https://orcid.org/0009-0001-0223-819X
Figures
Abstract
Horse owners and veterinarians report that from the age of 15, their horses can lose body condition and be more susceptible to diseases. Large intestinal microbiome changes may be involved. Indeed, microbiota is crucial for maintaining the condition and health of herbivores by converting fibres into nutrients. This study aimed to compare the faecal microbiome in horses aged from 6 to 30 years old (yo), living in the same environment and consuming the same diet, in order to assess whether the parameters changed linearly with age and whether there was a pivotal age category. Fifty horses were selected from the same environment and distributed across four age categories: 6–10 (n = 12), 11–15 (n = 11), 16–20 (n = 13), and 21–30 (n = 14) yo. All horses had no digestive problems, had teeth suitable for consuming their feed, and were up to date with their vaccination and deworming programmes. After three weeks of constant diet (ad libitum hay and 860 g of concentrate per day), one faecal sample per horse was collected on the same day. The bacterial communities’ richness and intra-sample diversity were negatively correlated with age. There was a new distribution of non-beneficial and beneficial taxa, particularly in the 21–30 yo category. Although the faecal concentration of short-chain fatty acids remained stable, the acetate proportion was negatively correlated with age while it was the opposite for the proportions of butyrate, valerate, and iso-valerate. Additionally, the faecal pH was negatively correlated with age. Differences were more pronounced when comparing the 6–10 yo and 21–30 yo categories. The values of the parameters studied became more dispersed from the 16–20 yo category onwards, which appeared as a transitional moment, as it did not differ significantly from the younger and older categories for most of these parameters. Our data suggest that the microbiome changes with age. By highlighting the pivotal age of 16–20, this gives the opportunity to intervene before individuals reach extremes that could lead to pathological conditions.
Citation: Baraille M, Buttet M, Grimm P, Milojevic V, Julliand S, Julliand V (2024) Changes of faecal bacterial communities and microbial fibrolytic activity in horses aged from 6 to 30 years old. PLoS ONE 19(6): e0303029. https://doi.org/10.1371/journal.pone.0303029
Editor: Franck Carbonero, Washington State University - Spokane, UNITED STATES
Received: January 10, 2024; Accepted: April 17, 2024; Published: June 3, 2024
Copyright: © 2024 Baraille et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files. All raw sequence read files are available from the NCBI database (accession number: PRJNA1090926).
Funding: The study was funded by the Sandgrueb Foundation, which focuses on animal welfare. The funder Sandgrueb-Stiftung (animal protection foundation) contributed to the study design, data analysis and preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Within the last decade, the proportion of horses aged over 20 years increased from 7% to 12% in France [1,2], from 8% to 17% in Switzerland [3,4], and from 8% to 11% in the USA [5]. In the same time, horse owners and veterinarians report that from the age of 15, their horses can lose body condition and be more susceptible to infectious or chronic diseases [6–8]. Due to a lack of related data, the recommendations for preventive healthcare measures in elderly horses are often extrapolated from those used in adult horses [6]. To provide relevant recommendations, particularly nutritional ones, it is essential to understand certain physiological mechanisms linked to the ageing process in horses [9].
The large intestinal (LI) microbiome plays a key role in digestion, contributing to maintaining the condition and health of the host. Indeed, only microbes are able to degrade fibre which mainly compose the herbivore diet [10]. This degradation enables the release of microbial metabolites (i.e. short chain fatty acids (SCFAs) and lactate). The latter are sources of energy for the horse, and also support various mechanisms that promote good immunity [11,12].
Research into changes of the human LI microbiome that occur with age has been ongoing for the last two decades and is mainly approached by targeting the faecal bacterial communities. The faecal bacterial communities of elderly people appear less diverse and rearranged compared to those of healthy adult people [13–17]. Most studies agree that the abundance of pathobionts increases (i.e. Desulfovibrio, Enterobacteriaceae, Eggerthella) whereas some commensal taxa carrying essential functions disappear (i.e. Roseburia, Faecalibacterium, Bifidobacterium) [18–27]. These changes linked with ageing have also been implicated with the development of frailty [28–30] and diseases such as Alzheimer [31] and Parkinson [32,33]. From about 20 studies included in different reviews [13,14,16,17], only 5 compared microbial activity parameters between elderly and younger people [34–38]. Among these, solely 3 report differences between age categories. In the faeces of elderly people, the percentage of dry matter (DM), the concentrations of tryptophan, indole, iso-valerate, iso-butyrate, iso-capronate, L-lactate and pentadecylic acid are lower while the concentrations of propionate, palmitic and arachidic acids are higher in comparison with younger people [34,36,38].
If the composition of the LI bacterial communities erodes over the lifespan in horses as it does in humans, the LI microbial activity, particularly the fibrolytic activity, may be altered, leading to clinical issues reported by owners [6]. However, in horses, changes in the faecal bacterial communities with age are still little documented. Currently, 2 out of 5 available scientific publications related to this topic mention a decrease in bacterial communities diversity [39,40], whereas 3 highlight a change in bacterial communities structure [40–42] with age. Among these changes, there appears to be a higher abundance of non-beneficial taxa (i.e. Proteobacteria, Amplicon Sequence Variant (ASV) assigned to the Eggerthellaceae family) and a lower abundance of potentially beneficial taxa (i.e. genus Fibrobacter, ASV assigned to the Ruminococcaceae family) in elderly horses compared to younger ones. [40,41]. Finally, among these 5 studies, 2 have also investigated microbial activity parameters [41,42]. One of these reports a higher faecal pH in the elderly compared to adult horses [41], while the second observes no difference between the same age categories [42]. In both studies, no difference is observed in faecal SCFAs concentrations between the investigated age categories. These scattered initial results make it difficult to draw conclusions on the involvement of the LI microbiome in clinical signs that appear during the ageing process in horses.
As the age at which a horse is deemed old remains unclear, we decided to follow a cohort of horses aged from 6 to 30 years old (yo), living in the same environment and consuming the same diet. Using a linear approach, our first aim was to assess the change of the faecal bacterial communities with age, while evaluating parameters of microbial fibrolytic activity. Moreover, in human research, classifying people according to their age is frequently applied, especially since the category of centenarians is studied to identify markers of extreme longevity. Therefore, based on age categorisation, our second aim was to evaluate whether there was a pivotal age in horses from which some of the parameters studied differed markedly.
