Impact of commercial gut health interventions on caecal metagenome and broiler performance

22

23 Maintaining gut health is a persistent and unresolved challenge in the poultry industry. Given the 24 critical role of gut health in chicken performance and welfare, there is a pressing need to identify 25 effective gut health intervention (GHI) strategies to ensure optimal outcomes in poultry farming. In 26 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint this study, across three broiler production cycles, we compared the metagenomes and performance 27 of broilers provided with ionophores as control against birds subjected to five different GHI 28 combinations involving vaccination, probiotics, prebiotics, essential oils, and reduction of ionophore 29 use. 30

31 Using a binning strategy, 84 (≥75% completeness, ≤ 5% contamination) metagenome-assembled 32 genomes (MAGs) from 118 caecal samples were recovered and annotated for their metabolic 33 potential. The majority of these (n = 52, 61%) had a differential response across all cohorts and are 34 associated with the performance parameter - European Poultry Efficiency Factor (EPEF). The control 35 group exhibited the highest EPEF, followed closely by the cohort where probiotics are used in 36 conjunction with vaccination. The use of Probiotics B, a commercial Bacillus strain-based 37 formulation, was determined to contribute to the superior performance of birds. GHI 38 supplementation generally affected abundance of microbial enzymes relating to carbohydrate and 39 protein digestion, and metabolic pathways relating to energy, nucleotide synthesis, short-chain fatty 40 acid synthesis, and drug-transport systems. These shifts are hypothesised to differentiate 41 performance among groups and cycles, highlighting the beneficial role of several bacteria including 42 Rikenella microfusus and UBA7160 species. 43

44 All GHIs are shown to be effective methods for gut microbial modulation, with varying influences on 45 MAG diversity, composition and microbial functions. These metagenomic insights greatly enhance 46 our understanding of microbiota-related metabolic pathways, enabling us to devise strategies 47 against enteric pathogens related to poultry products and presenting new opportunities to improve 48 overall poultry performance and health. 49

Chicken gut microbiome, metagenomics, gut health, broiler performance, shotgun 50 sequencing, probiotics 51 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint

52 The gastrointestinal tract of chickens harbours a complex and dynamic microbial community 53 collectively known as the gut microbiota. This microbiota, along with its corresponding genetic 54 material, forms the gut microbiome which is recognised for its significance in both health and 55 metabolism in its host [1]. The majority of the microbiome consists of a diverse set of bacteria which 56 can be classified as either commensal, pathogenic or beneficial to the host; of which their occurrence 57 and interactions can be influenced by a variety of factors, such as genetics, age, environment, diet, 58 and administration of feed additives [1–4]. Interestingly, these same factors have also been 59 established to directly impact overall health and performance of the chicken host [5–7], indicating a 60 possible link between gut microbiota composition and broiler performance. Indeed, one study 61 identified 24 bacterial species to be differentially abundant between broilers with high and low feed 62 conversion ratios (FCR) [8]. Furthermore, in our previous work, we demonstrated that extrinsic 63 parameters including stocking density, percentage of protein and energy in diet, and Omega -3 64 supplementation are able to modulate key microbiome members involved in nutrition and 65 metabolism, subsequently affecting growth and feed efficiency in the host [9]. 66 67 Due to this accumulating evidence supporting the importance of gut health in poultry performance , 68 there has been a significant rise in interest on modulation of gut microbiota for improved animal 69 health, productivity, and food safety [10]. Historically, growth promoters have utilised for enhanced 70 feed efficiency, while also decreasing illness and death rates from both overt and hidden diseases ; 71 hence, [11]. These drugs are purported to achieve these benefits by altering the gut microbiota, 72 resulting in decreased nutrient utilisation by microbes, increased absorption of nutrients through 73 thinned host gut walls, and reduction in inflammatory stress [11–13]. Ionophores are growth 74 promoters that have been safely used as the approved and standard intervention to maintain gut 75 health in poultry in the last decades, especially for the effective prevention of economically important 76 diseases such as coccidiosis and necrotic enteritis [14–16]. Currently, a myriad of new gut health 77 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint interventions (GHIs) with similar effects are being considered as supplements to further improve 78 overall bird health and performance. These include, but are not limited to prebiotics, probiotics, 79 phytogenic substances, organic acids, essential oils, and enzymes [17]. Each of these GHIs has their 80 own mechanism of action and corresponding effects on chickens, with different subtypes within each 81 group. Descriptions of these GHIs have been detailed previously in recent reviews [18,19]. Options are 82 further expanded by combinatory use of multiple GHIs throughout a single production cycle for their 83 potential synergistic effects [20,21]. However, there is limited information on their effects on the gut 84 microbiome in poultry. 85 86 Research in order to investigate the influence of the GHIs has mainly involved the use of 87 metataxonomic sequencing of the microbiota through amplification of the 16S rRNA gene marker [22]. 88 Use of this method has proved ground -breaking for our understanding of gut health; however, taxa 89 identification of less abundant and unknown species as well as the characterising of their metabolic 90 capacity (functional profiles) remains challenging [22–24]. These limitations can hinder our 91 comprehension of their connection to broiler health and performance . In contrast, with shotgun 92 sequencing (which involves indiscriminate sequencing of all random DNA segments within a sample), 93 a higher resolution of microbial genomes enables these features to be identified, allowing a deeper 94 understanding of relevant metabolic functions of bacteria [24]. For instance, Chen et al., (2023) [25] 95 used metagenomic methods in order to understand the role of gut microbiota in fat regulation in 96 chickens. From this, they were able to observe the presence of differential carbohydrate active 97 enzymes (CAZymes) and functional metabolic modules between high and low abdominal fat chickens. 98 Applying this concept, an exploration of the metagenome of chickens given various GHIs will help us 99 understand the gut microbial functions of chickens that promote an improved health and performance 100 of broilers. Therefore, the aim of this study is to characterise how GHIs impact the gut microbiome in 101 relation to performance through shotgun metagenomic sequencing, in comparison to the standard 102 use of ionophores. 103 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint

104 Ethics statement 105 All animal trials were reviewed and conducted in accordance with the Animal Welfare and Ethical 106 Review Board of London School of Hygiene and Tropical Medicine, UK (Reference 2023-05). Poultry 107 farm management and industry plant processing activities were conducted following Moy Park Ltd 108 standard operating protocols which are compliant with UK animal handling laws and regulations 109 (Craigavon, UK) (Moy Park Ltd, 2022, 2023) [26 – 27]. As part of standard commercial practices of 110 the company, all birds were subjected to stunning before slaughter and subsequent carcass 111 processing [28]. 112 113 Experimental design and sampling 114 Three broiler production cycles (C0, C1, C2) were implemented in a Moy Park Ltd affiliated 115 commercial farm in Northern Ireland between June and October 2022. In each cycle, a total of 116 18,000 Ross-308 mix-sexed broilers were raised in an automated commercial poultry house and 117 provided a four-stage standard commercial diet regimen based on Aviagen specifications for Ross 118 broilers [29]. This was composed of a starter diet (S, 0-11 days), grower diet (G, 11-23 days), finisher 119 diet (F, 23-31 days) and a withdrawal diet (W, 32 days until clearing at day 40). Birds were 120 distributed into 6 groups (1 control, 5 GHI groups), wherein each group is allocated into 6 pens 121 containing 500 birds each. For our control group, we administered ionophores which is a safe and 122 legally accepted method for control of coccidiosis [15,16]. For our treatment groups, we adapted 123 different gut health strategies which involves combination of GHIs; this was designed to optimise 124 and maximise the differences on gut health and performance as based on the study of Granstad et 125 al., (2020). GHI treatment groups for C0 and C1 includes: T2 – Coccidiosis vaccine (V); T3 – V + 126 Bacillus strain probiotics A (PA), T4 – V + PA + reduced crude protein (-1%) in G/F/W diets; T5 – V + 127 PA + Essential oil and T6 – V + PA + and ionophore in F diet. Birds in C2 had similar treatment design 128 with two differences: a different GHI – Bacillus strain Probiotics B (PB) was utilised instead of PA in 129 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint T2 to T5, and essential oils in T5 was replaced by prebiotics. Birds were provided ad libitum access to 130 feed and water. 131 132 At the end of the production cycle (Day 40), caecal samples were collected within 5-10 minutes of 133 the slaughtering process in a Moy Park Ltd industry plant. Intact caeca (N = 120) from C1 (n = 60) and 134 C2 (n = 60) were obtained through aseptic incision from the rest of the GIT and were then 135 transferred into sterile 50ml tubes and stored in a polystyrene container with frozen icepacks. All 136 specimens were immediately sent to the laboratory for storage at -80°C until further use for DNA 137 extraction (Illustrated in Figure 1). 138 139 140 Figure 1. Overview of study design. C: Production cycle, S: Starter, G: Grower, F: Finisher, W: Withdrawal, 141 PA: Probiotic A, PB: Probiotic B, IP: Ionophore, V: Coccidiosis vaccine 142 3 Broiler Production Cycles 36 pens per cycle (6 pens per treatment group) 1 pen = 500 40 days PERFORMANCE PARAMETERS (N = 108 pens) House clearing (at poultry factory) 10 caecal samples per treatment group for C1 (n = 60) and C2 (n = 60) SHOTGUN METAGENOMIC SEQUENCING (N = 120 DNA samples) AGE GROWTH STUDY DESIGN .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint 143 DNA extraction, library preparation, and shotgun metagenomic sequencing 144 Microbial DNA was extracted from caecal chyme, using the QIAamp® PowerFecal Pro DNA Kit 145 (Qiagen, Germany) according to manufacturer’s instructions. Total DNA was eluted in 50 µl of 146 elution buffer and stored at -80°C. Initial DNA concentration was measured using Nanodrop ND-147 1000 (NanoDrop Technologies, Inc., Wilmington, US). 148 Sequencing libraries were generated using a modified Illumina DNA Prep tagmentation approach 149 (Illumina, Inc., Cambridge, UK) described previously [30]. Tagmentation was performed as follows: a 150 master mix composed of 0.5 µl bead linked transposomes 0.5 µl tagmentation buffer and 4 µl 151 nuclease free distilled water was created for each sample (2 µl), placed in a 96-well plate, and run on 152 a thermocycler at 55°C for 15 mins. Another PCR master mix using the Kapa2G Fast Hot Start PCR kit 153 (Sigma-Aldrich, Gillingham, UK) was then generated and transferred into a 96-well plate, to which 5 154 µl of P7 and P7 of Nextera XT Index Kit v2 index primers (Illumina, Cambridge, UK) and 7 µl of the 155 previous tagmentation reaction were added. The plate was run on the thermocycler with conditions: 156 72°C for 3 mins, 95°C for 1 min, 14 cycles of 95°C for 10 secs, 55°C for 20 secs and 72°C for 3 mins. 157 Quality control of multiplex barcoding was performed on a D5000 ScreenTape using the Agilent 158 Tapestation 4200 (Agilent, Wardbronn, Germany). Next, barcoded libraries were quantified on a 159 Qubit 3.0 instrument (Invitrogen, Paisley, UK), pooled in equivalent concentrations in a tube, and 160 washed with 0.5-0.7X solid phase reversible immobilisation KAPA Pure Beads (Roche, Wilmington, 161 US). In order to calculate final pool molarity, the pooled library was quantified using Qubit 3.0 and 162 on a D5000 High Sensitive ScreenTape. After library qualification, the library was sequenced using 163 the NovaSeq 6000 System, Paired-end 150 bp (Illumina, Cambridge, UK). 164 165 Bird performance and health monitoring 166 The performance parameters included bird weight (BW, kg of body weight), average daily gain (ADG, 167 grams feed per day), corrected food conversion ratio (FCR) at 2kg BW, and total mortality (MT). 168 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint These measurements were taken as mean average per pen at clearing day (40d) which were 169 conducted in line with typical industrial practices. Contact dermatitis measures which included 170 footpad dermatitis (FPD) lesion scores (FPDS), FPD prevalence (FPDP), and hockburn (HB) lesion 171 scores (HBS) and prevalence (HBP) were also taken before slaughter as conducted previously [5]. To 172 estimate overall performance, European Production Efficiency Factor (EPEF) was calculated based on 173 preliminary recorded performance measures [31]. 174 175 Bioinformatic analysis 176 Recovery of Metagenomic-Assembled Genomes (MAGs) 177 A total of 120 metagenomic samples were processed – from which adapter trimmed reads were 178 generated by the sequencing centre. Reads were subjected to quality trimming using Sickle v1.200 179 [32]. Trimming involved removing reads where the average phred below 20 and retaining paired end 180 reads with a post-trimming length exceeding 50 bp. Two samples (one from T3 and one from T4 in 181 C2) were excluded due to non-recovery of reads, resulting in a total of 118 samples which generated 182 2,588,938,595 reads. Forward and reverse reads were then aggregated and subjected to collective 183 assembly using MEGAHIT [33]. Assembly parameters used were --k-list 27,47,67,87 --kmin-1pass -m 184 0.95 --min-contig-len 1000. This gave us a total of 1,276,325 contigs, a total of 3,101,580,806 bases 185 (bp), maximum of 403,439 bp, average length of 2,430 bp, and an N50 score of 2,724 bp. Assemblies 186 were then subjected to binning via the MetaWRAP pipeline [34], wherein three algorithms namely 187 metabat2 [35], MaxBin [36], and CONCOCT [37] were utilised. Bins from each of the algorithms were 188 consolidated using the MetaWRAP framework, resulting in a total of 308 bins. For estimation of 189 completion (COM) and contamination (CON) metrics of each MAG, CheckM was used on all bins 190 [38]. We retained bins with more than 75% and less than 5% contamination to give a final set of 84 191 MAGs. The summary statistics of these MAGs are provided in Supplement file 1. 192 Taxonomic and functional annotation 193 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint For metabolic function and taxonomic assessment of each MAG, the METABOLIC pipeline was 194 employed [39]. Within its framework, taxonomic classification of bins was incorporated using GTDB-195 TK [40], whilst functional annotations were recovered using Kyoto Encyclopedia of Genes and 196 Genomes (KEGG) for metabolic function modules and submodules [41], dbCAN2 for carbohydrate 197 active enzymes (CAZymes) [42], custom hidden Markov model databases for nutrient cycles [43] and 198 MEROPS for proteases [44]. To obtain taxonomic and functional coverages per sample, read 199 coverages (proportion of each bin per sample) were multiplied with each feature coverages 200 (returned from METABOLIC). From this, we derived the sample-wise abundance tables: dbCAn2 (n = 201 118 samples x 117 CAZyme IDs), KEGG Modules (n = 118 samples x 251 module IDs), KEGG 202 Submodules (n = 118 samples x 964 submodule IDs) and MEROPS (n = 118 samples x 108 203 peptidases). 204 205 Phylogenetic tree generation 206 To construct a phylogenetic tree of MAGs, we used GToTree [45] that involves detection of Single 207 Copy Genes (SCGs) in MAGs and multisequence alignment. Specifically, we used the bacteria and 208 archaea HMM set which covers 25 SCGs. MAGs that had very few hits for these SCG were removed, 209 resulting in a phylogeny recovery for a total of 65 MAGs. For assessment of novelty of MAGs, the 210 Genome Tree Toolkit was utilised [46], wherein phylogenetic gain (PG) for each MAG against other 211 MAGs in the tree was estimated. 212 213 Statistical analyses 214 All tests were performed in R [47]. The normality of data was assessed using Shapiro-Wilk test [48]. 215 To determine significant differences between treatment groups, we employed analysis of variance 216 (ANOVA) and pairwise t-test with Bonferroni correction for normally distributed data, while Kruskal-217 Wallis testing and posthoc Dunn testing with Bonferroni correction (p < 0.05) for non-normal 218 distributed data [47]. For performance data, when significant p-values (p < 0.05) were obtained in 219 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint ANOVA, statistical groupings were evaluated using Duncan Multiple Range Test (DMRT) through the 220 agricolae package [49]. 221 222 To evaluate the individual effects of GHIs across the three cycles, we performed a generalised linear 223 model analysis via the penalised maximum likelihood method using the glmnet package [50]. 224 Specifically, we used the least absolute shrinkage and selection operator (LASSO) or L1 penalty 225 model by setting the regularisation parameter or alpha = 1 and cross validation folds (nfolds) = 10. 226 This generalised model was used to prevent overfitting of models due to multicollinearity and 227 sparsity of covariates (such as vaccination and ionophore use in our study) and to determine which 228 covariate fits best as a predictor of the outcome of interest. The model is represented as: 229 𝑚𝑖𝑛𝑖𝑚𝑖𝑠𝑒 &∑ (𝑦! − 𝛽" − ∑ 𝑥!# 𝛽# $ #%& - '( !%& + 𝜆 ∑ 0𝛽#0$ #%& 1 , where: 230 𝑁: the number of observations, 𝑝: the number of predictors, 𝑦!: the outcome variable for the 𝑖-th 231 observation, 𝑥!