The pig gut microbiota is a complex ecosystem composed of microbial guilds, which until recently were unknown. Here, we decomposed the pig fecal microbiota of 648 healthy pigs during the transition (n=400) and growing-finishing (n=248) periods in co-occurring bacterial communities defined as pig enterosignatures (pig ES). Our results indicate that fecal microbial ecosystems can accurately be described by combinations of at least six pig ES, driven by Prevotella (ES-Prev), Treponema (ES-Trep), Lactobacillus (ES-Lact), Clostridium (ES-Clost), Streptococcus (ES-Strep), and UBA2810 (ES-UBA2) genera. We observed a dynamic shift in the composition of pig ES, where ES-Prev, ES-Strep, and ES-Lact seem to be core components, while ES-Trep, ES-Clost, and ES-UBA play age-specific roles. Our results suggest partial genetic control by the host, with heritabilities of pig ES composition ranging from 0.30 to 0.42. Furthermore, our findings indicate that stress can impact pig ES assembly, decreasing ES-Lact abundance while increasing the prevalence of ES-Strep. We noted a positive link between ES-Prev and growth at 60 days, which later evolved negatively, impacting feed efficiency during the finish-growing period. Additionally, we observed a negative association of ES-Strep and a positive impact of ES-Lact on growth and feed efficiency during the growing period. Remarkably, a negative association between the abundance of ES-Lact and levels of hair cortisol was also observed during this period. Overall, our findings confirm the key ecological roles of Prevotella, Treponema, Lactobacillus, and Streptococcus from previous studies in pigs and reveal novels links with relevant porcine physiological and performance traits. Finally, despite pig ES proving valuable in dissecting microbial communities into assemblies of underlying microbial guilds, showcasing its potential as a novel tool in breeding and precision farming. Our findings emphasize the importance of customizing microbial consortia to meet the nutritional and health requirements at each stage or category of the porcine production chain.
Research Article
Enterosignatures of the Fecal Pig Microbiota: Exploring Determinants and Revealing Host-Performance Consequences
https://doi.org/10.21203/rs.3.rs-3978889/v1
This work is licensed under a CC BY 4.0 License
Version 1
posted
You are reading this latest preprint version
Global pig production is one of greatest in livestock breeding industry. However, pig production is faced with rising feed costs and the emergence of pig diseases, while having to meet high standards of animal welfare, product quality and sustainability. It has been shown that health, nutrition, and growth of pigs are mediated by the gut microbiota (Vasquez et al. 2022; Ramayo-Caldas et al. 2022). A deepen knowledge of the genetic basis of gut microbiota and its relationship to genetic and environmental determinants is crucial for a sustainable and improved pig livestock. The gastrointestinal tract (GIT) microbiome is a diverse ecosystem that contains microorganisms from several kingdoms such as bacteria, archaea, protozoa, fungi and virome (Billaud et al. 2021). Diet and management have a major impact in the ecosystem composition while host-genetic has a more taxa-specific impact. Thus, the composition of gut microbiome varies depending on several factors such as diet, age, geography, season of the years, and host genetics (Fan et al. 2020, Nicholson et al. 2012, Yang et al. 2016, Ramayo-Caldas et al. 2016, Reverter et at. 2021, Ramayo-Caldas et al. 2020).
In livestock, particularly in pigs, many studies have focused on taxonomic decomposition, including the identification of distinct taxonomic levels that are the most prevalent across samples. Particularly, by analysing the relative abundances of genera, samples can be classified into distinct clusters known as “enterotypes” (Arumugam at al. 2011). To identify those enterotypes, a distance matrix is calculated based on the relative abundance of each genus in each sample using methods such as Jensen-Shannon divergence to measure the similarity between two probability distributions. Then, clustering is applied to the generated distance matrix using the partitioning around medoids (PAMs) algorithm, while the optimal cluster number is obtained using the Calinski-Harabasz (CH) index (Arumugam et al. 2011). Other algorithms such as Dirichlet multinomial mixture models (DMM) have also been employed, in studies of human gut microbiota. Enterotypes in animals have been detected in several species including pigs. Analysis performed on a cohort of 518 piglets indicated two enterotype-like groups, driven by either Ruminococcus and Treponema or Prevotella and Mitsuokella genera (Ramayo-Caldas et al. 2016). In another study, Lu et al. (2018) used a crossbred population of pigs to study at different times its faecal microbiota composition and diversity regarding host-genetics background. They also found two distinct enterotypes for the weaning animals, a Prevotella-enriched and an Escherichia-enriched communities. Moreover, two different sets of clusters were identified in a longitudinal analysis from post-weaning to finishing (Le Sciellour et al. 2019). At 52 days, the detected enterotypes were driven by Lactobacillus and Prevotella-Sarcina, while from 99 to 154 days the genus defining the enterotypes were Lactobacillus and Turicibacter-Clostridium. While clustering methods used for the enterotypes have contributed to understanding the diversity of pig gut microbiota, their implementation assumes that each of the identified communities is driven only by distinct dominant phylogenetic groups. Thus, they may fail to capture the co-existence of bacterial guilds under various conditions that can affect the assembly of the microbial ecosystem. Furthermore, the allocation of samples to each group results in a categorical variable. Despite these limitations, previous studies have reported association between the defined enterotypes and several performance traits including growth, body weight (BW), and average daily gain (ADG) (Mach et al. 2015, Ramayo-Caldas et al. 2016, Lu et al. 2018, Le Sciellour et al. 2019).
To overcome the aforementioned limitations of enterotyping, Frioux et al. (2023) introduced the complementary concept of enterosignature (ES), defined as microbial signatures of commonly co-occurring bacterial guilds. Compared to enterotypes, Frioux et al. (2023) proposal is based on the identification of microbial guilds composed of co-occurring genera that can be combined to accurately characterize gut microbial composition. Therefore, considering the intrinsic ecosystem information, and returning the proportional representation of bacterial assemblages rather than a categorical classification for each sample. They found that the human gut microbiome can be characterized as a combination of at least five complementary ESs in their metabolism. Human ESs are driven by the primary dominant taxa Bacteroides, Firmicutes, Prevotella, Bifidobacterium and Escherichia. The decomposition of human gut microbiome was performed by the non-negative matrix factorization (NMF, Lee et al. 1999), a multivariate approach where the latent variables are continuous rather than categorical. While the approach has been applied to human gut microbiome, from the best of our knowledge, the occurrence of ESs in the pig microbiota has not been investigated yet.
Here, we explore for the first time the existence of ESs in pig faecal microbiota. The primary objective of our study was to decompose faecal pig microbiota into bacterial guilds. Our focus extends to investigate the composition of pig ESs, identifying key divers and determining whether they exhibit stable or dynamic fluctuations across time. Additionally, we aim to explore the diversity of ES across different host and environmental determinants to reveal their association with host performance.
Experimental design and sample collection
Animal care and experimental procedures were carried out following national and institutional guidelines for the Good Experimental Practices and were approved by the IRTA Ethical Committee. A total of 648 healthy Duroc pigs from the same commercial genetic line, at two different ages, and allocated in two distinct farms were employed (Figure 1). Faecal samples were collected at 60 ± 8 days of age from 400 weaned piglets (201 males and 199 females) distributed across six batches on commercial farm conditions. Refer to (Ballester et al. 2020) and (Ramayo et al. 2021) for more information about animal management and data collection.