Materials and methods
The protocol was approved by the local animal experimentation ethics committee of the University of Burgundy (Comité d’éthique de l’expérimentation animale—Grand campus Dijon N°105) and met all requirements for ethical care and treatment of animals.
Experimental design, animals’ characteristics, and management
This study was conducted in July 2020 at a horse sanctuary located in France. Fifty horses aged from 6 to 30 yo, either mares or geldings, of different breeds were included in the study. These animals did not suffer from any digestive issue for at least 3 months prior to the study and were up to date with their deworming and vaccination programmes. Potential metabolic disorders were not tested. At least 5 weeks prior to sampling, horses were submitted to a dental examination and received dental treatment, if needed. Following this, the veterinary dentist scored the wear levels of incisors and molars on a 6-point scale ranging from “no wear” to “very significant wear”, and evaluated if each horse could have gripping or chewing problems. Horses’ weight was measured and body condition score (BCS) according to the Henneke scale [43] was determined by a single competent examiner to ensure standardised and objective assessment. Characteristics of the horse population are available in the supplementary data (S1 Table).
All horses had been housed in the same sanctuary for at least one year prior to the study. For the study, they were housed in groups, in dry lot paddocks. All horses were fed hay ad libitum and received an individual ration of 860 g of concentrate (Landmüsli–Typ Senior, Futtermühle Tock GmbH, Germany) once a day, for 3 weeks to habituate and stabilise the LI microbiome. The biochemical composition of hay and concentrate (Equine Complete; DairyOne, Ithaca, USA) is available in the supplementary data (S2 Table). During this period, they had free access to water and mineral salt block and their general condition was followed daily by the sanctuary’s caretakers.
Sample collection
After 3 weeks of constant diet, freshly voided faecal samples were obtained from each horse once on the same day. For each sample, the central portion was aliquoted in sterile microtubes and immediately frozen at -20°C and then -80°C to perform bacterial 16S ribosomal RNA gene (rRNA) sequencing analysis.
For each faecal sample, 10 g of faeces were weighed and dried at 70°C for 72 hours to determine the faecal DM. Faecal particles size was determined by sieving 150 g of faeces with water through 3 decreasing size sieves (2; 0.5 and 0.15 mm). The proportion of particles for each category was determined and related to DM. Faecal samples were filtered through a 100 μm nylon screen. The faecal pH was measured in the filtrate with a Cyberscan 500 pH-meter (Eutech Instruments, Strasbourg, France). After that, the filtrate was sampled in microtubes with or without a preservative solution (4.25% H3PO4 and 1.0% HgCl2) and frozen at -20°C to determine SCFAs and lactate concentrations respectively.
Bacterial 16S rRNA gene sequencing analysis
Faecal total DNA was extracted as described by Yu and Morrison [44]. Briefly, to lyse the cells, 0.25 g of faecal sample was bead-beaten with a mixture of sodium dodecyl sulfate (SDS), NaCl, and EDTA. Impurities and SDS were removed by precipitating them with ammonium acetate. Nucleic acids were then recovered by precipitating them with isopropanol. Genomic DNA was purified by sequentially digesting with RNase and Proteinase K, followed by the use of QIAamp columns. After spectrophotometric assessment of the quantity and purity of the DNA obtained (Eppendorf spectrophotometer, Hamburg, Germany), the V3-V4 hypervariable region of the 16S rRNA gene was amplified and sequenced as described by Grimm et al. [45]. Briefly, 2 consecutive polymerase chain reactions (PCR) were performed (PCR 1 for V3-V4 region amplification and PCR 2 to ligate Illumina adapters and index for sample identification). The PCR mix contained DNA, buffer, dNTPs, Taq polymerase and primers (PCR 1: F343 and R784; PCR 2: forward primer AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGAC and reverse primer CAAGCAGAAGACGGCATACGAGAT-Index-GTGACTGGAGT TCAGACGTGT). The PCR programme was as follows: 1 minute at 94°C, 30 (PCR 1) or 12 (PCR 2) x [94°C for 1 minute, 65°C for 1 minute and 72°C for 1 minute] and 10 minutes at 72°C. The correct V3-V4 region amplification after PCR 1 was verified by electrophoresis on a 2% agarose gel. The PCR products obtained were sequenced using an Illumina MiSeq run of 250-paired ends, according to the manufacturer’s instructions (Illumina Inc., San Diego, CA, United States) at Genotoul Bioinformatics Platform (Toulouse, France). 16S rDNA sequences were submitted to the NCBI Sequence Read Archive and can be found with the following accession number: PRJNA1090926.
FROGS (Find Rapidly OTU with Galaxy Solution) metabarcoding pipeline on the Galaxy server was used to perform bioinformatics analysis [46]. The first step was to assemble the raw data (R1 and R2 reads) and sort them to remove aberrant sequences, i.e. sequences without primers or out of range (<380 or >490 base pairs). Clusters were formed from the remaining sequences using the SWARM aggregation technique (distance = 1). Chimeric sequences were eliminated, and a filter was applied to retain only clusters present in at least 2 samples and with an abundance greater than or equal to 0.005% to obtain the ASV abundance table. Each ASV was aligned to the silva138.1 16S database using BLAST and was affiliated to the highest taxonomic possible rank. Finally, only affiliations with a percentage of identity and coverage greater than 90% and 99% respectively were retained for data analysis.
Firstly, the relative abundance of each bacterial phylum and genus was calculated by relating their abundance to the number of total sequences in the sample concerned. This allowed comparisons of abundance between all the samples. Secondly, the sequences of all samples were rarefied to the smallest number of sequences obtained. Based on this new dataset, the richness (i.e. number of observed ASV and Chao 1 index), intra-sample diversity (i.e. Inverse Simpson and Shannon indexes) and inter-sample diversity (i.e. Bray-Curtis distance) were calculated.
Determination of short chain fatty acids and lactate concentrations
SCFAs concentrations were determined as described by Jouany [47]. Briefly, filtered faecal samples were injected, under nitrogen, onto a 30 m x 0.25 mm diameter x 0.25 μm capillary column (Elite-FFAP column, PerkinElmer, Courtaboeuf, France) of gas-liquid chromatography coupled to a flame ionisation detector (Clarus, PerkinElmer, Courtaboeuf, France). The internal standard added to all filtered faecal samples was 4-methyl valeric acid (277827-25G, Sigma-aldrich, USA). A standard solution was used to determine the concentration of acetate (C2), propionate (C3), butyrate (C4), iso-butyrate (iC4), valerate (C5), and iso-valerate (iC5) in each filtered faecal sample. The addition of each SCFA gave the total SCFAs concentration. Each SCFA was expressed as a percentage of the total SCFAs.
