Иммунный транскриптомный профиль у взрослых женских свиней: доминантный статус оказывает большее влияние, чем экологическая обогащение.

Introduction

Decades of research in psychoneuroimmunology have demonstrated that the immune system and affective states influence each other [ 1 , 2 ]. In nonhuman animals, acute stressful events influence various immune features, such as antibody production, lymphocyte proliferation, cell trafficking, and the release of inflammatory mediators [ 3 – 6 ]. Chronic stress modulates the inflammatory response [ 7 – 9 ]. In humans, prolonged exposure to adverse conditions results in a repeatable pattern in the blood immune cell transcriptome, characterized by an increase in proinflammatory gene expression and a decrease in the expression of genes involved in type I interferon responses and antibody synthesis [ 10 – 13 ]. This pattern is known as the “Conserved Transcriptional Response to Adversity” (CTRA; 14). Interestingly, while the CTRA phenotype characterizes stressed and depressed individuals, the reversed phenotype could identify those with a better mood than the average population [ 14 ]. However, it remains unclear whether the transcriptomic patterns of psychological well-being observed in primates also exist in other species.

In the field of animal production, public concern regarding farm animal welfare is growing. On farms, production results, health and body condition indicators can be used to assess poor welfare, but the physiological variations associated with positive states of welfare have yet to be explored. Furthermore, while many physiological and behavioral tools are now available to identify responses to acute emotions, it is more difficult to assess the state of welfare associated to long-term stable situations. Identifying the genes or biological functions modulated during these situations could lead to a better understanding of the biological processes involved, and help to find new animal-based indicators of welfare. Housing animals in an enriched environment (EE) is a way to improve their welfare over the long term [ 15 ]. In pigs, EE can be achieved by the provision of rooting substrates, materials for manipulation, larger living spaces, and access to social partners or outdoor environments [ 16 ]. Changes in emotional sensitivity to reward loss and attention bias demonstrate the positive influence of EE on pigs’ affective states, which is an important component of the animal welfare [ 17 , 18 ]. Furthermore, studies have demonstrated the impact of EE on the immune system of pigs [ 19 – 22 ]. Beyond the physical environment, the social milieu is another crucial element to welfare. For instance, individuals’ hierarchical position within a stable social group modulates anxiety and stress responses [ 23 ]. In mice and pigs, an inverse relationship is observed between dominance and anxiety-like behaviors [ 24 , 25 ], and related neural mechanisms have been described [ 24 , 26 ]. In European breeding systems, sows are often housed in mixed parity groups during gestation, which contributes to the development of stable social structures that significantly influence their daily social and nonsocial behaviors [ 27 ] and stress responsiveness [ 28 ].

In this study, we investigated the variations in the peripheral blood mononuclear cell (PBMC) transcriptome associated with environmental enrichment and dominance status. For this, we conducted an experiment on a farm where sows were kept in either a conventional or an enriched environment, without ever changing systems during their successive reproductive cycles, which made it possible to study the effect of enrichment over the long term. During their gestation period, sows were housed in stable groups, which made it possible to study the effect of dominance status at the end of gestation, at a distance from the stressful moment when the group was formed (a few days before the insemination). Transcriptome measures were carried out at the end of gestation, but also at the start of lactation to assess the persistence of these effects once the sows had left the physical and social environment of their gestation period.

Methods

Experimental farm

The experiment was carried out at the experimental farm of the Chambre Régionale d’Agriculture de Bretagne (Saint-Nicolas-du-Pélem, France). On this demonstration farm, animals are raised and marketed for meat in the same way as on a conventional commercial farm. Young nulliparous sows (Landrace x Large White) are randomly assigned upon purchase to one of two independent reproductive units, which feature distinct housing environments. Sows remain in the same unit throughout their entire reproductive life. In the conventional (C) unit, sows are housed during gestation in rooms with concrete slatted floors, a minimum space of 2.4 m 2 per sow and two hanging metal chains as a minimal enrichment material. In the enriched (E) unit, sows are housed on deep straw bedding with greater space per sow (3.5 m 2 ). The animals enter the rooms at the gestation day (G) 0 and remain there until ten days before farrowing (G105). At G105, they are transferred to the maternity rooms, where they are allocated in individual farrowing crates on a slatted floor, and then remain with their litter until weaning. More details regarding the experimental units are showed in Fig. 1 A. At weaning, the sows return to the same room as in their previous gestation. The group of sows in each housing environment is composed of 17 to 24 animals of mixed parities, ranging from 0 to 7 gestations. Sows are individually fed through an automatic dispensing system, and management and feeding are similar for both units. Sows and piglets remain in the farm production cycle and are marketed for meat once reaching standard weight (for pigs) or culling age (for sows).

Fig. 1 The experimental ( A ) gestation conventional (C) and enriched (E) rooms and the maternity room, ( B ) timeline from the start of gestation (G0) until the 12th day of lactation (L12), and ( C ) number of multiparous sows for the investigated factors (housing environment, dominance status and parity)

Experimental design

For this study, we carried out observations and took samples from the animals present in the farm at the time of our visits without in any way interfering with the usual management described above. The experiment was carried out on two batches (Fig. 1 ): the first (B1) from November 2021 to February 2022 (B1-C and B1-E groups), and the second (B2) from March to July 2022 (B2-C and B2-E groups). These two batches were made up of a total of 71 animals (C, n = 36 and E, n = 35), among which 11 primiparous sows, and 60 multiparous sows (Supplementary Table S1 ). At sampling time, we selected a subset of 28 multiparous sows for biological samplings (7 sows per batch and per housing environment). These same animals were sampled at the three visits (G35 and G98, and day of lactation (L) 12). We opted for excluding primiparous sows (sows involved in their first gestation at our first visits) due their limited familiarity with their housing environments and due to their expected strong behavioral and physiological differences with other sows. In each batch, among the multiparous sows, we selected 14 sows that were confirmed pregnant and for which we tried to balance the parity ranks among C and E pens as best we could.

The decision to investigate not only the effect of enrichment but also the influence of dominance status was made a posteriori , as we discovered significant effects of this factor on the transcriptome. The method to classify sows in dominant, intermediate and subordinate categories within each social group is described in more details in Sect. 2.4 . The numbers of dominant, intermediate and subordinate sows among the 28 sampled sows was not perfectly balanced once this classification was not anticipated from the outset (see Fig. 1 C and Sect. 2.5 ). Among multiparous sows, there were 24 dominant, 21 intermediate, and 15 subordinate sows. Among the 28 sampled sows, there was 9 dominant sows, 11 intermediate sows, and 8 subordinate sows.

Because dominance and parity are strongly correlated in groups of sows [ 29 ], we also felt it was important to consider in this study the effects of parity. Among the multiparous sows, using the sample clustering obtained during the descriptive analysis of the transcriptomic data (Supplementary Fig. S1 ), we distinguished a posteriori two parity groups of the same size. The low-parity (LP, n = 29) group included sows in their second or third gestations (554 ± 73 days of age upon entering the maternity room), and the high-parity (HP, n = 31) group included sows involved in their fourth or subsequent gestations (867 ± 163 days of age upon entering the maternity room). Among the 28 sampled sows, we had 13 HP and 15 LP sows. The distribution of parity and dominance classes among animals of the B1-C, B1-E, B2-C and B2-E groups is shown in Fig. 1 and in Supplementary Table S1 .

Physiological analyses

Saliva and hair cortisol

Saliva samples were collected from the subset of 28 sows on days G35 and G98 at 09:00 h using cotton buds (Salivette ® , Sarstedt 51588 Nümbrecht, Germany). The animals chewed the material voluntarily until it was moistened. The samples were immediately placed on ice until centrifugation at 2500 × g for ten minutes at 4 °C and frozen at -20 °C until cortisol analysis. On G98, hair samples were obtained from the neck area by shaving close to the skin with electric hair clippers, avoiding the inclusion of hair follicles. The hair samples were stored at room temperature in hermetically sealed bags and immediately frozen at -20 °C upon arrival at the laboratory for subsequent analysis. Cortisol was extracted from hair samples according to the method outlined by Heimbürge and colleagues [ 30 ]. Both salivary and hair cortisol concentrations were analyzed in duplicate with a luminescence immunoassay kit (LIA, IBL, Hamburg, Germany) following the manufacturers’ instructions.

Blood collection, cell count and PBMC isolation

Blood samples were collected from the subset of 28 sows on G98 and L12. An experienced veterinarian collected blood samples (three EDTA tubes of 10 mL per sow) from the anterior vena cava on G98 and the mammary vein on L12. On G98, animal keepers guided the sows individually to a separate room away from the group and maintained them with a nose sling while the sampler collected blood from the anterior vena cava. On L12, blood was collected from the mammary vein directly in the sows’ farrowing crates, requiring no additional containment. Blood samples were transported in a thermobox with refrigerated rolls and arrived at the laboratory within three hours after the sampling procedure.

The total numbers of lymphocytes, monocytes, neutrophils, eosinophils, and basophils were measured with an MS-9 ® hematology automatic cell counter (Melet Schloesing laboratories, Osny, France). To isolate the PBMC fraction, 9 ml of EDTA-treated blood was half diluted with PBS buffer and centrifuged for 30 min at 400 × g over a Ficoll density gradient (Histopaque 1077, Sigma‒Aldrich, Saint-Quentin Fallavier, France). Cells at the interphase were collected and washed, and the remaining red blood cells were lysed through an 8-minute incubation in a hypotonic solution. After two washes with PBS, the cells were counted, pellets of 2 million cells were prepared, and lysis buffer solution (Macherey Nagel, Düren, Germany) was added to the PBMC pellets, which were stored at -75 °C until RNA extraction.

RNA isolation, cDNA library preparation, and sequencing

RNA was isolated using the NucleoSpin RNA Kit according to the manufacturers’ instructions (Macherey‑Nagel, Düren, Germany). The RNA concentration was measured with a DS-11 nanospectrophotometer (Denovix, Wilmington, USA), and the RNA quality was checked with an Agilent 2100 bioanalyzer (Agilent Technologies, Santa Clara, USA). The RIN values ranged from 7 to 9.4. The RNA samples were then sent to the PGTB facility (Bordeaux, France) for library preparation and sequencing. Libraries were prepared from 500 ng of total RNA following the QIAseq Stranded mRNA Library Kit (Qiagen, Hilden, Germany) protocol and using 12 cycles for amplification. The final libraries were checked on a TapeStation 4200 (Agilent Technologies) using an HS D1000 kit and quantified by fluorescence with a Qubit 1X dsDNA High Sensitivity (Thermo Fisher Scientific, Waltham, USA) on a BioTek Synergy H1 (Agilent Technologies) before equimolar pooling. Sequencing was performed on a NextSeq 2000 (Illumina, San Diego, USA) using a P3 2 × 150 bp kit. Dragen RNA v3.8.4 software (Illumina) was used for mapping and annotation against the Sus scrofa reference genome (Sscrofa11.1, GCF_000003025.6).

