Intestinal Permeability, Gut Inflammation, and Gut Immune System Response Are Linked to Aging-Related Changes in Gut Microbiota Composition: A Study in Female Mice

Abstract

Aging entails changes at the cellular level that increase the risk of various pathologies. An association between gut microbiota and age-related diseases has also been attributed. This study aims to analyze changes in fecal microbiota composition and their association with genes related to immune response, gut inflammation, and intestinal barrier impairment. Fecal samples of female mice at different ages (2 months, 6 months, 12 months, and 18 months) and gene expression in colon tissue were analyzed. Results showed that the older mice group had a more diverse microbiota than the younger group. Additionally, the abundance of Cyanobacteria, Proteobacteria, Flavobacteriaceae, Bacteroides, Parabacteroides, Prevotellaceae_UCG-001, Akkermansia, and Parabacteroides goldsteinii increased with age. In contrast, there was a notable decline in Clostridiaceae, Lactobacillaceae, Monoglobaceae, Ligilactobacillus, Limosilactobacillus, Mucispirillum, and Bacteroides faecichinchillae. These bacteria imbalances were positively correlated with increased inflammation markers in the colon, including Tnf-α, Ccl2, and Ccl12, and negatively with the expression of tight junction genes (Jam2, Tjp1, and Tjp2), as well as immune response genes (Cd4, Cd72, Tlr7, Tlr12, and Lbp). In conclusion, high levels of diversity did not result in improved health in older mice; however, the imbalance in bacteria abundance that occurs with aging might contribute to immune senescence, inflammation, and leaky gut disease.

The proportion of older people around the world is increasing dramatically. Several diseases are associated with the process of aging as a result of the accumulation of a number of molecular and cellular damages (1). For example, as we age, the intestine, which plays a critical role in immunity and acts as a barrier for pathogens and toxins, decreases its ability to self-repair after damage resulting in a leaky gut (2).

It has been widely recognized that the intestinal barrier has a direct impact on health and disease. The intestinal epithelium is a single-cell layer connected by intercellular tight junction proteins such as claudins, junctional adhesion molecules, and zonula occludens. These proteins are downregulated by inflammatory cytokines, thus increasing the risk of endotoxemia and causing inflammation on a local and systemic level (3–5). In addition, inflammation and intestinal permeability have been linked with an imbalance of intestinal bacteria (dysbiosis) (6,7). A recent study in obese older mice showed that Lactobacillus and Enterococcus strains may contribute to decreasing gut inflammation and enhancing the expression of tight junctions (8). Physiological changes related to aging may have an impact on gut microbiota composition and diversity (9). Currently, the research on this topic is limited, and there are conflicting results regarding the association between aging and some bacteria because both an increase (10,11) and a decrease (4) have been observed in the same species. It has been shown, for example, that although some studies suggest Akkermansia abundance increases with age (12,13), others have found the exact opposite (10,14). There is evidence that this bacterium plays a significant role in improving immunomodulation, reducing inflammation, and maintaining gut barrier function (10,14). Interestingly, it has also been found that centenarians have higher levels of this bacteria than healthy adults (10).

The aim of this study was to analyze changes in fecal microbiota composition and their association with genes related to immune response, gut inflammation, and intestinal barrier impairment. We investigated fecal microbiota composition at different lifespan states: young (2 months old), adult (6 months old), middle-aged (12 months old), and old (18 months old). To reach this objective, samples from the OBELEX project were used. As part of this study, female mice were studied and our team found metabolic disorders including an increase in glucose and cholesterol levels, as well as changes in body composition such as an increase in visceral fat levels as the mice aged (15).