D- and L-lactate concentrations were measured using an enzymatic colorimetric assay kit (Megazyme International Ltd, Wicklow, Ireland), according to the manufacturer’s instructions and the modifications described by Grimm et al [48]. Optical density was measured at 340 nm (MRX Revelation Microplate Reader, Dynatech Laboratories, Guyancourt, France).
Statistical analysis
The homoscedasticity of the quantitative variables as a function of age was tested using a Fligner-Killeen test and linear regressions were plotted with a 95% confidence interval using R software to see if there was a linear change of some parameters with age. Pearson correlations between all the parameters studied and age were obtained using the PROC CORR procedure in SAS software.
Furthermore, to determine whether the studied parameters evolved differently from a pivotal age, four categories were compared: 6 to 10 yo (n = 12), 11 to 15 yo (n = 11), 16 to 20 yo (n = 13), 21 to 30 yo (n = 14). Gripping and chewing problems and wear levels of incisors and molars were compared between categories using a Fisher exact test on R software. For quantitative variables, the means of each category were compared using the PROC MIXED procedure (LSMEANS/PDIFF option with Tukey-Kramer adjustment) in the SAS software. Finally, principal coordinate analysis (PCoA) based on Bray-Curtis distances was performed on R software using the Phyloseq package to graphically evaluate the clustering of individual faecal bacterial communities by category. The impact of categories on the structure of the bacterial communities was evaluated by a permutational multivariate ANOVA (PERMANOVA). Galaxy Toulouse platform was used to perform a Linear discriminant analysis effect size (LEfSe) [49], which combined the Kruskal-Wallis sum-rank test to identify taxa with significant differences in abundance between categories (using all-against-all comparisons) and a linear discriminant analysis (LDA) to estimate the effect size of each differentially abundant taxon (with a threshold set at 3 log LDA scores).
For all statistical tests, changes and differences were considered significant at P≤0.05.
The software versions used are R 4.3.0 (The R Foundation for Statistical Computing) and SAS 9.3 (Statistical Analysis System Institute Inc, Cary, North Carolina).
Results
All horses remained in good condition and did not develop any health problems during the 3 weeks prior to faecal samples collection.
Dental check-up of horses and faecal particle sizes
There were no differences in prehension and mastication between age categories. However, the wear level of incisors and molars differed between the categories. Incisors were more worn in horses from 16 to 20 yo and 21 to 30 yo than in horses from 6 to 10 yo and 11 to 15 yo. Horses from 16 to 20 yo and 21 to 30 yo had more worn molars than horses from 6 to 10 yo (Table 1, S1 Fig).
The proportions of different faecal particle sizes were not correlated with age, except for those between 2 mm and 0.5 mm which were negatively correlated (r = -0.29, P = 0.038). No differences in proportions were identified between categories for other particle sizes (S3 Table).
Faecal bacterial communities
Of the 50 faecal samples, 5 were excluded from the bioinformatic analysis due to a number of sequences less than 2000 after filtering steps. In the remaining 45 samples, 719,225 sequences from the 16S rRNA V3-V4 region were recovered, with an average of 15,983 ± 8,467 sequences per sample. Following various filtering processes, 2,558 ASVs were identified. Rarefaction curves (S2 Fig) showed that, at ASV level, some samples had not plateaued, suggesting that a full sampling of these environments had not been achieved.
The richness (number of ASVs and Chao 1 index) and the intra-sample diversity (Inverse Simpson and Shannon indexes) were negatively correlated with age (Fig 1). Among the categories, horses from 21 to 30 yo had a lower number of ASVs compared to those from 6 to 10 yo and 11 to 15 yo (Fig 2A). No difference between categories was observed for the Chao1 index (Fig 2B). The Inverse Simpson index was lower in horses from 21 to 30 yo in comparison to those from 11 to 15 yo (Fig 2C). The Shannon index was lower in horses from 21 to 30 yo in comparison to those from 6 to 10 yo and 11 to 15 yo (Fig 2D).
Linear regressions illustrating the correlations between faecal bacterial richness (A: Number of ASVs; B: Chao 1 index) and intra-sample diversity (C: Inverse Simpson index, D: Shannon index) and age. Shaded areas represent 95% confidence intervals.
Faecal bacterial richness (A: Number of ASV; B: Chao 1 index) and intra-simple diversity (C: Inverse Simpson index, D: Shannon index) in horses according to the age category. The mean (cross), median (solid line) and interquartile ranges are indicated in each boxplot. The P values reported correspond to the comparison of each parameter between age categories. Asterisks indicate significant differences between two age categories (*: P < 0.05; **: P < 0.01).
PERMANOVA analysis highlighted a difference in inter-sample diversity between categories (P<0.001) and the PCoA shows the extent of these variations (Fig 3).
Ellipses represent 95% confidence intervals.
The 2,558 ASVs were assigned to 8 phyla, 14 classes, 28 orders, 48 families and 97 genera. Some ASVs have been multi-affiliated. The mean relative abundance of unknown ASVs was 0.2 ± 0.3% at the order level, 1.2 ± 0.9% at the family level and 30.6 ± 2.7% at the genus level. After analysing the phyla, we targeted the genera as they best approximate the functional potential of the bacteria. Correlation with age and the difference between categories were only tested on taxa with an average relative abundance greater than 0.1%.
Of the 8 phyla, 7 had relative abundances greater than 0.1%. Of these, 2 were negatively and 2 positively correlated with age (Table 2). Among the latter, only Bacteroidota (P = 0.035) was different between categories, with a lower relative abundance in horses from 21 to 30 yo than in those from 11 to 15 yo (S4 Table).
