Age-Related Changes in Intestinal Immunity and the Microbiome

Abstract

The gastrointestinal (GI) tract is a vitally important site for the adsorption of nutrients as well as the education of immune cells. Homeostasis of the gut is maintained by the interplay of the intestinal epithelium, immune cells, luminal antigens, and the intestinal microbiota. The wellbeing of the gut is intrinsically linked to the overall health of the host, and perturbations to this homeostasis can have severe impacts on local and systemic health. One factor which causes disruptions in gut homeostasis is age, and recent research has elucidated how critical systems within the gut are altered during the aging process. Intestinal stem cell proliferation, epithelial barrier function, the gut microbiota, and the composition of innate and adaptive immune responses are all altered in advanced age. The aging population continues to expand worldwide, a phenomenon referred to as the “Silver Tsunami,” and every effort must be made to understand how best to prevent and treat age-related maladies. Here, we review recent research about changes observed in the intestinal epithelium, the intestinal immune system, the microbiota, and how the aging gut interacts with and influences other organs such as the liver, lung, and brain. Better understanding these age-related changes and their impact on multi-organ interactions will aid the development of therapies to increase the quality of life for all aged individuals.

Graphical Abstract

Summary sentence:

Review of how advanced age alters the gut, the microbiome, and how these alterations impact overall health

Introduction

The global population is aging. According to the World Health Organization, 8.5% of the population is >65 years old, and this proportion is expected to nearly triple by 2050¹. This massive increase in the aged population will likely coincide with an increased strain on hospitals and assisted living facilities. Gaining a better understanding of how advanced age impacts overall health will allow for targeted therapies to alleviate some of this burden. One arena in which age is appreciated to have a large impact is the health of the gut. The intestine plays a critical role in health and wellbeing^{2, 3}, as well as critical illness⁴. Advanced age negatively impacts epithelial barriers⁵. In the gut, this involves exhaustion of intestinal stem cells⁶ along with aberrant proliferation and differentiation,⁷ causing an age-dependent delay in the recovery of this organ after injury. Research into non-invasive biomarkers of gut barrier dysfunction⁸ has allowed investigators to conduct studies in patients utilizing blood and feces to monitor gut barrier integrity and intestinal function⁹, and develop treatments to restore the barrier¹⁰. Intestinal fatty acid binding protein (iFABP) is a gut-specific bio-marker for intestinal epithelial damage that can be readily measured both in the blood and urine following intestinal ischemia¹¹ and necrotizing enterocolitis^12–14. Glutathione S-transferases in the blood and urine may also indicate intestinal epithelial damage¹⁵ but may also indicate damage to liver and kidneys⁹. Lipopolysaccharide (LPS) measured in the circulation indicates an increase in microbial translocation, but these tests are highly sensitive and prone to false-positives¹⁶. Endotoxin core antibody assays are also used as an indirect measure of increased bacterial translocation¹⁷, as is sCD14, which indicates an inflammaging-associated increase in monocyte activation^{18, 19}. iFABP, LPS, and sCD14 are all increased in advanced age in humans²⁰. Pro-inflammatory cytokines, such as IL-6, IL-15, and IL-8, and C-reactive protein, are also increased in advanced age as reviewed elsewhere²¹. Taken together, these findings describe a correlation between increased age-related inflammation and decreases in intestinal barrier function.

Along with the physical barrier of tight junctions that connect epithelial cells in the intestine, there are also chemical barriers generated by antimicrobial peptides (AMPs)²². These peptides are evolutionarily conserved, natural antibiotics produced by immune and epithelial cells in the gut. One AMP, regenerating islet-derived protein 3-gamma, or Reg3γ, is thought to prevent microbiota from invading intestinal epithelial cells²³ by creating spatial segregation of microorganisms within the gut²⁴. Reg3γ carries out its protective, bactericidal functions by binding to peptidoglycan in the cell wall of gram-positive bacteria²⁵ and forming a hexameric membrane-permeabilizing oligomeric pore²⁶. When produced in appropriate amounts by Paneth cells in the intestinal crypts of the small intestine, Reg3γ helps maintain homeostasis of the microbiome²³. Interestingly, AMP upregulation is regulated by cytokines and the expression of multiple AMPs is altered with advanced age²⁷. Moreover, evidence in elderly humans and in rodent models of aging show that even in the absence of injury there are dramatic changes in fecal microbiota relative to younger subjects^{28, 29}(discussed in greater detail below). Although there may be significant changes to the microbiota in age, these changes in specific gut microbial populations do not establish causation of disease. These observations reveal that multiple intestinal parameters are altered with age. Despite recent research efforts, our current understanding of the mechanisms behind these changes in the aging gut is still lacking. Here, we review age-related changes to the intestinal epithelium, immune cells within the gut, and how age impacts the interactions between the gut and distant organ systems such as the liver, lung, and brain.

I. Alterations in intestinal epithelial cells in aging

The intestinal epithelium is a single cell layer that serves as a physical barrier separating the microbiota and other luminal contents from the intestinal tissue. Intestinal epithelial cells (IECs) are connected by intercellular tight junction proteins such as occludin and zona occludens-1 (ZO-1), which serve to regulate the migration of materials between the cells³⁰. Maintenance of the intestinal barrier function is vital to the health of the host³¹. The compromise of this barrier coincides with increased prevalence of damage-associated molecular patterns (DAMPs), pathogen-associated molecular patterns (PAMPs)³², and microbe-associated molecular patterns (MAMPs)³³ in the intestine, all of which are associated with chronic immune activation, which has broad local and systemic consequences. The compromise of the intestinal barrier is also known as ‘leaky gut³⁴.’ Locally, impairment of the intestinal barrier has been linked to inflammatory bowel disease (IBD)³⁵, colorectal cancer³⁶, celiac disease³⁷, and metabolic disorders like obesity and diabetes³⁸. Recently, advances in available technology, such as intestinal organoids^{39, 40}, have allowed researches to interrogate the effects of aging on the intestinal barrier, as reviewed previously⁴¹.