The remaining 248 pigs (119 castrated males and 129 females) were reared during 2017 and 2019 at IRTA experimental farm in Monells (Girona, Spain). Faecal samples were taken at 190 ± 10 days of age when pigs were fed a finishing standard diet. For 134 out of the 248 pigs, BW (kg) and average daily feed intake (kg) were recorded using both a weighing scale and electronic feeding stations during the experimental period (INSENTEC). These comprehensive datasets were employed to calculate Average Daily Gain (ADG), Feed Conversion Ratio (FCR), and Residual Feed Intake (RFI), computed from a model after adjusted by the metabolic weight, ADG, and the backfat thickness at the end of the period. Moreover, 96 out of the 248 growing pigs, comprising 48 castrated males and 48 females, were allocated into eight pens of 12 animals (six castrated males and six females per pen). Five out of the 96 pigs were excluded, while the remaining 91 were distributed across two treatments.: T1 – control group (N=44), involving pigs not mixed during the growing-finishing period; and T2 – stress group (N=47), inducing social stress by mixing the pigs three times during the growing-finishing period. Cortisol (CORT) levels in hair samples of these pigs were analysed at the end of the experiment as follow: 150 mg of hair was weighted from each hair sample and placed into a 50-ml conical tube. After three washes with 3 ml of isopropanol, all samples were left to dry in an extractor hood during 12 h. Dried hair samples were cut into 2–3 mm pieces using scissors, and 50 mg were transferred into 2 ml eppendorf. One ml of methanol was added to each sample and incubated 18 h at 37 °C under moderate shaking (100 rpm). After incubation, extracted samples were centrifuged at 7000g for 2 min and 700 µl of supernatant was transferred to a new 1.5 ml tube. The supernatant was placed into a speed vac for 2 h to evaporate the methanol. The dried extracts were stored at −20 °C until analysis. Total concentrations of CORT were measured by ELISA kit (Cusabio Technology LLC., Bionova, Spain) with dried samples reconstituted with 210 µl of sample diluent. Samples were quantified by reference to standard curves constructed with known concentrations of pig cortisol dilutions of the Standard. Absorbance was read at 450 nm using an ELISA plate reader (Bio-Rad) and analysed using the Microplate manager 5.2.1 software (Bio-Rad).
Microbial DNA extraction, sequencing, and bioinformatics analysis
DNA was extracted from the fecal samples with the DNeasy PowerSoil Kit (QIAGEN, Hilden, Germany) following manufacturer’s instructions. Extracted DNA was sent to the Keck Center at the University of Illinois for Fluidigm sample preparation and paired-end (2 × 250 nt) sequencing on an Illumina NovaSeq (Illumina, San Diego, CA, USA). The 16S rRNA gene fragment was amplified using the primers V3_F357_N: 5'-CCTACGGGNGGCWGCAG-3' and V4_R805: 5'-GACTACHVGGGTATCTAATCC-3'. Sequences were analysed with Qiime2 (Bolyen et al. 2019). Barcode sequences, primers and low-quality reads (Phred scores of < 30) were removed. The quality control also trimmed sequences based on the expected amplicon length and removed chimaeras. Afterwards, sequences were processed into Amplicon Sequences Variants (ASVs) at 99% of identity. ASVs present in less than two samples and representing less than 0.0001% of the total counts were filtered out. Samples with less than 10,000 reads were also excluded. ASVs were classified to the lowest possible taxonomic level based on a primer-specific trained version of GreenGenes2 database (released October 2022) (McDonald et al. 2023).
Enterosignatures identification
Following Frioux et al (2023), to evaluate the existence of enterosignatures in pigs, a non-negative matrix factorization (NMF) approach was applied to the normalized genera relative abundance table. NMF decomposed the initial scaled table to two new ones: W, defined as the weight of the genera in ESs, and the second one, H, defined as the existence of ESs across samples (Supplementary Figure 1). Particularly, W represents the contribution of the genera (n in rows) to the resulted ESs (k in columns) when normalized by the columns. The optimal number of ESs was evaluated by performing a nine-fold bicross validation (Owen and Perry 2009) as it was suggested in Frioux et al (2023). For each of the 9-folds, the model uses 8/9 of the folds as the training set, while the performance was validated on the remaining data set (1/9). The process was repeated 200 times for a k number of ESs ranging from 2 to 30, each time the matrix was randomly shuffled. The optimal k of pig ES was determined by tracking the derived evaluation metrics corresponding to each ES, such as reconstruction error, explained variance, and cosine similarity. More specifically, the best k was considered once the gain of the previous metrics used to assess the quality of the decomposition, does not change. Pigs ES computation was performed by applying NMF with Python 3.9, SciPy v1.7.3. and Scikit-Learn v0.24.1 for 100 runs, a maximum number of 2000, random initialization, the multiplicative update solver and Kullback-Leibler divergence as a beta-loss function. Additionally, an L1 regularization parameter equal to 1.0 was selected and applied to both H and W matrices.
Enterosignature analyses
After reducing the putative number of clusters, a second step of the ESs evaluation was done using the enterosignatures_nmf.py script (https://gitlab.inria.fr/cfrioux/enterosignature-paper) for each of the two data set (400 and 248 samples). The script uses the previously estimated k optimal ESs as hyperparameter. In addition to this evaluation, we run the Python script for a range of k [2-10], keeping the output evaluation scores to define the optimal k. During each run, the NMF algorithm implemented the decomposition of the initial relative abundance of genera matrix, resulting in the two new complementary matrices (H and W). The driven genus in each k-ESs was identified based on the higher contribution value across all the genera, while the assignment probability of genera to ESs is retrieved by normalizing the W matrix row-wise. Thus, the ESs are assigned to each sample based on the higher relative abundance value obtained by the normalized H matrix. The same H matrix is used to define the number and set of ESs needed to explain most of variance in relation to host covariates or phenotypes (e.g. age, sex, BW, ADG, RFI). The accumulated relative abundance is calculated for a k of ESs in each sample. Then, we applied the same cut-off of 90% used in Frioux et al. (2023) to filter out k of ESs below this threshold.
Determinants of pig ES and association with host performance
We aim to explore the determinants influencing pig ES assembly, considering both host-specific and environmental factors, and examine associations between pig ES and host productive traits. Our dataset comprises 648 samples gathered at various ages (N=400 at 60 days vs N=248 at 190 days of life), from different farms (N=400 commercial vs N=248 experimental facilities), and from diverse experimental setups. Consequently, metadata information varies between the samples. To address this, we conducted separate analyses for the 400 and 248 samples based on the available information as outlined below:
1. The composition of ESs in the 400 piglets was first studied in two discrete traits: batch and sex effects. For the batch category, there were six different groups with number of samples 57, 68, 73, 71, 72, and 59. The sex category contains 201 male piglets and 199 females. For this type of traits, pie plots were generated to demonstrate the minimum number of ESs needed to explain most of the variance within each category, as well as the ES distribution across each category. Specifically, for each sample, the variance explained by the ES was considered as the corresponding abundance value. Then, the cumulative sum of the explained variance was calculated gradually, keeping the minimum number of ESs that explain 100% of the variance. The frequency of specific ES that appeared in the corresponding selected number was also calculated. Furthermore, the dynamics and association of ES abundance with age and BW were also investigated across the 400 samples using ggplot (Wickham, 2016) and geom_smooth() option with method “lm” to add a regression line, where “lm” stands for a classical linear model.