Of 97 genera, 50 had relative abundances greater than 0.1%. Of these, 11 were negatively and 7 positively correlated with age (Table 3). Among the latter, some differed between categories (S4 Table). Agathobacter (P = 0.025) was less abundant in horses from 21 to 30 yo than in those from 11 to 15 yo. Butyvibrio (P = 0.013) and Prevotella (P = 0.015) were less abundant in horses from 21 to 30 yo than in those from 6 to 10 yo. Desulfovibrio (P = 0.018), Lachnospiraceae ND3007 group (P = 0.003), Prevotellaceae Ga6A1 group (P = 0.003) and Roseburia (P = 0.026) were less abundant in horses from 16 to 20 yo and 21 to 30 yo than in those from 6 to 10 yo. On the other hand, Christensenellaceae R-7 group (P = 0.004) was more abundant in horses from 21 to 30 yo than in those from 6 to 10 yo and 11 to 15 yo. Rikenellaceae RC9 gut group (P = 0.009) was more abundant in horses from 21 to 30 yo than in those from 6 to 10 yo. Lachnospiraceae UCG 009 (P = 0.020) was more abundant in horses from 16 to 20 yo than in those from 6 to 10 yo. Finally, 2 genera that were not correlated with age, nevertheless showed differences between categories: Anaerovorax (P = 0.004) was more abundant in horses from 21 to 30 yo than in those from 6 to 10 yo and 16 to 20 yo and Oribacterium (P = 0.045) was less abundant in horses from 21 to 30 yo than in those from 11 to 15 yo.
In addition, we used LEfSe analysis to identify bacterial taxa that were strongly associated with each category compared with the others. The differences between categories are illustrated in a circular cladogram (Fig 4) and the LDA scores are grouped in the supplementary data (S5 Table). In horses from 6 to 10 yo, the class Lachnospirales and the families Bacteroidales RF16 group and Lachnospiraceae were overrepresented. In horses from 11 to 15 yo, the faecal microbiome was enriched in Bacteroidales class, Bacteroidales BS11 gut group and p-251-o5 families and Prevotellaceae Ga6A1 group, Lachnospiraceae UCG-003, Mailhella and Oribacterium genera. In horses from 16 to 20 yo, only the Muribaculaceae family differentiated this class from the others. Finally, the 21 to 30 yo category was distinguished from the others by an enrichment of the Proteobacteria phylum, the Gammaproteobacteria order, the Ruminococcaceae family and the Enterorhabdus genus.
From the centre outwards, the concentric circles represent the taxonomic level of phylum, class, order, family, and genus respectively.
Faecal microbial fibrolytic activity
Faecal DM was positively (Fig 5A), and pH negatively (Fig 5B) correlated with age. There was no correlation between the total SCFAs (Fig 5C) and lactate (Fig 5D) concentrations and age.
Linear regressions illustrating the correlations between faecal microbial activity parameters (A: DM; B: pH; C: Total SCFAs; D: Total lactate) and age. Shaded areas represent 95% confidence intervals.
When looking at the detail of the fermentation products, the faecal proportion of C2 was negatively and the faecal proportions of C4, iC5 and C5 were positively correlated with age (Table 4). There was no correlation between the C3, iC4, D- and L-lactate proportions and age (Table 4).
In terms of categories, faecal DM was greater (Fig 6A) and pH was lower (Fig 6B) for horses from 21 to 30 yo in comparison to those from 6 to 10 yo. The total SCFAs and lactate concentrations did not differ between categories (Fig 6C and 6D).
Faecal microbial activity parameters (A: DM; B: pH; C: Total SCFAs; D: Total lactate) in horses according to the age category. The mean (cross), median (solid line) and interquartile ranges are indicated in each boxplot. The P values reported correspond to the comparison of each parameter between age categories. Asterisks indicate significant differences between two categories (**: P < 0.01).
When looking at the detail of the fermentation products, the proportion of C4 was higher in the faeces of horses from 21 to 30 yo compared to all other categories (Table 4). The proportion of C5 was higher in the faecal sample of horses from 21 to 30 yo compared to those from 6 to 10 yo and 11 to 15 yo (Table 4). Additionally, the faecal proportion of iC5 was higher for horses from 21 to 30 yo than those from 6 to 10 yo (Table 4). There were no differences between the categories for C2, C3, IC4, D- and L-lactate proportions (Table 4).
Discussion
This study compared the faecal microbiome (i.e. faecal bacterial communities and microbial fibrolytic activity) of horses aged from 6 to 30 yo, to determine whether the parameters change linearly with age and whether there was a pivotal age category beyond which differences between age categories become more pronounced.
In humans, the country [50] and residence place [28] in which elderly people live, are described as a co-factor contributing to the observed differences in faecal bacterial communities. In horses, the location explains 6.4% of the variation in faecal bacterial communities [40], but not specifically in elderly individuals. In addition, reviews summarise and highlight the major effect of the diet on the large intestinal microbiota [51,52]. Thus, to limit the impact of the location and diet and highlight only the effect of age on the faecal bacterial communities, we selected horses kept at one location for at least 1 year and fed the same diet. Our results demonstrated that the faecal bacterial richness and intra-sample diversity were negatively correlated with age, confirming recent observations [40]. Additionally, intra-sample diversity indexes were lower in the 21 to 30 yo category compared to the younger categories. These results are in accordance with data from a previous study comparing a category of horses aged between 19 and 29 with a category of horses aged between 9 and 12 [39]. In humans, it has been suggested that a loss of diversity favours the expansion of the abundance of non-beneficial taxa [17], and both are strongly associated with an unhealthy ageing. It has been linked with increased frailty [29,30,53,54] and cognitive decline [55], as well as Parkinson’s [33] and Alzheimer’s [31] diseases. In horses, there is no grid to define frailty like those used in humans which are based on parameters such as appearance of illnesses (physical and mental) and painful conditions, taking of medication, reduction in grip strength and speed of movement, history of hospitalization, loss of body condition and autonomy [29,54]. Thus, when examining the effect of age in horses, it is still not possible to attribute the observed microbiome changes to frailty. Screening for metabolic diseases such as insulin dysregulation or Cushing’s syndrome could provide interesting information. Insulin dysregulation has a greater impact than age of ponies on the faecal bacterial communities [41]. In contrast, in ponies negative for Cushing’s syndrome, when two groups aged 21.5 ± 2.9 years and aged 9.8 ± 3.2 years are compared, there is no difference in intra-sample diversity and faecal microbial activity [42]. There is no existing data on the impact of Cushing’s syndrome on the bacterial communities in horse faeces. In humans, such metabolic disorder greatly reduces the intra-sample diversity and modifies the structure. Simultaneously, this is accompanied by an increase in the abundance of pathobionts (i.e. Proteobacteria including Escherichia-Shigella) and parallel reduction of the abundance of beneficial taxa (i.e. Blautia and Agathobacter), and a lower faecal propionate concentration [56]. The numerous correlations between the parameters studied in our study and age suggested that the changes observed in the microbiome were at least partly linked to ageing, although it cannot be guaranteed that they were linked to healthy ageing. The human healthy ageing microbiome is still in the early stages of being described [17,57]. In healthy elderly people, the decrease in the abundance of important taxa (i.e. Prevotella, Faecalibacterium, Coprococcus) and the increase in the abundance of pathobionts (i.e. Eggerthella, Streptococcus, Enterobacteriaceae) is compensated by the increase in the abundance of other beneficial taxa (i.e. Akkermansia, Christensenellaceae, Butyrivibrio). These taxa disappear when the elderly individual shifts from a healthy state to a state of physiological decline [57]. In our study, a rearrangement also appeared to be taking place, particularly in the 21 to 30 yo category. Simultaneously with the expansion of taxa known to carry pathobiont species (i.e. Clostridium sensu stricto 1 and, Proteobacteria including Gammaproteobacteria), there was a rearrangement in the abundance of certain beneficial taxa: Agathobacter, Blautia, Butyrivibrio, Lachnospiraceae ND3007, Prevotella, Prevotellaceae Ga6A11 group, and Roseburia make way for Christensenellaceae R-7 group, Rikenellaceae RC9 gut group and Ruminococcaceae. This rearrangement could be responsible for changes in essential functions carried out by the microbiome, in particular the degradation of dietary fibre, which provides energy and supports host health. To assess this, we measured fibrolytic activity in faeces.
