Effects of Multi-Strain Probiotics and Perilla frutescens Seed Extract Supplementation Alone or Combined on Growth Performance, Antioxidant Indices, and Intestinal Health of Weaned Piglets

Simple Summary Weaning piglets face stressors from changes in feed and environment, which affects their growth. To resolve this problem, we explored the separate effects of multi-strain probiotics and Perilla frutescens seed extract and their combined effect on weaning piglets. We found multi-strain probiotics or Perilla frutescens seed extract both improved the gain to feed ratio and antioxidant capacity. In addition, multi-strain probiotics improved jejunal villus height and the villus height/crypt depth ratio. Perilla frutescens seed extract improved ileal villus height. The interactive effects were observed in jejunal villus height and the villus height/crypt depth ratio, ileal villus height, and the gene expression of IL-1β and mucin2 in the intestinal mucosa. This study shows that using either multi-strain probiotics or Perilla frutescens seed extract alone is more effective than their combined use in weaning piglets. Abstract This study examined the effects of multi-strain probiotics (BL) and Perilla frutescens seed extract (PSE), alone or in combination, on weaning piglets. In total, 96 weaning piglets were allocated into four treatments: CON group (the basal diet), PSE group (basal diet + 1g/kg PSE), BL group (basal diet + 2 g/kg BL), and BL+PSE group (basal diet +1 g/kg PSE + 2 g/kg BL) according to a 2 × 2 factorial arrangement. The supplementation of BL or PSE improved the gain to feed ratio. Dietary BL reduced diarrhea occurrence and Escherichia coli, but increased Lactobacillus counts in the ileal digesta. Dietary PSE tended to increase Lactobacillus counts in the ileal digesta. Interactive effects were found in terms of ileal villus height, the gene expression of IL-1β, and malondialdehyde in the ileal mucosa. Dietary BL lowered malondialdehyde in the spleen, liver, and jejunal mucosa but increased the total antioxidant capacity (T-AOC) in the liver and ileum mucosa. The supplementation of PSE improved superoxide dismutase in serum and T-AOC in the liver, and reduced MDA in liver, spleen, and jejunum mucosa. Taken together, BL or PSE showed positive effects, improving growth and intestinal morphology and enhancing antioxidant capacity. However, their interaction showed no beneficial effects on the antioxidant indices and the intestinal morphology of weaned piglets.


Introduction
Probiotics are frequently used as additives in livestock feed as they are beneficial to the gut. The best-known probiotics are Bacillus and Lactobacillus. Acid-resistant Bacillus

Materials and Methods
This study was conducted at the experimental base in Ya'an (Sichuan, China). Animal experiments were approved by the Animal Care and Use Committee of Sichuan Agricultural University (Sichuan, China, approval code:20181215), and conducted in accordance with animal protection laws.

Animals, Design, and Housing
In total, 96 Duroc × (Landrace × Yorkshire) weaned piglets (30 ± 1 days old) weighing 8.20 ± 0.29 kg were divided into 4 treatment groups according to a 2 × 2 factorial arrangement. Each treatment group consisted of 6 pens with 4 pigs (2 male and 2 female). Diets included: (1) control diet, basal diet; (2) PSE diet, basal diet + 1 g/kg PSE; (3) BL, basal diet + 2 g/kg BL; and (4) basal diet + 1 g/kg PSE+2 g/kg BL. The basal corn-soybean diet was formulated to meet the nutritional needs of NRC [21] for piglets (7-25 kg) and was free from any antibiotics, as shown in Table 1. Piglets were housed in the pens (2 m × 1.5 m × 0.8 m) equipped with a nipple drinker and feeder with free access to water and feed. The temperature was kept at 26-28 • C and relative humidity was kept at 55-65%. . It contained concentrations of 3.5 × 10 9 CFU/g of B. subtilis, 1.75 × 10 9 CFU/g B. subtilis var natto, and 1.75 × 10 9 CFU/g L. sporogenes. They were all dormant spores. The PSE was provided by Chongqing Super Science & Technology Development Co., Ltd. It was extracted by supercritical CO 2 , containing ≥2 g/kg alpha-linolenic acid, ≥0.6 g/kg linoleic acid, and ≥0.14 g/kg flavonoids. In order to ensure the freshness of the feed and the activity of multi-strain probiotics, we formulated the feed once weekly and usually put it in cold storage.

