Microbial shifts in the porcine distal gut in response to diets supplemented with Enterococcus Faecalis as alternatives to antibiotics

Gut microbiota plays an important role in host health and nutrient digestion of animals. Probiotics have become one of effective alternatives to antibiotics enhancing animal health and performance through modulating gut microbiota. Previously, our research demonstrated that dietary Enterococcus Faecalis UC-100 substituting antibiotics enhanced growth and health of weaned pigs. To investigate the alterations of microbiota in the distal gut of pigs fed E. faecalis UC-100 substituting antibiotics, this study assessed fecal microbiota in pigs from different dietary treatments: the basal diet group, the E. faecalis group, and the antibiotic group on d 0, 14, and 28 of feeding through 16 S rRNA sequencing. Twenty-one phyla and 137 genera were shared by all pigs, whereas 12 genera were uniquely identified in the E. faecalis group on d 14 and 28. Bacterial abundance and diversity in the E. faecalis group, bacterial diversity in the antibiotic group, especially abundances of Fibrobacteres phylum and 12 genera in the E. faecalis group and antibiotics group were lower than that in the basal diet group on d 28. These results showed that microbial shifts in the porcine gut in response to diets containing E. faecalis were similar to the response to which containing antibiotics.

The total read length was 1.97 gigabases (GB), and the average read length per sample was 0.05 GB. On d 0, 14, and 28 of feeding, there were 201,303, 212,206, and 222,669 raw reads in pigs from the basal diet group; 222,402, 217,627, and 216,349 raw reads in pigs from the E. faecalis group; 196,387, 207,015, and 215,410 raw reads in pigs from the antibiotic group, respectively (Table S1).
After quality control, 1,846,755 high quality sequences were obtained. On average, 51,298 sequences were obtained per sample. On d 0, 14, and 28 of feeding, there were 6,073, 6,378, and 6,908 operational taxonomic units (OTUs) in pigs from the basal diet group; 6,032, 6,459, and 6,235 OTUs in pigs from the E. faecalis group; 5,446, 5,856, and 6,274 OTUs in pigs from the antibiotic group, respectively, based on 97% species similarity (Table S1). A total of 9,966 OTUs were identified from all fecal samples (Table S1). Most OTUs were shared among groups at the same age, only 953 and 749 OTUs were uniquely identified in pigs from the E. faecalis group on d 14 and 28 of feeding, respectively (Fig. S1A). In addition, 753 and 729 OTUs were uniquely identified in pigs from the antibiotic group on d 14 and 28 of feeding, respectively (Fig. S1B).

Shifts in microbial abundance and diversity after E. faecalis treatment. Good's coverage was at
least 96% for each group. The range of the calculated value for Ace value was 5,887-7,482 over the 3 sampling times (d 0, 14, and 28 of feeding). On d 28 of feeding, the bacterial abundance in pigs from the E. faecalis group was lower (P < 0.01) than those from the basal diet group, whereas there was no difference between the E. faecalis group and the antibiotic group (Fig. 1A). And no difference in bacterial abundance was detected between the antibiotic group and the basal diet group on either d 14 or 28 of feeding (Fig. 1A). Compared with d 0 of feeding, bacterial abundance in pigs from the basal diet group and the antibiotic group increased (P < 0.01) on d 28, whereas bacterial abundance in the E. faecalis group increased (P < 0.05) on d 14 (Fig. 1B). No difference in bacterial diversity was detected among 3 groups on either d 14 or 28 of feeding (Fig. 1C). Compared with d 0 of feeding, bacterial diversity increased (P < 0.05) on d 28 from the basal diet group, whereas no change in the E. faecalis group and the antibiotic group (Fig. 1D). Weighted UniFrac distances were used to estimate β -diversity and to compare the three diet groups on d 14 ( Fig. 2A,B and C) and d 28 of feeding (Fig. 2D,E and F). The PCoA plot of the weighted UniFrac distances showed that the three diet groups did not form distinct clusters on either d 14 or 28 of feeding, although the antibiotic group microbiota tended (P = 0.063) to separate from the Enterococcus faecalis group microbiota along principal coordinate 3 on d 14 ( Fig. 2B and C).

Shifts in community membership after E. faecalis treatment.
