Effect of Probiotics and Multi-Component Feed Additives on Microbiota, Gut Barrier and Immune Responses in Broiler Chickens During Subclinical Necrotic Enteritis

The withdrawal of antibiotic growth promoters from poultry feed has increased the risk of necrotic enteritis (NE) outbreaks. This study examined the effects of a probiotic (PROB) or probiotic/prebiotic/essential oil supplement (PPEO) during a subclinical NE challenge. On day (d) of hatch, 960 male broilers were randomized to four groups (8 pens/treatment, 30 birds/pen) including (1) negative control (NC): corn-soybean meal diet; (2) positive control (PC): NC + 20 g Virginiamycin/ton diet; (3) NC + 227 g PROB/ton diet; and (4) NC + 453 g PPEO/ton diet. One d after placement, birds were challenged by a coccidia vaccine to induce NE. Feed intake and body weights were measured on d 8 (NE onset) and end of each feeding period. On d 8, the small intestines of three birds/pen were examined for NE lesions. Jejunum samples and ileal mucosal scrapings from one bird/pen were respectively collected to measure mRNA abundance (d 8 and d 14) and profile the microbiota (d 8 and d 42). Data were analyzed in JMP or QIIME 2 and significance between treatments identified by LSD (P < 0.05). PROB and PPEO had significantly lower mortality (d 0–14) and NE lesion scores compared to NC. Feed conversion ratio was significantly lower in PC, PROB, and PPEO, while average daily gain was higher in PPEO and PC groups compared to NC from d 0–42. On d 8 and d 14, mRNA abundance of claudin-3 was higher in PPEO compared to NC. On d 14, compared to NC, mRNA abundance of sIgA and PGC-1α in PROB and PPEO were lower and higher, respectively. Compared to NC, PPEO increased mTOR abundance on d 14. On d 8, relative abundance of Clostridium sensu stricto 1, Ruminiclostridium9, Prevotellaceae, Prevotellaceae UCG-014, ASF356, and Muribaculaceae was higher in NC compared to PPEO and PROB, while Lactobacillus was lower in NC. Escherichia-Shigella had higher abundance in PC compared to PPEO and PROB. Collectively, these data indicate that during a subclinical naturally occurring NE, supplementation of PROB or PPEO supports performance and reduces intestinal lesions, potentially through modifying tight junction proteins, gut microbiota, immune responses, and cell metabolism.