Sample Collection
The individual fasted body weight (BW) was determined at 08:00 on days 8, 15, and 22. The feed allocation and refusals were recorded daily on a pen basis. Growth performance data included average daily gain (ADG), average daily feed intake (ADFI), and ADG/ADFI (G/F). Scores were given thorough observing signs of stool consistency, and the average stool score per pen was recorded twice daily. The fecal consistency score was determined using the 5-point judging method of Giang et al. [10]. Briefly, the feces consistency criteria were as follows: 0, hard bar or granular; 1, soft stools but shapeable; 2, unshaped; 3, watery stool. Scores of 2 and 3 were considered diarrhea.
On days 8, 15, and 22, blood was obtained by aseptic needle puncturing the front vena cava of one pig per pen (fasted for 6 h) whose body weight was closest to the mean of the treatment group and gathered into vacuum tubes containing heparin sodium (5 mL; Shandong Yongkang Medical Products Co., Ltd., Shandong, China). Serum was harvested by centrifuging blood samples at 4 • C, at 3000× g for 10 min and stored at −20 • C. When the feeding trial ended, six piglets in each treatment (whose individual body weight was closest to the mean of the treatment group) were euthanized with injection of pentobarbital sodium. Their intestine was stripped from the mesentery and immediately placed on ice. Approximately 2-cm length tissues cut from the middle of the duodenum, jejunum, and ileum were put into paraformaldehyde solution (4% v/v) for morphological analysis. Duodenal, jejunal, and ileal mucosa were collected for antioxidant analysis as described by Chen et al. [22]. The liver samples were taken from the largest lobe after removing the surface membrane and stored at −80 • C. Spleen tissues taken from the same place were frozen at −80 • C for antioxidant analysis. Individual digesta collected from the jejunum and ileum was stored in sterile Eppendorf tubes and rapidly stored at −80 • C rapidly until analysis of microbiota according to the description of Hu et al. [5].

Assay of Oxidant and Antioxidant Indices in Serum, Liver, and Spleen Tissues, and Small Intestine Mucosa
In terms of assay of oxidant and antioxidant indices, we used the method described by Chen et al. [22]. Briefly, tissue and mucosa samples (1.0 g) were homogenized in 5.0 mL of ice-cold phosphate buffer (pH 7.2-7.4) and centrifuged at 4000× g for 10 min. The obtained supernatants were stored at −20 • C until further analysis. The concentrations of malondialdehyde (MDA) and total antioxidant capacity (T-AOC), and activities of superoxide dismutase (SOD), glutathione peroxidase (GSH-Px), and catalase (CAT) in sera and supernatants were assayed with matching kits (Product Nos. A003-1-2, A015-2-1, A005-1-2, A001-3, and A007-1-1, Nanjing Institute of Jiancheng Biological Engineering, Nanjing, China). Inter-and intra-assay variations were kept below 15%. Protein concentrations in the tissues and mucosa (duodenum, jejunum, ileum, liver, and spleen) were detected using a BCA Protein Assay Kit (Beyotime Biotechnology Inc. Shanghai, China). All samples were analyzed in duplicate.

Small Intestine Morphology Analysis
As intestinal morphology plays a vital function in absorbing nutrients and changes after weaning [23], we evaluated the morphology of duodenum, jejunum, and ileum. As previously described [24], the fixed intestinal segments (jejunum and ileum) were rinsed using running water for 30 min and subsequently dehydrated with ethanol at varying concentrations. The tissues were cleared with xylene, embedded in wax, and sliced into 5 µm-thick slices using a Leica RM2235 microtome (Leica, Germany). Finally, these tissue slices were dewaxed and subjected to hematoxylin-eosin staining. For each well-oriented villus, 10 measurements were recorded for both villus height (VH) and crypt depth (CD) using Image Pro Plus 6.0 (Media Cybernetics, Inc., Bethesda, MD, USA). The average of these 10 measurements was used to represent the VH and CD for each tissue. The V/C ratio was calculated by dividing the VH value by the CD value.