Materials and Methods

Mice and Study Design

C57BL/6J female mice (Harlan Laboratories, Barcelona, Spain) were housed at the animal facilities of the University of Navarra in a specific pathogen-free environment under controlled conditions (22 ± 2°C, with a 12-hour light-dark cycle, relative humidity, 55 ± 10%). In total, 37 mice were fed ad libitum with a normal diet containing: 20% proteins, 67% carbohydrates, and 13% lipids (Harlan Teklad Global Diets, Harlan Laboratories, Indianapolis, IN, USA), and with free access to water. Mice were sacrificed at different periods: 2 months old (n = 10), 6 months old (n = 7), 12 months old (n = 10), and 18 months old (n = 10), and tissues and feces were collected and immediately stored at −80°C. It is important to note that this study is part of the OBELEX project of the University of Navarra, thus 3 mice from the 6-month-old group were sacrificed to obtain samples for other analyses. All experiments were performed by national animal care guidelines, and with the approval of the Ethics Committee for Animal Experimentation of the University of Navarra (Protocol 113-15), under the EU Directive 2010/63/EU.

cDNA Synthesis and Quantitative Real-Time PCR

RNA was isolated from colon tissue using TRIzol Reagent (Invitrogen, ThermoFisher Scientific, Waltham, MA, USA). RNA (1–5 g) was then incubated with DNase I (RapidOut DNA Removal kit, Thermo Fisher Scientific) at 37°C for 30 minutes. RNA quality and concentration were determined using the Nanodrop Spectrophotometer ND1000 (Nanodrop Technologies, Inc., Wilmington, NC, USA). Retrotranscription to cDNA was performed using High-Capacity cDNA Reverse Transcription (Applied Biosystems, Foster City, CA, USA) according to the manufacturer’s instructions. Then, quantitative PCR was performed using the Touch Real-Time PCR System (C1000 + CFX384, BIO-RAD, Hercules, CA, USA), and gene expression was analyzed using Power SYBR Green PCR Master Mix. Primers were obtained from the online PrimerBank and a previous study (16) (Supplementary Table S1). Relative gene expression was determined by the 2^−ΔΔCT method, and the expression data were normalized with the Gapdh gene expression. The samples were processed in duplicate, and the mean values were used for statistical analysis.

Fecal DNA Extraction and 16S rRNA Gene Sequencing

Microbiota analysis in stool samples was performed in collaboration with CIMA LAB Diagnostics (University of Navarra), which followed the established protocol. Briefly, the Maxwell RSC Fecal Microbiome DNA kit (Promega, Madison, WI, USA) in the Maxwell RSC Instrument was used to efficiently isolate DNA from OMNIgene. GUT kits following the manufacturer’s instructions; dsDNA characterization was performed with Qubit (ThermoFisher Scientific). Two PCR reaction protocols were used to prepare samples for sequencing the variable V3 and V4 regions of the 16S rRNA gene. The first PCR was performed with the next instructions: (1) 50 ng of dsDNA in 2.5 µL, (2) 15.5 µL of master mix, (3) 5 µL of F and 5 µL R primers (Forward Primer: 5ʹ TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCTACGGGNGGCWGCAG; Reverse Primer: 5ʹ GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGACTACHVGGGTATCTAATCC), and (4) the following PCR program within SimpliAmp Thermal Cycler (ThermoFisher Scientific): 95°C for 3 minutes, and 25 cycles of 95°C for 30 seconds, 55°C for 30 seconds, 72°C for 30 seconds, and finally, 72°C for 5 minutes, to later keep refrigerated at 4°C. The PCR product was purified using AMPURE and quantified using Qubit (ThermoFisher Scientific).

The second PCR was performed following the next instructions: (1) 15 µL of 1st PCR product, (2) 25 µL of master mix, (3) 5 µL of F and 5 µL of R primers (indexing step), and (4) the following PCR program within SimpliAmp Thermal Cycler (ThermoFisher Scientific): 95°C for 3 minutes, and 8 cycles of 95°C for 30 seconds, 55°C for 30 seconds, 72°C for 30 seconds, and finally, 72°C for 5 minutes, to later keep refrigerated at 4°C. The PCR product was purified using AMPURE and quantified using Qubit (ThermoFisher Scientific). The sequencing was performed on the Illumina MiSeq SY-410-1003 system according to the manufacturer’s protocols.

