–0.65 %BW 5 0.37 0.93 0.02–0.74 %BW 5 0.37 0.93 0.02-0.74 ∆HP 6 ∆HP 6 0.16 0.16 0.66 0.66 0.00–0.38 0.00-0.38 %HP 77 0.13 0.53 0.00–0.32 1 %HP 0.13 0.53 0.00-0.32 Mean of the marginal posterior distribution of the heritability; 2 Probability of the heritability estimate being 1 Mean than greater of the marginal 0.10; posterior 3 Highest posteriordistribution of the density interval heritability; at 95% 2 Probability of the heritability esti- of probability; 4 Body weight deviation from the mate being expected greater growth curvethan 0.10; 3 Highest of non-vaccinated pigsposterior at 28 daysdensity interval (DPV); post-vaccination at 95%5 of probability; Ratio between ∆BW 4 Body and 6 weight deviation the expected from the body weight at 28expected DPV given growth curve the growth of non-vaccinated curve pigs at 28 days of control pigs; Haptoglobin post-vaccina- increment at 4 DPV; 7 Ratio between tion ∆HP between (DPV); 5 Ratio and the basal ∆BW level andofthe haptoglobin. expected body weight at 28 DPV given the growth curve of control pigs; 6 Haptoglobin increment at 4 DPV; 7 Ratio between ∆HP and the basal level of hap- 4. Discussion toglobin. There is no consensus on the definition of resilience nor on how to measure it. Selec- tion for animals with a natural ability to cope with stressors and adapt to sudden changes has been proposed as a contribution to the sustainability of animal breeding [26]. However, in order to select these animals, we first need to be able to measure the trait. Different pro- ductivity and non-productivity indicators that reflect the disturbance caused by stressors have been used as indicators of resilience in several livestock species. Bai et al. [23] used quartiles of growth and the treatment rate in response to an infectious challenge to quantify the pigs’ resilience and considered animals with low number of treatments received and high growth rate as resilient. Other authors have defined resilience as the ability of animals to be minimally affected by perturbations and proposed resilience indicators based on the Animals 2021, 11, 2970 9 of 12 deviation of production traits such as the variance of deviations, the residual variance, and the skewness of deviations, considering individuals with the highest deviations as the most susceptible ones [5,6,8]. In the present work, we suggest novel resilience indicators based on the deviation from the expected BW given the growth curve of control pigs (∆BW) and the increment of the acute-phase protein HP at 4 DPV (∆HP), as a stress indicator. Animals that were able to maintain their productivity (positive values of ∆BW) and had a minor activation of HP (lower values of ∆HP) were deemed resilient. It is important to study the vaccine response since commercial animals are routinely vaccinated. The response to an attenuated Aujeszky vaccine was used to identify pigs with increased resilience. The vaccine response is an attractive alternative to natural infection because all animals can be vaccinated at the same dose, age, and time and, thus, response phenotypes can be collected with higher consistency and accuracy. This strategy has been previously used to study the genetic variability of the immune response to Aujeszky disease [27], influenza virus [28], European and American PRRS virus [29], and atrophic rhinitis [30]. Another advantage of using vaccines as a proxy of pathogen infection is that vaccinated and non-vaccinated animals can be reared in the same batch under the same environmental conditions, eliminating the confusion between treatment and batch. Vaccinated and control pigs were reared in five different batches where additional uncontrolled events occurred during the experiment and affected both E and C groups. In addition to the vaccine challenge, the first batch