Behavioral observations

Behavioral observations were collected from all sows of the B1-E, B1-C, B2-E, and B2-C groups. On G98, all sows were individually marked with a number sprayed on their back, and video footage was continuously recorded from G99 to G105. All the behavioral observations were conducted on the video footage by a single observer using The Observer XT 14.1 software (Noldus Information Technology B.V., The Netherlands).

On G99, G101, and G103, sows’ behaviors and postures were scored using 10-minute instantaneous scan sampling [ 31 ]. Scan sampling was performed during one morning (10:00 h − 13:00 h) and one afternoon (13:00 h − 16:00 h) session of three hours per day, resulting in 108 samples per sow. During each scan, the posture (standing, lying, sitting) and activity of each sow were scored (ethogram in supplementary Table S2 ). Seven categories of activities were identified: inactivity, pen exploration, locomotion, social interactions (aggression, nosing interactions, and oral manipulation of pen mates), self-oriented stereotypies (vacuum chewing, tongue rolling, and head waving), chewing the metal chains, and others (eating, drinking and maintenance behaviors). Postures and behavioral activities recorded over the six observation sessions were expressed as proportions of total scans.

From G99 to G103, social interactions were scored using continuous all-occurrence behavioral sampling [ 31 ] during one session of three hours (10:00 h -13:00 h) per day, resulting in 15 h of observation per sow. Four categories of social interactions were identified: aggression (bites, threats, chases, displacements, and head knocks), nosing snout, body exploration (nosing head, nosing body, and allogrooming), and other social interactions (mounts, nudges, oral manipulations, and head-over-body, ethogram in supplementary Table S3 ). Social interactions were averaged over the five observation days and were expressed as frequencies (occurrences per hour).

For each aggression, we identified the initiator, the receiver, the outcome (tie or clear winner/loser), and the winner of the interaction. The sow was considered the winner of the interaction if it induced a behavior of avoidance or flight from the other sow (see Responses to aggression in Supplementary Table S3 ). Aggressive interactions with a clear outcome (clear winner/loser) were used to determine dominance status by calculating the individual Elo score using the EloRating R package, as described in [ 29 ]. Briefly, dominance matrixes were built based on the outcome (winner/loser) of the aggressive interactions (the interactions with a tie outcome were not used). The Elo-score was calculated iteratively from a unique start score (default = 1000) and was modified according to the outcomes of the chronological sequence of the aggressive interactions over five days. Within each social group (B1-C, B1-E, B2-C, B2-E), sows were ranked according to their Elo-score and classified as dominant, intermediate, or subordinate, ensuring an equal distribution, with approximately one-third of the group members in each dominance category.

Statistical analyses

Rational

This study was initially designed to investigate the effect of enrichment on the behavior and transcriptome of sows, and we were surprised to discover the major effect of dominance status. We also felt it was important to make it clear that effects attributed to dominance could be, in fact, the result of parity rank and not the social context of the individuals. The decision to study these factors a posteriori created constraints in terms of the statistical model linked to the poorly balanced numbers between the dominance and parity groups. A total of 83% of the DOM sows belonged to the HP group (20 of 24 sows), while 80% of SUB sows were in the LP group (12 of 15 sows). Among the INT sows, 62% were LP, and 38% were HP. This imbalance in the numbers (very few DOM LP sows and SUB HP sows) meant that the dominance and parity rank variables could not be included in the same statistical model without the risk of generating aberrations in the statistical outputs. Consequently, we investigated the influence of dominance status and enrichment on one side, and parity and enrichment on the other side with two different statistical models. We compared the results of these two models to discern the genuine dominance status effects from those potentially associated with parity/aging.

Statistical analysis of cortisol concentrations, blood cell numbers, and behavioral variables

All the statistical analyses were performed using R version v. 4.1.2 [ 32 ]. For the behavioral variables, we analyzed the data of all multiparous sows ( n = 60), and for the physiological variables (cortisol concentration and blood cell count) we analyzed the data of the 28 selected multiparous sows used for biological sampling.

Linear mixed-effect models were employed for statistical analyses using the lmer function of the lme4 package. The effects reported in the results section for EE and dominance status were derived from the statistical model containing the environment (E vs. C), dominance status (DOM, INT, and SUB), their two-way interactions as fixed effects, and the batch as a random effect. For salivary cortisol and white blood cell counts, the models additionally included the sampling day as fixed effect and sow as random effect, in order to take into account the repeated measures on each sow. The normality of the distribution and the equality of residual variance were checked visually. Square root transformations were performed for salivary cortisol, blood cells numbers, and the frequency of nosing-snout, nosing-head, body exploration, and social activities. Arcsine square root transformations were performed for the proportion of pen exploration, chain chewing, walking, social interactions, stereotypies, and other behaviors (eating, drinking, and maintenance). Factor effects were considered significant if the p-value was < 0.05. After identifying significant factors using ANOVA, paired comparisons were carried out by post hoc Tukey tests using the emmeans package. The values presented in this study are reported as the back-transformed values of the lsmeans.

Analysis of differential gene expression

Gene count tables were imported into R, and gene symbol annotation was performed using the ‘sscrofa_gene_ensembl’ database extracted via the R package “biomaRt”. Hemoglobin A and B expression represented more than 5% of the total reads in four samples collected on G98 (E-B1) and one on L12 (C-B1), indicating red blood cell contamination in the PBMC preparation. To avoid possible confounding effects, these samples were excluded from downstream analysis. Normalization and differential gene expression analysis were conducted using the DESeq2 package following the standard procedure [ 33 ]. Genes with low expression (i.e., those with fewer than ten counts and absent in at least half of the samples) were excluded from the dataset. An outlier sample on L12 (C-B2) was excluded from the dataset, resulting in a final count of 24 animals on G98 and 26 on L12. The table showing the number of samples by enrichment group, parity rank and dominance status is available in the supplementary Table S1 .

Due to the configuration of the DESeq2 package, the matrix models consisted only of fixed variables, and it was not possible to declare repeated measures. Therefore, the data from G98 and L12 were analyzed separately. Two statistical models were constructed by time point. Model 1 included the environment (E and C), dominance status (DOM, INT, and SUB), and batch (B1 and B2) effects. Model 2 included the environment (E and C), parity (HP and LP), and batch (B1 and B2) effects. To obtain the desired comparisons (E vs. C or DOM vs. SUB from Model 1 or HP vs. LP from Model 2), the contrast argument of the results function was employed. Genes considered significantly differentially expressed (DEGs) presented a fold change (FC) greater than 1.2 or less than 0.83 (approximately 20% of differences between groups) and a BH-adjusted p value < 0.1 [ 34 ]. The DEG lists of up and downregulated DEGs put together resulting from the DOM vs. SUB and HP vs. LP contrasts were subjected to enrichment analysis with the DAVID bioinformatic database v2022q2 [ 35 ], using the Biological Processes (BP) Gene Ontology as the data source and the Homo sapiens background. Biological processes with at least 3 genes and p-values < 0.05 were considered enriched.

Results

A complete description of the data and raw results (blood cell count, RNA-Seq count tables, and results of the analyses with R and DAVID software) are presented in a data paper [ 36 ], and the datasets are available on Data INRAE’s institutional collection website (10.57745/NHE0MP).

Impact of environmental enrichment on cortisol concentration, white blood cell numbers and behaviors

Upon arrival at the farrowing unit, no significant differences in body weight were detected between the multiparous E sows (265 ± 8.8 kg, 711 ± 212 days, n = 29) and the C sows (279 ± 6.8 kg, 720 ± 196 days, n = 31; P > 0.05). E sows exhibited lower salivary cortisol concentrations than those in the C environment ( P = 0.004; 0.67 ± 0.05 ng/mL in the E group and 0.92 ± 0.05 ng/mL in the C group; Fig. 2 A). The sampling day did not influence salivary cortisol concentrations ( P > 0.05). The hair cortisol concentration was lower in the E sows than in their C counterparts ( P = 0.003; 2.58 ± 0.67 ng/g in the E group and 4.44 ± 0.64 ng/g in the C group). The blood numbers and percentages of lymphocytes, monocytes, and neutrophils were unaltered by EE on G98 and L12 ( P > 0.05 for all). Significant differences were observed between sampling days for nearly all blood cell types ( P < 0.05), excluding the lymphocyte number ( P > 0.05).

Regarding social behaviors, compared with C sows, E sows tended to exhibit a reduced frequency of aggression ( P = 0.06). Surprisingly, they also exhibited fewer snout-to-snout contacts ( P = 0.01; Fig. 2 B). No significant differences between housing environments were observed for body exploration or other social behaviors ( P > 0.05). Regarding the sows’ activity, E sows spent more time exploring the pen ( P < 0.001) and less time chewing the chain ( P < 0.001). They also spent less time engaged in stereotypic behaviors ( P = 0.04) than did the C sows (Fig. 2 C). No significant differences were detected between environments for time spent inactive, engaged in locomotor or social activities, or engaging in other behaviors (eating, drinking, and maintenance; P > 0.05; data not shown).

Fig. 2 Comparison of sows housed in enriched (E) or conventional (C) environments for ( A ) pooled salivary cortisol concentrations on gestation days 35 and 98, ( B ) frequencies (occurrences/hour) of negative (aggression) and positive (nosing the snout of another sow) social behaviors, and ( C ) proportion of scans spent exploring the pen, chewing the suboptimal chain toy, and exhibiting stereotypies at the end of gestation. * P < 0.05, ** P < 0.01, *** P < 0.001

Impact of dominance status on cortisol concentration, white blood cell numbers and behaviors

Multiparous DOM sows were heavier and older (304 ± 8.3 kg, 854 ± 201 days, n = 24) than INT (275 ± 8.1 kg, 664 ± 150 days, n = 21; P = 0.001) and SUB (237 ± 11.7 kg, 565 ± 113 days, n = 15; P < 0.001) sows at the maternity room entrance. There was an interaction effect between dominance status and sampling day on salivary cortisol ( P = 0.002; Fig. 3 A). On G35, DOM sows presented higher cortisol concentrations than INT sows ( P = 0.006), and this difference disappeared on G98 ( P > 0.05). For hair cortisol, no differences between the dominance groups were detected ( P > 0.05). The numbers and percentages of lymphocytes, monocytes, and neutrophils in the sows’ blood were unaltered by dominance status ( P > 0.05 for all).

