Skip to main content
Download PDF

Identification of age-associated microbial changes via long-read 16S sequencing

Gut Pathogens volume 16, Article number: 56 (2024) Cite this article

Abstract

Background

Age-related gut microbial changes have been widely investigated over the past decade. Most of the previous age-related microbiome studies were conducted on the Western population, and the short-read sequencing (e.g., 16S V4 or V3-V4 region) was the most common microbiota profiling method. We evaluated the gut compositional differences using the long-read sequencing approach (i.e., PacBio sequencing targeting the full-length V1-V9 regions) to enable a deeper taxonomic resolution and better characterize the gut microbiome of Singaporeans from different age groups.

Results

A total of 83 research participants were included in this study. Although no significant differences were detected in alpha and beta diversity, our study demonstrated several bacterial taxa with abundances that were significantly different across age groups. With young individuals as the reference group, Eggerthella lenta and Bacteroides uniformis were found to be significantly altered in the middle-aged group, while Catenibacterium mitsuokai and Bacteroides plebeius were significantly altered in the elderly group. These age-related differences in the gut microbiome were associated with aberrations in several predicted functional pathways, including dysregulations of pathways related to lipopolysaccharide and tricarboxylic acid cycle in older adults.

Conclusions

The utilization of long-read sequencing facilitated the identification of species- and strain-level differences across age groups, which was challenging with the partial 16S rRNA sequencing approach. Nevertheless, replication studies are warranted to confirm our findings, and if confirmed, further in vitro and in vivo studies are crucial to better understand the impact of the altered levels of age-related bacterial taxa. Additionally, the modest performance of strain-level taxonomic classification using 16S-ITS-23S gene sequences, likely due to the limited depth of currently available alignment databases, highlights the need for optimization and refinement in curating these databases for the long-read sequencing approach.

Background

According to the 2023 United Nations Report, the global aging population over the age of 65 is projected to reach 1.6 billion by 2050 [1]. While this significant increase in life expectancy is a remarkable achievement, it is accompanied by a rise in disability rates, potentially leading to higher healthcare costs. Understanding how environmental and biological factors contribute to unhealthy aging is therefore essential. Generally, the composition of the human gut microbiota remains relatively stable during early and mid-adulthood. However, as individuals age, the stability of the gut microbiome may be disrupted due to various factors such as biological aging processes, changes in health status, and diet [2, 3].

A growing body of evidence suggests alterations in gut microbial composition in aged individuals. However, current knowledge on aging and the gut microbiome is predominantly derived from studies conducted in the Western world [4, 5] and, to some extent, China [6, 7]. There is a lack of similar studies from ASEAN countries, such as Singapore, which has an ethnically diverse population (i.e., Chinese, Malay, and Indian) practicing contrasting lifestyles. Furthermore, most previous studies have utilized the 16S ribosomal ribonucleic acid (rRNA) sequencing method (e.g., V3-V4 region [5, 8, 9]), which has several limitations.

Short-read sequencing of the bacterial 16S rRNA gene is often used to profile the gut microbiome due to its cost-effectiveness. The 16S rRNA gene, which is 1.5 kb in length, covers nine hypervariable regions. Amplicon-based 16S rRNA gene sequencing allows the amplification of selected variable regions only. The limited read length of the short-read sequencing approach, combined with amplification bias, often results in low taxonomic resolution [10, 11]. To overcome these limitations, some studies have utilized shotgun metagenomics sequencing, which provides better sensitivity and resolution by fragmenting the whole genome into smaller pieces of deoxyribonucleic acid (DNA) and then reassembling the fragments using overlapping regions. However, shotgun metagenomics sequencing often requires significant computational power and is costly.

Long-read sequencing is a cost-effective alternative to shotgun metagenomics and offers better taxonomic resolution than short-read 16S rRNA sequencing. This method allows the entire 16S rRNA gene (i.e., all nine variable regions), extending to the Internal Transcribed Spacer (ITS) and 23S rRNA gene regions, to be sequenced. Although some may argue that long-read sequencing has lower accuracy (∼ 90%), the recent development of circular consensus sequencing (CCS) from PacBio and improved methodologies to remove polymerase chain reaction (PCR)-related errors have enabled researchers to discriminate sequences differing by a single nucleotide across an entire gene with an average length of up to 13.5 kb with 99.8% accuracy [11, 12]. This allows for the identification of intragenomic 16S gene copy variants within a taxon, enabling easier distinction of species and strains (sub-species) and preventing the over-estimation of bacterial diversity [10, 13].

Therefore, in this study, we utilized the long-read sequencing approach to investigate the gut microbiome profiles of individuals from different age groups, with the aim to better characterize and understand the gut microbial differences at species-level.

Methods

Data and sample collection

Singapore residents above age of 21 were recruited from AMILI’s Poop Saves Life campaign which was a public outreach program, where 83 participants were selected for this study. Participants were selected if they fulfilled the following criteria: (1) ethnicity of Chinese, Malay, or Indian, (2) no use of antibiotics three months prior to recruitment, (3) no regular use of proton pump inhibitors, and (4) absence of gastrointestinal diseases such as inflammatory bowel disease and colon disease. Written informed consent was obtained from all participants and the study was approved by AMILI Institutional Review Board (reference number 2020/0501) in accordance with ethical guidelines.

