TYPE Original Research PUBLISHED 06 September 2022 DOI 10.3389/fphys.2022.884925 Novel strategies to improve OPEN ACCESS chicken performance and EDITED BY Mariano Enrique Fernández-Miyakawa, Instituto Nacional de Tecnología welfare by unveiling Agropecuaria, Argentina REVIEWED BY host-microbiota interactions Hauke Smidt, Wageningen University and Research, Netherlands through hologenomics Brian B. Oakley, Western University of Health Sciences, Núria Tous 1†, Sofia Marcos 2†, Farshad Goodarzi Boroojeni 3, United States *CORRESPONDENCE Ana Pérez de Rozas 4, Jürgen Zentek 3, Andone Estonba 2, Antton Alberdi, Dorthe Sandvang 5, M. Thomas P. Gilbert 6,7, antton.alberdi@sund.ku.dk † Enric Esteve-Garcia 1, Robert Finn 8, Antton Alberdi 6* and These authors have contributed equally to this work and share first authorship Joan Tarradas 1 1 SPECIALTY SECTION Animal Nutrition, Institute of Agrifood Research and Technology (IRTA), Constantí, Spain, 2Applied This article was submitted to Avian Genomics and Bioinformatics, University of the Basque Country (UPV/EHU, Bilbao, Spain, 3Institute of Physiology, Animal Nutrition, Department of Veterinary Medicine, Freie Universität Berlin (FUB), Berlin, Germany, a section of the journal 4 Animal Health-CReSA, Institute of Agrifood Research and Technology (IRTA), Bellaterra, Spain, 5Chr. Frontiers in Physiology Hansen A/S, Animal Health Innovation, Hoersholm, Denmark, 6Center for Evolutionary Hologenomics, The GLOBE Institute, University of Copenhagen, Copenhagen, Denmark, 7University Museum, RECEIVED 27 February 2022 Norwegian University of Science and Technology (NTNU), Trondheim, Norway, 8European ACCEPTED 25 July 2022 Bioinformatics Institute (EMBL-EBI), Wellcome Genome Campus, Hinxton, United Kingdom PUBLISHED 06 September 2022 CITATION Tous N, Marcos S, Goodarzi Boroojeni F, Pérez de Rozas A, Zentek J, Estonba A, Sandvang D, Gilbert MTP, Fast optimisation of farming practices is essential to meet environmental Esteve-Garcia E, Finn R, Alberdi A and sustainability challenges. Hologenomics, the joint study of the genomic Tarradas J (2022), Novel strategies to improve chicken performance and features of animals and the microbial communities associated with them, welfare by unveiling host-microbiota opens new avenues to obtain in-depth knowledge on how host-microbiota interactions through hologenomics. Front. Physiol. 13:884925. interactions affect animal performance and welfare, and in doing so, improve doi: 10.3389/fphys.2022.884925 the quality and sustainability of animal production. Here, we introduce the COPYRIGHT animal trials conducted with broiler chickens in the H2020 project HoloFood, © 2022 Tous, Marcos, Goodarzi and our strategy to implement hologenomic analyses in light of the initial Boroojeni, Pérez de Rozas, Zentek, Estonba, Sandvang, Gilbert, Esteve- results, which despite yielding negligible effects of tested feed additives, provide Garcia, Finn, Alberdi and Tarradas. This is relevant information to understand how host genomic features, microbiota an open-access article distributed development dynamics and host-microbiota interactions shape animal welfare under the terms of the Creative Commons Attribution License (CC BY). and performance. We report the most relevant results, propose hypotheses to The use, distribution or reproduction in explain the observed patterns, and outline how these questions will be other forums is permitted, provided the addressed through the generation and analysis of animal-microbiota multi- original author(s) and the copyright owner(s) are credited and that the omic data during the HoloFood project. original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution KEYWORDS or reproduction is permitted which does animal performance, genomics, metagenomics, multi-omics, sustainability not comply with these terms. Frontiers in Physiology 01 frontiersin.org Tous et al. 10.3389/fphys.2022.884925 Introduction processes happening in each individual animal, rather than considering animals just as units that contribute to pen or With the ever-increasing human population on Earth, tank statistic averages. The first attempts to implement humanity is facing several major challenges to ensure long- individual-based multi-omic strategies in farm animals have term balance between natural resource use and environmental provided detailed understanding of feed-microbiota-animal conservation (Reid et al., 2010). Many of the current agricultural interactions (Andersen et al., 2021), by for example practices are not sustainable due to excessive carbon emission, demonstrating which bacteria with which genes are able to resource consumption, and waste production (Agovino et al., degrade which carbohydrates (Michalak et al., 2020). We 2019). Hence, there is an urgent need to transition into a more anticipate that such a mechanistic understanding of biological resilient and sustainable agriculture model, in which the processes will contribute to generating refined hypotheses, efficiency of production is improved, the use of antibiotics is predictive models and experimental treatments that will lead reduced, and the welfare of animals is ensured (Eyhorn et al., to a reduction of animals employed for research and an ultimate 2019). development of more optimal farming strategies. A fundamental step to optimise farming practices is to obtain HoloFood (HoloFood, 2019) is a multi-partner in-depth understanding on the biological functioning of the H2020 project that is pioneering such an approach, among animal production systems (Messerli et al., 2019). These other farming systems, on broiler chickens. The project aims systems often include biological elements beyond the actual at developing and implementing joint multi-omic analyses of animal that is being produced, among which host-associated animals and their associated microorganisms to generate in- microorganisms stand out due to their relevance for the optimal depth knowledge of animal-microbiota interactions, and in doing biological functioning of most animals (McFall-Ngai et al., 2013). so, improve the quantity, quality and safety of the produced food, Intestinal microorganisms not only modulate nutrient intake the sustainability of the production process and the welfare of (Diaz Carrasco et al., 2019), but also shape intestinal immune and animals. To set the baseline to an upcoming series of publications inflammatory processes (Zhou et al., 2020), intervene on host that implement such multi-omic analyses, here we introduce the systemic growth parameters (Fraune and Bosch, 2010), and even strategic vision, and experimental work conducted to generate influence host behaviour (Johnson and Foster, 2018). biological samples and associated performance results of broiler Microorganisms colonise the animal gut as soon as it is chickens in the project HoloFood. We present the main exposed to the environment (Sprockett et al., 2018), and performance results, discuss their relevance, and relate them develop communities with complex spatial and temporal to future multi-omic analyses that HoloFood partners will dynamics (Debray et al., 2021), which continuously interact conduct to address the variety of biological questions raised with the host animal (Khan et al., 2019; Ansari et al., 2020). from the initial screening of animal performance. Animal-microbiota interactions have so far remained largely unexplored because of the limited capacity of scientists to properly characterise and analyse the key elements partaking Material and methods in this interplay due to their excessive complexity (Alberdi et al., 2021). However, the development of high-throughput DNA Animals and housing sequencing and mass spectrometry technologies, linked to higher computing capacity and development of powerful The study consisted of three identical experiments (A, B, and bioinformatic tools, is changing this scenario (Graw et al., C). In each trial a total of 960 days-of-hatch broiler chicks 2021). Today we are not only able to characterise the entire belonging to two fast growing genetic lines (Ross308® and genetic information of animals and their associated Cobb500®) from two hatcheries (to increase genetic microorganisms (namely the hologenome), but we can also variability) were allocated at 24 pens upon arrival. The name quantify how genes are expressed, which proteins are of genetic lines has been blinded and each one is described along synthesised and what metabolites result from enzymatic the manuscript under the letter X or Y. In order to avoid the reactions happening in the gut (Nyholm et al., 2020). Such a possible influence of the parent stock, birds were distributed in holo-omic approach that considers multiple omic layers of both such a way that each replicate received the same number of animals and associated microorganisms, is starting to unveil broilers from each hatchery tray. Each pen had a total surface of biological features and patterns that have remained hidden so 2.25 m2 with 40 birds per pen. Each pen was provided with one far (Alberdi et al., 2021). individual hopper feeder and two nipple drinkers. The barn is This technological revolution can contribute to many sources windowless and provided with automatic environment control of variability that have been so far attributed to background with a gas heating system by screens and ventilation