This experiment was conducted at the Research Farm of Animal Nutrition Institute, Sichuan Agricultural University, Ya'an, China. The experimental protocol used in the present study followed the current laws regarding animal protection (Approval No. 20200156).

This study utilized 30 large white × landrace crossbred sows, which were closely matched in terms of body weight (228.16 ± 6.16 kg), backfat thickness (12.88 ± 0.26 mm), and parity (5–7). Following artificial insemination, sows were randomly divided into two groups: control (CON) or pectin (PEC). Each group contained 15 replicated pens, each of which accommodated one sow. The CON diet was formulated on a corn and soybean meal diet to provide 3.38 Mcal digestible energy/kg and 12.08 g/kg total dietary fiber according to the nutrient requirements of gestating sows of National Research Council (NRC) (20) (Table 1).

The gestational diets were meticulously crafted to maintain consistent nutrient content across all groups. For the CON group, the feeding regimen varied by gestation stage: 2.4 kg/day from 0 to 30 d, 2.3 kg/day from 30 to 90 d, and 2.7 kg/day from 90 to 110 d of gestation. In contrast, sows in the PEC group were fed a diet supplemented with 30 g/kg pectin. The scale of feeding in this group was 2.47, 2.37, and 2.78 kg/d at different stages of gestation. Sows were fed at 08:30 and 14:30 h, and water was provided ad libitum. On day 110 of pregnancy, all sows were cleaned and transferred to individual farrowing pens equipped with slatted floors and warm pads for the newborn piglets. To standardize litter sizes to 8–12 piglets per nest, cross-fostering was performed within 24 h post-farrowing, exclusively among sows from the same treatment group. Lactation feed intake began at 2 kg/day, gradually increasing by 0.5 kg/day for the first 5 days post-partum, thereafter allowing ad libitum diet access. Both sows and piglets had unrestricted access to water. All piglets were weaned on day 21 of lactation.

A commercial ELISA kit (Meimian, Jiangsu, China) was used for determining the fecal β-G activity. The feces samples were pretreated to meet the requirements of the kit. Firstly, 9 mL of phosphate-buffered saline (PBS) was added to 1 g of feces and homogenized by vortex. Then, mixtures were centrifuged at 2,500 × g at 4°C for 20 min to collect the supernatants. Finally, supernatants were measured for β-G activity with a kit.

A commercial ELISA kit (R&D Company, USA) was selected to measure the concentration of 17β-estradiol in the plasma of sows on day 110 of gestation.

Within a sterile laminar flow cabinet, swabs previously stored at −80°C were removed and immersed in 10 mL precooled PBS within 15 mL centrifuge tubes for 30 min to ensure the complete elution of the vaginal secretion on swabs. The resultant liquid was then diluted at 10⁻⁷ in PBS. From this dilution, 10 μL was decanted into 15 mL Man Rogosa Sharp (MRS) agar (Hopebiol, Qingdao, China) and subsequently incubated anaerobically at 37°C for 72 h. Anaerobic conditions were preserved using anaerobic atmosphere generation bags (Hopebiol, Qingdao, China) in Anaero Jars (Fisher, Hampton, NH, USA). After 72 h of incubation, the resulting colonies were enumerated, and the counts were expressed as the decimal logarithms of the colony-forming units per milliliter (log CFU/mL).

The microbial diversity was determined as described in an earlier research from our group (21). In brief, the bacterial genomic DNA from frozen sow stool, vaginal swabs, and colon contents of newborn piglet were extracted using the DNA kit (Omega Bio-Tek, Norcross, GA, USA). Subsequently, Majorbio Company (Shanghai, China) performed the sequencing and data analysis on the Illumina HiSeq 2000 platform. The same operational taxonomic unit (OTU) was defined as sequence similarity reaching 97%. The taxonomy of each OTU was conducted with Ribosomal Database Project (RDP) classifier program against the Silva 16S rRNA database (Release138: http://www.arb-silva.de). Categorical data was compared by calculating α-diversity and β-diversity. The Simpson and Shannon indexes were used to determine the difference of α-diversity. The principal coordinate analysis (PCoA) based on weighted Unifrac distance matrix was used to summarize the β-diversity (22, 23) Analysis of similarities (Anosim) was used to access differences among the microbial communities. All analyses from clustering to α- and β-diversity were performed in QIIME (V1.9.1) and displayed in R software (V3.3.1).

