Maternal supplementation with a casein hydrolysate and yeast beta-glucan from late gestation through lactation improves gastrointestinal health of piglets at weaning

Improving maternal nutrition during pregnancy/lactation is a promising strategy to maximise the intestinal health of piglets undergoing abrupt weaning under commercial production conditions. This experiment investigated the effects of maternal supplementation of a casein hydrolysate and yeast β-glucan (CH-YBG) from day 83 of gestation until weaning (day 28) on sow faecal microbial populations and measures of piglet gastrointestinal health parameters at weaning. Sows (n = 10 sows/group) were assigned to: (1) control diet, and (2) control diet + CH-YBG. Maternal supplementation increased the abundance of the phylum Firmicutes, including members Lactobacillus in the sows faeces, with a concomitant increase in the caecal abundance of Lactobacillus in the weaned piglets compared to the controls. Piglets weaned from the supplemented sows had increased villus height in the duodenum (P < 0.05) and increased villus height to crypt depth ratio in the jejunum, as well as a decreased expression of the proinflammatory cytokine genes (IL6/TNF/TGFB), the tight junction gene CLDN3 and the mucin gene MUC2 in the duodenum/jejunum compared to the controls (P < 0.05). In conclusion, maternal CH-YBG supplementation during pregnancy/lactation improved microbial, structural, and inflammatory measures of gastrointestinal health of piglets at weaning. This is a promising strategy to alleviate the challenges that occur with early abrupt weaning in commercial pig production.

Commercial pigs are subjected to significant levels of stress at weaning. This causes an initial reduction in feed intake, contributing to adverse gut morphological and functional changes such as villous atrophy and crypt hyperplasia. This in turn, increases epithelial permeability and inflammation 1,2 . Alongside an immature immune system, these factors favour the proliferation of pathogenic bacteria, including, enterotoxigenic Escherichia coli, thus increasing overall susceptibility to infection and post-weaning diarrhoea 3 . Supplementation of dietary zinc oxide (ZnO) at pharmacological doses improves growth and controls the proliferation of pathogenic bacteria post-weaning. However, the EU have begun the phasing out of pharmacological doses of ZnO (Commission Implementing Decision of 26.6.2017, C (2017) 4, 529 Final) alongside further restrictions on the use of infeed medication (Regulation (EU) 2019/6) in weaner pig diets by 2022 due to their association with antimicrobial resistance.
Recent research regarding the enhancement of performance and gut health, in pigs, has focussed on postweaning dietary manipulation. However, a more effective alternative may be supplementing the diet of the gestating and/or lactating sow to enhance the pre-weaning health status of the piglet as this may have a positive influence on post-weaning performance 4,5 . The gastrointestinal tract (GIT) of the piglet is colonized immediately Management of piglets. All farrowings were supervised. Between 6 and 12 h after the birth of the last piglet, litter size was adjusted by cross-fostering piglets within sow diets to ensure that sows nursed a similar number of piglets (n = 12), and this was maintained throughout the suckling period until weaning. The piglets body weight (BW) was recorded at birth and weaning, and the average daily gain (ADG) calculated. The piglets received an intramuscular injection of Fe-dextran (Ferdex 100; Medion Farma Jaya) on day 7 after birth. No creep feed was offered to the piglets throughout the lactation period, and piglets did not have access to the sows' feed.