INTRODUCTION
Necrotic enteritis (NE) is a significant enteric disease in poultry with considerable economic effect on profitability. The annual financial losses to the poultry industry worldwide are estimated at $6 billion (1). NE presents a unique challenge, being a complex enteric disease that often leads to either clinical (acute) or subclinical (chronic) form (2,3). The latter typically results in poor performance (reduced feed intake, weight gain, and eventually higher feed conversion ratio) with low mortality rates and represents the greatest economic impact on poultry production (4,5). Predisposing factors such as Eimeria and fishmeal would lead to disturbances in the microbiota providing a favorable ecological environment and nutrients that allow the proliferation of Clostridium perfringens, thus subclinical NE occurrence (5,6).
Gut barrier function is an important factor for the optimum health and performance of broiler chickens. Maintenance of gut integrity and function is a highly energy demanding process, thus energy deficits due to alteration in the intestinal metabolism and mitochondrial functionality may cause intestinal damage and injury (7). In addition, during infection and inflammation (e.g., NE), cell proliferation in the intestine accelerates to replace damaged enterocytes (8). This epithelial cell renewal along the crypt-villus axis is regulated by the mechanistic target of rapamycin (mTOR) signaling pathway. This may partly be due to its ability to affect the antioxidant capacity and protein synthesis of intestinal epithelial cells (9,10). In many intestinal diseases, cell loss from the villus exceeds the regenerative capacity of the crypts (11) and shedding of epithelial cells may cause transient gaps or microerosions in the epithelial barrier resulting in increased intestinal permeability (11). Tight junctions (TJ) are protein complexes that seal the paracellular space between adjacent cells in the gut, maintaining epithelial barrier function (12). However, these barriers have a dynamic structure that remodel in response to signals from the cells or external stimuli such as nutrients, and pathogenic and commensal bacteria (13). C. perfringens, the pathogen responsible for NE, produces enterotoxins that bind to TJ proteins, mainly claudins (CLDN)-3 and -4 (14,15), which eventually leads to pore formation, an increase in paracellular permeability, and cytotoxicity (15,16). In addition to being barriers, TJ contain essential components of the signaling networks that guide diverse cell behaviors and functions and regulate epithelial proliferation and differentiation (17). As a result, TJ signaling plays an important role in general cell responses to stress. Zonula occluden (ZO) proteins are linked to the maintenance of junctional integrity in response to osmotic stress or changes in cell shape during morphogenesis in non-avian models including zebrafish and Caenorhabditis elegans (18). Non-avian studies evidenced relationships between immune responses and gut barrier function. As an example, interleukin (IL)-17 and IL-22 are tissue-signaling cytokines that favor protection and regeneration of cells in barrier organs such as the gastrointestinal tract (19). Cytokines could affect TJ structure and function, thus mediating paracellular permeability and exposure of tissues to luminal antigens in the gastrointestinal tract (12,20).
One of the additive groups that has been tested for their efficacy as alternatives to antibiotic growth promoters are probiotics (21). Supplementation of probiotics promotes gut health and performance by reducing the number of pathogens in the gastrointestinal tract through competitive exclusion or production of antimicrobial peptides, thus improving intestinal maturation and integrity and modifying immune responses (22)(23)(24)(25). Despite several reports on the effect of probiotics, less attention has been paid to the effect of multi-component additives or comparison of these additives with probiotics as alternatives to antibiotic growth promoters especially during NE challenge. In a recent study, supplementation of a multicomponent additive had no effect on NE lesions in the small intestine of broiler chickens, but improved FCR during the starter and grower periods (26).
Therefore, this study was conducted in order to evaluate the effects of a probiotic or a multi-component additive on performance, pathology, RNA expression of TJ proteins, immune responses, and cell energy metabolism-related genes, as well as ileal mucosa microbial profile during a naturally occurring NE challenge model.

Birds, Diets, and Management
This project was approved and conducted under the guidelines of the Virginia Tech Institutional Animal Care and Use Committee (IACUC # 18-136).
A total of 960 one-day-old Cobb 500 male broiler chickens were acquired from a local hatchery. Prior to placement, birds were weighed in groups of 30 each and allocated to one of 32 pens with dimensions of 1. The PROB consisted of Bacillus licheniformis spores (3.2 × 10 9 CFU/g) while the PPEO supplement contained B. licheniformis spores (3.2 × 10 9 CFU/g), 276 ppm mannan-oligosaccharides and β-glucans (1,3 and 1,6)/kg, and 8,000 ppm capsaicin (from chili peppers) and curcuma (from turmeric)/kg. Composition of the basal diet is detailed in Table 1. One large batch of the basal diet was mixed, and the amount needed for each treatment separated to which the appropriate amounts of supplements were added according to treatments. The mash diets were subsequently crumbled (starter period) or pelleted (grower and finisher periods). Each pen was equipped with a bucket-type feeder and a nipple drinker line with fresh wood shavings as litter (∼7.5 cm-deep). Birds had ad libitum access to water and feed from placement (d 0) until the end of the study (d 42). Lighting schedule

Mortality
Starting at placement, birds were monitored twice a day. For each dead bird, the date, body weight and cause of death were recorded. This procedure continued throughout the study (up to d 42) to record mortality/treatment for each period thus allowing for adjusting performance parameters against daily mortality.

Performance
Upon arrival (d 0), birds were weighed in groups of 30 and assigned to each pen. Subsequently, birds were weighed on d 8 (7 days after the coccidiosis challenge, which was also the peak mortality) and at the end of starter (d 14), grower (d 28), and finisher (d 42) periods. Additionally, feed consumption was also recorded on a per pen basis on d 8, 14, 28, and 42. Finally, adjusted average daily gain (ADG), average daily feed intake (ADFI) and feed conversion ratio (FCR) were calculated for each period (days 0-8, 9-14, 15-28, and 29-42) and also for the cumulative experimental period (d 0-42).