In horses, poor dentition, i.e. an incorrect occlusal angle in the premolar area, reduces apparent fibre digestibility [58] and dental correction improves their digestibility [59]. To avoid bias in the present study, the horses’ dentition was checked and treated. Inevitable differences in the wear of incisors and molars between age categories were highlighted, but the ability to grip and chew was similar between age categories as judged by the veterinary dentist. Wear issues did not appear to affect fibre digestibility between age categories. Faecal concentrations of total SCFAs and lactate remained stable with age, confirming previous observations [41,42]. However, acidification of the faecal environment was suggested by the negative correlation between pH and age. Acidification in culture-based experiments is known to have a major impact on some beneficial fibrolytic species and fibre degradation [60]. In our study, we found that the proportion of faecal acetate was negatively correlated with age while it was the opposite for the proportions of butyrate, valerate and iso-valerate. This suggested that, despite the acidification of the faecal environment, the function of fibre degradation did not decline with age, but rather reorganised, probably due to the rearrangement of bacterial communities that were no longer able to degrade the same type of fibre. It would have been valuable to measure which type of fibre was the most degraded as a function of age, based on enumeration of fibrolytic microorganisms or measurement of fibre digestibility, in order to draw conclusions on this point.
In the 21 to 30 yo category, the pH was lower than in the younger categories. In this category, there was a higher proportion of butyrate, valerate and iso-valerate without a significant decrease in other acids. Butyrate is particularly important as it is essential for maintaining the integrity of the intestinal mucosa, by nourishing the colonocytes that make it up [61]. In our study, this raises the question of a higher production of butyrate and/or an accumulation of this SCFA due to a lower absorption by the intestinal mucosa. At this point, our results are contradictory and do not allow us to conclude. Indeed, on the one hand, we measured a negative correlation with age of some butyrate-producing taxa such as Roseburia [62] and Agathobacter [63]. Such a decline is consistent with the human literature [64]. On the other hand, the faecal dry matter was positively correlated with age, which could reflect more water absorption in the large intestine, and consequently more SCFAs absorption, since water absorption is always accompanied by SCFAs absorption in the large intestine of horses [65]. However, faecal dry matter modification could also be related to a longer transit time [65] or to a lower water consumption. To confirm the changes in SCFAs absorption with age, it might be informative to measure blood SCFAs. Measurement of the transit time and the daily water consumption could also be valuable data.
In humans, the composition of the microbiota changes gradually throughout life, with no specific chronological threshold at which changes occur suddenly [66]. In horses, our data suggest the same pattern as we observed that many parameters of the faecal bacterial communities and microbial activity were correlated with age. However, even if the change appeared to be gradual, the inter-individual variability was increasingly observed, with visually greater variations from the 16 to 20 yo for several parameters (i.e. PCoA, pH, dry matter, total SCFAs). This observation is consistent with findings in humans, where a greater inter-individual variation in faecal bacterial composition in elderly people is observed compared with younger ones [67]. This suggests that changes do not follow the same slope for all individuals, which can explain why some individuals reach a very advanced chronological age in good health, while others lose health earlier. In addition, the 16–20 yo category did not differ significantly from the younger ones (6 to 10yo and 11 to 15 yo) and the older one (21 to 30 yo) regarding most of the microbiome parameters we monitored. It is therefore possible that this 16–20 yo category represented the point at which individuals made the transition towards a pathological state or extreme longevity for the end of their lives. This could explain why some horse owners report more clinical problems after the age of 15 [6,7].
Conclusion
In horses aged between 6 and 30 years, which have lived in the same environment for a long time and have teeth good enough to consume their feed, we showed that the bacterial communities lose richness and diversity and reorganise themselves with age. Fibrolytic activity in faeces did not appear to be reduced, but also reorganised with age. However, the lack of data on the state of health (i.e. frailty, metabolic disorders) of horses did not allow us to claim that these changes were solely age-related. Defining the parameters of the microbiome of elderly horses that have aged in good health would make it possible to discern the change in the faecal microbiome that predisposes to a pathological condition. Finding out whether there was a pivotal age at which it was relevant to study the horse microbiome was a major challenge. To our knowledge, there is no published data on this topic. From 16 to 20 yo, the values of the microbiome parameters became increasingly scattered, suggesting that from this point on, some individuals reached extremes. This new understanding may allow for intervention before this pivotal age, for example through an adapted diet or by administering biotics (including prebiotics, probiotics, and postbiotics), to influence the microbiome change towards healthy ageing and avoid pathological conditions reported by horse owners.
Supporting information
S1 Fig. Gripping and chewing ability and wear level of incisors and molars by age category.
