Microbiomes other than the gut: inflammaging and age-related diseases

The microbiota of the oral cavity is extremely diverse containing as many as 700 or more species [65‐68], of which the vast majority belong to the phyla Firmicutes, Actinobacteria, Bacteroidetes, Proteobacteria, Fusobacteria [68‐70], and Spirochaetes [68]. Bacterial colonization in the oral cavity and oropharynx occurs mainly on the lips, teeth, cheeks, subgingival and supragingival surfaces, hard and soft palate, and tonsils [68]. The Human Microbiome Project sampled many of these locations and found that, in most, the dominant genera were Streptococcus, followed by Haemophilus, Actinomyces, and Prevotella in the buccal mucosa (cheek), supragingival, and subgingival plaque, respectively [66]. It is thought that there is a “core microbiome” present in the majority of individuals which consists of Actinomyces, Atopobium, Corynebacterium, Rothia, Bergeyella, Capnocytophaga, Prevotella, Granulicatella, Streptococcus, Veillonella, Campylobacter, Cardiobacterium, Haemophilus, Neisseria, TM7, and Fusobacteria [68]. Compared with the gut, the relationship between the oral microbiome and aging is not as well studied [65]. Ogawa et al. analyzed the oral microbiome of elderly individuals living in a nursing home (EN) and those that live independently (control) and found the EN group was less diverse at the phyla level but not at the genus level. The EN group had a higher relative abundance of Actinomyces, Streptococcus, Bacilli, Selenomonas, Veillonella, Haemophilus, and a lower relative abundance of Prevotella, Leptotrichia, Campylobacter, and Fusobacterium compared with the controls [71]. Furthermore, Singh et al. analyzed oral microbiomes of healthy aging (HA) and non-healthy aging (NHA) individuals and found that HA had a higher alpha diversity than NHA. The only genus that was more abundant in HA of the overall most abundant genera was Neisseria. Haemophilus, Fusobacterium, and Capnocytophaga were all increased in HA, but were lower abundance genera. [72].

Inflammaging likely plays a role in the relationship between aging and oral microbiome [73]. During aging, many changes occur in the oral cavity that can lead to chronic inflammation, which can increase an individual’s susceptibility to oral disease [74]. Additionally, age is considered a risk factor for oral diseases, such as periodontal disease, which, in the USA affects around 60% of adult population [73]. Although a causative role has not been demonstrated yet, it is interesting to mention that periodontal pathology has been associated with atherosclerosis, suggesting that bacteria from the oral cavity may play a role in the onset of atherosclerosis and cardiovascular disease [75]. Feres et al. analyzed the subgingival microbiota within different age groups and found that in healthy individuals, there were no differences in the amounts of the analyzed taxa within the age groups but noted that the older adults (> 64) trended towards an increased abundance of three F. nucleatum ssp. However, in individuals with periodontal disease, the younger group (< 35) had an increased amount of the P. gingivalis and T. forsythia compared with the older groups. Furthermore, the oldest group had higher levels of four Actinomyces, especially Actinomyces naeslundii and Actinomyces oris [73].

While much is known about the nasal microbiome early in life, there is less research over the nasal microbiome in elderly individuals. Bomar et al. stated that one study found in elderly individuals the nasal microbiota has a high abundance of Streptococcus, and a study analyzing the nasal microbiota of elderly individuals with Parkinson’s disease (PD) found that in both PD and healthy controls, the composition of the nasal microbiota resembled that of the middle-aged adult. An additional study analyzing the nasal microbiota of elderly individuals found no difference in diversity between those living in a nursing home and those living independently. However, the relative abundance of Lactobacillus reuteri, Streptococcus, Staphylococcus epidermidis, and Rothia mucilaginosa were increased in individuals residing in the nursing home [76]. Koskinen et al. examined the relationship between the nasal microbiota and olfactory function. They found that Faecalibacterium and Porphyromonas strongly correlated with a reduction of olfactory function, and Corynebacterium members correlated with a reduction in odor discrimination and threshold. Interestingly, they also found that compared with the normal threshold scores, the individuals with lower scores had a more diverse microbiome [77]. Additionally, Rullo et al. characterized both the oral and nasal microbiome in newly diagnosed neovascular age-related macular degeneration (AMD) and healthy controls. In the oral microbiome, Propionibacteriales, Rothia, Staphylococcus, and Cornyebacteriaceae were increased in AMD, while Fusobacterium and Bacilli were higher in controls. In the nasal microbiome, Actinomycetaceae, Gemella, Proteobacteria, Actinomyces, and Veillonella were significantly higher in AMD, and when compared with controls, Streptococcus underwent the largest relative shift in AMD. Although not relatively abundant, Burkholderiales were also significantly increased in AMD, while Clostridia were increased in the control group [78]. More studies are desired to elucidate the relationship between the oral and nasal microbiome and aging in order to prevent upper respiratory tract infection [76] and subsequent lower respiratory tract, such as lung (which is covered in the next section) infections.