2. On the other hand, concerning the 248 piglets raised under experimental conditions, our initial focus was on examining the diversity of ESs across sex, encompassing 119 males and 129 females. Subsequently, we conducted two independent analyses to explore the progression of pig ESs based on quantitative traits:
2.1 The dataset comprising 134 samples was utilized to investigate the relationship between pig ES with ADG and feed efficiency traits (FCR, RFI).
2.2 The impact of stress challenge on pig ES assignment was evaluated based on the dataset consisting of 91 samples (47 stress vs. 44 control). Additionally, this dataset was employed to assess the association between ES abundance and levels of hair cortisol.
Genotyping, and estimation of genetic parameters
After evaluating host-specific and environmental factors affecting pig ES, the potential genetic impact of the host was investigated using the 400 samples collected at 60 ± 8 days of age. The utilization of this dataset with a larger sample size than the growing pigs dataset is expected to improve the power and the accuracy of estimated genetic parameters, including heritability (h2) and the genetic correlation between pig ES and BW at the time of sampling (60 days). The Porcine 70 k GGP Porcine HD Array (Illumina, San Diego, CA) was used to genotype the 400 animals. The quality control excluded SNPs with minor allele frequencies < 5%, rates of missing genotypes above 10%, and SNPs that did not map to the porcine reference genome (Sscrofa11.1 assembly). The h2 of ES was estimated by applying Reproducing Kernel Hilbert Space (RKHS, Gianola et al. 2006) from the BGLR package (Pérez and de Los Campos, 2014) separately for each of them using the following model:
Enterosignatures of the pig fecal microbiota
A total of 1,755 Amplicon Sequence Variants (ASVs) were identified from the 16S rRNA gene sequence data of 648 pigs. In agreement with previous reports (Xiao et al., 2016; Chen et al., 2021a), the dominant bacterial phyla were Firmicutes and Bacteroidetes. Supplementary Figure 2(A) shows the boxplot of the explained microbial variance obtained for the 400 pigs dataset with different k values ranging from 2 to 30. The explained variance plateaued at approximately k=5, showing minimal fluctuations in the mean value with a total explained variance of the original genus abundance around 74%. Afterwards, we accessed the optimal k by computing NMF on the whole dataset, for k varying from 2 to 10, separately and keeping the corresponding explained variance, cosine similarity, and reconstruction error. Supplementary Figure 2(B) shows that at k=5, the evaluation metrics were stabilized with no significant fluctuations observed. More specifically, the model of five pig ES explained 91% variance of the original genus abundances with cosine similarity equal to 0.95. Thus, based on this, we propose the existence of at least five pig ES at 60 days.
The taxonomic composition of the five pig ES was characterized by: ES-Prev (mainly dominated by Prevotella), ES-Trep (Treponema), ES-Lact (Lactobaicllus), ES-Clos (Clostridium), and ES-Stre (Streptococcus). Following the original proposal (Frioux et al. 2023), the driven genus in each ES was identified based on its highest contribution as is indicated by the normalized W matrix. Figure 2 shows the contents of each enterosignature displaying only genera with percentage contribution higher than 4% at least in one ES. As summarized in Figure 2, the combination of bacterial guilds differed dramatically across pig ES in terms of genera and composition. For example, Clostridim was found at percentage 1% in ES-Prev whereas at 60% in ES-Clos. Moreover, some pig ESs such as ES-Clos, were driven by one or two fully dominated genera, while others were determined by many genera with moderate or weak contributions, such as ES-Trep. The full composition of the five ESs can be seen in Supplementary Figure 3 and Supplementary file 1. Particularly, a significant contribution of genera with composition lower than 4% can be seen in ES-Prev and ES-Trep, while in the other three ES the effect of the driven genera was dominant. In addition, Figure 3 shows the evolution of genera assignment to each ES with the chosen number of ES. As can be seen, ES-Prev, ES-Trep, and ES-Lact are the most prominent across the different ranks (from k=2 to k=10), exhibiting the larger prevalence and strongest associations to the corresponding combination of genera at 60 days of age.
Next, we followed the same pipeline to investigate the number and taxonomic composition using the dataset comprising 248 samples of 190-day-old pigs. Supplementary Figure 4 resumes the explained variance (Supplementary Figure 4A) and the evaluation score (Supplementary Figure 4B). In this case, the evaluation score suggested the optimal NMF solution at k=4. The four ESs explained 87% of the original genus abundances, with cosine similarity equal to 94%, and reconstruction error 10.54. Based on this model, the primary ES driving the variance in the 248 pig microbiota samples were: ES-Prev (Prevotella), ES-Lact (Lactobacillus), ES-UBA2 (UBA2810), and ES-Strep (Streptococcus). The combinations of co-occurring genera and their relative genus-level composition can be found in Figure 4. The full taxonomic composition of the four ES can be found in Supplementary file 2.
Figure 4 shows that ES-Prev and ES-Strep are characterized by multiple genera, with the driven genus demonstrating a moderate association to the ES, whereas ES-Lact and ES-UBA2 are mostly represented by their corresponding dominant genera (Lactobacillus and UBA2810, respectively), found in high percentage. This can be observed in more detail in Supplementary Figure 5. Specifically, genera of relative composition lower than 4% can be found in abundance in ES-Prev and ES-Strep, both resulting in multiple bacteria guilds. On the other hand, ES-Lact and ES-UBA2 represent bacteria guilds observed frequently by the dominant Lactobacillus and UAB2810 genera. Finally, the evolution of genera assignments for the 248 samples is depicted in Figure 5. The plot shows the evolution for a rank of decomposition ranging from 4 to 6 since NMF did not decompose the relative abundance matrix for lower than 4 or larger than 6 ranks.
Determinants and consequences of pig faecal ES at 60 days
Beyond the number of pig ES and their taxonomic composition, we aim to understand how pig ESs are assembled and evolve under the impact of host and environmental factors. We analysed each dataset related to different discrete and quantitative traits. Supplementary Figure 6 shows results regarding the number and frequency of ESs across the batch of the 400 samples from 60d-old pigs. In most groups, three ESs were adequate to explain the variance within these samples. The primary ES across the different groups appeared to be ES-Prev, followed by ES-Trep and ES-Lact. It can also be observed that ES-Strep and ES-Clos are abundant in two out of the six batches. On the other hand, results from assessing sex effects confirmed the distribution of the five ES in both males and females (Supplementary Figure 7). As done by Frioux et al. (2023), we also explored the interaction between pig ESs dynamics and age. These observations can help us to further investigate the ecological conditions impacting the pig faecal microbiota. As illustrated in Figure 6, despite the small range, ES-Prev persistence decreased with age, while on the contrary ES-Trep was significantly and positively associated with it. Significant associations were also found between BW at 60 days and the abundance of ES, as illustrated in Figure 7. ES-Prev exhibited a strong positive association with BW (P=0.0006) (Supplementary Figure 8), whereas ES-Trep displayed an opposing pattern with a negative association with BW (P=0.025). It is noteworthy that these two primary ESs have established a robust exclusion pattern, predominantly observed in the dynamic evolution of age and BW traits.
Determinants and implications of pig ES during growing-finishing period
A comparable assessment of sex effects was conducted based on the dataset of 248 pigs at 190 days old. Concerning the sex effect, and similar to the observed patterns at 60-days, all ESs identified by the NMF model were consistently observed in females and castrated males (Supplementary Figure 9).