Determination of Enzyme Activity in Digestive Species
As previously described [5]. Frozen digesta was weighed and diluted (1:9, w/v) with phosphate buffer before cells were disrupted using an ultrasonic homogenizer (Scientz-48L, Ningbo, China). The supernatant was collected after centrifugation (3500× g, 10 min, 4 • C) and used for detecting the activity of lipase, trypsin, and amylase using the corresponding kits with the catalog numbers of A054-2-1, A045-2-2, and C016-1-1 (Jiancheng Bioengineering Institute, Nanjing, China). Parallel determination was carried out for individual indices. Coefficients of inter-and intrasample variations were all controlled within 12%.

Total RNA Extraction and Quantitative Real-Time PCR
To assess the effects of BL and PSE on the gut barrier genes and inflammatory factors, we measured the gut mucosal genes according to the method of Hu et al. [5]. The total RNAs of ileum mucosa (about 0.1 g) were extracted by Trizol (TaKaRa Biotechnology Co., Ltd., Dalian, China) following the manufacturer's instructions. The extracted RNA was evaluated by a NanoDrop 2000 (Thermo Scientific, Wilmington, DE, USA) with the OD 260 /OD 280 ratio in the range of 1.8 to 2.0, and quality was assessed via gel electrophoresis (1% w/v agarose). After dilution, 1 µg RNA was reverse transcribed using HiScript ® III RT SuperMix (+gDNA wiper, Vazyme Biotech Co., Ltd., Nanjing, China). The primers (Table S1) for mucin1, mucin2, zonula occludens-1 (ZO-1), claudin1, and β-actin were designed via Prime Premier 6.0 and synthesized by Sangon Biotech Co., Ltd. (Chengdu, China). Real-time quantitative PCR analysis was carried out using SYBR qPCR Master Mix (Vazyme Biotech Co., Ltd., Nanjing, China) and an ABI 7500 real-time PCR instrument QuantStudio 6 (Thermo, USA) with a reaction volume of 20 µL. For real-time PCR, samples underwent denaturation at 95 • C for 30 s, followed by 40 cycles of 95 • C for 10 s and 30 s at 60 • C. Individual samples were assayed in triplicate. Relative expression levels of target genes were calculated using the 2 −∆∆Ct method [25] by normalization to β-actin.

Microbial DNA Isolation and RT-qPCR
The total bacterial DNA in ileal chyme (n = 6 for each treatment) was extracted with the Stool DNA Kit (Omega Bio-Tek Doraville, CA, USA) according to the manufacturer's manual. Primers and probes (Table S2) were designed via Primer Premier 5.0 (Premier Biosoft International, Palo Alto, CA, USA).
Microbial RT-qPCR analysis and measurement was carried out as described by Chen et al. [26]. The total number of bacteria was determined with CFX-96 RT-qPCR and SYBR ® -qPCR Master Mix (Vazyme Biotech Co., Ltd., Nanjing, China) in a reaction volume of 25 µL, while the numbers of Lactobacillus, Bifidobacterium, Bacillus, and E. coli were determined with CFX-96 RT-qPCR with SuperReal PreMix (TIANGEN Biotechnology Co., Ltd., Tianjin, China) in a reaction volume of 20 µL. Microbial data were converted into log 10 -for analysis referring to the standard curve made by Chen et al. [26].

Statistical Analysis
Data were analyzed as a 2 × 2 factorial using the mixed procedure of SAS 9.4 (Cary, NC, USA). The BL, PSE, and interaction of BL and PSE were fixed effects, and the pigs were random effects. Duncan's multiple comparison was used to separate means when interactive effects were significant. For serum parameters, diarrhea scores, ADG, ADFI, and G/F, the individual pen was regarded as an experimental unit. To analyze microbiota, relative gene expression, and antioxidant indicators of tissue, every euthanized pig was a statistical unit. Results were considered significant when p < 0.05 while 0.05 < p < 0.1 was considered to be a tendency.

Growth Performance
Dietary PSE tended to reduce G/F (p = 0.063) but BL tended to improve G/F during days 8-14 (p = 0.068). Dietary BL increased the G/F of piglets (p = 0.034) from days 1 to 21, as shown in Table 2, while PSE increased the G/F (p = 0.006) from days 15 to 21. However, the combined supplementation of BL and PSE showed no interactive influence on the growth performance of piglets (p > 0.10).