#: the value of the 𝑗 -th predictor for the 𝑖-th observation, 𝛽": intercept term, 𝛽#: are 232 the coefficients for the parameter, and 𝜆: the regularisation parameter controlling the strength of 233 the penalty term [51]. This model indicates that the penalty term forces some of the beta 234 coefficients to go to zero when their corresponding predictors are not significant. For our model, we 235 included the following as predictors: Ionophore: All stages (as “Yes”) or “No” , Ionophore: Finisher 236 (as “Yes”) only or “No” , Vaccination: “Yes” or “No” , Probiotic A: “Yes” or “No”, Probiotic B: “Yes” or 237 “No”, Essential oil: “Yes” or “No”, Reduced crude protein (-1%): “Yes” or “No”, Prebiotic: “Yes” or 238 “No”, wherein “No” was used as reference for all covariates. For outcome variables, we used the 239 parameters: 40d ADG, 40d BW, EPEF, corrected 2kg FCR (2kgFCR), FPDP (%), HBP (%), and total MT 240 (%). 241 242 For microbial diversity assessment, different functions of the vegan package [52] were employed. 243 For alpha diversity, we estimated richness (R) (using the rarefy function), Shannon entropy (H) and 244 Simpson (Si) (using the diversity function), Fisher alpha (FA) (using the fisher.alpha function), and 245 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint Pielou’s evenness (PE) (using the specnumber (S) function for formula: PE = ) *+, (/)). After confirming 246 their normal distribution, ANOVA was then employed to determine significance differences between 247 treatment groups [47]. For beta diversity, we employed Bray Curtis dissimilarity index analysis using 248 the vegdist function of the vegan package, followed by principal component analysis using the R 249 base function cmdscale [47,52]. The separation between groups was tested with permutational 250 analysis of variance (PERMANOVA) through the vegan command adonis2 [52]. 251 To find the relationship between the individual MAGs and each treatment group (C1 and C2 done 252 separately) as well as the relationship between individual MAGs and performance parameters – 253 EPEF and MT, we employed Generalised Linear Latent Variable Model (GLLVM) regression analysis as 254 described previously [53]. This model is an extension of the basic generalised linear model wherein 255 the mean abundances (for i-th sample and j-th MAG) are regressed against the covariates (T, 256 treatment groups or performance/health trait) by incorporating latent variables 𝒖𝒊 as: 257 𝑔9𝜇!#; = 𝜂!# = 𝛼! + 𝛽"# + 𝒙! 2𝜷# + 𝒖! 2𝜽#, 258 where 𝜷# are coefficients specific to each MAGs associated with individual covariate, 𝜽# are the 259 coefficients associated with the latent variable, and 𝛽"# are MAG-specific intercepts, while 𝛼! are 260 optional sample effects (which can either be random or fixed effects) and 𝒖! 2𝜽#, are random effects 261 (latent variables). The recovered 𝜷# coefficients form a relationship between the covariate of 262 interest j and the abundance of a microbe i with a positive value suggesting an increase in the 263 covariate causes an increase in abundance whilst the negative value shows a converse response. The 264 95% confidence intervals of these coefficients were generated and where they crossed the 0 265 boundary, they were deemed insignificant. We conducted this GLLVM analysis through the gllvm 266 package, where we specified use of the negative binomial distribution and the variational 267 approximation method [54]. 268 269 For comparison of grouped features between treatment groups, sample-wise abundance tables 270 were initially subjected into normalisation by total sum scaling (TSS) (number of reads for each MAG 271 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint divided by total of reads per sample) and subsequently via centralised log ratio (CLR) method using 272 the logratio.transfo function (log-ratio transformation) of the mixOmics package [55]. Meanwhile, 273 for individual feature statistical comparison between treatment groups, sample-wise abundance 274 tables were subjected into differential expression analysis using the DESeq2 package with default 275 settings (test: Negative binomial Wald test, type of fitting of dispersions to the mean intensity: 276 parametric, with p < 0.05) [56]. Individual CAZyme function categorisation was adapted from 277 dbCAN2 annotated substrate information [42] and ontology of MetaCyc database [57], with 278 supplementing information from the CAZy database [58] and CAZypedia [59]. Identification of 279 enzymes and subsequent KO enzyme mapping for production pathways of short-chain fatty acids 280 (SCFA): acetate, butyrate and propionate were adapted from previous metagenomic studies [60,61]. 281 Identified enzymes were then matched to KEGG submodules using the KEGG database [41]. 282 283 For visualisation, ggplot2 was used for generation of plots (line graphs, barplots, boxplots, 284 sankeyplot) [62], while ComplexHeatMap was utilised for heatmap clustering (we used Euclidean 285 clustering) [63]. For mapping of phylogenetic tree, we utilised the packages ape and ggtree for 286 manipulation and layering of other MAG features: Guanosine-Cytosine (GC) content, novelty (PG), 287 and quality score (computed using formula: COM – 5 x CON) [64,65]. 288 289

290 Administration of different GHIs impacted overall bird performance. 291 Growth performance is one of the most important indicators of nutrition and health in poultry. 292 According to Table 1, significant differences between treatment groups were only observed in FCR, 293 EPEF, and HB measures (ANOVA, p < 0.05). Estimation of corrected FCR of C0 and C2 has shown T1 294 to have best FCR. EPEF was observed to be highest in T1, followed by T3, but was not significant in 295 C0 and C1. Across all cycles, performance of C2 can be considered most superior as EPEF, BW and 296 ADG of C2 are consistently higher than the C0 and C1 (ANOVA, p <0.05, Table 2). However, C2 also 297 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint demonstrated the highest values and significant between-treatment differences in HB scores and 298 prevalence, in which T1 and T2 exhibited lowest, whilst T4 exhibited the highest HB metrics (ANOVA, 299 p < 0.05). 300 301 Table 1. Performance and health parameters of birds across treatment groups grouped according to 302 cycles. 303 T1 T2 T3 T4 T5 T6 SEM MSE F-stat P-val C0 BW 2.435 2.405 2.436 2.397 2.383 2.376 0.011 0.005 0.819 0.546 FCR 1.368b 1.393ab 1.389ab 1.418a 1.410a 1.412a 0.005 0.001 2.891 0.030 ADG 60.876 60.118 60.902 59.933 59.580 59.410 0.282 2.949 0.819 0.546 MT 3.500 2.867 3.133 3.167 3.033 3.167 1.249 1.249 0.209 0.956 EPEF 406.007 398.163 402.015 389.479 390.510 388.676 2.662 244.196 1.312 0.285 FPDS 0.027 0.020 0.027 0.010 0.007 0.010 0.003 0.000 1.440 0.239 FPDP 2.667 2.000 2.667 1.000 0.667 1.000 0.314 3.333 1.440 0.239 HBS 0.010 0.013 0.007 0.000 0.007 0.003 0.002 0.000 1.200 0.333 HBP 1.000 1.333 0.667 0.000 0.667 0.333 0.178 1.111 1.200 0.333 C1 BW 2.417 2.418 2.439 2.412 2.378 2.436 0.0101 0.004 0.731 0.606 FCR 1.425 1.426 1.415 1.447 1.436 1.426 0.004 0.001 1.146 0.358 ADG 58.945 58.967 59.480 58.832 58.009 59.419 0.248 2.300 0.731 0.606 MT 1.767 2.200 2.600 2.867 1.833 2.433 0.128 0.501 2.250 0.075 EPEF 383.464 383.409 386.183 371.883 377.150 381.341 2.150 167.042 0.975 0.449 FPDS 0.020 0.010 0.043 0.020 0.017 0.027 0.004 0.001 0.516 0.762 FPDP 1.667 1.000 3.333 2.000 1.333 2.000 0.308 3.556 0.725 0.610 HBS 0.030 0.027 0.033 0.037 0.023 0.023 0.005 0.001 0.973 0.450 HBP 2.000 2.000 2.667 2.667 2.000 1.667 0.312 3.533 0.962 0.456 C2 BW 2.648 2.625 2.629 2.603 2.586 2.597 0.013 0.006 0.540 0.744 FCR 1.398d 1.428bc 1.417c 1.459a 1.442a 1.442ab 0.004 0.000 17.473 0.000 ADG 64.591 64.026 64.120 63.477 63.062 63.338 0.307 3.633 0.540 0.744 MT 2.000 2.567 2.900 2.033 2.800 2.367 1.011 0.166 0.857 0.521 EPEF 429.986a 412.178bc 416.671ab 401.411c 399.630c 403.987bc 2.557 141.778 5.617 0.001 FPDS 0.230 0.243 0.213 0.260 0.263 0.260 0.012 0.006 0.401 0.844 FPDP 17.333 17.667 16.667 19.333 19.333 19.000 0.945 36.222 0.218 0.952 HBS 0.153b 0.153b 0.200ab 0.243a 0.190ab 0.210ab 0.009 0.002 3.214 0.019 HBP 13.000b 12.000b 15.333ab 17.667a 14.000ab 15.667ab 0.562 9.133 2.727 0.038 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint Cycle 0 (C0), Cycle 1 (C1) and Cycle 2 (C2) (N = 108 pens, 36 pens/cycle). Performance parameters 304 taken as mean average per pen: Bird weight (BW), Average daily gain (ADG) - grams per day, Feed 305 conversion ratio (FCR), Total mortality (MT), (E) European Production Efficiency Factor (EPEF), 306 Footpad dermatitis score (FPDS) and prevalence (FPDP), and Hockburn score (HBS) and prevalence 307 (HBP). Significant differences between treatment groups (ANOVA, p <0.05). Different letters denote 308 significant differences using DMRT grouping (p <0.05). SEM: Standard error of the mean, MSE: Error 309 Mean Sum of Squares; F-stat (ANOVA F-statistic). 310 Table 2. Performance and health parameters of birds across cycles. 311 C0 C1 C2 MSE F-stat P-val BW 2.405b ± 0.011 2.417b ± 0.01 2.615a ± 0.013 0.005 106.555 0.000 FCR 1.399b ± 0.005 1.429a ± 0.004 1.390b ± 0.005 0.001 19.345 0.000 ADG 60.136b ± 0.282 58.942c ± 0.248 63.769a ± 0.307 2.826 80.499 0.000 MT 3.144a ± 0.175 2.283b ± 0.128 2.444b ± 0.166 0.896 8.420 0.000 EPEF 395.808b ± 2.662 380.572c ± 2.15 410.644a ± 2.557 218.944 37.176 0.000 FPDS 0.017b ± 0.003 0.023b ± 0.004 0.245a ± 0.012 0.002 282.571 0.000 FPDP 1.667b ± 0.314 1.889b ± 0.308 18.222a ± 0.945 13.046 248.782 0.000 HBS 0.007c ± 0.002 0.029b ± 0.005 0.192a ± 0.009 0.001 283.873 0.000 HBP 0.667c ± 0.178 2.167b ± 0.312 14.611a ± 0.562 5.348 394.411 0.000 Cycle 0 (C0), Cycle 1 (C1) and Cycle 2 (C2) (N = 108 pens, 36 pens/cycle). Performance parameters 312 taken as mean average per pen: Bird weight (BW), Average daily gain (ADG) - grams per day, Feed 313 conversion ratio (FCR), Total mortality (MT), (E) European Production Efficiency Factor (EPEF), 314 Footpad dermatitis score (FPDS) and prevalence (FPDP), and Hockburn score (HBS) and prevalence 315 (HBP). Significant differences between treatment groups (ANOVA, p <0.05). Different letters denote 316 significant differences using DMRT grouping (p <0.05). MSE: Error Mean Sum of Squares; F-stat 317 (ANOVA F-statistic). 318 319 To evaluate the overall effects of the individual GHI components across the three cycles, we 320 employed LASSO regression. As shown in Table 3, all GHI components were revealed to be significant 321 predictors in at least one of the performance parameters. Notably, EPEF was largely influenced by all 322 variables while MT was not associated with any. Analysis showed that ionophore administration at 323 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint all stages can largely increase EPEF, but slightly decrease FCR, while the opposite is demonstrated by 324 administration of ionophores at only during finisher stage. Vaccination was shown to have only slight 325 negative effect on EPEF, while use of Probiotic A is shown to have negatively affect almost all 326 parameters. In contrast, use of Probiotic B is generally associated with positive changes in respect to 327 all parameters, apart from FCR and MT. The remaining predictors were shown to have large negative 328 impacts on EPEF but very low positive impact on FCR. 329 Table 3. Lasso regression of performance and health parameters of broilers across all cycles (C0, C1 330 and C2), against sources of variability where the treatment components are coded as 1/0, 331 corresponding to “Yes”/”No”. The model assumes an L1 penalty term which forces the beta 332 coefficient of insignificant predictors to become zero. Significant beta coefficients in blue and red, 333 indicate decrease and increase in performance parameter, respectively. 334 Predictors Performance parameters ADG BW EPEF FCR FPDP HBP MT (Intercept) 61.207 2.489 397.937 1.400 7.354 5.294 2.624 Ionophore: All stages as Yes 0.000 0.000 8.517 -0.019 0.000 0.000 0.000 Ionophore: Finisher only as Yes 0.000 0.000 -10.152 0.019 0.000 0.000 0.000 Vaccinated: Yes 0.000 0.000 -0.002 0.000 0.000 0.000 0.000 Probiotic A: Yes -1.438 -0.069 -3.313 0.000 -5.481 -3.651 0.000 Probiotic B: Yes 1.931 0.099 17.376 -0.018 10.534 9.644 0.000 1 % reduced crude protein: Yes 0.000 0.000 -13.893 0.034 0.000 0.000 0.000 Essential oil: Yes -0.427 -0.017 -10.748 0.022 0.000 0.000 0.000 Prebiotic: Yes 0.000 0.000 -15.617 0.024 0.000 0.000 0.000 Average daily gain (ADG) - grams per day, Bird weight (BW), European Production Efficiency Factor 335 (EPEF), Feed conversion ratio (FCR), FPDP (Footpad dermatitis prevalence), HBP (Hockburn 336 prevalence), Mortality (MT %) 337 338 Administration of different GHIs resulted in shifts in microbial community structure and diversity. 339 The sequencing analysis of 118 caecal samples yielded a total of 2.6 B reads, from which a total of 84 340 MAGs with greater than 75% completeness and less than 5% contamination were recovered 341 Specifically, 83 of these MAGs are detected in C1, while 78 are present in C2. All included MAGs 342 were identified as bacterial species which represents 7 unique phyla, with Firmicutes_A as the most 343 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint common designated phylum among all samples (Figures 2 and 3, Supplement file 2). This is followed 344 by Firmicutes and Bacteroidota, while MAGs belonging to Proteobacteria was not present in any of 345 the groups in C2 (Figure 2A). The global microbiota abundance is dominated by MAGs distributed 346 across 57 identified genera accounting for 96% of the population. Among these, Lactobacillus is the 347 predominant genus, while its member Lactobacillus crispatus is the most common species. However, 348 predominant MAGs at genus and species levels were observed to largely vary across treatments and 349 cycles (Figure 2, Supplement file 2). For instance, Ruminococcus_G was observed to have highest 350 proportion in T1 and T3; Anaerobutyricum in T2, T5 and T6; and Alisipes for T6, while Lactobacillus 351 was consistently highest in proportion among C2 treatment groups. Comparison of F/B ratio 352 revealed significant differences between treatment groups in C1 only wherein T5 is observed to have 353 a higher ratio than T4 (Figure 2D). Approximately 77% of the MAGs (n = 65) contained SCGs for 354 phylogenetic mapping, showing 4 major groupings which are dominated by Firmicutes_A species. 355 Among these MAGs, bin.108 (CAG-267 sp001917135) is revealed as the most novel species (Figure 356 4). 357 358 359 T1 T2 T3 T4 T5 T6 T1 T2 T3 T4 T5 T6 0.00 0.25 0.50 0.75 1.00Proportions CYCLE 1 (n = 60) CYCLE 2 (n = 58) T1 T2 T3 T4 T5 T6 T1 T2 T3 T4 T5 T6 0.00 0.25 0.50 0.75 1.00Proportions T1 T2 T3 T4 T5 T6 T1 T2 T3 T4 T5 T6 0.00 0.25 0.50 0.75 1.00Proportions T1 T2 T3 T4 T5 T6 T1 T2 T3 T4 T5 T6 0.00 0.25 0.50 0.75 1.00Proportions g__Lactobacillus g__Anaerostipes g__Ruminococcus_G g__Mediterraneibacter g__Alistipes g__Unidentified genus g__Anaerobutyricum g__Blautia g__Erysipelatoclostridium g__Limosilactobacillus g__Ligilactobacillus g__Phocaeicola g__CAG−110 g__UMGS1994 g__Streptococcus g__Blautia_A g__OF09−33XD g__Rikenella g__Odoribacter Others PROPORTION (SPECIES) CYCLE 1 (n = 60) CYCLE 2 (n = 58) CYCLE 1 (n = 60) CYCLE 2 (n = 58) T1 T2 T3 T4 T5 T6 T1 T2 T3 T4 T5 T6 0.00 0.25 0.50 0.75 1.00Proportions g__Lactobacillus g__Anaerostipes g__Ruminococcus_G g__Mediterraneibacter g__Alistipes g__Unidentified genus g__Anaerobutyricum g__Blautia g__Erysipelatoclostridium g__Limosilactobacillus g__Ligilactobacillus g__Phocaeicola g__CAG−110 g__UMGS1994 g__Streptococcus g__Blautia_A g__OF09−33XD g__Rikenella g__Odoribacter Others T1 T2 T3 T4 T5 T6 T1 T2 T3 T4 T5 T6 0.00 0.25 0.50 0.75 1.00Proportions g__Lactobacillus g__Anaerostipes g__Ruminococcus_G g__Mediterraneibacter g__Alistipes g__Unidentified genus g__Anaerobutyricum g__Blautia g__Erysipelatoclostridium g__Limosilactobacillus g__Ligilactobacillus g__Phocaeicola g__CAG−110 g__UMGS1994 g__Streptococcus g__Blautia_A g__OF09−33XD g__Rikenella g__Odoribacter Others T1 T2 T3 T4 T5 T6 T1 T2 T3 T4 T5 T6 0.00 0.25 0.50 0.75 1.00Proportions g__Lactobacillus g__Anaerostipes g__Ruminococcus_G g__Mediterraneibacter g__Alistipes g__Unidentified genus g__Anaerobutyricum g__Blautia g__Erysipelatoclostridium g__Limosilactobacillus g__Ligilactobacillus g__Phocaeicola g__CAG−110 g__UMGS1994 g__Streptococcus g__Blautia_A g__OF09−33XD g__Rikenella g__Odoribacter Others T1 T2 T3 T4 T5 T6 T1 T2 T3 T4 T5 T6 0.00 0.25 0.50 0.75 1.00Proportions g__Lactobacillus g__Anaerostipes g__Ruminococcus_G g__Mediterraneibacter g__Alistipes g__Unidentified genus g__Anaerobutyricum g__Blautia g__Erysipelatoclostridium g__Limosilactobacillus g__Ligilactobacillus g__Phocaeicola g__CAG−110 g__UMGS1994 g__Streptococcus g__Blautia_A g__OF09−33XD g__Rikenella g__Odoribacter Others T1 T2 T3 T4 T5 T6 T1 T2 T3 T4 T5 T6 0.00 0.25 0.50 0.75 1.00Proportions g__Lactobacillus g__Anaerostipes g__Ruminococcus_G g__Mediterraneibacter g__Alistipes g__Unidentified genus g__Anaerobutyricum g__Blautia g__Erysipelatoclostridium g__Limosilactobacillus g__Ligilactobacillus g__Phocaeicola g__CAG−110 g__UMGS1994 g__Streptococcus g__Blautia_A g__OF09−33XD g__Rikenella g__Odoribacter Others T1 T2 T3 T4 T5 T6 T1 T2 T3 T4 T5 T6 0.00 0.25 0.50 0.75 1.00Proportions g__Lactobacillus g__Anaerostipes g__Ruminococcus_G g__Mediterraneibacter g__Alistipes g__Unidentified genus g__Anaerobutyricum g__Blautia g__Erysipelatoclostridium g__Limosilactobacillus g__Ligilactobacillus g__Phocaeicola g__CAG−110 g__UMGS1994 g__Streptococcus g__Blautia_A g__OF09−33XD g__Rikenella g__Odoribacter Others T1 T2 T3 T4 T5 T6 T1 T2 T3 T4 T5 T6 0.00 0.25 0.50 0.75 1.00Proportions Phylum p__Firmicutes_A p__Firmicutes p__Bacteroidota p__Cyanobacteria p__Desulfobacterota p__Firmicutes_B p__Proteobacteria PROPORTION (GENUS) PROPORTION (PHYLUM) A C C T1 T2 T3 T4 T5 T6 T1 T2 T3 T4 T5 T6 0.00 0.25 0.50 0.75 1.00Proportions Unidentified at species level s____ Lactobacillus crispatus s__ Alistipes sp900021155 s__ Lactobacillus gallinarum s__ Limosilactobacillus vaginalis s__ Ligilactobacillus salivarius s__ Mediterraneibacter sp002314255 s__ Phocaeicola vulgatus s__ Lactobacillus johnsonii s__ Streptococcus alactolyticus s__ CAG−110 sp900546915 s__ Rikenella microfusus s__ Odoribacter splanchnicus s__ Erysipelatoclostridium spiroforme s__ UMGS929 sp900546875 s__ Tidjanibacter inops_A s__ Enterocloster sp900547035 s__ Erysipelatoclostridium sp002160495 s__ CAG−631 sp900762575 s__ Bacteroides fragilis Others T1 T2 T3 T4 T5 T6 T1 T2 T3 T4 T5 T6 0.00 0.25 0.50 0.75 1.00Proportions Unidentified at species level s____ Lactobacillus crispatus s__ Alistipes sp900021155 s__ Lactobacillus gallinarum s__ Limosilactobacillus vaginalis s__ Ligilactobacillus salivarius s__ Mediterraneibacter sp002314255 s__ Phocaeicola vulgatus s__ Lactobacillus johnsonii s__ Streptococcus alactolyticus s__ CAG−110 sp900546915 s__ Rikenella microfusus s__ Odoribacter splanchnicus s__ Erysipelatoclostridium spiroforme s__ UMGS929 sp900546875 s__ Tidjanibacter inops_A s__ Enterocloster sp900547035 s__ Erysipelatoclostridium sp002160495 s__ CAG−631 sp900762575 s__ Bacteroides fragilis Others T1 T2 T3 T4 T5 T6 T1 T2 T3 T4 T5 T6 0.00 0.25 0.50 0.75 1.00Proportions Unidentified at species level s____ Lactobacillus crispatus s__ Alistipes sp900021155 s__ Lactobacillus gallinarum s__ Limosilactobacillus vaginalis s__ Ligilactobacillus salivarius s__ Mediterraneibacter sp002314255 s__ Phocaeicola vulgatus s__ Lactobacillus johnsonii s__ Streptococcus alactolyticus s__ CAG−110 sp900546915 s__ Rikenella microfusus s__ Odoribacter splanchnicus s__ Erysipelatoclostridium spiroforme s__ UMGS929 sp900546875 s__ Tidjanibacter inops_A s__ Enterocloster sp900547035 s__ Erysipelatoclostridium sp002160495 s__ CAG−631 sp900762575 s__ Bacteroides fragilis Others Kruskal−Wallis, p = 0.0026 *** 0 10 20 30 T1 T2 T3 T4 T5 T6 Treatment Firmicutes:Bacteroidetes ratio Kruskal−Wallis, p = 0.061 0 10 20 30 40 T1 T2 T3 T4 T5 T6 Treatment Firmicutes:Bacteroidetes ratio CYCLE 1 (n = 60) CYCLE 2 (n = 58) D .