Basic socio-demographic information and medical history, including age and body mass index (BMI), were collected from each participant. Information on daily dietary intake were recorded using a locally-validated Food Frequency Questionnaire (FFQ) [14]. Stool samples were collected in DNA/RNA shield stool collection kit (Zymo Research, Irvine, California, USA). The stool collection kit and a detailed instruction sheet on stool collection (Additional File 1) were provided to the participants. After sample collection, stool samples were sent to the laboratory via post. All samples reached the laboratory within a week from sample collection. Upon receiving, the fecal samples were aliquoted and frozen at -80 Â°C until further sample processing.

Sample preparation

Total DNA was extracted from the fecal samples using the QIAmp PowerFecal Pro DNA extraction kit (Qiagen, Hilden, Germany) following the manufacturer’s instructions. The full-length 16S-ITS-23S rRNA gene was amplified by PCR using consensus sequences 27F 5’-AGRRTTYGATYHTDGYTYAG-3’ and 23SR 5’-AGTACYRHRARGGAANGR-3′. Amplicon libraries were created using the Shoreline Complete StrainID kit (Shoreline Biome, Farmington, Connecticut, US) and sequenced on PacBio Sequel II system.

Sequence and predictive functional analyses

The raw 16S-ITS-23S rRNA gene sequences were first quality-filtered and denoised using DADA2 version 1.22.0 [15] and subsequently imported into SBanalyzer software for strain-level microbial identification using the Athena database. For the analysis of 16S V1-V9 gene sequences, the raw 16S-ITS-23S rRNA gene sequences were first trimmed to the 16S V1-V9 region using cutadapt version 3.7 [16] based on primer sequences F27 (AGRGTTYGATYMTGGCTCAG) and R1492 (AAGTCGTAACAAGGTARCY). The trimmed sequences were then quality-filtered and denoised using DADA2 version 1.22.0 [15]. After that, filtered reads were grouped into amplicon sequence variants (ASVs). To maximize the assignment rate, each ASV was first taxonomically classified based on GTDB r207 [17], followed by Silva v138 [18], and lastly RefSeq + RDP [19, 20]. The microbial gene content was then inferred from the taxa abundance using PICRUSt2 [21].

ASV sequences corresponding to the differentially abundant bacterial species detected using the 16S V1-V9 dataset were extracted and searched against the ASV taxonomy table of the 16S-ITS-23S dataset. The 16S-ITS-23S ASVs that have similar gene sequences and that were assigned to the same species were analyzed to obtain strain-level information.

Statistical analysis

All statistical analyses were carried out using R version 4.1.2. All tests were considered significant at p-value < 0.05. All p-value adjustments were based on the Benjamini-Hochberg procedure. Alpha diversity measurements (Shannon, Chao1, and Pielou’s index) were computed using the microbiome R package version 1.16.0 [22]. Beta diversity measures were inferred by first transforming the data based on centered log-ratio to account for compositionality [22, 23]. The Aichitson distance was used for principal coordinate analysis (PCoA) and permutational multivariate analysis of variance (PERMANOVA) (1000 permutations). PERMANOVA was carried out using the adonis2 function in vegan version 2.6.2 [24]. Differentially abundant taxa/pathways across different age groups (adjusted p-value < 0.05) were identified based on the analysis of compositions of microbiomes with bias correction 2 (ANCOM-BC2) method [25].

Results

Subjects’ characteristics

A total of 83 participants were included in this study. Following the age cut-offs used by several prior studies [9, 26, 27], participants were grouped into different age groups, as follows: (1) young (age < 36; n = 25), (2) middle-aged (36 ≤ age < 56; n = 30), and (3) old (age ≥ 56; n = 28). The demographics and basic characteristics of the cohort are summarized in Table 1. There were no significant differences in sex, BMI, race, smoking status and daily dietary intake among the three groups included in this study.

Table 1 Demographics and characteristics of subjects included in the study
Full size table

Comparisons of taxonomy classification of 16S-ITS-23S vs. 16S V1-V9 datasets

The original 16S-ITS-23S sequencing data yielded a total of 2,327,728 gene sequences (mean = 28,045; standard deviation = 8,494; min = 11,147; max = 49,313). After the alignment against Athena database, nine phyla, 14 classes, 26 orders, 45 families, 83 genera, and 144 species were detected (Additional File 2: Supplementary Table 1). Of the 2,740 features, 1,843 (67.3%) and 1,111 (40.5%) reached species- and strain-level resolution, respectively. Notably, there were only 59 different strains found in the output of this sequencing dataset. Given that bacterial strains can differ by only a few bases, the limited strain-level information obtained using the 16S-ITS-23S sequences is likely due to the limited depth of the Athena database.

To maximize the utility of our data, we trimmed the 16S-ITS-23S sequences to the 16S V1-V9 region, aiming to identify more distinct species. This 16S V1-V9 sequencing dataset yielded 1,937,509 gene sequences, with an average of 23,343 sequences per sample (ranging from 9,127 to 44,382, with a standard deviation of 8,783). After quality filtering using the DADA2 R-package and assigning taxonomy based on three different reference databases, we identified eight phyla, 19 classes, 40 orders, 76 families, 182 genera, and 299 species. Of the 2,307 ASVs, 2,031 (88.0%) reached species-level resolution (Additional File 2: Supplementary Table 1). Since the taxonomic classification of 16S V1-V9 data provided more information, we proceeded the subsequent analyses using this dataset, unless otherwise specified.