Dominance status significantly influenced the body exploration of congeners ( P = 0.01), nosing-snout ( P = 0.02) and other social behaviors ( P = 0.02). DOM sows expressed fewer of these behaviors than did SUB sows ( P = 0.01, 0.04 and 0.02; Fig. 3 B). Dominance status did not have a significant impact on the frequency of aggression ( P > 0.05). With respect to the time budget, dominance status affected locomotion ( P = 0.02), the expression of other behaviors (eating, drinking and maintenance behaviors; P = 0.04), and social activities ( P = 0.06). DOM and INT sows presented a lower frequency of locomotion than did SUB sows ( P = 0.08 and P = 0.02, respectively). Additionally, INT sows tended to perform fewer social activities than did SUB sows ( P = 0.07; Fig. 3 C). No differences were detected between the dominance groups for pen exploration, chewing the chain, stereotypies, or inactivity ( P > 0.05; data not shown).

Fig. 3 Comparison of dominant (DOM), intermediate (INT) or subordinate (SUB) sows’ levels of ( A ) salivary cortisol concentrations on gestation days 35 and 98, ( B ) frequencies (occurrences/hour) of body exploration of congeners and positive (nosing the snout of another sow) social behaviors, and ( C ) proportion of scans for time spent locomoting and engaged in other activities (eating, drinking and ingesting) and in social interactions at the end of gestation. a, b,c Means with different superscripts differ ( P < 0.05)

Impact of environmental enrichment on the PBMC transcriptome

At both time points, we examined the contrast of gene expression in E sows relatively to C sows. On G98, among the 12,260 genes tested (E, n = 10 vs. C, n = 14), 31 were differentially expressed. Ten genes were downregulated (FC within the range [0.37:0.80]), while 21 were upregulated in E sows (FC within the range [1.21:8.2]; Fig. 4 ). Based on UniProtKB keyword annotation, two upregulated genes ( LDHB and ATP5ME ) and one downregulated gene ( PXMP4 ) in E sows were implicated in energy metabolism. Additionally, genes associated with chromatin organization exhibited varied expression levels ( C2orf15 , TET2 , C17orf49 and H2AZ1 ). Several upregulated DEGs were linked to protein synthesis, processing, and transport ( SELENOS , RPLP1 , PFDN6 , and SEC61B ), rRNA and mRNA processing ( EXOSC3 and BCAS2 ), transcriptional regulation ( AHRR , HINT1 , and HMGN5 ), mitosis ( BORA ) and DNA damage repair ( PDRG1 ). Furthermore, some genes related to innate immunity ( TLR6 and CD55 ) were downregulated.

On L12, among the 12,268 tested genes (E, n = 14 vs. C, n = 12), only four were differentially expressed: three downregulated ( OTUB2 , PIK3R6 , RNF39 ; FC in [0.5:0.74]) and one upregulated ( ENSSSCG00000033790 ; FC = 9.21). OTUB2 is associated with protein modification, PIK3R6 is associated with the PI3K-AKT signaling pathway, and RNF39 is part of major histocompatibility complex class I. The expression of the uncharacterized gene ( ENSSSCG00000033790 ) was upregulated on both G98 and L12.

Fig. 4 Heatmaps displaying differentially expressed genes (DEGs) on gestation day 98 (G98) and lactation day 12 (L12) in the PBMC transcriptome of sows housed in either enriched (E) or conventional (C) environments. Each column on the heatmap represents a sample. The color key indicates the Z score normalization performed on the normalized read counts. Column dendrograms illustrate hierarchical clustering results based on DEG expression

Impact of dominance status and parity on the PBMC transcriptome

Among the sows sampled for transcriptome analysis, most dominant sows were from the high-parity group (8 of 9 sows), and most subordinates were from the low-parity group (7 of 8). However, the reverse was not true: among the high-parity animals, there were animals in all three social classes (8 DOM, 4 INT, 1 SUB), and this was also true for the low-parity sows (1 DOM, 7 INT, 7 SUB). Given the considerable overlap, which makes it impossible to include both factors within the same statistical model, we adopted separate statistical models (model 1 for dominance, model 2 for parity). This comparison enabled us to dissociate, among the DEGs and functions obtained from Model 1, those strictly linked to dominance from those more related to parity or age.

Regarding the dominance status, we present here only the contrast of gene expression in dominant sows relatively to subordinate sows. Comparisons with sows of intermediate status are not presented. For parity, we examined the contrast of gene expression in high-parity sows relatively to low-parity sows. Dominance status influenced 449 DEGs on G98 (DOM, n = 6 vs. SUB, n = 8) and 25 DEGs on L12 (DOM, n = 10 vs. SUB, n = 7). Parity influenced 542 DEGs on G98 (HP, n = 11 vs. LP, n = 13) and 91 DEGs on L12 (HP, n = 13 vs. LP, n = 13; Table 1 ). The complete gene lists can be found in the data paper [ 36 ]. The dominance status and parity DEG lists on G98 had 180 genes in common (approximately 40% of the dominance status DEGs), while 17 genes were shared on L12 (approximately 68% of the dominance status DEGs). Additionally, four DEGs for dominance were consistent across both dates and independent of parity influence (up: CDH23 , ENSSSCG00000033104 ; down: CR1 , C4BPB , CEACAM1 ).

Table 1 Numbers of differentially expressed genes (DEGs) when comparing dominant relatively to subordinate and high- relatively to low-parity sows on gestation day 98 (G98) and lactation day 12 (L12) G98 L12 Common DEGs Downregulated Upregulated Downregulated Upregulated G98 and L12 Dominance status 233 FC [0.1:0.83] 216 FC [1.2:3.21] 18 FC [0.29:0.73] 7 FC [1.34:2.58] 18 Parity 313 FC [0.2:0.82] 229 FC [1.2:3.6] 63 FC [0.32:0.82] 28 FC [1.22:2.99] 69 Shared DEGs 101 79 15 2

Impact of dominance status and parity on immune-related functions

Of the 449 DEGs identified in the dominance list (up and downregulated genes together), 244 were implicated in 106 enriched biological processes. Among the 542 DEGs identified in the parity list, 217 were associated with 63 enriched biological processes. Twenty-one of them were affected both by dominance status and parity (Table 2 ), while 85 enriched processes were exclusive to dominance status (Table 3 ), and 42 exclusives to parity (not shown). Biological processes influenced by both dominance and parity included: adaptive immunity, innate immunity, cytokine response, cellular process, signaling, developmental process, and protein modification (Table 2 ). For both factors, DEGs associated with T-cell differentiation and activation, macrophage differentiation and cellular responses to IL-4 and IL-5 were affected, which demonstrated that these functions might be altered by the parity/age of the animals. Biological processes critical for correctly addressing immune responses, such as cell‒cell adhesion, cell migration, and lamellipodium assembly, were also similarly impacted.

Table 2 Biological processes affected by both dominance and parity of the sows on gestation day 98. For each biological process, we present the up and downregulated genes in DOM sows relatively to SUB sows and in HP sows relatively to LP sows Biological process 1 Dominance status Parity Downregulated Upregulated Downregulated Upregulated Immune system (general) Immune response CCR6 , CCR7 , CCR9 , CD22 , CD79B , CRHR , CTSG , CXCR5 , ENPP CCL14 , CD1D , CD4 , CD8A , CTSS , FCGR2B , FCGR3A , IL1RAP , IL1RL1 , TENM1 , XBP1 BPI , CCR7/9 , CRHR1 , CTSG , ENPP1 , GPR183 , IL18R1 , PAMR1 , TNFSF11 , ZAP70 BMPR1A , CCL14/23 , CCR8 , CD1D , CD4/86/96/244 , CD8A , CXCR3 , FCGR3A , IL1RAP , IL1RL1 , MR1 , MS4A2 , TNFRSF4 Inflammatory response ALOX15 , APOL3 , CCR7 , CD40 , CHST1 , CHST2 , GATA3 , NFKB2 , TBXA2R AIF1 , CAMK4 , CCL14 , F2RL1 , FCGR2B , IL1RAP , IL1RL1 , IL1RL APOL3 , CHST1/2 , CCR7 , GATA3 , GGT5 , IL18R1 , PLA2G4B CCL14/23 , CD96 , CXCR3 , F2RL1 , IL1RAP , IL1RL1/2 , MS4A2 , PTGS1 , TAC4 , TNFRSF4 , TSPAN2 , XCL1 T-cell differentiation GATA3 , SOX4 , ZAP70 CD4 , LIPA , RUNX2 CHD7 , GATA3 , SOX4 , ZAP70 CD4 , IL7R , RUNX2 T-cell activation TREML2 , CD48 , LY9 CD4 , CD8A CD48 , SLAMF7 , TREML2 , ZAP70 CD4 , CD8A , CD86 , DPP4 Macrophage differentiation GATA3 , ZBTB46 CD4 , IL1RL1 GATA3 , ZBTB46 CD4 , IL1RL1 Cytokine response Cellular response to IL-4 CORO1A , GATA3 HSPA5 , XBP1 GATA3 , LEF1 MRC1 , XCL1 Reg. of IL-5 production GATA3 IL1RAP , IL1RL1 CRLF2 , GATA3 IL1RAP , IL1RL1 Cellular processes Macromolecular complex assembly CD40 , CD247 , RIC3 , TFAP4 GCH1 , HFE , IL1RAP , PICALM CD247 , DLG4 , HES1 , NHERF2 , TEAD2 , RIC3 CAPZA2 , DMD , HFE , IL1RAP , SKAP2 Lamellipodium assembly ABLIM3 , AJUBA , VAV2 CCDC88A , FGD4 , ITGB1 ABLIM3 , ARHGEF4 CCDC88A , FGD4 , ITGB1 Cellular response to BMP stimulus ACVR2A , ADAMTS7 , GATA3 BMPR2 , RUNX2 ACVR2A , GATA3 BMPR1A , RUNX2 Cell migration BRAT1 , CASS4 , CEACAM1 , CORO1A , FMNL3 , FSCN2 , GPC1 , JUP , L1CAM , PIK3C2B , PRAG1 , VAV2 ADAM9 , CCDC88A , DOCK1 , ITGB1 , MEGF9 , MET , PAK1 , SDC2 CASS4 , FMNL3 , FSCN2 , HES1 , JUP , L1CAM , PIK3C2B , RHOV CD63 , CCDC88A , ITGB1 , MEGF9 , PAK1 Cell‒cell adhesion ACTN2 , CEACAM1 , CASS4 , JUP , L1CAM , LY9 , MCAM , MPZL3 , SIGLEC10 , THY1 ADAM9 , ADGRE1 , CD4 , CD22 , DST , EMB , ITGA4 , ITGB1 , NRP2 , PDZD2 , VCAN CDH24 , DLG4 , DSP , ITGA9 , JUP ADGRV1 , ANXA2 , CD58 , S100A11 , TJP2 , TMEM47 Signaling Pos. reg. of PKB (Akt1) signaling CASS4 , CCR7 , ERBB3 , GATA3 , TNFSF11 , TSPYL5 F2RL1 , FLT3 , HIP1 , ITGB1 , MET , MFHAS1 , PDGFB , TGFBR1 , XBP1 CASS4 , CCR7 , ERBB3 , GATA3 , PIK3AP1 , TNFSF11 EGF , FLT3 , F2RL1 , ITGB1 , PDGFB Cell surface receptor signaling pathway CD247 , CD27 , CRHR1 , EPHA1 , FCRL1 , FCRL3 , FCRL4 , TNFRSF13B ADGRE1 , ADGRV1 , CD4 , CD8A , FCGR2B , FCGR3A , IL13RA1 , MET , MS4A10 , TSPAN12 ADGRL1 , CD247 , CD27 , CRHR1 , FCRL4 , TNFRSF25 , TNFSF11 ADGRE1 , ADGRV1 , CD4 , CD86 , CD8A , CLEC2D , CXCR3 , FCGR3A , IL13RA1 , IL7R , MS4A2 , SCTR Calcium-mediated signaling CCR6 , CCR7 , CCR9 , CXCR5 , TNFSF11 , TRPM8 , ZAP70 CD4 , FCGR3A , MCTP1 , RCAN2 CCR7 , CCR9 , PLA2G4B , TNFSF11 , TRPM8 , ZAP70 CALM2 , CCR8 , CD4 , CXCR3 , FCGR3A , MCTP1 , RCAN2 1 Only four pathways, related to protein metabolism, ion homeostasis and developmental process, are not presented in this table