Intestinal stem cells

All subtypes of the intestinal epithelium are derived from intestinal stem cells (ISCs) that reside in intestinal crypts. ISCs reside either in the crypt base as crypt base columnar cells, or at the +4 position from the base of the crypt as +4 label retaining stem cells, which can repopulate the crypt base columnar cells should they be lost due to damage⁴² (Figure 1). ISCs asymmetrically divide to give rise to two daughter cells, a nascent ISC and a rapidly dividing and maturing cell that occupies the so-called transit amplifying compartment. The ISC and their progeny enable the intestinal epithelium to renew itself every 3–5 days⁴³. As the transit amplifying cells divide and mature, most intestinal epithelial cells travel up the “epithelial conveyer” of the crypt, differentiate, and eventually are sloughed off into the lumen⁴⁴. As the host ages, however, the capacity for proliferation and self-renewal of ISCs is greatly diminished, as shown by decreased growth of intestinal organoids derived from old mice when compared to young mice⁶. In addition to decreased growth, ISCs in older mice also highly express a pro-apoptotic gene profile, indicating decreased survival of these cells⁶. As a consequence of decreased growth and survival, aged mice also exhibit a decreased ability to heal from experimentally induced intestinal damage⁴⁵.

Figure 1. Alterations in the intestinal epithelium and immune cells with advanced age.

The intestinal epithelium is arranged into crypts and villi. Crypt base columnar stem cells reside at the base of the crypt and asymmetrically divide to give rise to stem cells and rapidly dividing cells that make up the transit amplifying region, which differentiate as they move up the crypt. The +4 label retaining, or “reserve,” stem cells reside at the boundary between the crypt base columnar stem cells and the transit amplifying region and can repopulate the crypt base columnar stem cells if they are lost. As the host enters advanced age, a multitude of changes occur, leading to overall increased inflammation.

GIP: Glucose-dependent insulinotropic polypeptide/gastric inhibitory polypeptide

ILC : Innate lymphoid cell

IgA : Immunoglobulin A

The self-renewal capacity of ISCs is maintained by a gradient of canonical Wnt proteins that increases at the base of the intestinal crypts^{46, 47}. This Wnt gradient is provided by subepithelial mesenchymal cells^{48, 49} and functions to maintain expression of stem-cell markers in the ISC niche⁵⁰. The loss of this canonical Wnt gradient at the base of the crypt leads to a complete ablation of intestinal epithelium⁵⁰, highlighting the vital role of intestinal Wnts. Although the exact mechanisms are not yet known, canonical Wnt proteins, including Wnt3a, are decreased in ISCs, Paneth cells, and subepithelial mesenchymal cells in aged mice when compared to their younger counterparts⁵¹. Further, supplementing intestinal organoid cultures derived from aged mice and humans with Wnt3a restored their growth capacity to that of organoids generated from younger subjects⁵¹. One explanation for the altered expression of Wnt and Wnt targets is age-related epigenetic changes. One study using murine organoids derived from the small intestine of aged mice found that the stem cell marker lgr5 was epigenetically silenced by histone H3 lysine 27 trimethylation, causing a decrease in cell proliferation within those organoids⁵². Another study used long term culture of murine colonic organoids to mimic aging and found that the organoids underwent multiple epigenetic changes that led to activated Wnts and mirrored a pro-cancer phenotype⁵³.

As the epithelia proliferate, they migrate to the top of the crypt and become exposed to decreasing concentrations of canonical Wnt and increasing levels of soluble factors called bone-morphogenetic proteins (BMPs)^{54, 55}. BMPs function to inhibit the expression of canonical Wnts, allowing for epithelial cell differentiation and maturation into a number of different cell lineages with important functions in addition to being a physical barrier.

Absorptive enterocytes

Absorptive enterocytes are the most abundant of the IECs. These cells express catabolic enzymes on their luminal surface to digest a variety of molecules, including water, ions, and nutrients, to allow for cell uptake⁵⁶. Enterocytes also express Toll-like receptors (TLRs) 2, 3, 4, 5, and 9^{57, 58} that, when engaged, initiate production of cytokines and chemokines that activate nearby immune cells⁵⁷. In aged individuals, however, there is an altered production of cytokines by the intestinal epithelium. Notably, when human epithelial biopsies taken from the ileum of elderly donors were exposed to flagellin, CXCL8 production was decreased⁵⁹, although there was no notable age-related reduction in TLR5. CXCL8 is an important leukocyte chemoattractant produced by intestinal epithelium in response to bacterial entry⁶⁰, and its reduction in the aged ileum suggests a decreased ability to respond to bacterial infection. Enterocyte TLR engagement also stimulates production of soluble mediators that exert antimicrobial effects, including iNOS⁶¹ and β-defensin⁶². The production of β-defensin, along with other C-type lectins Reg3β, Reg3γ, angiogenin, and resistin-like molecule beta (Relmβ) are significantly up-regulated in the ileum of aged mice⁶³. Reg3β and Reg3γ are constitutively expressed in the intestines and display antimicrobial activity when upregulated in response to bacterial sensing^64–66. Angiogenin, a potent stimulator of angiogenesis, acts within the intestines to promote IEC survival and proliferation⁶⁷. Relmβ maintains IEC barrier function and regulates expression of Reg3β and Reg3γ⁶⁸. Increased levels of these C-type lectins in aged mice suggest a constant stress response and an attempt to restore the intestinal barrier. Enterocytes also express MHC class II, allowing the intestinal epithelium to act as nonconventional antigen presenting cells⁶⁹. Absorptive enterocytes in the duodenum of healthy, elderly human donors were shown to have increased levels of apoptosis and proliferation⁷⁰, perhaps contributing to an impairment of their usual functions.