suffered from metabolic stress due to a failure in feed distribution, while batches 2 and 5 were exposed to an additional stress due to disease (gastric ulcers outbreak and respiratory disease). As a consequence, these batches had the lowest ADG. These results were expected since affected animals have a reduced feed intake and redirect energy to fight against diseases, and therefore, they convert food less efficiently into edible products [31]. On the other hand, batches 2, 3, and 5 had the highest HP concentration at 4 DPV, indicating that, in addition to the vaccine challenge, the batches were exposed to different stressful conditions. The HP concentrations after the challenge were lower than those in the literature for Aujeszky disease [21,22]. Carpintero et al. [22] and Parra et al. [21] reported greater average HP responses (3.96 and 1.81 mg/mL, respectively). One possible explanation could be that the animals were challenged with the Aujeszky disease vaccine three and two times, respectively, whereas, in our case, pigs were challenged with one dose of the attenuated Aujeszky disease vaccine. Vaccinated pigs had a reduced growth compared to control pigs. On average, the observed BW28 of challenged pigs was lower than expected. Besides, the HP concentration was 5% higher at 4 DPV. Among the 445 challenged pigs, 25 were deemed resilient and 33 were deemed susceptible. The observed BW28 for the resilient pigs was 6.6% (+3.45 kg) higher than expected, while the observed BW28 for the susceptible pigs was 11.7% (−6 kg) lower than expected. Regarding the HP response, the HP concentration highly increased after the challenge in the S group. Thus, our approach allowed us to separate two groups with extreme responses to challenge. Resilience was also assessed based on the combination of %BW and %HP. The concordance was high between both methods, indicating that ∆BW and ∆HP are not sensitive to the animal’s expected BW nor the basal level of HP and are consequently potential indicators of resilience. Moreover, the concordance was low between the classifications obtained by the combination of “∆BW and ∆HP” and the observed BW28 , indicating that ∆BW and ∆HP do not only capture the differences in the observed BW. Finally, the concordance was moderate between the resilience classification indicated by the combination of “∆BW and ∆HP” and “∆BWADG and ∆HP”. Thus, pigs could be consistently classified into R, A, and S based on ∆BWADG and ∆HP without using a control group in the same batch of animals. The ∆BWADG indicator would enable the application of the proposed resilience criteria under practical conditions where it is not possible or desirable to have non-vaccinated animals such as in the selection of nucleus farms. Animals 2021, 11, 2970 10 of 12 Since there is no consensus on the quantification method of resilience, completely dif- ferent approaches may be applied depending on the resilience definition and the objectives of the study or breeding program. In this work, resilience was studied as a qualitative trait (R, A and S) and the classification of the pigs’ resilience was based on the quartiles of ∆BW and ∆HP as suggested by Bai et al. [23]. Nevertheless, resilience was also analyzed as a quantitative trait (R = α ∗ ∆BW + β ∗ ∆HP) with our database and similar results were obtained compared with the qualitative ones (Supplementary Materials), suggesting that our qualitative approach is robust enough to set up resilience criteria. The heritability estimates of the resilience indicators were moderate for ∆BW and %BW and low-to-moderate for ∆HP and %HP. Heritabilities for ∆BW at 28 DPV have not been reported before, but our estimated value is similar to the heritability of BW reported in Duroc pigs at 180 days of age (0.31) [32]. Our heritability estimates