Relation with feed efficiency
The dataset comprising 134 samples was employed to explore the relationship of pig ES with ADG and feed efficiency traits, obtaining different association patterns (Figure 8). ES-Prev (P=0.0028) and ES-Strep (P=3.37E-6) were negatively associated with ADG. Conversely, positive association trends with ADG were found for ES-Lact (P=0.061) and ES-UBA (P=0.009). The negative effect of ES-Prev on feed efficiency was further supported by its positive association with RFI (P=0.023) and FCR (P=0.042).
Impact of stress on pig ES and association with cortisol in hair
As stress can influence faecal microbial ecosystems, we investigated the impact of stress challenges on the dominance of primary pig ESs during the growing-finishing period (Figure 9). The largest differences between stress and control groups were characterized by a depletion of ES-Lact and increased abundance of ES-Strep in stressed animals compared to those in the control group. In agreement with this finding, a negative and significant association (P=0.0023) was found between the abundance of ES-Lact and the levels of the stress-related hormone cortisol in hair (Figure 10).
Genetic parameters of pig ES
Regarding the putative host-genetic control on pig ES composition at 60 days of life, Figure 11 shows the mean and standard deviation of heritability posterior distributions for each ES. The highest heritability values can be observed with decreasing order for ES-Trep [h2=0.42], ES-Lact [h2=0.41], ES-Clos [h2=0.40], ES-Prev [h2=0.38], and ES-Strep [h2=0.31]. To be noted, our findings suggest that an important fraction of generic variance in pig ESs can be explained by the additive genomic value based on SNPs genotypes. Table 1 summarizes the phenotypic and genetic correlations among pig ES at 60 days. ES-Prev and ES-Trep showed the most robust negative correlation at both phenotypic (r=-0.712) and genetic levels (rg=-0.861). Additionally, negative, and significant genetic correlations were found between the abundance of ES-Strep and ES-Lact (rg=-0.647), as well as between ES-Trep and ES-Clost (rg=-0.647). Intriguingly, a positive and significant genetic correlation was noted between ES-Prev and ES-Strep (rg=0.642). Furthermore, consistently with the observed association with growth during the transition period, negative genetic correlations of ES-Trep (rg=-0.430) and ES-Strep (rg=-0.853) with BW were observed.
Here we present the first study exploring the existence of ESs within the faecal microbiota of pigs. Our results suggest that pig faecal microbiota can be described as combinations of at least six ESs, dominated either by Prevotella, Treponema, Lactobacillus, Clostridium, Streptococcus, or UBA2810 genera. Like in human microbiome (Frioux et al. 2023), we observed the dynamic nature of pig ES, with shifts occurring in response to the host's age. From our observations at two different production-categories (transition at 60 days and growing-finishing period at 190 days), we suggest Prevotella, Streptococcus, and Lactobacillus pig ES as core components of pig faecal microbiota. Meanwhile, Treponema, Clostridium, and UBA2810 seem to play an age-specific role in the dynamic and assembly of the pig faecal ecosystem. Furthermore, in agreement with previous reports (Ramayo-Caldas et al. 2016; Larzul et al. 2023) our results confirm strong co-exclusion between Prevotella and Treponema, playing a key ecological role in the faecal microbial ecosystem of 60- day-old pigs.
Our results also reveal a partial genetic control of the host, alongside the influence of age and management factors like stress on the assembly of pig ES. Estimated heritability was found in the range of 0.30 to 0.42. Thus, our findings agree with other studies in pigs where low to moderate heritability values of specific genera abundances have been reported (Ramayo-Caldas et al. 2020, Grieneisen et al. 2021, Aliakbari et al. 2021, Déru et al. 2022, Larzul et al. 2023). Furthermore, in agreement with Larzul et al. 2023, Treponema, the main driver of ES-Trep showed the larger heritability.
Regarding the impact of stress, we observed that prolongated stress challenges can reduce the abundance of ES-Lact while increasing the prevalence of ES-Strep. Moreover, supporting its positive links with well-being in growing pigs, we observed a negative association between the abundance of ES-Lact and levels of cortisol in hair. Our findings align with previous research, which has consistently shown that Lactobacillus, the central genera of ES-Lact, is positively associated with beneficial health outcomes (Upadhaya & Kim, 2022a; Lee et al., 2023). Numerous studies also indicated a decrease in relative abundance of Lactobacillus under diverse stress challenges, such as heat stress (Xia et al., 2021; Hu et al., 2022a; Ringseis & Eder, 2022), weaning, and social stress (Jiang et al., 2020; Saladrigas-García et al., 2021; Upadhaya & Kim, 2022). Furthermore, research involving various species from the genus Lactobacillus and Limosilactobacillus reuteri (formerly classified as Lactobacillus reuteri) as probiotic supplements demonstrated a reduction in cortisol levels within the treated groups when compared to the control groups (Wang et al., 2013; Burdick Sanchez et al., 2019; Yang et al., 2020; Jiang et al., 2020). This reduction in cortisol levels is attributed to the regulation of the hypothalamic-pituitary-adrenal axis and the immune system, as well as the inhibition of inflammatory processes (Galbo & Kall, 2016; Bauer & Teixeira, 2019; Dillon & Wu, 2021). These findings suggested that higher abundance of Lactobacillus and Limosilactobacillus may contribute to lower cortisol levels in pigs, and are consistent with our findings. Therefore, we hypothesize that ES-Lact may be employed as an indicator of well-being during the growing-finishing period in pigs. Their putative beneficial effect to reduce stress during the porcine growing period deserves further research, even though Megasphaera, another member of ES-Lact, has demonstrated a positive association with low lesion scores in weaned pigs (Correa et al., 2023). The relationship between Megasphaera and social stress is less clear, with some studies suggesting the relative abundance of Megasphaera as a control indicator (Nguyen et al., 2023), and others showing a suggestive (p = 0.06) reduction of Megasphaera abundance in stressed groups (Li et al., 2018).
Finally, concerning stress indicators, our results report an increased abundance of ES-Strep in the stressed pigs. These results agree well with those of previous studies showing a risen relative abundance of Streptococcus and Treponema in social stress models (Nguyen et al., 2023) and heat stress models (Hu et al., 2022; Xiong et al., 2022). Paraprevotella, which is also a member of ES-Strep, has been proposed as a potential biomarker for depression and other mental disorders in humans, including attention deficit hyperactivity disorder (Barandouzi et al., 2020; Wu et al., 2022). In a separate study, Gilroy et al. in 2021 identified Candidatus Cryptobacteroides through metagenomic analyses of the chicken intestinal microbiome. However, the role of Cryptobacteroides in pig microbiota remains unclear, requiring further research.
Pig enterosignatures are associated to host-performance
We have investigated the association between pig ES and host-production traits, focusing separately on BW at 60days, as well as ADG, RFI, and FCR across the growing-finishing period. Consistently with earlier findings by Mach et al. (2015) and Ramayo-Caldas et al. (2016), our results confirm the positive link of Prevotella with BW and growth rate at 60 days. This association can be attributed to the Prevotella's remarkable capability to ferment complex polysaccharides (Ellekilde, 2014) and metabolize dietary fibre, resulting in the production of substantial amounts of short-chain fatty acids (SCFAs) (Ramayo-Caldas et al., 2016). Interestingly, UBA956 genera, which belongs to ES-Prev, also ferments dietary polysaccharides producing SCFAs (Flint et al., 2008). This process may contribute to the reduction of intestinal inflammation, thereby ensuring improved intestinal absorption capacity and increasing feed efficiency (Yang et al, 2017).