More and more evidence show that there are also diverse resident microbes in healthy lungs. The main phyla are Proteobacteria, Firmicutes, Bacteroidetes, and Actinobacteria [79]. The microbiome of the lung tissue is still largely unknown, and many studies have linked changes in the lung microbiome to the development of chronic lung diseases, such as cystic fibrosis (CF) or chronic obstructive pulmonary disease (COPD). However, the complex relationship between lung microbiota and disease remains to be elucidated. The surface of healthy lung is a dynamic environment, and debris and microorganisms from the mouth and nose continue to enter this respiratory organ, and the ciliary cells of the bronchus can remove these debris and invading bacteria through rhythmic movement.

In general, the lung microbiota is diverse, and it varies greatly among specific individuals and is dominated by 9 core genera: Prevotella, Sphingomonas, Pseudomonas, Acinetobacter, Clostridium, Megasphaera, Veillonella, Staphylococcus, and Streptococcus [80]; some of them such as Prevotella, Veillonella, and Streptococcus were also frequently observed in the oral cavity [81]. Although the respiratory system and its mucosa are interconnected, the lung microbiota has very distinctive features that set it apart from the upper respiratory tract [82].

Studies have shown several factors that are related to changes in lung microbiology, such as air pollutions, smoking, aging, and diseases. For instance, air pollution caused an increase in the relative abundance of potentially pathogenic bacterial groups such as Streptococcus and Neisseria [83]. The characteristics of the lung microbiome also change with the natural process of aging. Evidence has shown the loss of diversity of lung microbiota with increasing age and lung disease severity [79, 84], and antibiotic exposure was strongly associated. Aging-related immune dysfunction also affects the lung microbiome [84, 85]. A recent analysis of the lung microbiota of 167 severe asthma patients also revealed significant differences among patients with different inflammatory phenotypes [86].

The human skin is inhabited by a large and diverse community of both bacteria and fungi that contributes to the protection against invading pathogens and educates the IS [87, 88]. The composition of the skin microbiome primarily depends on the physiology of the skin site: for instance, sebaceous sites are usually dominated by Propionibacterium members, whereas moist areas, such as feet and elbow’s bend, are usually populated by Staphylococcus and Corynebacterium [89, 90]. This is related to the fact that microbes inhabiting skin are selected based on their ability to utilize resources present in sweat, sebum, and/or the debris of dead skin cells present in the outermost layer of human epidermis [88].

Being that the skin is such an exposed environment, it would be natural to think of its microbiome composition as much less stable than that of more protected environments such as gut or vaginal ecosystem; conversely, it has been proven by longitudinal studies that skin microbiome composition is stable, especially in sebaceous sites, e.g., the forehead [91]. Studies on individuals of different ages have provided evidences that human skin microbiome stabilizes its compositional structure around the age of 3 years, similarly to what happens for the GM [92], but it undergoes a dramatic restructuring at the time of puberty, when changes in hormone concentration stimulate sebum production [93]. Age-related changes in skin microbiome are interesting because of the many skin disorders associated with the puberty transition as well as the different propensity for atopy shown by the children and adult’s skin [88].