A total of 21 phyla were shared by pigs from all groups ( Fig. S2 A), as follows: Actinobacteria, Bacteroidetes, Caldiserica, Chlamydiae, Chloroflexi, Crenarchaeota, Cyanobacteria, Deferribacteres, Euryarchaeota, Fibrobacteres, Firmicutes, Fusobacteria, Lentisphaerae, Planctomycetes, Proteobacteria, Spirochaetes, Synergistetes, Tenericutes, Thermi, TM7, and Verrucomicrobia. Of them, Firmicutes was the most dominant among the 21 phyla (P < 0.01) in the samples, and comprised more than 85% of the total sequences. Tenericutes and Bacteroidetes were the 2nd and 3rd dominant phyla which comprised only 1% of the total sequences. The bacterial abundance of Fibrobacteres in pigs from the E. faecalis group was lower (P < 0.01) than that in the basal diet group on d 28 of feeding, while no difference in bacterial abundance of Fibrobacteres was detected between the E. faecalis group and the antibiotic group (Fig. 3A,B). Meanwhile, the bacterial abundance of Fibrobacteres in pigs from the antibiotic group tended (P = 0.063) to be lower than that in the basal diet group on d 28 of feeding (Fig. 3A).
At the genus level, a total of 137 genera were identified from all samples (Fig. S2 B). The 11 most abundant genera, containing more than 85% of the total sequences, were Lactobacillus, Bulleidia, Clostridium, Streptococcus, Chlamydia, Coprococcus, Oscillospira, Eubacterium, Treponema, Ruminococcus, and Blautia. Of them, Chlamydia was a member of the phylum Chlamydiae, Treponema belongs to the phylum Spirochaetes and the other 9 genera belong to the phylum Firmicutes. Among the 11 most abundant genera, Lactobacillus, Bulleidia, and Clostridium were the most predominant genera, accounting for 23, 13, and 11% of total sequences, respectively. Most genera were shared among 3 groups at the same age, only 9 genera (Anaerovorax, Brevibacterium, Deinococcus, Facklamia, Ignatzschineria, Mycoplasma, Pedobacter, Sphingobium and Vibrio) and 3 genera (Burkholderia, Paraprevotella and Stenotrophomonas) were uniquely identified in pigs from the E. faecalis group on d 14 and 28 of feeding, respectively (Table 1). Only 10 (Acetobacter, Aequorivita, Anaerococcus, B-42, Holdemania, HTCC, Mycobacterium, Sporanaerobacter, Tsukamurella and Veillonella) and 1 (Ramlibacter) genera were uniquely identified in pigs from the antibiotic group on d 14 and 28 of feeding, respectively ( The abundance of 12 genera (Acholeplasma, Arcobacter, Caldicoprobacter, Desulfotomaculum, Ignatzschineria, KSA1, Leptolyngbya, Natronincola_Anaerovirgula, Pseudomonas, Pseudoramibacter_Eubacterium, Tepidimicrobium, and Tissierella_Soehngenia) in pigs from both the E. faecalis group and the antibiotic group were lower (P < 0.05) than that in the basal diet group on d 28 of feeding, while no difference in bacterial abundance of these genera were detected between the E. faecalis group and the antibiotic group (Tables 3,4). The abundance of 5 genera (Anaerococcus, Fibrobacter, Megasphaera, Selenomonas and Sharpea) changed (P < 0.05) in pigs from the E. faecalis group than that in the basal diet group, and the abundance of 4 genera (Bacillus, Sphaerochaeta, Vibrio and Zhouia) changed (P < 0.05) in pigs from the antibiotic group than that in the basal diet group on d 28 (Table 5). However, no difference in bacterial abundance of these 9 genera was detected between the E. faecalis group and the antibiotic group (Table 5).

Discussion
Dietary supplementation of E. faecalis strains, as a probiotic, has become one of effective alternatives to the use of antibiotics to increase health and growth performance of pigs 24,25 as it has been shown that probiotics can affect gut microbiota which plays an important role in health and nutrient digestion in pigs [1][2][3][4] . Although many studies have examined the impact of antibiotics on the gut microbiota in pigs 1,4,[26][27][28][29] , there is very little information on how consumed E. faecalis affect the entire porcine gut microbiota 24,30 . Because many E. faecalis strains can inhibit pathogen by producing bacteriocin [19][20][21][22] , we try to study if E. faecalis UC-100 administration could induce alteration of the gut microflora to directly or indirectly impact porcine health and performance. Meanwhile, it is important to explore if E. faecalis UC-100 administration could cause alteration of the composition or activity of the host normal microbiota to exclude the possibility of the occurrence of undesirable microbiota changes before we applicate the E. faecalis UC-100 in pig production.