Bioinformatics and Statistical Analysis

Bioinformatics analyses were performed in collaboration with the Bioinformatics Platform of CIMA (University of Navarra). 16S rRNA sequences obtained were filtered following quality criteria of the OTU processing pipeline LotuS2 (17) (release 2.19). This pipeline includes UPARSE de novo sequence clustering (18), removal of chimeric sequences and phix contaminants (19) for the identification of OTUs (Operational Taxonomic Units), and OTU abundance matrix generation. Finally, taxonomy was assigned using lambda aligner (20) and SILVA 16S/18S database (21) achieving up to species-level sensitivity.

Data raw counts were normalized (center log-ratio) with the ALDEx R-script package. Shannon and Simpson’s indexes were calculated at the genus level using the microbiome R-script package to determine the alpha diversity. Beta diversity was determined and visualized in a non-metric multidimensional scaling (NMDS) plot based on Bray–Curtis distances at the OTUs level using the ecodist R-script package. For multiple comparisons, bacteria not found in at least half the samples from each group were discarded. Relative abundance was used for the graphical representation. Kruskal–Wallis’s test and the post hoc Bonferroni test were used to adjust multiple comparisons.

For genes, the one-way ANOVA test was used when the data were normally distributed. Kruskal–Wallis’s test was used when the data were not normal. The Tukey (parametric) and Tamhane (nonparametric) tests post hoc were performed to determine differences between groups. Associations between relative abundance and gene expression were calculated using Spearman’s rank correlation coefficient. The p value < .05 was considered statistically significant and correlations were adjusted by Benjamini–Hochberg’s FDR controlling procedure (q value < 0.05).

The analyses were performed using RStudio 2022.07.1 for macOS, and gene graphs were performed with GraphPad Prism 9.1.2 for macOS. Data are shown as means ± SEM unless otherwise noted.

Results

Diversity and Composition

To determine how age affects the gut microbiota, we first evaluated differences in alpha (a summary of the microbial community richness and evenness in each sample (22)) and beta diversity (a measure of interindividual diversity assessing the similarity between communities (22)) at 2, 6, 12, and 18 months of age. According to the Shannon and Simpson indexes, alpha diversity significantly increased as the mice aged, as shown in Figure 1A and B. Beta diversity was analyzed by NMDS based on Bray–Curtis distances at the OTUS level, which revealed that age clusters are different (Figure 1C).

Figure 1.

Comparative analysis of alpha diversity at genus level in young (2 months), adult (6 months), middle-aged (12 months), and old (18 months) female mice: (A) Shannon index and (B) Simpson index. Data are presented as mean ± SEM; (C) non-metric multidimensional scaling (NMDS) at the OTUS level based on Bray–Curtis distances showing dissimilarities between groups. (D) Relative abundance at the phylum level. *p < .05 versus 2 months. OTUS = operational taxonomic units; SEM = standard error of the mean.

Next, different taxonomic levels were analyzed (phylum, family, genus, and species). As detailed in Figure 1D  Bacteroidetes and Firmicutes were the most abundant phyla in all groups. Furthermore, the results of the study indicated that older groups had a significantly increased abundance of the phylum Cyanobacteria and Proteobacteria (Figure 2A and B). In addition, the abundance of the Flavobacteriaceae family, well-known pathogens (23), increased from 6 months old onwards (Figure 2C). In contrast, there was a notable decline in the Clostridiaceae, Lactobacillaceae, and Monoglobaceae families, members of the Firmicutes phylum (Figure 2D–F).

Figure 2.

Relative abundance of some phyla (A and B) and families (C–F) differs by age in female mice. Data are presented as mean ± SEM. f = family, p = phyla. ***p < .001, **p < .01, *p < .05 versus 2 months. SEM = standard error of the mean.

At the genus level, aging promoted higher levels of Bacteroides, Parabacteroides, Prevotellaceae_UCG-001, and Akkermansia (Figure 3A–D). On the contrary, aging was accompanied by lower levels of Ligilactobacillus, Limosilactobacillus, and Mucispirillum (Figure 3E–G). Finally, at the species level, we found an unexpected increase in Parabacteroides goldsteinii in middle-aged and old mice (Figure 3H), which has been associated with improving intestinal integrity and reducing inflammation (24). Additionally, we found an increase, as mice aged, in the abundance of Bacteroides faecichinchillae, which has been poorly studied (Figure 3I).