by noise, such as host microgenetic and microbiota variation among depression. The room also has programmable lighting, individuals, to be surfaced and included in the analyses (Alberdi provided by TL tubes evenly distributed. The temperature et al., 2021). This requires an increased attention on the biological program was adjusted according to the standard program Frontiers in Physiology 02 frontiersin.org Tous et al. 10.3389/fphys.2022.884925 FIGURE 1 Experimental, sampling and analytical design of the HoloFood study focused on broiler chickens. (A) Three experiments with identical study design were conducted in 2019. (B) Each experiment contained 12 factor combinations of three dietary treatments, two genetic lines and two sexes. (C) Each combination was replicated twice per experiment, yielding 24 pens per experiment and 72 pens in total. (D) Each pen contained 40 chickens in the beginning of the experiment, 18 experimental animals that were randomly selected and tagged the first day of the experiment, and 22 more animals that provided commercial-like density conditions. (E) Six chickens were euthanised at day 7, six more at day 21 and six more at day 35 for collecting samples for performance, multi-omic and complementary analyses. (F) From each animal, 14 tissue and digesta samples were collected for a variety of analyses (see Table 1 for details), and complementary organ samples were also collected for future analyses. used in the farm: from 0 to 2 days the temperature increased from experimental design deviates from conventional studies aimed 32 to 34°C; from 3 to 7 days the temperature was reduced to at testing the effect of feed additives on performance, as it was 29–31°C and continued decreasing for 3°C per week afterwards designed seeking to maximise inter-individual genetic variability, until reaching 21°C. The lighting program was 24 h of light the and in doing so increase the probability to identify interactions first 2 days, 18 h of light until 7 days, and 14 h of light per day between animal and microbial genomic features. The six animals afterwards. The litter was fresh wood shavings. All birds were to be slaughtered at days 7, 21, and 35 within each pen were vaccinated against Avian Infectious Bronchitis and Gumboro randomly selected and marked at day 0 to avoid observer biases diseases according to the vaccination program usually practised in subsequent samplings. at the hatchery. Moreover, a set of 240 Cobb500 animals from one of the hatcheries in experiment 1 were vaccinated against Marek disease. Diets, additives and feeding The study had a randomized complete block design with a factorial 2 × 2 × 3 arrangement according to broiler line (X or Y), Aimed at maximising the effects of the feed additives, the BD sex (male or female) and dietary treatment: 1) basal diet (BD), 2) was designed as a pro-inflammatory diet, using wheat (a cereal BD plus a probiotic additive (PR), and 3) BD plus a phytobiotic rich in non-starch polysaccharides (NSP) with a more than 50% additive (PH) (overview in Figure 1). The three trials were carried inclusion) and soybean meal as main ingredients without the out in different seasons of the year: spring (trial A), summer (trial addition of enzymes, antibiotics, or coccidiostats. Diets were B) and autumn (trial C). Treatments were randomly assigned to formulated according to birds’ requirements and commercial one pen of each block (2 blocks per experiment), so that each practices in three different periods: starter (0–9 days), grower treatment had 2 replicates (pens) per experiment (6 in total) with (10–23 days) and finisher (24–37 days). Feeds were presented in 40 animals per pen (20 animals from each hatchery). The crumble form for the starter period and in 3 mm pellets later on. Frontiers in Physiology 03 frontiersin.org Tous et al. 10.3389/fphys.2022.884925 Composition of diets and estimated nutrient contents are TABLE 1 Overview of the biological samples collected from each individual animal within the project HoloFood and their presented in Supplementary Table S1. The PR treatment corresponding multi-omic and complementary analyses with data consisted of BD with a mixture of three strains, namely included in this article bolded. Bacillus subtilis DSM 32324, B. subtilis DSM 32325 and B. Sample Analyses amyloliquefaciens DSM 25840 (level of inclusion at 0.75 g/kg feed, >3.20 × 109 CFU/g), which has been recently authorised as a Blood Lipopolysaccharide and acute phase proteins zootechnical product for poultry species (EFSA Panel on Ileum tissue Mucus production Additives and Products or Substances used in Animal Feed Histology et al., 2020). The PH treatment consisted of BD with a Targeted amplification of inflammatory markers phytobiotic additive obtained from white grapes and Shotgun chicken transcriptomics containing 78% of procyanidins and 22% of polyphenols as Ileum mucosa Shotgun chicken transcriptomics active ingredients (level of inclusion at 0.75 g/kg feed). 16S rRNA amplicon sequencing Batch feed samples were taken from each production for Ileum content Shotgun metagenomics proximate analysis (Association of Official Analytical Chemists, 16S rRNA amplicon sequencing 2000, moisture -dry matter-by oven drying–method 2-, nitrogen Shotgun metatranscriptomics -crude protein-by combustion -Dumas method-, ether extract on Chicken genomics a Soxtec system -method 3B- and ash after muffle furnace Metabolomics incineration -method 12) and to quantify concentrations of Caecum tissue Mucus production probiotic and phytobiotic additives. Data of analytical Histology composition of diets is shown in Supplementary Table S1. In Targeted amplification of inflammatory markers addition, water and litter samples were collected in each pen at Shotgun chicken transcriptomics days 0, 7, 21, and 35 for microbiological and litter quality Caecum mucosa Shotgun chicken transcriptomics analyses. Moreover, three samples of each feed per diet period 16S rRNA amplicon sequencing were taken from three different bags for microbiological Caecum content Pathogen detection through PCR assessment. Shotgun metagenomics 16S rRNA amplicon sequencing Shotgun metatranscriptomics Animal monitoring Chicken genomics Metabolomics Animals were counted and weighed by hatchery tray upon Feathers Corticosterone measurement arrival and by replicate at days 7, 21, and 35. Growth performance was monitored to replicate the same days. Pen level (40 chickens/pen) analyses included average body weight (BW), daily gain (ADG), daily feed intake (ADFI), feed euthanised according to RD 53/2013 (Spain), following conversion ratio (FCR) and European production efficiency the ethical requirements established. After the euthanasia, a factor (EPEF). The incidence and severity of footpad total of 14 samples were collected from each animal to dermatitis per pen was subjectively evaluated by trained measure individual key performance indicators (KPIs - personnel at days 7, 21 and 35 according the 5-point scale iBW, concentration of corticosterone (COR) in feathers, described by Butterworth (Butterworth and Welfare Quality concentration of acute phase proteins (C-reactive protein Consortium and Normalisatie-instituut, 2009): 0) no evidence (CRP) and chicken haptoglobin-like protein (PIT54)), and of footpad dermatitis; 1 and 2) minimal evidence of footpad lipopolisaccharides (LPS) concentration in plasma, and dermatitis; 3 and 4) evidence of footpad dermatitis. Dead animals pathogens detection in caecal contents (Salmonella spp., were weighted, and the most probable cause of death recorded. Campylobacter spp., and Clostridium spp.)) as well as to Animals (laggards) excluded from the trial during the first week generate multi-omic and complementary analyses (Table 1). were not considered. Measured KPIs aimed at quantifying quantity, quality and safety of the produced food, the sustainability of the production process and the welfare of the produced Animal sampling animals. Sections of ileum and cecum, intestinal content from ileum and cecum, feathers, and blood were obtained. At days 7–8, 21–22 and 35–37 (multiple days were necessary Different aliquots were distributed and properly stored for due to workload), 6 animals per pen were randomly selected, downstream analyses. Liver, pancreas, thymus, bursa of individually weighed (iBW), evaluated for footpad dermatitis Fabricius, brain, and spleen samples were also obtained for (data not shown), euthanised and sampled. Animals were future analyses (Figure 1F). Frontiers in Physiology 04 frontiersin.org Tous et al. 10.3389/fphys.2022.884925 In addition, at day 37 (commercial slaughtering age), Acute phase proteins in plasma 6 extra animals per pen were randomly selected, euthanised Chicken haptoglobin-like protein (PIT54) was quantified and subjected to evaluation of meat quality traits (carcass, using the haptoglobin ELISA Kit (Ref. ABIN1563052, abdominal fat, breast, and leg yield). The oxidative stability of antibodies-online.com) and C-reactive protein (CRP) was the thigh muscle was determined over a period of 7 days from measured using the CRP ELISA Kit (Ref. ABIN4947413, randomly selected three animals used for meat quality antibodies-online.com) in accordance with manufacturer assessment. instructions. The data on each individual animal derived from the experiments were analysed at three