All obtained raw sequence datasets have been uploaded to the NCBI Sequence Read Archive (SRA) with the accession number PRJNA754492.

Commercial ELISA kits (MEIMIAN, Jiangsu, China) were used for determining the concentrations of plasma IL-1β, IL-6, IL-10, and TNF-α of newborn piglets following respective instructions.

A quantitative real-time polymerase chain reaction (qPCR) was used to analyze mRNA expression levels of IL-1β, IL-6, IL-8, IL-10, and TNF-α in piglet intestinal samples. The specific primer sequences used for the real-time PCR analysis are listed in Supplementary Table 1. β-Actin was utilized as an internal control, co-amplified alongside the target genes to facilitate normalization and quantification of gene expression. All samples were measured in duplicate. The relative mRNA abundance of the genes detected in the intestinal samples was calculated with the 2–^ΔΔCt method.

The data shown in the study were expressed as mean ± standard error of the mean (SEM) or mean with cluster standard errors. The activity of sows fecal β-G, sow plasma E2, and gene expression of cytokines in jejunum, ileum, cecum, and colon of piglets were analyzed by an independent-samples T-test after normality tests. Statistical significance was assessed by SPSS 22.0 (IBM SPSS Company, Armonk, NY, USA), and a value of P < 0.05 was considered statistically significant, whereas 0.05 ≤ P < 0.10 was considered a tendency.

Normality tests were performed on the microbial data at first. If the data fit the normal distribution, Student's t-test will be used for analysis. If not, statistical analyses for the relative abundance of the vaginal and fecal microbiota will be performed with Wilcoxon rank-sum test with R (Version 3.3.1) stats package and Python scipy package. Statistical significance was considered for a value of P < 0.05, whereas 0.05 ≤ P < 0.10 was considered a tendency.

The concentration of E2 in plasma of sows in the pectin group was substantially higher than that in the control group at day 110 of gestation (P < 0.05, Figure 1A).

Dietary supplementation with pectin for sows during pregnancy significantly increased the activity of β-G in feces (P < 0.05, Figure 1B).

The top four dominated phyla were Firmicutes, Proteobacteria, Bacteroidota, and Actinobacteriota in both the groups (Figure 2A). Bar plots (Figure 2B) were created to illustrate the ratio of sow vaginal microbiota at genus level at day 110 of gestation. Genus Lactobacillus was present to the extent of 5.72% in the PEC group but only 1.85% in the CON group.

PCoA was used to analyze the β-diversity between samples. No difference was observed at the phylum level, while significant differences were found at both species level and OTU level (Figures 2C, D). At 110 days of gestation, the vaginal microbiota of PEC sows were closer within the group, while the distribution was more discrete in the CON group.

At the phylum level, Verrucomicrobiota was notably increased in PEC (P < 0.05, Supplementary Table 2), while at the genus level, the relative abundance of Lactobacillus (P < 0.05) and UCG-005 (P < 0.05) in vagina microbiota were increased notably in PEC sows at day 110 of pregnancy (Table 2). At the species level, the relative abundance of Lactobacillus amylovorus was significantly higher (P < 0.01) in the PEC group than in the CON group. Meanwhile the abundance of Lactobacillus reuteri and Lactobacillus salivarius was also increased significantly (P < 0.05) in the PEC group than in the CON group (Figure 3).

The relative abundance of Myroides odoratimimus, Actinobacillus rossii, and Streptococcus Str.122 was also significantly lower in the PEC group than those in the CON group (P < 0.05, Figure 4).

Through flat colony culturing and counting with MRS agar, the concentration of vaginal Lactobacillus in the PEC group was detected significantly higher than in the CON group (P < 0.05, Figure 5).

As shown in Table 3, no difference was determined in the growth performance of suckling piglets between treatments.

The α-diversity of colon microbiota of newborn piglets was significantly affected by maternal pectin supplementation. The Simpson index in the PEC group was notably lower than that in the CON group (P < 0.05, Figure 6A). Moreover, the Shannon index was higher in the PEC group than the CON group (P < 0.1, Figure 6B).

The microbial composition in the colon of newborn piglets is obviously different at both the phylum (Figure 7A) and species (Figure 7B) level. Beta-diversity, evaluated by PCoA based on weighted Unifrac distance, revealed that the colon microbiota of newborn piglets formed a well-defined cluster and an obvious separation was seen at the phylum level (P = 0.050, Figure 7C). But the result at the species level displayed no differences (P > 0.1, Figure 7D).