for HP are within the range of those reported in non-specific pathogen-free pig farms (0.20) [33] and in growing pigs (0.14) [34]. Noteworthy, the experimental sample size limits the accuracy of the heritability estimates. However, P0.10 (i.e., the probability of the heritability being greater than 0.10) showed that the resilience indicators are genetically controlled and consequently, may be improved through selective breeding but the genetic correlations of these resilience indicators will be robustly established in follow-up studies, with a bigger number of animals, before including them in the breeding selection program. 5. Conclusions Altogether, we propose ∆BW and ∆HP as novel resilience indicators in growing pigs. The suggested indicators are easy to measure, genetically controlled, and show substantial variability between animals. Thus, they may be improved through selective breeding. This approach may be applied to quantify resilience in other species using different infectious and non-infectious challenges. Moreover, genomic studies on R and S animals can help in elucidating the molecular basis of the different responses in R and S animals. This work is a starting point of the study of the resilience in pigs. Further analyses are needed to validate our findings. Supplementary Materials: The following are available online at https://www.mdpi.com/article/10.339 0/ani11102970/s1. Figure S1. Projection of the resilient, average and susceptible groups obtained with the first (Q1 ) and third (Q3 ) quartiles of ∆BW and ∆HP on the plane defined by the first (Q1 ) and third (Q3 ) quartiles of BW at 28 days post-vaccination (DPV). Individuals were colored according to their group classification using the criterion from Figure 2 to visualize concordance between both methods. ∆BW: body weight deviation from the expected growth curve of control pigs at 28 DPV, ∆HP: haptoglobin increment at 4 DPV. Figure S2. Projection of the resilient, average and susceptible groups obtained with the first (Q1 ) and third (Q3 ) quartiles of ∆BW and ∆HP on the plane defined by ∆HP and ∆BWADG . Individuals were colored according to their group classification using the criterion from Figure 2 to visualize concordance between both methods. ∆BW: body weight deviation from the expected growth curve of control pigs at 28 days post-vaccination (DPV), ∆HP: haptoglobin increment at 4 DPV, ∆BWADG : Body weight deviation from the expected BW at 28 DPV estimated based on each pig’s average daily before challenge. Figure S3. Distribution of the quantitative resilience indices Table S1. Correlations between the resilience indicators based on quantitative indices. Author Contributions: Conceptualization, L.F. and R.N.P.; Methodology, L.F., A.B., R.N.P., J.E., R.R.-F. and H.L.; Formal analysis, H.L.; Data curation, R.A., J.R., M.D. and H.L.; Writing—original draft preparation, L.F., A.B., R.N.P., J.E., R.R.-F., R.A., J.R., M.D. and H.L.; Writing—review and editing, L.F., R.N.P., J.E., R.R.-F., R.A., M.D., J.R. and H.L.; Funding acquisition, L.F. and R.N.P. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by the Spanish Ministry of Science, Innovation, and Universi- ties and the European Union Regional Development Funds (grant RTI2018-097700-B-I00). Houda Laghouaouta is a recipient of a PhD scholarship from the Government of Catalonia (2020FI_B 00171). Animals 2021, 11, 2970 11 of 12 Institutional Review Board Statement: All experimental procedures were approved by the Ethics Committee for Animal Experimentation of the University of Lleida with reference number CEEA 02-04/18. Informed Consent Statement: Not applicable. Data Availability Statement: The data presented in this study are available on request from the corresponding author. Acknowledgments: We acknowledge the support provided by Selección Batallé SA during the experiments at all the included farms. Conflicts of Interest: The authors declare no conflict of interest. References 1. FAO. Food and Agriculture Organization of the United Nations; FAOSTAT Statistical Database: Rome, Italy, 2019. 