Analogously to pubescent individuals, skin structure and physiology change for elderly along with aging-related endogenous intrinsic factors, e.g., changes in cellular metabolisms, immunosenescence, and altered hormone condition [94]. Changes in skin structure also depend upon lifestyle choices and environmental challenges taken during the whole adult life, including cumulative UV exposure, smoking, and pollution [95]. These factors together usually determine a decrease in sweat, sebum, and immune homeostasis, resulting in alterations in skin physiology (e.g., pH, lipid composition). These physiological changes ultimately provide alterations in the microenvironment that affect the skin microbiome composition, especially in relation to the decrease in sebum production [90] and the occurrence of skin ulcers in bedridden elderly [96]. However, the literature focusing on the skin microbiome in elderly population is far from comprehensive, with few studies on limited populations currently available [97‐99]. Ying and colleagues (2015) focused on rural and urban populations from Shanghai area (China) and provided evidences that, even if aging had an effect on skin microbiome composition, the rural/urban environment was the most relevant driver for this exposed human microbiome. Shibagaki and colleagues, on the contrary, focused on a small population of healthy Japanese women: besides identified a number of bacterial species that showed differential abundances between older and younger women, the authors provided evidences, confirmed later on by Wu and colleagues that age-related alterations in skin microbiome are site dependent. In forehead, cheek, and forearm, the author found an age-associated decline in Propionibacterium abundance, proposedly related to the decrease in sebaceous gland activity which is typical of older age. Indeed, Wu et al. confirmed not only in Sardinian elderly (Italy) but also in centenarians from the same area that Propionibacterium in forehead and palm microbiomes decreases along with the advancement of age, with other genera (i.e., Prevotella, Rothia, and Veillonella) becoming overrepresented. Wu et al. also took into account the eukaryotic skin population alongside the bacterial one: skin fungi population seemed to be less affected by the advancement of age, with Malassezia consistently dominating the various skin sites. On the contrary, it has been demonstrated that aging significantly affects the proportion of the abundance of the Archaea counterpart of the skin microbiome, which increases in older age in relation to the lower sebum levels and reduced moisture [100].

More studies, on larger and more geographically spread populations, will offer a comprehensive view of the microbiome changes that occur along with the aging process on the human skin, ultimately providing useful and exploitable information in the field of treating and preventing age-related skin disorders. For instance, elderly is a subgroup of patients with distinct atopic dermatitis manifestation with respect to atopic infants, children, and adults [101], with the atopic problem possibly opening the way to bacterial skin infections [102]. Since it is known that skin microbiome, and in particular an increase colonization by Staphylococcus aureus, contributes to the exacerbation of atopic dermatitis [103], it is necessary to understand if and how age-related modification in the skin microbiome, in the different sites, can favor a skin microecosystem in which atopic manifestations, as well as their infectious consequences, are promoted.

Vaginal microbiome is probably the most studied human microbial ecosystem after the gut, because of its well-known relationship with the women health status [104]. Consistently across the whole literature, a vaginal environment dominated by Lactobacillus species is associated with vaginal health, with this group of bacteria being considered as keystone for the ecological balance of the vaginal environment. Lactobacillus species are responsible for the production of metabolites such as lactic acid and hydrogen peroxide, which contribute to the maintenance of the healthy value of vaginal pH, as well as the creation of a microenvironment in which colonization by anaerobic and microaerophilic pathogens is prevented [105‐107]. The recent extensive application of NGS to vaginal samples across different population and physiological conditions allowed for the categorization of vaginal microbiome into a discreet number (5 to 8) of community state type (CST), characterized by different degree of dominance of different Lactobacillus species (i.e., L. iners, L. crispatus, L. gasseri, L. jensenii) or by the absence of such dominance. CSTs deprived of a strong Lactobacillus dominance and enriched in other, often anaerobic, bacteria (e.g., Streptococcus, Atopobium, Megasphaera, Prevotella) were associated to a higher probability of disease or poor health outcome, such as bacterial vaginosis and pre-term delivery [108‐111]. In spite of these few possible configurations, the vaginal ecosystem strikes as particularly dynamic and undergoes compositional and functional changes along the woman life, in relation to hormonal changes, the most evident being puberty, pregnancies, and the beginning of menopause. Smoking, diet, hygiene, and sexual practices add complexity to the description of the vaginal microbiome dynamics [104, 112]. While puberty is known to represent the moment of the most dramatic changes in the vaginal microenvironment [112], pregnancy is of outmost interest because of the proven contribution of vaginal microbiome to pre-term labor and delivery [110, 111]; menopause-related changes in vaginal ecosystem have started to interest research only during the last decade [113‐118].