Diarrhea Scores and Intestinal Microbes
As shown in Figure 1, BL supplementation reduced (p < 0.001) the diarrhea scores of piglets from days 8 to 14 and from days 1 to 21, whereas there was no interactive effect of BL and PSE on the diarrhea scores (p > 0.10). Dietary PSE tended to reduce diarrhea during days 1-7 (p = 0.068). Piglets fed on diets containing BL had a higher (p = 0.006, Table 3) Lactobacilli count and a lower E. coli count in the ileum chyme (p = 0.008). The PSE diet tended to boost the proliferation of Lactobacilli (p = 0.057).

Intestinal Mucosa Morphology and Digestive Enzymes
As shown in Table 4, dietary BL boosted the VH and V/C ratio in the jejunum (p < 0.05). The supplementation of PSE in the diet increased the VH in the ileum (p = 0.003) and tended to increase the ileal V/C ratio (p = 0.088). However, the interactive effects of BL and PSE were detrimental to the jejunal V/C ratio and ileal VH (p < 0.001). Additionally, there

Intestinal Mucosa Morphology and Digestive Enzymes
As shown in Table 4, dietary BL boosted the VH and V/C ratio in the jejunum (p < 0.05). The supplementation of PSE in the diet increased the VH in the ileum (p = 0.003) and tended to increase the ileal V/C ratio (p = 0.088). However, the interactive effects of BL and PSE were detrimental to the jejunal V/C ratio and ileal VH (p < 0.001). Additionally, there were interactive effects on VH and CD in the jejunum (p < 0.05), but neither increase nor decrease was found compared with other groups. In Table 5, PSE elevated the amylase activity in the jejunum (p = 0.011), while dietary BL induced higher trypsin activity in the ileum (p = 0.001). The interaction of BL and PSE decreased lipase activity in the jejunum (p = 0.049), but had the tendency to increase the trypsin activity in the ileum (p = 0.082).

Oxidation and Antioxidant Indices in Serum, Liver, Spleen, and Small Intestine
In Table 7, PSE improved (p = 0.001) SOD activity on day 21, decreased the MDA concentrations in the serum on day 7(p = 0.011), and tended to decrease MDA concentration on day 14 (p = 0.094). The interactive effect of BL and PSE tended to increase SOD on day 21 and GSH-Px activity on day 14 (p < 0.10). There was an interaction in MDA concentration on day 7 (p = 0.046). The dietary BL had the tendency to reduce the MDA concentration on day 14 but tended to increase the GSH-Px concentration on day 14 (0.05 < p < 0.1). Piglets fed BL had higher T-AOC concentrations in the liver (p = 0.016, Table 8) and lower MDA concentrations in the spleen (p = 0.003), whereas piglets fed on the PSE diet had higher T-AOC concentrations in the liver and lower MDA concentrations in the liver and spleen (p < 0.01). Additionally, in Table 9, PSE lowered MDA concentration in the jejunal mucosa (p = 0.024, Table 9), while BL increased T-AOC concentrations in the ileal mucosa (p = 0.019) and tended to reduce MDA concentration in the jejunal mucosa (p = 0.090). An interactive effect was observed in the MDA concentration in the jejunal mucosa, in which the MDA was higher in the +PSE+BL group compared with the +PSE-BL group, but lower compared with the -BL-PSE group (p < 0.05).   1 Values were presented as means ± SEM, n = 6. 2 BL, multi-strain probiotics; PSE, Perilla frutescens seed extract; −PSE−BL, basal diet; +PSE−BL, basal diet + 1 g/kg Perilla frutescens seed extract; +BL−PSE, basal diet + 2 g/kg multi-strain probiotics; +BL + PSE, basal diet+2 g/kg multi-strain probiotics + 1 g/kg Perilla frutescens seed extract. A,B,C Different letters on the shoulders of peers indicate significant differences (p < 0.05).

Discussion
Weaning stress results in diarrhea, oxidant stress, poor immune function, and growth arrest [26]. Functional probiotics and plant extracts have been reported to ease these symptoms when weaning [7,[26][27][28]. Therefore, it was necessary to explore the effects of BL and PSE alone and combined on weaned piglets.