Figure 3.

Relative abundance of the main age-related genus (A–G) and species (H–I) shifts in female mice. Data are presented as mean ± SEM. g = genus; s = species. *p < .05, **p < .01, ***p < .001 versus 2 months; ^#p < .05, ^##p < .01 versus 6 months. SEM = standard error of the mean.

Gut Permeability

We estimated potential changes in gut permeability through the analysis of the expression of different tight junction genes. Intestinal epithelial cells are connected by tight junctions (claudins, junctional adhesion molecules, and zonula occludens), which maintain intestinal barrier integrity (3). Age-related changes are evident in our results (Figure 4A). Claudin 8 (Cldn8) increased significantly between 2 months and 6 months (p = .04). The junctional adhesion molecule 2 (Jam2) decreased as the mice aged, as observed when comparing 2 months versus 18-month-old mice (p = .04). The tight junction protein 1 (Tjp1) remains stable in adult and middle-aged mice but exhibited a significant downregulation in old mice (2 months vs 18 months, p = .008). However, the tight junction protein 2 (Tjp2) was downregulated earlier (2 months vs 6 months, p = .01) and remained at this level until 18 months old (p = .01).

Figure 4.

Age-related changes in the expression of genes related to (A) intestinal barrier function, (B) inflammatory marker genes, and (C) immunity genes assessed in colon tissue of female mice. Data are presented as mean ± SEM. *p < .05, **p < .01, ***p < .001 versus 2 months; ^#p < .05 versus 6 months; ^Δp < .05 versus 12 months. (D) Correlation plots for the gene expression represent the Spearman coefficient in positive (blue) and negative (red) colors, and the larger the circle, the stronger the correlation. f = family, g = genus, p = phyla, s = species; SEM = standard error of the mean.

The Spearman correlation coefficient was calculated to find associations between bacterial abundance and gene expression (Figure 4D). We found a positive correlation between Jam2 gene expression and bacteria that were present in high levels in younger mice compared with older mice, such as the Lactobacillaceae (rho = 0.4, p = .03) and Monoglobaceae families (rho = 0.5, p = .001), as well as the Ligilactobacillus genus (rho = 0.4, p = .03). Moreover, the following taxa were present in low percentages in the younger groups and there was a negative correlation between Jam2 and them: Cyanobacteria (rho = −0.6, p = .001), Proteobacteria (rho = −0.6, p = .002), Akkermansia (rho = −0.4, p = .04), Bacteroides (rho = −0.5, p = .01), Parabacteroides (rho = −0.5, p = .01), Prevotellaceae_UCG-001 (rho = −0.7, p < .001), B faecichinchillae (rho = −0.5, p = .01), and P goldsteinii (rho = −0.4, p = .049). Regarding Tjp1, we found only a negative correlation with Bacteroides increase (rho = −0.4, p = .04). There were, however, statistically significant positive correlations between Tjp2 and high levels of the Clostridiaceae (rho = 0.5, p = .005), Lactobacillaceae (rho = 0.5, p = .005), and Monoglobaceae families (rho = 0.4, p = .03), and the Ligilactobacillus genus (rho = 0.6, p < .001). Further, the results indicate that Tjp2 was negatively correlated with several bacteria that increased as mice aged, including Cyanobacteria (rho = −0.6, p = .001), Proteobacteria (rho = −0.5, p = .01), Flavobacteriaceae (rho = −0.04, p = .03), Bacteroides (rho = −0.4, p = .02), Parabacteroides (rho = −0.5, p = .01), Prevotellaceae_UCG-001 (rho = −0.4, p = .04), B faecichinchillae (rho = −0.6, p = .002), and P goldsteinii (rho = 0.4, p = .01).