Goblet cells

Goblet cells are specialized mucus-secreting IECs that reside in both the small and large intestines⁷¹. The mucus layer is composed primarily of Muc2, a highly O-glycosylated protein produced by goblet cells. The mucus layer of the small intestine forms a single, loose layer that extends from the tips of the intestinal villi down to the base of the crypts⁷². The colon, which houses a much higher density of luminal microbes than the small intestine, has a higher number of goblet cells, resulting in two distinct mucus layers. The inner, firm mucus layer is comprised of polymerized, microbe free Muc2 that is anchored to the epithelium⁷³. Muc2 associated with the inner mucus layer is proteolytically processed by the host and microbiota to produce the outer, loose mucus layer that houses a variety of microbial species. The mucus layer is part of a larger structure called the glycocalyx, which is a filamentous mesh of glycolipids and glycoproteins that forms a barrier between the epithelium and bacteria⁷⁴. How aging impacts the gut mucus layer appears to differ between spatial regions of the small intestine. The gastric and duodenal mucus layer is not significantly different between young and old human subjects⁷⁵ but was slightly thicker in the ileum of aged mice, coinciding with an increase in goblet cells⁷⁶. Despite this increase in goblet cells and mucus production in the ileum of aged mice, there is an increase in epithelial associated bacteria, indicating a decline in the protection offered by the mucus layer⁷⁷. The most striking difference in mucus production in mice is seen in the colon, where there is a stark decline in the thickness of the mucus layer^{77, 78} and number of goblet cells⁷⁷, which paralleled broad changes in the expression of immune cell markers and the composition of the microbiota⁷⁹. Interestingly, male mice are more susceptible to age-related reductions in colonic mucus thickness than female mice⁷⁹, perhaps leading to increased susceptibility to dysbiosis and related inflammation.

Enteroendocrine cells

Enteroendocrine cells are another subtype of IECs that secrete a variety of intestinal hormones such as cholecystokinin and glucagon-like peptides (GLP) −1 and −2 that help to control the digestive function of the gut⁸⁰. These intestinal hormones also regulate the production of the neurotransmitter serotonin⁸⁰. In addition to these roles, cholecystokinin also regulates differentiation⁸¹ and cytokine production of CD4+ T-cells⁸² and B cells⁸³. In aging mice, there is an increased number and activity of K cells, a specialized enteroendocrine cell population that secretes glucose-dependent insulinotropic polypeptide/gastric inhibitory polypeptide (GIP), a molecule that increases fat accumulation⁸⁴. In the ileum of older humans, there is an increase in the number of enterochromaffin cells⁸⁵, which primarily secrete serotonin⁸⁰. This increase is theorized to be a compensatory mechanism to attempt to rescue age-related attenuation in afferent nerve sensitivity⁸⁵.

M cells

Microfold or M cells are specialized epithelial cells in the small intestine that allow for immune cell sampling of luminal content antigens by closely associated antigen-presenting cells. These antigen-presenting cells then deliver acquired luminal antigens to the Peyer’s Patches (PP) in the small intestine and are critical for the induction of antigen-specific immuno-globulin A (IgA) production⁸⁶. The number of mature M cells in the follicle-associated epithelium in PP of aged mice are significantly reduced, leading to compromised uptake of particulate antigen from the gut lumen⁸⁷. Additionally, there is a marked reduction in antigen-specific antibody production by intestinal B cells, reduced T-cell proliferation, and reduced antigen specific T-cell cytokine production in aged mice⁸⁸, which parallels the reduction in M cell numbers.

Paneth cells

Paneth cells are another specialized epithelial cell in the small intestine that produce AMPs and have granules containing IL-17A^{44, 89} and Wnt3⁴⁶. While Paneth cells may be seen during idiopathic inflammatory bowel disease in the colon, they are generally localized to the small intestine⁹⁰. Aged mice have decreased small intestinal expression of lysozyme, a Paneth cell marker⁷⁷, indicating a possible reduction in their numbers. However, Paneth cells from aged mice have increased production of Notum, a Wnt inhibitor that impairs regeneration of the aged intestinal epithelium⁹¹.

II. Age associated changes in gut immune cells

As mentioned above, the gut is a vitally important site for the education of immune cells⁹². Immune cell education within the gut is dependent on the presence of microbes⁹², a bolus of which neonates receive during vaginal birth⁹³. The microbial community remains relatively unstable in humans until around 2 or 3 years of age^94–96. During this period of time, microbial stimulation of immune cells supports the development of PP, mesenteric lymph nodes, and isolated lymphoid follicles⁹⁷. Further, this microbial stimulation promotes the maturation and recruitment of B and T cells into PP and the lamina propria (LP), a loose connective tissue layer that resides under the epithelium⁹⁸. In the ileum and colon of aged mice, however, there is a significant reduction in innate and adaptive immune genes, including tlr4, cd3ε, cd4, and cd8⁷⁷, indicating a reduction in overall T cells in the aging intestine. In addition to these alterations, neutrophil recruitment in response to Clostridium difficile infection is markedly reduced in middle-aged mice⁹⁹, leading to increased mortality. Innate lymphoid cells (ILCs) are also an important cell subset within the intestines of mice and humans and play a critical role in the maintenance of barrier function and the initiation of immune responses against pathogens¹⁰⁰. In aged human donors, it was discovered that ILC3s, the ILCs that phenotypically mirror Th17 cells, are decreased in the intestine when compared to young donors¹⁰¹. In aged mice, however, the major difference in intestinal ILCs appears to be an increase in ILC2s¹⁰², the ILCs that phenotypically mirror Th2 cells.

Gut microbes induce the development of gut-resident T-regulatory cells (Tregs) in the mesenteric lymph nodes that are vitally important in establishing oral tolerance to ingested food antigens and commensal microbes¹⁰³. Likewise, antigens from the gut microbiota promote the differentiation of B cells into IgA-producing plasma cells¹⁰⁴; secretory IgA promotes gut homeostasis through a number of mechanisms including bacterial disruption, neutralization of bacterial toxins, and blocking the bacterial invasion of the lumen^105–107. Aged mice display a drastic reduction in intestinal IgA expression⁷⁷, contributing to the attenuation of its protective effects.

The microbiota also contains potentially pathogenic bacteria that can contribute to gastrointestinal inflammation and perpetuation of IBD if they become too numerous, or are able to travel into the tissue where they are sensed by the immune system¹⁰⁸, perhaps as a consequence of disrupted barrier function in aged individuals. The outgrowth of opportunistic pathogens, or microbes, which may be beneficial in normal abundance that can become pathogenic when overgrown, within the gut microbiota correlates with autoimmune disease outside of the intestines. For example, the outgrowth of Prevotella copri correlates with increased susceptibility to auto-reactive T-cell mediated rheumatoid arthritis^{109, 110}. Further, the gut microbiome is required for the induction and progression of mouse experimental auto-immune encephalomyelitis (EAE)¹¹¹. The composition of microbiota in the gut modulates the effectiveness of a variety of cancer immunotherapies through regulating the differentiation of T cell subsets^112–115.