Comparisons of bacterial richness (alpha diversity) and composition (beta diversity) across age groups

No significant differences were detected in all diversity indices across the three age groups (Wilcoxon rank-sum test, adjusted p-value > 0.05; Fig. 1A). However, participants of older age generally exhibited higher alpha diversity than younger participants.

Fig. 1
figure 1

Comparisons of alpha diversity and beta diversity across age groups. (A) Alpha diversity is measured using the Shannon diversity index, Chao1 diversity index, and Pielou’s evenness index. No significant differences were detected in any of the diversity indices across the three age groups (Wilcoxon rank-sum test, adjusted p > 0.05). (B) PCoA demonstrating insignificant separation in the gut microbial composition of individuals from different age groups (PERMANOVA; R² = 0.028, p = 0.098). Each data point represents a single sample. The distance between data points reflects the differences between samples

Full size image

PERMANOVA revealed that age groups explained about 2.8% of the overall variation in gut microbiome composition, although this was not statistically significant (R² = 0.028, p-value = 0.098). Additionally, the complementary PCoA analysis showed modest separation among age groups (Fig. 1B). Age (on a continuous scale) (R² = 0.015, p-value = 0.089), as well as the rest of the potential confounding factors, including sex (R² = 0.012, p-value = 0.399), BMI (R² = 0.016, p-value = 0.155), race (R² = 0.025, p-value = 0.435), smoking status (R² = 0.029, p-value = 0.117), and various dietary parameters, did not significantly contribute to differences in gut microbial composition (Table 2).

Table 2 Effects of age groups and various potential confounders on gut microbiome compositional differences
Full size table

Differentially abundant taxa across age groups

Using young participants as the reference group, we identified twelve and ten bacterial species that were differentially abundant in the middle-aged and old groups, respectively (adjusted p-value < 0.05, Fig. 2A). However, only four differentially abundant taxa (i.e., Eggerthella lenta [adjusted p-value = 0.049], Bacteroides uniformis [adjusted p-value = 0.033], Catenibacterium mitsuokai [adjusted p-value = 0.004] and Bacteroides plebeius [adjusted p-value = 0.030]) passed sensitivity tests for the pseudo-count addition. This step helps to mitigate the risk of inflated false-positive rates due to the implementation of pseudo-count in ANCOM-BC2 [25]. E. lenta was significantly increased in the middle-aged group, while B. uniformis was significantly reduced in the same group. On the other hand, the abundances of B. plebeius and C. mitsuokai were significantly increased and reduced, respectively, in the old group. Interestingly, when the ASV gene sequences of the four differentially abundant bacterial species were compared against the 16S-ITS-23S output, we identified E. lenta DSM 2243 to be exclusively present in the middle-aged subgroup. However, it is important to note that the bacterial strain was detected in only 8/30 (26.7%) of them (Additional File 2: Supplementary Table 2).

Fig. 2
figure 2

Differentially abundant bacterial species and predicted functional pathways. The analysis was conducted using the ANCOM-BC2, with young individuals as the reference group. The columns represent age groups, while the rows show differentially abundant (A) bacterial taxa or (B) predicted functional pathways. Cell colors indicate log fold changes, with blue representing a significant reduction and red representing a significant increase (darker colors correspond to greater log fold changes). The log fold changes relative to the reference (i.e., young) group are noted in each cell. Values in white font indicate log fold changes that are significant at an adjusted p-value < 0.05, while values in bold black font indicate log fold changes that are significant at an adjusted p-value < 0.05 and passed sensitivity tests for pseudo-count addition

Full size image

At the genus level, Olsenella was found to be significantly increased in middle-aged participants (adjusted p-value < 0.001), while Holdemanella (adjusted p-value < 0.001) was significantly increased and Catenibacterium (adjusted p-value = 0.004) was significantly reduced in the old subgroup. Additionally, bacteria from the Atopobiaceae family were found to be increased in the gut microbiome of middle-aged individuals (adjusted p-value < 0.001). All four differentially abundant bacterial taxa had an adjusted p-value < 0.05 and passed the sensitivity tests. The complete list of bacterial classes, orders, families, and genera detected to be differentially abundant in the middle-aged and old groups, with young individuals as the reference, is provided in Additional File 2: Supplementary Fig. 1.

Differentially abundant bacterial taxa were also identified in pairwise comparisons between all possible combinations of the age groups. Bacterial species with abundances that were significantly different between any two groups (adjusted p-value < 0.05), as well as passed the sensitivity tests, are listed in Supplementary Table 3 (Additional File 2), along with their respective log fold changes and adjusted p-values.