Growth Performance
This study revealed that BL supplementation improved G/F of weaned piglets, but had no effect on ADFI during days 8-14 and 1-21. This result was in agreement with similar studies of Bacillus and Lactobacillus probiotics in weaned piglets [8,29,30]. As previously reported, Bacillus spp.-based probiotics can stimulate the activity of non-starch polysaccharide degrading enzymes and proteases, and augments nutrient digestibility [3]. Therefore, we guessed that the improved trypsin activity by BL may stimulate the decomposition of proteins and then increase nutrient absorption. Sequentially, the tendency improvement of ADG by BL may be contacted with lower diarrhea score during days 8-14, as probiotics can effectively prevent diarrhea in weaned piglets [31]. Likewise, PSE supplementation also improved the G/F from day 15 to day 21. Plant extracts containing essential oils, flavonoids, and hydroxybenzene can improve growth performance [32], which may result from improved nutrient digestibility. Reportedly, alpha-linolenic acid and alpha-linoleic acid in the PSE are easily absorbed due to the high digestibility of fat and protein [33]. In the present study, improved amylase activity by PSE was more conducive to decomposing starch to easily absorb maltose and glucose [34]. PSE tended to reduce diarrhea scores during days 1-7. However, PSE tended to reduce G/F during days 8-14 under the premise that there were no effects on the diarrhea score. As there was no research of PES on piglets, the most likely explanation was that in the former, the piglets had a maladjustment for 1 g/kg PSE.
The feed conversion ratio is a vital parameter in assessing the economic efficiency of animal husbandry. In this study, BL or PSE were shown to boost growth performance. This may be due to better gut characteristics, including healthy and stable histomorphology, balanced intestinal microflora, and better intestinal barrier integrity.

Diarrhea Scores and Microbes
The present study revealed that piglets fed multi-strain probiotics had lower diarrhea scores. The results are in agreement with the beneficial effects of Bacillus spores and Lactobacillus supplementation on reducing the incidence of diarrhea in weaned piglets [35,36]. In the report by Du et al. [36], Bacillus WS-1 prevented diarrhea and deaths of newborn piglets caused by E. coli injections, as Bacillus strains are well known to secrete surfactin, which acts against pathogenic bacteria. In the present study, the decreased diarrhea scores in BL-fed piglets were probably associated with the reduced E. coli counts and the increased Lactobacillus counts. The modulation of intestinal microorganisms by oral probiotics has been proven effective to reduce harmful E. coli counts and enrich healthy bacteria [30,37]. Reportedly, PSE can inhibit E. coli proliferation in vitro [38], whereas it showed no role in adjusting E. coli numbers in vivo in this study, but tended to boost Lactobacillus counts in the late stage of the trial. However, the latest report revealed that the Perilla extract had no inhibition on E. coli or Salmonella in vitro [39]. On the one hand, the Perilla variety, harvest season, and planting environment will have impacts on its composition [39]. On the other hand, when entering the body, PSE may be digested by digestive enzymes and absorbed by intestinal epithelial cells, so they cannot play a role in the hindgut.

Intestinal Mucosa Morphology
Weaning piglets are unsuitable for the abrupt changes in diet and line -haul with empty, resulting in reduced VH and increased CD [23,40]. In physiology, intestinal histomorphology is vital in assessing intestinal health and digestive function. We found that BL increased VH in the jejunum and ileum and increased the jejunal V/C ratio, which was in agreement with the results of Sayan et al. [41], who found that sucking piglets orally administered with L. salivarius had higher VH and V/C values in the jejunum and ileum. Meanwhile, piglets fed on diets containing PSE had numerically higher VH and V/C ratios in the ileum. Similar results were also reported by Wang et al. [29] with respect to the effects of herbs on weaning piglets. Plant extracts can stimulate the reconstruction of intestinal microvilli by activating the cytokeratin gene [42], possibly explaining the effects of PSE on morphology. The VH reflects the epithelial turnover speed and is related to the activation of cell mitosis [43], showing the protective effects of BL and PSE on epithelial mucosa. B. subtilis could prompt the development of wave-like villus arrays that increase the absorption area and prolong residence time in the digestive tract [44], which may explain the higher G/F in the BL groups.