III. Aging and the gut microbiome

Humans and commensal microbial communities have co-evolved over millions of years, developing a complex and mutually beneficial metabolic dialogue that regulates physiological processes and maintains homeostasis. The importance of this relationship is perhaps most evident in the lower GI tract, where bacteria play roles in nutrient absorption, immune function, synthesis of essential vitamins, drug processing, circadian rhythms, and insulin signaling, among other critical functions¹¹⁶. Impressive numbers help to illustrate the scale of this interdependence: It is generally estimated that the human gut is home to 10¹³–10¹⁴ microorganisms, outnumbering human cells in the entire body. Together, they weigh approximately 1–2 kg, similar to organs like the liver and brain, and contain over 100 times as many genes as in the human genome^{117, 118}.

As humans age, the host-microbiome relationship undergoes changes that have important implications for frailty and disease¹¹⁹. In most cases, these changes are thought to reflect dysbiosis – imbalances in microbial species and a reduction in overall microbial diversity that is detrimental to host fitness. However, given that gut bacteria do not age in the same way as host organs and the numerous variables that influence their relative abundance, a fundamental question about this relationship arises. Do changes in the gut microbiome contribute to phenotypic changes in the host over time, or vice versa? In other words, it is still unclear whether specific age-related changes in microbiota occur independently, as a maladaptive response to other age-related physiologic changes in the host, or as a beneficial, compensatory response¹²⁰. Answering this question and identifying mechanistic links to disease and overall functional decline are current areas of research focus. While it is difficult to isolate age-dependent effects, longitudinal human studies and animal models utilizing fecal microbiota transplant (FMT) have been useful. Ultimately, this field seeks to identify ways of manipulating the gut microbiome to buffer the normal effects of aging, reduce age-related disease burden, and extend the human lifespan.

Spatial variation of the human gut microbiome in form and function

The gut microbiome is dominated by bacteria, but also includes fungi, viruses, and archaea. Bacteria belonging to the phyla Firmicutes and Bacteroidetes are the most abundant, accounting for 80–90% of the total microbiota⁹⁴. Proteobacteria, Actinobacteria, Fusobacteria, Cyanobacteria, and Verrucomicrobia are less abundant. At the genus, species, and strain level, significant variation exists between individuals, influenced by several environmental and genetic factors.

Microbiome composition and function also vary based on location in the lower GI tract. Distinct spatial niches are created by proximal-to-distal gradients of oxygen, dietary nutrients, and pH, leading to colonization with bacteria engaging in distinct activities¹²¹. The small intestine mainly contains facultative anaerobes that compete with host epithelial transporters for dietary nutrients. The large intestine has far greater microbial diversity, density, and host-microbiome interactions. The majority of colonic bacteria are obligate anaerobes that metabolize insoluble complex carbohydrates, generating absorbable, high-energy short-chain fatty acids (SCFAs), essential amino acids, and vitamins, including B12 and folate, and secondary bile acids¹²². SCFAs play crucial roles in gut health, as discussed later in this section, but are produced in lower quantities and variable ratios in elderly versus young individuals^{28, 123}. Colonic bacteria also play a role in the activation of satiety pathways¹²⁴, suggesting that age-related changes may play a role in chronically reduced appetite seen in the elderly.

Age-related dysbiosis: changes in microbiome composition in the elderly

Microbial colonization of the GI tract starts at birth. In infants and children, method of delivery, breastfeeding, and antibiotic exposure can all influence the microbiome, with effects lasting into adulthood^{125, 126}. The gut is initially colonized by facultative anaerobes, which creates a reduced environment appropriate for later colonization with obligate anaerobes¹²⁶. Microbiome diversity continues to increase until approximately the age of 3 years, after which, it is relatively stable, with variations attributed to environmental factors such as diet or antibiotic exposure⁹⁴. However, as individuals age, a mirrored process of senescence occurs, leading to declines in microbial diversity and stability¹²⁷.

The elements that separate age-related dysbiosis from a ‘healthy’ gut microbiome are somewhat controversial, though overarching patterns have been identified by human population studies. Key phenomena include compositional instability, reduced overall diversity, and an increase in pro-inflammatory opportunistic pathogens^{128, 129}, linked to declining immune function,²⁸ and increased frailty in elderly populations^{119, 130}. Specific trends include an increase in facultative anaerobic bacteria and a decrease in SCFA-producers in the colon¹²⁷. In addition, some phylum and genus-level patterns have emerged in association with health, disease, and lifespan. Dysbiosis and clinical disease have generally been associated with an increase in Proteobacteria¹²⁹, a phylum containing pathogenic bacteria associated with intestinal inflammation^{129, 131}, and a corresponding decrease in Firmicutes. Pathogenic bacteria are normally kept in check by commensal microbes and AMPs secreted by both commensals and host cells. One such mechanism involves intestinal expression of Reg3β, an AMP that plays a role in maintaining homeostasis of the gut microbiome as described above^{23, 24, 132}. Changes in the relative abundance of Bacteroidetes are much more variable, with some studies reporting higher levels in elderly subjects,^{119, 128, 133} and others reporting lower levels^{123, 134–137}. Studies that observed an age-dependent increase in Bacteroidetes also noted a corresponding decrease in Firmicutes^{119, 128, 133}.

Investigators have also noted trends in less abundant bacterial populations with age. A seminal study profiled the gut microbiomes of 161 elderly individuals using 16S rDNA sequencing and found that the genera Bacteroides, Alistipes, and Parabacteroides comprised 8–27% of the microbiome in a younger cohort, but more than 50% in those over the age of 65 years¹¹⁹. Of note, there was also relatively greater inter-individual, genus-level variability seen in the elderly gut microbiome. In addition, several studies have shown that the abundance of Bifidobacteria, which are generally believed to exert gut health benefits, decreases in the elderly^{123, 138–141}. As with Bacteroidetes, investigators have reported variable results for bacteria belonging to the genus Lactobacilli. Some groups report increases in old age^{123, 136}, while others report decreases¹³⁹. These differences can partially be explained by significant inter-individual variations in diet, lifestyle, and environmental exposures. Technical differences in measuring relative microbial abundances between studies may also be a contributing factor¹²⁷. Ongoing advances in sequencing and metagenomic technology will undoubtedly enhance the accurate and reproducible measurement of microbial species abundance and allow us to gain a better understanding of aging and the microbiome.