different levels. The Lipopolysaccharide in plasma first level included the analyses of measurements obtained Lipopolysaccharide concentration in plasma was determined in the farm (e.g., iBW). The second level comprised analyses using Pierce LAL Chromogenic Endotoxin Quantification Kit conducted a posteriori from samples obtained during the trials (Ref. 88282; Thermofisher, United States) in accordance with in order to control the health, welfare of animals (e.g., ELISA manufacturer instructions. for CRP, PIT54, LPS and COR quantification, and PCR for pathogen detection), and meat quality traits (e.g., meat pH of litter oxidative stability). The third level included multi-omic Three samples of 50 g of litter per pen were collected for analyses, which characterises the animal-microbiota system pH determination at days 7, 21 and 35 avoiding the areas near at the highest level of breadth and resolution. These analyses and below the feeders and drinkers. The three samples collected will include whole-genome sequencing of chicken genomes, from the same pen and day were pooled and homogenised and metagenomics of microbial communities (meta) the moisture was determined in a sub-sample of 100 g according transcriptomics and metabolomics. This third level of data to the AOAC method Association of Official Analytical will be carried out in the next steps of the project. Chemists, 2000 method 925.09). For pH analysis, a subsample of 10 g was placed in a beaker with 100 ml of distilled water, shaked with a glass rod and allowed to stand for 30 min. The Analytical procedures pH value was obtained using a pH metre (Crison, L’Hospitalet de Llobregat, Spain). Pathogen detection Opportunistic infections of different zoonotic pathogens Oxidative stability of meat (Salmonella, Campylobacter and Clostridium spp.) were Samples (5 g of muscle from the thigh) were homogenised controlled in caecal contents from sampled animals at days with an aqueous 7.5% trichloroacetic acid solution, filtered and 7, 21 and 35 by conventional PCR (absence/presence). brought to 20 ml. To proceed, 5 ml of the extraction solution and Salmonella spp. was detected using the primers 5′-GTG 5 ml of 0.02 M thiobarbituric acid were mixed and boiled for AAATTATCGCCACGTTCGGGCAA-3′ and 5′-TCATCG 15 min and then cooled in cold water. Absorbance of the peak CACCGTCAAAGGAACC-3′, which are specific for the was measured at 525 nm as malondialdehyde production in an InvA gene of Salmonella, following the conditions ultraviolet-visible spectrophotometer (Shimadzu, Japan) using described in Rahn et al. (1992). For the detection of the third derivative of the spectrum between 425 and 650 to Clostridium perfringens, the primers 5′-AAGATTTGTAAG correct the baseline. The 1,1,3,3-tetraethoxypropane was used as GCGCTT-3′ and 5′-ATTTCCTGAAATCCACTC-3′ specific standard (Botsoglou et al., 1994; Ruiz et al., 1999). for alpha-toxin gene present in all strains of this bacteria species were used. The specific amplification program was as follows: 94°C/4’; (94°C/1′, 55°C/1′, 72°C/1′20″)x35; 72°C/15’; Statistical analyses 4°C/end. The presence of Campylobacter spp. was detected using the primers 5′-TTGGAAACGACTGCTAATACTCTA- Data was explored to discard any possible outlier according to 3′ and 5′-AGCCATTAGATTTCACAAGAGACT-3′, which the Kolmogorov-Smirnov test (Massey, 1951). As no outliers were amplify a specific segment of 16S ribosomal RNA gene considered, the statistical analysis included all data. The GLIMMIX specific of Campylobacter spp. The specific amplification procedure of SAS software (SAS/STAT 14.1; SAS Institute Inc., Cary, program was as follows: 94°C/4’ (94°C/1′, 48°C/2′, 72°C/1′) NC, United States) was used to perform the analysis of the different x35; 72°C/15’; 4°C/end. variables. In the case of ELISA determinations, when the limit of detection was not reached, the missing values were replaced by the Corticosterone in feathers limit of detection (L)/√2 (Hornung and Reed, 1990). The statistical Corticosterone was determined in accordance with the model used is shown below: method described by Bortolotti et al. (2008). yijklm  μ + Ti + Bk +S1 + Em + γj +eijklm Frontiers in Physiology 05 frontiersin.org Tous et al. 10.3389/fphys.2022.884925 FIGURE 2 Overview of main results. (A) Body weight differences across dietary treatments, namely basal diet (BD), BD plus probiotic (PR) and BD plus phytobiotic (PH), at day 35. (B) Body weight differences across lines and sexes at day 35. (C) Body weight progression of the three experiments, with detailed overview of days 7, 21 and 35. (D) Corticosterone (COR) levels measured in feathers at different days, sexes, and genetic lines. (E) Linear correlation between COR levels and body weight at the three time points. (F) C-reactive protein (CRP), avian haptoglobin-like protein (PIT54) and lipopolysaccharide (LPS) levels in plasma across time points in different dietary treatments. (G) CRP, PIT54 and LPS levels in plasma across time points in different genetic lines. (H) CRP, PIT54 and LPS levels in plasma across time points in different sexes. (I) CRP, PIT54 and LPS levels in plasma across time points in different experiments. (J) Hierarchical decomposition of observed body weight variation within pens, among pens and among experiments. Where yijklm is the response variable, Ti is the dietary Results treatment effect, Bk is the broiler line effect, Sl is the biological sex effect, Em is the experiment effect, γj is the random block In the following, we present the main results obtained from pen- effect, and eij is the error of the experimental unit. The level analyses, and the first two levels of the individual analyses experimental unit for the statistics and the tables presented sorted by topic. We explain their biological relevance based on was the pen. Results in Supplementary Tables S2–S12 are current knowledge, and we project the potential of multi-omic expressed as least square means ± standard error. Differences analyses that will be conducted in the project HoloFood on top were considered significant at p < 0.05, while those at p < 0.1 are of these results to address the most relevant pending questions. reported as tendencies. Hierarchical decomposition of the variance was carried out using ANOVA-estimation of variance components as Effect of dietary treatments implemented in the fitVCA function of the R package VCA (Schuetzenmeister and Dufey, 2017), using iBW as response No significant effect of administered dietary additives was variable, and pen and experiment as explanatory variables. observed in the performance of the animals (Figure 2A; Frontiers in Physiology 06 frontiersin.org Tous et al. 10.3389/fphys.2022.884925 FIGURE 3 Overview of holo-omic analyses that will be conducted in the H2020 project HoloFood to deepen into the results outlined in this manuscript and address the questions raised and beyond. Supplementary Tables S2-S12). Dietary treatments neither Males showed higher COR levels than females in the period induced any significant change in the acute phase protein 0–35 days (13.8 vs. 11.7 pg/mg; p < 0.001; SE = 0.91) and also at values measured in plasma (Figure 2F, Supplementary Table day 7 and at day 35 (p < 0.001 and p = 0.028, respectively) S7). The only parameter affected by dietary treatments was COR (Figure 2D, Supplementary Table S6). In the last period of the (Supplementary Table S6), which was increased with the PH diet experiment (at day 35), the levels of CRP in plasma were higher at day 7, but these levels were not prolonged over time nor in females than males (1.31 vs. 1.18 ng/mg; SE = 0.08; p = 0.045) decreased compared to the BD. (Figure 2H, Supplementary Table S7B). Effect of broiler line Effect of age and development Broiler line X reached significantly higher BW, ADG, and Corticosterone accumulation in feathers increased as animals ADFI and lower FCR than line Y at day 35 (2232 g and 1878 g for grew (Supplementary Table S6; Figure 2E), while the levels of BW (standard error (SE) = 18.8), 62.6 and 52.4 g for ADG (SE = PIT54 and CRP in plasma peaked at day 7, and decreased 0.54), 96.6 and 81.1 g for ADFI (SE = 0.80), and 1.497 and through time (Supplementary Table S7B; Figures 2F,G). 