For biological processes specific to dominance, a significant number of DEGs were implicated in the adaptive humoral immune response and B-cell activation. The genes involved in these functions were mainly downregulated in DOM sows suggesting a global reduction in those functions (table 3 ). Other processes associated with B-cell differentiation, proliferation and immunoglobulin production were also affected. We found some upregulated genes associated with endoplasmic reticulum functions, which might indicate increased activity of this organelle. The analysis further revealed biological processes linked to the inflammatory response, with the proinflammatory cytokine TNF-α and IL-1 pathways being associated with a greater number of upregulated genes. Notably, signaling pathways associated with inflammation, such as I-kappaB kinase/NF-kappaB signaling, mitogen-activated protein kinase cascade, and the c-Jun N-terminal kinases, were also exclusively affected. Finally, other signaling pathways (PKC and PI3K-AKT) and functions related to protein metabolism (protein processing, peptidyl-serine phosphorylation, regulation of protein stability, and ubiquitination) were also altered by dominance status.

Table 3 Presentation of a subset of biological processes affected by sows’ dominance status exclusively on gestation day 98 (G98). For each biological process, we present the up and downregulated genes in DOM sows relatively to SUB sows Biological process 1 Downregulated DEGs Upregulated DEGs p value Adaptive immunity Humoral immune response CCR6 , GATA3 , GPR183 , PAX5 , POU2AF1 - 0.04 B-cell activation CD22 , CD40 , CD79A , CXCR5 , FCRL1 , ZAP70 - < 0.001 Neg. reg. of B-cell receptor signaling pathway FCRL3 , CD22 FCGR2B 0.01 Neg. reg. of B-cell proliferation BLK , MNDA , TNFRSF13B FCGR2B 0.006 B-cell differentiation CD79A/B , LYL1 , PAX5 ITGA4 , ITGB1 , FLT3 0.01 Neg. reg. of immunoglobulin production CD22 , CR1 , FCRL3 FCGR2B < 0.001 IgG-mediated phagocytosis PLA2G6 , VAV2 FCGR2B , PAK1 0.008 T-cell mediated immunity CR1 CD8A , JAG1 0.04 Positive regulation of T-cell differentiation CD27 , GATA3 , ZAP70 IL1RL2 , XBP1 0.006 Innate immunity Dendritic cell chemotaxis CCR6 , CCR7 , GPR183 - 0.04 Cytokine response Cellular response to TNF ADAMTS7 , CD40 , GATA3 CD58 , CCL14 , GBP1 , NFE2L2 , NR1D1 , RORA 0.007 Cellular response to IL-1 ADAMTS7 , CD40 CCL14 , GBP1 , NR1D1 , RORA 0.02 Endoplasmic reticulum ER unfolded protein response DERL3 ATF6 , CREB3L2 , HSPA5 , NFE2L2 , XBP1 0.003 ERAD pathway BRSK2 ATF6 , SEL1L , XBP1 0.007 ER stress-induced upregulation of transcription - ATF6 , HSPA5 , XBP1 0.02 Signaling Reg. small GTPase mediated signal transduction ARHGEF9 , PLEKHG4B , SWAP70 , VAV2 AMOT , ARHGAP32 , FGD4 , PLEKHG3 , RALGAPA2 0.003 Pos. reg. of SMAD protein signal transduction ACVR2A , BMP8A BMPR2 , HFE , TGFBR1 , TGFBR2 0.003 Pos. reg. of PI3K activity CCR7 , CD19 , VAV2 FLT3 , PDGFB 0.003 Pos. reg. of JNK cascade CCR7 , CD27 , SH3RF3 , TNFSF11 CRACR2A , F2RL1 , FCGR2B , MFHAS1 0.003 Pos. reg. of MAP kinase activity AJUBA , CD40 , TNFSF11 ADAM9 , FLT3 , PDGFB , TENM1 0.004 Pos. reg. of I-kappaB kinase/NF-kappaB AJUBA , APOL3 , CCR7 , CD40 , TNFSF11 CD4 , F2RL1 , NDFIP2 , S100A4 , TAB3 0.02 1 Only pathways relevant for immune cells are presented here. In total, 85 functional pathways were significantly affected by dominance and not by parity

Due to the limited number of DEGs found on L12 for dominance status and parity, we did not submit the genes to enrichment analysis. The 17 genes related to both factors included genes linked to cytokines ( IL9R , TNFSF11 ), T-cell receptor ( TRDC ), neutrophil granules ( CTSG ), transcription factors ( GATA3 ), transport ( SLC4A4 ), signaling ( RRAGD , PDZD2 ), cell cycle control ( LZTS1 ), adhesion ( CASS4 , L1CAM ), proliferation and differentiation ( BLK , IGF2BP3 ), erythrocytes ( RHEX ), coiled-coil domain-containing ( CCDC158 ) and one uncharacterized gene ( ENSSSCG00000023790 ). By investigating the functions of the eight genes exclusive to dominance status, we identified one gene associated with a macromolecular complex that binds to both microtubules and cytoplasmic dynein ( ACTR1B ), one associated with phospholipid transport ( PITPNM3 ), one associated with an immune cell receptor ( CD8A ), two associated with cell-adhesion genes ( CEACAM1 and CDH23 ), two associated with a complement family gene ( C4BPB and CR1 ) and one associated with the nonannotated gene ENSSSCG00000033104 . These last five genes were consistently affected by dominance across both dates (G98 and L12).

Discussion

Environmental enrichment benefited sows’ welfare

The present study investigated the effects of environmental enrichment and dominance status on the peripheral blood mononuclear cell transcriptome in gestating and lactating sows. In this study, we used gestating sows because they are adult animals, with usually less infectious problems than younger animals that could interfere with their immune function. Furthermore, using multiparous sows who were exposed to the enriched or poor environments for successive gestations allowed investigating the long-term effects of enrichment over months and not those of novelty. Our findings highlight the positive impact of environmental enrichment on sows’ welfare and are in accordance with previous research [ 37 – 39 ]. The lower cortisol concentrations in saliva and hair, coupled with decreased stereotypic behaviors among the enriched group, were likely a result of reduced frustration and boredom attributed to the provision of a foraging substrate [ 40 ]. Moreover, sows are rationed on feed to prevent excessive weight gain during gestation, and the ingestion of straw potentially helps sows alleviate hunger [ 39 ]. Consistent with other studies in pigs, environmental enrichment (straw) stimulated the exploratory behavior of the pen while reducing the usage of chains, which are suboptimal enrichment materials [ 17 , 41 ]. Similarly, the decreased social exploration (nosing-snout) could be because the sows could largely satisfy their exploration needs on the straw and therefore turn away from exploring conspecifics. Additionally, sows in the enriched environment experienced fewer negative social interactions, possibly attributed to increased space allowance, thereby decreasing aggressive encounters and facilitating avoidance behaviors [ 17 , 42 ]. We assume that the positive impact of environmental enrichment on cortisol, stereotypies and exploratory behavior reveals an improved welfare.

Environmental enrichment modulated a few genes associated with anxiety and inflammation

Although environmental enrichment significantly impacted sows’ cortisol and behavior, it had a modest effect on the PBMC transcriptome on G98, influencing only 31 DEGs. This seems to be in agreement with investigations in other species [ 43 – 45 ]. Two genes directly involved in the immune response were downregulated in sows in the enriched environment: TLR6 and CD55 . In addition to their immune roles, these genes can be linked to central nervous system functions. Toll-like receptors (TLRs), which are expressed on antigen-presenting cells, recognize pathogen-associated molecular patterns [ 46 ], and their expression have been associated with major depressive disorders [ 47 , 48 ]. CD55 recognizes C4b and C3b complement fragments, reducing complement system activation [ 49 ]. Alterations in complement proteins and activation have been reported in the blood of individuals with various psychiatric disorders in humans [ 50 – 52 ], learned fear responses in mice [ 53 ] and in pigs with distinct coping styles [ 22 ]. Two other genes associated with neuropsychiatric disorders, HINT1 and MCTP1 , were upregulated and downregulated, respectively, in the enriched group. In mice, HINT1 mutations have been linked to increased anxiety [ 54 ], and its’ expression in the brain changes under stressful conditions, such as social isolation or altered light exposure [ 55 , 56 ]. MCTP1 overexpression has been proposed as a candidate gene for anxiety in mice and bipolar disorder in humans [ 57 ]. Evaluating the expression of these genes and their network in other contexts will be crucial in investigating their potential role in revealing positive or negative welfare states.

Other genes not directly linked to immune function, but associated with proinflammatory conditions were upregulated in sows in the enriched environment: PIGF [ 58 ], AHRR [ 59 ], and ZCCHC9 [ 60 ]. The upregulation of LDHB , which is involved in converting pyruvate to lactate during glycolysis, and the downregulation of PXMP4 , which is associated with lipid metabolism [ 61 ], also suggest a metabolic profile of the proinflammatory response [ 62 ].