Predictive functional analysis

With young individuals as the reference, predictive functional analysis revealed a total of eight metabolic pathways that differed significantly across age groups (adjusted p-value < 0.05; Fig. 2B). Among these, three pathways were significantly downregulated in the middle-aged subgroup, including (1) superpathway of UDP-N-acetylglucosamine-derived O-antigen building blocks biosynthesis, (2) superpathway of 2,3-butanediol biosynthesis and (3) chondroitin sulfate degradation I. The remaining five functional pathways were differentially abundant in the elderly group, with the superpathway of (Kdo)2-lipid A biosynthesis, superpathway of heme b synthesis from glycine, aerobactin biosynthesis and 3-phenylpropanoate degradation being significantly increased, while L-histidine degradation II being significantly downregulated. However, it is important to note that these pathways did not pass the ANCOM-BC2 sensitivity tests.

Discussion

In this study, we investigated the association between age and gut microbial composition in individuals residing in Singapore. To the best of our knowledge, this is the first full-length 16S rRNA gene assessment of the gut microbiota in a multi-ethnic country. Previous studies evaluating the effect of age on the gut microbiome primarily used a short-read 16S rRNA gene sequencing approach (e.g., V3-V4 region: ∼460 base pairs [5, 8, 9, 28]). The limited resolution of short-read sequencing often fails to detect bacterial changes at the species/strain level, which can affect data interpretation. In our work, we utilized the long-read sequencing approach (i.e., PacBio sequencing) to explore age-related gut microbiome alterations for the first time. In addition to replicating previous findings, our study unveiled several novel differentially abundant taxa and predicted functional pathways associated with age.

Overall, no significant differences were detected in alpha diversity across all age groups, although there was an increasing trend with age. de la Cuesta-Zuluaga et al. (2019) reported a minimal association between gut biodiversity and age in a Chinese cohort, contrasting with the significant positive correlation found in cohorts from the United States, United Kingdom, and Colombia [29]. This finding supports our results, as nearly all participants in our study are Chinese (n = 78; 94.0%). However, it is well established that individuals of the same ethnicity may exhibit different gut microbial profiles when residing in different geographical regions [30]. Therefore, factors such as ethnicity, geographical location, and dietary habits must be controlled when investigating the gut microbiome.

To ensure a more robust interpretation of age-related gut microbiome changes, we investigated the effects of various potential confounders, including sex, BMI, ethnicity, smoking status, and daily dietary intake, on the overall gut microbiome composition of our cohort. None of these potential confounders significantly contributed to the overall gut microbiome differences in the cohort. We also did not observe significant effects of age groups or age (on a continuous scale) on the overall gut microbial composition of the participants, which was further supported by the modest separation observed in the PCoA analysis. While these insignificant findings could be attributed to the study’s limited sample size, the lack of significant differences in gut microbial composition across age groups is consistent with results from two independent Japanese studies [9, 28]. Although the exact reasons for discrepancies with other age-related gut microbiome studies are unclear, this could be partly due to their Asian origin or the health of participants (i.e., participants were relatively healthy). Further investigations are warranted to better understand these observations.

Despite the insignificant differences in alpha and beta diversity, several differentially abundant bacterial taxa were detected among the age groups. For instance, a notable decrease in Bacteroides uniformis was observed in the gut microbiome of middle-aged individuals, while Bacteroides plebeius was significantly elevated in the old group. These findings align with multiple earlier studies, which reported contrasting findings on the abundances of Bacteroides in the gut microbiome of individuals from different age groups. Some studies documented increased levels in younger adults [4, 8], while others reported higher abundances in the elderly group [28, 31]. Interestingly, recent investigations have also linked the differential abundance of Bacteroides to the overall health status of the study group [5, 32]. Collectively, these findings underscore the importance of employing sequencing techniques that provide precise taxonomic assignments down to the species/strain level, thereby facilitating a more comprehensive delineation of the gut microbiome and permitting a more accurate understanding of the microbial ecology associated with specific conditions.

In the pairwise comparison between middle-aged and old groups, we found that elderly individuals exhibited a significantly higher abundance of Klebsiella pneumoniae (as well as the Klebsiella genus) in their gut microbiome. The elevated level of K. pneumoniae in older individuals is believed to be associated with factors such as increased use of medication [33] or inflammation linked to interleukin-6 [34], both of which are common in older individuals. This bacterium, which is known to be a pathogen, may contribute to health issues frequently observed in this age group. Given the potential health implications, future investigation should aim to validate this finding through larger, longitudinal studies. If confirmed, further research should explore the specific implications of changes in K. pneumoniae level in older adults.

Besides replicating previous research findings, our study revealed several novel differentially abundant bacterial species that have not been previously reported. These include increased abundances of Eggerthella lenta in middle-aged participants and reduced levels of Catenibacterium mitsuokai in elderly individuals. E. lenta, a bacterium belonging to the Coriobacteriaceae family, is known as an opportunistic pathogen implicated in various conditions and infections [35]. This gut bacterium is also involved in the inactivation of the cardiac drug digoxin [36]. While E. lenta may negatively impact human health, it is crucial to obtain strain-level information to elucidate its exact role in the aging-related gut microbiome. C. mitsuokai, on the other hand, is generally considered part of the normal human gut microbiome [37]. Previous studies have linked C. mitsuokai with dyslipidemia and insulin resistance [38], and a higher abundance of the Catenibacterium genus has been associated with a potentially lower risk of frailty [39]. Altogether, these findings suggest potential health implications related to changes in the levels of C. mitsuokai in the gut microbiome. The observed reduction of C. mitsuokai in elderly individuals of our cohort could either reflect age-related alterations in gut microbiome composition or represent a compensatory response to the health changes commonly seen in old individuals. With the increasing emphasis on independent replication of scientific findings, especially in the field of gut microbiome where numerous confounders exist [40], further investigations are warranted to confirm these findings and explore their roles in relation to human age.

