Age-associated changes in the composition and diversity of swine gut microbiomes have been described, but most of this literature focuses on the early life period and/or commercial pigs being produced for entry into the food chain. We add to this body of knowledge by describing age-related fecal microbiome, fungal and AMR changes in piglets destined to become breeding stock; we contextualize these age-related changes in relation to important physiological stages including first estrus, parturition, lactation, and weaning of the first litter (i.e., 3 weeks to 52 week of age). Relatively few longitudinal studies have been conducted to describe the dynamics of the gut microbiome in gravid hosts, including in both pigs and humans [7], with conflicting results reported in recent studies of pregnant women. The findings of one study [14] suggested that the microbiomes of various body sites (including fecal content) remained relatively stable, while the findings of a previous study showed a significant shift in the including fecal microbiome during the course of gestation [1]. In humans, performing longitudinal studies to profile microbial changes from a young age through gestation and pregnancy is challenging, and differences in overall results might be caused by many potential confounders, including environment and diet, which are almost impossible to control in human studies. In our study, female pigs were used as an animal model and were fed and housed similarly across all study subjects, providing insight into the dynamics of the fecal microbiome and AMR from the youngest age period through first estrus and weaning (end of lactation). Future analysis is needed to understand how our results compare to human microbiome dynamics, but the use of pigs as a model for humans is well-established and our findings suggest that the swine fecal microbiome matures in a stereotypic manner, which is a beneficial characteristic for an animal model system.
The fecal bacterial communities in female pigs exhibited an age-driven pattern, characterized by a rapid shift in composition and enterotypes from early life to the time of estrus (3–32 weeks of age), highlighting the crucial role of this period in microbiome development [30]. These findings are consistent with our previous studies, which demonstrated rapid assembly and increasing diversity of fecal microbial communities with age in adult pigs [18, 31]. Despite changes in housing, diet, and production practices, the fecal microbiome composition showed similar dynamics of richness and diversity between early life and young age in both our current study of female breeding pigs and our previous cohort of production pigs [31], suggesting age as a key factor influencing fecal microbiome composition. Contrary to our hypothesis, the microbiome profile of females in this current study did not significantly change from estrus to lactation (32–72 weeks of age). The majority of pig fecal samples maintained the same enterotypes from parturition to weaning, with only a few transitioning to the Streptococcus-dominated enterotype. This stability was consistent across various taxonomic levels, with only modest changes in specific taxa during lactation. Similar fecal microbiome stability has been reported in lactating sows [13], possibly indicating functional similarities in the fecal microbiome during these stages. Further research is needed to elucidate the factors influencing the stability of the fecal microbiome during lactation, a critical period for sows and their piglets [32].
Age-related changes in the fecal microbiome have been associated with variations in AMR in fecal bacteria and gene quantities in pigs [18, 31]. When we performed phenotypic assessments of AMR within specific genera, we observed the highest counts of coliforms and enterococci with AMR; subsequently, these values also declined during the first 3 months and then varied greatly depending on the drug class. The highest resistance was detected against tetracyclines, and lower resistance was recorded against chloramphenicol. As the pigs progressed to estrus, increasing trends in the counts of coliforms resistant to the tested antimicrobial classes, including ceftriaxone, tetracyclines, macrolides, and phenicols, and the counts of enterococci resistant to the tested antimicrobial classes, namely, quinolones and tetracyclines, were observed.
Further, quantities of AMR genes (both absolute levels and the levels standardized to 16S rRNA gene copies) — specifically, a tetracycline resistance gene (tetA) and 3rd -generation cephalosporin resistance gene (blaCTXM) copies — within the bacterial community, were initially higher during the early life period, in the absence of antibiotic use and even when the fecal microbiome was least diverse. Later, all the values decreased during the first 2-2.5 months of age. Interestingly, the abundance of the tet(A) gene was elevated from 10 to 22 weeks of age/prior to breeding, but the blaCTX-M gene did not exhibit this trend, which persisted until the time of estrus; however, as the pigs progressed to parturition, the levels all started to decline again. The presence of bacteria with AMR in the pigs fecal microbiome does not necessarily indicate that the individuals were exposed to those specific antimicrobials but could instead be attributed to the transfer of AMR genes through horizontal or vertical transmission between bacteria [33] and/or a shift in bacterial types to those harboring resistance genes. Thus, we suspect that the variation in the quantities among these stages could be attributed to age-related changes in the underlying bacterial population within the fecal microbiome [18, 34]. However, while we could not confidently link AMR gene-associated taxa, the consistent shift of microbial enterotypes from the early life to growing stages further suggests the possible association between underlying microbiome composition and associated fecal AMR dynamics.