Since dietary supplementation of 200 g/t E. faecalis UC-100 showed the benefits similar to antibiotics supplementation from the previous manuscript 25 , samples from the E. faecalis UC-100 group(200 g/t), the basal diet group and the positive control diet group collected on d 0, 14, and 28 of feeding were used to determine alteration of the distal gut microbiota population in response to the treatment with E. faecalis substitute for antibiotics. We obtained 1,846,755 high-quality sequences from all samples, and the read counts were greater than those in previous studies in pigs 1,4,28 . Moreover, according to Good's coverage index (96%) of each sample, the modified sequences were comprehensively enough to cover most bacterial diversity.
Venn diagrams were generated to make qualitative comparisons among E. faecalis group, antibiotic group and basal diet group at the same age. Most of the OTUs were shared between groups at the same age (Fig. S1A), which indicates that unique OTUs to each group were more likely to be found as less abundant OTUs, and this result is consistent with previous study 1 . In the basal diet group, bacterial abundance and diversity were increased with age, and these results are in accordance with a previous study 31 . Whereas, the increase of bacterial diversity   and abundance were inhibited in pigs from the E. faecalis group compared with the basal diet group on d 28 of feeding. Mechanisms whereby E. faecalis UC-100 decrease bacterial diversity and abundance may be that E. faecalis strains can produce bacteriocin to inhibit pathogen and modulate other gut microbiota [19][20][21][22] . Although the increase of bacterial diversity was inhibited in pigs from the antibiotic group compared with the basal diet group on d 28 of feeding, no difference in bacterial abundance and diversity was detected between the antibiotic group and the basal diet group, and the addition of antibiotic to the swine diet did not shift the overall microbial community structure (β-diversity indices) on either d 14 or 28 of feeding. These results were similar to a previous study which showed that overall microbial community structure, microbial abundance and diversity in weaned pigs were not affected by chlortetracycline treatment 28 . Poole et al. 32 also found no significant effect on a-diversity when pigs were fed with chlortetracycline for 28 days. In contrast, Looft et al. 33 34 . Meanwhile, dietary antibiotics also tend to decrease the bacterial abundance of Fibrobacteres on d 28. These results are similar to results of a previous study where the authors showed that weaned pigs treated with tylosin had a lower proportion of Fibrobacteres sequences than those in the control group 28 . O'Toole PW et al. 35 reported that consumption of probiotic may modulate the microbiota by competing for nutritional substrates, and by altering the dynamics of carbohydrate utilization by individual microbiota components. It can be speculated that E. faecalis UC-100 may modulate the Fibrobacteres by competing for nutritional substrates such as cellulose.
Lactobacillus, Bulleidia and Clostridium were the most predominant genera in the present study. These results are similar to a previous study where they showed that Lactobacillus and Clostridium were the most dominant genera in pig fecal samples 31 . Although most genera were shared among groups at the same age, 12 genera were uniquely identified in pigs from the E. faecalis group on d 14 or 28 of feeding. Of these unique genera, Anaerovorax functions to reduce the susceptibility to Campylobacter infection in humans 36 , Deinococcus is safely used as a feed supplement for hens 37 , and Paraprevotella may contribute to host health 38 . Conversely, it was found that Achromobacte and Gemella were specific to the basal diet group on d 14 and 28 of feeding, and several species of these 2 genera are opportunistic pathogens that affect humans 39,40 . In addition, 11 genera including Veillonella were uniquely identified in pigs from the antibiotic group on d 14 or 28 of feeding. Some Veillonella species have the function of utilization of macro-and micro-nutrients and may contribute to the regulation of host metabolism and body weight in human gut 41 . The existence of the unique beneficial genera in the E. faecalis group or the antibiotic group and the unique opportunistic pathogens in basal diet group may be a potential factor related to decreased incidence of diarrhea and increased body weight gain in the E. faecalis group and the antibiotic group.