Furthermore, our results confirmed a negative association of Streptococcus (Aliakbari et al. 2022; McCormack et al. 2019) and positive of ES-Lact (Dowarah et al. 2017; Gardiner et al. 2020;) with growth and porcine feed efficiency during the growing-finishing period. Intriguingly, even though Prevotella is known to be beneficial after weaning, our results also showed that, together with ES-Strep, the abundance of ES-Prev was negatively associated to ADG at the growing-finish period. Additionally, ES-Prev exhibited an unfavourable impact on feed efficiency, displaying a positive link with RFI and FCR. In agreement with our findings, an increased abundance of Prevotella (mainly Prevotella copri) in lowly efficient pigs had been reported by Jiang et al. (2021). A role of Prevotella on fat deposition within a commercial Duroc population has been documented (Chen et al. 2021b). In growing pigs, the growth of lean tissue demands less energy than fat deposition, leading to a more efficient transformation of feed into muscle. Consequently, we hypothesise that the key role of Prevotella on fat deposition during the growing period might partially explain the observed adverse association between ES-Prev and feed efficiency traits.
Can ES information be incorporated in breeding programs or precision farming?
Modulating the gut microbiome represents a promising strategy for enhancing animal welfare and productivity, thus contributing to livestock production sustainability. This can be achieved through a combination of breeding and nutritional strategies, like for example the use of prebiotics, probiotics, or postbiotics. As proved by Frioux et al. (2023), and supported by our study, the fecal microbial ecosystems can be accurately described by combinations of ESs, explaining in our case between the 87% and 91% variance of the original genera abundance. Therefore, ES can provide useful information to reduce the multidimensionality of microbial ecosystems. We also showed the heritable nature of pig ES composition and its association with host performance. Thus, pig ES meets two crucial criteria for integration into breeding programs, potentially enhancing the prediction of breeding values and host performance, but also serving as a tool in microbiome-tailored precision farming. Based on our results, we stress the need of developing microbial consortiums that consider the nutritional and health requirements at each pig production stage. The observed association patterns of ES-Prev represent a good example of what might benefit during transition may not be beneficial during the pig growing-finishing period. This can be explained by the fact that young piglets may need microbial signatures to support their developing immune systems, or to facilitate the adaptation to new cereal-based diets after weaning. Meanwhile, during the growing-finishing period, pigs might benefit from a different combination of strains that promote nutrient absorption, protein degradation, and gut health. Therefore, we anticipated that unless accurate and evidence-based recommendations ensuring that pre/pro/postbiotic formulations are supported by studies demonstrating their efficacy for specific stage, there will be a law of diminishing returns applying microbiome-derived strategies to improve health, well-being and increase sustainability in livestock productions.
Finally, we recognize some limitations of our study, such as the limited taxonomic resolution achieved by metataxonomic approaches targeting the 16S rRNA gene with short reads. Moreover, our model is based on a dataset of a Duroc pig population at 60 days and at the end of the growing-finishing period (190 days). As a result, it is reasonable to assume that pig ES linked with specific porcine breeds, developmental stages, or management conditions such as the use of antimicrobials are not captured by our analysis. Consequently, we anticipate that combinations of additional microbial guilds might be of relevance in the assembly of pig faecal microbial ecosystems, thereby contributing to overall health, well-being, and sustainable livestock production.
The pig faecal microbiota of our Duroc population is characterized by a combination of at least six enterosignatures, dominated by either Prevotella, Treponema, Lactobacillus, Clostridium, Streptococcus, or UBA2810. Pig ES assembly changes dynamically with the age of the host, is impacted by stress, and partly shaped by host's genetics. There are relevant associations between the abundance of these pig ES communities and the host productive performance, which can diverge with the host’s age. Overall, our findings offer novel insights into the pig gut microbiota assembly, suggesting the usefulness of the enterosignatures concept in breeding and precision farming, but also underscoring the relevance of customizing microbial consortia based on the nutritional and health requirements at each stage of the porcine production chain.
NMF
Non-negative Matrix Factorization
RKHS
Reproducing Kernel Hilbert Space
GRM
Genetic Relationship Matrix
ASV
Amplicon Sequence Variant
ES
Enterosignatures
Prev
Prevotella
Trep
Treponema
Lact
Lactobacillus
Clos
Clostridium
Stre
Streptococcus
UBA2
UBA2810
BW
Body Weight
ADG
Average Daily Gain
RFI
Residual Feed Intake
FCR
Feed Conversion Ratio
Ethics approval and consent to participate
Animal care and experimental procedures were carried out following national and institutional guidelines for the Good Experimental Practices and were approved by the IRTA Ethical Committee. Consent to participate is not applicable in this study.
Consent for publication
Not applicable.
Availability of supporting data
The raw sequencing data employed in this article has been submitted to the NCBI’s Sequence Read Archive (https://www.ncbi.nlm.nih.gov/sra); and BioProject under accession number: PRJNA608629.
Competing Interests
The authors declare no competing interests
Funding
ITV was funded by the HoloRuminant (101000213) and GEroNIMO (101000236) H2020 projetcs. AC was funded with a PhD fellowship (PRE2022-101829) awarded by the Spanish Ministry of Education and Culture. YRC received the “Ramon y Cajal” grant (RYC2019-027244-I) funded by MCIN/AEI/ https://doi.org/10.13039/501100011033 and by “ESF Investing in your future”. Part of the research presented in this publication was funded by Grants AGL2017-88849-R and PID2020-112677RB-C21 awarded by the Spanish Ministry of Economy and Competitiveness.
Author Contributions
ITV and YRC designed the study. MB carried out DNA extraction. MB, RQ, ACFO, LBC and YRC performed the sampling. ITV and YRC analyzed the data. ITV, RR, AC, MB, MP, JPS, MP, RQ, AD and YRC interpreted the results and wrote the manuscript. All the authors read and approved the final version of the manuscript.
Acknowledgements
The authors warmly thank all technical staff from Selección Batallé S.A, for providing the animal material, and specially to Josep Reixach, for their collaboration during the experimental protocols and the sampling. We are grateful to David Pérez-Pascual for valuable discussions and critical reading of the manuscript.
- Aliakbari A, Zemb O, Billon Y, Barilly C, Ahn I, Riquet J, Gilbert H. (2021). Genetic relationships between feed efficiency and gut microbiome in pig lines selected for residual feed intake. J Anim Breed Genet. 138(4):491-507. doi: 10.1111/jbg.
- Aliakbari A, Zemb O, Cauquil L, Barilly C, Billon Y, Gilbert H. (2022). Microbiability and microbiome-wide association analyses of feed efficiency and performance traits in pigs. Genet Sel Evol. 54(1):29. doi: 10.1186/s12711-022-00717-7.
- Amadeu RR, Cellon C, Olmstead JW, Garcia AA, Resende MF, Muñoz PR. (2016). AGHmatrix: R Package to Construct Relationship Matrices for Autotetraploid and Diploid Species: A Blueberry Example. Plant Genome. 9(3). doi: 10.3835/plantgenome2016.01.0009.