Dominance status influenced genes related to inflammation and the fate of B cells

Dominant sows displayed greater cortisol levels on G35 than other sows, which has already been observed by others, as has the disappearance of these effects later in gestation [ 63 ]. The impact of dominance on cortisol is highly dependent on the social context. For example, in primates, social instability of the group and social history, rather than dominance itself, could be responsible for greater perceived stress and higher cortisol concentrations [ 64 ]. Higher cortisol levels in dominant sows might be related to the cost of maintaining their social position. In addition to cortisol, the dominant sows were behaviorally less active (less social exploration, less nose‒nose interaction, less locomotory behavior), which might reflect the fact that they were also the oldest. Both cortisol and behavioral differences between dominants and subordinate sows indicated that dominance has a wide influence on the way an individual behaves and perceives its environment. This may have an indirect influence on their immune cells [ 63 , 65 ].

In groups of sows, higher social ranks is often the attribute of the oldest sows, while animals with lower social ranks are younger [ 29 ]. This is also what we observed in the groups of sows used for this study. It was therefore highly likely that part of the effects attributed here to dominance were in fact the result of another hidden factor, namely the reproductive history of the sows. One of the weaknesses of our experimental design was that it was impossible to include the ‘parity’ and ‘dominance’ factors in the same statistical model. For this reason, we analyzed successively the transcriptomic dataset with two separate models, one analyzing the effect of dominance (DOM, INT and SUB levels), and one considering the effect of parity (the HP group containing mainly DOM and INT sows, and the LP group mainly SUB and INT sows). We attributed the differences in gene expression observed in the first model to dominance when these genes did not appear in the list of genes affected by parity in the second model. However, this approach is not powerful and our results must be considered with caution. We identified a strong global signature of dominance status on the gene expression data, with 269 genes differentially expressed by dominance but not by parity at G98. In particular, the dominant sows displayed increased expression of genes and biological functions linked to inflammatory cytokines (IL-1 and TNF-α) and inflammatory signaling pathways (JNK and MAP kinases), suggesting a greater stress level in the dominant sows, which is also in agreement with their greater cortisol concentration on G35. The effect of dominance on blood mononuclear cell transcriptomic profiles has also been reported in primates, but contrary to our findings, a high proinflammatory gene expression was associated with a lower dominance status [ 66 , 67 ]. The same has been reported in mice [ 68 ] and in humans [ 11 , 69 ]. The discrepancies between species might reflect the different implications of being dominant or subordinate depending on social groups and species.

Our findings showed that other enriched biological processes were exclusively associated with dominance status. Many DEGs related to B lymphocyte activity were downregulated in the dominant sows. It is possible that some variations in the expression of genes linked to the B-cell lineage were simply due to lower B-cell proportions in the PBMCs of dominant sows than subordinate sows. We did not assess lymphocyte subset proportions in this study, and deconvolution tools for porcine blood cells are not currently available. However, the global downregulation of some genes specific to B cells, such as Fc receptor-like molecules ( FCRL1 , FCRL3 and FCRL4 ), suggests a potential decrease in the number of circulating B cells. Prior studies have indicated that dominance status influences blood B-cell numbers in pregnant sows [ 63 ] and male mice [ 68 ] but in opposing directions (e.g., higher cell counts in middle-ranking or dominant animals). In contrast to B lymphocyte activation and proliferation functions, the process of B-cell differentiation into plasma cells seemed to be enhanced in dominant sows. The distinct transcriptomes of B cells and plasma cells are regulated by opposing sets of transcription factors: those that maintain the B-cell program, including PAX5 , BACH2 , and BCL6 , and those that support plasma cell differentiation, such as IRF4 , BLIMP1 , and XBP1 [ 70 ]. In the dominant sows, PAX5 was downregulated, while PRDM1 (or BLIMP1 ) and XBP1 were upregulated. Furthermore, highly secretory cell types, such as antibody-secreting cells, exhibit increased endoplasmic reticulum (ER) stress and activation of the unfolded protein response (UPR) pathway. Accordingly, we identified enriched ER-related biological processes exclusively in the dominance comparison. The effect of dominance on immunoglobulin production in sows has already been reported several times in the literature, although the direction of this effect varies from one study to another [ 65 , 71 ]. Nonetheless, a shift in B lymphocyte functions was not observed when comparing low- to high-parity sows, suggesting an effect specific to dominance status. Interestingly, both the inflammatory response and immunoglobulin production are functions that are affected in the CTRA in humans [ 72 ].

Some effects apparently associated with dominance were more likely due to sows’ parity

For both dominance status and parity, DEGs associated with T-cell differentiation and activation, macrophage differentiation and cellular responses to IL-4 and IL-5 were affected, which demonstrated that these functions might be altered by the parity or age of the animals. The dominant sows in this study were also the oldest and heaviest and had higher parity ranks, which are characteristics that have been reported before [ 29 ]. However, it is important to note that all females involved in the study were relatively young: low-parity sows were very young females concluding their growth phase, while the oldest high-parity sows were approximately three years old, which is still very young from the point of view of the animal, considering that a pigs’ lifespan can reach 15 years. Conversely, the transcriptomic responses noted in high-parity sows seemed more connected to the accumulation of pregnancies rather than aging. During pregnancy and lactation, the concentrations of hormones such as estrogen, progesterone, prolactin and leptin vary sharply and influence immune cells [ 73 ]. In addition to short-term effects, successive gestations have irreversible effects on female physiology. In women, high multiparity has been linked to adverse health effects [ 74 , 75 ] and may lead to accelerated epigenetic aging [ 76 ]. Hormonal variations due to pregnancy have also been shown to interfere with cytokine production patterns during aging, resulting in increased IL-4, IL-6, and IFN-γ production in mice [ 77 ]. This parity-related effect could explain our observations in high-parity sows, where functions such as “cellular response to IL-4” and “positive regulation of IFN-γ production” were affected.

The imprint of social and environmental factors fades during lactation

The effects of environmental enrichment and dominance status were relatively short-term. Three weeks after the transfer to the maternity room, the number of genes affected by both factors decreased from 31 to 449 on G98 to four and 25 on L12, respectively. In the maternity rooms, both treatment groups were placed in individual farrowing crates with their piglets. Consequently, their previous dominance status was probably no longer relevant, and the effects of dominance status on gene expression also disappeared. The disappearance of the dominance effects aligns with observations in rhesus macaques, which indicated that the gene expression profile associated with dominance was not fixed and could change if the dominance rank changed [ 67 ]. Moreover, farrowing crates restrict sows’ movement and natural behavior, potentially contributing to the disappearance of environmental enrichment effects. However, a similar reduction in parity-related DEGs (from 542 on G98 to 91 DEGs on L12) suggested that these fading effects might also be due to the substantial metabolic and endocrine reprogramming occurring between pregnancy and lactation. Although not discussed in this study, this transition resulted in a significant transcriptional shift, as observed in cows [ 78 ].

Conclusion

In conclusion, this study showed that environmental enrichment, despite having positive effects on the welfare of sows, had a limited impact on their blood transcriptome. On the other hand, we were able to demonstrate the major effects of the dominance status of sows within their social group. Some effects, notably on T-cell responses, were likely tied to parity rank, while others, such as the impact on B lymphocyte differentiation and some inflammatory clues, seemed specific to dominance. This study is a reminder that, for social animals kept in captivity, social relationships remain extremely powerful factors influencing brain and cognitive processes and, by ricochet, immune function. Nevertheless, because parity rank and dominance rank often vary collinearly, further studies would be useful to better dissociate the role of each of these two factors, one psychosocial, the other physiological, in the final immune phenotype of female mammals of reproductive age.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Supplementary Material 2

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Acknowledgements

The authors are grateful to C Gerard and N Villain, engineers of Brittany’s Regional Agricultural Chamber (CRAB), and C Connan and P Lirzin, staff of the experimental farm of CRAB, for their help in following up the experiment, sampling, and data collection. We also thank C Trombani, veterinarian (Breizh Pig, France), for his help during saliva and blood samplings. We acknowledge as well R Comte, C Guerin, L Le Normand, F Mayeur-Nickel, and P Touanel for their technical assistance and C Diot for his help in analyzing RNA-Seq data (PEGASE, INRAE, Saint-Gilles, France). We are also grateful to the sows who took part in this study.

Author contributions

M Mescouto Lopes: conceived and designed the analysis, collected the data, contributed data or analysis tools, performed the analysis, and wrote the paper. C Clouard: conceived and designed the analysis, collected the data, contributed data or analysis tools, and performed the analysis. A Vincent: collected the data, contributed data, or analysis tools. F Thomas: collected the data. F Hérault, I Louveau and R Resmond: contributed data or analysis tools. H Jammes: conceived and designed the analysis. E Merlot: conceived and designed the analysis, collected the data, performed the analysis, wrote the paper, and acquired funding. All the authors have read and approved the final manuscript.

Funding

This work was supported by INRAE (SANBA Health and Welfare of Farmed Livestock program) and the Brittany Region.

Data availability

Sequencing data is available in the NCBI Gene Expression Omnibus (GEO; https://www.ncbi.nlm.nih.gov/geo) repository under accession number GSE234561 . The blood assays, RNA-seq count tables, and results of the analyses with R and DAVID software are available in a data paper (https://doi.org/10.1016/j.anopes.2024.100078). The corresponding tables have been deposited at Recherche Data Gouv repository (https://entrepot.recherche.data.gouv.fr/dataverse/Thesis_Mescouto_Lopes) under accession number https://doi.org/10.57745/NHE0MP.