Collection of piglet tissue and digesta samples at weaning. On the day of weaning (day 28), one male piglet per sow of average birth weight (which was identified at parturition) was euthanized after a lethal injection with pentobarbitone sodium, Euthatal solution, 200 mg/ml, (Merial Animal Health, Essex, UK) at a rate of 0.71 ml/kg BW to the cranial vena cava to humanely euthanize the animal. Euthanasia was performed by a competent person in a separate room away from sight and sound of the other pigs. The entire intestinal tract was immediately removed. Sections from the duodenum (located 10 cm distal from the stomach), the jejunum (60 cm from the stomach) and the ileum (15 cm from the caecum) were excised and fixed in 10% neutralbuffered formalin. Digesta from the caecum and colon was collected and stored in sterile containers (Sarstedt, Wexford, Ireland) and immediately frozen at − 20 °C for subsequent 16 s rRNA sequencing. In addition, tissue samples were taken from the duodenum, jejunum and ileum for the purpose of measuring the expression of cytokines, digestive enzymes, nutrient transporters, mucins and tight junctions using QPCR. Tissue sections (1 cm) from the duodenum, jejunum, and ileum were cut out, dissected along the mesentery, emptied, and rinsed using sterile phosphate buffered saline (Oxoid, Hampshire, UK). The tissue sections were stripped of  Table 2. Dissociation curves were generated to confirm the specificity of the resulting PCR products. The QPCR assay efficiencies were established by plotting the cycling threshold (CT) values derived from fourfold serial dilutions of cDNA against their arbitrary quantities and only assays exhibiting 90-110% efficiency and single products were used in this study. Normalised relative quantities were obtained using the qbase PLUS software (Biogazelle, Ghent, Belgium) from stable reference genes; B2M, ACTB and PPIA. These genes were selected as reference genes based on their M value (< 1.5) generated by the GeNorm algorithm within GeNorm. The genes analysed in the current study are as follows: SLC15A1 (previously known as PEPT1); FABP2; SLC2A1 (previously known as GLUT1); SLC5A1 (previously known as  Illumina sequencing. Extraction of the bacterial DNA from the caecal digesta samples and high-throughput sequencing of the V3-V4 hypervariable region of the bacterial 16S rRNA gene was performed on an Illumina MiSeq platform according to their standard protocols (Eurofins Genomics, Ebersberg, Germany). Briefly, the V3-V4 region was PCR-amplified using universal primers containing adapter overhang nucleotide sequences for forward and reverse index primers. Amplicons were purified using AMPure XP beads (Beckman Coulter, Indianapolis, IN) and set up for the index PCR with Nextera XT index primers (Illumina, San Diego, CA). The indexed samples were purified using AMPure XP beads, quantified using a fragment analyser (Agilent, Santa Clara, CA), and equal quantities from each sample were pooled. The resulting pooled library was quantified using the Bioanalyzer 7500 DNA kit (Agilent, Santa Clara, CA) and sequenced using the v3 chemistry (2 × 300 bp paired-end reads).
Bioinformatic. The bioinformatic analysis of the resulting sequences was performed by Eurofins Genomics (Ebersberg, Germany) using the open-source software package (version 1.9.1) Quantitative Insights into Microbial Ecology (QIIME) 19 . All raw reads passing the standard Illumina chastity filter were demultiplexed according to their index sequences (read quality score > 30). The primer sequences were clipped from the starts of the raw forward and reverse reads. If primer sequences were not perfectly matched, read pairs were removed to retain only high-quality reads. Paired-end reads were then merged if possible, to obtain a single, longer read that covers the full target region using the software FLASH 2.2.00 20 . Pairs were merged with a minimum overlap size of 10 bp to reduce false-positive merges. The forward read was only retained for the subsequent analysis steps when merging was not possible. Merged reads were quality filtered according to the expected length and known length variations of the V3-V4 region (ca. 445 bp). The ends of retained forward reads were clipped to a total read length of 285 bp to remove low quality bases. Merged and retained reads containing ambiguous bases were discarded. The filtered reads (merged and quality clipped retained forward reads) were used for the microbiome profiling. Chimeric reads were identified and removed based on the de-novo algorithm of UCHIME 21 as implemented in the VSEARCH package 22 . The remaining set of high-quality reads were processed using minimum entropy decomposition (MED) to partition reads to operational taxonomic units (OTU) 23,24 . DC-MEGABLAST alignments of cluster representative sequences to the NCBI nucleotide sequence database were performed for taxonomic assignment (from phylum to species) of each OTU. A sequence identity of 70% across at least 80% of the representative sequence was the minimal requirement for considering reference sequences. Abundances of bacterial taxonomic units were normalized using lineage-specific copy numbers of the relevant marker genes to improve estimates 25 . The normalized OTU table combined with the phenotype metadata and phylogenetic tree comprised the data matrix. This matrix was then input into the phyloseq package within R (http:// www.r-proje ct. org; version 3.5.0). The dynamics of richness and diversity in the microbiota were computed with the observed, Chao1, ACE, Shannon, Simpson, InvSimpson and Fisher indices. The Simpson and Shannon indices of diversity account for both richness and evenness parameters. The beta diversity measurements are a measure of separation of the phylogenetic structure of the OTU in one sample compared with all other samples. This was estimated by normalising the data so taxonomic feature counts were comparable across samples. Several distance metrics were considered, to calculate the distance matrix of the different multidimensional reduction methods. These included weighted/unweighted UniFrac distance and non-phylogenetic distance metrics (i.e., Bray-Curtis, Jensen-Shannon divergence and Euclidian) using phyloseq in R 26,27 . Differential abundance testing was performed on tables extracted from the phyloseq object at phylum, family, and genus level. The data was analysed using the PROC Glimmix procedure within Statistical Analysis Software (SAS) 9.4 (SAS Institute, Cary, NC, USA). The model assessed the effect of 'treatment' , with the piglet (sow pen) as the experimental unit. 10 sows and 10 piglets per treatment group were used for the statistical analysis of the relative bacterial abundances. Results are presented using Benjamini-Hochberg (BH) adjusted P-values.