Necrotic Enteritis Lesion Scores
On d 8, three birds per pen were selected based on average body weight of the pen, euthanized by cervical dislocation, and the small intestines were examined for NE lesions and scored based on a 0-4 scale system. Each section of the small intestine, i.e., duodenum, jejunum, and ileum, was scored separately by personnel blinded to the treatments. The lesion scoring criteria used were as follows (27) Table 2. Each reaction was performed in a total volume of 20 µL in duplicate wells. Product specificity was confirmed by analysis of the melting curves generated by the 7500 software (version 2.0.3). mRNA abundance was analyzed using glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as an endogenous control. Average mRNA abundance relative to GAPDH for each sample was calculated using the 2 − Ct method (29) with the calibrator for each gene being the average Ct value from the negative control group.

16S rRNA Gene Sequencing
On d 8 and d 42, one bird/pen was euthanized, the ileum excised, mucosal scrapings collected, snap-frozen in liquid nitrogen and then stored at −80 • C until further analysis. Bacterial DNA was extracted from mucosal scrapings using the Qiagen QIAamp R PowerFecal R DNA kit (Qiagen GmbH, Hilden, Germany) based on the supplier's instructions. Quality and quantity of extracted DNA were measured using gel electrophoresis and spectrophotometer. All DNA samples were diluted to 300 ng/µl and 50 µl of each sample were sent to the Virginia Tech Biocomplexity Institute for sequencing. The V4 region of the 16S rRNA gene was amplified using the Illumina MiSeq kit V3 (600-cycle format, 25 million reads, read length 2 × 300). Earth Microbiome Project primer sets 515F (30) and 806R (31) generating 390 bp amplicons were used. Data were analyzed using QIIME 2 software version 2019.10 (32). The reads were denoised using the DADA2 pipeline. The sequences flagged as non-chimeras were retained, sequence reads with >99% identity were clustered into a single operational taxonomic unit (OTU) and SILVA database version 132 was used to assign 16S rRNA sequence reads to OTU. To remove sequencing effort heterogeneity, samples for d 8 and d 42 were rarefied to 10,891 and 29,818 sequences, respectively. These were the lowest sequence reads within each sampling day.
Alpha diversity was measured using the Shannon index (H), which accounts for both abundance (richness) and evenness of the species present. To estimate the similarity of microbial community structure between groups (β-diversity), principal coordinate analysis (PCoA) based on weighted UniFrac distance (bray-curtis) were performed. To assess the association between microbial community and dietary treatment within each sampling day, pairwise PERMANOVA analysis implemented in QIIME2 was performed on an unweighted UniFrac distance matrix of 72 samples. The significance of PERMANOVA was obtained by 999 permutation tests.
Relative abundance of taxa for each sampling day and treatment were calculated at the phylum and genus levels.
The distinctive taxa between treatment groups were identified with linear discriminant analysis effect size (LEfSe) which was performed through Galaxy/Hutlab workflow platform (33). In order to predict the function of ileal mucosa microbiota, data analysis was performed through the Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt) pipeline (34). Then, PICRUST output for the level 3 of the Kyoto Encyclopedia of Genes and Genomes (KEGG) were analyzed and illustrated with statistical analysis of the taxonomic and functional profiles (STAMP) software version 2.1.3 (35).

Statistical Analyses
Statistical analysis for all data (except 16s sequencing) was performed using the ANOVA procedure of JMP software (2013) For each gene, the primer sequence for forward (F) and reverse (R) (5 ′ -3 ′ ) primers, the amplicon size (bp) and the NCBI Accession number (Acc) used for the primer design are listed. a Secretory immunoglobulin A. b Peroxisome proliferator-activated receptor-γ coactivator 1α. c Mechanistic target of rapamycin. d AMP-activated protein kinase-α1. e Glyceraldehyde 3-phosphate dehydrogenase.
and significance between treatments (P < 0.05) determined by the LSD test. The statistical model for data analysis is outlined below: Yij = µ + Ai + eij Yij = measured value for each observation (data) µ = grand mean Ai = treatment effect eij = experimental error.

Mortality
Mortality was significantly (P = 0.011) reduced in PROB, PPEO and PC groups during the d 0-14 period compared to NC ( Table 3). There were no significant differences in mortality between treatments during the overall (d 0-42) experimental period. However, compared to NC, PC and PPEO reduced d 0-42 mortality by 38 and 33%, respectively (although not statistically different likely due to variability among replicate pens).