1.542 for FCR (SE = 0.015), respectively; Figure 2B and Supplementary Table S2). The accumulation of COR in feathers was significantly higher Effect of zoonotic pathogens for line X (13.91 pg/mg) than for line Y (11.54 pg/mg) (SE = 0.914; p < 0.001) (Figure 2D, Supplementary Table S6) and Targeted detection of three common zoonotic pathogens, individual COR levels were also higher in animals with higher namely Salmonella, Clostridium and Campylobacter spp. was iBW (Figure 2E). PIT54 levels in plasma were also higher in line performed to detect whether natural colonisation of the chicken X at days 7 (p = 0.0687) and 21 (p = 0.001) than in line Y (+35.5% intestines occurred during the trials. Only one animal in trial A and and +25.4%, respectively) (Figure 2G, Supplementary Table S7B) another in trial B were positive for Clostridium, being all the and for the whole experiment (p = 0.0354; +27.5%). animals negative for Salmonella and Campylobacter. However, in trial C, the analyses revealed that 13.7% of the sampled animals at day 7 were positive for Salmonella, but the colonisation vanished Effect of biological sex as the animals grew. Moreover, 99% of the birds sampled from day 21 onwards presented a Campylobacter colonisation, with no Male chickens exhibited larger average BW than females at all prevalence differences between dietary treatments. The detection time points, reaching an 8.7% larger BW at day 35 (2157 and of Campylobacter was correlated with a drop in BW from day 1953 g, respectively (SE = 13.8); Figure 2B). The rest of 21 onwards compared to the two previous trials (Figure 2C). In performance traits were also improved in males compared to addition, a peak of CRP values in plasma was detected in females (ADG, ADFI, FCR, and EPEF) (Supplementary Table S2). Campylobacter positive animals at day 21 (Figure 2I). Frontiers in Physiology 07 frontiersin.org Tous et al. 10.3389/fphys.2022.884925 Hierarchical variance decomposition of Whelan et al., 2019). Accordingly, the overall performance was chicken body weight 13.8% lower than the expected from reference performance tables (Cobb-Vantress, 2018; Aviagen, 2019), yet with no differences Due to our interest in generating systemic characterisation of between treatments. individual chickens, we explored how the variance of iBW data for COR measured in feathers is used as a biomarker of each combination of factors (i.e., treatment, biological sex and accumulative stress (Bortolotti et al., 2008), as chronic levels genetic line) was distributed across the three hierarchical levels of of COR are associated with detrimental effects on growth and the study: 1) variation across the six animals sampled in each pen at related biological traits (Scanes, 2016). The increase of COR with each time point, 2) variation between the two pen replicates within the PH diet at day 7, and the disappearance of this effect each experiment, and 3) variation among the three experiments. throughout the study contrasts with previous studies that The average coefficient of variation for iBW of the six sampled reported a reduction of serum and feather COR in broilers animals within each pen was 11.08%, with maximum values fed with polyphenol extracts (Gong et al., 2018; Gopi et al., 2020). reaching 22.5% (Supplementary Table S12). At day 35, the To delve into the reasons under, among other questions, the average BW difference between the largest and smallest animal lack of positive effect of the feed additives on commonly assessed sampled in each pen was over 31% of the mean value. The intrapen nutritional parameters, HoloFood will generate deep genome- variability explained most (76.3%) of the variance observed within resolved metagenomic datasets (Almeida et al., 2019; Pasolli combinations of factors (Figure 2J). However, its weight with et al., 2019) from the ileal and caecal content. While most respect to experiment factor decreased with the age of the animal. chicken-associated microbiota research is being conducted using targeted sequencing (Mohd Shaufi et al., 2015; Jurburg et al., 2019; Ocejo et al., 2019), the first shotgun-sequencing based studies have Discussion recently been published (Glendinning et al., 2020; Gilroy et al., 2021). The metagenome-assembled genome (MAG) catalogue of Effect of dietary treatments bacteria associated with broiler chickens generated in HoloFood, will not only complement such efforts for the high-resolution Bacillus spp. are commonly used as additives in broiler characterisation of chicken-associated microbiomes, but will also production (Irta, 2015), and the specific probiotic strains tested enable in-depth study of strain-level microbiota dynamics in the in our study have been previously shown to improve performance analysed production context through combining taxonomic and and physiological traits in broilers (Molnár et al., 2011; Goodarzi direct functional inferences. The reconstructed bacterial genomes Boroojeni et al., 2018). However, contrasting observations that will be functionally annotated, thus directly inferring the metabolic align with our results have also been reported (Li et al., 2018), capacities (e.g., complex polysaccharide degradation, short-chain which could be explained by varying experimental conditions, fatty acid (SCFA) production) of strains, and acknowledging the such as the specific Bacillus strain employed, the dose, or the basal aggregated functional landscape of the entire community present diet. The supplementation of poultry diets with multi-strain in each animal (Shaffer et al., 2020). HoloFood will also generate Bacillus probiotic products is in general associated with whole genome sequences of the probiotic strains through hybrid competitive exclusion of common pathogens, improved nutrient short- and long-read DNA sequencing to ensure highest-quality digestion and absorption through the production of exogenous genome reconstructions of the tested probiotics (Wick et al., 2017), enzymes, enhanced intestinal morphology, and the modulation of and perform a pangenome analysis with other sequenced and relevant immune system pathways (Ramlucken et al., 2020; annotated Bacillus strains to identify bacterial genes that could Tarradas et al., 2020). Regarding the phytobiotic, multiple confer beneficial functional capabilities each of the strains. This modes of action have been attributed to additives containing will enable us to understand the specific means of action through polyphenols and procyanidins, including antioxidant and anti- which each strain can interact with the microbiota and various inflammatory properties, promotion of beneficial bacteria in the intestinal features of the host. All these analyses will enable gut, and enhancement of nutrient absorption through binding of ascertaining the relative abundances of Bacillus probiotics in dietary proteins and carbohydrates (Chamorro et al., 2019; Hasted different intestinal segments, and measuring whether the et al., 2021). additives trigger broad-as well as fine-scale taxonomic and/or As the beneficial effect of feed additives is usually not evident functional changes in the microbiota of broilers, that could when animals grow under optimal conditions (Vilà et al., 2010), we contribute to explain the observed results. deliberately induced a challenging condition through increasing the amount of dietary soluble NSP. These compounds are known to have deleterious effects on the bird’s health and performance Effect of broiler line through increasing intestinal viscosity and hampering nutrient digestibility (Raza et al., 2019), and can thus maximise the Although theoretically both lines tend to perform similarly in beneficial effects of feed additives (Bortoluzzi et al., 2019; terms of growth and final BW (Cobb-Vantress, 2018; Aviagen, Frontiers in Physiology 08 frontiersin.org Tous et al. 10.3389/fphys.2022.884925 2019; Livingston et al., 2020), line X reached significantly higher in the performance differences observed. Such chicken genomic BW, ADG, and ADFI and lower FCR than line Y at day 35. The data will be coupled with the aforementioned microbial differences observed on performance between lines could metagenomic information, which will allow exploring whether partially be explained by the higher initial BW of line X and how host-microbiota interactions orchestrate different (44.5 g) at day 0 than line Y (40.8 g) (SE = 0.31; physiological responses to nutritional stress in the two broiler Supplementary Table S2). A retrospective analysis of the lines, as previously reported in other taxa (Ma et al., 2019). breeders’ features showed that the age of breeders was higher for line X (49.6 ± 10.1 weeks) than for line Y (44.1 ± 11.4 weeks), which probably caused the observed difference between the Effect of biological sex initial BW (Iqbal et al., 2017). However, the differences between lines at day 35 did not disappear even when BW at The differences between biological sex among chickens day 0 was included as a covariate in the statistical model, observed in the current experiment are expected (Cobb- suggesting that other factors contributed to shape the Vantress, 2018; Aviagen, 2019), as growth patterns of male differences observed between both lines. Therefore, chickens outperform that of females (Aggrey, 2002; Hausman performance results suggest that line X could exhibit a higher et al., 2014; Livingston et al., 2020). However, unlike in previous resistance to NSP or a more active feeding behaviour than line Y. studies (Carbajal et al., 2014), males showed higher COR levels The higher accumulation of COR in feathers in the broiler line than females, supporting again that COR cannot directly be X or in the animals with a higher iBW could be explained by an negatively associated with welfare and performance of animals increased growth rate promoting deposition of COR in feathers (Jimeno et al., 2018). On the other hand, females showed higher (Jimeno et al., 2018), in contrast to what would be expected from a levels of CRP than in males at day 35. CRP is an acute phase stress indicator (Carbajal et al., 2014). On the other hand, PIT54 is protein used as a highly sensitive marker of inflammation and an acute phase protein with an important inhibitory role in tissue damage (O’reilly and Eckersall, 2014). Some immune inflammation processes (Wicher and Fries, 2006), which is biomarkers including CRP are influenced by biological sex in rapidly increased in the blood as a response to infectious agents other production animals