The P values reported correspond to the comparison of each parameter between age categories. Asterisks indicate significant differences between two age categories (*: P < 0.05; **: P < 0.01, ***: P < 0.001).
https://doi.org/10.1371/journal.pone.0303029.s001
(TIF)
S2 Fig. Rarefaction curves at the amplicon sequence variant level showing the depth of sequencing of the faecal bacterial communities (obtained after the last filtering step before data normalisation).
https://doi.org/10.1371/journal.pone.0303029.s002
(TIF)
S1 Table. Individual characteristics of the horses involved in the study.
https://doi.org/10.1371/journal.pone.0303029.s003
(XLSX)
S2 Table. Biochemical composition of hay and concentrate.
https://doi.org/10.1371/journal.pone.0303029.s004
(XLSX)
S3 Table. Correlations (Pearson correlation coefficients, r) between age and proportions of particle sizes in faeces, and LSMean of each proportion according to the age category.
P values in bold indicate that there was a significant correlation with age or a significant difference in LSMeans between age categories. LSMeans: Least square means. SEM: Standard error of the mean.
https://doi.org/10.1371/journal.pone.0303029.s005
(XLSX)
S4 Table. LSMean values of each phylum and genus relative abundance that were significantly correlated with age (Table 3) and/or differed between age categories.
P values in bold indicate that there was a significant difference in LSMeans between age categories. For each row, LSmeans with different superscripts differ (P < 0.05). LSMeans: Least square means. SEM: Standard error of the mean.
https://doi.org/10.1371/journal.pone.0303029.s006
(XLSX)
S5 Table. Linear discriminant analysis scores of the bacterial taxa enriched in the faecal microbiome of different age category.
P values in bold indicate the dominance of the taxa in the age category concerned. LDA: Linear discriminant analysis.
https://doi.org/10.1371/journal.pone.0303029.s007
(XLSX)
Acknowledgments
We thank the technical staff for the animal care and laboratory analyses (Gut Aiderbichl team in Trévol, Lab To Field team in Créancey, PMB team from the PAM UMR of the Institut Agro Dijon in Dijon).
References
- 1. Ifce-OESC. Combien d’équidés en France? 2019. Available from: https://www.ifce.fr/wp-content/uploads/2019/07/IFCE_OESC_Note_thematique-Effectifs-equides_juillet2019_V2.pdf.
- 2. Doliguez P, Le Masne L, Delerue M. Gestion pratique du vieux cheval. 2023. Available from: https://equipedia.ifce.fr/sante-et-bien-etre-animal/soin-prevention-et-medication/prevention/gestion-pratique-du-vieux-cheval.
- 3. Agroscope. La filière équine Suisse: les chiffres clefs. Bilan 2016. 2017. Available from: https://www.cofichev.ch/Htdocs/Files/v/6062.pdf/Publications-autres/HNS/Bericht_2016_fr_DEF-20171205.pdf.
- 4. Agroscope. La filière équine Suisse: les chiffres clefs. Aperçu 2019. 2019. Available from: https://www.cofichev.ch/Htdocs/Files/v/6113.pdf/Publications-autres/Kennzahlen-Pferdebranche-F.pdf.
- 5. APHIS. Age-related Trends in Demographics of Equids in the United States. 2018. Available from: https://www.aphis.usda.gov/animal_health/nahms/equine/downloads/equine15/Equine15_is_TrendsAge.pdf.
- 6. Ireland J, Clegg P, McGowan C, McKane S, Pinchbeck G. A cross-sectional study of geriatric horses in the United Kingdom. Part 2: Health care and disease. Equine Veterinary Journal. 2011;43(1):37–44. pmid:21143632
- 7. Gehlen H, Lilge S, Merle R, Steinborn S. Einfluss des Pferdealters auf die Wahl der Haltung, des Managements und der Dienstleistungsansprüche von Pferdebesitzern. Berliner und Münchener Tierärztliche Wochenschrift. 2021. Available from: https://www.vetline.de/einfluss-des-pferdealters-auf-die-wahl-der-haltung-des-managements-und-der.
- 8. Ireland J, Clegg P, McGowan C, McKane S, Chandler K, Pinchbeck G. Disease prevalence in geriatric horses in the United Kingdom: Veterinary clinical assessment of 200 cases. Equine Veterinary Journal. 2012;44(1):101–6. pmid:21668494
- 9. Brosnahan M, Paradis M. Demographic and clinical characteristics of geriatric horses: 467 cases (1989–1999). Journal of the American Veterinary Medical Association. 2003;223(1):93–8. pmid:12839071
- 10. Julliand V, Grimm P. The microbiome of the horse hindgut: History and current knowledge. Journal of Animal Science. 2016;94(6):2262–74.