Gut Inflammation

Afterward, some inflammation markers in colon tissue were assessed (Figure 4B). There was an increase in Tnf-α expression as the mice grew older (2 months vs 18 months, p = .001). Several studies have suggested that excessive production of proinflammatory cytokines, such as TNF-α, may contribute to the increase in intestinal permeability by modulating tight junction proteins (3,4,9), which is consistent with the correlation between Tnf-α and Tjp1 found in this study (rho = −0.5, p = .02), as shown in Figure 4D. In contrast, IL-6 plays a key role in maintaining intestinal epithelial development and homeostasis. It has been suggested that IL-6 stimulates intestinal epithelial cells to secrete claudin-2 (25). In the present study, mice showed a significant decline in IL-6 expression between 12 months and 18 months of age (p = .04). In addition, there was a positive correlation between IL-6 and Cldn2 (rho = 0.6, p < .001), see Figure 4D. Similarly, IL-10 upregulates the expression of tight junction proteins by suppressing proinflammatory cytokine secretion (26). In the current study, Il-10 expression was reduced in older mice (2 months vs 18 months, p = .001). Moreover, Il-10 was positively correlated with tight junction gene expression such as Cldn2 (rho = 0.6, p = .02), Jam2 (rho = 0.6, p = .01), and Tjp1 (rho = 0.7, p = .004), see Figure 4D. On the other hand, CCL2 and its homolog CCL12 have been implicated in inflammatory conditions, and high levels of CCL2 have been linked to shortening life expectancy (27). We found a marked increase with aging in the expression of both Ccl2 and Ccl12 genes (2 months vs 18 months, p = .003 and p = .03, respectively).

Furthermore, we searched for associations between these inflammatory genes and bacteria. As detailed in Figure 4D, Tnf-α was positively correlated with an increase in the proportion of the Proteobacteria phylum (rho = 0.4, p = .02), the Flavobacteriaceae family (rho = 0.4, p = .049), Akkermansia (rho = 0.5, p = .003), and Parabacteroides genus (rho = 0.5, p = .01); and the species B (rho = 0.5, p = .004), and P goldsteinii (rho = 0.6, p = .001). Conversely, the low abundance of Clostridiaceae (rho = −0.5, p = .01), Lactobacillaceae (rho = −0.4, p = .03), and Monoglobaceae families (rho = −0.5, p = .01) exhibited a negative association with Tnf-α as well as the Ligilactobacillus (rho = −0.5, p = .01), and Mucispirillum genus (rho = −0.6, p < .001). Additionally, a positive correlation was found between the chemokine Ccl2 and high levels found in older groups of Akkermansia (rho = 0.5, p = .01) as well as P goldsteinii (rho = 0.5, p = .01). By contrast, a negative correlation was found with the decline in the abundance of Lactobacillaceae (rho = −0.4, p = .03), Monoglobaceae families (rho = −0.5, p = .01), and Ligilactobacillus (rho = −0.5, p = .01), and Mucispirillum genus (rho = −0.6, p < .001). Significant correlations were not found between age-related changes in bacterial abundance and the gene expression of the antiinflammatory cytokine Il-10.

Immune Response

Lastly, some genes were analyzed to assess the gut immune response (Figure 4C). It has been proposed that gut microbiota regulate intestinal immunity, and the immune system can also be influenced by age (5). A total of 6 immune genes were studied from colon tissue. The expression of Cd4, which is required for initiation or enhancement of T-cell activation (28), was downregulated in older mice (p = .02). Similarly, Cd72 which plays a role in B-cell proliferation and differentiation (29) was decreased in the older group (p = .02). Moreover, toll-like receptors (Tlr) have a key role in pathogen recognition and innate immune response by activating inflammatory signaling pathways, stimulating immunoglobulin A (IgA) production, maintaining tight junctions, and promoting antimicrobial peptide expression (30). There was a significant increase in the Tlr4 gene expression in adult mice (2 months vs 6 months, p < .001), and it remained high in middle-aged and older mice. Interestingly, this study showed that Tlr7 expression was downregulated from 6 to 18 months (p = .041), as well as Tlr12 from 2 months to 18 months (p < .001). Finally, the lipopolysaccharide-binding protein (LBP) enhances the recognition of endotoxin and bacterial surfaces by the immune system (31). According to the results of this study, Lbp downregulation in mice began at 6 months of age (2 months vs 6 months, p < .04).