Gut Barrier Gene Expression
Weaning induces intestinal epithelial barrier damage, resulting in increased intestinal permeability, the invasion of intestinal pathogens, and decreased nutrient digestibility [44,45]. Therefore, an intact mucosal barrier is important to ensure nutrient absorption. This barrier is established by tight junction proteins, including claudin and occludin and intracellular linker proteins, such as ZOs [45]. In the current study, PSE upregulated the gene expression of mucin2 but downregulated that of IL-1β. Phenolic acid and flavonoid constituents have a protective function on the gene expression of ZO-1 and occludin because it is a potent inhibitor of protein tyrosine kinases that prevents tight junction protein from being phosphorylated [46]. Mucin secreted by goblet cells is regarded as a contributor to regulating intestinal inflammation by protecting against bacterial and parasitic invasions [45]. The high expression of the pro-inflammatory cytokine IL-1β disturbs tight junctions [29]. Interestingly, BL also decreased IL-1β gene expression. Strangely, PSE tended to improve the gene expression of TNF-α, the pro-inflammatory cytokines. The effects of PSE on gut tight junction proteins in weaning piglets have not been explored, and merit more attention in future studies.

Antioxidant Capacity
Oxidant stress results in weaker growth performance and lower nutrient digestibility in weaned piglets [47]. Changes in diet, transport, and vaccine injection are major stressors, leading to the excessive production of free radicals that exceeds the removal rate of the antioxidant system, which leads to inflammation and damages the intestinal barrier. The levels of SOD, CAT, and GSH-Px are important antioxidant enzymes, and T-AOC in the body reflects the body's ability to eliminate oxygen radicals. SOD converts reactive oxygen species into hydrogen peroxide; next, CAT and GSH-Px play roles in degrading unstable hydrogen peroxide to water and oxygen [22,48]. Another index is MDA, one of the final products of lipid peroxidation as well as an excellent oxidative stress marker [48]. In the present study, BL tended to augment GSH-Px and increased T-AOC capacity in the liver and ileum and reduced the MDA concentrations in the spleen and jejunum. Numerous studies have shown that Bacillus spp.-based probiotics improved T-AOC of sows and weaned lambs [48,49]. Probiotics enhanced the antioxidant capacity of animals by utilizing their antioxidant enzyme system of SOD, CAT, and NADH-oxidase or stimulating the host's antioxidant system as well as probiotics-mediated antioxidant signaling pathways [50]. PSE induced higher SOD activity in serum and higher T-AOC capacity in the liver and decreased the MDA concentrations in the liver, spleen, and jejunum mucosa. Perilla extract consists of natural antioxidants, such as apigenin, citral, luteolin, perillaldehyde, quercetin, and rosmarinic acid, which chelate metal ions that produce free radicals and inhibit some enzymes involved in radical generation, due to special benzenoid rings and phenolic hydroxyl groups [51,52]. The results were linked to reports that plant extracts containing polyphenols and flavonoids strengthened CAT activity and the gene expression of SOD1 and decreased MDA concentrate [53]. The enhancive antioxidant indices may reflect the protective effects of BL and PSE on the spleen, which involves the immune system; the liver, which participates in detoxification and metabolism [54]; and the intestinal mucosa, which maintains the intestinal barrier [55].

Interactive Effect
Interestingly, when piglet diets were supplemented with both BL and PSE, antagonistic effects on ileal VH and the jejunal V/C ratio were observed. Analogously, Zangeronimo et al. [56] certified that 1 g/kg plant extract mixed with 3 g/kg probiotics in the diet of weaned piglets decreased the VH compared with 0.5g/kg plant extract and 3g/kg of probiotics. The lack of increase in E. coli count in this period excluded the possibility of damage by microorganisms, and litter deeper CD predicted the possibility of intestinal inflammation [57], as proved by higher MDA concentrations in the jejunum and serum when BL and PSE were used together in this study. Maybe the unaffected growth performance was due to the above-mentioned and the decline in lipase activity. However, there was a contradiction between the downregulation of IL-1β expression, the upward trend of Claudin-1, the downward trend of ZO-1, and the increasing tendency in GSH-Px and SOD in serum by their interaction. These conflicting indicators are currently unexplained. Studies have found that PSE has a great antibacterial effect in vitro [58], but we are not sure of the specific action in the body.

Conclusions
This study provides the first evidence that PSE administration to weaned piglets increases G/F, strengthens the antioxidant capacity of the serum, liver, and jejunal mucosa, and ameliorates ileal morphology and barrier function. Feeding BL promoted G/F of weaned piglets, which may be associated with an enhanced antioxidant capacity in the liver, spleen, and ileal mucosa, and an improved jejunal morphology, and modulated the counts of E. coli and Lactobacillus. However, the combination of BL and PSE had no effect on growth performance, possibly because of the antagonistic effect on antioxidant capacity and intestinal morphology, but this remains to be determined.