One approach used to study connections between age and the microbiome is to examine changes in gut flora at the limits of age. Centenarians and supercentenarians (those who have reached or surpassed the ages of 100 and 110 years, respectively), have been the focus of several investigations attempting to identify factors contributing to extreme longevity, a surrogate measure of health. The lower GI tracts of these individuals appear to be colonized by a disproportionately high number of health-associated bacterial populations, such as SCFA-producers^142–144. Biagi et al. examined the fecal microbiome of young adults (22–48 years old), elderly adults (65–75 years old), and semi-supercentenarians (105–109 years old) in an Italian population¹⁴². They found that the fecal microbiota in all age groups was dominated by just three families – Bacteroidaceae, Lachnospiraceae, and Ruminococcaceae. However, the relative abundance of bacteria belonging to these families decreased with age. In other words, extreme aging in humans was associated with an increase in certain other bacterial genera, including Bifidobacterium, Akkermansia, and Christensenellaceae. One possible explanation for this shift is that these bacteria may promote healthy aging. Another striking finding was that the overall difference in microbiome composition between elderly individuals and centenarians appear to be greater than that between young adults and elderly adults, even though the latter two groups were separated in age to a greater degree.

Signatures of age-related disease in the microbiome

Many clinical diseases, some of which are strongly associated with advanced age, have distinct gut microbial signatures. Advanced age is an important independent risk factor for type 2 diabetes (T2D)¹⁴⁵ and a substantial body of literature describes the role of the gut microbiome in its pathophysiology. Reports somewhat vary with respect to the association between T2D specific taxonomic groups, but Gurung et al. recently reviewed 42 human studies on the topic and identified trends. Commonly-reported findings included a negative association between T2Dand the genera of Bifidobacterium, Bacteroides, Faecalibacterium, Akkermansia, and Roseburia, and a positive association with Ruminococcus, Fusobacterium, and Blautia¹⁴⁶. A recent cohort study in Northern China involving 16S rRNA sequencing revealed a significant decrease in overall gut microbiota diversity and butyrate-producing bacteria, such as Bifidobacterium and Akkermansia in diabetic patients compared with healthy controls¹⁴⁷. Colorectal cancer is another example. On average, it presents in the 6^th and 7^th decades of life, and the gut microbiome appears to be involved in GI tumorigenesis. Animal models have helped identify associations with Fusobacterium nucleatum, Bacteroides fragilis, Streptococcus bovis, Heliobacter pylori, Enterococcus faecalis, and certain strains of Escherichia coli^148–150. Furthermore, recent data suggest that the microbiome composition affects the efficacy and toxicity of cancer immunotherapy, opening the door for synergistic therapies that target gut microbiota¹⁵¹. The association between age-related disease and taxonomic patterns in the gut flora has other implications for therapeutics¹⁵². In brief, it is possible that the gut microbiota could be used in the future as biomarkers of specific diseases, and that modulation of the microbiome through FMT or antibiotic, probiotic, or prebiotic therapies could one day be used as a means to prevent or treat disease in individual patients.

In addition to identifying patterns of dysbiosis in association with specific age-related diseases, researchers have found that certain microbial signatures are associated with functional frailty and general ill-health in the elderly^{28, 153}. Even after adjusting for confounding variables like age and medications, trends in microbiota composition correlate with clinical measures of frailty¹⁵⁴. Claesson et al. reported that the abundance of bacteria belonging to the Oscillibacter and Alistipes genera and Eubacteriaceae family, as well as the loss of Lactobacillus and Faecalibacterium, is associated with frailty in elderly people²⁸. These individuals also appear to have fewer producers of beneficial SCFAs and polyamines when compared to healthy elderly people¹⁵⁵. Some factors common to aging-related dysbiosis, frailty, and disease states include inflammation¹³⁷ and suboptimal diet¹⁵⁶. It has been established that consumption of fiber-rich foods decreases in the elderly¹⁵⁷ and that a year-long Mediterranean diet intervention can alter the gut microbiome in older individuals, reduce frailty, and improve overall health status¹⁵⁸.

It is important to note that human studies that recognize associations between specific gut microbial populations and aging or disease do not establish causation. It is still largely unclear whether all changes in the microbiome associated with age represent dysbiosis or whether some are simply adaptive responses to other age-related physiologic or behavioral changes. Likewise, it is uncertain whether dysbiosis identified in association with various disease states is a cause or consequence. In an effort to establish mechanistic links, investigators have used animal models of aging with shorter lifecycles and less complex microbiomes than humans, including C. elegans, Drosophila species, zebrafish, and murine strains. Gnotobiotic models, in particular, have helped to elucidate the contributions of specific microbiota components in regulating host metabolism¹¹⁶. A comprehensive review of animal models of aging and host-microbiota interactions is beyond the scope of this paper but can be found elsewhere^{131, 133, 159}.

Regarding mechanism, some investigators have suggested that the host immune system develops an inflammatory response to certain commensal bacteria over time. This is sometimes seen after GI infections, in which microbiota-specific T cells are activated and go on to form memory cells, leading to the loss of mucosal immune tolerance to beneficial commensals¹⁶⁰. Chronic inflammation resulting from this process could then lead to intestinal degradation and a higher risk of age-associated diseases¹⁶¹. Another plausible mechanistic link involves insulin-like growth factor 1 (IGF1)¹⁶². Yan et al. found that the transfer of pathogen-free microbiota into germ-free mice induces high levels of IGF1 and that supplementation with beneficial SCFAs leads to the upregulation of IGF1¹⁶². The growth hormone - IGF1 axis is involved in the activation of canonical aging pathways,^{163, 164} suggesting a connection between the gut microbiome and aging.

Role of the microbiome in intestinal barrier function and ‘inflammaging’

As noted above, one of the most important functions of the intestinal epithelium is maintaining a barrier to gut luminal content, while allowing for selective absorption of dietary nutrients and innate and adaptive immune response. Commensal gut bacteria help maintain both cellular and junctional components of the barrier and regulate tissue immunity. Hence, one of the key consequences of aging-related dysbiosis is a ‘leaky gut’ and subsequent inflammation, which contributes to further dysbiosis in a feed-forward loop (Figure 2)^{131, 165, 166}.

Figure 2. The ‘leaky gut’ model of aging and inflammation.