Declarations

Ethics approval

The experiment was carried out at the experimental farm of the Chambre Régionale d’Agriculture de Bretagne (CRAB) at Crécom (Saint - Nicolas du Pélem, France) and complied with the EU Directive 2010/63/EU for animal experiments. This study was approved by the French Regional Ethics Committee in Animal Experiments no. 007 and by the French Ministry of Higher Education and Research (APAFIS#27038-2020090315423669 v2).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

References

1. MacGillivray DM Kollmann TR The role of environmental factors in modulating immune responses in early life Front Immunol 2014 5 434 10.3389/fimmu.2014.00434 25309535 PMC4161944 MacGillivray DM, Kollmann TR. The role of environmental factors in modulating immune responses in early life. Front Immunol. 2014;5:434. 25309535 10.3389/fimmu.2014.00434 PMC4161944

2. Muscatell KA. Social psychoneuroimmunology: Understanding bidirectional links between social experiences and the immune system. Brain Behav Immun [Internet]. 2021;93:1–3. https://www.sciencedirect.com/science/article/pii/S0889159120324764 10.1016/j.bbi.2020.12.023 33383146

3. Dhabhar FS Miller AH McEwen BS Spencer RL Effects of stress on immune cell distribution. Dynamics and hormonal mechanisms J Immunol 1995 154 10 5511 27 10.4049/jimmunol.154.10.5511 7730652 Dhabhar FS, Miller AH, McEwen BS, Spencer RL. Effects of stress on immune cell distribution. Dynamics and hormonal mechanisms. J Immunol. 1995;154(10):5511–27. 7730652

4. Dhabhar FS Enhancing versus suppressive effects of stress on immune function: implications for immunoprotection versus immunopathology Allergy Asthma Clin Immunol 2008 4 1 2 11 10.1186/1710-1492-4-1-2 20525121 PMC2869337 Dhabhar FS. Enhancing versus suppressive effects of stress on immune function: implications for immunoprotection versus immunopathology. Allergy Asthma Clin Immunol. 2008;4(1):2–11. 20525121 10.1186/1710-1492-4-1-2 PMC2869337

5. Dhabhar FS Viswanathan K Short-term stress experienced at time of immunization induces a long-lasting increase in immunologic memory Am J Physiol - Regul Integr Comp Physiol 2005 289 3 58–3 738 44 10.1152/ajpregu.00145.2005 15890793 Dhabhar FS, Viswanathan K. Short-term stress experienced at time of immunization induces a long-lasting increase in immunologic memory. Am J Physiol - Regul Integr Comp Physiol. 2005;289(3 58–3):738–44. 10.1152/ajpregu.00145.2005 15890793

6. Steptoe A Hamer M Chida Y The effects of acute psychological stress on circulating inflammatory factors in humans: a review and meta-analysis Brain Behav Immun 2007 21 7 901 12 10.1016/j.bbi.2007.03.011 17475444 Steptoe A, Hamer M, Chida Y. The effects of acute psychological stress on circulating inflammatory factors in humans: a review and meta-analysis. Brain Behav Immun. 2007;21(7):901–12. 17475444 10.1016/j.bbi.2007.03.011

7. Kanitz E Tuchscherer M Puppe B Tuchscherer A Stabenow B Consequences of repeated early isolation in domestic piglets (Sus scrofa) on their behavioural, neuroendocrine, and immunological responses Brain Behav Immun 2004 18 1 35 45 10.1016/S0889-1591(03)00085-0 14651945 Kanitz E, Tuchscherer M, Puppe B, Tuchscherer A, Stabenow B. Consequences of repeated early isolation in domestic piglets (Sus scrofa) on their behavioural, neuroendocrine, and immunological responses. Brain Behav Immun. 2004;18(1):35–45. 14651945 10.1016/s0889-1591(03)00085-0

8. Sutherland MA Niekamp SR Rodriguez-Zas SL Salak-Johnson JL Impacts of chronic stress and social status on various physiological and performance measures in pigs of different breeds J Anim Sci 2006 84 3 588 96 10.2527/2006.843588x 16478950 Sutherland MA, Niekamp SR, Rodriguez-Zas SL, Salak-Johnson JL. Impacts of chronic stress and social status on various physiological and performance measures in pigs of different breeds. J Anim Sci. 2006;84(3):588–96. 16478950 10.2527/2006.843588x

9. Avitsur R Stark JL Dhabhar FS Sheridan JF Social stress alters splenocyte phenotype and function J Neuroimmunol 2002 132 1–2 66 71 10.1016/S0165-5728(02)00310-7 12417435 Avitsur R, Stark JL, Dhabhar FS, Sheridan JF. Social stress alters splenocyte phenotype and function. J Neuroimmunol. 2002;132(1–2):66–71. 12417435 10.1016/s0165-5728(02)00310-7

10. Mehta D, Voisey J, Bruenig D, Harvey W, Morris CP, Lawford B et al. Transcriptome analysis reveals novel genes and immune networks dysregulated in veterans with PTSD. Brain Behav Immun [Internet]. 2018;74(August):133–42. 10.1016/j.bbi.2018.08.014 10.1016/j.bbi.2018.08.014 30189241

11. Miller GE Chen E Fok AK Walker H Lim A Nicholls EF Low early-life social class leaves a biological residue manifested by decreased glucocorticoid and increased proinflammatory signaling Proc Natl Acad Sci 2009 106 34 14716 21 10.1073/pnas.0902971106 19617551 PMC2732821 Miller GE, Chen E, Fok AK, Walker H, Lim A, Nicholls EF, et al. Low early-life social class leaves a biological residue manifested by decreased glucocorticoid and increased proinflammatory signaling. Proc Natl Acad Sci. 2009;106(34):14716–21. 19617551 10.1073/pnas.0902971106 PMC2732821

12. Cole SW Conti G Arevalo JMG Ruggiero AM Heckman JJ Suomi SJ Transcriptional modulation of the developing immune system by early life social adversity Proc Natl Acad Sci U S A 2012 109 50 20578 83 10.1073/pnas.1218253109 23184974 PMC3528538 Cole SW, Conti G, Arevalo JMG, Ruggiero AM, Heckman JJ, Suomi SJ. Transcriptional modulation of the developing immune system by early life social adversity. Proc Natl Acad Sci U S A. 2012;109(50):20578–83. 23184974 10.1073/pnas.1218253109 PMC3528538

13. O’Connor MF, Schultze-Florey CR, Irwin MR, Arevalo JMG, Cole SW. Divergent gene expression responses to Complicated Grief and Non-complicated Grief. Brain Behav Immun [Internet]. 2014;37:78–83. 10.1016/j.bbi.2013.12.017 10.1016/j.bbi.2013.12.017 PMC4015725 24380850

14. Cole SW. The Conserved Transcriptional Response to Adversity. Curr Opin Behav Sci [Internet]. 2019;28:31–7. 10.1016/j.cobeha.2019.01.008 10.1016/j.cobeha.2019.01.008 PMC6779418 31592179

15. Snodgrass JG, Lacy MG, Dengah HJF, Polzer ER, Else RJ, Arevalo JMG et al. Positive mental well-being and immune transcriptional profiles in highly involved videogame players. Brain Behav Immun [Internet]. 2019;82:84–92. https://www.sciencedirect.com/science/article/pii/S0889159119304210 10.1016/j.bbi.2019.07.035 PMC6800642 31376495

16. Mason G, Clubb R, Latham N, Vickery S. Why and how should we use environmental enrichment to tackle stereotypic behaviour? Appl Anim Behav Sci [Internet]. 2007;102(3):163–88. https://www.sciencedirect.com/science/article/pii/S0168159106001900

17. Godyń D Nowicki J Herbut P Effects of environmental enrichment on pig welfare—a review Animals 2019 9 6 383 10.3390/ani9060383 31234475 PMC6616547 Godyń D, Nowicki J, Herbut P. Effects of environmental enrichment on pig welfare—a review. Animals. 2019;9(6):383. 31234475 10.3390/ani9060383 PMC6616547

18. Luo L Reimert I Middelkoop A Kemp B Bolhuis JE Effects of Early and Current Environmental Enrichment on Behavior and Growth in pigs Front Vet Sci 2020 7 June 1 14 32582773 10.3389/fvets.2020.00268 PMC7287207 Luo L, Reimert I, Middelkoop A, Kemp B, Bolhuis JE. Effects of Early and Current Environmental Enrichment on Behavior and Growth in pigs. Front Vet Sci. 2020;7(June):1–14. 32582773 10.3389/fvets.2020.00268 PMC7287207

19. Luo L, Reimert I, de Haas EN, Kemp B, Bolhuis JE. Effects of early and later life environmental enrichment and personality on attention bias in pigs (Sus scrofa domesticus). Anim Cogn [Internet]. 2019;22(6):959–72. 10.1007/s10071-019-01287-w 10.1007/s10071-019-01287-w PMC6834757 31250144

20. Ernst K, Tuchscherer M, Kanitz E, Puppe B, Manteuffel G. Effects of attention and rewarded activity on immune parameters and wound healing in pigs. Physiol Behav [Internet]. 2006;89(3):448–56. https://www.sciencedirect.com/science/article/pii/S0031938406002903 10.1016/j.physbeh.2006.07.001 16904140

21. Luo L Van Dixhoorn IDE Reimert I Kemp B Bolhuis JE Parmentier HK Effect of enriched housing on levels of natural (auto-)antibodies in pigs co-infected with porcine reproductive and respiratory syndrome virus (PRRSV) and Actinobacillus pleuropneumoniae Vet Res 2017 48 1 1 10 10.1186/s13567-017-0481-y 29126442 PMC5681819 Luo L, Van Dixhoorn IDE, Reimert I, Kemp B, Bolhuis JE, Parmentier HK. Effect of enriched housing on levels of natural (auto-)antibodies in pigs co-infected with porcine reproductive and respiratory syndrome virus (PRRSV) and Actinobacillus pleuropneumoniae. Vet Res. 2017;48(1):1–10. 29126442 10.1186/s13567-017-0481-y PMC5681819

22. Merlot E Vincent A Thomas F Meunier-Salaün MC Damon M Robert F Health and immune traits of basque and large White pigs housed in a conventional or enriched environment Animal 2012 6 8 1290 9 10.1017/S1751731112000080 23217232 Merlot E, Vincent A, Thomas F, Meunier-Salaün MC, Damon M, Robert F, et al. Health and immune traits of basque and large White pigs housed in a conventional or enriched environment. Animal. 2012;6(8):1290–9. 23217232 10.1017/S1751731112000080

23. Reimert I, Rodenburg TB, Ursinus WW, Kemp B, Bolhuis JE. Selection based on indirect genetic effects for growth, environmental enrichment and coping style affect the immune status of pigs. PLoS ONE. 2014;9(10). 10.1371/journal.pone.0108700 PMC4183536 25275507

24. Olsson IAS Westlund K More than numbers matter: the effect of social factors on behaviour and welfare of laboratory rodents and non-human primates Appl Anim Behav Sci 2007 103 3–4 229 54 10.1016/j.applanim.2006.05.022 Olsson IAS, Westlund K. More than numbers matter: the effect of social factors on behaviour and welfare of laboratory rodents and non-human primates. Appl Anim Behav Sci. 2007;103(3–4):229–54.

25. Yin YY, Lai ZK, Yan JZ, Wei QQ, Wang B, Zhang LM et al. The interaction between social hierarchy and depression/anxiety: involvement of glutamatergic pyramidal neurons in the medial prefrontal cortex (mPFC). Neurobiol Stress. 2023;24(November 2022). 10.1016/j.ynstr.2023.100536 PMC10085780 37057073

26. Ruis MAW Te Brake JHA Engel B Buist WG Blokhuis HJ Koolhaas JM Implications of coping characteristics and social status for welfare and production of paired growing gilts Appl Anim Behav Sci 2002 75 3 207 31 10.1016/S0168-1591(01)00191-5 Ruis MAW, Te Brake JHA, Engel B, Buist WG, Blokhuis HJ, Koolhaas JM. Implications of coping characteristics and social status for welfare and production of paired growing gilts. Appl Anim Behav Sci. 2002;75(3):207–31.