Consistent with previous findings [33], the phyla Firmicutes and Bacteroidetes were most abundant in the pig gut across all age points, regardless of the physiological stage (i.e., early life to parturition and weaning). Consistent with the findings of a previous study [1, 13], relative abundance of Actinobacteria gradually increased during gestation. Notably, relative abundance of Proteobacteria did not increase as previously reported [1] but was instead similar to that observed by Liu et al. [13]. A striking observation in this study was that relative abundance of Treponema (phylum Spirochaetota), one of the dominant taxa in enterotype 4, increased rapidly during early age through parturition. Although a majority of the members of the Treponema genus were not classified at the species level, of those classified to species, we observed a higher relative abundance of Treponema bryantii as pigs progressed to estrus. Members of the genus Treponema include both commensal bacteria and pathogens [35], with T. bryantii recently linked to feed efficiency in sows [36] and reported to be present in the pig gut at the end of gestation [7]. A previous study found a significant interaction between host sex hormones and the spirochetes Treponema spp. [37], which was also associated with a specific period of the estrus cycle; however, further studies are needed to understand the specific roles of Treponema spp. during gestation. The relative abundance of Bifidobacterium (phylum Actinobacteria) also increased with age. Bifidobacterium is considered a beneficial bacterium and research has shown that the hormone progesterone can directly influence the composition of the gut microbiome in pregnant women, leading to an increase in the relative abundance of Bifidobacterium during late pregnancy [38]. Relative abundance of Lactobacillus spp. was depleted during the period from 3 weeks of age to estrus, but their relative abundance increased with the time of weaning of first litter. These findings highlight the potential association of species-specific bacteria with responses during different physiological stages, such as estrus and parturition in female pigs.
Although the bacterial constituents of gut microbial communities have been studied extensively, fecal-associated fungi and their roles are poorly understood in humans [23] as well as in pigs [39]. In our study, fewer fecal-associated fungi were detected across all age points, and their diversity did not follow the same age-related pattern as the fecal bacterial community (Fig. 5). In our study, Ascomycota and Basidiomycota are two dominant phyla and Candida, Kazachstania, Issatchenkia, and Diutina are the most abundant genera across all samples. These fungal taxa also reported on the gut mycobiome of the Human Microbiome Project healthy cohort (HMP) [40, 41]. Interestingly, we noticed a higher abundance of fungal genera, including Candida and Kawachstania spp., at parturition. Our study revealed that approximately 20% of fungal ASVs were classified as known fungal genera, and previous work from the HMP mycobiome study [41] reported a similarly lacking taxonomic information. ITS sequencing provides greater resolution of fungal constituents of the microbiome; however, the lower numbers of fungi in the GI tract and several technical challenges, including the sparse database and accuracy of taxonomic information, present challenges to identifying key fungal features. This suggests standardizing fungal databases, DNA isolation and sequencing methods, and bioinformatics data analysis will be crucial for comprehensive mycobiome data analysis in the future.
Strengths and limitations
Despite the increase in microbiome research, there is still insufficient published research on microbiome dynamics and AMR, particularly from the early life period through estrus and lactation/weaning in females. The major strengths of this study include the following: a) its longitudinal design; importantly, we followed animals from the 3 weeks of age through estrus, parturition and weaning (3–53 weeks of age, i.e., parity 1) to capture the trajectories of the gut microbiome and mycobiome and during the entire period of first parity, using sequence-based techniques (16S rRNA gene and ITS based); and b) the quantification of AMR fecal bacteria within the bacterial community using a culture-based approach that provides broader insight into whether the antimicrobial-resistant fecal bacterial population changes from the early life period through parturition and weaning.
However, we note some potential limitations. Although 16S rRNA sequencing is widely used to characterize the microbial community across different sample types, this approach also has limitations, including the inability to classify all sequence features at the species level [42]. For instance, in our study of 142 ASVs that were classified as Treponema, only four (and none of the Bifidobacterium ASVs) were classified at the species level. Thus, future studies using multiple approaches, including species-level identification (e.g., with shotgun sequencing), will provide additional insights into bacterial species and associated AMR genes, supplementing those gained from this study. Similarly, changes in the maternal microbiome occur beyond the gut; thus, microbiomes at other body sites, including vaginal microbiomes[43], along with the hormonal profile should be considered for future studies, and such studies will undoubtedly shed light on how microbes interact with each other and their association with health. We acknowledge the limitations of performing microbial culture-based phenotypic AMR analysis on just two indicator bacteria that may not represent the true source population (i.e., the gut microbial community); however, these two broad categories of gram-negative and gram-positive bacteria are representative of clinically important pathogens, and their resistance profiles are highly relevant.