such as pigs (Gutiérrez et al., 2018). or physiological stressors (O’reilly and Eckersall, 2014). PIT54 in However, the reasons behind biological sex differences for the chicken plasma binds free haemoglobin to inhibit inflammatory markers are still unknown. haemoglobin-mediated oxidation of lipid and protein (Ahn In HoloFood we will aim at further understanding these et al., 2019). Antinutritional effects of NSP are related to a differences in growth between both sexes, as well as other reduction of BW and FCR, and can trigger a mild chronic questions mentioned above, by generating whole-genome inflammation in the gut (Cardoso Dal Pont et al., 2020). Host transcriptomic data (RNAseq) to identify gene expression gut inflammatory response produces reactive oxygen species differences in the intestinal tissues. While intestinal expression (ROS), which cause oxidative stress and have the potential to of targeted genes is routinely measured in animal sciences damage host tissue (Costantini and Møller, 2009). The increased (Slawinska et al., 2019; Farahat et al., 2021), how genome- antioxidant activity through the high levels of PIT54 in line X wide gene expression varies across sexes and intestinal could explain the better performance compared with line Y sections is still largely unknown. As intestinal gene expression through the amelioration of antinutritional effects of NSP. is known to be modulated by the microbiota (Volf et al., 2017), The observed differences between broiler lines point to and biological sex also contributes to shaping microbial differences in systemic responses to the pro-inflammatory communities (Lee et al., 2017), we will aim at detecting diet, which probably yielded the BW and associated KPI associations between host expression patterns and microbial differences between the two genetic lines. The whole-genome communities to ascertain host-microbiota interactions related analyses we will conduct in HoloFood will enable deepening into to sex differences. For instance, we will analyse expression of host these observations through unveiling genome-wide differences genes involved in cholesterol (precursor of steroids, and thus between both lines. We will perform whole genome resequencing related to COR) absorption, such as NPC1L1 and ABCG5/ to generate single nucleotide polymorphism (SNP) profiles of ABCG8, which have been shown to be modulated by the all individuals (Li et al., 2009). This will enable testing whether microbiota in rodents (Zhong et al., 2015). Sample collection both lines have genetic differences in key genes related to in HoloFood was extended to organs beyond the gastrointestinal inflammatory responses induced by a high dietary tract, including liver and brain, which are involved in appetite concentration of NSP, as well as key metabolic pathways such regulation and other processes related to the gut-brain axis as steroid hormone biosynthesis (Kanehisa et al., 2021). In (Cryan et al., 2019). The analysis of gene expression in such addition, we will also generate de-novo reference genomes organs along with gut processes, will enable delving into the (Baker, 2012) and Hi-C maps (Lieberman-Aiden et al., 2009) relationship between host genetics and microbiota factors with of both lines, to explore the effects of structural genome variants nutrient metabolism in shaping feeding behaviour and related (e.g., copy number variations (CNVs), translocations, inversions) performance differences in broiler chickens. Frontiers in Physiology 09 frontiersin.org Tous et al. 10.3389/fphys.2022.884925 Effect of age and development (Awad et al., 2015, 2018). In accordance with these observations, the detection of Campylobacter was correlated with a drop in BW Animal development is linked to multiple changes in analysed from day 21 onwards. In addition, the peak of CRP detected in metrics. For instance, COR in feathers increased as animals grew, Campylobacter positive animals at day 21, would most probably mirroring previous observations (Nordquist et al., 2020). The pattern indicate systemic reaction to the colonisation (Liu et al., 2019; observed by the levels of PIT54 and CRP in plasma (peaked at day Zhang et al., 2020). Other opportunistic pathogens from the 7 and decreased through time), exhibit a trend that could be linked to Campylobacterales order, such as Helicobacter brantae, which vaccination as well as to microbiota development. On the one hand, may be present along with Campylobacter but undetected with vaccines are known to increase concentration of acute phase proteins targeted approaches, might likewise be involved in performance and stress markers during the first days after administration (Kaab drop (Kollarcikova et al., 2019). etal.,2018),andallchickensintheexperimentwerevaccinatedagainst The bacterial genome catalogue we will build in HoloFood will Avian Infectious Bronchitis and Gumboro diseases at hatch. On the not only enable us to ascertain whether the natural Campylobacter other hand, early microbial colonisation of the intestine is also known colonisation was due to a single or multiple strains (Chaloner et al., to boost the development of the immune system (Broom and Kogut, 2014), but also to characterise the entire catalogues of genes of these 2018) through, for example, the production of the intestinal mucus strains and thus identify potential virulence factors that seem to be layer (Duangnumsawang et al., 2021), which provides the first inducing an inflammatory response. We will be able to measure protective shield preventing a direct access of pathogenic bacteria whether the Campylobacter colonisation correlated with any to the epithelial surface. systemic change in the microbiota and in the intestinal response HoloFood will also generate and analyse microbiota-wide gene of the animals, through combining gene expression data of chickens expression in the chicken intestine, as well as metabolites that play and microorganisms as well as metabolomic information. essential roles in the host-microbiota interplay, such as SCFAs (van Campylobacter induces the expression of various host pro- der Hee and Wells, 2021). Our study design, which entails inflammatory cytokines through the activation of Toll-like euthanising animals at each sampling point, prioritises spatial receptor 4 (TLR4) and TLR21 signalling pathways (de Zoete resolution of intestinal sections over temporal development, et al., 2010). Our analyses will enable measuring changes in the which complicates tracking the temporal development of expression levels of genes involved in these signalling pathways individual animals. However, sampling at three different time between Campylobacter positive and negative animals, to deepen points will provide an overview of how much microbiota into the effect of Campylobacter in chicken welfare and performance. development varies across individuals (Sprockett et al., 2018; Debray et al., 2021; Ballou et al., 2016; Jurburg et al., 2019). Shotgun metatranscriptomic data will complement the Hierarchical variance decomposition of metagenomic information, thus providing not only an overview chicken body weight of the relative abundances of different bacterial taxa in different time points, but also displaying the gene expression patterns of the In the current study, the mentioned individual variability was bacteria. In addition, metabolomic data will enable validating higher than previously reported (Vasdal et al., 2019; Lundberg et al., whether the activated metabolic pathways are translated into 2021), although chicken BW variation can increase if animals are different levels of SCFA concentrations. We will measure how subjected to challenging diets (Gous, 2018). The reduction of the the expression of microbial genes involved in carbohydrate intrapen variability with the age of animals, would probably be due metabolism and SCFA production vary across development, to augmenting environmental effects and potentially microbiota and how these changes are associated with animal growth and development factors. Although the three experiment replicates were changes observed in acute phase protein levels. identical, and abiotic conditions were controlledin the farm, these were conducted in spring, summer, and autumn, which could have entailed slight differences in the temperature and humidity of the barn. Effect of zoonotic pathogens The whole-genome analyses we will perform in HoloFood will enable ascertaining whether the interindividual genetic variability Campylobacter and Salmonella are usually considered mere could be related to the observed dispersion of the data. Genotype- commensals in poultry, but they cause the highest numbers of phenotype association studies in commercial chicken lines have foodborne diseases in humans globally (European Food Safety identified several important genomic regions that explain a Authorityand European Centre for Disease Prevention and percentage of the BW variation (Tarsani et al., 2019; Wang et al., Control, 2021). Recent studies have nevertheless shown that 2020; Dadousis et al., 2021). In addition, the variability in the Campylobacter colonisation in chickens can cause gut intestinal microbiota can intensify differences in performance microbiota alterations and intestinal damage that occasionally between individuals from the same population (Yan et al., 2017; facilitates