- Arumugam M, Raes J, Pelletier E, Le Paslier D, Yamada T, Mende DR, Fernandes GR, Tap J, Bruls T, Batto JM, Bertalan M, Borruel N, Casellas F, Fernandez L, Gautier L, Hansen T, Hattori M, Hayashi T, Kleerebezem M, Kurokawa K, Leclerc M, Levenez F, Manichanh C, Nielsen HB, Nielsen T, Pons N, Poulain J, Qin J, Sicheritz-Ponten T, Tims S, Torrents D, Ugarte E, Zoetendal EG, Wang J, Guarner F, Pedersen O, de Vos WM, Brunak S, Doré J; MetaHIT Consortium; Antolín M, Artiguenave F, Blottiere HM, Almeida M, Brechot C, Cara C, Chervaux C, Cultrone A, Delorme C, Denariaz G, Dervyn R, Foerstner KU, Friss C, van de Guchte M, Guedon E, Haimet F, Huber W, van Hylckama-Vlieg J, Jamet A, Juste C, Kaci G, Knol J, Lakhdari O, Layec S, Le Roux K, Maguin E, Mérieux A, Melo Minardi R, M'rini C, Muller J, Oozeer R, Parkhill J, Renault P, Rescigno M, Sanchez N, Sunagawa S, Torrejon A, Turner K, Vandemeulebrouck G, Varela E, Winogradsky Y, Zeller G, Weissenbach J, Ehrlich SD, Bork P. (2011). Enterotypes of the human gut microbiome. Nature. 473(7346):174-80. doi: 10.1038/nature09944.
- Ballester M, Ramayo-Caldas Y, González-Rodríguez O, Pascual M, Reixach J, Díaz M, Blanc F, López-Serrano S, Tibau J, Quintanilla R. (2020). Genetic parameters and associated genomic regions for global immunocompetence and other health-related traits in pigs. Sci Rep. 10(1):18462. doi: 10.1038/s41598-020-75417-7.
- Barandouzi ZA, Starkweather AR, Henderson WA, Gyamfi A, Cong XS. (2020). Altered Composition of Gut Microbiota in Depression: A Systematic Review. Frontiers in Psychiatry, 11, 541. https://doi.org/10.3389/fpsyt.2020.00541.
- Bauer ME, & Teixeira AL. (2019). Inflammation in psychiatric disorders: What comes first? Annals of the New York Academy of Sciences, 1437(1), 57–67. https://doi.org/10.1111/nyas.13712.
- Billaud M, Lamy-Besnier Q, Lossouarn J, Moncaut E, Dion MB, Moineau S, Traoré F, Le Chatelier E, Denis C, Estelle J, Achard C, Zemb O, Petit MA. Analysis of viromes and microbiomes from pig fecal samples reveals that phages and prophages rarely carry antibiotic resistance genes. ISME Commun. 2021 Oct 25;1(1):55. doi: 10.1038/s43705-021-00054-8.
- Bolyen E, Rideout JR, Dillon MR, Bokulich NA, Abnet CC, Al-Ghalith GA, Alexander H, Alm EJ, Arumugam M, Asnicar F, Bai Y, Bisanz JE, Bittinger K, Brejnrod A, Brislawn CJ, Brown CT, Callahan BJ, Caraballo-Rodríguez AM, Chase J, Cope EK, Da Silva R, Diener C, Dorrestein PC, Douglas GM, Durall DM, Duvallet C, Edwardson CF, Ernst M, Estaki M, Fouquier J, Gauglitz JM, Gibbons SM, Gibson DL, Gonzalez A, Gorlick K, Guo J, Hillmann B, Holmes S, Holste H, Huttenhower C, Huttley GA, Janssen S, Jarmusch AK, Jiang L, Kaehler BD, Kang KB, Keefe CR, Keim P, Kelley ST, Knights D, Koester I, Kosciolek T, Kreps J, Langille MGI, Lee J, Ley R, Liu YX, Loftfield E, Lozupone C, Maher M, Marotz C, Martin BD, McDonald D, McIver LJ, Melnik AV, Metcalf JL, Morgan SC, Morton JT, Naimey AT, Navas-Molina JA, Nothias LF, Orchanian SB, Pearson T, Peoples SL, Petras D, Preuss ML, Pruesse E, Rasmussen LB, Rivers A, Robeson MS 2nd, Rosenthal P, Segata N, Shaffer M, Shiffer A, Sinha R, Song SJ, Spear JR, Swafford AD, Thompson LR, Torres PJ, Trinh P, Tripathi A, Turnbaugh PJ, Ul-Hasan S, van der Hooft JJJ, Vargas F, Vázquez-Baeza Y, Vogtmann E, von Hippel M, Walters W, Wan Y, Wang M, Warren J, Weber KC, Williamson CHD, Willis AD, Xu ZZ, Zaneveld JR, Zhang Y, Zhu Q, Knight R, Caporaso JG. (2019). Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat Biotechnol. 37(8):852-857. doi: 10.1038/s41587-019-0209-9.
- Burdick Sanchez, NC, Carroll JA, Broadway PR, Bass BE, Frank JW. (2019). Supplementation of a Lactobacillus acidophilus fermentation product can attenuate the acute phase response following a lipopolysaccharide challenge in weaned pigs. Animal, 13(1), 144–152. https://doi.org/10.1017/S1751731118001222.
- Chen C, Fang S, Wei H, He M, Fu H, Xiong X, Zhou Y, Wu J, Gao J, Yang H, Huang L. (2021b). Prevotella copri increases fat accumulation in pigs fed with formula diets. Microbiome. 9(1):175. doi: 10.1186/s40168-021-01110-0.
- Chen C, Zhou Y, Fu H, Xiong X, Fang S, Jiang H, Wu J, Yang H, Gao J, Huang L. (2021a) Expanded catalog of microbial genes and metagenome-assembled genomes from the pig gut microbiome. Nat Commun. 12(1):1106. doi: 10.1038/s41467-021-21295-0.
- Correa F, Luise D, Palladino G, Scicchitano D, Brigidi P, Martelli PL, Babbi G, Turroni S, Litta G, Candela M, Rampelli S, Trevisi, P. (2023). Influence of body lesion severity on oxidative status and gut microbiota of weaned pigs. Animal, 17(6), 100818. doi: 10.1016/j.animal.2023.100818.
- Dang T, Kumaishi K, Usui E, Kobori S, Sato T, Toda Y, Yamasaki Y, Tsujimoto H, Ichihashi Y, Iwata H. (2022). Stochastic variational variable selection for high-dimensional microbiome data. Microbiome. 10(1):236. doi: 10.1186/s40168-022-01439-0.
- Déru V, Tiezzi F, Carillier-Jacquin C, Blanchet B, Cauquil L, Zemb O, Bouquet A, Maltecca C, Gilbert H. (2022). Gut microbiota and host genetics contribute to the phenotypic variation of digestive and feed efficiency traits in growing pigs fed a conventional and a high fiber diet. Genet Sel Evol. 54(1):55. doi: 10.1186/s12711-022-00742-6.
- Dillon EL, & Wu G. (2021). Cortisol enhances citrulline synthesis from proline in enterocytes of suckling piglets. Amino Acids, 53(12), 1957–1966.
- Dowarah R, Verma AK, Agarwal N. (2017). The use of Lactobacillus as an alternative of antibiotic growth promoters in pigs: A review. Anim Nutr. 3(1):1-6. doi: 10.1016/j.aninu.2016.11.002.
- Ellekilde M, Selfjord E, Larsen CS, Jakesevic M, Rune I, Tranberg B, Vogensen FK, Nielsen DS, Bahl MI, Licht TR, Hansen AK, Hansen CH. (2014). Transfer of gut microbiota from lean and obese mice to antibiotic-treated mice. Sci Rep. 4:5922. doi: 10.1038/srep05922.
