The animal experiment was approved by the Animal Experiment Committee of Nanjing Agricultural University (Nanjing, Jiangsu province, China), in accordance with the Regulations for the Administrations of Affairs Concerning the Experimental Animals (The State Science and Technology Commission of China, 1988). All experiments were performed in accordance with the approved guidelines and regulations.

Ten Yorkshire×Landrace sows were artificially inseminated by one Duroc boar. The sows had delivered on the same day. The piglets suckled the colostrum for three days after birth. On the 4th day after birth, the piglets were adjusted to 8 per litter; the average weight of piglets was 1.87±0.05 kg. Ten litters of piglets were randomly divided into the control group (sow milk, SM) and the treatment group (CM), with 5 litters in each group (n = 5). The piglets in the SM group were reared only by the sow, while the piglets in the CM group were supplemented with a CM supplement along with suckling from d 4 to d 28 of age. Piglets in the CM group had ad libitum access to milk supplement via a nipple system. The CM supplement was dissolved in 50°C warm water (150 g/L) and was provided via a nipple system twice a day. At 9:00 and 19:00 every day during the experimental period, we supplemented the milk supplement with a weight that is same amount of average daily feed intake (ADFI) in the nipple system, and then weighed the rest milk supplement in the nipple system when we added milk supplement next time. Then, we calculated the intake of CM supplement every day. On d 28, one piglet was randomly picked from each litter, and then slaughtered for sampling (n = 5). The remaining piglets were weaned, and the piglets in the same litter were housed in a pen, therefore each group had five pens (n = 5). All the piglets were switched to a common diet for 7 days. On d 35, one piglet was randomly picked from each litter, and then slaughtered for samples (n = 5). The piglets had free access to feed and water throughout the experimental period. The body weight (BW) of each piglet in the SM group or CM group was individually recorded on d 4, 14, 21, 28, and 35. Then, we utilized those data to calculate average daily gain (ADG) of the SM group and CM groups. The nutritional values of the milk supplement and SM are shown in Table 1 [10–13].

At days 28 and 35, after fasting overnight, one piglet from each litter (5 piglets from each group, n = 5) was euthanised with sodium pentobarbital (100 mg/kg BW). The abdomen was opened and the segments (stomach, duodenum, jejunum, ileum, cecum, and colon) were identified and ligated before separation. The digesta in the jejunum was collected and mixed for measuring the pH value. Then, the digesta was immediately stored in liquid nitrogen for further analysis. Five cm jejunum tissue were fixed in 4% paraformaldehyde (100 mmol/L phosphate buffer, pH 7.4) for morphology analysis. The mucosa was isolated by gentle scraping with a glass slide, snap frozen in liquid nitrogen, and then stored in −80°C for future disaccharidase activities and gene expression analysis.

After fixing for 24 h, formalin-fixed samples were embedded in paraffin, sectioned at a 5 μm thickness, and stained with hematoxylin and eosin for histologic examination to determine villus length, crypt depth. The villus height and crypt depth were measured and analyzed by using NIS-Elements BR software (version 2.20; Nikon, Tokyo, Japan).

Each sample was homogenized in 1.0 mL cold saline mixed with protease inhibitors (Roche, Shanghai, China) on ice. Protein concentration was determined using a BCA Protein Assay Kit (Beyotime, Shanghai, China). The activity of the disaccharidase (maltase, lactase, and sucrase) was determined using commercial kits (Jiancheng Bioengineering Institute, Nanjing, China). Enzyme activity was described as U per mg of protein.

The concentration of lactic acid was determined using commercial kits (Jiancheng Bioengineering Institute, China). The ammonia N concentration was determined by the colorimetric method described by Nyachoti et al [14]. The concentrations of SCFAs including acetate, propionate, butyrate, valerate and branched chain fatty acids were analyzed by gas chromatography (Shimadzu GC-14B, Tokyo, Japan) [15]. A Nukol Capillary GC Column (Sigma-Aldrich, Milwaukee, WI, USA) and an FID detector were used as described by Mao et al [16]. The temperature of injector, column and detector were 110°C, 150°C, and 180°C, respectively.

The total bacterial DNA was extracted from the jejunal digesta (0.1 g) according to a bead beating method described by Zoetendal et al [17]. The concentration of extracted DNA was determined using a Nano-Drop 1000 spectrophotometer (Thermo Scientific Inc., Wilmington, DE, USA). To analyze the taxonomic composition of bacterial community in the jejunal digesta, universal primers 319F (ACTCCTACGGGA GGCAGCAG) and 806R (GGACTACHVGGGTWTCTAAT) targeting the V3–V4 region of the 16S rRNA gene were chosen for the amplification and subsequent pyrosequencing of the polymerase chain reaction (PCR) products. The PCR amplification was performed using the following program: initial denaturation step at 95°C for 2 min, followed by 25 cycles at 95°C for 30 s, 55°C for 30 s, and 72°C for 30 s and a final extension at 72°C for 5 min. All amplicons were visualized on 2% agarose gels and were purified using the Axyprep DNA Gel Extraction Kit (Axygen Biosciences, Union City, CA, USA) following the recommended instructions and were quantified using Quantifluor TM-ST (Promega, Madison, WI, USA). The purified amplicons were paired-end sequenced on an Illumia Miseq platform according to a standard protocol [18].