Necrotic Enteritis Lesion Scores
Typical with the challenge model applied, lesion scores were mostly prevalent in the duodenum portion and less pronounced in the distal sections of the small intestine ( Table 4). Dietary supplementation of PROB and PPEO significantly (P = 0.013) reduced lesion scores in the duodenum compared to the NC birds. However, there was no difference in NE lesion scores for jejunum and ileum among the treatments.

Performance
All performance results are presented in    (Figure 1). However, abundance of CLDN-1, OCLN, ZO-1, and ZO-2 were not affected by any treatment. All treatment groups reduced (P = 0.034) mRNA abundance of IFN-γ compared to NC (Figure 2) while mTOR abundance was only higher (P = 0.015) for PPEO compared to NC on d 14 (Figure 3).

Diversity and Composition of Microbiota
There was no difference between treatments with regard to alpha diversity on d 8; however, alpha diversity was different between NC and PPEO (P = 0.082) on d 42 (Supplementary Table 1; Supplementary Figures 1A, 2A). β-diversity PCoA plots showed a trend of separation of microbial communities among some groups on d 8 and d 42 (Supplementary Figures 1B, 2B). PERMANOVA analyses on d 8 revealed that the composition of microbial communities in PPEO was different (P < 0.10) with NC and PC groups (Supplementary Table 2). On d 42, the composition of microbial communities in PC was different (P < 0.10) from PPEO and PROB groups (Supplementary Table 2).
The relative abundance of amplicon sequence variants of ileal scrapings microbiota was analyzed at different ranking levels from phylum to genus. The dominant phyla across the groups were Firmicutes, Bacteroidetes, Proteobacteria, Actinobacteria, Tenericutes, and Epsilonbacteraeota together contributing more than 99 and 82% of the whole phyla on d 8 and d 42, respectively ( Table 6). Birds supplemented with additives had a higher abundance of Firmicutes and a lower abundance of Bacteroidetes compared to NC and PC on d 8. However, differences in relative abundance of the dominant bacterial phyla among the treatments decreased with age ( Table 6).
On d 8, relative abundance of Rominiclostridium 9, Oscillibacter, RuminococcaceaeUCG_014, ASF356, Clostridium sensu stricto 1, Prevotellacea, and Muribaculaceae was higher in NC compared to PROB and PPEO groups (Figures 4A,C). On the other hand, relative abundance of Lactobacillus was lower in the NC birds compared to the PROB and PPEO treatments. On d 42, relative abundance of Bacteroides, Clostridiales vadin BB60 group, Prevotellaceae NK3B31 group, Desulfovibrio, Butyricicoccus, Helicobacter, Sellimonas, and Filimonas was higher in NC compared to PROB and PPEO groups (Figures 4B,D). Interestingly, relative abundance of Lactobacillus was significantly lower (Figures 5A,C) while relative abundance of Escherichia-Shigella was numerically  higher (data not shown) in PC compared to PPEO and PROB groups on d 8. Furthermore, relative abundance of several bacterial genera and families were different in PC compared to PROB and PPEO groups on d 42 (Figures 5B,D).

Predicted Functions of Ileal Mucosal Microbiota
Prediction of microbial functions showed distinctive results among the treatment groups. There were 53 and 26 pathways at KEGG level 3 with distinctive enrichment between NC and PPEO groups on d 8 and d 42, respectively (Figures 6A,B). There were 66 and 14 pathways at KEGG level 3 with distinctive enrichment between NC and PROB groups on d 8 and d 42, respectively (Figures 7A,B). Among these, the NC microbiota had higher numbers of functional genes involved in lipopolysaccharide (LPS) biosynthesis, LPS biosynthesis proteins, DNA replication proteins, membrane and intracellular structural molecules, protein kinases, oxidative phosphorylation, and bacterial secretion system compared to PROB and PPEO. Interestingly, functional genes involved in both butanoate metabolism and propanoate metabolism were enriched in PROB compared to NC, while PPEO only enriched functional genes involved in propanoate metabolism compared to NC on d 8. There were 47 and 14 pathways at KEGG level 3 with distinctive enrichment between the PC and PPEO groups on d 8 and d 42, respectively (Figures 8A,B), while 50 and 3 pathways with distinctive enrichment were identified between the PC and PROB groups on d 8 and d 42, respectively (Figures 9A,B). Among these, the PROB group had higher numbers of functional genes involved in lysine biosynthesis, butanoate metabolism, retinol metabolism, glycolysis/gluconeogenesis and synthesis and degradation of ketone bodies; while LPS biosynthesis, LPS biosynthesis proteins, bacterial secretion system, lipid biosynthesis proteins, and carbohydrate digestion and absorption were enriched in PC compared to PPEO and PROB groups on d 8. On d 42, lysosome, proteasome, peroxisome, antigen processing and presentation were enriched in PC compared to PPEO; however, ribosome biogenesis and glycerophospholipid metabolism were enriched in PPEO.