A correlation analysis was conducted, as before, to determine a relationship between immunity gene expression and microbiota composition (Figure 4D). Significant positive correlations were found between Cd4 and the bacteria that diminish with aging such as Clostridiaceae (rho = 0.5, p = .02), and Lactobacillaceae families (rho = 0.5, p = .007), as well as Ligilactobacillus (rho = 0.6, p = .003), and Mucispirillum genus (rho = 0.6, p = .005). In contrast, those that increased in the older mice showed a negative correlation with Cd4 gene expression: Cyanobacteria (rho = −0.4, p = .02), Akkermansia (rho = −0.5, p = .02), Parabacteroides (rho = −0.6, p = .002), Prevotellaceae_UCG-001 (rho = −0.4, p = .02), B faecichinchillae (rho = −0.6, p = .001), P goldsteinii (rho = −0.5, p = .003). Regarding Cd72, there was a positive correlation with the increase in the Mucispirillum genus (rho = 0.5, p = .02), and a negative correlation with low levels of Flavobacteriaceae family (rho = −0.4, p = .03). Concerning the Tlr7 gene, bacteria that in young mice groups were more abundant compared to the older group and had a positive correlation with the Clostridiaceae (rho = 0.5, p = .05), Lactobacillaceae (rho = 0.5, p = .01), and Monoglobaceae (rho = 0.6, p = .005) families, and the Ligilactobacillus genus (rho = 0.6, p = .002). On the contrary, bacteria increased in older mice such as the Cyanobacteria (rho = −0.6, p = .001), Proteobacteria (rho = −0.5, p = .009), Akkermansia (rho = −0.5, p = .01), Bacteroides (rho = −0.4, p = .03), Parabacteroides (rho = −0.6, p < .001), Prevotellaceae_UCG-001 (rho = −0.6, p = .002), B faecichinchillae (rho = −0.6, p = .001), and P goldsteinii (rho = −0.6, p < .001) showed a negative correlation with the Tlr7 mRNA expression. The Tlr12 expression was positively correlated with bacteria that were most abundant among young mice groups including the families Clostridiaceae (rho = 0.4, p = .03), Lactobacillaceae (rho = 0.5, p = .01), and Monoglobaceae (rho = 0.5, p = .005). On the contrary, Tlr12 was negatively correlated with bacteria that have distinguished aging in mice: the Flavobacteriaceae family (rho = −0.4, p = .04), the Akkermansia (rho = −0.4, p = .03), and Parabacteroides genus (rho = −0.4, p = .03), as well as the B faecichinchillae (rho = −0.5, p = .007), and P goldsteinii (rho = −0.4, p = .03) species. Finally, Lbp gene expression was positively correlated with bacteria that declined as mice age: family Clostridiaceae (rho = 0.6, p = .006), Lactobacillaceae (rho = 0.5, p = .01), and Monoglobaceae (rho = 0.6, p = .001), as well as the genus Ligilactobacillus (rho = 0.6, p = .005), and Mucispirillum (rho = 0.6, p = .004); and negatively correlated with bacteria found in younger mice at low levels: Cyanobacteria (rho = −0.6, p = .002), and Proteobacteria (rho = −0.6, p = .005) phyla, the Flavobacteriaceae family (rho = −0.5, p = .01), the genus Akkermansia (rho = −0.6, p = .001), Bacteroides (rho = −0.6, p = .006), Parabacteroides (rho = −0.7, p < .001), Prevotellaceae_UCG-001 (rho = −0.6, p = .004), and species such as P goldsteinii (rho = −0.6, p = .001), and B faecichinchillae (rho = −0.7, p < .001).