- 11. Den Besten G, Van Eunen K, Groen A, Venema K, Reijngoud D, Bakker B. The role of short-chain fatty acids in the interplay between diet, gut microbiota, and host energy metabolism. Journal of Lipid Research. 2013;54(9):2325–40. pmid:23821742
- 12. Yao Y, Cai X, Fei W, Ye Y, Zhao M, Zheng C. The role of short-chain fatty acids in immunity, inflammation and metabolism. Critical Reviews in Food Science and Nutrition. 2022;62(1):1–12. pmid:33261516
- 13. Cheng J, Palva A, De Vos W, Satokari R. Contribution of the Intestinal Microbiota to Human Health: From Birth to 100 Years of Age. Current Topics in Microbiology and Immunology. 2013;358:323–346. pmid:22094893
- 14. Biagi E, Candela M, Turroni S, Garagnani P, Franceschi C, Brigidi P. Ageing and gut microbes: Perspectives for health maintenance and longevity. Pharmacological Research. 2013;69(1):11–20. pmid:23079287
- 15. Nagpal R, Mainali R, Ahmadi S, Wang S, Singh R, Kavanagh K, et al. Gut microbiome and aging: Physiological and mechanistic insights. Nutrition and Healthy Aging. 2018;4:267–285. pmid:29951588
- 16. Badal V, Vaccariello E, Murray E, Yu K, Knight R, Jeste D, et al. The Gut Microbiome, Aging, and Longevity: A Systematic Review. Nutrients. 2020;12(12):3759. pmid:33297486
- 17. Kim M, Benayoun B. The microbiome: An emerging key player in aging and longevity. Translational Medicine of Aging. 2020;4:103–16. pmid:32832742
- 18. Park S, Kim K, Ahn Y, Jeong J, Huh C, Kim D. Comparative analysis of gut microbiota in elderly people of urbanized towns and longevity villages. BMC Microbiol. 2015;15(1):49. pmid:25887483
- 19. Biagi E, Franceschi C, Rampelli S, Severgnini M, Ostan R, Turroni S, et al. Gut Microbiota and Extreme Longevity. Current Biology. 2016;26(11):1480–5. pmid:27185560
- 20. Odamaki T, Kato K, Sugahara H, Hashikura N, Takahashi S, zhong Xiao J, et al. Age-related changes in gut microbiota composition from newborn to centenarian: a cross-sectional study. BMC Microbiol. 2016;16(1):90. pmid:27220822
- 21. Iwauchi M, Horigome A, Ishikawa K, Mikuni A, Nakano M, Xiao J, et al. Relationship between oral and gut microbiota in elderly people. Immunity Inflam & Disease. 2019;7(3):229–36. pmid:31305026
- 22. Kim B, Choi C, Shin H, Jin S, Bae J, Han M, et al. Comparison of the Gut Microbiota of Centenarians in Longevity Villages of South Korea with Those of Other Age Groups. Journal of Microbiology and Biotechnology. 2019;29(3):429–40. pmid:30661321
- 23. Tuikhar N, Keisam S, Labala R, Imrat , Ramakrishnan , Arunkumar M, et al. Comparative analysis of the gut microbiota in centenarians and young adults shows a common signature across genotypically non-related populations. Mechanisms of Ageing and Development. 2019;179:23–35. pmid:30738080
- 24. Wang N, Li R, Lin H, Fu C, Wang X, Zhang Y, et al. Enriched taxa were found among the gut microbiota of centenarians in East China. PLoS ONE. 2019;14(10):e0222763. pmid:31639130
- 25. Wu L, Zeng T, Zinellu A, Rubino S, Kelvin D, Carru C. A Cross-Sectional Study of Compositional and Functional Profiles of Gut Microbiota in Sardinian Centenarians. mSystems. 2019;4(4):e00325–19. pmid:31289141
- 26. Xu C, Zhu H, Qiu P. Aging progression of human gut microbiota. BMC Genomics. 2019;20(1):798. pmid:31672122
- 27. La-ongkham O, Nakphaichit M, Nakayama J, Keawsompong S, Nitisinprasert S. Age-related changes in the gut microbiota and the core gut microbiome of healthy Thai humans. 3 Biotech. 2020;10(6):276. pmid:32537376
- 28. Claesson M, Jeffery I, Conde S, Power S, O’Connor E, Cusack S, et al. Gut microbiota composition correlates with diet and health in the elderly. Nature. 2012;488(7410):178–84. pmid:22797518
- 29. Jackson M, Jeffery I, Beaumont M, Bell J, Clark A, Ley R, et al. Signatures of early frailty in the gut microbiota. Genome Med. 2016;8(1):8. pmid:26822992
- 30. Maffei V, Kim S, Blanchard E, Luo M, Jazwinski S, Taylor C, et al. Biological Aging and the Human Gut Microbiota. The Journals of Gerontology: Series A. 2017;72(11):1474–82. pmid:28444190
- 31. Haran J, Bhattarai S, Foley S, Dutta P, Ward D, Bucci V, et al. Alzheimer’s Disease Microbiome Is Associated with Dysregulation of the Anti-Inflammatory P-Glycoprotein Pathway. mBio. 2019;10(3):e00632–19. pmid:31064831
- 32. Nishiwaki H, Hamaguchi T, Ito M, Ishida T, Maeda T, Kashihara K, et al. Short-Chain Fatty Acid-Producing Gut Microbiota Is Decreased in Parkinson’s Disease but Not in Rapid-Eye-Movement Sleep Behavior Disorder. mSystems. 2020;5(6):e00797–20. pmid:33293403
- 33. Heinzel S, Aho V, Suenkel U, Von Thaler AK, Schulte C, Deuschle C, et al. Gut Microbiome Signatures of Risk and Prodromal Markers of Parkinson Disease. Annals of Neurology. 2021;90(3). pmid:34021620
- 34. Hopkins M, Sharp R, Macfarlane G. Variation in human intestinal microbiota with age. Digestive and Liver Disease. 2002;34:S12–8. pmid:12408433
- 35. Woodmansey E, McMurdo M, Macfarlane G, Macfarlane S. Comparison of Compositions and Metabolic Activities of Fecal Microbiotas in Young Adults and in Antibiotic-Treated and Non-Antibiotic-Treated Elderly Subjects. Appl Environ Microbiol. 2004;70(10):6113–22. pmid:15466557
- 36. Tiihonen K, Tynkkynen S, Ouwehand A, Ahlroos T, Rautonen N. The effect of ageing with and without non-steroidal anti-inflammatory drugs on gastrointestinal microbiology and immunology. Br J Nutr. 2008;100(1):130–7. pmid:18279548
- 37. An R, Wilms E, Smolinska A, Hermes G, Masclee A, De Vos P, et al. Sugar Beet Pectin Supplementation Did Not Alter Profiles of Fecal Microbiota and Exhaled Breath in Healthy Young Adults and Healthy Elderly. Nutrients. 2019;11(9):2193. pmid:31547291
- 38. Ruiz‐Ruiz S, Sanchez‐Carrillo S, Ciordia S, Mena M, Méndez‐García C, Rojo D, et al. Functional microbiome deficits associated with ageing: Chronological age threshold. Aging Cell. 2020;19(1):e13063. pmid:31730262
- 39. Dougal K, de la Fuente G, Harris P, Girdwood S, Pinloche E, Geor R, et al. Characterisation of the faecal bacterial community in adult and elderly horses fed a high fibre, high oil or high starch diet using 454 pyrosequencing. PLoS ONE. 2014;9(2):e87424. pmid:24504261
- 40. Theelen M, Luiken R, Wagenaar J, Sloet van Oldruitenborgh-Oosterbaan M, Rossen J, Zomer A. The Equine Faecal Microbiota of Healthy Horses and Ponies in The Netherlands: Impact of Host and Environmental Factors. Animals. 2021;11(6):1762. pmid:34204691
- 41. Morrison P, Newbold C, Jones E, Worgan H, Grove-White D, Dugdale A, et al. The Equine Gastrointestinal Microbiome: Impacts of Age and Obesity. Front Microbiol. 2018;9:3017. pmid:30581426
- 42. Morrison P, Newbold C, Jones E, Worgan H, Grove-White D, Dugdale A, et al. Effect of age and the individual on the gastrointestinal bacteriome of ponies fed a high-starch diet. PLoS ONE. 2020;15(5):e0232689. pmid:32384105
- 43. Henneke D, Potter G, Kreider J, Yeates B. Relationship between condition score, physical measurements and body fat percentage in mares. Equine Veterinary Journal. 1983;15(4):371–2. pmid:6641685
- 44. Yu Z, Morrison M. Improved extraction of PCR-quality community DNA from digesta and fecal samples. BioTechniques. 2004;36(5):808–12. pmid:15152600
- 45. Grimm P, Combes S, Pascal G, Cauquil L, Julliand V. Dietary composition and yeast/microalgae combination supplementation modulate the microbial ecosystem in the caecum, colon and faeces of horses. Br J Nutr. 2020;123(4):372–82.