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint Figure 2. Sample-wise proportion of 84 MAGs recovered from C1 (n = 60) and C2 (n = 58). 360 Proportions at (A) Phylum, (B) Genus (genera with <2% prevalence grouped into “Others”) and (C) 361 Species levels (species with <1% prevalence grouped into “Others”), ranked from most dominant to 362 least, grouped per treatment/cycle. Plot also shows (D) Firmicutes:Bacteroidota ratio, grouped per 363 cycle and treatment, with significant differences based on Kruskal-Wallis (p < 0.05) and pairwise 364 Dunn testing with Bonferroni correction: p < 0.001). 365 366 Next, we hypothesised that GHI groups may influence in the microbial community diversity, which 367 we then estimated through use of various alpha and beta diversity metrics (Figure 3). In C1, 368 microbial communities of T3 samples were found richer and possessed more even distribution than 369 other groups (T1, T2, T4, and T6) as indicated by significantly higher FA, R and PE values. Meanwhile 370 in C2, the majority of alpha metrices of T5 and T6 were significantly elevated compared to T1 and T3. 371 The principal components (PC1 and PC2) of our PCoA plot on Bray-Curtis estimates explained a 372 considerable portion of the variance (14% and 11%, respectively), from which a discernible 373 separation (PERMANOVA, p < 0.001, R2 = 27.84%) between C1 and C2 groups was observed. While 374 separation between treatment groups is not distinct, slight clustering of groups can be observed: T2 375 and T3 (in C1) are separated from the other groups, while T1, T2 and T6 (in C2) are separated from 376 the rest. 377 378 379 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint 380 Figure 3. Phylogeny and taxonomic diversity of MAGs recovered from C1 (n = 60) and C2 (n = 58). (A) 381 Phylogenetic tree of 65 MAGs recovered via GToTree using 25 bacterial and archaeal specific single 382 copy genes. The tree also features G-C content, Quality index (genome completion – 5 x genome 383 contamination), and Novelty (represented by phylogenetic gain (PG) values calculated using GTDB 384 toolkit). (B) 10 most novel MAGs shown (indicated by high PG). Finally, alpha diversity (C) is 385 represented by Fisher alpha, Pielou’s evenness, Rarefied richness, Shannon and Simpson index, with 386 ANOVA significance: p < 0.001 (***), p < 0.01 (**), p < 0.05 (*), (D) Beta diversity is represented by 387 PCoA plot of Bray-Curtis indices, with PERMANOVA (p <0.001). 388 389 bin.19bin.179bin.76 bin.52bin.138 bin.93 bin.163 bin.74bin.168 bin.11 bin.251 bin.119 bin.279 bin.208 bin.114 bin.17 bin.152bin.267bin.120bin.125bin.258 bin.193 bin.9 bin.112 bin.280 bin.98 bin.254 bin.65 bin.151 bin.140 bin.249 bin.84bin.260bin.262 bin.61 bin.148 bin.239 bin.156 bin.14 bin.106 bin.92 bin.215 bin.289 bin.27 bin.205 bin.81 bin.122 bin.83 bin.135 bin.145 bin.137bin.34bin.5bin.213 bin.286bin.91 bin.96 bin.214 bin.4 bin.266 bin.79 bin.26 bin.108 bin.200 bin.211 g__GCA −900066135;s__ s__CAG −267 sp001917135 s__UMGS946 sp900544965 s__Merdibacter sp900754715 s__UMGS929 sp900546875 g__OEMS01;s__ s__Lactobacillus johnsonii g__UMGS1370;s__ s__Lactobacillus crispatus f__UBA7702;g__;s__ s__Rikenella microfusus g__UBA5446;s__ s__Mediterraneibacter sp002314255 s__CAG −313 sp900539265 o__Bacillales;f__;g__;s__ g__Agathobaculum;s__ g__CAG −822;s__ s__Ligilactobacillus salivarius g__Anaerostipes;s__ g__Intestinimonas;s__ s__CAG −110 sp900546915 s__CAG −314 sp000437915 s__UMGS403 sp900541565s__An172 sp002160515 s__Streptococcus alactolyticus g__UMGS1994;s__ g__UMGS1370;s__ s__UMGS1688 sp900554085 g__CAG−269;s__ s__Tidjanibacter inops_A f__Gastranaerophilaceae;g__;s__ s__Blautia_A sp002159835 g__Anaerotignum;s__ s__UMGS268 sp900540175 g__UBA1685;s__ s__CAG −631 sp900762575 s__Lactobacillus gallinarum g__Bilophila;s__ g__Ruminococcus_D;s__s__UBA1409 sp002305045 f__RUG14156;g__;s__ s__Limosilactobacillus vaginalis g__OF09−33XD;s__ f__DTU072;g__;s__ s__Erysipelatoclostridium spiroforme s__Odoribacter splanchnicus g__Ruminococcus_G;s__ s__Alistipes communis f__Gastranaerophilaceae;g__;s__ s__Alistipes sp900021155 g__UBA1777;s__ g__UBA1409;s__ g__CAG −302;s__ g__CAG −110;s__ s__UMGS403 sp900541975 f__RUG14156;g__;s__ s__Enterocloster sp900547035 g__Romboutsia;s__ g__UBA3818;s__ s__Massilicoli timonensis s__Bacteroides fragilis g__UBA7160;s__ s__UBA5446 sp004553625 s__Phocaeicola vulgatus s__Erysipelatoclostridium sp002160495 20 ** * *** *** * ** * * ** * * ** * * * * * *** ** ** ** ** * * ** Fisher alpha ** Pielou's evenness ** Richness * Shannon Simpson C1_T1 C1_T2 C1_T3 C1_T4 C1_T5 C1_T6 C2_T1 C2_T2 C2_T3 C2_T4 C2_T5 C2_T6 C1_T1 C1_T2 C1_T3 C1_T4 C1_T5 C1_T6 C2_T1 C2_T2 C2_T3 C2_T4 C2_T5 C2_T6 C1_T1 C1_T2 C1_T3 C1_T4 C1_T5 C1_T6 C2_T1 C2_T2 C2_T3 C2_T4 C2_T5 C2_T6 C1_T1 C1_T2 C1_T3 C1_T4 C1_T5 C1_T6 C2_T1 C2_T2 C2_T3 C2_T4 C2_T5 C2_T6 C1_T1 C1_T2 C1_T3 C1_T4 C1_T5 C1_T6 C2_T1 C2_T2 C2_T3 C2_T4 C2_T5 C2_T6 0.85 0.90 0.95 2.5 3.0 3.5 4.0 7 8 9 10 0.7 0.8 0.9 1.0 1.1 10 15 20 25 30Observed Values ** * *** *** * ** * * ** * * ** * * * * * *** ** ** ** ** * * ** Fisher alpha ** Pielou's evenness ** Richness * Shannon Simpson C1_T1 C1_T2 C1_T3 C1_T4 C1_T5 C1_T6 C2_T1 C2_T2 C2_T3 C2_T4 C2_T5 C2_T6 C1_T1 C1_T2 C1_T3 C1_T4 C1_T5 C1_T6 C2_T1 C2_T2 C2_T3 C2_T4 C2_T5 C2_T6 C1_T1 C1_T2 C1_T3 C1_T4 C1_T5 C1_T6 C2_T1 C2_T2 C2_T3 C2_T4 C2_T5 C2_T6 C1_T1 C1_T2 C1_T3 C1_T4 C1_T5 C1_T6 C2_T1 C2_T2 C2_T3 C2_T4 C2_T5 C2_T6 C1_T1 C1_T2 C1_T3 C1_T4 C1_T5 C1_T6 C2_T1 C2_T2 C2_T3 C2_T4 C2_T5 C2_T6 0.85 0.90 0.95 2.5 3.0 3.5 4.0 7 8 9 10 0.7 0.8 0.9 1.0 1.1 10 15 20 25 30Observed Values −0.4 −0.2 0.0 0.2 −0.4 −0.2 0.0 0.2 Dim1 (14.48%) Dim2 (11.15%) C1_T1 C1_T2 C1_T3 C1_T4 C1_T5 C1_T6 C2_T1 C2_T2 C2_T3 C2_T4 C2_T5 C2_T6 PERMANOVA (p<0.001) Taxon PG Groups bin.108:s_CAG-267 sp001917135 0.561 p__Proteobacteria bin.249:s_UMGS268 sp900540175 0.402 p__Firmicutes bin.254:s__CAG-631 sp900762575 0.381 p__Firmicutes bin.140:s__CAG-313 sp900539265 0.353 p__Firmicutes bin.83:g__Romboutsia;s__ 0.252 p__Firmicutes_A bin.135:o__UBA7702;f__UBA7702;g__;s__ 0.224 p__Firmicutes_B bin.279:f__DTU072;g__;s__ 0.223 p__Firmicutes_A bin.267:s__Limosilactobacillus vaginalis 0.207 p__Firmicutes bin.145:o__Bacillales;f__;g__;s__ 0.197 p__Firmicutes bin.114:s_UMGS929 sp900546875 0.183 p__Firmicutes_A −0.4 −0.2 0.0 0.2 −0.4 −0.2 0.0 0.2 Dim1 (14.48%) Dim2 (11.15%) C1_T1 C1_T2 C1_T3 C1_T4 C1_T5 C1_T6 C2_T1 C2_T2 C2_T3 C2_T4 C2_T5 C2_T6 PERMANOVA (p<0.001) Novelty (Phylogenetic Gain) 30 40 50 60 GC content (%) Quality (Completeness – 5 x Contamination) A B C D C1_T1 C1_T2 C1_T3 C1_T4 C1_T5 C1_T6 C2_T1 C2_T2 C2_T3 C2_T4 C2_T5 C2_T6 0.00 0.25 0.50 0.75 1.00 T reatment Proportions Phylum p__Firmicutes_A p__Firmicutes p__Bacteroidota p__Cyanobacteria p__Desulfobacterota p__Firmicutes_B p__Proteobacteria PHYLUM .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint Metagenomic-assembled genome composition is associated with administration of different GHIs 390 and broiler performance 391 To further understand individual influence of GHI groups on microbiome and performance, we 392 conducted GLLVM regression analysis of each MAG (using T1 as reference predictor). A total of 43 393 and 38 MAGs (59 for both) were observed to have a significant association with GHI groups in C1 and 394 C2, respectively (Figure 4, Supplement file 1). For C1, the majority of these MAGs (n = 41/43, 95%) 395 exhibited a negative association with GHI administration, denoting decrease in abundance compared 396 to T1. From this, T2 had the greatest number of decreased MAGs, followed by T3 and T6, while 397 increase of 5 MAGs were seen in T4. Notable MAGs to consistently change across groups in C1 were 398 those belonging to Bacteroidales (bin.291 Alistipes megaguti, bin.34 Alistipes communis, bin.294 399 Alistipes sp900290115, bin.234 Barnesiella viscericola), as well as bin.99 (Limolactobacillus) and 400 bin.74 (CAG-110). Bin.34 is also observed to have the lowest GLLVM coefficient, followed by bin.92 401 (g_UBA7160) and bin.4 (f_Gastrophilaceae). In addition, all Cyanobacteria MAGs were observed to 402 show negative association with GHI groups. 403 In contrast, a higher number of positively associated MAGs (n = 15) was observed in C2 than in C1 (n 404 = 7), from which T5 has the highest count (n = 8). Meanwhile, 25 MAGs were seen to have negative 405 association with GHI groups in C2, with T6 having the highest number of negatively associated MAGs 406 (n = 15). Bin.74 (g_CAG-110) is observed to be dynamic across all GHI groups, as well as the having 407 the highest coefficient among all MAGs in C2; while bin.208 (UMGS1688 sp90054085) is shown to 408 have the strongest negative association (Figure 4B). 409 For performance, we considered EPEF and MT to represent overall broiler health and performance 410 (EPEF are highly correlated with the other measured performance parameters). Our analysis 411 revealed that EPEF is positively associated with 40 MAGs and negatively associated with 33 MAGs, 412 while MT is associated with only 30 MAGs (10 positive and 20 negative associated MAGs). Bin.137 413 (Rikenella microfusus) is noted to have the highest GLLVM coefficient against EPEF, followed by 414 bin.92 (g_UBA7160) and bin.58 (Butyricicoccus). Bin.234 (Barnesiella viscericola) is observed to have 415 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint the strongest negative association with EPEF but also the strongest positive association with MT 416 (Figure 4C). 417 Interestingly, at least one of the Lactobacillaceae MAGs (bin.258 Lactobacillus gallinarum, bin.99 418 Limolactobacillus, bin.152 Ligilactobacillus, bin.125 Lactobacillus crispatus, bin.267 Limolactobacillus 419 salivarius, bin.120 Lactobacillus johnsonii) are negatively associated across the GHI groups in both 420 cycles (except for T4) but are also seen to have positive association with EPEF, indicating that 421 decreased levels of Lactobacillaceae MAGs in GHIs may contribute to diminished EPEF values. Similar 422 patterns were also seen for other MAGs such as bin.34, bin.137, bin.65, and bin.84. Furthermore, 423 numerous MAGs enriched in GHI groups but negatively associated with EPEF were also identified 424 (Cyanobacteria MAGs, bin.200: UMGS1993, bin.279: f_DTU072 in C2) which signifies that elevated 425 levels of these MAGs can contribute to decreased levels of EPEF. 426 427 428 Figure 4. Taxonomic classification and differential analysis of 84 MAGs recovered from C1 (n = 60) 429 and C2 (n = 58). (A) A sankey plot illustrating the classification of the bacterial species at various 430 taxonomic ranks. The figure also includes general linear latent variable model (GLLVM) results 431 showing association of MAGs with (B) treatment groups (with C1 and C2 done separately) and (C) 432 C1 C2 T2 T3 T4 T5 T6 T2 T3 T4 T5 T6 EPEF bin.291 bin.34 bin.294 bin.286 bin.234 bin.96 bin.91 bin.5 bin.137 bin.213 bin.266 bin.214 bin.79 bin.4 bin.26 bin.254 bin.151 bin.65 bin.140 bin.258 bin.99 bin.152 bin.125 bin.267 bin.120 bin.193 bin.112 bin.9 bin.98 bin.280 bin.145 bin.249 bin.200 bin.211 bin.97 bin.127 bin.83 bin.218 bin.58 bin.84 bin.163 bin.61 bin.168 bin.19 bin.93 bin.262 bin.179 bin.74 bin.55 bin.260 bin.110 bin.138 bin.76 bin.292 bin.52 bin.148 bin.122 bin.106 bin.239 bin.14 bin.302 bin.27 bin.225 bin.301 bin.215 bin.205 bin.227 bin.7 bin.289 bin.81 bin.31 bin.156 bin.24 bin.198 bin.92 bin.17 bin.208 bin.114 bin.279 bin.11 bin.251 bin.119 bin.135 bin.108 MT matrix_31 −10 −5 0 5 10 matrix_50 −0.2 −0.1 0 0.1 0.2 matrix_51 −5 0 5 p__Proteobacteria = 1 p__Firmicutes_B = 1 p__Firmicutes_A = 50 p__Firmicutes = 17 p__Desulfobacterota = 1 p__Cyanobacteria = 4 p__Bacteroidota = 10 c__Alphaproteobacteria = 1 c__Dehalobacteriia = 1 c__Clostridia_A = 7 c__Clostridia = 43 c__Bacilli = 17 c__Desulfovibrionia = 1 c__Vampirovibrionia = 4 c__Bacteroidia = 10 o__RF32 = 1 o__UBA7702 = 1 o__Christensenellales = 7 o__Lachnospirales = 19 o__Oscillospirales = 18 o__Peptostreptococcales = 3 o__TANB77 = 3 o__Acholeplasmatales = 1 o__Bacillales = 1 o__Erysipelotrichales = 4 o__Lactobacillales = 7 o__ML615J−28 = 1 o__RF39 = 2 o__RFN20 = 1 o__Desulfovibrionales = 1 o__Gastranaerophilales = 4 o__Bacteroidales = 10 f__CAG−239 = 1 f__UBA7702 = 1 f__CAG−74 = 1 f__CAG−138 = 2 f__DTU072 = 1 f__CAG−314 = 2 f__CAG−917 = 1 f__Lachnospiraceae = 18 f__Anaerotignaceae = 1 f__Butyricicoccaceae = 3 f__Oscillospiraceae = 7 f__Ruminococcaceae = 5 f__Acutalibacteraceae = 3 f__Anaerovoracaceae = 2 f__Peptostreptococcaceae = 1 f__CAG−508 = 3 f__Anaeroplasmataceae = 1 f__ = 1 f__Erysipelatoclostridiaceae = 2 f__Erysipelotrichaceae = 2 f__Streptococcaceae = 1 f__Lactobacillaceae = 6 f__CAG−313 = 1 f__UBA660 = 2 f__CAG−631 = 1 f__Desulfovibrionaceae = 1 f__Gastranaerophilaceae = 2 f__RUG14156 = 2 f__Rikenellaceae = 6 f__Bacteroidaceae = 2 f__Barnesiellaceae = 1 f__Marinifilaceae = 1 bin.108:CAG−267 sp001917135 = 1 bin.135:g__;s__ = 1 bin.119:OEMS01;s__ = 1 bin.251:UBA1685;s__ = 1 bin.11:UMGS946 sp900544965 = 1 bin.279:g__;s__ = 1 bin.114:UMGS929 sp900546875 = 1 bin.208:UMGS1688 sp900554085 = 1 bin.17:CAG−314 sp000437915 = 1 bin.92:UBA7160;s__ = 1 bin.198:Anaerobutyricum;s__ = 1 bin.24:Fusicatenibacter;s__ = 1 bin.156:Anaerostipes;s__ = 1 bin.31:Eubacterium_G;s__ = 1 bin.81:Enterocloster sp900547035 = 1 bin.289:Ruminococcus_G;s__ = 1 bin.7:Blautia_A;s__ = 1 bin.227:Mediterraneibacter;s__ = 1 bin.205:UMGS1370;s__ = 1 bin.215:Blautia_A sp002159835 = 1 bin.301:Anaerostipes;s__ = 1 bin.225:Blautia;s__ = 1 bin.27:OF09−33XD;s__ = 1 bin.302:Mediterraneibacter;s__ = 1 bin.14:Mediterraneibacter sp002314255 = 1 bin.239:Anaerotignum;s__ = 1 bin.106:GCA−900066135;s__ = 1 bin.122:UMGS1370;s__ = 1 bin.148:Agathobaculum;s__ = 1 bin.52:UBA1777;s__ = 1 bin.292:Gemmiger;s__ = 1 bin.76:UMGS403 sp900541975 = 1 bin.138:UBA5446;s__ = 1 bin.110:Butyricicoccus;s__ = 1 bin.260:Ruminococcus_D;s__ = 1 bin.55:g__;s__ = 1 bin.74:CAG−110;s__ = 1 bin.179:UMGS403 sp900541565 = 1 bin.262:UBA1409 sp002305045 = 1 bin.93:UBA5446 sp004553625 = 1 bin.19:An172 sp002160515 = 1 bin.168:CAG−110 sp900546915 = 1 bin.61:UBA1409;s__ = 1 bin.163:Intestinimonas;s__ = 1 bin.84:UBA3818;s__ = 1 bin.58:Butyricicoccus;s__ = 1 bin.218:CAG−145 sp900754795 = 1 bin.83:Romboutsia;s__ = 1 bin.127:g__;s__ = 1 bin.97:CAG−269;s__ = 1 bin.211:CAG−269;s__ = 1 bin.200:UMGS1994;s__ = 1 bin.249:UMGS268 sp900540175 = 1 bin.145:g__;s__ = 1 bin.280:Erysipelatoclostridium spiroforme = 1 bin.98:Erysipelatoclostridium sp002160495 = 1 bin.9:Massilicoli timonensis = 1 bin.112:Merdibacter sp900754715 = 1 bin.193:Streptococcus alactolyticus = 1 bin.120:Lactobacillus johnsonii = 1 bin.267:Limosilactobacillus vaginalis = 1 bin.125:Lactobacillus crispatus = 1 bin.152:Ligilactobacillus salivarius = 1 bin.99:Limosilactobacillus;s__ = 1 bin.258:Lactobacillus gallinarum = 1 bin.140:CAG−313 sp900539265 = 1 bin.65:CAG−302;s__ = 1 bin.151:CAG−822;s__ = 1 bin.254:CAG−631 sp900762575 = 1 bin.26:Bilophila;s__ = 1 bin.4:g__;s__ = 1 bin.79:g__;s__ = 1 bin.214:g__;s__ = 1 bin.266:g__;s__ = 1 bin.213:Tidjanibacter inops_A = 1 bin.137:Rikenella microfusus = 1 bin.5:Alistipes sp900021155 = 1 bin.91:Bacteroides fragilis = 1 bin.96:Phocaeicola vulgatus = 1 bin.234:Barnesiella viscericola = 1 bin.286:Odoribacter splanchnicus = 1 bin.294:Alistipes sp900290115 = 1 bin.34:Alistipes communis = 1 bin.291:Alistipes megaguti = 1 Phylum Class Order Family id x A B C (n= 7) (n= 8) (n= 17) (n= 32) (n= 84) T2 T3 T4 T5 T6 CYCLE 1 (n = 60) GENUS/ SPECIES FAMILY ORDER CLASS PHYLUM C1C2 C1_T2 C1_T3 C1_T4 C1_T5 C1_T6 C2_T2 C2_T3 C2_T4 C2_T5 C2_T6 bin.291 bin.34 bin.294 bin.286 bin.234 bin.96 bin.91 bin.5 bin.137 bin.213 bin.266 bin.214 bin.79 bin.4 bin.26 bin.254 bin.151 bin.65 bin.140 bin.258 bin.99 bin.152 bin.125 bin.267 bin.120 bin.193 bin.112 bin.9 bin.98 bin.280 bin.145 bin.249 bin.200 bin.211 bin.97 bin.127 bin.83 bin.218 bin.58 bin.84 bin.163 bin.61 bin.168 bin.19 bin.93 bin.262 bin.179 bin.74 bin.55 bin.260 bin.110 bin.138 bin.76 bin.292 bin.52 bin.148 bin.122 bin.106 bin.239 bin.14 bin.302 bin.27 bin.225 bin.301 bin.215 bin.205 bin.227 bin.7 bin.289 bin.81 bin.31 bin.156 bin.24 bin.198 bin.92 bin.17 bin.208 bin.114 bin.279 bin.11 bin.251 bin.119 bin.135 bin.108 ADG BW EPEF FCR_2kg FPDP HBP MT matrix_26 −10 −5 0 5 10 matrix_27 −10 −5 0 5 10 10 EPEF MT CYCLE 2 (n = 58) T2 T3 T4 T5 T6 C1 C2 C1_T2 C1_T3 C1_T4 C1_T5 C1_T6 C2_T2 C2_T3 C2_T4 C2_T5 C2_T6 bin.291 bin.34 bin.294 bin.286 bin.234 bin.96 bin.91 bin.5 bin.137 bin.213 bin.266 bin.214 bin.79 bin.4 bin.26 bin.254 bin.151 bin.65 bin.140 bin.258 bin.99 bin.152 bin.125 bin.267 bin.120 bin.193 bin.112 bin.9 bin.98 bin.280 bin.145 bin.249 bin.200 bin.211 bin.97 bin.127 bin.83 bin.218 bin.58 bin.84 bin.163 bin.61 bin.168 bin.19 bin.93 bin.262 bin.179 bin.74 bin.55 bin.260 bin.110 bin.138 bin.76 bin.292 bin.52 bin.148 bin.122 bin.106 bin.239 bin.14 bin.302 bin.27 bin.225 bin.301 bin.215 bin.205 bin.227 bin.7 bin.289 bin.81 bin.31 bin.156 bin.24 bin.198 bin.92 bin.17 bin.208 bin.114 bin.279 bin.11 bin.251 bin.119 bin.135 bin.108 ADG BW EPEF FCR_2kg FPDP HBP MT matrix_26 −10 −5 0 5 10 matrix_27 −10 −5 0 5 10 - 0.2 - 0.1 0 0.1 0.2 C1C2 T2T3T4T5T6T2T3T4T5T6EPEF bin.291 bin.34 bin.294 bin.286 bin.234 bin.96 bin.91 bin.5 bin.137 bin.213 bin.266 bin.214 bin.79 bin.4 bin.26 bin.254 bin.151 bin.65 bin.140 bin.258 bin.99 bin.152 bin.125 bin.267 bin.120 bin.193 bin.112 bin.9 bin.98 bin.280 bin.145 bin.249 bin.200 bin.211 bin.97 bin.127 bin.83 bin.218 bin.58 bin.84 bin.163 bin.61 bin.168 bin.19 bin.93 bin.262 bin.179 bin.74 bin.55 bin.260 bin.110 bin.138 bin.76 bin.292 bin.52 bin.148 bin.122 bin.106 bin.239 bin.14 bin.302 bin.27 bin.225 bin.301 bin.215 bin.205 bin.227 bin.7 bin.289 bin.81 bin.31 bin.156 bin.24 bin.198 bin.92 bin.17 bin.208 bin.114 bin.279 bin.11 bin.251 bin.119 bin.135 bin.108 MT matrix_31 −10 −5 0 5 10 matrix_50 −0.2 −0.1 0 0.1 0.2 matrix_51 −5 0 5 - 5 0 5 Coef Coef .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint performance and health parameters – EPEF (European Performance Efficiency Factor) and MT (Total 433 mortality). Coef: GLLVM coefficients in blue (0), representing negative or positive 434 coefficients (or decrease or increase) for each MAG against GHI treatments (T2, T3, T4, T5, T6) as 435 predictors, in comparison to T1 (reference). Coefficients in white indicates that coefficient is 436 insignificant (no association), while grey indicates that the MAG was not recovered in that 437 treatment. 