At the phylum level, the relative abundance of Proteobacteria in piglets' colon was significantly lower in the PEC group, while the relative abundance of Firmicutes tended to be higher than in the CON group (Table 4).

The concentration of IL-6 in the plasma of newborn piglets was notably decreased in the PEC group than in the CON group (P < 0.05, Figure 8).

The effects of gestational dietary pectin supplementation on the relative mRNA expression of cytokine in the intestine of piglets are shown in Figure 9. In jejunum, no difference was detected (Figure 9A). But in ileum, the relative mRNA expression of TNF-α was found to be significantly higher in the CON group than in the PEC group (P < 0.05; Figure 9B). The relative IL-10 mRNA expression in the cecum was notably higher in the PEC group than in the CON group (P < 0.01). IL-1β was found to be expressed lower in the PEC group compared to the CON group (P < 0.05; Figure 9C). The relative expression of TNF-α in the colon (Figure 9D) was significantly lower (P < 0.05) in the PEC group than in the CON group, while in the CON group, the expression of IL-10 was higher (P < 0.1).

The results of morphometric measurements in the duodenum, jejunum, and ileum in neonatal piglets are shown in Table 5. The villus height and V/C ratio of duodenum, ileum, and jejunum in the PEC group was extremely significantly higher than that in the CON group (P < 0.001). The crypt depth of ileum in the CON group was higher than in the PEC group (P < 0.1). Meanwhile, the muscular layer thickness of duodenum in the PEC group was markedly higher than that in the CON group (P < 0.05).

The estrobolome has been defined as the gene repertoire of the microbiota in the gut that are capable of regulating estrogens. The conjugated estrogens existing in the gut can be deconjugated by bacterial secreted β-G into estrogen, which will be reabsorbed through the epithelial cells to the blood stream (24, 25), and distributed to different distant organs like the epithelium in the vagina (26). Microorganisms like Faecalibacterium prausnitzii, Ruminococcus, Clostridium, and so on demonstrated their ability to produce β-G (27–30). The observed increase in fecal β-G activity within the pectin group corroborates our hypothesis. This finding aligns with Licht et al. (19), who demonstrated that incorporating 7% apple pectin into rat's diet for 14 weeks substantially elevates β-G activity in rat feces. Similarly, other researchers (18) contended that in the presence of a pectin–protein complex isolated from cabbage, the activity of β-G was significantly enhanced in vitro. Furthermore, the addition of wheat flour aleurone, a fermentable dietary fiber sourced from the outer layer of wheat grains, has been reported to significantly boost β-G activity in the cecum of rats (12).

In our study, we observed a significant increase in the concentration of estradiol (E2) in plasma by day 110 of gestation with the supplementation of pectin to sows during gestation. The elevation in the E2 levels may be influenced by the activity of β-G, which was modified by alterations in the gut microbiota as mentioned. This is in line with the findings in postmenopausal women, who, despite their ovaries not producing estrogen, exhibit higher E2 concentrations in their blood when their diets are rich in soluble fiber (31). Conversely, some studies have demonstrated that increasing composition of fiber in diets can reduce the level of free E2 in the bloodstream (32–34). This theory is plausible as fiber intake can increase stool bulk, facilitating the premature excretion of conjugated estrogen metabolites that originate from enterohepatic circulation. For example, Jiang (35) reported that crude fiber intake can remarkably increase the concentration of free E2 in feces, but not in saliva, which indicates that fiber intake might facilitate the deconjugation of E2 in the hindgut, but estrogen was excreted in the expanded feces as a result of insoluble fiber before being reabsorbed into the bloodstream.