Age-related intestinal microbial dysbiosis leads to reductions in commensal bacteria-derived short-chain fatty acids (SCFAs), increases in lipopolysaccharide (LPS) production, and decreased secondary bile acid production. Loss of SCFAs results in degradation of the intestinal barrier, leakage of bacterial products and other pro-inflammatory luminal and mucosal debris, leading to inflammation and further dysbiosis in a feed-forward cycle.

TNFα: tumor necrosis factor alpha

CRP: C-reactive protein

IL-6, IL-1β, IL-18: Interleukin-1, Interleukin-1 beta, Interleukin-18

The ‘leaky gut’ hypothesis describes a process in which advanced aging, malnutrition, chronic stress, and other harmful factors lead to changes in the microbial populations of the gut. These changes are invariably detrimental and often include reductions in the relative abundance of SCFA-producing bacteria. SCFAs, including butyrate, acetate, and propionate, are generated by bacterial fermentation of dietary fibers and resistant starch. They have roles in immunomodulation, energy balance, and have other far-reaching systemic effects but are also necessary for the maintenance of local gut integrity^167–169. For instance, butyrate enhances the intestinal barrier by facilitating tight junction assembly and acting as the primary fuel source for colonocytes^{170, 171}. The main butyrate producing-bacteria in the human gut belong to the phylum Firmicutes, in particular Faecalibacterium prausnitzii, Clostridium leptum, Eubacterium rectale, and Roseburia species¹⁷¹. As the intestinal barrier breaks down, the epithelium allows unregulated passage of luminal content into the bloodstream, including bacterial cells and fragments, such as bacterial endotoxins, and dietary metabolites. This results in immune activation, inflammation, and even autoimmunity^{171, 172}. The contribution of aging to this process is evidenced by studies showing higher levels of intestinal permeability and epithelial pro-inflammatory cytokine levels in older patients compared to younger counterparts⁵⁹. Ultimately, local inflammation drives further dysbiosis. Beneficial lifestyle factors, such as high-fiber diets, caloric restriction, and exercise, as well as endogenous anti-inflammatory mechanisms, can provide resistance to this feed-forward cycle.

In addition to decreasing levels of SCFAs, age-related dysbiosis can lead to altered levels of other beneficial bacterial metabolites, including secondary bile acids and essential vitamins that are important for immune function. It can also result in higher levels of LPS. In particular, LPS is produced by Enterobacteriaceae, a family of Proteobacteria, whose relative abundance increases in age-related dysbiosis. LPS is recognized by host TLR-4 receptors and results in an increased local expression of inflammatory cytokines¹⁷³. In combination with a ‘leaky gut,’ this can lead to endotoxemia and systemic inflammation. Finally, AMPs, including α-defensins, lysozyme C, phospholipases, C-type lectin, and Reg3γ, have a critical role in innate immunity and help to keep pathogenic bacteria in check. A subset of these AMPs – bacteriocins – are produced by commensal bacteria, presenting another mechanism driving the feed-forward dysbiosis cycles.

The link between age-dependent changes in gut microbiota and systemic inflammation has been established in animal studies in which germ-free mice co-housed with old, but not young conventionally raised mice, leads to increased intestinal permeability and inflammation¹⁷⁴. Gut epithelial barrier integrity has also been shown to decrease as a function of age in C. elegans¹⁷⁵ and Drosophila¹⁷⁶ models, as well as in humans⁵⁹.

Implications of age-related dysbiosis for disease and therapy

The connection between aging and gut microbiome has immense implications for human health. Just as enrichment in certain SCFA-producing bacteria is associated with longevity, decreases in overall microbiota richness appear to be a predictor of mortality in older populations¹⁷⁷. Other groups have suggested, for example, that the microbial dysbiosis-related inflammation and immune dysfunction, in particular deficits in immune surveillance, may contribute to increased cancer risk in the elderly¹⁵⁵. The unregulated leakage of LPS and other bacterial products increases the production of IL-1β, IL-6, TNFα, and interferons systemically, contributing to a chronic pro-inflammatory state in the elderly termed “inflammaging”^{155, 161}, as noted above. In this state, the ability of myeloid cells to remove dysplastic, senescent, apoptotic, or malfunctioning cells is impaired¹⁹, creating an environment conducive for various types of cancer. Alterations of the microbiome with age also influence the homeostasis of various gut-systemic axes. For instance, investigators have shown that gut microbiota are key regulators of brain development and neurodegeneration. Mediators of the gut-brain cross-talk include the vagus nerve, gut-derived hormones, gut-derived neurotransmitter precursors (e.g., tryptophan), or neurotransmitters themselves (e.g., dopamine, acetylcholine, and γ-Aminobutyric acid), and SCFAs^178–183. Imbalances in these mediators may explain why dysbiosis has been implicated in the pathogenesis of several age-related neuropsychiatric diseases, including inflammation-driven Alzheimer’s disease, schizophrenia, and depression.

Interventions aimed at avoiding or treating the myriad of age-related diseases associated with dysbiosis could take many forms. Fecal microbiota transplant, the transfer of stool or stool-derived microbiota from one individual to another, has been used with significant success in the treatment of opportunistic Clostridium difficile infections and its clinical indications are likely to expand in the future, to include inflammatory and autoimmune diseases^{184, 185}. Barcena et al. showed that in a mouse model of progeria with prominent dysbiosis, FMT from wild type mice enhanced healthspan and overall lifespan and restored secondary bile acids¹⁸⁶. A future challenge in this field will be to develop and test mixtures of engineered microbiota that could emulate the microbiome of young, healthy individuals following FMT. Prebiotics, namely fibrous substrates for bacterial SCFA production, probiotics, and selective antibiotics, have also been studied extensively, though with relatively less success. It is worth noting that antibiotics cause significant changes in microbiota composition¹⁸⁷, and a study that evaluated 2007–2009 Medicare Part D data found that patients aged ≥65 years used more antimicrobials, at 1.10 per person per year, compared to 0.88 antimicrobials used per person per year in patients aged 0–64 years¹⁸⁸.