Feed analysis. The feed samples were milled through a 1 mm screen (Christy and Norris Hammer Mill, Chelmsford, England) and retained for chemical analysis. The gross energy (GE) content was determined using an adiabatic bomb calorimeter (Parr Instruments, Moline, IL USA) as previously described 28 . The dry matter (DM) content of the feed was determined after drying overnight at 104 °C. Feed samples were analysed for crude ash (AOAC method 942.05 29 ) and nitrogen (N × 6.25; AOAC method 990.03 29 ). All samples were measured in duplicate.
Statistical analysis. The data on performance, intestinal morphology and gene expression were initially checked for normality using the univariate procedure of Statistical Analysis Software 30 . The data on performance, intestinal morphology, and gene expression were analysed using the GLM procedure within SAS. The model assessed the effect of treatment, with the sow as the experimental unit. For piglet diarrhoea score, the data was analysed by repeated-measures analysis using the PROC MIXED procedure of SAS. All data presented in the tables are expressed as least-square means with their standard errors. The probability level that denoted significance was P < 0.05. Data are presented as least-square means and standard error of the mean.

Results
Gestation length and suckling piglet growth performance. The effect of maternal supplementation on gestation length and piglet body weight and ADG is presented in Table 3. There was no effect of maternal dietary supplementation on gestation length (115. The effects of maternal dietary supplementation on the sow faecal and piglet caecal and colonic microbiota. Bacterial richness and diversity. The effects of maternal supplementation on the measures of beta diversity and alpha diversity are presented in Fig. 1 and Table 4, respectively. Beta diversity in the piglets caecal and colonic microbiome differed to that of the sow (P < 0.05) based on Permanova analysis and through visualisation using the Bray Curtis distance matrix and multi-dimensional scaling. Maternal dietary supplementation had no effect on the Observed, Chao1, ACE, Shannon, Simpson, InvSimpson and Fisher index measures of alpha diversity (P > 0.05) in either the sows faeces or the piglets caecal and colonic digesta.
Differential bacterial abundance analysis. All data on bacterial abundances at phylum, family and genus level are provided in Tables 5 and 6.
Family. At family level, dietary supplementation increased the relative abundance of Lactobacillaceae (P < 0.0001) and Lachnospiraceae (P < 0.05) compared with the control group. Dietary supplementation decreased the relative abundance of Christensenellaceae (P < 0.05), Planococcaceae (P < 0.05) and Moraxellaceae (P < 0.0001) compared to the control.
Piglets caecal and colonic microbiota. Phylum. In the caecum and colon, there were six bacterial phyla identified with Firmicutes being the dominant phyla followed by Actinobacteria, Proteobacteria, Verrucomicrobia, Fusobacteria, and Bacteroidetes. The relative abundance of Actinobacteria was decreased (P < 0.05) in the caecum and colon of piglets from supplemented sows compared to piglets from control sows.
Family. In the caecum, the relative abundance of Lactobacillacea and Oscillospiraceae were increased (P < 0.05), while the relative abundance of Coriobacteriaceae, Ruminococcaceae and Akkermansiaceae decreased (P < 0.05) in piglets from supplemented sows compared to piglets from control sows. In the caecum and colon, the relative abundance of Erysipelotrichaceae and Christensenellaceae were increased (P < 0.05), while the relative abundance of Atopobiaceae was decreased (P < 0.05) in piglets from supplemented sows compared to piglets from control sows. www.nature.com/scientificreports/ Gene expression in the small intestine. The intestinal expression of genes related to nutrient digestion and absorption, mucosal barrier function, and immunity in the duodenum, jejunum and ileum of piglets are presented in Table 7.
Expression of genes involved in inflammation and the epithelial barrier. The expression of IL6 (interleukin-6, P < 0.05), TNF (tumor necrosis factor, P < 0.05) and TGFB (transforming growth factor beta-1, P < 0.05) were down-regulated in the duodenum and jejunum of piglets suckling the supplemented sows compared with those suckling the control sows. The expression of CLDN3 (claudin-3, P < 0.01) was up-regulated in the duodenum and jejunum of piglets suckling the supplemented sows compared with those suckling the control sows. The expression of IL10 (interleukin-10, P < 0.01) was down-regulated in the jejunum of piglets suckling the supple- www.nature.com/scientificreports/ mented sows compared with those suckling the control sows. The expression of CXCL8 (interleukin 8, P < 0.01) and CLDN3 (P < 0.0001) were upregulated in the jejunum of piglets suckling the supplemented sows compared with those suckling the control sows.