bacterial colonisation of extraintestinal organs, which Shah et al., 2019; Wen et al., 2021). Ultimately, we will aim at may eventually lead to a reduced animal performance and welfare identifying chicken genetic variants associated with microbiota Frontiers in Physiology 10 frontiersin.org Tous et al. 10.3389/fphys.2022.884925 changes with noteworthy impact on performance. These chicken experimental procedures approved by Ethical Committee of genomic and microbial metagenomic analyses will enable Generalitat de Catalunya, Spain (Proceeding number 10226). ascertaining to which degree the observed intrapen and interexperiment variation can be attributed to differences in the genetic features of chickens and microbial communities. Author contributions NT, EEG, DS, AA, and JT contributed to the conception and Conclusion design of the study. NT, SM, FGB, APR, EEG, AA, and JT performed the material preparation, data collection, and analysis. All authors To conclude, the range of analytical approaches outlined in this contributed to the article and approved the submitted version. article will give us the opportunity to showcase the strength of implementing new multi-omic approaches to address relevant questions for farming practices. Many questions that would Funding remain unanswered by employing traditional techniques can now be addressed using these new technologies, and most importantly, This research was funded by the European Union’s Horizon new questions that were not so far set out (e.g., the reasons for intra- Research and Innovation Programme under grant agreement No. pen variability) can be now proposed. We believe that the 817729 (HoloFood, Holistic solution to improve animal food hologenomic approach being implemented in HoloFood will help production through deconstructing the biomolecular interactions us move from “Does factor X affect KPI Y?” to “How and why does between feed, gut microorganisms and animals in relation to factor X affect KPI Y?”. That is to say, transitioning from a trial-and- performance parameters). error approach to a knowledge-based strategy in which understanding biological processes that underlie the administration of feeds, additives or drugs, as well as the Acknowledgments observed interindividual variation, is prioritised. We will address We would like to thank the researchers, farm workers and all the questions outlined in this article and more through the volunteers who participated in the design, maintenance and collaboration of multiple academic and industrial partners, aimed at execution of the animal trials, as well as in the collection and pioneering the large-scale implementation of hologenomics in processing of the samples. MTPG acknowledges the Danish animal farming (Figure 3). Although HoloFood will generate and National Research Foundation through the grant DNRF143 analyse one of the largest multi-omic datasets in farm animals with “A Center for Evolutionary Hologenomics”. characterisation of hundreds of specimens, we acknowledge the mathematical challenges of analysing such a complex and hyper- dimensional dataset. The dimensionality of the data will be reduced Conflict of interest by leveraging the hierarchical structure of biology itself (e.g., enzymes embedded within metabolic pathways), as well as using The authors declare that the research was conducted in the the most advanced feature selection approaches to identify the most absence of any commercial or financial relationships that could relevant molecular elements. Ongoing data analyses, which will be be construed as a potential conflict of interest. published in upcoming articles, will show us how far we can reach, what are the limitations of this novel approach, and how the field can best advance to make the most of the new technologies for a more Publisher’s note secure, ethical and sustainable food production. All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated Data availability statement organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or The original contributions presented in the study are claim that may be made by its manufacturer, is not guaranteed or included in the article/Supplementary Material, further endorsed by the publisher. inquiries can be directed to the corresponding author. Supplementary material Ethics statement The Supplementary Material for this article can be found The animal study was reviewed and approved by This experiment online at: https://www.frontiersin.org/articles/10.3389/fphys. followed the EU principles for animal care and experimentation and 2022.884925/full#supplementary-material Frontiers in Physiology 11 frontiersin.org Tous et al. 10.3389/fphys.2022.884925 References Aggrey, S. E. (2002). Comparison of three nonlinear and spline regression models gut microbial metabolism and intestinal barrier in broiler chickens. Food Funct. 10, for describing chicken growth curves. Poult. Sci. 81, 1782–1788. doi:10.1093/ps/81. 1444–1454. doi:10.1039/c8fo02465k 12.1782 Cobb-Vantress (2018). Cobb 500: Broiler Performance & Nutrition Supplement. Agovino, M., Casaccia, M., Ciommi, M., Ferrara, M., and Marchesano, K. (2019). Available at: https://www.cobb-vantress.com/assets/5a88f2e793/Broiler- Agriculture, climate change and sustainability: The case of EU-28. Ecol. Indic. 105, Performance-Nutrition-Supplement.pdf (Accessed February 1, 2022). 525–543. doi:10.1016/j.ecolind.2018.04.064 Costantini, D., and Møller, A. P. (2009). Does immune response cause oxidative Ahn, J., Woodfint, R. M., Lee, J., Wu, H., Ma, J., Suh, Y., et al. (2019). Comparative stress in birds? A meta-analysis. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 153, identification, nutritional, and physiological regulation of chicken liver-enriched 339–344. doi:10.1016/j.cbpa.2009.03.010 genes. Poult. Sci. 98, 3007–3013. doi:10.3382/ps/pez057 Cryan, J. F., O’Riordan, K. J., Cowan, C. S. M., Sandhu, K. V., Bastiaanssen, T. F. Alberdi, A., Andersen, S. B., Limborg, M. T., Dunn, R. R., and Gilbert, M. T. P. S., Boehme, M., et al. (2019). The microbiota-gut-brain Axis. Physiol. Rev. 99, (2021). Disentangling host–microbiota complexity through hologenomics. Nat. 1877–2013. doi:10.1152/physrev.00018.2018 Rev. Genet. 23, 281–297. doi:10.1038/s41576-021-00421-0 Dadousis, C., Somavilla, A., Ilska, J. J., Johnsson, M., Batista, L., Mellanby, R. J., Almeida, A., Mitchell, A. L., Boland, M., Forster, S. C., Gloor, G. B., Tarkowska, et al. (2021). A genome-wide association analysis for body weight at 35 days A., et al. (2019). A new genomic blueprint of the human gut microbiota. Nature 568, measured on 137, 343 broiler chickens. Genet. Sel. Evol. 53, 70. doi:10.1186/s12711- 499–504. doi:10.1038/s41586-019-0965-1 021-00663-w Andersen, T. O., Kunath, B. J., Hagen, L. H., Arntzen, M. Ø., and Pope, P. B. de Zoete, M. R., Keestra, A. M., Roszczenko, P., and van Putten, J. P. M. (2010). (2021). Rumen metaproteomics: Closer to linking rumen microbial function to Activation of human and chicken toll-like receptors by Campylobacter spp. Infect. animal productivity traits. Methods 186, 42–51. doi:10.1016/j.ymeth.2020.07.011 Immun. 78, 1229–1238. doi:10.1128/IAI.00897-09 Ansari, I., Raddatz, G., Gutekunst, J., Ridnik, M., Cohen, D., Abu-Remaileh, M., Debray, R., Herbert, R. A., Jaffe, A. L., Crits-Christoph, A., Power, M. E., and et al. (2020). The microbiota programs DNA methylation to control intestinal Koskella, B. (2021). Priority effects in microbiome assembly. Nat. Rev. Microbiol. 20, homeostasis and inflammation. Nat. Microbiol. 5, 610–619. doi:10.1038/s41564- 109–121. doi:10.1038/s41579-021-00604-w 019-0659-3 Diaz Carrasco, J. M., Casanova, N. A., and Fernández Miyakawa, M. E. (2019). Association of Official Analytical Chemists (2000). Official methods of analysis. Microbiota, gut health and chicken productivity: What is the connection? 17th Edn. Washington, DC. Microorganisms 7, E374. doi:10.3390/microorganisms7100374 Aviagen (2019). Ross 308: Performance objectives. Available at: http://eu.aviagen. Duangnumsawang, Y., Zentek, J., and Goodarzi Boroojeni, F. (2021). com/tech-center/download/1339/Ross308-308FF-BroilerPO2019-EN.pdf. Development and functional properties of intestinal mucus layer in poultry. (Accessed February 1, 2022). Front. Immunol. 12, 745849. doi:10.3389/fimmu.2021.745849 Awad, W. A., Molnár, A., Aschenbach, J. R., Ghareeb, K., Khayal, B., Hess, C., EFSA Panel on Additives and Products or Substances used in Animal Feed et al. (2015). Campylobacter infection in chickens modulates the intestinal epithelial (FEEDAP)Bampidis, V., Azimonti, G., de Lourdes Bastos, M., Christensen, H., and barrier function. Innate Immun. 21, 151–160. doi:10.1177/1753425914521648 ® Dusemund, B. (2020). Safety and efficacy of GalliPro Fit (Bacillus subtilis DSM 32324, Bacillus subtilis DSM 32325 and Bacillus amyloliquefaciens DSM 25840) for Awad, W. A., Hess, C., and Hess, M. (2018). Re-Thinking the chicken- all poultry species for fattening or reared for laying/breeding. EFSA J. 18, e06094. Campylobacter jejuni interaction: A review. Avian Pathol. 