Given the importance of understanding the functional implications of changes in the gut microbiome across different age groups, we conducted a predictive functional analysis using PICRUSt2. Through differential abundance analysis, we identified several metabolic pathways that may be impacted by alterations in the gut microbiome. Notably, we observed dysregulated pathways related to lipopolysaccharide (LPS) biosynthesis, including the superpathway of UDP-N-acetylglucosamine-derived O-antigen building blocks biosynthesis and the superpathway of (Kdo)2-lipid A biosynthesis, in older individuals. LPS are outer membrane components of gram-negative bacteria, composed of three main domains: lipid A, the core oligosaccharide, and the O antigen [41]. The dysregulations of LPS-related pathways observed in this study suggest the presence of gut inflammation in older individuals, potentially activated by LPS through toll-like receptor (TLR)-4 signaling [42]. Interestingly, we also observed a significant reduction in chondroitin sulfate degradation in middle-aged individuals. The degradation and fermentation of chondroitin sulfate by the human gut microbiota produce significant amounts of short-chain fatty acids, which are also known to regulate gut permeability and inflammation [43]. Additionally, elderly individuals demonstrated downregulated degradation of L-histidine and increased degradation of 3-phenylpropanoate. Aberrations of these two metabolic pathways could lead to the production of compounds like fumarate, pyruvate and acetoacetate, which are involved in the tricarboxylic acid (TCA) cycle, potentially disrupting central metabolism [44, 45]. However, the functional analysis conducted in this study is predicted based on microbial abundances, and the differentially abundant predicted functional pathways did not pass the ANCOM-BC2 sensitivity tests, and thus requires cautious interpretation. Further in vitro and in vivo studies are warranted to better elucidate the functional implications of age-related microbial changes in the gut.

Our study is not free from limitations. Although we managed to identify several age-related differentially abundant bacterial species, most of our results were derived using the 16S V1-V9 sequences. This approach was necessitated by the modest taxonomic classification of 16S-ITS-23S sequences down to species/strain level, which is likely attributed to the limited depth of currently available alignment databases. For instance, using Athena database (16S-ITS-23S), only 67.3% and 40.5% of the features reached species- and strain-level resolution, respectively.

Furthermore, the small sample size of our study may impact the robustness of detecting microbiome differences across age groups, thereby requiring cautious interpretation and further validation in future studies. The use of the FFQ provides only a snapshot of the participants’ nutritional intake and may not fully capture the complexity of diet-microbiota interactions in this study. Additionally, the age cut-offs used in this study were primarily based on the quartiles and median age of the overall cohort, as well as the cutoffs employed by a few previous studies. The lack of consistent definitions of age cut-offs for young, middle-aged, and old adults in gut microbiome research was similarly highlighted by a previous study [46]. Lastly, the effect of age (as a continuous variable) on the gut microbiome was not explored in this study due to the lack of significance in beta diversity and differential abundance analyses.

Conclusions

Although no significant differences were observed in alpha and beta diversity, our study revealed several age-related differentially abundant bacterial taxa using the long-read sequencing method. Specifically, we observed altered abundances of Eggerthella lenta (enriched) and Bacteroides uniformis (reduced) in middle-aged individuals, while the old individuals exhibited significant alterations in the levels of Bacteroides plebeius (enriched) and Catenibacterium mitsuokai (reduced). These changes in the gut microbiome across age groups had some functional implications, such as dysregulations of LPS-related pathways and metabolic pathways involved in the TCA cycle. Further studies are warranted to confirm our findings and to better elucidate the impact of age-related changes in the gut microbiome. Importantly, the higher taxonomic resolution provided by the long-read sequencing technology used in this study allowed for more specific identification of bacterial taxa that may have been overlooked in previous studies employing the more commonly used short-read sequencing method. However, the limited depth of currently available alignment databases for 16S-ITS-23S sequences hindered the detection of a broader range of bacterial species/strains. Further optimization and refinement in the curation of alignment databases for 16S-ITS-23S sequences are required to resolve the issues faced in this study, as well as to enhance the accuracy and comprehensiveness of microbial identification in future research.

Data availability

The demultiplexed 16S-ITS-23S gene sequences generated and anaylzed in this study are available at the NCBI Sequence Read Archive (SRA) under BioProject ID PRJNA983707.

Abbreviations

rRNA:

Ribosomal ribonucleic acid

DNA:

Deoxyribonucleic acid

ITS:

Internal Transcribed Spacer

CCS:

Circular consensus sequencing

PCR:

Polymerase chain reaction

BMI:

Body mass index

FFQ:

Food Frequency Questionnaire

ASV:

Amplicon sequence variant

PCoA:

Principal coordinate analysis

PERMANOVA:

Permutational multivariate analysis of variance

LPS:

Lipopolysaccharides

TLR:

Toll-like receptor

TCA:

Tricarboxylic acid

References

  1. Wilmoth JR, Bas D, Mukherjee S, Hanif N. World social report 2023: leaving no one behind in an ageing world. UN; 2023.