DISCUSSION
Necrotic enteritis occurs in clinical (acute) or subclinical (chronic) form (2,3). During subclinical NE, chronic damage to the intestinal mucosa leads to poor performance (reduced FI, weight gain and eventually increased FCR) with or without low mortality (4,5). The NE challenge in this study was subclinical as manifested by low mortality and lesion scores in all the treatments. Better performance in PROB and PPEO groups may be due to the modulated immune responses (lower mRNA abundance of IFN-γ, sIgA, IL-10, and IL-17), better gut health (lower lesion scores and higher mRNA abundance of CLDN-3), and modified energy and protein metabolism (higher abundance of PGC-1α and mTOR). Immune responses are energy demanding processes that divert nutrients from growth,  subsequently reducing performance. Therefore, in PROB and PPEO birds, nutrients could have been directed toward growth rather than mounted immune responses in NC birds. Previously, B. licheniformis spores reduced NE mortality and lesion score, and improved FCR compared to the C. perfringens-challenged non-medicated birds (36).
Furthermore, supplementation of B. subtilis to the diet of broiler chickens challenged with NE significantly decreased lesion scores and improved FCR compared to challenged control birds (37)(38)(39). Gut health plays a pivotal role in the chickens' ability to reach their genetic potential and low FCR, while diminished   feed efficiency and profitability would be the result of enteric diseases such as NE (4,5). In this study, birds in the PPEO and PROB groups had significantly lower NE lesion scores in the duodenum compared to NC. Lower lesion scores reflect a stronger gut barrier against C. perfringens and may explain the lower mortality rates in those birds compared to NC birds during the early (d 0-14) challenge period (5% compared to 1.66 and 0.8%, respectively). An intact intestinal epithelium prevents entry of potential pathogens and leads to optimal health and performance of birds as a result of proper nutrient absorption and utilization (40). Tight junction proteins are critical components of gut integrity and make up a barrier in the paracellular space. However, these barriers have a dynamic structure that remodel in response to signals from the cells or external stimuli such as nutrients, and pathogenic and commensal bacteria (13). Furthermore, TJ proteins are binding sites for C. perfringens enterotoxins (CPE) (41), and understanding their changes during NE is a promising aspect of research. In this study, supplementation of PPEO led to higher mRNA abundance of CLDN-3 on d 8 and d 14 compared to NC, while CLDN-3 abundance was higher in PROB fed birds only on d 8. Particularly, the CLDN family proteins are structural components of TJ proteins and characterized by their extracellular domain to which CPE attach and influence gut integrity (15,41). Higher abundance of CLDN-3 in PROB and PPEO groups coincided with lower lesion scores and better FCR in these groups. Therefore, one explanation would be that these products alleviated the negative effects of NE, thus the higher mRNA abundance of CLDN-3. In other words, lower abundance of CLDN-3 in the NC birds might be a defense mechanism to reduce the attachment sites for CPE in order to reduce its negative effects. We have previously shown the possibility of TJ modification in chickens through supplementation of probiotics under a field-like NE challenge (42). Curcumin (a component of PPEO) can increase the mRNA abundance of ZO-1 and CLDN-1 in Caco-2 cells and alleviate IL-1β-induced disorganization of ZO-1, CLDN-1, and CLDN-7 in intestinal epithelial cells, leading to improved barrier function and reduced paracellular permeability (43,44). This might be a reason for the effect of PPEO on CLDN-3 mRNA abundance on d 14, while there was no effect of PROB alone.