- Fan C, Zhang L, Fu H, Liu C, Li W, Cheng Q, Zhang H, Jia S, Zhang Y. (2020). Enterotypes of the Gut Microbial Community and Their Response to Plant Secondary Compounds in Plateau Pikas. Microorganisms. 8(9):1311. doi: 10.3390/microorganisms8091311.
- Flint HJ, Bayer EA, Rincon MT, Lamed R, White BA. (2008). Polysaccharide utilization by gut bacteria: potential for new insights from genomic analysis. Nat Rev Microbiol. 6(2):121-31. doi: 10.1038/nrmicro1817.
- Frioux C, Ansorge R, Özkurt E, Ghassemi Nedjad C, Fritscher J, Quince C, Waszak SM, Hildebrand F. (2023). Enterosignatures define common bacterial guilds in the human gut microbiome. Cell Host Microbe. 31(7):1111-1125.e6. doi: 10.1016/j.chom.2023.05.024.
- Galbo H, & Kall L. (2016). Circadian variations in clinical symptoms and concentrations of inflammatory cytokines, melatonin, and cortisol in polymyalgia rheumatica before and during prednisolone treatment: A controlled, observational, clinical experimental study. Arthritis Research & Therapy, 18(1), 174. https://doi.org/10.1186/s13075-016-1072-4.
- Gardiner GE, Metzler-Zebeli BU, Lawlor PG. (2020). Impact of Intestinal Microbiota on Growth and Feed Efficiency in Pigs: A Review. Microorganisms. 8(12):1886. doi: 10.3390/microorganisms8121886.
- Gianola D, Fernando RL, Stella A. (2006). Genomic-assisted prediction of genetic value with semiparametric procedures. Genetics. 2006 Jul;173(3):1761-76. doi: 10.1534/genetics.105.049510.
- Gilroy R, Ravi A, Getino M, Pursley I, Horton DL, Alikhan N-F, Baker D, Gharbi K, Hall N, Watson M, Adriaenssens EM., Foster-Nyarko E, Jarju S, Secka A, Antonio M, Oren A, Chaudhuri RR, La Ragione R, Hildebrand F, Pallen M J. (2021). Extensive microbial diversity within the chicken gut microbiome revealed by metagenomics and culture. PeerJ, 9, e10941. https://doi.org/10.7717/peerj.10941.
- Grieneisen L, Dasari M, Gould TJ, Björk JR, Grenier JC, Yotova V, Jansen D, Gottel N, Gordon JB, Learn NH, Gesquiere LR, Wango TL, Mututua RS, Warutere JK, Siodi L, Gilbert JA, Barreiro LB, Alberts SC, Tung J, Archie EA, Blekhman R. (2021). Gut microbiome heritability is nearly universal but environmentally contingent. Science. 373(6551):181-186. doi: 10.1126/science.aba5483.
- Holmes I, Harris K, Quince C. (2012). Dirichlet multinomial mixtures: generative models for microbial metagenomics. PLoS One. 7(2):e30126. doi: 10.1371/journal.pone.0030126.
- Hu C, Patil Y, Gong D, Yu T, Li J, Wu L, Liu X, Yu Z, Ma X, Yong Y, Chen J, Gooneratne R, Ju, X. (2022). Heat Stress-Induced Dysbiosis of Porcine Colon Microbiota Plays a Role in Intestinal Damage: A Fecal Microbiota Profile. Frontiers in Veterinary Science, 9. https://www.frontiersin.org/articles/10.3389/fvets.2022.686902.
- Jiang H, Fang S, Yang H, Chen C. (2021). Identification of the relationship between the gut microbiome and feed efficiency in a commercial pig cohort. J Anim Sci. 99(3):skab045. doi: 10.1093/jas/skab045.
- Jiang X, Lu N, Zhao H, Yuan H, Xia D, Lei H. (2020). The Microbiome–Metabolome Response in the Colon of Piglets Under the Status of Weaning Stress. Frontiers in Microbiology, 11. https://www.frontiersin.org/articles/10.3389/fmicb.2020.02055.
- Le Sciellour M, Renaudeau D, Zemb O. (2019). Longitudinal Analysis of the Microbiota Composition and Enterotypes of Pigs from Post-Weaning to Finishing. Microorganisms. 7(12):622. doi: 10.3390/microorganisms7120622.
- Lee DD, Seung HS. (1999). Learning the parts of objects by non-negative matrix factorization. Nature. 401(6755):788-91. doi: 10.1038/44565.
- Lee JH, Kim S, Kim ES, Keum GB, Doo H, Kwak J, Pandey S, Cho JH, Ryu S, Song M, Cho J H, Kim S, Kim HB. (2023). Comparative analysis of the pig gut microbiome associated with the pig growth performance. Journal of Animal Science and Technology, 65(4), 856–864. https://doi.org/10.5187/jast.2022.e122.
- Li Y, Guo Y, Wen Z, Jiang X, Ma X, Han X. (2018). Weaning Stress Perturbs Gut Microbiome and Its Metabolic Profile in Piglets. Scientific Reports, 8, 18068. https://doi.org/10.1038/s41598-018-33649-8.
- Lu D, Tiezzi F, Schillebeeckx C, McNulty NP, Schwab C, Shull C, Maltecca C. (2018). Host contributes to longitudinal diversity of fecal microbiota in swine selected for lean growth. Microbiome. 6(1):4. doi: 10.1186/s40168-017-0384-1.
- Mach N, Berri M, Estellé J, Levenez F, Lemonnier G, Denis C, Leplat JJ, Chevaleyre C, Billon Y, Doré J, Rogel-Gaillard C, Lepage P. (2015) Early-life establishment of the swine gut microbiome and impact on host phenotypes. Environ Microbiol Rep. 7(3):554-69. doi: 10.1111/1758-2229.12285.
- McCormack UM, Curião T, Metzler-Zebeli BU, Magowan E, Berry DP, Reyer H, Prieto ML, Buzoianu SG, Harrison M, Rebeiz N, Crispie F, Cotter PD, O'Sullivan O, Gardiner GE, Lawlor PG. (2019). Porcine Feed Efficiency-Associated Intestinal Microbiota and Physiological Traits: Finding Consistent Cross-Locational Biomarkers for Residual Feed Intake. mSystems. 4(4):e00324-18. doi: 10.1128/mSystems.00324-18.
- McDonald D, Jiang Y, Balaban M, Cantrell K, Zhu Q, Gonzalez A, Morton JT, Nicolaou G, Parks DH, Karst SM, Albertsen M, Hugenholtz P, DeSantis T, Song SJ, Bartko A, Havulinna AS, Jousilahti P, Cheng S, Inouye M, Niiranen T, Jain M, Salomaa V, Lahti L, Mirarab S, Knight R. (2023). Greengenes2 unifies microbial data in a single reference tree. Nat Biotechnol. 10.1038/s41587-023-01845-1. doi: 10.1038/s41587-023-01845-1.
- Nguyen TQ, Martínez-Álvaro M, Lima J, Auffret MD, Rutherford KMD, Simm G, Dewhurst R J, Baima ET, Roehe R. (2023). Identification of intestinal and fecal microbial biomarkers using a porcine social stress model. Frontiers in Microbiology, 14, 1197371. https://doi.org/10.3389/fmicb.2023.1197371.
- Nicholson JK, Holmes E, Kinross J, Burcelin R, Gibson G, Jia W, Pettersson S. (2012). Host-gut microbiota metabolic interactions. Science. 336(6086):1262-7. doi: 10.1126/science.1223813.