Moreover, it was found that the bacterial abundance of 12 genera were increased as pigs aged in the basal diet group, but decreased in both the E. faecalis and antibiotic group on d 28. Of these 12 genera, many species of Pseudomonas 42 , Acholeplasma 43 Table 5. The bacterial abundances of 9 distinct genera were compared among the basal diet group, the Enterococcus faecalis group and the antibiotic group on d 28 of feeding, whereas these genera shifts caused by the Enterococcus faecalis group was different with the antibiotic group.
antibiotic 1 and E. faecalis [19][20][21][22][23] can reduce or inhibit the presence of opportunistic pathogens, and this may be the reason that the bacterial abundance of these genera decreased and then caused the decreased incidence of diarrhea and the increased body weight gain in both E. faecalis UC-100 and antibiotic group. And the microbial shifted in the porcine gut in response to diets fed E. faecalis were similar to the response to dietary supplementation of antibiotics, indicating that E. faecalis UC-100 could be a potential alternative to the use of antibiotics in pigs to promote health and growth of host. In addition, the bacterial abundance of 4 genera including Fibrobacter and Megasphaera were decreased, and Selenomonas were increased only in the pigs from the E. faecalis group compared with the basal diet group on d 28. A decrease in Fibrobacter genus was consistent with the decrease in Fibrobacteres phyla in E. faecalis group on d 28 of feeding. Some species of Megasphaera may cause diarrhea 48 , and Selenomonas was related to obesity 49 . The decrease of Megasphaera and the increase of Selenomonas may cause the decreased incidence of diarrhea and the increased body weight gain in the E. faecalis group. Meanwhile, the bacterial abundance of Bacillus and Sphaerochaeta, were decreased and the bacterial abundance of Vibrio and Zhouia were increased only in pigs from the antibiotic group compared with the basal diet group. Some species of Bacillus may cause cutaneous, gastrointestinal, and inhalation anthrax 50 , its decreased abundance may cause the decreased incidence of diarrhea in the antibiotic group. Previous study 51 showed that bacteria of Vibrio might lead to development of acute gastroenteritis characterized by diarrhea, headache, vomiting, nausea, and abdominal cramps, its increased abundance in the antibiotic group might be due to its resistance to antibiotic 52 .
It was interesting to note that administration of E. faecalis UC-100 did not increase the abundance of Enterococcus genus. The lack of an effect on Enterococcus genera is probably due to the insufficient contribution of the Enterococcus faecalis strain, as E. faecalis UC-100 after intake still accounted for a minor part of the Enterococcus community in our samples.

Conclusion
The abundance and diversity of the gut microbiota in pigs of the E. faecalis group and the bacterial diversity in the antibiotic group were inhibited on d 28 of feeding. Most genera were shared among groups at the same age, 12 and 11 genera were uniquely identified in pigs from the E. faecalis group and the antibiotic group on d 14 or 28 of feeding, respectively. Several species of these unique genera can be beneficial to host health. The abundance of Fibrobacteres phylum and 12 genera including Fibrobacter and some opportunistic pathogens in pigs from both the E. faecalis group and the antibiotic group were lower than that in the basal diet group on d 28 of feeding. These results showed that microbial shifts in the porcine gut in response to diets fed E. faecalis were similar to the response to dietary supplementation of antibiotics, indicating that E. faecalis can be a potential alternative to the use of antibiotics in pigs.

Materials and Methods
Probiotics. E Animals and sample collection. This experiment was approved by Animal Care and Use Committee of Nanjing Agricultural University. All procedures and the use of animals were carried out in accordance with the Guide for the Care and Use of Laboratory Animals prepared by the Institutional Animal Care and Use Committee of Nanjing Agricultural University, Nanjing, China.