During pregnancy, E2 can not only maintain the pregnancy status but also improve the vaginal microbiota to be dominated by Lactobacillus. Elevated levels of E2 promote the maturation and deposition of glycogen in vaginal epithelial cells. Glycogen from exfoliated and lysed epithelial cells is catabolized by α-amylase in the vaginal lumen to smaller polymers that are subsequently metabolized to lactic acid by Lactobacillus. Lactic acid acidifies the vaginal milieu, favoring the proliferation of Lactobacillus and inhibiting the growth of infection-associated organisms (36–40). Thus, some studies aimed to improve vaginal microbes during pregnancy through nutrition methods. Yang (41) tried to improve the vaginal microbiota of pregnant women by giving them Lactobacillus rhamnosus GR-1, and Lactobacillus reuteri RC-14 orally, but they were found to be ineffective in increasing microbes. However, supplying Lactobacillus rhamnosus GR-1 and Lactobacillus reuteri RC-14 in capsules to pregnant women was suggested to significantly reduce vaginal Streptococcus agalactiae colonization (42). But almost all of the above studies were only providing pregnant women with probiotics directly and did not look into the effects of non-microbial prebiotics on vaginal microbiota, while the present study indicates that it is possible to increase the relative abundance of Lactobacillus in the vagina of sows with pectin-supplied diet.

Two species of bacteria were also found to be reduced remarkably in vagina in the PEC group. Myroides odoratimimus is associated with urinary tract inflammation in humans (43, 44), which indicated that supplying sows with pectin during gestation may decrease the risk of urinary tract inflammation. Mayor et al. (45) certified that Actinobacillus rossii was isolated from aborted piglets, uterus, and vaginal discharge since 1989 in USA. Holyoake and Thompson (46) isolated A. rossii from an aborted piglet as well, which means these abortions are probably caused by A. rossii, or there might be some relationship between abortion and A. rossii. Therefore, we further explored the intestinal health of the new piglet generation.

Providing pectin-containing diet to the sows during pregnancy has a significant impact on the colonic microbial diversity and their composition in the neonatal piglets. Compared with the CON group, the colonic flora of piglets in PEC has a higher Shannon index and a lower Simpson index, which means a higher microbial diversity. The β-diversity of the colonic microbiota of piglets from each group was found with a significant discrimination through PCoA. The relative abundance of phylum Proteobacteria and species Escherichia–Shigella, both generally regarded as harmful bacteria, was lower in the PEC group than in the CON group. In addition, studies have shown that part of mother's vaginal microbiota will be colonized in the intestines of the offspring through vaginal delivery (47). And in the sows and piglets, the mode of delivery and early nutrition modulate microbial colonization (48). This suggests that if the mother has vaginal microbiota dominated by probiotics before giving birth, the microbiota of newborn intestine may be colonized with probiotics and may provide a better condition to the offspring. Surprisingly, Streptococcus_suis enriched in the CON sows' vagina was found to be notably higher in the colon of the CON offspring, but Lactobacillus, which also were notably enriched in the vagina of sows in the PEC group did not show any difference in the colon of piglets from both groups. This may indicate that the transmission of microorganisms between sows and piglets does not entirely depend on the contact with vaginal mucus during vaginal delivery. Recently, sequencing has indicated that placenta and umbilical cord blood is not sterile, and the microbiota within the neonate's meconium shares significant similarity with that of the placenta, thus suggesting that maternal transfer of microbiota is possible and might occur during gestation (49–51). Therefore, the maternal microbiota plays a vital role in the composition of the intestinal flora of the offspring.

Maternal nutrition is known to impact the development of the intestine (52), with the morphology of the intestine serving as a key indicator of its health. Villus height, in particular, plays a crucial role in determining the nutrient absorption capacity of the small intestine. The ratio of villus height to crypt depth (V/C) is regarded as a reliable measure for assessing the nutrient absorption potential of the small intestine. In this study, offspring of mothers supplemented with pectin during pregnancy exhibited increased villus height and V/C ratio, suggesting an enhanced ability to absorb nutrients. Similarly, existing research indicates that maternal diets high in soluble fiber during gestation can alter the intestinal microbiota, enhance growth performance, and reduce intestinal permeability in piglets (53).

Accumulating evidence indicates that maternal inflammation also has long-term consequences for offspring by affecting the intrauterine environment (54). In the present study, we found that piglets from the PEC group exhibited lower plasma IL-6 levels. IL-6 is a pleiotropic cytokine participating in the physiology of virtually every organ system (55). We can not only find the change of inflammatory cytokines in the piglet plasma but also in their intestine. Although no differences were found in the jejunum, they gradually emerged from the ileum. The improvement of intestinal immune function of newborn piglets is thought to be mediated by microorganisms (56, 57). A study showed a similar result of a maternal soluble fiber diet increasing the plasma concentrations of anti-inflammatory factors, IL-10, in the offspring, and both sows and piglets were found to have lower plasma TNF-α levels with soluble fiber intake (21).