Caloric restriction and other targeted dietary changes may be the most straightforward approach to mitigate dysbiosis as individuals age¹⁸⁹. Avoidance of a ‘western’ high-fat diet, processed foods, and red meat, in favor of a Mediterranean diet, high in whole grains and polyunsaturated fats have shown promise^{190, 191}. Zhang et al. showed that a long-term low fat, 30% calorie restricted-diet in mice led to enrichment in Lactobacillus with age, decreased LPS levels, and prolonged lifespan¹⁹². Finally, elderly individuals exhibit a phenomenon called ‘anabolic resistance,’ in which greater amounts of dietary protein are required to stimulate muscle synthesis and maintenance of lean body mass, and the aging microbiome is one of many factors implicated in this. Changes in the gut microbiota can impact the bioavailability of dietary amino acids, and changes in the quantity and type of ingested protein, often seen in the elderly, can alter the relative abundance of bacterial genera, especially protein-fermentors^{189, 193–195}. In particular, plant protein is associated with higher proportions of Bifidobacterium, Lactobacillus, and Roseburia, while animal protein is associated with more Bacteroides, Alistipes, Bilophila and Ruminococcus^{189, 196}. These findings suggest that directed dietary interventions could be used to buffer age-related dysbiosis.

Ultimately, dysbiosis is complex and likely both a result of and a response to inflammation and aging. When developing therapies, it is important to remember that humans are super-organisms comprised of human and microbial cells. Investigations should consider the possibility that not all age-related changes in microbiota are detrimental. Some may reflect adaptations to other drivers of aging, evolved, for instance, to optimize nutrient absorption or adjust the basal metabolic level.

IV. Interactions between aging organ systems

Gut-liver axis

The gut is capable of communicating with a number of organs, including the liver, lung, and brain, and it is likely that age-associated changes in the gut microbiome and intestinal barrier function also affect these organs. The gut-liver axis refers to the bi-directional communication between the gut, including its microbiome, and the liver. This pathway allows for critical communication between the two organs in response to dietary and environmental factors, as reviewed by Tripathi et al¹⁹⁷. The liver communicates with the gut by producing and releasing bile acids into the intestine via the biliary tract. In the gut, the intestinal microbiome metabolizes dietary components, bile acids, and other environmental factors that can travel to the liver via the portal tract and regulate a number of metabolic functions. However, when the intestinal barrier is compromised, increased translocation of bacterial derived MAMPs to the liver occurs. MAMPs can then bind to pattern recognition receptors (PRRs) on hepatic non-parenchymal cells, including Kupffer cells and hepatic stellate cells, leading to activation of pro-inflammatory and pro-fibrotic signaling cascades³³. Although the liver is a robust organ, a number of physiological changes have been shown to occur with aging, including increases in hepatic expression of the pro-inflammatory cytokines il6, tnfa and il1b, along with increased hepatic steatosis¹⁹⁸ and an increase in acute phase proteins¹⁹⁹. Interestingly, there is emerging evidence that these age-associated changes in the liver may be a result of impaired gut function. For example, supplementation with the gut brush border enzyme intestinal alkaline phosphatase in aged mice resolves not only aging-induced gut barrier dysfunction but also lowers age-associated increases in liver enzyme levels²⁰⁰. Furthermore, these observations suggest that increased leakage of LPS and other MAMPs from the gut contributes to the aged liver phenotype. Increased endotoxin levels are found in plasma from aged mice, and deletion of the LPS-binding protein (LPB) ameliorates the observed age-related liver inflammation²⁰¹. Importantly, the aged liver appears to be sensitized to LPS, as LPS-induced up-regulation of the pro-inflammatory IL-1β/inflammasome pathway is increased in the liver of aged rats²⁰². Whether or not impaired gut function and increased translocation of bacterial endotoxins contribute to age-related liver dysfunction in humans remains to be determined but warrants further study.

Gut-lung axis

The bi-directional link between the gut and lung has recently become more appreciated. The effects of the gut microbiome on lung immunity have been reviewed elsewhere^203–205. The gut and lung share many similar physiological characteristics at their mucosal surfaces, and it has been shown that immune cells readily traffic between these sites²⁰⁶. Further, the lung contains its own unique microbiome that affects many aspects of host immunity^{207, 208}, and there is microbial continuity along the aerodigestive tract²⁰⁹. Studies carried out in germ-free, or antibiotic-treated animals have shown that these animals have increased susceptibility and aberrant immune responses following infection with various respiratory pathogens as reviewed by Dumas et al.²¹⁰ Additionally, the function of alveolar macrophages, the tissue-resident phagocytes of the lung, is impaired in the absence of a gut microbiota²¹¹. Interestingly, one study found that the antibacterial function of murine alveolar macrophages can be restored to basal levels through the addition of NOD-like receptor agonists via the GI tract²¹¹. Furthermore, there is inflammatory cross-talk between the gut and lungs in diseases that primarily affect one or the other organs, such as Crohn’s disease (gut) and chronic obstructive pulmonary disease (COPD; lung)^{212, 213}.

In advanced age, heightened bacterial translocation believed to result from a ‘leaky gut,’ contributes to persistent inflammation and organ dysfunction, including pulmonary dysfunction. Lungs from aged mice and cells from elderly human donors showed sustained basal p38 MAPK activation and a delayed inflammatory response to LPS, which may contribute to age-related inflammatory lung injury²¹⁴. In addition, various chronic pulmonary conditions disproportionately affect the elderly, including COPD. COPD is the 3^rd leading cause of global death, with an estimated 3 million yearly mortalities. 92% of these fatalities are in people over 60 years of age, and 73% are in those over 70 years of age²¹⁵. Indeed, advanced age has become recognized as a clinical risk factor that may contribute to gut dysbiosis and the development of COPD (reviewed in²¹⁶). A proposed intervention to target the gut-lung axis and improve health outcomes associated with COPD is to increase dietary fiber²¹⁶. Given the importance of the gut-lung axis in health and disease, further studies that target the gut microbiome to treat chronic and acute respiratory conditions in elderly subjects are indeed merited.