2. Colditz, I.G.; Hine, B.C. Resilience in farm animals: Biology, management, breeding and implications for animal welfare. Anim. Prod. Sci. 2016, 56, 1961–1983. [CrossRef] 3. Iung, L.H.D.S.; Carvalheiro, R.; Neves, H.H.D.R.; Mulder, H.A. Genetics and genomics of uniformity and resilience in livestock and aquaculture species: A review. J. Anim. Breed. Genet. 2020, 137, 263–280. [CrossRef] 4. Martínez-Miró, S.; Tecles, F.; Ramón, M.; Escribano, D.; Hernández, F.; Madrid, J.; Orengo, J.; Martínez-Subiela, S.; Manteca, X.; Cerón, J.J. Causes, consequences and biomarkers of stress in swine: An update. BMC Vet. Res. 2016, 12, 171. [CrossRef] 5. Berghof, T.V.L.; Bovenhuis, H.; Mulder, H.A. Body Weight Deviations as Indicator for Resilience in Layer Chickens. Front. Genet. 2019, 10, 1216. [CrossRef] 6. Putz, A.M.; Harding, J.C.S.; Dyck, M.K.; Fortin, F.; Plastow, G.S.; Dekkers, J.C.M. Novel Resilience Phenotypes Using Feed Intake Data From a Natural Disease Challenge Model in Wean-to-Finish Pigs. Front. Genet. 2019, 9, 660. [CrossRef] [PubMed] 7. Argente, M.J.; García, M.L.; Zbyňovská, K.; Petruška, P.; Capcarová, M.; Blasco, A. Correlated response to selection for litter size environmental variability in rabbits’ resilience. Animal 2019, 13, 2348–2355. [CrossRef] 8. Elgersma, G.G.; de Jong, G.; van der Linde, R.; Mulder, H.A. Fluctuations in milk yield are heritable and can be used as a resilience indicator to breed healthy cows. J. Dairy Sci. 2018, 101, 1240–1250. [CrossRef] 9. Poppe, M.; Bonekamp, G.; van Pelt, M.L.; Mulder, H.A. Genetic analysis of resilience indicators based on milk yield records in different lactations and at different lactation stages. J. Dairy Sci. 2021, 104, 1967–1981. [CrossRef] 10. Dantzer, R.; Cohen, S.; Russo, S.J.; Dinan, T.G. Resilience and immunity. Brain. Behav. Immun. 2018, 74, 28–42. [CrossRef] [PubMed] 11. Webster, J.I.; Tonelli, L.; Sternberg, E.M. Neuroendocrine regulation of immunity. Annu. Rev. Immunol. 2002, 20, 125–163. [CrossRef] [PubMed] 12. Glaser, R.; Kiecolt-Glaser, J.K. Stress-induced immune dysfunction: Implications for health. Nat. Rev. Immunol. 2005, 5, 243–251. [CrossRef] 13. Guy, S.Z.Y.; Li, L.; Thomson, P.C.; Hermesch, S. Quantifying the health challenges in an Australian piggery using medication records for the definition of disease resilience1. J. Anim. Sci. 2019, 97, 1076–1089. [CrossRef] 14. Chen, Y.; Tibbs Cortes, L.E.; Ashley, C.; Putz, A.M.; Lim, K.S.; Dyck, M.K.; Fortin, F.; Plastow, G.S.; Dekkers, J.C.M.; Harding, J.C.S.; et al. The genetic basis of natural antibody titers of young healthy pigs and relationships with disease resilience. BMC Genom. 2020, 21, 648. [CrossRef] 15. Murata, H.; Shimada, N.; Yoshioka, M. Current research on acute phase proteins in veterinary diagnosis: An overview. Vet. J. 2004, 168, 28–40. [CrossRef] 16. Cray, C. Acute phase proteins in animals. In Progress in Molecular Biology and Translational Science; Academic Press: Cambridge, MA, USA, 2012; Volume 105, pp. 113–150. 17. Klauke, T.N.; Piñeiro, M.; Schulze-Geisthövel, S.; Plattes, S.; Selhorst, T.; Petersen, B. Coherence of animal health, welfare and carcass quality in pork production chains. Meat Sci. 2013, 95, 704–711. [CrossRef] 18. Fraile, L.; Saco, Y.; Grau-Roma, L.; Nofrarías, M.; López-Soria, S.; Sibila, M.; Callén, A.; Bassols, A.; Segalés, J. Serum haptoglobin dynamics in pigs vaccinated or not vaccinated against porcine circovirus type 2. Porc. Health Manag. 