Discussion

Gut microbiota composition could be considered a hallmark of aging (32). Although numerous studies in this field have been conducted, gut microbiota composition in aging as well as their relationship to age-related health decline remain poorly understood (33,34). We examined the gut microbiota composition of female mice in 4 age groups (young, adult, middle-aged, and old) and their association with markers of intestinal permeability, colon inflammation, and intestinal immune response. Results of this study indicate that the older group had a more diverse microbiota than the younger groups, which agrees with other studies conducted on male mice (4,13,35,36). Unfortunately, studies in female mice have been limited on this topic. We found a study conducted by Cao et al., in female mice, where a reduction in bacterial diversity was observed, but their diet had a higher carbohydrate content compared with our study. Evaluating the effect of the impact of the diet has not been the objective of the present study, the animals were fed with a standard diet. Based on this, it is possible that a healthy diet could prevent bacterial diversity decline in aged mice. Further research is needed to assess the impact of different diets on aging and gut microbiota composition and its link to healthy aging (12).

In our study, the abundance of pathogens such as the Cyanobacteria and Proteobacteria phyla increased with aging. A study conducted by Bárcena et al. (10) found similar results using mouse models of progeria (a genetic disorder that speeds up aging) in both sexes, as well as the study of Liu et al. (4) with humans and mice samples. Furthermore, we found the abundance of these bacteria was positively correlated with gut inflammation, a weak gut barrier, and a diminished immune response. Cyanobacteria produce toxins that lead to neurodegenerative diseases and have also been linked to irritable bowel disease (37). To our knowledge, this is the first time that their relationship with intestinal permeability and weak immune response was observed, which points to the importance of conducting further research to understand the role of this phylum in human health. Many bacteria that belong to the Proteobacteria phylum are pathogenic to humans as well (38). Hexaacylated form of lipopolysaccharide produced by these bacteria may cause chronic inflammation (39) a characteristic of aging that could be related to the colon inflammation observed in our study.

Furthermore, we found a higher abundance in old mice of the Flavobacteriaceae family, the Bacteroides, Parabacteroides, and Prevotellaceae_UGC-100 genus, and the P goldsteinii, and B faecichinchillae species, which are members of the Bacteroidetes phylum. They were all negatively correlated with the expression of genes related to immunity. A recent study by Qiu et al. (40) showed a positive correlation between Flavobacteriales and immunosuppressive biomarkers in zebrafish. However, the Flavobacteriaceae family has not been extensively studied, thus more research is necessary to clarify its role. According to our knowledge, this is the first time this family has been linked to aging and intestinal permeability, thus the increase of these bacteria in old mice could influence the expression of Tjp2 in the colon.

Regarding Bacteroides, they are gram-negative bacteria and can modulate their cell-wall polysaccharides to alter their immunity recognition capabilities (41). Therefore, this genus is able to avoid the immune response of the host, which explains the negative correlation with the immunity-related genes found in this study. In a recent study, supplementation with a Parabacteroides strain prevented the disruption of tight junction proteins by inhibiting NF-Kβ activation and lowering inflammation levels (42). Additionally, P goldsteinii produces acetic and succinic acids as end-products of glucose metabolism and has been negatively correlated with liver inflammation (43,44). In contrast, in the present study, this bacterium was found in high abundance in old mice, also the behavior of this bacteria showed a link with the inflammation markers and with a decrease in tight junction expression; therefore, the increase in the presence of this strain might be related to the intestinal disease of aging. The study led by Guo et al. (45) also reported significantly higher levels of the Parabacteroides genus in frail older persons suggesting that this genus may play a crucial role in the development of diseases related to aging. Continuing research on this bacterium could potentially serve as an indirect marker for evaluating health status in aging.

In contrast, we observed that the abundance of the Clostridiaceae, Lactobacillaceae, and Monoglobaceae families, as well as the Ligilactobacillus genus, were reduced in older mice. Interestingly, the present study also suggests that as mice age, their tight junction expression decreases, and there is a link between this and the decrease in the Clostridiaceae family which could be explained by the production of the metabolites of these bacteria. It is widely known that Clostridium bacteria are butyrate producers, according to the study by Fang et al. (46), butyrate enhances the expression of tight junctions in obese mice. Our results are also consistent with studies that evaluated the effect of supplementation with a strain of bacteria belonging to the Clostridiaceae family (Clostridium butyricum), which not only showed enhanced expression of tight junction proteins but decreased colon inflammation in obese mice and in a mouse model of induced colitis (47,48).