Raw fastq files were de-multiplexed and quality-filtered using QIIME (version 1.70, CU-Boulder, CO, USA) with standard criteria as described by Sun et al [19]. Operational taxonomic units (OTUs) were clustered with a 97% similarity cut-off using UPARSE (version 7.1 http://drive5.com/uparse/), and the chimeric sequences were identified and removed using UCHIME. The bacterial diversity was assessed using the observed species, Simpson, Chao1 and Shannon indices.

The raw sequencing data were submitted to the Sequence Read Archive under accession number SRP107026. The data were analyzed with SPSS 20.0 and shown as the means with standard error of the mean. The student’s t-test and Mann-Whitney U test were used to assess the differences between treatments, with the litter as the experimental unit for growth performance, and the piglet as the experimental unit for other indices (n = 5). The normality of the distribution of variables was tested by the Shapiro-Wilk test. The variables that had a non-normal distribution were analyzed using the nonparametric methods. The t-test and the Mann-Whitney U test were used to analyze the data that had a normal (the data of jejunal morphology, disaccharidase activities and the gene expression) or non-normal distribution (microbial data) on d 28 and d 35, respectively. Significant differences were declared when p<0.05, while p values between 0.05 and 0.10 were considered to be a tendency.

The villus height, crypt depth and villus height to crypt depth ratio are shown in Table 2. On d 28 and d 35 of age, no significant difference in villus height, crypt depth and villus height to crypt depth ratio was observed between the CM group and the SM group.

The disaccharidase activities in jejunal mucosa of piglets are presented in Figure 1. On d 28 of age (Figure 1A), the activity of lactase in the jejunum was significantly decreased in the CM group, while the activity of sucrase and maltase to lactase ratio were significantly increased (p<0.05). On d 35 of age, the activity of maltase in the jejunum was significantly increased in the CM group (p<0.05), and maltase to lactase ratio tended to increase in the CM group (p = 0.065, Figure 1B).

The bacterial metabolites in jejunal digesta are presented in Table 3 and 4. On d 28 (Table 3), the concentrations of acetate and the total SCFAs were significantly increased in the jejunal digesta in the CM group (p<0.05). Piglets in the CM group had increased lactic acid concentration (p = 0.07). On d 35 of age (Table 4), the pH value significantly decreased in the jejunal digesta, and the concentration of lactic acid in the jejunal digesta statistically increased in the CM group (p = 0.04). Moreover, no obvious difference was observed in the concentrations of ammonia N and total SCFAs in the jejunal digesta of the piglets in the SM group and CM group.

In this study, the number of average raw sequences detected in each group was at least 52,629 reads with 51,369 valid sequences (Supplementary Table S1). The overall number of OTUs detected was 17,756 at least in each sample, based on a 97% sequence similarity between reads. The rarefaction curves (Supplementary Figure S1) were generated by MOTHUR via plotting the number of reads versus the number of observed species and tended to approach the saturation plateau. The curves show that the sequencing amount was sufficient to evaluate the bacterial community in both groups. As shown in Supplementary Table S2, on d 28 of age, no statistical difference was observed in the diversity indices (Simpson) and richness estimators (observed species and Chao 1) of jejunal bacteria between SM group and CM group (p>0.05). However, compared with the SM group, the indices (Shannon) tended to increase in the CM group (p = 0.07). On d 35, no statistical difference was observed in the diversity indices (Shannon and Simpson) and richness estimators (observed species and Chao 1) of jejunal bacteria between the SM group and the CM group.

The bacterial community was assessed at different taxonomic levels. At the phylum level, as shown in Figure 2, the Firmicutes was the predominant phylum in all samples by d 28 of age, accounting for 95.97% of the total 16S rRNA gene sequences. Proteobacteria was the second dominant phylum, represented by 3.00% of total 16S rRNA gene sequences, followed by the bacteria from phyla Bacteroidetes (0.25%), and Actinobacteria (0.54%). By d 35 of age, the Firmicutes was the predominant phylum in all samples, accounting for 77.97% of the total 16S rRNA gene sequences. Cyanobacteria/Chloroplast was the second dominant phylum, represented by 11.78% of total 16S rRNA gene sequences, followed by the bacteria from phyla Actinobacteria (5.72%), Proteobacteria (4.09%). No statistical difference in the bacterial community at the phylum level was observed between the SM group and the CM group.

At the genus level, as shown in Figure 3, by d 28 of age, the predominant genera were the Lactobacillus, Streptococcus, Veillonella, and Clostridium XI, while, there was no statistical difference between the SM group and the CM group (p>0.05). By d 28 of age (Figure 4A), piglets in the CM group had a higher abundance of Clostridium XI, Tuicibater, and Moraxella in the jejunum than those of piglets in the SM group (p<0.05). By d 35 of age (Figure 4B), the Lactobacillus, Streptophyta, Streptococcus, and Clostridium XI were predominant. In addition, the abundance of Lactobacillus was significantly increased, while the abundance of Streptococcus and Blautia was statistically decreased in the CM group by d 35 of age (p<0.05).