Higher lesion scores in the NC birds was accompanied by higher relative abundance of Rominiclostridium 9, Oscillibacter, RuminococcaceaeUCG_014, ASF356, Clostridium sensu stricto 1, Prevotellacea, and Muribaculaceae compared to PROB and PPEO on d 8. Conversely, relative abundance of Lactobacillus was lower in NC compared to other treatments except PC. Prevotellaceae can degrade mucus oligosaccharides resulting in the disruption of intestinal mucosal barrier and intestinal inflammation (45). In addition, higher proportion of Prevotellaceae has been detected in the microbiota of NLRP6 KO mice with high risk for colitis (46). Higher relative abundance of Prevotellaceae in the NC group might have led to the disruption of the mucosal barrier rendering it prone to bacterial pathogens, which coincided with higher FCR and lesion scores in this group. Several studies reported the suppression of Lactobacillus at the time of C. perfringens challenge (6,24,47). Certain pathogenic bacteria have the ability to attach to the intestinal mucosa; however, probiotic bacteria such as Lactobacillus and Bifidobacterium prevent adherence of pathogens to the intestinal mucosa via competitive exclusion thus displacing pathogenic bacteria (48,49). Accordingly, higher relative abundance of Lactobacillus might have led to lower NE lesions in the jejunum of broilers in the PROB and PPEO groups. A mixed C. perfringens and Eimeria challenge led to a significant increase in Clostridium sensu stricto 1 and reduction in Lactobacillus with the concurrent increase in NE lesions (50); the authors concluded that overgrowth of C. sensu stricto 1 is associated with NE. As reported earlier, supplementation of feed additives in this study reduced relative abundance of C. sensu stricto 1 in the ileal mucosal scrapings of broiler chickens, which might be a reason for better gut health and lower NE lesions in these groups. C. perfringens infection decreased (numerically) the relative abundance of the Peptostreptococcaceae family, and increased that of the Enterobacteriaceae family, mainly composed of Escherichia-Shigella, in the ileum of broiler chickens (24). This group of bacteria is mainly composed of opportunistic pathogens and higher presence might be an indicator of disrupted gut health that allows the overgrowth of such pathogens. Further, C. perfringens and its co-infection with Eimeria increased relative abundance of Escherichia-Shigella in the jejunum of broiler chickens (50). In addition, other reports indicated a positive correlation between the relative abundance of Escherichia-Shigella with NE occurrence in broiler chickens (51). In this study, the relative abundance of Lactobacillus was significantly lower, while relative abundance of Escherichia-Shigella was numerically higher in the PC birds compared to other groups. Supplementation of virginiamycin in combination with monensin to broiler chickens' diet increased E. coli, Coprococcus, and Anaeroflum, while Roseburia, Lactobacillus, and Enterococcus decreased in the ceca (52). Previous study showed that a probiotic (L. acidophilus) improved intestinal health by numerically decreasing the relative abundance of Escherichia-Shigella in the ileum (24).
Lactobacillus are among the predominant bacterial genera in the gastrointestinal tract of broiler chickens (53), and provide both direct and indirect beneficial effects. The direct effects include immunomodulation via attachment and interaction with enterocytes, antagonistic activity against pathogens by  production of lactate thus lowering pH and making the gastrointestinal tract environment unsuitable for acid-sensitive pathogens, and competitive exclusion mechanisms along with the production of bacteriostatic and bactericidal substances (54)(55)(56). Supplementation of PROB and PPEO increased the relative abundance of Lactobacillus in the ileal mucosa, thereby possibly contributing to their positive effect on gut health and performance of the birds. Likewise, dietary supplementation of a probiotic to broiler chickens challenged with NE increased abundance of Lactobacilliaceae and Clostridiaceae families in the cecal digesta compared with the challenged control group (39).
Supplementation of PROB enriched the predicted metabolism of butanoate and propanoate in the ileal bacteria compared to NC on d 8. This might be due to increased relative abundance of Firmicutes in PROB compared to NC. Butyrate producers mostly belong to the Firmicutes phylum (57). In addition, Bifidobacterium and Lactobacillus are the substrate providers for butyrate producers (58). Optimal butyrate production relies on the presence of butyrate-producing bacteria and various others including lactate-producing bacteria that cross-feed butyrate producers (59,60). Butyrate could enhance epithelial regeneration by stimulating villus growth; however, it does not inhibit C. perfringens (61,62). Absorption of butyrate and propionate by chicken cecal mucosa could improve host energy metabolism and improve performance (63). Similar to this study, supplementation of B. licheniformis enriched butanoate metabolism in the microbiota of broiler chickens challenged with NE compared to challenged non-supplemented group (64). Further, dietary supplementation of a prebiotic (but not probiotic and symbiotic) to laying hens enriched cecal microbial genes involved in butanoate and propanoate metabolism (63).