438 439 Administration of different GHIs resulted in shifts in metabolic functions. 440 Since significant differences in MAGs were observed across treatments, we then investigated the 441 impact of GHIs on metabolic functions. Enzymes including CAZymes and proteases are important for 442 metabolism and reproduction of microbial species, hence they may also play crucial roles in nutrition 443 and digestive physiology of chickens. As seen in Figure 4, we detected a total of 128 CAZymes 444 belonging to two major families – glycoside hydrolase (GH) and polysaccharide lyase (PL), with the 445 former being more dominant. Both CAZyme families were significantly different in abundance across 446 C1 treatment groups, revealing T5 to have highest GH abundance but also lowest PL abundance (p < 447 0.05, ANOVA) (Figure 5A). Meanwhile, there were no significant differences between groups within 448 C2, but they have a relatively higher overall CAZyme abundance than C1 groups. Furthermore, 449 clustering analysis exhibited separation of T4 and T5 from the other groups in C2, and T5 in C2 450 (Figure 5B). 451 The most common enzyme across all treatments is observed to be GH013 which is detected in 27 452 bacterial families in our study. A total of 24 enzymes are noted to have significantly different 453 abundance between T1 and GHI groups T4 and T5 (C1) (Wald Test, p <0.05, Figure 5B). For T4, 9 454 enzymes with activity against pectin were significantly lower than T1, while 11 enzymes with activity 455 against starch/sugars and other carbohydrates are significantly higher. Similarly, 15 enzymes with 456 activity against pectin and hemicellulose are downregulated in T5; these same enzymes are 457 observed to be consistently present in the family Bacteroidaceae (Figure 5E). In C2, only GH15 is 458 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint observed to have a significantly difference (upregulated in T2, T3, and T5) which has capacity for 459 sugars/starch digestion. 460 Differential abundance of proteases across treatments was also investigated in this study. A total of 461 108 protease families distributed across 9 protease types were detected in our samples (Figure 6). 462 Metallo peptidases were detected as the most common and most diverse protease catalytic type, 463 with M38 family being the most dominant across 47 metallo families. Other detected catalytic types 464 include inhibitors and threonine peptidases which were both significantly different in abundance in 465 C1, where threonine abundance was specifically observed significantly lower in T4 compared to T1 466 (Dunn Test with Bonferroni correction, p <0.05). In C2, abundances of cysteine peptidases were 467 markedly disparate (Kruskal Sum Rank test, p < 0.05), with T4 having the highest value among all 468 groups but although insignificant. At family level, divergent clustering of T4 and T5 from other 469 groups were seen in C1; In contrast, for C2, T1 and T4 are seen to be more similar in abundance than 470 the other groups. Across all treatment groups, C38 (Cysteine) is consistently the most abundant 471 peptidase, followed by M38 and S33. However, diminished levels of M28X were evident in T4 and T5 472 while elevated levels of M93 and M28A were observed in T4 in C1. Notably, all these differentiated 473 families are present in Bacteroidaceae family (Figure 6E). 474 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint 475 Figure 5. Carbohydrate-active enzymes (CAZyme) gene abundance recovered from C1 (n = 60) and 476 C2 (n = 58). (A) Mean normalised abundance of CAZyme gene IDs; Red and blue colour of heatmap 477 cells indicates high and low abundance, respectively); IDs in bold and underlined indicates significant 478 increase or decrease in abundance compared to T1 based on Wald-Test (DESEq2). CAZyme IDs 479 grouped according to (B) substrate/function based on the dbCAN2 and MetaCyc databases) across 480 treatment groups and cycle. Number of MAGs containing CAZyme genes, grouped according to (C) 481 phylum and (D) family taxonomic ranks, (E) Comparison of total glycoside hydrolase (GH) and 482 polysaccharide lyase (PL) abundances across treatment groups and cycles, based on ANOVA and 483 posthoc pairwise t-testing with Bonferroni correction. Normalisation method: Total Sum Scaling and 484 C2 C1 T6 T2 T3 T1 T4 T5 T6 T3 T2 T1 T4 T5 Sugars/StarchCelluloseHemicellulosePectinBeta −glucans FructansOther glucansOthers GH013 GH002 GH043 GH003 GH109 GH001 GH025 GH032 GH036 GH031 GH073 GH077 GH020 GH092 GH023 GH029 GH004 GH065 GH018 GH097 GH028 GH030 GH078 GH033 GH051 GH105 GH042 GH094 GH084 GH016 GH038 GH005 GH130 GH127 GH095 GH027 GH133 GH057 GH170 GH008 GH026 GH089 GH035 GH088 GH171 GH136 GH146 GH110 GH125 GH123 GH112 GH101 GH154 GH085 GH024 PL008 GH141 GH010 GH115 GH015 PL012 GH079 PL033 GH068 GH076 GH144 GH159 GH120 GH053 GH039 GH106 GH067 GH165 GH063 GH140 GH108 GH137 GH142 GH143 GH138 PL001 PL029 PL011 PL010 PL042 PL026 GH139 PL004 GH148 GH102 GH091 PL022 GH161 GH103 GH151 PL035 GH098 PL009 GH128 GH081 GH055 GH050 GH163 GH117 GH037 GH049 GH064 GH011 GH111 GH116 PL037 GH074 GH059 PL017 PL038 PL006 GH066 p__Firmicutes_Ap__Bacteroidotap__Firmicutesp__Cyanobacteriap__Desulfobacterotap__Proteobacteriap__Firmicutes_Bf__Lachnospiraceaef__Rikenellaceaef__Bacteroidaceaef__Butyricicoccaceaef__CAG.508f__Acutalibacteraceaef__Erysipelatoclostridiaceaef__CAG.74f__Barnesiellaceaef__Marinifilaceaef__Erysipelotrichaceaef__CAG.917f__DTU072f__CAG.138f__CAG.314f__Anaeroplasmataceaef__Gastranaerophilaceaef__RUG14156f__UBA660f__Streptococcaceaef__CAG.313f__CAG.631f__Peptostreptococcaceaef__Anaerovoracaceaef__Anaerotignaceaef__f__UBA7702f__Desulfovibrionaceaef__CAG.239f__Ruminococcaceaef__Oscillospiraceaef__Lactobacillaceae TSS+CLR −4 −2 0 2 4 matrix_15 0 0.5 1 MAG counts 0 20 40 60 ******** **** **** ** **** **** **** **** **** **** **** **** **** **** **** *** **** **** *** **** **** *** **** **** **** **** **** **** **** **** **** **** **** **** **** **** CYCLE 2 (n = 58) TOTAL GLYCOSIDE HYDROLASE (GH) AND POLYSACCHARIDE LYASE (PL) ABUNDANCE ************ ** ****Anova, p = 3.2e−11 1.5 2.0 2.5 3.0 T1 T2 T3 T4 T5 T6 Treatment GH (TSS + CLR) ************** ****Anova, p = 3.2e−11 −2.5 −2.0 −1.5 T1 T2 T3 T4 T5 T6 Treatment PL (TSS + CLR) Anova, p = 0.45 2.1 2.4 2.7 T1 T2 T3 T4 T5 T6 Treatment GH (TSS + CLR) Anova, p = 0.45 −2.7 −2.4 −2.1 T1 T2 T3 T4 T5 T6 Treatment PL (TSS + CLR) ************ ** ****Anova, p = 3.2e−11 1.5 2.0 2.5 3.0 T1 T2 T3 T4 T5 T6 Treatment GH (TSS + CLR) ************** ****Anova, p = 3.2e−11 −2.5 −2.0 −1.5 T1 T2 T3 T4 T5 T6 Treatment PL (TSS + CLR) Anova, p = 0.45 2.1 2.4 2.7 T1 T2 T3 T4 T5 T6 Treatment GH (TSS + CLR) Anova, p = 0.45 −2.7 −2.4 −2.1 T1 T2 T3 T4 T5 T6 Treatment PL (TSS + CLR) CYCLE 1 (n = 60) C2 C1 T6 T2 T3 T1 T4 T5 T6 T3 T2 T1 T4 T5 Sugars/Starch Cellulose Hemicellulose Pectin Beta −glucansFructans Other glucans Others GH013 GH002 GH043 GH003 GH109 GH001 GH025 GH032 GH036 GH031 GH073 GH077 GH020 GH092 GH023 GH029 GH004 GH065 GH018 GH097 GH028 GH030 GH078 GH033 GH051 GH105 GH042 GH094 GH084 GH016 GH038 GH005 GH130 GH127 GH095 GH027 GH133 GH057 GH170 GH008 GH026 GH089 GH035 GH088 GH171 GH136 GH146 GH110 GH125 GH123 GH112 GH101 GH154 GH085 GH024 PL008 GH141 GH010 GH115 GH015 PL012 GH079 PL033 GH068 GH076 GH144 GH159 GH120 GH053 GH039 GH106 GH067 GH165 GH063 GH140 GH108 GH137 GH142 GH143 GH138 PL001 PL029 PL011 PL010 PL042 PL026 GH139 PL004 GH148 GH102 GH091 PL022 GH161 GH103 GH151 PL035 GH098 PL009 GH128 GH081 GH055 GH050 GH163 GH117 GH037 GH049 GH064 GH011 GH111 GH116 PL037 GH074 GH059 PL017 PL038 PL006 GH066 p__Firmicutes_A p__Bacteroidota p__Firmicutes p__Cyanobacteria p__Desulfobacterota p__Proteobacteria p__Firmicutes_B f__Lachnospiraceae f__Rikenellaceae f__Bacteroidaceae f__Butyricicoccaceae f__CAG.508 f__Acutalibacteraceae f__Erysipelatoclostridiaceae f__CAG.74 f__Barnesiellaceae f__Marinifilaceae f__Erysipelotrichaceae f__CAG.917 f__DTU072 f__CAG.138 f__CAG.314 f__Anaeroplasmataceae f__Gastranaerophilaceae f__RUG14156 f__UBA660 f__Streptococcaceae f__CAG.313 f__CAG.631 f__Peptostreptococcaceae f__Anaerovoracaceae f__Anaerotignaceae f__ f__UBA7702 f__Desulfovibrionaceae f__CAG.239 f__Ruminococcaceae f__Oscillospiraceae f__Lactobacillaceae TSS+CLR −4 −2 0 2 4 matrix_15 0 0.5 1 MAG counts 0 20 40 60 **** ******** **** ** **** **** **** **** **** **** **** **** **** **** **** *** **** **** *** **** **** *** **** **** **** **** **** **** **** **** **** **** **** **** **** **** C2 C1 T6 T2 T3 T1 T4 T5 T6 T3 T2 T1 T4 T5 Sugars/StarchCelluloseHemicellulosePectinBeta −glucans FructansOther glucansOthers GH013 GH002 GH043 GH003 GH109 GH001 GH025 GH032 GH036 GH031 GH073 GH077 GH020 GH092 GH023 GH029 GH004 GH065 GH018 GH097 GH028 GH030 GH078 GH033 GH051 GH105 GH042 GH094 GH084 GH016 GH038 GH005 GH130 GH127 GH095 GH027 GH133 GH057 GH170 GH008 GH026 GH089 GH035 GH088 GH171 GH136 GH146 GH110 GH125 GH123 GH112 GH101 GH154 GH085 GH024 PL008 GH141 GH010 GH115 GH015 PL012 GH079 PL033 GH068 GH076 GH144 GH159 GH120 GH053 GH039 GH106 GH067 GH165 GH063 GH140 GH108 GH137 GH142 GH143 GH138 PL001 PL029 PL011 PL010 PL042 PL026 GH139 PL004 GH148 GH102 GH091 PL022 GH161 GH103 GH151 PL035 GH098 PL009 GH128 GH081 GH055 GH050 GH163 GH117 GH037 GH049 GH064 GH011 GH111 GH116 PL037 GH074 GH059 PL017 PL038 PL006 GH066 p__Firmicutes_Ap__Bacteroidotap__Firmicutesp__Cyanobacteriap__Desulfobacterotap__Proteobacteriap__Firmicutes_Bf__Lachnospiraceaef__Rikenellaceaef__Bacteroidaceaef__Butyricicoccaceaef__CAG.508f__Acutalibacteraceaef__Erysipelatoclostridiaceaef__CAG.74f__Barnesiellaceaef__Marinifilaceaef__Erysipelotrichaceaef__CAG.917f__DTU072f__CAG.138f__CAG.314f__Anaeroplasmataceaef__Gastranaerophilaceaef__RUG14156f__UBA660f__Streptococcaceaef__CAG.313f__CAG.631f__Peptostreptococcaceaef__Anaerovoracaceaef__Anaerotignaceaef__f__UBA7702f__Desulfovibrionaceaef__CAG.239f__Ruminococcaceaef__Oscillospiraceaef__Lactobacillaceae TSS+CLR −4 −2 0 2 4 matrix_15 0 0.5 1 MAG counts 0 20 40 60 ******** **** **** ** **** **** **** **** **** **** **** **** **** **** **** *** **** **** *** **** **** *** **** **** **** **** **** **** **** **** **** **** **** **** **** **** YES NO 0 20 40 60PHYLUM (n = 7) FAMILY (n = 32) SUBSTRATE A B C D E Number of MAGs .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint Centralised Log Ratio (TSS + CLR). Significance: p<0.0001 (****), p < 0.001 (***), p < 0.01 (**), p < 485 0.05 (*). White and red significance indicates decrease and increase of MAGs in treatments T2 to T6 486 compared to T1, respectively. 487 488 489 Figure 6. Mean normalised abundance of proteases across treatment groups and cycles. (A) Mean 490 abundance of individual protease families (MEROPS ID) grouped according to protease type in 491 accordance to MEROPS database [44]. MEROPS protease ID in bold and underlined indicates 492 significant increase or decrease in abundance compared to T1, based on Wald-Test (using DESEq2). 493 Number of MAGs with protease gene grouped according to taxonomic ranks (B) phylum and (C) 494 Metallo Serine Cysteine Aspartic Unknown Inhibitor Threonine Asparagine Mixed T4 T2 T3 T5 T1 T6 P−value TSS+CLR −4 −2 0 2 4 Metallo Serine Cysteine Aspartic Unknown Inhibitor Threonine Asparagine Mixed T1 T3 T2 T5 T6 T4 P−value TSS+CLR −4 −2 0 2 4 CYCLE 1 (n = 60) CYCLE 2 (n = 58) * * * A B C D TOTAL PROTEASE FAMILY ABUNDANCE C2 C1 T6 T2 T3 T1 T4 T5 T6 T3 T2 T1 T4 T5 Sugars/Starch Cellulose Hemicellulose Pectin Beta −glucansFructans Other glucans Others GH013 GH002 GH043 GH003 GH109 GH001 GH025 GH032 GH036 GH031 GH073 GH077 GH020 GH092 GH023 GH029 GH004 GH065 GH018 GH097 GH028 GH030 GH078 GH033 GH051 GH105 GH042 GH094 GH084 GH016 GH038 GH005 GH130 GH127 GH095 GH027 GH133 GH057 GH170 GH008 GH026 GH089 GH035 GH088 GH171 GH136 GH146 GH110 GH125 GH123 GH112 GH101 GH154 GH085 GH024 PL008 GH141 GH010 GH115 GH015 PL012 GH079 PL033 GH068 GH076 GH144 GH159 GH120 GH053 GH039 GH106 GH067 GH165 GH063 GH140 GH108 GH137 GH142 GH143 GH138 PL001 PL029 PL011 PL010 PL042 PL026 GH139 PL004 GH148 GH102 GH091 PL022 GH161 GH103 GH151 PL035 GH098 PL009 GH128 GH081 GH055 GH050 GH163 GH117 GH037 GH049 GH064 GH011 GH111 GH116 PL037 GH074 GH059 PL017 PL038 PL006 GH066 p__Firmicutes_A p__Bacteroidota p__Firmicutes p__Cyanobacteria p__Desulfobacterota p__Proteobacteria p__Firmicutes_B f__Lachnospiraceae f__Rikenellaceae f__Bacteroidaceae f__Butyricicoccaceae f__CAG.508 f__Acutalibacteraceae f__Erysipelatoclostridiaceae f__CAG.74 f__Barnesiellaceae f__Marinifilaceae f__Erysipelotrichaceae f__CAG.917 f__DTU072 f__CAG.138 f__CAG.314 f__Anaeroplasmataceae f__Gastranaerophilaceae f__RUG14156 f__UBA660 f__Streptococcaceae f__CAG.313 f__CAG.631 f__Peptostreptococcaceae f__Anaerovoracaceae f__Anaerotignaceae f__ f__UBA7702 f__Desulfovibrionaceae f__CAG.239 f__Ruminococcaceae f__Oscillospiraceae f__Lactobacillaceae TSS+CLR −4 −2 0 2 4 matrix_15 0 0.5 1 MAG counts 0 20 40 60 **** ******** **** ** **** **** **** **** **** **** **** **** **** **** **** *** **** **** *** **** **** *** **** **** **** **** **** **** **** **** **** **** **** **** **** **** Aspartic Cysteine Inhibitor Metallo Asparagine Mixed Serine Threonine Unknown T1 T2 T3 T4 T5 T6 TSS+CLR −4 −20 2 4 Aspartic Cysteine Inhibitor Metallo Asparagine Mixed Serine Threonine Unknown T1 T2 T3 T4 T5 T6 TSS+CLR −4 −20 2 4 Aspartic Cysteine Inhibitor Metallo Asparagine Mixed Serine Threonine Unknown T1 T2 T3 T4 T5 T6 TSS+CLR −4 −20 2 4 P−value P−value 0 0.01 0.02 0.03 >0.5 P−value P−value 0 0.01 0.02 0.03 >0.5 * C2 C1 Asparagine Serine Unknown Cysteine Metallo Aspartic Threonine Inhibitor Mixed T1 T4 T6 T5 T3 T2 T1 T2 T3 T6 T4 T5 N06 N10 S11 S09X S14 S33 S26A S08A S12 S16 S54 S66 S46 S15 S09C S01C S09B S51 S09A S49B S37 S78 S13 S41A S55 S24 S49A S26X S28 U32 U57 C26 C59 C69 C01B C60B C44 C56 C108 C40 C60A C82A C15 C01A C11X C39 M38 M24A M20A M20D M50B M18 M20B M16B M20C M13 M42 M29 M79 M15B M19 M24B M03B M41 M48A M03A M20F M23B M16C M32 M01 M78 M48C M55 M17 M14A M15A M10B M14C M103 M28X M75 M48B M93 M16A M14B M20X M28A M03C M10A M26 M82 M95 A08 A25 A24X A24A A31 A36 T05 T01B T02 T01A T03 I87 I51 I04 P01 p__Firmicutes_Ap__Firmicutesp__Bacteroidotap__Cyanobacteriap__Desulfobacterotap__Proteobacteriap__Firmicutes_Bf__Lachnospiraceaef__Lactobacillaceaef__Rikenellaceaef__Oscillospiraceaef__Ruminococcaceaef__Acutalibacteraceaef__Butyricicoccaceaef__CAG.508f__Bacteroidaceaef__Erysipelatoclostridiaceaef__Erysipelotrichaceaef__CAG.138f__Anaerovoracaceaef__Streptococcaceaef__Marinifilaceaef__Barnesiellaceaef__CAG.314f__Desulfovibrionaceaef__Peptostreptococcaceaef__Anaerotignaceaef__CAG.74f__UBA7702f__ f__CAG.239f__CAG.917f__DTU072f__CAG.313f__Anaeroplasmataceaef__CAG.631f__UBA660f__RUG14156f__Gastranaerophilaceae **** **** ** **** Number of MAGs PHYLUM (n = 60) FAMILY (n = 32) 0 20 40 60 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint family, (D) Mean abundance of total protease grouped according to family across groups and cycles, 495 statistical differences were based on Kruskal-Sum Rank and posthoc pairwise Dunn testing with 496 Bonferroni correction. Normalisation method: Total Sum Scaling and Centralised Log Ratio (TSS + 497 CLR). Significance: p<0.0001 (****), p < 0.001 (***), p < 0.01 (**), p < 0.05 (*), white and red 498 significance indicates decrease and increase of MAGs in GHI treatments (T2, T3, T4, T5, T6) 499 compared to T1, respectively. 500 501 Gut microorganisms engage in complex metabolic interactions, potentially involving various 502 production of metabolites that modulate host physiology. Based on our analysis, we determined 503 that GHI administration has also an impact on other KEGG modules including various transport and 504 transporter systems, amino acid metabolism and nucleic acid metabolism pathways (Figure 8). 505 Significantly varied abundances across treatment groups in 25 and 15 different KEGG function 506 module categories in C1 and C2 were observed (p < 0.05, Kruskal-Rank Sum test). Heatmap 507 clustering also exhibits the divergence of T6 from the other groups in C1, whilst both T5 and T6 have 508 partitioned from the others in C2. However, compared to T1, abundance in GHI groups were 509 significantly downregulated in 4 module categories namely phosphotransferase system and ATP 510 synthesis in C2 – T6, serine and threonine metabolism in C2 – T5, and co-factor and vitamin 511 metabolism in C1 – T5, while methane metabolism function is upregulated in C1 – T3. Furthermore, 512 differential analysis (T1 as reference) of individual modules revealed decreased abundance of 513 antibiotic related transport systems in GHI groups C1 – T6 (M00747) and C2 – T3 (M00817, M00708) 514 and increase of organic compound biosynthesis and transport, and regulatory system modules in C1 515 – T4 (M00020, M00364, M00365), C2 – T6 (M00213), C2 – T5 (M00521). 516 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint 517 Figure 7. KEGG metabolic module abundances across treatment and cycles. Mean normalised 518 abundance of (A) KEGG module categories and respective p-values based on Kruskal-Wallis testing 519 (B) Selected KEGG modules across treatment groups and cycles (C1 and C2). Red and blue colour of 520 heatmap cells indicates high and low abundance, respectively. KEGG category and module IDs in 521 bold and underlined indicates significant difference in abundance compared to T1 based on posthoc 522 pairwise Dunn testing with Bonferroni correction and DESEq2, respectively. Normalisation method: 523 Total Sum Scaling and Centralised Log Ratio (TSS + CLR). Significance: p<0.0001 (****), p < 0.001 524 (***), p < 0.01 (**), p < 0.05 (*). Significance in white and red indicates decrease and increase of 525 MAGs in GHI treatments (T2, T3, T4, T5, T6) compared to T1, respectively. 