2015, 1, 3. [CrossRef] 19. Lipowski, A. Evaluation of efficacy and safety of Aujeszky’s disease vaccines. Pol. J. Vet. Sci. 2006, 9, 75–79. [PubMed] 20. Saco, Y.; Fraile, L.; Giménez, M.; Pato, R.; Montoya, M.; Bassols, A. Haptoglobin serum concentration is a suitable biomarker to assess the efficacy of a feed additive in pigs. Animal 2010, 4, 1561–1567. [CrossRef] [PubMed] 21. Parra, M.D.; Fuentes, P.; Tecles, F.; Martínez-Subiela, S.; Martínez, J.S.; Muñoz, A.; Cerón, J.J. Porcine acute phase protein concentrations in different diseases in field conditions. J. Vet. Med. Ser. B Infect. Dis. Vet. Public Health 2006, 53, 488–493. [CrossRef] [PubMed] 22. Carpintero, R.; Alonso, C.; Piñeiro, M.; Iturralde, M.; Andrés, M.; Le Potier, M.F.; Madec, F.; Álava, M.Á.; Piñeiro, A.; Lampreave, F. Pig major acute-phase protein and apolipoprotein A-I responses correlate with the clinical course of experimentally induced African swine fever and Aujeszky’s disease. Vet. Res. 2007, 38, 741–753. [CrossRef] Animals 2021, 11, 2970 12 of 12 23. Bai, X.; Putz, A.M.; Wang, Z.; Fortin, F.; Harding, J.C.S.; Dyck, M.K.; Dekkers, J.C.M.; Field, C.J.; Plastow, G.S.; Canada, P.G. Exploring Phenotypes for Disease Resilience in Pigs Using Complete Blood Count Data From a Natural Disease Challenge Model. Front. Genet. 2020, 11, 216. [CrossRef] 24. Thrusfield, M.; Christley, R.; Brown, H.; Diggle, P.J.; French, N.; Howe, K.; Kelly, L.; O’Connor, A.; Sargeant, J.; Wood, H. Veterinary Epidemiology; John Wiley & Sons, Ltd.: Chichester, UK, 2018; ISBN 9781118280249. 25. Legarra, A.; Varona, L.; de Maturana, E.L. TM: Threshold Model. 2008. Available online: http://genoweb.toulouse.inra.fr/ ~{}alegarra/tm_folder/ (accessed on 1 September 2021). 26. Friggens, N.C.; Blanc, F.; Berry, D.P.; Puillet, L. Review: Deciphering animal robustness. A synthesis to facilitate its use in livestock breeding and management. Animal 2017, 11, 2237–2251. [CrossRef] 27. Rothschild, M.F.; Hill, H.T.; Christian, L.L.; Warner, C.M. Genetic differences in serum-neutralization titers of pigs after vaccination with pseudorabies modified live-virus vaccine. Am. J. Vet. Res. 1984, 45, 1216–1218. [PubMed] 28. Zanella, R.; Gava, D.; de Oliveira Peixoto, J.; Schaefer, R.; Ciacci-Zanella, J.R.; Biondo, N.; da Silva, M.V.G.B.; Cantão, M.E.; Ledur, M.C. Unravelling the genetic components involved in the immune response of pigs vaccinated against influenza virus. Virus Res. 2015, 210, 327–336. [CrossRef] 29. Abella, G.; Pena, R.N.; Nogareda, C.; Armengol, R.; Vidal, A.; Moradell, L.; Tarancon, V.; Novell, E.; Estany, J.; Fraile, L. A WUR SNP is associated with European Porcine Reproductive and Respiratory Virus Syndrome resistance and growth performance in pigs. Res. Vet. Sci. 2016, 104, 117–122. [CrossRef] [PubMed] 30. Meeker, D.L.; Rothschild, M.F.; Christian, L.L.; Warner, C.M.; Hill, H.T. Genetic Control of Immune Response to Pseudorabies and Atrophic Rhinitis Vaccines: II: Comparison of Additive Direct and Maternal Genetic Effects. J. Anim. Sci. 1987, 64, 414–419. [CrossRef] [PubMed] 31. Rauw, W.M. Immune response from a resource allocation perspective. Front. Genet. 2012, 3, 267. [CrossRef] [PubMed] 32. Ros-Freixedes, R.; Reixach, J.; Tor, M.; Estany, J. Expected genetic response for oleic acid content in pork. J. Anim. Sci. 2012, 90, 4230–4238. [CrossRef] 33. Clapperton, M.; Diack, A.B.; Matika, O.; Glass, E.J.; Gladney, C.D.; Mellencamp, M.A.; Hoste, A.; Bishop, S.C. Traits associated with innate and adaptive immunity in pigs: Heritability and associations with performance under different health status conditions. Genet. Sel. Evol. 2009, 41, 54. [CrossRef] 34. Henryon, M.; Heegaard, P.M.H.; Nielsen, J.; Berg, P.; Juul-Madsen, H.R. Immunological traits have the potential to improve selection of pigs for resistance to clinical and subclinical disease. Anim. Sci. 2006, 82, 596–606. [CrossRef]