Concerning Lactobacillaceae family members, the study conducted by Ahmadi et al. (8), demonstrated that supplementation with some strains of Lactobacillus and Enterococcus can increase taurine levels. As a result, leaky gut is reduced by stimulating tight junction expression in the colon, and inflammation is improved. Similarly, Wong et al. (49) observed an improvement in colon inflammation by Lactobacillus casei supplementation, via increasing taurine-conjugated bile acids. In this regard, this study showed that the Lactobacillaceae family, which decreased with age, had a positive correlation with tight junction expression and a negative association with inflammation genes. In addition, Ligilactobacillus is a genus member of the Lactobacillaceae family that has gained recognition for its health benefits for the host. In this sense, our analysis revealed a negative correlation between this genus and proinflammatory cytokines. Moreover, Yao et al. (50) examined the effect of Ligilactobacillus salivaris on colon inflammation and determined that this strain could reduce inflammatory cytokines and increase antiinflammatory ones. Thus, maintaining adequate levels of this family may prove beneficial in preventing age-related health decline.

Despite the limited knowledge of Monoglobaceae, it has been shown that Monoglobus pectinilyticus, a member of this family, can degrade pectin, which may reduce the severity of colitis (51). Additionally, pectin degradation appears to be capable of modulating intestinal immunity by increasing the expression of IgA (52). This could explain the association of the Monoglobaceae family with a better immune response found in the present study. The immune response of older adults may be affected as a result of the decline in the abundance of Monoglobaceae bacteria.

The present study indicated that the Akkermansia proportion increased with age. This rise was associated with the high expression of gut inflammation markers and the suppression of immune-related genes in the colon. Our findings are consistent with those observed by Van der Lugt et al. (16) who observed that Akkermansia muciniphila supplementation resulted in the downregulation of Cd4 gene expression in aging mice. It has been observed in other studies that Akkermansia levels increase among centenarians (10,11), suggesting that these microorganisms may contribute to longevity, although there is not enough evidence to clarify their role in the maintenance of health.

This study provides insight into the diversity of gut microbiota composition as well as the relationship between the major hallmarks of aging and their association with the different bacteria found in the gut at different stages of life. The main limitation of this study was that only female mice were used in the experiment. However, several previous studies have been carried out only on male mice, which allows discussion and comparison with them. As an example, one research project using male mice found the same result, the decline in Lactobacillus and Bacteroides abundance by aging, and on the contrary, they found that Akkermansia abundance was higher in young mice (4). Further research is needed to confirm if this difference could be attributed to gender.

In summary, the older mice exhibited more gut microbiota diversity, but this was not translated into better health. On the contrary, old mice were losing levels of mutualist bacteria and increased the pathogens, which can contribute to triggering age-related diseases such as immunosenescence, inflammaging, and leaky gut. More investigations are necessary to establish the adequate level of each bacterium in order to find a new therapy that contributes to improving the quality of life in the older population.

Funding

This study was funded by Ministerio de Ciencia e Innovación/Agencia Estatal de Investigación – European Regional Development Fund (MINECO/FEDER) of the Government of Spain and European Union, respectively (BFU2015-65937-R); by CIBER - Centro de Investigación Biomédica en Red- (CIBEROBN - CB12/03/30002), Instituto de Salud Carlos III; and through the Department of University, Innovation and Digital Transformation, Government of Navarra (PI026 BIOAGEMT).

Conflict of Interest

None.

Author Contributions

P.E.G.-M.: Carried out molecular biology techniques, analyzed the data, and writing—original draft preparation; E.F.-S.: carried out animal experiments, writing review, and editing; M.S.: carried out molecular biology techniques, writing review, and editing; N.S.: carried out animal experiments, writing review, and editing; M.J.M.-A.: study design, funding acquisition, writing review, and editing; P.G.-M.: contributed to study design, funding acquisition, supervision, writing—original draft preparation, review, and editing. All authors contributed to the article and approved the submitted version.