The experimental design and animal feeding procedure have been described previously 25 . Briefly, 150 newly weaned pigs (Duroc × Landrace × Yorkshire, 25 days of age at 8.4 ± 0.2 kg body weight, weaned at day 25) were allotted to 5 dietary treatments based on a randomized complete block design with gender and initial body weight as blocks. Each dietary treatment had 4 pens (replicates), and each pen had 7 or 8 pigs. Dietary treatments represent a basal diet, 3 test diets containing E. faecalis UC-100 at various levels (100, 200, and 400 g/t, respectively), or a positive control diet containing multiple antibiotics (bacitracin zinc 40 g/t, aureomycin 75 g/t, and colistin 20 g/t). Dietary treatments were given to pigs for 28 days. All pens were decontaminated and disinfected for 7 days before the pigs moved in to ensure minimal bacterial contamination. The building was temperature-controlled (26.3 ± 2 °C) during the study. Feed and water were available ad libitum for all pigs. Diet composition and nutrient contents are provided in the supplementary material (Table S2). The experiment was divided into 2 phases: phase I (from d 0 to d 14 of feeding) and phase II (from d 14 to d 28 of feeding). Fecal samples were collected from 1 randomly selected pig in all pens by rectal massage on d 0, 14 and 28 of feeding and then stored at − 80 °C before DNA extraction. Each group had the same ratio between barrows and gilts. Because 200 g/t E. faecalis UC-100 showed the benefits similar to antibiotic supplementation from the previous manuscript 25 , only 36 fecal samples from the E. faecalis UC-100 group(200 g/t), the basal diet group and the positive control diet group collected on d 0, 14, and 28 of feeding were used in the current study based on the objective of investing microflora changes in the porcine distal gut in response to the treatment with E. faecalis UC-100 as alternatives to antibiotics. Gut microbiota population in fecal samples were assessed through 16 S rRNA gene sequencing.
DNA extraction, PCR amplification of 16 S rRNA gene, amplicon sequence and sequence data processing. Microbial genomic DNA was extracted from 220 mg of each fecal sample using a TIANamp Stool DNA Kit (Spin Column, Cat. no. DP328) according to the manufacturer's recommendation. Successful DNA isolation was confirmed by agarose gel electrophoresis 31 .
The V4 hypervariable regions of 16 S rRNA gene were amplified by PCR using the barcoded fusion primers referred to previous study 53 . The primer sequences were 520 F 5-AYTGGGYDTAAAGNG-3 and 802 R 5-TACNVGGGTATCTAATCC-3. The PCR condition was as follows: initial denaturation at 94 °C for 4 min; 94 °C denaturation for 30 s, 50 °C annealing for 45 s, and 72 °C extension for 30 s, repeated for 25 cycles; final extension at 72 °C for 5 min. The PCR amplicon products were separated on 0.8% agarose gels and extracted from the gels. Only PCR products without primer dimers and contaminant bands were used for sequencing by synthesis. Barcoded V4 amplicons were sequenced using the paired-end method by Illumina MiSeq with a 7-cycle index read. Sequences with an average phred score lower than 30, ambiguous bases, homopolymer runs exceeding 6 bp, primer mismatches, or sequence lengths shorter than 100 bp were removed. Only sequences with an overlap longer than 10 bp and without any mismatch were assembled according to their overlap sequence. Reads that could not be assembled were discarded. Barcode and sequencing primers were trimmed from the assembled sequence 31 .
Taxonomy classification and sequence analysis. Taxon-dependent analysis was conducted using the Greengene database 54 . Greengenes is a quality controlled, comprehensive 16 S reference database and taxonomy based off a de novo phylogeny that provides standard operational taxonomic unit sets. OTUs were counted for each sample to express the richness of bacterial species with an identity cutoff of 97%. Low abundance OTUs (fewer than 5 reads) were filtered out of our analysis 55 . The OTU abundance of each sample was generated at genus level. The mean length of all effective bacterial sequences without primers was 227 bp. The abundance count at the genus level was log 2 transformed and then normalized as follows: from each log-transformed measure, the arithmetic mean of all transformed values was subtracted, and the difference was divided by the standard deviation of all log-transformed values for a given sample. After this procedure, the abundance profiles for all samples exhibited a mean of 0 and a standard deviation of 1.
A Venn diagram was generated to compare OTUs between groups, and the bacterial community indices applied here included Ace and Good's coverage. The bacterial abundance is shown by Ace. Good's coverage estimates what percent of the total species is represented in a sample. The bacterial diversity is shown by the number of OTUs. β-diversity was calculated using weighted UniFrac distance and displayed using principal coordinate analysis (PCoA) 28 .

Data analysis.
Only high-quality sequences obtained after quality control analysis were used in present analysis which were uploaded to QIIME for further study 54 . All effective bacterial sequences were compared to the Greengene databases using the best hit classification option to classify the abundance count of each taxon. The sequence length was archived by QIIME. The abundance and diversity indices were generated using Mothur with an OTU identity cutoff of 97% after implementing a pseudo-single linkage algorithm 1 . For all parameters, data were compared using a one-way analysis of variance (ANOVA) at the end of each bioassay. A mean comparison was performed using Fisher's least significant difference test (LSD) and the Duncan multiple range test with a significance level of P < 0.05.