47, 352–363. doi:10. doi:10.2903/j.efsa.2020.6094 1080/03079457.2018.1475724 European Food Safety Authority, and European Centre for Disease Prevention Baker, M. (2012). De novo genome assembly: What every biologist should know. and Control (2021). The European Union one health 2019 zoonoses report. EFSA J. Nat. Methods 9, 333–337. doi:10.1038/nmeth.1935 19, e06406. doi:10.2903/j.efsa.2021.6406 Ballou, A. L., Ali, R. A., Mendoza, M. A., Ellis, J. C., Hassan, H. M., Croom, W. J., Eyhorn, F., Muller, A., Reganold, J. P., Frison, E., Herren, H. R., Luttikholt, L., et al. (2016). Development of the chick microbiome: How early exposure influences et al. (2019). Sustainability in global agriculture driven by organic farming. Nat. future microbial diversity. Front. Vet. Sci. 3, 2. doi:10.3389/fvets.2016.00002 Sustain. 2, 253–255. doi:10.1038/s41893-019-0266-6 Bortolotti, G. R., Marchant, T. A., Blas, J., and German, T. (2008). Corticosterone Farahat, M., Ibrahim, D., Kishawy, A. T. Y., Abdallah, H. M., Hernandez- in feathers is a long-term, integrated measure of avian stress physiology. Funct. Ecol. Santana, A., and Attia, G. (2021). Effect of cereal type and plant extract 22, 494–500. doi:10.1111/j.1365-2435.2008.01387.x addition on the growth performance, intestinal morphology, caecal Bortoluzzi, C., Serpa Vieira, B., de Paula Dorigam, J. C., Menconi, A., Sokale, A., microflora, and gut barriers gene expression of broiler chickens. Animal 15, Doranalli, K., et al. (2019). Bacillus subtilis DSM 32315 supplementation attenuates 100056. doi:10.1016/j.animal.2020.100056 the effects of Clostridium perfringens challenge on the growth performance and Fraune, S., and Bosch, T. C. G. (2010). Why bacteria matter in animal intestinal microbiota of broiler chickens. Microorganisms 7, E71. doi:10.3390/ development and evolution. Bioessays. 32, 571–580. doi:10.1002/bies.200900192 microorganisms7030071 Gilroy, R., Ravi, A., Getino, M., Pursley, I., Horton, D. L., Alikhan, N.-F., et al. Botsoglou, N. A., Fletouris, D. J., Papageorgiou, G. E., Vassilopoulos, V. N., Mantis, A. (2021). Extensive microbial diversity within the chicken gut microbiome revealed J., and Trakatellis, A. G. (1994). Rapid, sensitive, and specific thiobarbituric acid method by metagenomics and culture. PeerJ 9, e10941. doi:10.7717/peerj.10941 for measuring lipid peroxidation in animal tissue, food, and feedstuff samples. J. Agric. Food Chem. 42, 1931–1937. doi:10.1021/jf00045a019 Glendinning, L., Genç, B., Wallace, R. J., and Watson, M. (2020). Metagenomic analysis of the cow, sheep, reindeer and red deer rumen. bioRxiv. Broom, L. J., and Kogut, M. H. (2018). The role of the gut microbiome in shaping the immune system of chickens. Vet. Immunol. Immunopathol. 204, 44–51. doi:10. Gong, K., Chen, L., Li, X., Sun, L., and Liu, K. (2018). Effects of germination 1016/j.vetimm.2018.10.002 combined with extrusion on the nutritional composition, functional properties and polyphenol profile and related in vitro hypoglycemic effect of whole grain corn. Butterworth, A., and Welfare Quality Consortium and Normalisatie-instituut, N. J. Cereal Sci. 83, 1–8. doi:10.1016/j.jcs.2018.07.002 (2009). Assessment protocol for poultry. Lelystad Netherlands [Delft]: Welfare Quality ; NEN. Goodarzi Boroojeni, F., Vahjen, W., Männer, K., Blanch, A., Sandvang, D., and Zentek, J. (2018). Bacillus subtilis in broiler diets with different levels of energy and Carbajal, A., Tallo-Parra, O., Sabes-Alsina, M., Mular, I., and Lopez-Bejar, M. protein. Poult. Sci. 97, 3967–3976. doi:10.3382/ps/pey265 (2014). Feather corticosterone evaluated by ELISA in broilers: A potential tool to evaluate broiler welfare. Poult. Sci. 93, 2884–2886. doi:10.3382/ps.2014-04092 Gopi, M., Dutta, N., Pattanaik, A. K., Jadhav, S. E., Madhupriya, V., Tyagi, P. K., et al. (2020). Effect of polyphenol extract on performance, serum biochemistry, skin Cardoso Dal Pont, G., Farnell, M., Farnell, Y., and Kogut, M. H. (2020). Dietary pigmentation and carcass characteristics in broiler chickens fed with different cereal factors as triggers of low-grade chronic intestinal inflammation in poultry. sources under hot-humid conditions. Saudi J. Biol. Sci. 27, 2719–2726. doi:10.1016/j.sjbs. Microorganisms 8, E139. doi:10.3390/microorganisms8010139 2020.06.021 Chaloner, G., Wigley, P., and Humphrey, S. (2014). Dynamics of dual infection Gous, R. M. (2018). Nutritional and environmental effects on broiler uniformity. with Campylobacter jejuni strains in chickens reveals distinct strain-to-strain World’s. Poult. Sci. J. 74, 21–34. doi:10.1017/S0043933917001039 variation in infection ecology. Appl. Environ. Microbiol. 80, 6366. doi:10.1128/ AEM.01901-14 Graw, S., Chappell, K., Washam, C. L., Gies, A., Bird, J., Robeson, M. S., 2nd, et al. (2021). Multi-omics data integration considerations and study design for biological Chamorro, S., Romero, C., Brenes, A., Sánchez-Patán, F., Bartolomé, B., Viveros, systems and disease. Mol. Omics 17, 170–185. doi:10.1039/d0mo00041h A., et al. (2019). Impact of a sustained consumption of grape extract on digestion, Frontiers in Physiology 12 frontiersin.org Tous et al. 10.3389/fphys.2022.884925 Gutiérrez, A. M., Montes, A., Gutiérrez-Panizo, C., Fuentes, P., and De La Cruz- on Drosophila developmental traits under nutritional stress. iScience 19, 436–447. Sánchez, E. (2018). Gender influence on the salivary protein profile of finishing pigs. doi:10.1016/j.isci.2019.07.048 J. Proteomics 178, 107–113. doi:10.1016/j.jprot.2017.11.023 Massey, F. J. (1951). The Kolmogorov-smirnov test for goodness of fit. J. Am. Stat. Hasted, T.-L., Sharif, S., Boerlin, P., and Diarra, M. S. (2021). Immunostimulatory Assoc. 46, 68–78. doi:10.1080/01621459.1951.10500769 potential of fruits and their extracts in poultry. Front. Immunol. 12, 641696. doi:10. McFall-Ngai, M., Hadfield, M. G., Bosch, T. C. G., Carey, H. V., Domazet-Lošo, 3389/fimmu.2021.641696 T., Douglas, A. E., et al. (2013). Animals in a bacterial world, a new imperative for Hausman, G. J., Barb, C. R., Fairchild, B. D., Gamble, J., and Lee-Rutherford, L. the life sciences. Proc. Natl. Acad. Sci. U. S. A. 110, 3229–3236. doi:10.1073/pnas. (2014). Gene expression profiling in adipose tissue from growing broiler chickens. 1218525110 Adipocyte 3, 297–303. doi:10.4161/adip.29252 Messerli, P., Murniningtyas, E., and Eloundou-Enyegue, P. (2019). Global HoloFood (2019). European Commission. Holistic solution to improve animal sustainable development report 2019: The future is now–science for achieving food production through deconstructing the biomolecular interactions between feed, sustainable development. Available at: http://pure.iiasa.ac.at/id/eprint/16067/1/ gut microorganisms and animals in relation to performance parameters. Available at: 24797GSDR_report_2019.pdf (Accessed August 2, 2022). https://cordis.europa.eu/project/id/817729 (Accessed August 22, 2022). Michalak, L., Gaby, J. C., Lagos, L., La Rosa, S. L., Hvidsten, T. R., Tétard-Jones, Hornung, R. W., and Reed, L. D. (1990). Estimation of average concentration in C., et al. (2020). Microbiota-directed fibre activates both targeted and secondary the presence of nondetectable values. Appl. Occup. Environ. Hyg. 5, 46–51. doi:10. metabolic shifts in the distal gut. Nat. Commun. 11, 5773. doi:10.1038/s41467-020- 1080/1047322X.1990.10389587 19585-0 Iqbal, J., Mukhtar, N., Rehman, Z. U., Khan, S. H., Ahmad, T., Anjum, M. S., et al. Mohd Shaufi, M. A., Sieo, C. C., Chong, C. W., Gan, H. M., and Ho, Y. W. (2015). (2017). Effects of egg weight on the egg quality, chick quality, and broiler Deciphering chicken gut microbial dynamics based on high-throughput 16S rRNA performance at the later stages of production (week 60) in broiler breeders. metagenomics analyses. Gut Pathog. 7, 4. doi:10.1186/s13099-015-0051-7 J. Appl. Poult. Res. 26, 183–191. doi:10.3382/japr/pfw061 Molnár, A. K., Podmaniczky, B., Kürti, P., Tenk, I., Glávits, R., Virág, G., et al. Irta (2015). Review of immune stimulator substances/agents that are (2011). Effect of different concentrations of Bacillus subtilis on growth susceptible of being used as feed additives: Mode of action and performance, carcase quality, gut microflora and immune response of broiler identification of end-points for efficacy assessment. EFSA Support. Publ. 12, chickens. Br. Poult. Sci. 52, 658–665. doi:10.1080/00071668.2011.636029 905E. doi:10.2903/sp.efsa.2015.EN-905 Nordquist, R. E., Zeinstra, E. C., Dougherty, A., and Riber, A. B. (2020). Effects of Jimeno, B., Hau, M., and Verhulst, S. (2018). Corticosterone levels reflect dark brooder rearing and age on hypothalamic vasotocin and feather corticosterone variation in metabolic rate, independent of “stress. Sci. Rep. 8, 13020. doi:10. levels in laying hens. Front. Vet. Sci. 7, 19. doi:10.3389/fvets.2020.00019 1038/s41598-018-31258-z Nyholm, L., Koziol, A., Marcos, S., Botnen, A. B., Aizpurua, O., Gopalakrishnan, Johnson, K. V.-A., and Foster, K. R. (2018). Why does the microbiome affect S., et al. (2020). Holo-Omics: Integrated host-microbiota multi-omics for basic and behaviour? Nat. Rev. Microbiol. 16, 647–655. doi:10.1038/s41579-018-0014-3 applied biological research. iScience 23, 101414. doi:10.1016/j.isci.2020.101414 Jurburg, S. D., Brouwer, M. S. M., Ceccarelli, D., van der Goot, J., Jansman, A. O’reilly, E. L., and Eckersall, P. D. (2014). Acute phase proteins: A review of their J. M., and Bossers, A. (2019). Patterns of community assembly in the developing function, behaviour and measurement in chickens. World’s. Poult. Sci. J. 70, 27–44. chicken microbiome reveal rapid primary succession. Microbiologyopen 8, e00821. doi:10.1017/S0043933914000038 doi:10.1002/mbo3.821 Ocejo, M., Oporto, B., and Hurtado, A. (2019). 