526 527 Administration of different GHI can impact short-chain fatty acid production pathways 528 Due to the known relevance of metabolic pathways involved in the production of short-fatty acids 529 (SCFA) including acetate, butyrate and propionate on gut health, we also explored the effects of GHI 530 on SCFA production. As shown in Figure 9, the majority of KEGG submodules related to SCFA 531 production were present in our samples (excluding M00377+02). We also identified MAGs to have 532 C2 C1 T6 T5 T3 T2 T1 T4 T5 T6 T4 T2 T3 T1 Saccharide, polyol, and lipid transport system Central carbohydrate metabolism ABC−2 type and other transport systems Phosphotransferase system (PTS) Metallic cation, iron−siderophore and vitamin B12 transport system Mineral and organic ion transport system Phosphate and amino acid transport system Bacterial secretion system Cofactor and vitamin metabolism ATP synthesis Purine metabolism Ribosome Two−component regulatory system Branched−chain amino acid metabolism Carbon fixation Aminoacyl tRNA Drug efflux transporter/pump Peptide and nickel transport system Fatty acid biosynthesis and degradation RNA polymerase DNA polymerase RNA processing Other carbohydrate metabolism Aromatic amino acid metabolism Arginine and proline metabolism Pyrimidine metabolism Lysine metabolism Terpenoid backbone biosynthesis Cysteine and methionine metabolism Polyketide sugar unit biosynthesis Histidine metabolism Serine and threonine metabolism Drug resistance Lipopolysaccharide metabolism Lipid metabolism Polyamine biosynthesis Repair system Nitrogen metabolism Other amino acid metabolism Methane metabolism C1 C2 TSS+CLR −5 0 5 matrix_37 0 0.02 0.04 0.06 C2 C1 ABC−2 type and other transport systems ATP synthesis Cofactor and vitamin metabolism Drug efflux transporter/pump Methane metabolism Phosphotransferase system (PTS) Saccharide, polyol, and lipid transport system Serine and threonine metabolism Terpenoid backbone biosynthesis Two−component regulatory system M00747: Bacitracin transport system M00817: Lantibiotic transport system M00157: F−type ATPase, prokaryotes and chloroplasts M00159: V/A−type ATPase, prokaryotes M00144: NADH:quinone oxidoreductase, prokaryotes M00153: Cytochrome d ubiquinol oxidase M00149: Succinate dehydrogenase, prokaryotes M00150: Fumarate reductase, prokaryotes M00151: Cytochrome bc1 complex respiratory unit M00120: Coenzyme A biosynthesis, pantothenate => CoA M00119: Pantothenate biosynthesis, valine/L−aspartate => pantothenate M00126: Tetrahydrofolate biosynthesis, GTP => THF M00122: Cobalamin biosynthesis, cobinamide => cobalamin M00127: Thiamine biosynthesis, AIR => thiamine−P/thiamine−2P M00115: NAD biosynthesis, aspartate => NAD M00123: Biotin biosynthesis, pimeloyl−ACP/CoA => biotin M00140: C1−unit interconversion, prokaryotes M00572: Pimeloyl−ACP biosynthesis, BioC−BioH pathway, malonyl−ACP => pimeloyl−ACP M00125: Riboflavin biosynthesis, GTP => riboflavin/FMN/FAD M00573: Biotin biosynthesis, BioI pathway, long−chain−acyl−ACP => pimeloyl−ACP => biotin M00577: Biotin biosynthesis, BioW pathway, pimelate => pimeloyl−CoA => biotin M00121: Heme biosynthesis, glutamate => protoheme/siroheme M00116: Menaquinone biosynthesis, chorismate => menaquinone M00708: Multidrug resistance, PatAB transporter M00345: Formaldehyde assimilation, ribulose monophosphate pathway M00276: PTS system, mannose−specific II component M00269: PTS system, sucrose−specific II component M00270: PTS system, trehalose−specific II component M00273: PTS system, fructose−specific II component M00279: PTS system, galactitol−specific II component M00266: PTS system, maltose/glucose−specific II component M00271: PTS system, beta−glucoside−specific II component M00280: PTS system, glucitol/sorbitol−specific II component M00268: PTS system, alpha−glucoside−specific II component M00275: PTS system, cellobiose−specific II component M00274: PTS system, mannitol−specific II component M00806: PTS system, maltose−specific II component M00303: PTS system, N−acetylmuramic acid−specific II component M00265: PTS system, glucose−specific II component M00272: PTS system, beta−glucoside (arbutin/salicin/cellobiose)−specific II component M00283: PTS system, ascorbate−specific II component M00277: PTS system, N−acetylgalactosamine−specific II component M00287: PTS system, galactosamine−specific II component M00807: PTS system, galactose−specific II component M00809: PTS system, glucose−specific II component M00610: PTS system, D−glucosaminate−specific II component M00278: PTS system, sorbose−specific II component M00764: PTS system, fructoselysine/glucoselysine−specific II component M00267: PTS system, N−acetylglucosamine−specific II component M00213: L−Arabinose transport system M00018: Threonine biosynthesis, aspartate => homoserine => threonine M00020: Serine biosynthesis, glycerate−3P => serine M00364: C10−C20 isoprenoid biosynthesis, bacteria M00365: C10−C20 isoprenoid biosynthesis, archaea M00521: CiaH−CiaR two−component regulatory system T1 T5 T4 T2 T6 T3 T4 T6 T5 T1 T3 T2 TSS+CLR −4 −2 0 2 4 C2 C1 ABC−2 type and other transport systems ATP synthesis Cofactor and vitamin metabolism Drug efflux transporter/pump Methane metabolism Phosphotransferase system (PTS) Saccharide, polyol, and lipid transport system Serine and threonine metabolism Terpenoid backbone biosynthesis Two−component regulatory system M00747: Bacitracin transport system M00817: Lantibiotic transport system M00157: F−type ATPase, prokaryotes and chloroplasts M00159: V/A−type ATPase, prokaryotes M00144: NADH:quinone oxidoreductase, prokaryotes M00153: Cytochrome d ubiquinol oxidase M00149: Succinate dehydrogenase, prokaryotes M00150: Fumarate reductase, prokaryotes M00151: Cytochrome bc1 complex respiratory unit M00120: Coenzyme A biosynthesis, pantothenate => CoA M00119: Pantothenate biosynthesis, valine/L−aspartate => pantothenate M00126: Tetrahydrofolate biosynthesis, GTP => THF M00122: Cobalamin biosynthesis, cobinamide => cobalamin M00127: Thiamine biosynthesis, AIR => thiamine−P/thiamine−2P M00115: NAD biosynthesis, aspartate => NAD M00123: Biotin biosynthesis, pimeloyl−ACP/CoA => biotin M00140: C1−unit interconversion, prokaryotes M00572: Pimeloyl−ACP biosynthesis, BioC−BioH pathway, malonyl−ACP => pimeloyl−ACP M00125: Riboflavin biosynthesis, GTP => riboflavin/FMN/FAD M00573: Biotin biosynthesis, BioI pathway, long−chain−acyl−ACP => pimeloyl−ACP => biotin M00577: Biotin biosynthesis, BioW pathway, pimelate => pimeloyl−CoA => biotin M00121: Heme biosynthesis, glutamate => protoheme/siroheme M00116: Menaquinone biosynthesis, chorismate => menaquinone M00708: Multidrug resistance, PatAB transporter M00345: Formaldehyde assimilation, ribulose monophosphate pathway M00276: PTS system, mannose−specific II component M00269: PTS system, sucrose−specific II component M00270: PTS system, trehalose−specific II component M00273: PTS system, fructose−specific II component M00279: PTS system, galactitol−specific II component M00266: PTS system, maltose/glucose−specific II component M00271: PTS system, beta−glucoside−specific II component M00280: PTS system, glucitol/sorbitol−specific II component M00268: PTS system, alpha−glucoside−specific II component M00275: PTS system, cellobiose−specific II component M00274: PTS system, mannitol−specific II component M00806: PTS system, maltose−specific II component M00303: PTS system, N−acetylmuramic acid−specific II component M00265: PTS system, glucose−specific II component M00272: PTS system, beta−glucoside (arbutin/salicin/cellobiose)−specific II component M00283: PTS system, ascorbate−specific II component M00277: PTS system, N−acetylgalactosamine−specific II component M00287: PTS system, galactosamine−specific II component M00807: PTS system, galactose−specific II component M00809: PTS system, glucose−specific II component M00610: PTS system, D−glucosaminate−specific II component M00278: PTS system, sorbose−specific II component M00764: PTS system, fructoselysine/glucoselysine−specific II component M00267: PTS system, N−acetylglucosamine−specific II component M00213: L−Arabinose transport system M00018: Threonine biosynthesis, aspartate => homoserine => threonine M00020: Serine biosynthesis, glycerate−3P => serine M00364: C10 −C20 isoprenoid biosynthesis, bacteria M00365: C10−C20 isoprenoid biosynthesis, archaea M00521: CiaH−CiaR two−component regulatory system T1 T5 T4 T2 T6 T3 T4 T6 T5 T1 T3 T2 TSS+CLR −4 −2 0 2 4 C2 C1 ABC−2 type and other transport systems ATP synthesis Cofactor and vitamin metabolism Drug efflux transporter/pump Methane metabolism Phosphotransferase system (PTS) Saccharide, polyol, and lipid transport system Serine and threonine metabolism Terpenoid backbone biosynthesis Two−component regulatory system M00747: Bacitracin transport system M00817: Lantibiotic transport system M00157: F−type ATPase, prokaryotes and chloroplasts M00159: V/A−type ATPase, prokaryotes M00144: NADH:quinone oxidoreductase, prokaryotes M00153: Cytochrome d ubiquinol oxidase M00149: Succinate dehydrogenase, prokaryotes M00150: Fumarate reductase, prokaryotes M00151: Cytochrome bc1 complex respiratory unit M00120: Coenzyme A biosynthesis, pantothenate => CoA M00119: Pantothenate biosynthesis, valine/L−aspartate => pantothenate M00126: Tetrahydrofolate biosynthesis, GTP => THF M00122: Cobalamin biosynthesis, cobinamide => cobalamin M00127: Thiamine biosynthesis, AIR => thiamine−P/thiamine−2P M00115: NAD biosynthesis, aspartate => NAD M00123: Biotin biosynthesis, pimeloyl−ACP/CoA => biotin M00140: C1−unit interconversion, prokaryotes M00572: Pimeloyl−ACP biosynthesis, BioC−BioH pathway, malonyl−ACP => pimeloyl−ACP M00125: Riboflavin biosynthesis, GTP => riboflavin/FMN/FAD M00573: Biotin biosynthesis, BioI pathway, long−chain−acyl−ACP => pimeloyl−ACP => biotin M00577: Biotin biosynthesis, BioW pathway, pimelate => pimeloyl−CoA => biotin M00121: Heme biosynthesis, glutamate => protoheme/siroheme M00116: Menaquinone biosynthesis, chorismate => menaquinone M00708: Multidrug resistance, PatAB transporter M00345: Formaldehyde assimilation, ribulose monophosphate pathway M00276: PTS system, mannose−specific II component M00269: PTS system, sucrose−specific II component M00270: PTS system, trehalose−specific II component M00273: PTS system, fructose−specific II component M00279: PTS system, galactitol−specific II component M00266: PTS system, maltose/glucose−specific II component M00271: PTS system, beta−glucoside−specific II component M00280: PTS system, glucitol/sorbitol−specific II component M00268: PTS system, alpha−glucoside−specific II component M00275: PTS system, cellobiose−specific II component M00274: PTS system, mannitol−specific II component M00806: PTS system, maltose−specific II component M00303: PTS system, N−acetylmuramic acid−specific II component M00265: PTS system, glucose−specific II component M00272: PTS system, beta−glucoside (arbutin/salicin/cellobiose)−specific II component M00283: PTS system, ascorbate−specific II component M00277: PTS system, N−acetylgalactosamine−specific II component M00287: PTS system, galactosamine−specific II component M00807: PTS system, galactose−specific II component M00809: PTS system, glucose−specific II component M00610: PTS system, D−glucosaminate−specific II component M00278: PTS system, sorbose−specific II component M00764: PTS system, fructoselysine/glucoselysine−specific II component M00267: PTS system, N−acetylglucosamine−specific II component M00213: L−Arabinose transport system M00018: Threonine biosynthesis, aspartate => homoserine => threonine M00020: Serine biosynthesis, glycerate−3P => serine M00364: C10−C20 isoprenoid biosynthesis, bacteria M00365: C10−C20 isoprenoid biosynthesis, archaea M00521: CiaH−CiaR two−component regulatory system T1 T5 T4 T2 T6 T3 T4 T6 T5 T1 T3 T2 TSS+CLR −4 −2 0 2 4 C1 C2 T1 T2 T3 T4 T5 T6 T1 T2 T3 T4 T5 T6 Central carbohydrate metabolism Other carbohydrate metabolism Arginine and proline metabolism Lysine metabolism Cysteine and methionine metabolism Serine and threonine metabolism Branched−chain amino acid metabolism Aromatic amino acid metabolism Histidine metabolism Other amino acid metabolism Pyrimidine metabolism Purine metabolism Lipopolysaccharide metabolism Lipid metabolism Fatty acid biosynthesis and degradation Terpenoid backbone biosynthesis Cofactor and vitamin metabolism Polyamine biosynthesis ATP synthesis Carbon fixation Methane metabolism Nitrogen metabolism Ribosome RNA polymerase Mineral and organic ion transport system Saccharide, polyol, and lipid transport system Phosphate and amino acid transport system ABC−2 type and other transport systems Peptide and nickel transport system Metallic cation, iron−siderophore and vitamin B12 transport system DNA polymerase Phosphotransferase system (PTS) Repair system Bacterial secretion system Aminoacyl tRNA RNA processing Two−component regulatory system Drug resistance Drug efflux transporter/pump Polyketide sugar unit biosynthesis C1 C2 TSS+CLR −5 0 5 matrix_28 0 0.02 0.04 0.06-5 0 5 TSS +CLR C2 C1 T6 T5 T3 T2 T1 T4 T5 T6 T4 T2 T3 T1 Saccharide, polyol, and lipid transport system Central carbohydrate metabolism ABC−2 type and other transport systems Phosphotransferase system (PTS) Metallic cation, iron−siderophore and vitamin B12 transport system Mineral and organic ion transport system Phosphate and amino acid transport system Bacterial secretion system Cofactor and vitamin metabolism ATP synthesis Purine metabolism Ribosome Two−component regulatory system Branched−chain amino acid metabolism Carbon fixation Aminoacyl tRNA Drug efflux transporter/pump Peptide and nickel transport system Fatty acid biosynthesis and degradation RNA polymerase DNA polymerase RNA processing Other carbohydrate metabolism Aromatic amino acid metabolism Arginine and proline metabolism Pyrimidine metabolism Lysine metabolism Terpenoid backbone biosynthesis Cysteine and methionine metabolism Polyketide sugar unit biosynthesis Histidine metabolism Serine and threonine metabolism Drug resistance Lipopolysaccharide metabolism Lipid metabolism Polyamine biosynthesis Repair system Nitrogen metabolism Other amino acid metabolism Methane metabolism C1 C2 TSS+CLR −5 0 5 matrix_26 0 0.02 0.04 0.06 KEGG MODULE CATEGORY P-VALUES A B *** C1 C2 T1 T2 T3 T4 T5 T6 T1 T2 T3 T4 T5 T6 Central carbohydrate metabolism Other carbohydrate metabolism Arginine and proline metabolism Lysine metabolism Cysteine and methionine metabolism Serine and threonine metabolism Branched−chain amino acid metabolism Aromatic amino acid metabolism Histidine metabolism Other amino acid metabolism Pyrimidine metabolism Purine metabolism Lipopolysaccharide metabolism Lipid metabolism Fatty acid biosynthesis and degradation Terpenoid backbone biosynthesis Cofactor and vitamin metabolism Polyamine biosynthesis ATP synthesis Carbon fixation Methane metabolism Nitrogen metabolism Ribosome RNA polymerase Mineral and organic ion transport system Saccharide, polyol, and lipid transport system Phosphate and amino acid transport system ABC−2 type and other transport systems Peptide and nickel transport system Metallic cation, iron−siderophore and vitamin B12 transport system DNA polymerase Phosphotransferase system (PTS) Repair system Bacterial secretion system Aminoacyl tRNA RNA processing Two−component regulatory system Drug resistance Drug efflux transporter/pump Polyketide sugar unit biosynthesis C1 C2 TSS+CLR −5 0 5 matrix_28 0 0.02 0.04 0.06-4 0 4 TSS +CLR KEGG MODULE ID ** * * *** * *** *** * **** **** * ** Phosphotransferase system Cofactor and vitamin metabolism ATP synthesis Serine and threonine metabolism Methane metabolism M00747: Bacitracin transport system M00817: Lantibiotic transport system M00708: Multidrug resistance, PatAB transporter M00213: L-Arabinose transport system M00020: Serine biosynthesis, glycerate-3P => serine M00364: C10– C20 isoprenoid biosynthesis, bacteria M00521: CiaH-CiaR two-component regulatory system M00365: C10– C20 isoprenoid biosynthesis, archaea .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint the complete sets of modules for Acetyl-CoA pathway (48/84 MAGs) and butyrate pathways 533 (bin.145); However, none of our MAGs possessed complete sets of submodules for Wood-Ljungdahl 534 and propionate pathways. Bin.92 and bin.215 have the highest number of submodules (n = 5), while 535 MAGs bin.91 and bin.234 are missing M00173+03 – submodule for pyruvate carboxylase (pyc). 536 Across our treatment groups, 11 out of the included 19 submodules are significantly varied, wherein 537 propionate submodules were observed as mostly affected. From this, post hoc pairwise comparison 538 revealed a single elevation of M00377+03 in C1-T5. Furthermore, distinct separation of T5 in C1 and 539 distancing of T1 and T6 from other groups in C2 were observed. 540 541 542 Figure 8. Metagenomic abundance of elements of short-chain fatty acids (SCFA) pathways. 543 (A) Selected MAGs containing most complete SCFA submodules. Blue and white colour indicates 544 presence and absence of element, respectively; (B) Heatmap of normalised abundances of SCFA 545 enzyme corresponding module IDs across treatment groups and cycles (C1 and C2), with significance 546 based on Kruskal-Rank Sum test (p <0.05 in blue). TSS + CLR: Total Sum Scaling and Centralised Log 547 Ratio. Red and blue colour of heatmap cells indicates high and low abundance, respectively; (C) SCFA 548 pathways showing KEGG orthology (KO) numbers and enzyme names (in red). Submodule and 549 significance in red: significant increase based on pairwise Dunn test with Bonferroni correction (p < 550 fhs (K01938) Formate Formyl-THF foID (K01491) Methylene=THF Methenyl-THF Methyl-THF Acetyl-CoA THF Acetyl phosphate ACETATE MTHFR (K00297) acsE (K15023) acsB (K14138) pta (K00625) ackA (K00925) Pyruvate Acetyl-CoA por (K00169, K00170) Aceto-acetyl CoA atoB (K00626) Crotonyl -CoA 3-hydroxybutyryl - CoA Butyryl -CoA Butyryl -P BUTYRATE croR (K17865) hbd bcd (K00248) ptb buk CO2M00377 + 02 fdhA/fdhB (K05299, K15022) M00377+03 M00377+04 M00377+05 M00377+06 M00377+07 M00357+02 M00357+01 M00374+01 M00374+13 M00374+03 M00013+01 M00173+03 M00173+04 M00173+05 M00173+06 M00009+05 M00376+06 M00376+05 M00373+09 Pyruvate Oxaloacetate Malic acid Fumaric acid Succinate Succinyl -CoA S-Methyl malonyl -CoA Propanoyl -CoA pyc (K01958, K01959, K01960) mdh (K00024) fum (K01676, K01677, K01678) sdhA (K00239, K00240) sucDC /sarC (K01902, K01903, K18118) MUTAB/mcmA12 (K01847, K01848, K01849) MCEE (K05606) pccAB (K01965, K01966) Methylmalonyl -CoA PROPIONATEpct Propionyl -CoA mmdA SUCCINATE PATHWAY CLASSICAL PATHWAY WOOD - LJUNGDAHL PATHWAY ACETATEBUTYRATEPROPIONATE Acetyl -CoA pathway C A B P bin.106bin.148bin.215bin.227bin.24bin.286bin.292bin.301bin.302bin.31bin.92bin.145bin.234bin.91C1 C2 C1 C2 M00377+03 M00377+04 M00377+05 M00377+06 M00377+07 M00357+02 M00357+01 M00374+01 M00374+13 M00373+03 M00013+01 M00173+03 M00173+04 M00173+05 M00173+06 M00009+05 M00376+06 M00376+05 M00373+09 T5 T1 T6 T2 T3 T4 T1 T2 T3 T5 T4 T6 TSS+CLR 0 0.5 1 TSS+CLR −2 −1 0 1 2 Gene occurrence Present AbsentA TSS + CLR -4 -2 0 2 4 C D E bin.148 bin.106 bin.215 bin.227 bin.24 bin.301 bin.302 bin.31 bin.92 bin.286 bin.292 bin.145 bin.239 bin.91 bin.96 bin.234 bin.110 bin.148 bin.11 bin.251 bin.108 C1 C2 M00377+03 M00377+04 M00377+05 M00377+06 M00377+07 M00357+02 M00357+01 M00374+01 M00374+13 M00373+03 M00013+01 M00173+03 M00173+04 M00173+05 M00173+06 M00009+05 M00376+06 M00376+05 M00373+09 T1 T2 T3 T4 T5 T6 T1 T2 T3 T4 T5 T6 TSS+CLR 0 0.5 1 TSS+CLR −4 −2 0 2 4 B P−value P−value 0 0.01 0.02 0.03 >0.5 * .