Ileum. The expression of IFN (interferon gamma, P < 0.05) was downregulated in the ileum of piglets suckling the supplemented sows compared with those suckling the control sows.
Piglet small-intestinal morphology. The effect of maternal supplementation on piglet small intestinal morphology is presented in Table 8.
In the duodenum, piglets suckling the supplemented sows had larger VH (P < 0.05) compared with those suckling the control sows. In the ileum, piglets suckling supplemented sows had a larger VH:CD ratio (P < 0.05) compared with those suckling the control sows.

Discussion
In the present study, it was hypothesised that maternal supplementation with a casein hydrolysate and yeast β-glucan (CH-YBG) from day 83 of gestation would alter the GIT microbiota of the sow and modulate selected intestinal microbial populations, the inflammatory response, and aspects of intestinal health of her progeny at weaning. Maternal supplementation with CH-YBG altered the sow's faecal microbiota by increasing the relative abundance of bacterial members of the Firmicutes phylum, including Lactobacillus. The positive response observed in piglets suckling the CH-YBG supplemented sows, such as increased abundance of numerous Firmicutes, including Lactobacillus and Christensenella in the caecal microbiota, improved intestinal morphology, decreased expression of proinflammatory cytokine genes, and increased expression of nutrient transporter, tight junction, and mucin genes at the time of weaning supports this hypothesis.
The early colonization of the piglets intestinal microbiota plays an essential role in the development of the piglets intestinal immune system, barrier function and nutrient absorption 31,32 . The major microbial changes in this experiment were observed within the Firmicutes, with supplemented sows having a higher abundance of beneficial bacteria such as Lactobacillus compared to the control sows. The abundance of Lactobacillus, Christensenella and Erysipelotrichaceae was also increased in the progeny of supplemented sows at 28 days of age. These genera are positively correlated with a healthy gut, with low numbers associated with inflammatory disorders such as inflammatory bowel disease 33,34 and Crohns disease 35,36 . The increase in Lactobacillus in the sow is likely to be attributable to the casein hydrolysate's role as a substrate for the growth of Lactobacilli, as casein hydrolysate supplementation increased the growth of Lactobacillus in rats and pigs 37,38 . Both Lactobacillus and Christensenella species have strong anti-inflammatory properties and play an important role in maintaining microbial symbiosis, thereby contributing to the establishment of an "optimal" microbiota in piglets at weaning. These results indicate that maternal supplementation positively influenced the microbial composition in the faeces of the sow and in the large intestine of her progeny and thus, may play a role in improving the health of piglets at weaning.
The small intestine plays an essential role in the digestion and absorption of nutrients and serves as a key regulator of growth. Under conventional practice, abrupt weaning is often associated with undesirable morphological Table 4. The effect of dietary treatment on measures of alpha diversity (least-square means with their standard errors). Supplemented = casein hydrolysate + yeast β-glucan supplementation, and control = basal lactation diet. a Ten sows per treatment group. b 10 piglets per treatment group. www.nature.com/scientificreports/ changes, such as villous atrophy and crypt hyperplasia, leading to functional changes such as a decrease in nutrient digestion and absorption 2,39 . Thus, supplements that positively influence intestinal health should aid in the amelioration of the intestinal dysfunction associated with weaning stress. Supplementation with bacterial members of the genus Lactobacillus has been shown to improve intestinal health in broilers through the improvement of small intestinal villous height and crypt depth 40,41 . In the present study, pigs weaned from supplemented sows had improved VH in the duodenum and VH to CD ratio in the ileum. The improvement in intestinal health in response to maternal supplementation, consequently enhanced the digestive and absorptive capacity of the intestinal epithelium resulting in an upregulation of nutrient transporters in the small intestine. Maternal dietary supplementation with CH-YBG increased the expression of fatty acid transporter FABP2, peptide transporter SLC15A1, glucose transporters SLC2A1, SLC2A2, SLC2A5, and SGLT1, amino acid transporter SLC1A4, and sodium transporter SLC5A8 and had a tendency to increase solute carrier SLC16A. This suggests there is increased availability and transport of fatty acids, sugars, sodium, peptides and amino acids in the duodenum and jejunum of pigs suckling supplemented sows. The upregulation of these nutrient transporters indicates a positive functional adaptation in the pigs, and potentially improving their ability to digest a diet primarily comprised of carbohydrates and protein post-weaning.