27. Zhou T, Sandi C, Hu H. Advances in understanding neural mechanisms of social dominance. Curr Opin Neurobiol [Internet]. 2018;49:99–107. 10.1016/j.conb.2018.01.006 10.1016/j.conb.2018.01.006 29428628

28. Pacheco E Lopez M Salak-Johnson JL Social Status differentially affects behavioral and Immunological Outcomes of Group-Kept Sows Fed different dietary Fiber using different length feeding barriers Front Anim Sci 2021 2 September 1 12 Pacheco E, Lopez M, Salak-Johnson JL. Social Status differentially affects behavioral and Immunological Outcomes of Group-Kept Sows Fed different dietary Fiber using different length feeding barriers. Front Anim Sci. 2021;2(September):1–12.

29. Zanella AJ Brunner P Unshelm J Mendl MT Broom DM The relationship between housing and social rank on cortisol, β-endorphin and dynorphin (1–13) secretion in sows Appl Anim Behav Sci 1998 59 1–3 1 10 10.1016/S0168-1591(98)00115-4 Zanella AJ, Brunner P, Unshelm J, Mendl MT, Broom DM. The relationship between housing and social rank on cortisol, β-endorphin and dynorphin (1–13) secretion in sows. Appl Anim Behav Sci. 1998;59(1–3):1–10.

30. Lanthony M, Danglot M, Špinka M, Tallet C. Dominance hierarchy in groups of pregnant sows: characteristics and identification of related indicators. Appl Anim Behav Sci. 2022;254(June).

31. Heimbürge S Kanitz E Tuchscherer A Otten W Within a hair’s breadth – factors influencing hair cortisol levels in pigs and cattle Gen Comp Endocrinol 2020 288 113359 10.1016/j.ygcen.2019.113359 31830475 Heimbürge S, Kanitz E, Tuchscherer A, Otten W. Within a hair’s breadth – factors influencing hair cortisol levels in pigs and cattle. Gen Comp Endocrinol. 2020;288:113359. 31830475 10.1016/j.ygcen.2019.113359

32. Martin P, Bateson P. Recording methods. Measuring Behaviour: an introductory guide. Cambridge University Press; 2007. pp. 48–61.

33. Core TRD. A Language and Environment for Statistical Computing [Internet]. Vol. 2, R Foundation for Statistical Computing. Vienna, Austria; 2018. p. https://www.R-project.org. Available from: http://www.r-project.org .

34. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol [Internet]. 2014 Dec 5 [cited 2023 Jun 21];15(12):1–21. https://genomebiology.biomedcentral.com/articles/ 10.1186/s13059-014-0550-8 10.1186/s13059-014-0550-8 PMC4302049 25516281

35. Benjamini Y, Hochberg Y. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. J R Stat Soc Ser B [Internet]. 1995 Jan 1 [cited 2023 Jul 21];57(1):289–300. https://onlinelibrary.wiley.com/doi/full/ 10.1111/j.2517-6161.1995.tb02031.x

36. Dennis G, Sherman BT, Hosack DA, Yang J, Gao W, Lane HC et al. DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biol [Internet]. 2003 Aug 14 [cited 2023 Jun 21];4(5):1–11. https://genomebiology.biomedcentral.com/articles/ 10.1186/gb-2003-4-9-r60 12734009

37. Lopes MM, Vincent A, Thomas F, Clouard C, Comte R, Brien M et al. Data paper: Dataset describing the effects of environmental enrichment and sows’ characteristics on the peripheral blood mononuclear cell transcriptome. Anim - Open Sp [Internet]. 2024;3:100078. https://www.sciencedirect.com/science/article/pii/S2772694024000189

38. Merlot E Calvar C Prunier A Influence of the housing environment during sow gestation on maternal health, and offspring immunity and survival Anim Prod Sci 2017 57 8 1751 8 10.1071/AN15480 Merlot E, Calvar C, Prunier A. Influence of the housing environment during sow gestation on maternal health, and offspring immunity and survival. Anim Prod Sci. 2017;57(8):1751–8.

39. Merlot E Pastorelli H Prunier A Père MC Louveau I Lefaucheur L Sow environment during gestation: part I. Influence on maternal physiology and lacteal secretions in relation with neonatal survival Animal 2019 13 7 1432 9 10.1017/S1751731118002987 30468144 Merlot E, Pastorelli H, Prunier A, Père MC, Louveau I, Lefaucheur L, et al. Sow environment during gestation: part I. Influence on maternal physiology and lacteal secretions in relation with neonatal survival. Animal. 2019;13(7):1432–9. 30468144 10.1017/S1751731118002987

40. Quesnel H, Peuteman B, Père MC, Louveau I, Lefaucheur L, Perruchot MH et al. Effect of environmental enrichment with wood materials and straw pellets on the metabolic status of sows during gestation. Livest Sci [Internet]. 2019;229(September):43–8. 10.1016/j.livsci.2019.09.005

41. Merlot E, Meunier-Salaün MC, Peuteman B, Père MC, Louveau I, Perruchot MH et al. Improving maternal welfare during gestation has positive outcomes on neonatal survival and modulates offspring immune response in pigs. Physiol Behav. 2022;249(February). 10.1016/j.physbeh.2022.113751 35217067

42. Beattie VE O’Connell NE Kilpatrick DJ Moss BW Influence of environmental enrichment on welfare-related behavioural and physiological parameters in growing pigs Anim Sci 2000 70 3 443 50 10.1017/S1357729800051791 Beattie VE, O’Connell NE, Kilpatrick DJ, Moss BW. Influence of environmental enrichment on welfare-related behavioural and physiological parameters in growing pigs. Anim Sci. 2000;70(3):443–50.

43. Remience V Wavreille J Canart B Meunier-Salaün MC Prunier A Bartiaux-Thill N Effects of space allowance on the welfare of dry sows kept in dynamic groups and fed with an electronic sow feeder Appl Anim Behav Sci 2008 112 3–4 284 96 10.1016/j.applanim.2007.07.006 Remience V, Wavreille J, Canart B, Meunier-Salaün MC, Prunier A, Bartiaux-Thill N, et al. Effects of space allowance on the welfare of dry sows kept in dynamic groups and fed with an electronic sow feeder. Appl Anim Behav Sci. 2008;112(3–4):284–96.

44. Brod S Gobbetti T Gittens B Ono M Perretti M D’Acquisto F The impact of environmental enrichment on the murine inflammatory immune response JCI Insight 2017 2 7 e90723 10.1172/jci.insight.90723 28405616 PMC5374068 Brod S, Gobbetti T, Gittens B, Ono M, Perretti M, D’Acquisto F. The impact of environmental enrichment on the murine inflammatory immune response. JCI Insight. 2017;2(7):e90723. 28405616 10.1172/jci.insight.90723 PMC5374068

45. Rattazzi L, Piras G, Brod S, Smith K, Ono M, D’Acquisto F. Impact of enriched environment on murine T cell differentiation and gene expression profile. Front Immunol. 2016;7(SEP). 10.3389/fimmu.2016.00381 PMC5042968 27746779

46. Lansade L, Valenchon M, Foury A, Neveux C, Cole SW, Layé S et al. Behavioral and Transcriptomic Fingerprints of an Enriched Environment in Horses (Equus caballus). PLoS One [Internet]. 2014;9(12):e114384. 10.1371/journal.pone.0114384 10.1371/journal.pone.0114384 PMC4262392 25494179

47. Kawai T Akira S Signaling to NF-kappaB by toll-like receptors Trends Mol Med 2007 13 11 460 9 10.1016/j.molmed.2007.09.002 18029230 Kawai T, Akira S. Signaling to NF-kappaB by toll-like receptors. Trends Mol Med. 2007;13(11):460–9. 18029230 10.1016/j.molmed.2007.09.002

48. Figueroa-Hall LK Paulus MP Savitz J Toll-like receptor signaling in Depression Psychoneuroendocrinology 2020 121 104843 10.1016/j.psyneuen.2020.104843 32911436 PMC7883590 Figueroa-Hall LK, Paulus MP, Savitz J. Toll-like receptor signaling in Depression. Psychoneuroendocrinology. 2020;121:104843. 32911436 10.1016/j.psyneuen.2020.104843 PMC7883590

49. Hung Y-Y, Huang K-W, Kang H-Y, Huang GY-L, Huang T-L. Antidepressants normalize elevated toll-like receptor profile in major depressive disorder. Psychopharmacology. 2016;233(9):1707–14. 10.1007/s00213-015-4087-7 PMC4828490 26415953

50. Nicholson-Weller A, Wang CE. Structure and function of decay accelerating factor CD55. J Lab Clin Med [Internet]. 1994 Apr 1 [cited 2023 Sep 28];123(4):485–91. https://europepmc.org/article/med/7511675 7511675

51. Ji E, Boerrigter D, Cai HQ, Lloyd D, Bruggemann J, O’Donnell M et al. Peripheral complement is increased in schizophrenia and inversely related to cortical thickness. Brain Behav Immun [Internet]. 2022;101(November 2021):423–34. 10.1016/j.bbi.2021.11.014 10.1016/j.bbi.2021.11.014 34808287

52. Akcan U, Karabulut S, Tüzün E, Çakır S, İsmail Küçükali C. Bipolar disorder patients display reduced serum complement levels and elevated peripheral blood complement expression levels. Acta Neuropsychiatr [Internet]. 2017/04/12. 2018;30(2):70–8. https://www.cambridge.org/core/product/4CE70277F0BFBFF72BD4E4EE769ADCD4 10.1017/neu.2017.10 28399943

53. Luo X, Fang Z, Lin L, Xu H, Huang Q, Zhang H. Plasma complement C3 and C3a are increased in major depressive disorder independent of childhood trauma. BMC Psychiatry [Internet]. 2022;22(1):1–8. 10.1186/s12888-022-04410-3 10.1186/s12888-022-04410-3 PMC9706857 36447174

54. Westacott LJ, Humby T, Haan N, Brain SA, Bush EL, Toneva M et al. Complement C3 and C3aR mediate different aspects of emotional behaviours; relevance to risk for psychiatric disorder. Brain Behav Immun [Internet]. 2022;99(July 2021):70–82. 10.1016/j.bbi.2021.09.005 10.1016/j.bbi.2021.09.005 34543680

55. Varadarajulu J, Lebar M, Krishnamoorthy G, Habelt S, Lu J, Bernard Weinstein I et al. Increased anxiety-related behaviour in Hint1 knockout mice. Behav Brain Res [Internet]. 2011;220(2):305–11. 10.1016/j.bbr.2011.02.012 10.1016/j.bbr.2011.02.012 21316396