- 46. Escudié F, Auer L, Bernard M, Mariadassou M, Cauquil L, Vidal K, et al. FROGS: Find, Rapidly, OTUs with Galaxy Solution. Bioinformatics. 2018;34(8):1287–94. pmid:29228191
- 47. Jouany J. Volatile fatty acid and alcohol determination in digestive contents, silage juices, bacterial cultures and anaerobic fermentor contents. Sci Aliments. 1982;2:131–144.
- 48. Grimm P, Philippeau C, Julliand V. Faecal parameters as biomarkers of the equine hindgut microbial ecosystem under dietary change. Animal. 2017;11(7):1136–45. pmid:28065211
- 49. Segata N, Izard J, Waldron L, Gevers D, Miropolsky L, Garrett W, et al. Metagenomic biomarker discovery and explanation. Genome Biol. 2011;12(6):R60. pmid:21702898
- 50. Mueller S, Saunier K, Hanisch C, Norin E, Alm L, Midtvedt T, et al. Differences in Fecal Microbiota in Different European Study Populations in Relation to Age, Gender, and Country: a Cross-Sectional Study. Applied and Environment Microbiology. 2006;72(2):1027. pmid:16461645
- 51. Julliand V, Grimm P. The impact of diet on the hindgut microbiome. Journal of Equine Veterinary Science. 2017;52:23–8.
- 52. Garber A, Hastie P, Murray J. Factors influencing equine gut microbiota: current knowledge. Journal of Equine Veterinary Science. 2020;88:102943. pmid:32303307
- 53. Ghosh T, Das M, Jeffery I, O’Toole P. Adjusting for age improves identification of gut microbiome alterations in multiple diseases. eLife. 2020;9:e50240. pmid:32159510
- 54. Lim M, Hong S, Kim J, Nam Y. Association Between Gut Microbiome and Frailty in the Older Adult Population in Korea. The Journals of Gerontology: Series A. 2021;76(8):1362–8. pmid:33437992
- 55. Verdi S, Jackson M, Beaumont M, Bowyer R, Bell J, Spector T, et al. An Investigation Into Physical Frailty as a Link Between the Gut Microbiome and Cognitive Health. Front Aging Neurosci. 2018;10:398. pmid:30564113
- 56. Zhang Q, Hu W, Deng Y, Wan J, Wang Y, Jin P. Dysbiosis of gut microbiota and decreased propionic acid associated with metabolic abnormality in Cushing’s syndrome. Front Endocrinol. 2023;13:1095438. pmid:36755580
- 57. Ghosh T, Shanahan F, O’Toole P. The gut microbiome as a modulator of healthy ageing. Nat Rev Gastroenterol Hepatol. 2022;19:565–584. pmid:35468952
- 58. Ralston S, Foster D, Divers T, Hintz H. Effect of dental correction on feed digestibility in horses. Equine Veterinary Journal. 2001;33(4):390–3. pmid:11469773
- 59. Zwirglmaier S, Remler H ‐ P., Senckenberg E, Fritz J, Stelzer P, Kienzle E. Effect of dental correction on voluntary hay intake, apparent digestibility of feed and faecal particle size in horse. Animal Physiology Nutrition. 2013;97(1):72–9. pmid:22017617
- 60. Wang X, Li F, Zhang N, Ungerfeld E, Guo L, Zhang X, et al. Effects of supplementing a yeast culture in a pelleted total mixed ration on fiber degradation, fermentation parameters, and the bacterial community in the rumen of sheep. Animal Feed Science and Technology. 2023;296:115565.
- 61. Salvi P, Cowles R. Butyrate and the Intestinal Epithelium: Modulation of Proliferation and Inflammation in Homeostasis and Disease. Cells. 2021;10(7):1775. pmid:34359944
- 62. Tamanai-Shacoori Z, Smida I, Bousarghin L, Loreal O, Meuric V, Fong S, et al. Roseburia spp.: a marker of health? Future Microbiology. 2017;12(2):157–70. pmid:28139139
- 63. Rosero J, Killer J, Sechovcová H, Mrázek J, Benada O, Fliegerová K, et al. Reclassification of Eubacterium rectale (Hauduroy et al. 1937) Prévot 1938 in a new genus Agathobacter gen. nov. as Agathobacter rectalis comb. nov., and description of Agathobacter ruminis sp. nov., isolated from the rumen contents of sheep and cows. International Journal of Systematic and Evolutionary Microbiology. 2016;66(2):768–73. pmid:26619944
- 64. Mäkivuokko H, Tiihonen K, Tynkkynen S, Paulin L, Rautonen N. The effect of age and non-steroidal anti-inflammatory drugs on human intestinal microbiota composition. Br J Nutr. 2010;103(2):227–34. pmid:19703328
- 65.
Merritt A, Julliand V. Gastrointestinal physiology. In: Geor R, Harris P and Coenen M editors. Oxford: Saunders Elsevier; 2013. pp. 3–32.
- 66. O’Toole P, Jeffery I. Gut microbiota and aging. Science. 2015;350(6265):1214–5. pmid:26785481
- 67. Claesson M, Cusack S, O’Sullivan O, Greene-Diniz R, De Weerd H, Flannery E, et al. Composition, variability, and temporal stability of the intestinal microbiota of the elderly. Proc Natl Acad Sci USA. 2011;108(supplement_1):4586–91. pmid:20571116