Menopause causes modifications of the vaginal environment that include decrease of the mucus layer width, estrogen level, and glycogen production. Such physiological changes are accompanied by modifications in the resident microbiome that includes a depletion in the proportion of Lactobacillus members, as well as a general decrease in the absolute number of colonizing bacteria and, consequently, an increase in biodiversity and vaginal pH [112, 118]. The focus of the majority of the research in this field has been how these microbiome changes are connected with genitourinary symptoms that are experienced by a large number of aging post-menopausal women. Indeed, vulvovaginal atrophy (VVA) and genitourinary symptoms of menopause (GSM, including burning, dryness, irritation, and so on) are experienced by approximately half of western post-menopausal women [112, 119]. It was reported that post-menopausal women with none to mild symptoms had significantly higher Lactobacillus predominance, and consequently lower biodiversity, than those complaining of more severe vaginal symptoms, whose samples were found enriched in bacteria such as Prevotella, Porphyromonas, Peptoniphilus, and Bacillus [114]. Later on, Brotman and colleagues confirmed that post-menopausal women with the most severe VVA showed a vaginal CST belonging to the group IVA, i.e., the one without Lactobacillus dominance and enriched in Anaerococcus, Peptoniphilus, Prevotella, and Streptococcus. Even if such studies did not provide answer to the chicken-and-egg question (VVA and GSM-related changes in microbiome cause or are caused by menopausal symptoms?), it was a natural evolution of the research field to wonder if therapies improving menopause-associated disturbances also had some effect on vaginal microbiome. Indeed, several studies showed that hormone replacement therapy, besides being effective in reducing menopausal symptoms [120, 121], influences the vaginal microbiome in a positive manner, by increasing the Lactobacillus amount and favoring the re-establishment of a vaginal microenvironment more similar to the one found in pre-menopausal women, i.e., higher glycogen production and lower pH [112, 117, 118]. Based on this, even if the relationship between estrogen level and vaginal microbiome in menopausal women has yet to be thoroughly explored, it is suggested that the maintenance of a vaginal microbiome dominated by Lactobacillus is relevant for ensuring a good quality of life for post-menopausal women. For this reason, the possibility of probiotic usage, orally or locally administered, during the management of VVA in aging women has been proposed and explored [122, 123]. A few studies highlighted that Lactobacillus-based products have the ability to increase the Lactobacillus dominance in post-menopausal women [124‐126].

Advances in our understanding of human microbiota especially GM and host interaction has stimulated our interest in other mucosal sites such as the genito-urinary tract microbiome (GUTM). Bladder and lower urinary tract were misunderstood as sterile for a long time. Although in normal physiological conditions, the commensals in GUTM were less abundant but highly variable compared with that in the gut [127], there are several genera commonly observed, such as the most dominant genera Lactobacillus and Gardnerella in healthy female cohorts [128, 129] and Lactobacillus and Streptococcus in healthy male cohorts [130, 131]. However, there are not always consistent results among studies for the commensals in the GUTM, which may be caused by the different types of samples and detection methodology [132]. Despite significant differences in gross anatomy and physiology of the lower urinary tract for females and males [133], the urine microbiota in the male and female are dispersed clustered into several “urotypes” rather than have a clearly separated clustering [134]. It is worth noting that the diversity of urine microbiota in healthy men is larger than that of healthy women [134]. Although the urine microbiota may have linkage with vaginal microbiota and seminal microbiota (which was well summarized in review) [135], it was not strongly influenced by the distal regions of the urogenital tract. Whereas a significant overlap between the bladder and vaginal microbiota was identified in a recent study [127], which reveals an interconnected GUTM. Currently, there is no direct comparison between the bladder microbiota (or clean-catch urine microbiota) and seminal microbiota, while comparison of the bacterial communities in semen with those of first catch urine did demonstrate that these specimens shared one-third of species [136]. Till now, there remain some doubts on whether different urogenital sites harbor a unique microbiome.