- Owen AB, & Perry PO. (2009). Bi-Cross-Validation of the SVD and the Nonnegative Matrix Factorization. The Annals of Applied Statistics, 3(2), 564–594. http://www.jstor.org/stable/30244256.
- Pérez P, de los Campos G. (2014). Genome-wide regression and prediction with the BGLR statistical package. Genetics. 198(2):483-95. doi: 10.1534/genetics.114.164442.
- Ramayo-Caldas Y, Mach N, Lepage P, Levenez F, Denis C, Lemonnier G, Leplat JJ, Billon Y, Berri M, Doré J, Rogel-Gaillard C, Estellé J. (2016). Phylogenetic network analysis applied to pig gut microbiota identifies an ecosystem structure linked with growth traits. ISME J. 10(12):2973-2977. doi: 10.1038/ismej.2016.77.
- Ramayo-Caldas Y, Prenafeta-Boldú F, Zingaretti LM, Gonzalez-Rodriguez O, Dalmau A, Quintanilla R, Ballester M. (2020). Gut eukaryotic communities in pigs: diversity, composition and host genetics contribution. Anim Microbiome. 2(1):18. doi: 10.1186/s42523-020-00038-4.
- Ramayo-Caldas Y, Zingaretti LM, Pérez-Pascual D, Alexandre PA, Reverter A, Dalmau A, Quintanilla R, Ballester M. (2021). Leveraging host-genetics and gut microbiota to determine immunocompetence in pigs. Anim Microbiome. 3(1):74. doi: 10.1186/s42523-021-00138-9.
- Reverter A, Ballester M, Alexandre PA, Mármol-Sánchez E, Dalmau A, Quintanilla R, Ramayo-Caldas Y. (2021). A gene co-association network regulating gut microbial communities in a Duroc pig population. Microbiome. 9(1):52. doi: 10.1186/s40168-020-00994-8.
- Ringseis R, & Eder K. (2022). Heat stress in pigs and broilers: Role of gut dysbiosis in the impairment of the gut-liver axis and restoration of these effects by probiotics, prebiotics and synbiotics. Journal of Animal Science and Biotechnology, 13(1), 126. https://doi.org/10.1186/s40104-022-00783-3 .
- Larzul C, Estellé J, et al. (2023). Driving gut microbiota enterotypes through host genetics. Research Square; DOI: 10.21203/rs.3.rs-3081627/v1.
- Saladrigas-García M, D’Angelo M, Ko HL, Traserra S, Nolis P, Ramayo-Caldas Y, Folch JM, Vergara P, Llonch P, Pérez JF, Martín-Orúe SM. (2021). Early socialization and environmental enrichment of lactating piglets affects the caecal microbiota and metabolomic response after weaning. Scientific Reports, 11, 6113. https://doi.org/10.1038/s41598-021-85460-7.
- Upadhaya SD, & Ki IH. (2022Jiang). Maintenance of gut microbiome stability for optimum intestinal health in pigs – a review. Journal of Animal Science and Biotechnology, 13(1), 140. https://doi.org/10.1186/s40104-022-00790-4.
- Vasquez R, Oh JK, Song JH, Kang DK. (2022). Gut microbiome-produced metabolites in pigs: a review on their biological functions and the influence of probiotics. J Anim Sci Technol. 64(4):671-695. doi: 10.5187/jast.2022.e58.
- Wang An, Cai Cj, Zeng Xf, Zhang Fr, Zhang Gl, Thacker Pa, Wang Jj, Qiao S.Y. (2013). Dietary supplementation with Lactobacillus fermentum I5007 improves the anti-oxidative activity of weanling piglets challenged with diquat. Journal of Applied Microbiology, 114(6), 1582–1591. https://doi.org/10.1111/jam.12188.
- Wickham H. (2016). ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York. ISBN 978-3-319-24277-4, https://ggplot2.tidyverse.org.
- Wu J, Chai T, Zhang H, Huang Y, Perry SW, Li Y, Duan J, Tan X, Hu X, Liu Y, Pu J, Wang H, Song J, Jin X, Ji P, Zheng P, Xie P. (2022). Changes in gut viral and bacterial species correlate with altered 1,2-diacylglyceride levels and structure in the prefrontal cortex in a depression-like non-human primate model. Translational Psychiatry, 12(1), Article 1. https://doi.org/10.1038/s41398-022-01836-x.
- Xia B, Wu W, Fang W, Wen X, Xie J, Zhang, H. (2021). Heat stress-induced mucosal barrier dysfunction is potentially associated with gut microbiota dysbiosis in pigs. Animal Nutrition, 8, 289–299. https://doi.org/10.1016/j.aninu.2021.05.012.
- Xiao L, Estellé J, Kiilerich P, Ramayo-Caldas Y, Xia Z, Feng Q, Liang S, Pedersen AØ, Kjeldsen NJ, Liu C, Maguin E, Doré J, Pons N, Le Chatelier E, Prifti E, Li J, Jia H, Liu X, Xu X, Ehrlich SD, Madsen L, Kristiansen K, Rogel-Gaillard C, Wang J. (2016). A reference gene catalogue of the pig gut microbiome. Nat Microbiol. doi: 10.1038/nmicrobiol.2016.161.
- Xiong Y, Cao S, Xiao H, Wu Q, Yi H, Jiang Z, Wang L. (2022). Alterations in intestinal microbiota composition coincide with impaired intestinal morphology and dysfunctional ileal immune response in growing-finishing pigs under constant chronic heat stress. Journal of Animal Science and Biotechnology, 13(1), 1. https://doi.org/10.1186/s40104-021-00651-6.
- Yang H, Huang X, Fang S, He M, Zhao Y, Wu Z, Yang M, Zhang Z, Chen C, Huang L. (2017). Unraveling the Fecal Microbiota and Metagenomic Functional Capacity Associated with Feed Efficiency in Pigs. Front Microbiol. 8:1555. doi: 10.3389/fmicb.2017.01555.
- Yang J, Wang C, Huang K, Zhang M, Wang J, Pan X. (2020). Compound Lactobacillus sp. Administration ameliorates stress and body growth through gut microbiota optimization on weaning piglets. Applied Microbiology and Biotechnology, 104(15), 6749–6765. https://doi.org/10.1007/s00253-020-10727-4.
- Yang Y, Deng Y, Cao L. (2016). Characterising the interspecific variations and convergence of gut microbiota in Anseriformes herbivores at wintering areas. Sci Rep. 6:32655. doi: 10.1038/srep32655.
Table 1. Resume of phenotype (above diagonal) and genetic (below diagonal) correlations between the enterosignatures and body weight at 60-days
|
ES_Prev |
ES_Trep |
ES_Lact |
ES_Clos |
ES_Stre |
BW-60d |
|
|
ES_Prev |
1 |
-0.712 |
-0.166 |
-0.096 |
-0.094 |
0.149 |
|
ES_Trep |
-0.861 |
1 |
-0.243 |
-0.081 |
-0.208 |
-0.119 |
|
ES_Lact |
-0.262 |
-0.166 |
1 |
-0.354 |
-0.253 |
0.005 |
|
ES_Clos |
0.153 |
-0.550 |
-0.238 |
1 |
-0.161 |
-0.086 |
|
ES_Stre |
0.589 |
-0.289 |
-0.625 |
0.179 |
1 |
0.037 |
|
BW-60d |
0.380 |
-0.540 |
0.400 |
0.584 |
-0.783 |
1 |
No competing interests reported.