  2. Boehme M, Guzzetta KE, Bastiaanssen TF, Van De Wouw M, Moloney GM, Gual-Grau A, et al. Microbiota from young mice counteracts selective age-associated behavioral deficits. Nat Aging. 2021;1(8):666–76.

    Article  PubMed  Google Scholar 

  3. Ghosh TS, Shanahan F, O’Toole PW. The gut microbiome as a modulator of healthy ageing. Nat Reviews Gastroenterol Hepatol. 2022;19(9):565–84.

    Article  Google Scholar 

  4. Herzog EL, Wäfler M, Keller I, Wolf S, Zinkernagel MS, Zysset-Burri DC. The importance of age in compositional and functional profiling of the human intestinal microbiome. PLoS ONE. 2021;16(10):e0258505.

    Article  PubMed  CAS  PubMed Central  Google Scholar 

  5. Wilmanski T, Diener C, Rappaport N, Patwardhan S, Wiedrick J, Lapidus J, et al. Gut microbiome pattern reflects healthy ageing and predicts survival in humans. Nat Metabolism. 2021;3(2):274–86.

    Article  CAS  Google Scholar 

  6. Kong F, Hua Y, Zeng B, Ning R, Li Y, Zhao J. Gut microbiota signatures of longevity. Curr Biol. 2016;26(18):R832–3.

    Article  PubMed  CAS  Google Scholar 

  7. Wang F, Yu T, Huang G, Cai D, Liang X, Su H, et al. Gut microbiota community and its assembly associated with age and diet in Chinese centenarians. J Microbiol Biotechnol. 2015;25(8):1195–204.

    Article  PubMed  CAS  Google Scholar 

  8. Chaudhari DS, Dhotre DP, Agarwal DM, Gaike AH, Bhalerao D, Jadhav P, et al. Gut, oral and skin microbiome of Indian patrilineal families reveal perceptible association with age. Sci Rep. 2020;10(1):5685.

    Article  PubMed  CAS  PubMed Central  Google Scholar 

  9. Takagi T, Naito Y, Inoue R, Kashiwagi S, Uchiyama K, Mizushima K, et al. Differences in gut microbiota associated with age, sex, and stool consistency in healthy Japanese subjects. J Gastroenterol. 2019;54:53–63.

    Article  PubMed  Google Scholar 

  10. Gehrig JL, Portik DM, Driscoll MD, Jackson E, Chakraborty S, Gratalo D, et al. Finding the right fit: evaluation of short-read and long-read sequencing approaches to maximize the utility of clinical microbiome data. Microb Genomics. 2022;8(3):000794.

    Article  Google Scholar 

  11. Johnson JS, Spakowicz DJ, Hong B-Y, Petersen LM, Demkowicz P, Chen L, et al. Evaluation of 16S rRNA gene sequencing for species and strain-level microbiome analysis. Nat Commun. 2019;10(1):5029.

    Article  PubMed  PubMed Central  Google Scholar 

  12. Wenger AM, Peluso P, Rowell WJ, Chang P-C, Hall RJ, Concepcion GT, et al. Accurate circular consensus long-read sequencing improves variant detection and assembly of a human genome. Nat Biotechnol. 2019;37(10):1155–62.

    Article  PubMed  CAS  PubMed Central  Google Scholar 

  13. Sun D-L, Jiang X, Wu QL, Zhou N-Y. Intragenomic heterogeneity of 16S rRNA genes causes overestimation of prokaryotic diversity. Appl Environ Microbiol. 2013;79(19):5962–9.

    Article  PubMed  CAS  PubMed Central  Google Scholar 

  14. Whitton C, Ho JCY, Tay Z, Rebello SA, Lu Y, Ong CN, Van Dam RM. Relative validity and reproducibility of a food frequency questionnaire for assessing dietary intakes in a multi-ethnic Asian population using 24-h dietary recalls and biomarkers. Nutrients. 2017;9(10):1059.

    Article  PubMed  PubMed Central  Google Scholar 

  15. Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJA, Holmes SP. DADA2: high-resolution sample inference from Illumina amplicon data. Nat Methods. 2016;13(7):581–3.

    Article  PubMed  CAS  PubMed Central  Google Scholar 

  16. Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet J. 2011;17(1):10–2.

    Article  Google Scholar 

  17. Parks DH, Chuvochina M, Chaumeil P-A, Rinke C, Mussig AJ, Hugenholtz P. A complete domain-to-species taxonomy for Bacteria and Archaea. Nat Biotechnol. 2020;38(9):1079–86.

    Article  PubMed  CAS  Google Scholar 

  18. Yilmaz P, Parfrey LW, Yarza P, Gerken J, Pruesse E, Quast C, et al. The SILVA and all-species living tree project (LTP) taxonomic frameworks. Nucleic Acids Res. 2014;42(D1):D643–8.

    Article  PubMed  CAS  Google Scholar 

  19. Cole JR, Wang Q, Fish JA, Chai B, McGarrell DM, Sun Y, et al. Ribosomal database project: data and tools for high throughput rRNA analysis. Nucleic Acids Res. 2014;42(D1):D633–42.