Gut-brain axis

The gut-brain axis refers to the cross-talk between the central and enteric nervous systems, primarily dominated by signaling from the brain to the gut microbiota and vice versa, as reviewed elsewhere^{217, 218}. Importantly, the presence of gut microbiota is vital for the development and maturation of both the enteric and central nervous systems. For example, germ-free mice have impaired sensorimotor function²¹⁹ and memory formation²²⁰. Normal communication from the gut microbiota to the brain occurs through the vagus nerve, and severing the vagus nerve attenuates altered emotional behavior seen in mice infected with Lactobacillus rhamnosus¹⁷⁸. In addition to this normal cross-talk, mounting evidence from the past ten years has linked dysfunction in the intestinal epithelial barrier to a variety of neurological disorders. It is enticing to speculate that age-related increases in intestinal inflammation and permeability contribute to the pathogenesis of these neurological disorders. Occludin, an important intestinal tight-junctional protein, is reduced, and the normal distribution of zona occludens protein-1 is disrupted in the colonic epithelium of human patients with Parkinson’s Disease²²¹. Further, markers of intestinal inflammation and permeability, calprotectin, and alpha-1-antitrypsin/zonulin, respectively, were found to be increased in the fecal material of Parkinson’s patients²²², further implicating intestinal permeability in the pathogenesis of the disease. Gut leakiness has also been implicated in the pathogenesis of Alzheimer’s disease as previously reviewed by Köhler et al²²³, and multiple sclerosis as previously reviewed by Camara-Lemarroy et al.²²⁴. Finally, intestinal permeability is linked to depression and anxiety. Plasma markers of intestinal barrier dysfunction, zonulin, and iFABP, are increased in the plasma of those with depression and anxiety, which correlated with dysbiosis²²⁵. Whether these neurological disorders driven by alterations in gut permeability are further exacerbated by age warrants further study.

V. Possible systemic and/or clinical implications

Nutrient absorption, drug metabolism and efficacy

A common health condition that negatively impacts health and predicts premature mortality in the elderly is malnutrition^{226, 227}. While aging does not inevitably accompany malnutrition, malabsorption of nutrients does contribute to deficiencies in vitamin D, vitamin B₁₂, folate and anemia²²⁸. Clinical manifestations of malabsorption in the elderly, can be mild and the issue often goes undiagnosed²²⁹. The digestion and absorption of nutrients is largely controlled by the enteric nervous system (ENS), an integrated network of 200–600 million neurons within the myenteric and submucosal plexuses²³⁰. Importantly, the ENS modulates both the contractility and transit of material throughout the gut. Multiple studies have found enteric nerve fibers are negatively impacted with age and contribute to constipation and diarrhea^{229, 231}. Age-related changes of the ENS include neuronal loss²³² as well as neurodegeneration of nerve fibers^{233, 234}. Importantly, decreased innervation as well as degeneration of villi are thought to play predominate roles in malabsorption of material in the GI tract²²⁸.

An added consequence of aging is altered drug metabolism. The GI tract is a key component of first-pass metabolism and several studies have reported that drug clearance occurring via phase I metabolism typically decreases by 30–40% in individuals ≥65 years, while phase II metabolism is largely unaffected^{235, 236}. Collectively, altered metabolism can lead to prolonged drug half-lives and reduced drug clearance, but other confounding factors, including nutritional status and genetics, can also impact drug efficacy in the aged. Moreover, renal drug metabolism is highly impacted, leading to an approximate 50% reduction in drug clearance in the majority of the geriatric population²³⁶. Overall, adverse drug reactions are more frequent and have serious clinical implications in the aged²³⁷. Common prescriptions used in the elderly population include, but are not limited to, antibiotics, anticoagulants, digoxin, diuretics, hypoglycemics agents, antineoplastic agents, and non-steroidal anti-inflammatory drugs (NSAIDs) which collectively contribute to 60% of adverse drug reactions (ADR) leading to hospital admission and 70% of ADRs occurring post-admission²³⁷. Importantly, combination therapies may synergistically increase toxicity. For example, using NSAIDs increases the risk for peptic ulcers by 10% in aged individuals, while combination therapy of corticosteroids and NSAIDs increase the risk of peptic ulcers by 15 fold²³⁸. While drug metabolism largely takes place in the liver, the small intestine does express CYP450s, including CYP3A4, that eliminate a large proportion of drugs prior to reaching systemic circulation^{239, 240}. Intestinal expression of CYP3A4 does not appear to change with age in rodents²⁴¹; however, data are lacking on the expression CYP450 isoforms in the intestine of humans. As the gut microbiota are becoming increasingly recognized as an important contributor to xenobiotic metabolism^{242, 243}, and since the microbiome changes with age, it will be important for future studies to examine the complex interactions between the host, microbiota, and aging, and their relative contributions to drug metabolism.

Concluding remarks

As the global population continues to age, we are presented with an opportunity to conduct impactful research into treatments that will increase not only the length but the quality of life. Aging comes with a host of associated diseases that can impact every organ system, and recent scientific advancements have highlighted the aging gut as a focal point that interacts with and influences other organ systems. More focus should be placed on a better understanding of age-related alterations in the gut and how those changes can negatively impact overall health. However, the benefits of these studies would not be limited to increasing the quality of life in the aging population. Perturbations in gut barrier function can occur due to traumatic injury²⁴⁴, alcohol consumption²⁴⁵, combination burn injury and alcohol consumption^{246, 247}, and autoimmune diseases such as IBD. Lessons learned from the aging gut may yield novel therapies to treat or prevent systemic diseases linked to gut dysfunction in humans of all ages.

List of abbreviations:

AMP

Antimicrobial peptide

BMP

Bone-morphogenetic protein

DAMP

Damage-associated molecular pattern

EAE

Experimental auto-immune encephalomyelitis

FMT

Fecal microbiota transplant

GI

Gastrointestinal

GIP

Glucose-dependent insulinotropic polypeptide/gastric inhibitory polypeptide

GLP-1/2

Glucagon-like peptide 1/2

IBD

Inflammatory bowel disease

IEC

Intestinal epithelial cell

iFABP

Intestinal fatty acid binding protein

IgA

Immunoglobulin A

IGF1

Insulin-like growth factor 1

ILC

Innate lymphoid cell

ISC

Intestinal stem cell

LP

Lamina propria

LPS

Lipopolysaccharide

MAMP

Microbe-associated molecular pattern

NSAID

Non-steroidal anti-inflammatory drug

PAMP

Pathogen-associated molecular pattern

PP

Peyer’s patch

Reg3γ

Regenerating islet-derived protein 3-gamma

SCFA

Short-chain fatty acid

T2D

Type-2 diabetes

TLR

Toll-like receptor

TNFα

Tumor necrosis factor alpha

Treg

T-regulatory cell

ZO-1

Zona occludens-1