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint 0.05). Submodules in blue: not detected in metagenome dataset in this study. Enzyme names in 551 black: no matching submodule ID in KEGG database [41]. Enzymes are listed as follows: fdhA: 552 Formate dehydrogenase alpha subunit; fdhB: formate dehydrogenase beta subunit; fhs: formate 553 tetrahydrofolate ligase; foID: methylenetetrahydrofolate dehydrogenase; MTHFR: 554 methylenetetrahydrofolate reductase; acsE: 5-methyltetrahydrofolate corrinoid/iron sulfur protein 555 methyltransferase; acsB: acetyl-CoA synthase; pta: phosphate acetyltransferase; ackA: acetate 556 kinase; por: pyruvate ferredoxin oxidoreductase; atoB: acetyl-CoA C-acetyltransferase; hbd: 3-557 hydroxybutyryl-CoA dehydrogenase; croR: 3-hydroxybutyryl-CoA dehydratase; bcd: butyryl-CoA 558 dehydrogenase; ptb: phosphate butyryltransferase; pyc: pyruvate carboxylase; buk: butyrate kinase; 559 mdh; malate dehydrogenase; fum: fumarate hydratase; sdhA: succinate dehydrogenase; sucD: 560 succinyl-CoA synthetase alpha subunit; sucD: succinyl-CoA synthetase beta subunit; aarC: succinyl-561 CoA:acetate CoA-transferase; MUTAB; methylmalonyl-CoA mutase alpha and beta; mcmA1: 562 methylmalonyl-CoA mutase, N-terminal domain; mcmA2: methylmalonyl-CoA mutase, C-terminal 563 domain; mmdA: methylmalonyl-CoA decarboxylase subunit alpha; MCEE: epi; methylmalonyl-564 CoA/ethylmalonyl-CoA epimerase; pccA: propionyl-CoA carboxylase alpha chain; pccB: propionyl-565 CoA carboxylase beta chain; pct: propionate CoA-transferase. 566 567 MAGs associated with better performance have higher capacity for nutrient digestion and 568 metabolism 569 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint 570 Figure 9. Overview of significant metagenomic features of 52 MAGs associated with EPEF and GHIs. 571 EPEF: European Poultry Efficiency Factor, GH: Glycoside hydrolase, GHI: Gut health interventions, 572 GLLVM: Generalised latent linear variable model, MAG: Metagenome-assembled genomes, SCFA: 573 Short chain fatty acids, PL: Polysaccharide lyase. 574 575 Based on above analysis, we explored the gene abundance of several metagenomic features (which 576 were selected based on statistical significance among treatment groups) of 52 MAGs that were 577 revealed to be associated with both EPEF and composition in GHI groups (Figure 9). Clustering 578 analysis showed that MAGs with positive association with EPEF (EPEF+) have an overall higher 579 number of genes encoding for the various significant metabolic features. Specifically, the EPEF+ 580 group demonstrated higher genes in the PTS module, Butyrate and Propionate SCFA module while 581 similar abundance in CAZymes and proteases between the two EPEF groups can be seen. However, it 582 is interesting to point out that Rikenella microfusus, the MAG with highest GLLVM coefficient for 583 EPEF, do not possess any genes for PTS and butyrate production modules. 584 585 PositiveNegative bin.24:Fusicatenibacter;s__ bin.137:Rikenella microfusus bin.215:Blautia_A sp002159835 bin.92:Lachnospiraceae;g__UBA7160;s__ bin.19:Acutalibacteraceae;g__An172;s__An172 sp002160515 bin.34:Alistipes communis bin.125:Lactobacillus crispatus bin.258:Lactobacillus gallinarum bin.120:Lactobacillus johnsonii bin.156:Anaerostipes;s__ bin.110:Butyricicoccus;s__ bin.292:Gemmiger;s__ bin.58:Butyricicoccus;s__ bin.81:Enterocloster sp900547035 bin.93:Oscillospiraceae;g__UBA5446;s__UBA5446 sp004553625 bin.27:Lachnospiraceae;g__OF09−33XD;s__ bin.112:Merdibacter sp900754715 bin.99:Limosilactobacillus;s__ bin.267:Limosilactobacillus vaginalis bin.26:Bilophila bin.97:Clostridia;o__TANB77;f__CAG−508;g__CAG−269;s__ bin.84:Ruminococcaceae;g__UBA3818;s__ bin.151:Bacilli;o__RF39;f__UBA660;g__CAG−822;s__ bin.65:RF39;f__UBA660;g__CAG−302;s__ bin.96:Phocaeicola vulgatus bin.91:Bacteroides fragilis bin.234:Barnesiella viscericola bin.119:Christensenellales;f__CAG−74;g__OEMS01;s__ bin.291:Alistipes megaguti bin.286:Odoribacter splanchnicus bin.302:Mediterraneibacter;s__ bin.205:Lachnospiraceae;g__UMGS1370;s__ bin.294:Alistipes sp900290115 bin.279:Christensenellales;f__DTU072;g__;s__ bin.208:Christensenellales;f__CAG−917;g__UMGS1688;s__UMGS1688 sp900554085 bin.198:Anaerobutyricum;s__ bin.74:Oscillospiraceae;g__CAG−110;s__ bin.289:Ruminococcus_G;s__ bin.163:Intestinimonas;s__ bin.254:RFN20;f__CAG−631;g__CAG−631;s__CAG−631 sp900762575 bin.179:Acutalibacteraceae;g__UMGS403;s__UMGS403 sp900541565 bin.168:Oscillospiraceae;g__CAG−110;s__CAG−110 sp900546915 bin.83:Romboutsia;s__ bin.213:Tidjanibacter inops_A bin.79:Gastranaerophilales;f__RUG14156;g__;s__ bin.266:Gastranaerophilales;f__RUG14156;g__;s__ bin.152:Ligilactobacillus salivarius bin.11:Christensenellales;f__CAG−138;g__UMGS946;s__UMGS946 sp900544965 bin.218:Anaerovoracaceae;g__CAG−145;s__CAG−145 sp900754795 bin.239:Anaerotignum;s__ bin.108:RF32;f__CAG−239;g__CAG−267;s__CAG−267 sp001917135 bin.200:Clostridia;o__TANB77;f__CAG−508;g__UMGS1994;s__ GH PL Inhibitor Cysteine Threonine Phosphotransferase system (PTS) Cofactor and vitamin metabolism ATP.synthesis Serine and threonine metabolism Methane metabolism Saccharide, polyol, and lipid transport system Two-component regulatory system Terpenoid.backbone.biosynthesis Acetate Butyrate Propionate EPEF Coef C1 C2 Number of genes 0 50 100 150 200 EPEF association −0.6 −0.4 −0.2 0 0.2 CAZymes Proteases KEGG Metabolic modules SCFA Association with MAGs (GLLVM) PositiveNegative bin.24:Fusicatenibacter;s__ bin.137:Rikenella microfusus bin.215:Blautia_A sp002159835 bin.92:Lachnospiraceae;g__UBA7160;s__ bin.19:Acutalibacteraceae;g__An172;s__An172 sp002160515 bin.34:Alistipes communis bin.125:Lactobacillus crispatus bin.258:Lactobacillus gallinarum bin.120:Lactobacillus johnsonii bin.156:Anaerostipes;s__ bin.110:Butyricicoccus;s__ bin.292:Gemmiger;s__ bin.58:Butyricicoccus;s__ bin.81:Enterocloster sp900547035 bin.93:Oscillospiraceae;g__UBA5446;s__UBA5446 sp004553625 bin.27:Lachnospiraceae;g__OF09−33XD;s__ bin.112:Merdibacter sp900754715 bin.99:Limosilactobacillus;s__ bin.267:Limosilactobacillus vaginalis bin.26:Bilophila bin.97:Clostridia;o__TANB77;f__CAG−508;g__CAG−269;s__ bin.84:Ruminococcaceae;g__UBA3818;s__ bin.151:Bacilli;o__RF39;f__UBA660;g__CAG−822;s__ bin.65:RF39;f__UBA660;g__CAG−302;s__ bin.96:Phocaeicola vulgatus bin.91:Bacteroides fragilis bin.234:Barnesiella viscericola bin.119:Christensenellales;f__CAG−74;g__OEMS01;s__ bin.291:Alistipes megaguti bin.286:Odoribacter splanchnicus bin.302:Mediterraneibacter;s__ bin.205:Lachnospiraceae;g__UMGS1370;s__ bin.294:Alistipes sp900290115 bin.279:Christensenellales;f__DTU072;g__;s__ bin.208:Christensenellales;f__CAG−917;g__UMGS1688;s__UMGS1688 sp900554085 bin.198:Anaerobutyricum;s__ bin.74:Oscillospiraceae;g__CAG−110;s__ bin.289:Ruminococcus_G;s__ bin.163:Intestinimonas;s__ bin.254:RFN20;f__CAG−631;g__CAG−631;s__CAG−631 sp900762575 bin.179:Acutalibacteraceae;g__UMGS403;s__UMGS403 sp900541565 bin.168:Oscillospiraceae;g__CAG−110;s__CAG−110 sp900546915 bin.83:Romboutsia;s__ bin.213:Tidjanibacter inops_A bin.79:Gastranaerophilales;f__RUG14156;g__;s__ bin.266:Gastranaerophilales;f__RUG14156;g__;s__ bin.152:Ligilactobacillus salivarius bin.11:Christensenellales;f__CAG−138;g__UMGS946;s__UMGS946 sp900544965 bin.218:Anaerovoracaceae;g__CAG−145;s__CAG−145 sp900754795 bin.239:Anaerotignum;s__ bin.108:RF32;f__CAG−239;g__CAG−267;s__CAG−267 sp001917135 bin.200:Clostridia;o__TANB77;f__CAG−508;g__UMGS1994;s__ GH PL Inhibitor Cysteine Threonine Phosphotransferase system (PTS) Cofactor and vitamin metabolism ATP.synthesis Serine and threonine metabolism Methane metabolism Saccharide, polyol, and lipid transport system Two-component regulatory system Terpenoid.backbone.biosynthesis Acetate Butyrate Propionate EPEF Coef C1 C2 Number of genes 0 50 100 150 200 EPEF association −0.6 −0.4 −0.2 0 0.2 bin.24:Fusicatenibacter;s__ bin.137:Rikenella microfusus bin.215:Blautia_A sp002159835 bin.92:Lachnospiraceae;g__UBA7160;s__ bin.19:Acutalibacteraceae;g__An172;s__An172 sp002160515 bin.34:Alistipes communis bin.125:Lactobacillus crispatus bin.258:Lactobacillus gallinarum bin.120:Lactobacillus johnsonii bin.156:Anaerostipes;s__ bin.110:Butyricicoccus;s__ bin.292:Gemmiger;s__ bin.58:Butyricicoccus;s__ bin.81:Enterocloster sp900547035 bin.93:Oscillospiraceae;g__UBA5446;s__UBA5446 sp004553625 bin.27:Lachnospiraceae;g__OF09−33XD;s__ bin.112:Merdibacter sp900754715 bin.99:Limosilactobacillus;s__ bin.267:Limosilactobacillus vaginalis bin.26:Bilophila bin.97:Clostridia;o__TANB77;f__CAG−508;g__CAG−269;s__ bin.84:Ruminococcaceae;g__UBA3818;s__ bin.151:Bacilli;o__RF39;f__UBA660;g__CAG−822;s__ bin.65:RF39;f__UBA660;g__CAG−302;s__ bin.96:Phocaeicola vulgatus bin.91:Bacteroides fragilis bin.234:Barnesiella viscericola bin.119:Christensenellales;f__CAG−74;g__OEMS01;s__ bin.291:Alistipes megaguti bin.286:Odoribacter splanchnicus bin.302:Mediterraneibacter;s__ bin.205:Lachnospiraceae;g__UMGS1370;s__ bin.294:Alistipes sp900290115 bin.279:Christensenellales;f__DTU072;g__;s__ bin.208:Christensenellales;f__CAG−917;g__UMGS1688;s__UMGS1688 sp900554085 bin.198:Anaerobutyricum;s__ bin.74:Oscillospiraceae;g__CAG−110;s__ bin.289:Ruminococcus_G;s__ bin.163:Intestinimonas;s__ bin.254:RFN20;f__CAG−631;g__CAG−631;s__CAG−631 sp900762575 bin.179:Acutalibacteraceae;g__UMGS403;s__UMGS403 sp900541565 bin.168:Oscillospiraceae;g__CAG−110;s__CAG−110 sp900546915 bin.83:Romboutsia;s__ bin.213:Tidjanibacter inops_A bin.79:Gastranaerophilales;f__RUG14156;g__;s__ bin.266:Gastranaerophilales;f__RUG14156;g__;s__ bin.152:Ligilactobacillus salivarius bin.11:Christensenellales;f__CAG−138;g__UMGS946;s__UMGS946 sp900544965 bin.218:Anaerovoracaceae;g__CAG−145;s__CAG−145 sp900754795 bin.239:Anaerotignum;s__ bin.108:RF32;f__CAG−239;g__CAG−267;s__CAG−267 sp001917135 bin.200:Clostridia;o__TANB77;f__CAG−508;g__UMGS1994;s__ GH PL Inhibitor Cysteine Threonine Phosphotransferase system (PTS) Cofactor and vitamin metabolism ATP synthesis Serine and threonine metabolism Methane metabolism Saccharide, polyol, and lipid transport system Two component regulatory system Terpenoid backbone biosynthesis Acetate Butyrate Propionate EPEF C1 C2 Number of genes 0 50 100 150 200 EPEF association Negative Positive C1/C2 group association Absent in C1/C2 No association Positive/Negative association bin.24:Fusicatenibacter;s__ bin.137:Rikenella microfusus bin.215:Blautia_A sp002159835 bin.92:Lachnospiraceae;g__UBA7160;s__ bin.19:Acutalibacteraceae;g__An172;s__An172 sp002160515 bin.34:Alistipes communis bin.125:Lactobacillus crispatus bin.258:Lactobacillus gallinarum bin.120:Lactobacillus johnsonii bin.156:Anaerostipes;s__ bin.110:Butyricicoccus;s__ bin.292:Gemmiger;s__ bin.58:Butyricicoccus;s__ bin.81:Enterocloster sp900547035 bin.93:Oscillospiraceae;g__UBA5446;s__UBA5446 sp004553625 bin.27:Lachnospiraceae;g__OF09−33XD;s__ bin.112:Merdibacter sp900754715 bin.99:Limosilactobacillus;s__ bin.267:Limosilactobacillus vaginalis bin.26:Bilophila bin.97:Clostridia;o__TANB77;f__CAG−508;g__CAG−269;s__ bin.84:Ruminococcaceae;g__UBA3818;s__ bin.151:Bacilli;o__RF39;f__UBA660;g__CAG−822;s__ bin.65:RF39;f__UBA660;g__CAG−302;s__ bin.96:Phocaeicola vulgatus bin.91:Bacteroides fragilis bin.234:Barnesiella viscericola bin.119:Christensenellales;f__CAG−74;g__OEMS01;s__ bin.291:Alistipes megaguti bin.286:Odoribacter splanchnicus bin.302:Mediterraneibacter;s__ bin.205:Lachnospiraceae;g__UMGS1370;s__ bin.294:Alistipes sp900290115 bin.279:Christensenellales;f__DTU072;g__;s__ bin.208:Christensenellales;f__CAG−917;g__UMGS1688;s__UMGS1688 sp900554085 bin.198:Anaerobutyricum;s__ bin.74:Oscillospiraceae;g__CAG−110;s__ bin.289:Ruminococcus_G;s__ bin.163:Intestinimonas;s__ bin.254:RFN20;f__CAG−631;g__CAG−631;s__CAG−631 sp900762575 bin.179:Acutalibacteraceae;g__UMGS403;s__UMGS403 sp900541565 bin.168:Oscillospiraceae;g__CAG−110;s__CAG−110 sp900546915 bin.83:Romboutsia;s__ bin.213:Tidjanibacter inops_A bin.79:Gastranaerophilales;f__RUG14156;g__;s__ bin.266:Gastranaerophilales;f__RUG14156;g__;s__ bin.152:Ligilactobacillus salivarius bin.11:Christensenellales;f__CAG−138;g__UMGS946;s__UMGS946 sp900544965 bin.218:Anaerovoracaceae;g__CAG−145;s__CAG−145 sp900754795 bin.239:Anaerotignum;s__ bin.108:RF32;f__CAG−239;g__CAG−267;s__CAG−267 sp001917135 bin.200:Clostridia;o__TANB77;f__CAG−508;g__UMGS1994;s__ GH PL Inhibitor Cysteine Threonine Phosphotransferase system (PTS) Cofactor and vitamin metabolism ATP synthesis Serine and threonine metabolism Methane metabolism Saccharide, polyol, and lipid transport system Two component regulatory system Terpenoid backbone biosynthesis Acetate Butyrate Propionate EPEF C1 C2 Number of genes 0 50 100 150 200 EPEF association Negative Positive C1/C2 group association Absent in C1/C2 No association Positive/Negative association .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint

586 The microbial community structure of the gut has been proven crucial in both host health and 587 performance [1,8]. As such, supplementation of feed additives to diets and administration of other 588 health interventions have become common approaches to modulate intestinal microbial 589 communities and ensure the ideal growth and health of broilers [66]. Previously, we reported 590 alterations in the chicken gut microbiome in response to various production systems, using 16S rRNA 591 sequencing. Our findings revealed that inclusion of omega-3 in feed can result in increase in bacterial 592 genera associated with short-chain fatty acid (SCFA) production and affect levels of pathogenic 593 bacteria including Campylobacter levels through competitive exclusion [9]. Due to increasing 594 demand for poultry meat worldwide [67], there is a need to optimise gut health for improved feed 595 efficiency and overall health of broilers. Therefore, we designed modified gut health schemes across 596 three broiler production cycles and assessed their influence on the caecal microbiome using 597 metagenomic shotgun sequencing. Herein, we employed assembly of 84 high quality MAGs 598 recovered from 118 caecal samples and analysis of metabolic function capacity of the MAGs and 599 association with performance parameters and health. 600 Administration of ionophores and GHIs employed in this study have been previously reported to 601 modulate gut microbiota, with documented growth-promoting effects and microbial alterations 602 dependent on the specific formulation and application level (Cai et al., 2022; Maki et al., 2019; 603 Robinson et al., 2019; B. Wang et al., 2021). Within our study, we observed variations across the 604 three cycles, where C2 groups demonstrated best performance. Since Probiotic B was revealed to 605 have the highest positive impact on EPEF, we posit that the efficacy of Probiotic B may be a key 606 contributing factor to the comparatively superior performance observed in cycle 2 compared to the 607 other cycles provided with Probiotic A. These findings further underscore the importance of strain-608 specific effects and dosage considerations when implementing probiotic interventions in poultry 609 systems, as reported before [68,69]. However, cycle variations within the same production system 610 have also been reported previously and are hypothesised to occur due to variations in climate, 611 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint management, and fluctuations in microbiota of day-old chicks as affected by hatchery-to-farm 612 transfer [70]. Furthermore, some differences in MAG composition between C1 and C2 brought upon 613 by differences in the GHIs used may explain their disparities in performance. For instance, Rikenella 614 microfusus (bin.137), positively associated with EPEF but absent in C1, has been identified for its 615 potential probiotic effects, attributed to its role in producing short-chain carboxylic acids, which 616 contribute to maintaining cell structure integrity [71]. Conversely, Alistipes spp. (bin.291, bin.294) 617 and Phocaeicola vulgatus (bin.96), all observed to be negatively associated with EPEF and present in 618 only C1, have also been implicated in human health issues such as cancer, cardiovascular disorders 619 and inflammatory related diseases [72,73]. Another noteworthy MAG is Barnesiella viscericola 620 (bin.234), previously reported as an efficient coloniser of chicken caeca [74], but here observed as 621 negatively correlated with EPEF but positively correlated with MT, and absent in C2. 622 623 Comparison of groups have also revealed substantial differences in microbial diversity and 624 composition between GHI groups and the control (T1). Specifically, our analysis revealed that T1 and 625 T3 to have better overall performance but exhibited lower alpha diversity than other groups (in C2). 626 According to Coyte et al., (2015), high alpha diversity in the gut microbiome tends to destabilize 627 microbiome communities, potentially leading to decreased ecological stability which is the ability to 628 return to a natural state after a perturbation. Unstable gut microbial communities are then less likely 629 to maintain beneficial symbiotic relationships and may be more susceptible to disturbances or shifts 630 that could impact the host's health and productivity [75]. With this, the moderate microbial diversity 631 shown in T1 and T3 may indicate a more balanced and stable gut, thereby becoming supportive of 632 an optimal performance. Furthermore, these discrepancies can be explained by MAGs with 633 association with performance that were differentiated among groups. MAGs belonging to 634 Lactobacillaceae (bin.258 L. gallinarum, bin.99 Limosilactobacillus, bin.125 L. crispatus, bin.120 L. 635 johnsoniiI), Butyricicoccaceae (bin.58, bin.110 Butyricioccus spp.), Ruminococcaceae (bin.292 636 Gemmiger) and Lachnospiraceae (bin.81 Enterocloster sp900547035, bin.24 Fusicatenibacter), which 637 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint are families known for their beneficial roles in gut health such as SCFA production, gut integrity 638 promotion, and protection against pathogens such as Salmonella spp. [76–81], were identified 639 beneficial for performance but were decreased in different GHI groups in comparison to T1. 640 Meanwhile, MAGs including bin.254 CAG-631 sp900762575, bin.200 UMGS1994, bin.83 Romboutsia, 641 and bin.74 CAG-110, were enriched in GHI groups but are negatively associated with EPEF. Similarly, 642 increase in Romboutsia has also been noted in broilers given Bacillus subtilis and a coccidiosis 643 vaccine [82] and reported to also negatively impact performance of breeder broilers [83]. 644 645 The standard diet of broilers usually consists of approximately 70% carbohydrates, encompassing 646 starch, oligosaccharides, and non-starch polysaccharides (NSP) like cellulose, hemicellulose as well as 647 pectin [84]. These NSPs remain undigested by the host, serving as substrates for the gut microbiome. 