Many clinical situations such as urinary tract infection [137], interstitial cystitis [138], urinary incontinence [129], the formation of kidney stones (urolithiasis) [139], and even genito-urinary tract cancer such as bladder cancer [140], prostate cancer [141], and kidney cancer [142] have some correlations with the altered GUTM. Those clinical conditions are more relevant to the aging population, for example, the high incidence of urinary tract infection was 40% of men and 28% of women in their 70s [143, 144]. During the aging process, host physiological and lifestyle changes, for instance, the IS function declining, sexual activity frequencies decrease, more frequent medication exposure, may affect the GUTM. There are studies that demonstrated the age-related urine microbiota variations in female, similar with the GM which also showed age-related variations during aging [36, 145, 146]. A study that characterized the urinary microbiota in elderly (average 71.8 years) and younger females in China (average 50.0 years) has found significant differences between them [147]. The relative abundance of Lactobacillus and Bifidobacteria was negatively related to age, while Peptococcus was positively related to age. Moreover, the correlation between a higher level of Lactobacillus and diabetes was identified in the elderly, and lower levels of Peptoniphilus and Dialister were correlated with asymptomatic bacteriuria. While another study in the UK did not find significant correlations between age and diversity of the bladder microbiome in healthy females [148], the lactobacillus was observed related to the pre-menopausal females, and Mobiluncus was related with post-menopausal females. A study focused on the urine microbiota in the urinary incontinence female and control has found that urine microbiota formed six community types (urotype), which was not significantly associated with the urinary incontinence but was age-related [129]. The young females (< 51 years) have a higher proportion of individuals with Lactobacillus-dominated urotype structure (with relative abundance of Lactobacillus > 89%). Interestingly, the younger females without the Lactobacillus-dominated urotype were correlated with a significantly high incidence of urinary incontinence but not for the older females. Another similar study that surveyed the urine microbiota only in the urinary incontinence female showed that the younger females (average 55.8 years) have a higher incidence of being positively detected with urine microbes compared with the older female (average 61.3 years), and the Enterobacteriaceae-dominant urotypes were detected within females with an average age of 70 years compared with the Lactobacillus-dominated or Gardnerella-dominant urotypes with an average age around 54 years [149]. A more recent study describes the less disperse cluster of bladder microbiota in younger women (average 51 years) when compared with the older female (average 59 years) [150]. Enterobacteriaceae and other potential pathogens including Pseudomonas and Staphylococcus are consistent with clinical observations that the older female has an upward trend of getting urinary tract infection [151]. A study that used 16S rRNA qPCR and cultivation methods indicated that the “detectable” urine bacteria are not significantly associated with age; however, Jonquetella, Parvimonas, Proteiniphilum, and Saccharofermentans were found enriched in aged individuals over 70 years [152]. Furthermore, an association has been noted between male age and seminal bacteria; the presence of environmental bacteria such as Pseudomonas, Janthinobacterium, Gillisia, Flavobacterium, and Acidovorax was associated with older age instead of vaginal bacteria such as L. crispatus, L. iners, G. vaginalis, Dialister, Atopobium vaginae, and Mobiluncus curtisii that seemed to be associated with younger age [153]. Although there is still no comprehensive study of the aging-related urine microbiota, the impact of the microbiota on genito-urinary tract homeostasis and disease development is emerging. Future studies will improve our understanding of the causative relationship between the defined microbes and genito-urinary tract diseases. And undoubtedly, age should be considered when we try to use the urine microbiota as a predictor of disease or treatment.

A summary of the findings on the association of aging with changes in microbiomes from the different body sites described in this review is reported in Table 2.