    Article  PubMed  CAS  Google Scholar 

  20. O’Leary NA, Wright MW, Brister JR, Ciufo S, Haddad D, McVeigh R, et al. Reference sequence (RefSeq) database at NCBI: current status, taxonomic expansion, and functional annotation. Nucleic Acids Res. 2016;44(D1):D733–45.

    Article  PubMed  Google Scholar 

  21. Douglas GM, Maffei VJ, Zaneveld JR, Yurgel SN, Brown JR, Taylor CM, et al. PICRUSt2 for prediction of metagenome functions. Nat Biotechnol. 2020;38(6):685–8.

    Article  PubMed  CAS  PubMed Central  Google Scholar 

  22. Lahti L, Shetty S. Introduction to the microbiome R package. Preprint at https://microbiome github io/tutorials. 2018.

  23. Gloor GB, Macklaim JM, Pawlowsky-Glahn V, Egozcue JJ. Microbiome datasets are compositional: and this is not optional. Front Microbiol. 2017;8:294209.

    Article  Google Scholar 

  24. Oksanen J, Blanchet F, Friendly M, Kindt R, Legendre P, Mcglinn D et al. Vegan: community ecology package. Ordination methods, diversity analysis and other functions for community and vegetation ecologists, 05–26. Version 2.5-1. 2019.

  25. Lin H, Peddada SD. Multigroup analysis of compositions of microbiomes with covariate adjustments and repeated measures. Nat Methods. 2024;21(1):83–91.

    Article  PubMed  CAS  Google Scholar 

  26. Ghosh TS, Das M, Jeffery IB, O’Toole PW. Adjusting for age improves identification of gut microbiome alterations in multiple diseases. eLife. 2020;9:e50240.

    Article  PubMed  CAS  PubMed Central  Google Scholar 

  27. Woodmansey EJ, McMurdo ME, Macfarlane GT, Macfarlane S. Comparison of compositions and metabolic activities of fecal microbiotas in young adults and in antibiotic-treated and non-antibiotic-treated elderly subjects. Appl Environ Microbiol. 2004;70(10):6113–22.

    Article  PubMed  CAS  PubMed Central  Google Scholar 

  28. Odamaki T, Kato K, Sugahara H, Hashikura N, Takahashi S, Xiao J-z, et al. Age-related changes in gut microbiota composition from newborn to centenarian: a cross-sectional study. BMC Microbiol. 2016;16:1–12.

    Article  Google Scholar 

  29. de la Cuesta-Zuluaga J, Kelley ST, Chen Y, Escobar JS, Mueller NT, Ley RE, et al. Age-and sex-dependent patterns of gut microbial diversity in human adults. mSystems. 2019;4(4). https://doi.org/10.1128/msystems. 00261 – 19.

  30. Dwiyanto J, Ayub Q, Lee SM, Foo SC, Chong CW, Rahman S. Geographical separation and ethnic origin influence the human gut microbial composition: a meta-analysis from a Malaysian perspective. Microb Genomics. 2021;7(8):000619.

    Article  CAS  Google Scholar 

  31. La-Ongkham O, Nakphaichit M, Nakayama J, Keawsompong S, Nitisinprasert S. Age-related changes in the gut microbiota and the core gut microbiome of healthy Thai humans. 3 Biotech. 2020;10:1–14.

    Article  Google Scholar 

  32. Singh H, Torralba MG, Moncera KJ, DiLello L, Petrini J, Nelson KE, Pieper R. Gastro-intestinal and oral microbiome signatures associated with healthy aging. Geroscience. 2019;41(6):907–21.

    Article  PubMed  PubMed Central  Google Scholar 

  33. Leite G, Pimentel M, Barlow GM, Chang C, Hosseini A, Wang J et al. Age and the aging process significantly alter the small bowel microbiome. Cell Rep. 2021;36(13).

  34. Biagi E, Nylund L, Candela M, Ostan R, Bucci L, Pini E, et al. Through ageing, and beyond: gut microbiota and inflammatory status in seniors and centenarians. PLoS ONE. 2010;5(5):e10667.

    Article  PubMed  PubMed Central  Google Scholar 

  35. Dong X, Guthrie BG, Alexander M, Noecker C, Ramirez L, Glasser NR, et al. Genetic manipulation of the human gut bacterium Eggerthella lenta reveals a widespread family of transcriptional regulators. Nat Commun. 2022;13(1):7624.

    Article  PubMed  CAS  PubMed Central  Google Scholar 

  36. Haiser HJ, Gootenberg DB, Chatman K, Sirasani G, Balskus EP, Turnbaugh PJ. Predicting and manipulating cardiac drug inactivation by the human gut bacterium Eggerthella lenta. Science. 2013;341(6143):295–8.

    Article  PubMed  CAS  PubMed Central  Google Scholar 

  37. Kageyama A, Benno Y. Catenibacterium mitsuokai gen. nov., sp. nov., a gram-positive anaerobic bacterium isolated from human faeces. Int J Syst Evol MicroBiol. 2000;50(4):1595–9.

    Article  PubMed  Google Scholar 

  38. Brahe LK, Le Chatelier E, Prifti E, Pons N, Kennedy S, Hansen T, et al. Specific gut microbiota features and metabolic markers in postmenopausal women with obesity. Nutr Diabetes. 2015;5(6):e159–e.