16S rRNA amplicon sequencing Kaab, H., Bain, M. M., and Eckersall, P. D. (2018). Acute phase proteins and stress characterization of caecal microbiome composition of broilers and free-range slow- markers in the immediate response to a combined vaccination against Newcastle growing chickens throughout their productive lifespan. Sci. Rep. 9, 2506. doi:10. disease and infectious bronchitis viruses in specific pathogen free (SPF) layer chicks. 1038/s41598-019-39323-x Poult. Sci. 97, 463–469. doi:10.3382/ps/pex340 Pasolli, E., Asnicar, F., Manara, S., Zolfo, M., Karcher, N., Armanini, F., et al. Kanehisa, M., Sato, Y., and Kawashima, M. (2021). KEGG mapping tools for (2019). Extensive unexplored human microbiome diversity revealed by over 150, uncovering hidden features in biological data. Protein Sci. 31, 47–53. doi:10.1002/pro. 000 genomes from metagenomes spanning age, geography, and lifestyle. Cell 176, 4172 649. doi:10.1016/j.cell.2019.01.001 Khan, M. A. W., Stephens, W. Z., Mohammed, A. D., Round, J. L., and Kubinak, Rahn, K., De Grandis, S. A., Clarke, R. C., McEwen, S. A., Galán, J. E., Ginocchio, J. L. (2019). Does MHC heterozygosity influence microbiota form and function? C., et al. (1992). Amplification of an invA gene sequence of Salmonella typhimurium PLoS One 14, e0215946. doi:10.1371/journal.pone.0215946 by polymerase chain reaction as a specific method of detection of Salmonella. Mol. Cell. Probes 6, 271–279. doi:10.1016/0890-8508(92)90002-f Kollarcikova, M., Kubasova, T., Karasova, D., Crhanova, M., Cejkova, D., Sisak, F., et al. (2019). Use of 16S rRNA gene sequencing for prediction of new Ramlucken, U., Lalloo, R., Roets, Y., Moonsamy, G., van Rensburg, C. J., and opportunistic pathogens in chicken ileal and cecal microbiota. Poult. Sci. 98, Thantsha, M. S. (2020). Advantages of Bacillus-based probiotics in poultry 2347–2353. doi:10.3382/ps/pey594 production. Livest. Sci. 241, 104215. doi:10.1016/j.livsci.2020.104215 Lee, K.-C., Kil, D. Y., and Sul, W. J. (2017). Cecal microbiome divergence of Raza, A., Bashir, S., and Tabassum, R. (2019). An update on carbohydrases: broiler chickens by sex and body weight. J. Microbiol. 55, 939–945. doi:10.1007/ Growth performance and intestinal health of poultry. Heliyon 5, e01437. doi:10. s12275-017-7202-0 1016/j.heliyon.2019.e01437 Li, R., Li, Y., Fang, X., Yang, H., Wang, J., Kristiansen, K., et al. (2009). SNP Reid, W. V., Chen, D., Goldfarb, L., Hackmann, H., Lee, Y. T., Mokhele, K., detection for massively parallel whole-genome resequencing. Genome Res. 19, et al. (2010). Environment and development. Earth system science for global 1124–1132. doi:10.1101/gr.088013.108 sustainability: Grand challenges. Science 330, 916–917. doi:10.1126/science. 1196263 Li, C.-L., Wang, J., Zhang, H.-J., Wu, S.-G., Hui, Q.-R., Yang, C.-B., et al. (2018). Intestinal morphologic and microbiota responses to dietary Bacillus spp. in a broiler Ruiz, J. A., Pérez-Vendrell, A. M., and Esteve-García, E. (1999). Effect of β- chicken model. Front. Physiol. 9, 1968. doi:10.3389/fphys.2018.01968 carotene and vitamin E on oxidative stability in leg meat of broilers fed different supplemental fats. J. Agric. Food Chem. 47, 448–454. doi:10.1021/jf980825g Lieberman-Aiden, E., van Berkum, N. L., Williams, L., Imakaev, M., Ragoczy, T., Telling, A., et al. (2009). Comprehensive mapping of long-range interactions reveals Scanes, C. G. (2016). Biology of stress in poultry with emphasis on glucocorticoids and folding principles of the human genome. Science 326, 289–293. doi:10.1126/science. the heterophil to lymphocyte ratio. Poult. Sci. 95, 2208–2215. doi:10.3382/ps/pew137 1181369 Schuetzenmeister, and Dufey (2017). Vca: Variance component analysis 1.3. Liu, J., Gu, Z., Song, F., Zhang, H., Zhao, J., and Chen, W. (2019). Lactobacillus 3 [Software]. CRAN. R-project. org/package= VCA>[zitiert am 02.11. plantarum ZS2058 and lactobacillus rhamnosus GG use different mechanisms to Shaffer, M., Borton, M. A., McGivern, B. B., Zayed, A. A., La Rosa, S. L., Solden, L. M., prevent Salmonella infection in vivo. Front. Microbiol. 10, 299. doi:10.3389/fmicb. et al. (2020). DRAM for distilling microbial metabolism to automate the curation of 2019.00299 microbiome function. Nucleic Acids Res. 48, 8883–8900. doi:10.1093/nar/gkaa621 Livingston, M. L., Cowieson, A. J., Crespo, R., Hoang, V., Nogal, B., Browning, M., Shah, T. M., Patel, J. G., Gohil, T. P., Blake, D. P., and Joshi, C. G. (2019). Host et al. (2020). Effect of broiler genetics, age, and gender on performance and blood transcriptome and microbiome interaction modulates physiology of full-sibs chemistry. Heliyon 6, e04400. doi:10.1016/j.heliyon.2020.e04400 broilers with divergent feed conversion ratio. NPJ Biofilms Microbiomes 5, 24. Lundberg, R., Scharch, C., and Sandvang, D. (2021). The link between broiler doi:10.1038/s41522-019-0096-3 flock heterogeneity and cecal microbiome composition. Anim. Microbiome 3, 54. Slawinska, A., Dunislawska, A., Plowiec, A., and Radomska, M. (2019). doi:10.1186/s42523-021-00110-7 Modulation of microbial communities and mucosal gene expression in chicken Ma, D., Bou-Sleiman, M., Joncour, P., Indelicato, C.-E., Frochaux, M., Braman, intestines after galactooligosaccharides delivery in Ovo. PLoS. doi:10.1371/journal. V., et al. (2019). Commensal gut bacteria buffer the impact of host genetic variants pone.0212318 Frontiers in Physiology 13 frontiersin.org Tous et al. 10.3389/fphys.2022.884925 Sprockett, D., Fukami, T., and Relman, D. A. (2018). Role of priority effects in the Wen, C., Yan, W., Mai, C., Duan, Z., Zheng, J., Sun, C., et al. (2021). Joint early-life assembly of the gut microbiota. Nat. Rev. Gastroenterol. Hepatol. 15, contributions of the gut microbiota and host genetics to feed efficiency in chickens. 197–205. doi:10.1038/nrgastro.2017.173 Microbiome 9, 126. doi:10.1186/s40168-021-01040-x Tarradas, J., Tous, N., Esteve-Garcia, E., and Brufau, A. J. (2020). The control of Whelan, R. A., Doranalli, K., Rinttilä, T., Vienola, K., Jurgens, G., and Apajalahti, intestinal inflammation: A major objective in the research of probiotic strains as J. (2019). The impact of Bacillus subtilis DSM 32315 on the pathology, performance, alternatives to antibiotic growth promoters in poultry. Microorganisms 8, E148. and intestinal microbiome of broiler chickens in a necrotic enteritis challenge. Poult. doi:10.3390/microorganisms8020148 Sci. 98, 3450–3463. doi:10.3382/ps/pey500 Tarsani, E., Kranis, A., Maniatis, G., Avendano, S., Hager-Theodorides, A. L., and Wicher, K. B., and Fries, E. (2006). Haptoglobin, a hemoglobin-binding Kominakis, A. (2019). Discovery and characterization of functional modules plasma protein, is present in bony fish and mammals but not in frog and associated with body weight in broilers. Sci. Rep. 9, 9125. doi:10.1038/s41598- chicken. Proc. Natl. Acad. Sci. U. S. A. 103, 4168–4173. doi:10.1073/pnas. 019-45520-5 0508723103 van der Hee, B., and Wells, J. M. (2021). Microbial regulation of host physiology Wick, R. R., Judd, L. M., Gorrie, C. L., and Holt, K. E. (2017). Unicycler: Resolving by short-chain fatty acids. Trends Microbiol. 29, 700–712. doi:10.1016/j.tim.2021. bacterial genome assemblies from short and long sequencing reads. PLoS Comput. 02.001 Biol. 13, e1005595. doi:10.1371/journal.pcbi.1005595 Vasdal, G., Granquist, E. G., Skjerve, E., de Jong, I. C., Berg, C., Michel, V., et al. Yan, W., Sun, C., Yuan, J., and Yang, N. (2017). Gut metagenomic analysis reveals (2019). Associations between carcass weight uniformity and production measures prominent roles of Lactobacillus and cecal microbiota in chicken feed efficiency. Sci. on farm and at slaughter in commercial broiler flocks. Poult. Sci. 98, 4261–4268. Rep. 7, 45308. doi:10.1038/srep45308 doi:10.3382/ps/pez252 Zhang, B., Li, G., Shahid, M. S., Gan, L., Fan, H., Lv, Z., et al. (2020). Dietary Vilà, B., Esteve-Garcia, E., and Brufau, J. (2010). Probiotic micro-organisms: l-arginine supplementation ameliorates inflammatory response and alters gut 100 years of innovation and efficacy; modes of action. World’s. Poult. Sci. J. 66, microbiota composition in broiler chickens infected with Salmonella enterica 369–380. doi:10.1017/S0043933910000474 serovar Typhimurium. Poult. Sci. 99, 1862–1874. doi:10.1016/j.psj.2019.10.049 Volf, J., Polansky, O., Sekelova, Z., Velge, P., Schouler, C., Kaspers, B., et al. Zhong, C.-Y., Sun, W.-W., Ma, Y., Zhu, H., Yang, P., Wei, H., et al. (2015). (2017). Gene expression in the chicken caecum is dependent on microbiota Microbiota prevents cholesterol loss from the body by regulating host gene composition. Vet. Res. 48, 85. doi:10.1186/s13567-017-0493-7 expression in mice. Sci. Rep. 5, 10512. doi:10.1038/srep10512 Wang, Y., Cao, X., Luo, C., Sheng, Z., Zhang, C., Bian, C., et al. (2020). Multiple Zhou, H., Ni, W.-J., Meng, X.-M., and Tang, L.-Q. (2020). MicroRNAs as regulators of ancestral haplotypes harboring regulatory mutations cumulatively contribute to a QTL immune and inflammatory responses: Potential therapeutic targets in diabetic affecting chicken growth traits. Commun. Biol. 3, 472. doi:10.1038/s42003-020-01199-3 nephropathy. Front. Cell Dev. Biol. 8, 618536. doi:10.3389/fcell.2020.618536 Frontiers in Physiology 14 frontiersin.org