In the present study, maternal supplementation increased the expression of the tight junction gene, CLDN3 and the mucin gene MUC2. Claudin 3 is a tight junction integral protein that decreases paracellular permeability, thus preventing the translocation of luminal antigens 42 . Mucins provide a barrier against potential pathogens while also modulating the expression of inflammatory cytokines 43 . It is possible that the increased expression of CLDN3 reduced the translocation of bacteria across the intestinal barrier, while the increased expression of MUC2 reduced the adherence of bacteria to the intestinal cells, thereby reducing the risk of pigs to infection and diarrhoea. The increase in genes associated with intestinal barrier function was concomitant with a decrease in the expression of pro-inflammatory cytokine genes. Overall, these effects indicate a better host immune response with regard to potential environmental or pathogenic challenge at weaning time. Maternal supplementation from day 83 of gestation and throughout the lactation period decreased expression of the pro-inflammatory cytokine genes IL6, TGFB, TNF and IFNG. IL-6 and TNF are potent mediators of inflammation and play an important Table 5. The effect of maternal dietary treatment on the % bacterial abundance at phylum and family level (Least-square means with their standard errors). * Ten sows per treatment group. ¥ 10 piglets per treatment group.  44,45 . It is likely that the increase in the expression of the tight junction genes strengthened the intestinal barrier, thus limiting the ability of antigens to enter the body and prevent adverse immune responses. When the immune system is activated, and inflammation occurs, maintenance energy and amino acid requirements are increased, thereby reducing the proportion of daily nutrient intake directed towards growth 46 . Thus, the potential decrease in immune system activity in maternally supplemented pigs may result in an improvement in growth performance during the post-weaning period.
While it was not within the scope of this study to identify what effects each bioactive had on the gastrointestinal health parameters measured, it is worthy to note that the 5 kDa retentate (5kDaR) of a CH has established microbial activity and anti-inflammatory effects in vitro 9,47,48 . Furthermore, in a previous study the combination of the 5kDaR + YBG positively enhanced gastrointestinal parameters in pigs post-weaning, whereas the individual inclusion of either 5kDaR or YBG were not effective 10 . It is evident in the present study that maternal supplementation with a CH-YBG improved several gastrointestinal health parameters that influence performance. Thus, it may be anticipated that this would translate into an improvement in growth performance at weaning, however piglet weights at weaning were not influenced by sow diet. A similar effect was observed following maternal supplementation with seaweed-derived polysaccharides, containing laminarin and fucoidan 49 . Furthermore, maternal supplementation with Bacillus altitudinis 50 did not influence the growth performance of piglets at weaning, however maternal supplementation had an impact on the progeny's performance post-weaning. A further study to explore the effects of maternal supplementation with a CH-YBG on piglet growth performance post-weaning is warranted. Table 6. The effect of maternal dietary treatment on the % bacterial abundance at genus level (Least-square means with their standard errors). Supplemented = Casein hydrolysate + yeast β-glucan supplementation, and control = basal lactation diet. a Ten sows per treatment group. b 10 piglets per treatment group.

Conclusion
Maternal dietary supplementation with a combination of a casein hydrolysate + yeast β-glucan from day 83 of gestation to weaning had a positive effect on the microbial population of sows and their piglets, represented by a higher abundance of several bacterial members of the phylum Firmicutes, including Lactobacillus and Christensenella. Maternal supplementation enhanced piglet gastrointestinal health by improving intestinal morphology, reducing the expression of proinflammatory cytokine genes, and increasing the expression of nutrient transporter, tight junction, and mucin genes at weaning. These findings indicate that maternal casein hydrolysate + yeast β-glucan supplementation may be a promising strategy for improving the gastrointestinal health and function of piglets at weaning, making piglets more resilient to post-weaning challenges. Table 7. The effects of maternal dietary supplementation on the expression of nutrient transporter, immune marker and tight junction genes in piglet's duodenum, jejunum and ileum (least square means with their standard errors). SLC15A1/PEPT1, peptide transporter 1; FABP2, fatty acid binding protein 2; SLC2A1/ GLUT1, glucose transporter 1; SLC2A2/GLUT2, glucose transporter 2; SLC2A5/GLUT5, glucose transporter 5; SLC5A1/SGLT1, sodium glucose linked transporter 1; SLC5A8, solute carrier family 5 member 8; SLC16A, solute carrier family 16 member 1; SLC1A4, solute carrier family 1 member 4; TNF, tumor necrosis factor alpha; CXCL8, interleukin 8; IL6, interleukin 6; IL10, interleukin 10; IFNG, interferon gamma; MUC2, mucin-2; TGF-β, transforming growth factor beta; CLDN3, claudin 3; *A total of 10 replicates were used per treatment.