    Article  PubMed  CAS  PubMed Central  Google Scholar 

  39. Cui G, Li S, Ye H, Yang Y, Jia X, Lin M, et al. Gut microbiome and frailty: insight from genetic correlation and mendelian randomization. Gut Microbes. 2023;15(2):2282795.

    Article  PubMed  PubMed Central  Google Scholar 

  40. Schmidt TS, Raes J, Bork P. The human gut microbiome: from association to modulation. Cell. 2018;172(6):1198–215.

    Article  PubMed  CAS  Google Scholar 

  41. Whitfield C, Williams DM, Kelly SD. Lipopolysaccharide O-antigens—bacterial glycans made to measure. J Biol Chem. 2020;295(31):10593–609.

    Article  PubMed  CAS  PubMed Central  Google Scholar 

  42. Park BS, Lee J-O. Recognition of lipopolysaccharide pattern by TLR4 complexes. Exp Mol Med. 2013;45(12):e66–e.

    Article  PubMed  PubMed Central  Google Scholar 

  43. Wang Y, Ma M, Dai W, Shang Q, Yu G. Bacteroides salyersiae is a potent chondroitin sulfate-degrading species in the human gut microbiota. Microbiome. 2024;12(1):41.

    Article  PubMed  PubMed Central  Google Scholar 

  44. Zmuda AJ, Kang X, Wissbroecker KB, Freund Saxhaug K, Costa KC, Hegeman AD, Niehaus TD. A universal metabolite repair enzyme removes a strong inhibitor of the TCA cycle. Nat Commun. 2024;15(1):846.

    Article  PubMed  CAS  PubMed Central  Google Scholar 

  45. Martínez-Reyes I, Chandel NS. Mitochondrial TCA cycle metabolites control physiology and disease. Nat Commun. 2020;11(1):102.

    Article  PubMed  PubMed Central  Google Scholar 

  46. An R, Wilms E, Masclee AA, Smidt H, Zoetendal EG, Jonkers D. Age-dependent changes in GI physiology and microbiota: time to reconsider? Gut. 2018;67(12):2213–22.

    Article  PubMed  CAS  Google Scholar 

Download references

Acknowledgements

We would like to express our gratitude to all the volunteers in AMILI’s Poop Saves Life campaign for their participation in the research.

Funding

This study was funded by AMILI Pte Ltd, a privately-owned precision gut microbiome company based in Singapore, primarily financed by private investors and venture capital. The funding supported various aspects of the research, including recruitment, data collection, sample processing, DNA extraction, and sequencing.

Author information

Authors and Affiliations

  1. AMILI Pte Ltd, 89 Science Park Drive #03-09, The Rutherford, Lobby C, Singapore Science Park 1, Singapore, 118261, Singapore

    Kai Yee Toh, Priscilla Rajakumar & Jonathan Wei Jie Lee

  2. Department of Medicine, Faculty of Medicine, University of Malaya, Kuala Lumpur, Malaysia

    Tzi Shin Toh

  3. Pacific Biosciences of California, Menlo Park, CA, USA

    Khi Pin Chua

  4. Department of Medicine, Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore

    Jonathan Wei Jie Lee

  5. School of Pharmacy, Monash University Malaysia, Selangor, Malaysia

    Chun Wie Chong

  6. MUM Microbiome Research Centre, Monash University Malaysia, Selangor, Malaysia

    Chun Wie Chong

  7. Division of Gastroenterology and Hepatology, Department of Medicine , National University Health System, Singapore, 119228, Singapore

    Jonathan Wei Jie Lee

  8. iHealthtech, National University of Singapore, Singapore, 117599, Singapore

    Jonathan Wei Jie Lee

  9. SynCTI, National University of Singapore, Singapore, 117456, Singapore

    Jonathan Wei Jie Lee

Authors
  1. Kai Yee Toh
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Tzi Shin Toh
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Khi Pin Chua
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Priscilla Rajakumar
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Jonathan Wei Jie Lee
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Chun Wie Chong
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

CWC and JWJL conceptualized and designed the study. KYT acquired the data. TST and KPC processed the sequencing data and conducted the statistical analyses. TST and RPR wrote the first draft of the manuscript, and all authors read and approved the final manuscript.

Corresponding author

Correspondence to Kai Yee Toh.

Ethics declarations

Ethics approval

The study was approved by AMILI Institutional Review Board (reference number 2020/0501) in accordance with ethical guidelines.

Consent for publication

Not applicable.

Competing interests

KYT is an employee of AMILI Pte Ltd. JWJL is a co-founder of AMILI and serves as a member of the scientific advisory board. CWC and RPR are data consultants to AMILI. KPC is an employee of Pacific Biosciences. All other authors declare no financial or non-financial competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

13099_2024_650_MOESM1_ESM.pdf

Additional File 1: Stool collection instruction sheet (An instruction sheet provided to research participants, that explains the optimal method for collecting stool samples at home to ensure sample quality)

Additional File 2: Supplementary tables and figures

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Toh, K.Y., Toh, T.S., Chua, K.P. et al. Identification of age-associated microbial changes via long-read 16S sequencing. Gut Pathog 16, 56 (2024). https://doi.org/10.1186/s13099-024-00650-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s13099-024-00650-8

Keywords

Gut Pathogens

ISSN: 1757-4749