The essential oils thyme and anise decreased C. perfringens and E. coli counts in the small intestines, along with decreased intestinal lesion scores and significantly better FCR (65). Essential oils are constituents of the PPEO supplement used in this study and we observed lower relative abundance of Escherichia-Shigella in the ileum of birds in this treatment group. The probiotic B. subtilis DSM 32315 increased the abundance of Firmicutes while reduced the abundance of Bacteroidetes in the ceca of broiler chickens (66). Furthermore, this probiotic decreased the abundance of potentially harmful bacteria such as Vampirovibrio, Escherichia-Shigella, and Parabacteroides (66). Likewise, supplementation of PROB to the diet led to increased Firmicutes abundance and reduced Bacteroidetes abundance in the ileal mucosa of broiler chickens.
Mannooligosaccharides constitute one component of PPEO. In vitro studies showed high levels of propionate production from mannooligosaccharides fermentation (67), and production of short chain fatty acids reduced cecum pH, thus inhibiting the growth of Enterobacteriaceae (which includes Escherichia-Shigella) (68). Lower relative abundance of Escherichia-shigella in the PPEO and PROB groups compared to PC was associated with the reduction of LPS biosynthesis, LPS biosynthesis proteins, and bacterial secretion system in PROB and PPEO groups compared to NC and PC as indicated by KEGG pathway. LPS induce mucosal immune responses (69); thus, such responses to the overgrowth of opportunistic bacteria due to the dysbiosis caused by NE challenge might have been attenuated by PROB and PPEO supplementation.
Supplementation of PPEO but not PROB significantly increased mRNA abundance of PGC-1α and mTOR on d 14 compared to NC. Mitochondrial function is regulated by PGC-1α. PGC-1α modulates oxidative metabolism by increasing mitochondrial function and at the same time minimizing the build-up of oxidative respiration by-products such as reactive oxygen species (70). mTOR regulates protein and lipid synthesis, thus its signaling is of significant/central importance in regulating cell metabolism, growth, proliferation and survival (71). Renewal of intestinal epithelial cells along the cryptvillus axis is regulated by the mTOR signaling pathway. This may partly be due to its ability to affect the antioxidant capacity and protein synthesis of intestinal epithelial cells (9,10). Metabolism is controlled by sensing the internal (host) and external (presence of bacteria and nutrients in the gut lumen) signals that are integrated by mTOR and PGC-1α (71). These signals might lead to activation and proliferation of immune cells (such as T cells) leading to retarded growth as a consequence of a strong immune response (72). Therefore, mounted immune responses in the NC birds might be the cause of reduced ADG and increased FCR in this group. During injury or infection, archetypal proinflammatory cytokines (IL-1 and TNF-α) are rapidly released. Pro-inflammatory cytokines, including IL-1β and IFN-γ, could disrupt intestinal barrier function by affecting TJ structure (20,43,73). Here, IFN-y mRNA abundance was lower in PPEO and PROB compared to the NC group, demonstrating the protective effect of these supplements in reducing the potential damage caused by pro-inflammatory cytokines to the intestinal barrier.

CONCLUSION
Based on the presented findings, we can conclude that under a naturally occurring subclinical NE challenge model, supplementation of a PROB or PPEO to broiler diets could reduce d 8 lesion scores and decrease FCR during the cumulative (d 0-42) grow-out period. Additionally, birds fed diets supplemented with PPEO had significantly higher ADG during d 0-42. These changes might be due to higher mRNA abundance of TJ proteins, PGC-1α and mTOR, and modulated cytokine responses in these groups. Ileal mucosa microbial profiling and functional analysis revealed taxonomic and functional signatures associated with dietary supplementation of PPEO or PROB compared to NC and PC birds. Overall, compared to probiotic alone, concurrent supplementation of probiotic/prebiotic/essential oils resulted in better cumulative performance under this naturally occurring NE challenge model.