648 Consequently, gut microorganisms possess a diverse range of genes encoding enzymes known as 649 CAZymes, which facilitate the breakdown and metabolism of these polysaccharides [85]. CAZymes 650 are categorized into families, including Polysaccharide lyases (PL) and glycoside hydrolases (GH), 651 based on sequence similarities, although members within the same family may exhibit different 652 substrate specificities [85]. In our study, we observed preference of gut microbiota in T4 for “other 653 glucans” and “other glycans” and depletion of pectin bacterial specialists in T4 and T5 of C1. Since 654 availability of readily accessible growth substrates diminishes as it passes through the 655 gastrointestinal tract [86], we hypothesise that due to the limited availability of protein (amino 656 acids) in T4, more preferred substrates such as starch and pectin has been digested in the upper 657 intestines, leaving caecal microbiota to use other glycans as substrate. Meanwhile, depletion in T5 658 may be explained by the reduction of Bacteroidia bacteria (in T5), similarly reported by Ding et al., 659 (2022), which are microorganisms shown here to digest pectin. 660 661 Digestion of protein available in the diet is also of great importance for optimisation of gut health 662 [88]; However, there is limited information on its association with gut microbial functions in poultry. 663 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint Previous research has shown that approximately 20% of crude protein (CP) taken in by broilers goes 664 undigested due to insufficient concentrations of endogenous proteinases in the host [86,89]. 665 Consequently, undigested protein (or ileal bypass protein) which are fermented by gut microbiota in 666 the hindgut (caeca), can encourage with increased growth of Clostridium perfringens and production 667 of detrimental metabolites including ammonia, indoles, and phenols [86]. Hence, we included 668 reduction of CP as one of our gut health approaches in our study (as represented by T4). 669 Nonetheless, this group was shown to have significantly lower overall performance. Since restriction 670 of CP might have resulted in decrease of threonine [90,91], observed deficiency of threonine 671 proteases in T4 may have affected threonine intestinal absorption by the host. A large proportion of 672 host dietary threonine, known as the second (or third) limiting amino acid in broilers, is 673 predominantly used by the host for production of mucin, an important glycoprotein that preserves 674 the integrity of intestinal mucosa and function (Qaisrani et al., 2018). With this, it is hypothesised 675 that there could be impaired intestinal permeability in T4 broilers which may have then contributed 676 to overall poorer nutrient absorption, thereby affecting growth and performance. In addition, T4 in 677 C1 has also been revealed to have differentially abundant genes for metalloproteases, a family of 678 peptidases previously linked to have overactivity in patients with irritable bowel syndrome (Mills et 679 al., 2022). Meanwhile, the elevated levels of HB metrics in T4 of C2 was unexpected given that 680 reduction of CP in diets has commonly been associated with lower incidence of footpad lesions and 681 better litter quality [94,95]. However, significances in abundance among C2 groups were determined 682 for cysteine proteases, which are proteases renowned for their involvement in virulence and their 683 ability to induce inflammatory responses including atopic dermatitis in humans [96,97]. It is 684 interesting to note that C2 groups have both higher cysteine and HB levels than C1 groups (especially 685 C01A), also indicating possible link between cysteine levels and HB occurrence. 686 687 The gut microorganisms participate in the metabolism and uptake of numerous nutrients and play 688 crucial roles in preserving the integrity of the intestinal barrier, regulating the immune system, and 689 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint protecting against pathogen colonisation [2]. In this study, we primarily identified differentiation of 690 metabolic functions involved in energy production, nucleotide metabolism and drug transport 691 related pathways among treatment groups. This is in line with previous research that have shown 692 the following: 1) Probiotic supplementation in broilers can affect vitamin biosynthesis and other 693 energy related metabolic activities [98]; 2) Caecal microbial changes due to probiotics can affect 694 emissions such as nitrogen or ammonia [99]; 3) Antimicrobials can disrupt the nucleotide pool of 695 bacterial cells, resulting in increased nucleotide biosynthesis and elevated central carbon 696 metabolism [100]; and 4) By influencing the expression of microbial enzymes involved in pathways 697 linked to nucleotide, amino acid, carbohydrate, and energy metabolism, antimicrobials could 698 potentially steer metabolic flow, regulating bacterial proliferation or generating metabolites that 699 affect the host [101]. Specifically, we observed reduction of the phosphotransferase system, 700 cofactor and vitamin metabolism, and serine and threonine modules, in several GHI groups (T5 and 701 T6) compared to T1, which are pathways involved in nutrient absorption and defence against 702 infection in the host [102–104]. Increase in methane metabolism was also detected in T3 of C1 703 which was also similarly reported in other studies involving probiotic use in chickens [105,106]. 704 However, this is only attributed to one KEGG module (M00345) which was detected in bin.92 705 UBA7160 (Lachnospiraceae). We also observed the relative decrease in drug transport-related 706 modules namely bacitracin, lantibiotic and PatAB transport systems. Further research is needed to 707 confirm whether this can be explained by the similarity in pharmacological mechanisms of 708 ionophores to bacitracin and lantibiotics, both of which are antimicrobials that also prevent cell wall 709 synthesis and mainly act on gram-positive bacteria [107,108]. 710 711 SCFA, metabolites synthesised by caecal gut microorganisms from breakdown of dietary fibre, play 712 vital roles in improving metabolism, facilitating nutrient digestion and absorption, thereby 713 promoting optimal health, growth, and well-being in poultry [2]. From our analysis, we generally 714 observed gene differentiation of SCFA modules among our treatments, wherein T5 was observed to 715 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint have relatively higher gene abundance among other groups, especially in acetate production. We 716 speculate this can be due to the effect of essential oils given in T5, which coincides in the increase of 717 acetic and butyric acids in caecal of broilers given dietary oregano aqueous extracts [109]. This 718 finding, however does not confirm if the abundance of SCFA produced by gut microbes are optimal 719 for gut health and performance in broilers, since exorbitant amounts of SCFAs may activate the gut 720 microbiota-brain-cell axis response, resulting in either enteritis or other metabolic syndromes [110]. 721 Furthermore, greater amounts of propionate and butyrate acids were previously detected in birds 722 with low feed efficiency than those with high feed efficiency [111]. Nonetheless, a higher number of 723 EPEF+ MAGs were shown to possess at least one KEGG module associated to SCFA production, 724 compared to EPEF- MAGs, potentially indicating contribution of SCFA production capacity of caecal 725 gut microbiota to broiler performance. 726 727 Our study boasts several strengths, including the commercial farm set-up representing real-life 728 poultry industrial farming, the utilisation of shotgun metagenomic sequencing data, and thorough 729 assessments of performances characteristics. These aspects empowered us to delve into the 730 intricate composition and functions of gut microbiota concerning GHI administration with 731 meticulous resolution and effective control of potential confounding factors. As our sampling was 732 limited to a single genetic line of chickens and confined to caecal genetic study at one timepoint, we 733 missed the opportunity to observe the effects of GHI on early development, and its potential links to 734 temporal and spatial shifts of the chicken gut microbiome. For instance, a previous study by Gao et 735 al., (2017) demonstrated that maturation of gut microbiota is promoted by probiotic administration 736 whilst delayed by antibiotic use, highlighting the importance of broiler age in the use of 737 supplements. Additional study into other gut compartments, timepoints, and other metagenomic 738 features is therefore warranted. This includes deeper investigation of other gut microorganisms such 739 as of bacteriophages and fungi, and of other relevant microbial elements including CRISPR-Cas 740 systems, resistance, stress genes and virulence genes. In addition, future application of a multi-omics 741 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint approach involving proteomics, meta transcriptomics, and metabolomics may confirm several of our 742 hypotheses and uncover other areas we are not able to explore. Nevertheless, we believe our 743 research represents a novel and comprehensive comparative investigation of the chicken 744 metagenomic changes between ionophores and GHIs. 745 746

747 Metagenomics has allowed us to gain insights on the bacterial population in the chicken caeca. This 748

has been utilised to highlight the differences in the composition, diversity, and metabolic 749 functions of these caecal bacteria as influenced by different gut health schemes. Performing such 750 analyses, we explored the structure of the gut bacterial population in conjunction with relevant 751 metadata. We identified several MAGs such as Rikenella microfusus, UBA 7160 species and 752 Lactobacillus species as beneficial organisms, due to their positive association with EPEF and higher 753 capacity for metabolic functions. Such information will enhance our understanding of the highly 754 complex relationship between gut microbes and optimal performance. It will also enable us to 755 devise effective interventions and control strategies against enteric pathogens, which are important 756 members of the poultry gut microbiome. 757 758 Among the gut health strategies investigated in this study, we observed that use of Probiotics B in a 759 flock as observed in C2 enables better bird performance. Specifically, supplementation of Probiotics 760 B in conjunction with vaccination is observed as the best GHI strategy, resulting in a similar 761 performance to the control. However, we still observe the ionophore group to have the best 762 performance and is hypothesised to be due to their ability to reduce microbial competition, resulting 763 in a more efficient capture of nutrients by the gut microbiota and subsequently by the host. 764 Nonetheless, our results demonstrates that supplementation of GHIs are effective methods for 765 broiler gut modulation, with evidence of having various influences on both MAG composition and 766 feed related metabolic functions. Our data also suggests that excessive administration of GHIs may 767 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint not be beneficial for performance, highlighting the importance of careful selection of GHI type and 768 GHI combinations. These results significantly enhance our comprehension of microbiota-related 769 metabolic pathways, offering new avenues to improve overall performance and poultry health. 770 List of abbreviations 771 ADG: Average daily gain 772 ANOVA: Analysis of variance 773 BW: Bird weight 774 CAZymes: Carbohydrate active enzymes 775 COM: Completeness (Genome) 776 CON: Contamination (Genome) 777 CP: Crude protein 778 DMRT: Duncan Multiple Range Test 779 EPEF: European poultry efficiency factor 780 F: Finisher diet 781 FA: Fisher’s alpha 782 FCR: Feed conversion ratio 783 FPD: Footpad dermatitis 784 G: Grower diet 785 GC: Guanosine-cytosine 786 GH: Glycoside hydrolase 787 GHI: Gut health intervention 788 GLLVM: General linear latent variable model 789 H: Shannon’s index 790 HB: Hockburn 791 KO: KEGG Orthology 792 LASSO: Least absolute shrinkage and selection operator 793 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint MAG: Metagenome-assembled genomes 794 MT: Mortality 795 NSP: Non-starch polysaccharides 796 PA: Probiotics A 797 PB: Probiotics B 798 PCA: Principal component analysis 799 PCoA: Principal coordinate analysis 800 PG: Phylogenetic gain (Novelty) 801 PERMANOVA: Permutational multivariate ANOVA 802 PL: Polysaccharide lyase 803 R: Richness 804 S: Specnumber 805 Si: Simpson 806 SCFA: Short-chain fatty acids 807 SCG: Single Copy Genes 808 TSS + CLR: Total Sum Scaling and Centralised Log Ratio 809 W: Withdrawal diet 810 811 Declarations 812 Availability of data and material 813 The sequencing datasets generated and/or analysed during the current study are available in the 814 ENA repository Accession PRJEB75892. 815 Competing interests 816 AR and CH are employed by the company Moy Park. All other authors declare no competing 817 interests. 818 Funding 819 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint This work was supported by the Biotechnology and Biological Sciences Research Council (grant 820 number BB/T008709/1). 821 822 Authors' contributions 823 Conceptualisation: AR, CH, UL 824 Laboratory work: GP, BS, CK 825 Data curation: GP, UI 826 Formal analysis: GP, UI, BS 827 Writing – original draft: GP 828 Writing – review and editing: GP, OG, AP, CK, CH, DX, AR, UL, CN, UI, BS 829 830

831 We are grateful to Abbie Graham and Hugo Hanna for their contribution in sampling and 832 performance data collection. We also thank John Moore for his insights and comments on this study. 833 834

835 1. Bindari YR, Gerber PF . Centennial Review: Factors a;ecting the chicken gastrointestinal 836 microbial composition and their association with gut health and productive performance. 837 Poultry Science. 2022;101:101612. 838 2. Aruwa CE, Pillay C, Nyaga MM, Sabiu S. Poultry gut health – microbiome functions, 839 environmental impacts, microbiome engineering and advancements in characterization 840 technologies. J Animal Sci Biotechnol. 2021;12:119. 841 3. Dai X, Hinsu AT, Dadousis C, Pandit RJ, Crotta M, Limon G, et al. Genome-wide association 842 studies of chicken caecal microbiota. Proceedings of 12th World Congress on Genetics Applied 843 to Livestock Production (WCGALP) [Internet]. Rotterdam, the Netherlands: Wageningen 844 Academic Publishers; 2022 [cited 2023 Mar 1]. p. 2421–4. Available from: 845 https://www.wageningenacademic.com/doi/10.3920/978-90-8686-940-4_584 846 4. Pandit RJ, Hinsu AT, Patel NV , Koringa PG, Jakhesara SJ, Thakkar JR, et al. Microbial diversity 847 and community composition of caecal microbiota in commercial and indigenous Indian 848 chickens determined using 16s rDNA amplicon sequencing. Microbiome. 2018;6:115. 849 5. Baxter M, Richmond A, Lavery U, O’Connell NE. A comparison of fast growing broiler chickens 850 with a slower-growing breed type reared on Higher Welfare commercial farms. PLOS ONE. 851 2021;16:e0259333. 852 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint 6. Sasaki Y , Uemura R, Sekiguchi S, Takahashi T, Fujii Y , Sueyoshi M. An analysis of factors 853 a;ecting production performance in broiler flocks on Japanese commercial farms. British 854 Poultry Science. 2014;55:737–44. 855 7. Ndlebe L, Tyler NC, Ciacciariello M. E;ect of varying levels of dietary energy and protein on 856 broiler performance: a review. World’s Poultry Science Journal. 2023;79:449–65. 857 8. Stanley D, Hughes RJ, Moore RJ. Microbiota of the chicken gastrointestinal tract: influence on 858 health, productivity and disease. Appl Microbiol Biotechnol. 2014;98:4301–10. 859 9. McKenna A, Ijaz UZ, Kelly C, Linton M, Sloan WT, Green BD, et al. Impact of industrial 860 production system parameters on chicken microbiomes: mechanisms to improve performance 861 and reduce Campylobacter. Microbiome. 2020;8:128. 862 10. Yadav S, Jha R. Strategies to modulate the intestinal microbiota and their e;ects on nutrient 863 utilization, performance, and health of poultry. Journal of Animal Science and Biotechnology. 864 2019;10:2. 865 11. Feighner SD, Dashkevicz MP . Subtherapeutic levels of antibiotics in poultry feeds and their 866 e;ects on weight gain, feed e;iciency, and bacterial cholyltaurine hydrolase activity. Appl 867 Environ Microbiol. 1987;53:331–6. 868 12. Gaskins HR, Collier CT, Anderson DB. Antibiotics as Growth Promotants:mode of Action. 869 Animal Biotechnology. 2002;13:29–42. 870 13. Niewold TA. The Nonantibiotic Anti-Inflammatory E;ect of Antimicrobial Growth Promoters, 871 the Real Mode of Action? A Hypothesis. Poultry Science. 2007;86:605–9. 872 14. Chapman HD, Je;ers TK, Williams RB. Forty years of monensin for the control of coccidiosis 873 in poultry. Poultry Science. 2010;89:1788–801. 874 15. Commission Regulation (EC) No 1831/2003. REGULATION (EC) No 1831/2003 OF THE 875 EUROPEAN PARLIAMENT AND OF THE COUNCIL of 22 September 2003 on additives for use in 876 animal nutrition (Text with EEA relevance) OJ L 268, 18.10.2003, p. 29. 286:29 [Internet]. 2003. 877 Available from: http://data.europa.eu/eli/reg/2003/1831/2019-07-26. 878 16. Parker CD, Lister SA, Gittins J. Impact assessment of the reduction or removal of ionophores 879 used for controlling coccidiosis in the UK broiler industry. Veterinary Record. 2021;189:e513. 880 17. Ayalew H, Zhang H, Wang J, Wu S, Qiu K, Qi G, et al. Potential Feed Additives as Antibiotic 881 Alternatives in Broiler Production. Frontiers in Veterinary Science [Internet]. 2022 [cited 2023 882 Oct 30];9. Available from: https://www.frontiersin.org/articles/10.3389/fvets.2022.916473 883 18. Kalia VC, Shim WY , Patel SKS, Gong C, Lee J-K. Recent developments in antimicrobial growth 884 promoters in chicken health: Opportunities and challenges. Science of The Total Environment. 885 2022;834:155300. 886 19. Zhu Q, Sun P , Zhang B, Kong L, Xiao C, Song Z. Progress on Gut Health Maintenance and 887 Antibiotic Alternatives in Broiler Chicken Production. Frontiers in Nutrition [Internet]. 2021 [cited 888 2024 Jan 4];8. Available from: https://www.frontiersin.org/articles/10.3389/fnut.2021.692839 889 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint 20. Granstad S, Kristo;ersen AB, Benestad SL, Sjurseth SK, David B, Sørensen L, et al. E;ect of 890 Feed Additives as Alternatives to In-feed Antimicrobials on Production Performance and 891 Intestinal Clostridium perfringens Counts in Broiler Chickens. Animals. 2020;10:240. 892 21. Ren H, Vahjen W, Dadi T, Saliu E-M, Boroojeni FG, Zentek J. Synergistic E;ects of Probiotics 893 and Phytobiotics on the Intestinal Microbiota in Young Broiler Chicken. Microorganisms. 894 2019;7:684. 895 22. Choi KY , Lee TK, Sul WJ. Metagenomic Analysis of Chicken Gut Microbiota for Improving 896 Metabolism and Health of Chickens — A Review. Asian Australas J Anim Sci. 2015;28:1217–25. 897 23. Durazzi F , Sala C, Castellani G, Manfreda G, Remondini D, De Cesare A. Comparison 898 between 16S rRNA and shotgun sequencing data for the taxonomic characterization of the gut 899 microbiota. Sci Rep. 2021;11:3030. 900 24. Jovel J, Patterson J, Wang W, Hotte N, O’Keefe S, Mitchel T, et al. Characterization of the Gut 901 Microbiome Using 16S or Shotgun Metagenomics. Frontiers in Microbiology [Internet]. 2016 902 [cited 2023 May 7];7. Available from: 903 https://www.frontiersin.org/articles/10.3389/fmicb.2016.00459 904 25. Chen Y , Akhtar M, Ma Z, Hu T, Liu Q, Pan H, et al. Chicken cecal microbiota reduces 905 abdominal fat deposition by regulating fat metabolism. npj Biofilms Microbiomes. 2023;9:1–16. 906 26. Moy Park Ltd. The Moy Park Farming Way [Internet]. 2022 [cited 2024 Jun 30]. Available from: 907 https://moypark.com/agriculture 908 27. Moy Park Ltd. Code of Conduct and Ethics [Internet]. Pilgrims; 2023 [cited 2024 Jun 30]. 909 Available from: https://moypark.com/footer/compliance-codes-and-policies 910 28. Moy Park Ltd. Animal Welfare Conduct and Practice - Responsible Management [Internet]. 911 2024 [cited 2024 Jun 30]. Available from: https://moypark.com/case-studies/animal-welfare-912 conduct-and-practice 913 29. Aviagen. Ross Broiler: Nutrition Specifications [Internet]. Aviagen. 2022 [cited 2024 Jul 31]. 914 Available from: https://en.aviagen.com/assets/Tech_Center/Ross_Broiler/Ross-915 BroilerNutritionSpecifications2022-EN.pdf 916 30. Baker D. Quadram Institute Best Practice in Microbiome Research: Shotgun Metagenomic 917 Illumina Library Preparation v1.0 [Internet]. 2022 [cited 2024 Jul 31]. Available from: 918 https://quadram.ac.uk/wp-content/uploads/2021/06/6.-Shotgun-Metagenomic-Illumina-919 Library-Preparation-v1.0.pdf 920 31. Aviagen. Ross Broiler Management Book [Internet]. 2018 [cited 2024 Jul 31]. Available from: 921 https://aviagen.com/assets/Tech_Center/Ross_Broiler/Ross-BroilerHandbook2018-EN.pdf 922 32. Joshi N, Fass J. Sickle: A sliding-window, adaptive, quality-based trimming tool for FastQ 923 files [Internet]. 2011. Available from: https://github.com/najoshi/sickle 924 33. Li D, Liu C-M, Luo R, Sadakane K, Lam T-W. MEGAHIT: an ultra-fast single-node solution for 925 large and complex metagenomics assembly via succinct de Bruijn graph. Bioinformatics. 926 2015;31:1674–6. 927 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint 34. Uritskiy GV , DiRuggiero J, Taylor J. MetaWRAP—a flexible pipeline for genome-resolved 928 metagenomic data analysis. Microbiome. 2018;6:1–13. 929 35. Kang DD, Li F , Kirton E, Thomas A, Egan R, An H, et al. MetaBAT 2: an adaptive binning 930 algorithm for robust and e;icient genome reconstruction from metagenome assemblies. PeerJ 931 [Internet]. 2019 [cited 2024 Mar 8];7. Available from: 932 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6662567/ 933 36. Wu Y-W, Tang Y-H, Tringe SG, Simmons BA, Singer SW. MaxBin: an automated binning 934