56. Dang Yhui, Liu P, Ma R, Chu Z, Liu Y, ping, Wang J et al. bei,. HINT1 is involved in the behavioral abnormalities induced by social isolation rearing. Neurosci Lett [Internet]. 2015;607:40–5. 10.1016/j.neulet.2015.08.026 10.1016/j.neulet.2015.08.026 26300541

57. Zhou Y, Zhang HK, Liu F, Lei G, Liu P, Jiao T, et al. Altered light conditions contribute to abnormalities in emotion and cognition through HINT1 dysfunction in C57BL/6 mice. Front Behav Neurosci. 2018;12(June):1–10. 10.3389/fnbeh.2018.00110 PMC6002487 29937721

58. Ashbrook DG, Williams RW, Lu L, Hager R. A cross-species genetic analysis identifies candidate genes for mouse anxiety and human bipolar disorder. Front Behav Neurosci. 2015;9(July). 10.3389/fnbeh.2015.00171 PMC4486840 26190982

59. Kim KJ Cho CS Kim WU Role of placenta growth factor in cancer and inflammation Exp Mol Med 2012 44 1 10 9 10.3858/emm.2012.44.1.023 22217448 PMC3277893 Kim KJ, Cho CS, Kim WU. Role of placenta growth factor in cancer and inflammation. Exp Mol Med. 2012;44(1):10–9. 22217448 10.3858/emm.2012.44.1.023 PMC3277893

60. Vogel CFA Ishihara Y Campbell CE Kado SY Nguyen-Chi A Sweeney C A protective role of aryl hydrocarbon receptor repressor in inflammation and tumor growth Cancers (Basel) 2019 11 5 1 17 10.3390/cancers11050589 PMC6563059 31035533 Vogel CFA, Ishihara Y, Campbell CE, Kado SY, Nguyen-Chi A, Sweeney C, et al. A protective role of aryl hydrocarbon receptor repressor in inflammation and tumor growth. Cancers (Basel). 2019;11(5):1–17. 10.3390/cancers11050589 PMC6563059 31035533

61. Li L Meng H Wang X Ruan J Tian X Meng F Low ZCCHC9 gene expression in Peripheral Blood May be an Acute myocardial infarction genetic molecular marker in patients with stable coronary atherosclerotic disease Int J Gen Med 2022 15 February 1795 804 10.2147/IJGM.S346335 35210844 PMC8863191 Li L, Meng H, Wang X, Ruan J, Tian X, Meng F. Low ZCCHC9 gene expression in Peripheral Blood May be an Acute myocardial infarction genetic molecular marker in patients with stable coronary atherosclerotic disease. Int J Gen Med. 2022;15(February):1795–804. 35210844 10.2147/IJGM.S346335 PMC8863191

62. Fletcher JM Jordan MA Snelgrove SL Slattery RM Dufour FD Kyparissoudis K Congenic analysis of the NKT Cell Control Gene Nkt2 implicates the peroxisomal protein Pxmp4 J Immunol 2008 181 5 3400 12 10.4049/jimmunol.181.5.3400 18714012 Fletcher JM, Jordan MA, Snelgrove SL, Slattery RM, Dufour FD, Kyparissoudis K, et al. Congenic analysis of the NKT Cell Control Gene Nkt2 implicates the peroxisomal protein Pxmp4. J Immunol. 2008;181(5):3400–12. 18714012 10.4049/jimmunol.181.5.3400

63. Soto-Heredero G Gómez de las Heras MM Gabandé-Rodríguez E Oller J Mittelbrunn M Glycolysis – a key player in the inflammatory response FEBS J 2020 287 16 3350 69 10.1111/febs.15327 32255251 PMC7496292 Soto-Heredero G, Gómez de las Heras MM, Gabandé-Rodríguez E, Oller J, Mittelbrunn M. Glycolysis – a key player in the inflammatory response. FEBS J. 2020;287(16):3350–69. 32255251 10.1111/febs.15327 PMC7496292

64. Schalk C Pfaffinger B Schmucker S Weiler U Stefanski V Pregnancy-associated alterations of peripheral blood immune cell numbers in domestic sows are modified by social rank Animals 2019 9 3 1 15 10.3390/ani9030112 PMC6466169 30909543 Schalk C, Pfaffinger B, Schmucker S, Weiler U, Stefanski V. Pregnancy-associated alterations of peripheral blood immune cell numbers in domestic sows are modified by social rank. Animals. 2019;9(3):1–15. 10.3390/ani9030112 PMC6466169 30909543

65. Vandeleest JJ, Winkler SL, Beisner BA, Hannibal DL, Atwill ER, McCowan B. Sex differences in the impact of social status on hair cortisol concentrations in rhesus monkeys (Macaca mulatta). Am J Primatol. 2020;82(1). 10.1002/ajp.23086 PMC6980377 31876328

66. Hjarvard BM Larsen ON Juul-Madsen HR Jørgensen E Jensen KH Social rank influences the distribution of blood leukocyte subsets in female growing pigs Scand J Lab Anim Sci 2009 36 4 309 20 Hjarvard BM, Larsen ON, Juul-Madsen HR, Jørgensen E, Jensen KH. Social rank influences the distribution of blood leukocyte subsets in female growing pigs. Scand J Lab Anim Sci. 2009;36(4):309–20.

67. Snyder-Mackler N, Sanz J, Kohn JN, Brinkworth JF, Morrow S, Shaver AO et al. Social status alters immune regulation and response to infection in macaques. Science (80-) [Internet]. 2016 Nov 25 [cited 2023 Jul 21];354(6315):1041–5. https://www.science.org/doi/ 10.1126/science.aah3580 10.1126/science.aah3580 PMC5498102 27885030

68. Tung J Barreiro LB Johnson ZP Hansen KD Michopoulos V Toufexis D Social environment is associated with gene regulatory variation in the rhesus macaque immune system Proc Natl Acad Sci 2012 109 17 6490 5 10.1073/pnas.1202734109 22493251 PMC3340061 Tung J, Barreiro LB, Johnson ZP, Hansen KD, Michopoulos V, Toufexis D, et al. Social environment is associated with gene regulatory variation in the rhesus macaque immune system. Proc Natl Acad Sci. 2012;109(17):6490–5. 22493251 10.1073/pnas.1202734109 PMC3340061

69. Lee W Milewski TM Dwortz MF Young RL Gaudet AD Fonken LK Distinct immune and transcriptomic profiles in dominant versus subordinate males in mouse social hierarchies Brain Behav Immun 2022 103 130 44 10.1016/j.bbi.2022.04.015 35447300 Lee W, Milewski TM, Dwortz MF, Young RL, Gaudet AD, Fonken LK, et al. Distinct immune and transcriptomic profiles in dominant versus subordinate males in mouse social hierarchies. Brain Behav Immun. 2022;103:130–44. 35447300 10.1016/j.bbi.2022.04.015

70. Cole SW, Hawkley LC, Arevalo JM, Sung CY, Rose RM, Cacioppo JT. Social regulation of gene expression in human leukocytes. Genome Biol [Internet]. 2007;8(9):R189. 10.1186/gb-2007-8-9-r189 10.1186/gb-2007-8-9-r189 PMC2375027 17854483

71. Nutt SL, Taubenheim N, Hasbold J, Corcoran LM, Hodgkin PD. The genetic network controlling plasma cell differentiation. Semin Immunol [Internet]. 2011;23(5):341–9. 10.1016/j.smim.2011.08.010 10.1016/j.smim.2011.08.010 21924923

72. Rudine AC, Sutherland MA, Hulbert L, Morrow JL, McGlone JJ. Diverse production system and social status effects on pig immunity and behavior. Livest Sci. 2007;111(1–2):86–95.

73. Mor G Cardenas I The Immune System in pregnancy: a unique complexity Am J Reprod Immunol 2010 63 6 425 33 10.1111/j.1600-0897.2010.00836.x 20367629 PMC3025805 Mor G, Cardenas I. The Immune System in pregnancy: a unique complexity. Am J Reprod Immunol. 2010;63(6):425–33. 20367629 10.1111/j.1600-0897.2010.00836.x PMC3025805

74. Dribe M. Long-term effects of childbearing on mortality: Evidence from pre-industrial Sweden. Popul Stud (NY) [Internet]. 2004;58(3):297–310. 10.1080/0032472042000272357 10.1080/0032472042000272357 15513285

75. Zeng Y Ni ZM Liu SY Gu X Huang Q Liu JA Parity and all-cause mortality in women and men: a dose-response Meta-analysis of Cohort studies Sci Rep 2016 6 1 11 26758416 10.1038/srep19351 PMC4725925 Zeng Y, Ni ZM, Liu SY, Gu X, Huang Q, Liu JA, et al. Parity and all-cause mortality in women and men: a dose-response Meta-analysis of Cohort studies. Sci Rep. 2016;6:1–11. 26758416 10.1038/srep19351 PMC4725925

76. Ryan CP Hayes MG Lee NR McDade TW Jones MJ Kobor MS Reproduction predicts shorter telomeres and epigenetic age acceleration among young adult women Sci Rep 2018 8 1 11100 10.1038/s41598-018-29486-4 30038336 PMC6056536 Ryan CP, Hayes MG, Lee NR, McDade TW, Jones MJ, Kobor MS, et al. Reproduction predicts shorter telomeres and epigenetic age acceleration among young adult women. Sci Rep. 2018;8(1):11100. 30038336 10.1038/s41598-018-29486-4 PMC6056536

77. Barrat F Lesourd B Boulouis HJ Thibault D Vincent-Naulleau S Gjata B Sex and parity modulate cytokine production during murine ageing Clin Exp Immunol 1997 109 3 562 8 10.1046/j.1365-2249.1997.4851387.x 9328137 PMC1904767 Barrat F, Lesourd B, Boulouis HJ, Thibault D, Vincent-Naulleau S, Gjata B, et al. Sex and parity modulate cytokine production during murine ageing. Clin Exp Immunol. 1997;109(3):562–8. 9328137 10.1046/j.1365-2249.1997.4851387.x PMC1904767

78. Laporta J Rosa GJM Naya H Carriquiry M Liver functional genomics in beef cows on grazing systems: novel genes and pathways revealed Physiol Genomics 2014 46 4 138 47 10.1152/physiolgenomics.00120.2013 24326346 Laporta J, Rosa GJM, Naya H, Carriquiry M. Liver functional genomics in beef cows on grazing systems: novel genes and pathways revealed. Physiol Genomics. 2014;46(4):138–47. 24326346 10.1152/physiolgenomics.00120.2013