to recover individual genomes from metagenomes using an expectation-maximization 935 algorithm. Microbiome. 2014;2:1–18. 936 37. Alneberg J, Bjarnason BS, de Bruijn I, Schirmer M, Quick J, Ijaz UZ, et al. Binning 937 metagenomic contigs by coverage and composition. Nat Methods. 2014;11:1144–6. 938 38. Parks DH, Imelfort M, Skennerton CT, Hugenholtz P , Tyson GW. CheckM: assessing the 939 quality of microbial genomes recovered from isolates, single cells, and metagenomes. Genome 940 Research. 2015;25:1043. 941 39. Zhou Z, Tran PQ, Breister AM, Liu Y , Kieft K, Cowley ES, et al. METABOLIC: high-throughput 942 profiling of microbial genomes for functional traits, metabolism, biogeochemistry, and 943 community-scale functional networks. Microbiome. 2022;10:33. 944 40. Chaumeil P-A, Mussig AJ, Hugenholtz P , Parks DH. GTDB-Tk: a toolkit to classify genomes 945 with the Genome Taxonomy Database. Bioinformatics. 2020;36:1925–7. 946 41. Kanehisa M, Goto S. KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res. 947 2000;28:27–30. 948 42. Zhang H, Yohe T, Huang L, Entwistle S, Wu P , Yang Z, et al. dbCAN2: a meta server for 949 automated carbohydrate-active enzyme annotation. Nucleic Acids Res. 2018;46:W95–101. 950 43. Anantharaman K, Brown CT, Hug LA, Sharon I, Castelle CJ, Probst AJ, et al. Thousands of 951 microbial genomes shed light on interconnected biogeochemical processes in an aquifer 952 system. Nat Commun. 2016;7:1–11. 953 44. Rawlings ND, Barrett AJ, Bateman A. MEROPS: the peptidase database. Nucleic Acids 954 Research. 2010;38:D227. 955 45. Lee MD. GToTree: a user-friendly workflow for phylogenomics. Bioinformatics. 956 2019;35:4162–4. 957 46. Parks DH. Genome Tree Toolkit [Internet]. 2015 [cited 2024 Mar 8]. Available from: 958 https://github.com/donovan-h-parks/GenomeTreeTk 959 47. R Core Team. R: A language and environment for statistical computing [Internet]. Vienna, 960 Austria: R Foundation for Statistical Computing; 2022. Available from: https://www.R-961 project.org/ 962 48. Shapiro SS, Wilk MB. An Analysis of Variance Test for Normality (Complete Samples). 963 Biometrika. 1965;52:591–611. 964 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint 49. Mendiburu FD, Simon R. Agricolae - Ten years of an open source statistical tool for 965 experiments in breeding, agriculture and biology [Internet]. PeerJ Inc.; 2015 Sep. Report No.: 966 e1748. Available from: https://peerj.com/preprints/1404 967 50. Friedman JH, Hastie T, Tibshirani R. Regularization Paths for Generalized Linear Models via 968 Coordinate Descent. J Stat Soft. 2010;33:1–22. 969 51. James G, Witten D, Hastie T, Tibshirani R. An Introduction to Statistical Learning: with 970 Applications in R [Internet]. New York, NY: Springer US; 2021 [cited 2024 Mar 12]. Available 971 from: https://link.springer.com/10.1007/978-1-0716-1418-1 972 52. Dixon P . VEGAN, A Package of R Functions for Community Ecology. Journal of Vegetation 973 Science. 2003;14:927–30. 974 53. Ijaz UZ, Ameer A, Saleem F , Gul F , Keating C, Javed S. Specialty grand challenge: how can we 975 use integrative approaches to understand microbial community dynamics? 2024 [cited 2024 976 Jun 24]; Available from: https://www.frontiersin.org/journals/systems-977 biology/articles/10.3389/fsysb.2024.1432791/full 978 54. Niku J, Hui FKC, Taskinen S, Warton DI. gllvm: Fast analysis of multivariate abundance data 979 with generalized linear latent variable models in r. Methods in Ecology and Evolution. 980 2019;10:2173–82. 981 55. Rohart F , Gautier B, Singh A, Cao K-AL. mixOmics: An R package for ‘omics feature selection 982 and multiple data integration. PLOS Computational Biology. 2017;13:e1005752. 983 56. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-984 seq data with DESeq2. Genome Biology. 2014;15:550. 985 57. Caspi R, Billington R, Keseler IM, Kothari A, Krummenacker M, Midford PE, et al. The 986 MetaCyc database of metabolic pathways and enzymes - a 2019 update. Nucleic Acids Res. 987 2020;48:D445–53. 988 58. Drula E, Garron M-L, Dogan S, Lombard V , Henrissat B, Terrapon N. The carbohydrate-active 989 enzyme database: functions and literature. Nucleic Acids Res. 2022;50:D571–7. 990 59. The CAZypedia Consortium. Ten years of CAZypedia: a living encyclopedia of carbohydrate-991 active enzymes. Glycobiology. 2018;28:3–8. 992 60. Peng P , Lu Y , Bosma TNP , Nijenhuis I, Nijsse B, Shetty SA, et al. Metagenomic- and 993 cultivation-based exploration of anaerobic chloroform biotransformation in hypersaline 994 sediments as natural source of chloromethanes. bioRxiv. 2019;858480. 995 61. Wang L, Zhang G, Li Y , Zhang Y . E;ects of High Forage/Concentrate Diet on Volatile Fatty 996 Acid Production and the Microorganisms Involved in VFA Production in Cow Rumen. Animals. 997 2020;10:223. 998 62. Wickham H. ggplot2: Elegant Graphics for Data Analysis [Internet]. Springer-Verlag New 999 York; 2016. Available from: https://ggplot2.tidyverse.org 1000 63. Gu Z, Eils R, Schlesner M. Complex heatmaps reveal patterns and correlations in 1001 multidimensional genomic data. Bioinformatics. 2016;32:2847–9. 1002 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint 64. Paradis E, Claude J, Strimmer K. APE: Analyses of Phylogenetics and Evolution in R 1003 language. Bioinformatics. 2004;20:289–90. 1004 65. Yu G, Smith DK, Zhu H, Guan Y , Lam TT-Y. g g t r e e : a n r p a c k a g e f o r v i s u a l i z a t i o n a n d 1005 annotation of phylogenetic trees with their covariates and other associated data. Methods in 1006 Ecology and Evolution. 2017;8:28–36. 1007 66. Maki JJ, Klima CL, Sylte MJ, Looft T. The Microbial Pecking Order: Utilization of Intestinal 1008 Microbiota for Poultry Health. Microorganisms. 2019;7:376. 1009 67. FAO. Meat Market Review: Emerging trends and outlook. Rome, Italy: Food and Agriculture 1010 Organization of the United Nations; 2023. 1011 68. Behnsen J, Deriu E, Sassone-Corsi M, Ra;atellu M. Probiotics: Properties, Examples, and 1012 Specific Applications. Cold Spring Harb Perspect Med. 2013;3:a010074. 1013 69. Jha R, Das R, Oak S, Mishra P . Probiotics (Direct-Fed Microbials) in Poultry Nutrition and 1014 Their E;ects on Nutrient Utilization, Growth and Laying Performance, and Gut Health: A 1015 Systematic Review. Animals. 2020;10:1863. 1016 70. Johnson TJ, Youmans BP , Noll S, Cardona C, Evans NP , Karnezos TP , et al. A Consistent and 1017 Predictable Commercial Broiler Chicken Bacterial Microbiota in Antibiotic-Free Production 1018 Displays Strong Correlations with Performance. Appl Environ Microbiol. 2018;84:e00362-18. 1019 71. Zhang Y , Xing H, Bolotnikov G, Krämer M, Gotzmann N, Knippschild U, et al. Enriched 1020 Aptamer Libraries in Fluorescence-Based Assays for Rikenella microfusus-Specific Gut 1021 Microbiome Analyses. Microorganisms. 2023;11:2266. 1022 72. Gilliland A, Chan JJ, De Wolfe TJ, Yang H, Vallance BA. Pathobionts in Inflammatory Bowel 1023 Disease: Origins, Underlying Mechanisms, and Implications for Clinical Care. Gastroenterology. 1024 2024;166:44–58. 1025 73. Parker BJ, Wearsch PA, Veloo ACM, Rodriguez-Palacios A. The Genus Alistipes: Gut Bacteria 1026 With Emerging Implications to Inflammation, Cancer, and Mental Health. Front Immunol 1027 [Internet]. 2020 [cited 2024 Mar 25];11. Available from: 1028 https://www.frontiersin.org/journals/immunology/articles/10.3389/fimmu.2020.00906/full 1029 74. Kubasova T, Kollarcikova M, Crhanova M, Karasova D, Cejkova D, Sebkova A, et al. Gut 1030 Anaerobes Capable of Chicken Caecum Colonisation. Microorganisms. 2019;7:597. 1031 75. Coyte KZ, Schluter J, Foster KR. The ecology of the microbiome: Networks, competition, and 1032 stability. Science. 2015;350:663–6. 1033 76. Beresford-Jones BS, Suyama S, Clare S, Soderholm A, Xia W, Sardar P , et al. Enterocloster 1034 clostridioformis induces host intestinal epithelial responses that protect against Salmonella 1035 infection [Internet]. bioRxiv; 2023 [cited 2024 Mar 25]. p. 2023.07.20.549886. Available from: 1036 https://www.biorxiv.org/content/10.1101/2023.07.20.549886v1 1037 77. Forbes JD, Chen C, Knox NC, Marrie R-A, El-Gabalawy H, de Kievit T, et al. A comparative 1038 study of the gut microbiota in immune-mediated inflammatory diseases—does a common 1039 dysbiosis exist? Microbiome. 2018;6:221. 1040 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint 78. Jin M, Kalainy S, Baskota N, Chiang D, Deehan EC, McDougall C, et al. Faecal microbiota 1041 from patients with cirrhosis has a low capacity to ferment non-digestible carbohydrates into 1042 short-chain fatty acids. Liver International. 2019;39:1437–47. 1043 79. Lundberg R, Scharch C, Sandvang D. The link between broiler flock heterogeneity and cecal 1044 microbiome composition. Animal Microbiome. 2021;3:54. 1045 80. Walter J, O’Toole PW. Microbe Profile: The Lactobacillaceae. Microbiology (Reading). 1046 2023;169:001414. 1047 81. Wang Y , Wang C, Huang J, Xie M, Li X, Fu L. Butyricicoccus plays a key role in mediating the 1048 antagonism between probiotic and antibiotic on food allergy. Food and Agricultural 1049 Immunology. 2019;30:446–61. 1050 82. Cai H, Luo S, Zhou Q, Yan Z, Liu Q, Kang Z, et al. E;ects of Bacillus subtilis and coccidiosis 1051 vaccine on growth indices and intestinal microbiota of broilers. Poultry Science. 1052 2022;101:102091. 1053 83. Yang Z, Zhang C, Wang J, Celi P , Ding X, Bai S, et al. Characterization of the Intestinal 1054 Microbiota of Broiler Breeders With Di;erent Egg Laying Rate. Front Vet Sci [Internet]. 2020 1055 [cited 2024 Mar 25];7. Available from: 1056 https://www.frontiersin.org/articles/10.3389/fvets.2020.599337 1057 84. Tejeda O, Kim W. Role of Dietary Fiber in Poultry Nutrition. Animals (Basel). 2021;11:461. 1058 85. Cantarel BL, Coutinho PM, Rancurel C, Bernard T, Lombard V , Henrissat B. The 1059 Carbohydrate-Active EnZymes database (CAZy): an expert resource for Glycogenomics. Nucleic 1060 Acids Res. 2009;37:D233–8. 1061 86. Apajalahti J, Vienola K. Interaction between chicken intestinal microbiota and protein 1062 digestion. Animal Feed Science and Technology. 2016;221:323–30. 1063 87. Ding Y , Hu Y , Yao X, He Y , Chen J, Wu J, et al. Dietary essential oils improves the growth 1064 performance, antioxidant properties and intestinal permeability by inhibiting bacterial 1065 proliferation, and altering the gut microbiota of yellow-feather broilers. Poult Sci. 1066 2022;101:102087. 1067 88. Bartlett A, Kleiner M. Dietary protein and the intestinal microbiota: An understudied 1068 relationship. iScience. 2022;25:105313. 1069 89. Ndazigaruye G, Kim D-H, Kang C-W, Kang K-R, Joo Y-J, Lee S-R, et al. E;ects of Low -Protein 1070 Diets and Exogenous Protease on Growth Performance, Carcass Traits, Intestinal Morphology, 1071 Cecal Volatile Fatty Acids and Serum Parameters in Broilers. Animals. 2019;9:226. 1072 90. Mao X, Zeng X, Qiao S, Wu G, Li D. Specific roles of threonine in intestinal mucosal integrity 1073 and barrier function. FBE. 2011;3:1192–200. 1074 91. Qaisrani SN, Ahmed I, Azam F , Bibi F , Saima, Pasha TN, et al. Threonine in broiler diets: an 1075 updated review. Annals of Animal Science. 2018;18:659–74. 1076 92. Mills RH, Dulai PS, Vázquez-Baeza Y , Sauceda C, Daniel N, Gerner RR, et al. Multi-omics 1077 analyses of the ulcerative colitis gut microbiome link Bacteroides vulgatus proteases with 1078 disease severity. Nat Microbiol. 2022;7:262–76. 1079 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint 93. Soussou S, Jablaoui A, Mariaule V , Kriaa A, Boudaya H, Wysocka M, et al. Serine proteases 1080 and metalloproteases are highly increased in irritable bowel syndrome Tunisian patients. Sci 1081 Rep. 2023;13:17571. 1082 94. Lemme A, Hiller P , Klahsen M, Taube V , Stegemann J, Simon I. Reduction of dietary protein in 1083 broiler diets not only reduces n-emissions but is also accompanied by several further benefits. 1084 Journal of Applied Poultry Research. 2019;28:867–80. 1085 95. van Harn J, Dijkslag MA, van Krimpen MM. E;ect of low protein diets supplemented with free 1086 amino acids on growth performance, slaughter yield, litter quality, and footpad lesions of male 1087 broilers. Poultry Science. 2019;98:4868–77. 1088 96. Cau L, Williams MR, Butcher AM, Nakatsuji T, Kavanaugh JS, Cheng JY , et al. Staphylococcus 1089 epidermidis protease EcpA can be a deleterious component of the skin microbiome in atopic 1090 dermatitis. Journal of Allergy and Clinical Immunology. 2021;147:955-966.e16. 1091 97. Oido-Mori M, Rezzonico R, Wang P-L, Kowashi Y , Dayer J-M, Baehni PC, et al. 1092 Porphyromonas gingivalis Gingipain-R Enhances Interleukin-8 but Decreases Gamma 1093 Interferon-Inducible Protein 10 Production by Human Gingival Fibroblasts in Response to T-Cell 1094 Contact. Infection and Immunity. 2001;69:4493–501. 1095 98. Rodrigues DR, Briggs W, Du; A, Chasser K, Murugesan R, Pender C, et al. Cecal microbiome 1096 composition and metabolic function in probiotic treated broilers. PLOS ONE. 1097 2020;15:e0225921. 1098 99. Feng Y , Wu X, Hu D, Wang C, Chen Q, Ni Y . Comparison of the E;ects of Feeding Compound 1099 Probiotics and Antibiotics on Growth Performance, Gut Microbiota, and Small Intestine 1100 Morphology in Yellow-Feather Broilers. Microorganisms. 2023;11:2308. 1101 100. Yang JH, Wright SN, Hamblin M, McCloskey D, Alcantar MA, Schrübbers L, et al. A White-1102 Box Machine Learning Approach for Revealing Antibiotic Mechanisms of Action. Cell. 1103 2019;177:1649-1661.e9. 1104 101. Zou A, Nadeau K, Xiong X, Wang PW, Copeland JK, Lee JY , et al. Systematic profiling of the 1105 chicken gut microbiome reveals dietary supplementation with antibiotics alters expression of 1106 multiple microbial pathways with minimal impact on community structure. Microbiome. 1107 2022;10:127. 1108 102. Becker B, Wottawa F , Bakr M, Koncina E, Mayr L, Kugler J, et al. Serine metabolism is 1109 crucial for cGAS-STING signaling and viral defense control in the gut. iScience. 2024;27:109173. 1110 103. Biesalski HK. Nutrition meets the microbiome: micronutrients and the microbiota. Annals 1111 of the New York Academy of Sciences. 2016;1372:53–64. 1112 104. Deutscher J, Francke C, Postma PW. How Phosphotransferase System-Related Protein 1113 Phosphorylation Regulates Carbohydrate Metabolism in Bacteria. Microbiology and Molecular 1114 Biology Reviews. 2006;70:939–1031. 1115 105. Khan S, Chousalkar KK. Functional enrichment of gut microbiome by early 1116 supplementation of Bacillus based probiotic in cage free hens: a field study. anim microbiome. 1117 2021;3:50. 1118 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint 106. Zhu C, Gong L, Huang K, Li F , Tong D, Zhang H. E;ect of Heat-Inactivated Compound 1119 Probiotics on Growth Performance, Plasma Biochemical Indices, and Cecal Microbiome in 1120 Yellow-Feathered Broilers. Front Microbiol [Internet]. 2020 [cited 2024 Mar 28];11. Available 1121 from: 1122 https://www.frontiersin.org/journals/microbiology/articles/10.3389/fmicb.2020.585623/full 1123 107. Islam MR, Nagao J-I, Zendo T, Sonomoto K. Antimicrobial mechanism of lantibiotics. 1124 Biochem Soc Trans. 2012;40:1528–33. 1125 108. Phillips I. The use of bacitracin as a growth promoter in animals produces no risk to human 1126 health. Journal of Antimicrobial Chemotherapy. 1999;44:725–8. 1127 109. Zhang F , Yang J, Zhan Q, Shi H, Li Y , Li D, et al. Dietary oregano aqueous extract improves 1128 growth performance and intestinal health of broilers through modulating gut microbial 1129 compositions. Journal of Animal Science and Biotechnology. 2023;14:77. 1130 110. Liu L, Li Q, Yang Y , Guo A. Biological Function of Short-Chain Fatty Acids and Its Regulation 1131 on Intestinal Health of Poultry. Front Vet Sci. 2021;8:736739. 1132 111. He Z, Liu R, Wang M, Wang Q, Zheng J, Ding J, et al. Combined e;ect of microbially derived 1133 cecal SCFA and host genetics on feed e;iciency in broiler chickens. Microbiome. 2023;11:198. 1134 112. Gao P , Ma C, Sun Z, Wang L, Huang S, Su X, et al. Feed-additive probiotics accelerate yet 1135 antibiotics delay intestinal microbiota maturation in broiler chicken. Microbiome. 2017;5:91. 1136 1137 1138 .CC-BY-NC 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted August 2, 2024. ; https://doi.org/10.1101/2024.08.02.606333doi: bioRxiv preprint