Priority effects, nutrition and milk glycan-metabolic potential drive Bifidobacterium longum subspecies dynamics in the infant gut microbiome

View article
Microbiology

Main article text

 

Introduction

Material and Methods

Amsterdam Infant Microbiome Study cohort.

Study population and sample collection

DNA extraction and sequencing

Metagenomics reads pre-processing

MAG reconstruction

Public metagenomic data acquisition

Bifidobacterial (sub)species-profiling

Bifidobacterial strain-tracking

B. longum meta-pangenome and phylogenomics tree

B. longum MAGs subspecies differentiation

Functional and metabolic enrichment analyses

Profiling of HMG-utilization potential

Modeling and selection of abundance-associated features

Quantification of priority effects

Data analysis and statistics

Data availability

Results

B. longum subspecies replacement between one and six months of age

Targeted glycoside hydrolase and transporter profile analysis reveals inter- and intra-subspecies variability in HMG-utilization potential

Maternal effects, nutrition and HMG-utilization potential drive B. longum subspecies abundance dynamics at one and six months of age

Identification of potential priority effects

Discussion

Conclusions

Supplemental Information

Sequencing depth of AIMS fecal samples

Overview of the number of reads retained per sample. For each bar, the colors indicate the proportion of reads that passed (green) and failed (red) quality control as well as reads mapping to human DNA (blue). Failed and human reads were subsequently filtered out the dataset.

DOI: 10.7717/peerj.18602/supp-1

Pangenomic analysis of B. longum subspecies associated with the maternal and infant gut

Each layer represents a B. longum MAG and the dark coloration within each layer indicates the abundance of a gene cluster (GC). The subsequent 8 layers correspond to various statistics related to the analysis i.e. the number of contributing genomes per GC, number of genes per GC and different levels of KEGG and COG functional annotations. Additionally, we report MAG’s total length, completeness, redundancy, GC total number and content and HMG-profile affiliation: BLinfantis-HMG (purple, number of genomes=14), BLlongum-HMGsimple (green,n=24), BLlongum-HMGcomplex (orange, n=25), reference genomes (red, n=4). The dendrogram above the layers indicates the similarities among genomes based on GC presence/absence. Two major clusters are visible: a BLinfantis (purple) and BLlongum (dark yellow) cluster.

DOI: 10.7717/peerj.18602/supp-2

Phylogenomics tree of B. longum subspecies associated to the maternal and infant gut

Phylogenomics tree based on concatenated alignment of 172 single-copy core genes from 63 B. longum MAGs, four B. longum reference genomes (BLlongum = 2, BLinfantis = 2) and one B. adolescentis genome (outer group) used in this study. Differentiation of B. longum subspecies was based on phylogenomics clustering, where MAGsclustering with a specific reference genome were subsequently assigned to the respective subspecies.

DOI: 10.7717/peerj.18602/supp-3

B. longum subspecies differentiation using the MetaPhlAn-B.infantis database of subspecies-specific marker genes (Ennis et al., 2024).

The heatmap shows the distribution of B. longum subspecies-specific marker genes (119 BLinfantis and 128 BLlongum; MetaPhlAn-B.infantis database) among 63 MAGs from AIMS and publicly available studies. Clustering of MAGs based on the copy-number of marker genes was used to differentiate BLlongum and BLinfantis MAGs and validate subspecies assignment achieved through our phylogenomics method.

DOI: 10.7717/peerj.18602/supp-4

Comparison of BLlongum abundances estimated using sylph v.0.6.1 and MetaPhlAn v4.0.6 with the amended MetaPhlAn-B.infantis database

The correlation dot plot shows the relative abundances of BLlongum (orange) and BLinfantis (light blue) detected in AIMS infants and estimated using sylph v.0.6.1 and MetaPhlAn v4.0.6 with the amended MetaPhlAn-B.infantis database.

DOI: 10.7717/peerj.18602/supp-5

AIMS infant gut microbiota composition at 1 and six months of age

The bargraphs summarize the relative abundances of most prevalent and abundant Bifidobacterium spp. and bacterial phyla identified in metagenomics stool samples of A) 1 and B) six months old infants from the AIMS cohort.

DOI: 10.7717/peerj.18602/supp-6

Homology heatmap of GH2-encoding genes

The heatmap shows the clustering of GH2-encoding genes based on their sequence alignment identity (%), obtained from multiple pairwise protein sequence alignments. Sequence clusters were considered to represent the same GH variant (‘final_annotation’) if their average alignment identity was ≥ 60%.

DOI: 10.7717/peerj.18602/supp-7

Homology heatmap of GH20-encoding genes

The heatmap shows the clustering of GH20-encoding genes based on their sequence alignment identity (%), obtained from multiple pairwise protein sequence alignments. Sequence clusters were considered to represent the same GH variant (’final_annotation’) if their average alignment identity was ≥ 60%.

DOI: 10.7717/peerj.18602/supp-8

Homology heatmap of GH29-encoding genes

The heatmap shows the clustering of GH29-encoding genes based on their sequence alignment identity (%), obtained from multiple pairwise protein sequence alignments. Sequence clusters were considered to represent the same GH variant (’final_annotation’) if their average alignment identity was ≥ 60%.

DOI: 10.7717/peerj.18602/supp-9

Homology heatmap of GH33-encoding genes

The heatmap shows the clustering of GH33-encoding genes based on their sequence alignment identity (%), obtained from multiple pairwise protein sequence alignments. Sequence clusters were considered to represent the same GH variant (’final_annotation’) if their average alignment identity was ≥ 60%.

DOI: 10.7717/peerj.18602/supp-10

Homology heatmap of GH38-encoding genes

The heatmap shows the clustering of GH38-encoding genes based on their sequence alignment identity (%), obtained from multiple pairwise protein sequence alignments. Sequence clusters were considered to represent the same GH variant (’final_annotation’) if their average alignment identity was ≥ 60%.

DOI: 10.7717/peerj.18602/supp-11

Homology heatmap of GH42-encoding genes

The heatmap shows the clustering of GH42-encoding genes based on their sequence alignment identity (%), obtained from multiple pairwise protein sequence alignments. Sequence clusters were considered to represent the same GH variant (’final_annotation’) if their average alignment identity was ≥ 60%.

DOI: 10.7717/peerj.18602/supp-12

Homology heatmap of GH85-encoding genes

The heatmap shows the clustering of GH85-encoding genes based on their sequence alignment identity (%), obtained from multiple pairwise protein sequence alignments. Sequence clusters were considered to represent the same GH variant (’final_annotation’) if their average alignment identity was ≥ 60%.

DOI: 10.7717/peerj.18602/supp-13

Homology heatmap of GH112-encoding genes

The heatmap shows the clustering of GH112-encoding genes based on their sequence alignment identity (%), obtained from multiple pairwise protein sequence alignments. Sequence clusters were considered to represent the same GH variant (’final_annotation’) if their average alignment identity was ≥ 60%.

DOI: 10.7717/peerj.18602/supp-14

Identification of potential priority effects

(A) Bargraphs summarize the relative abundances of BLlongum in the gut of one and six-month-old infants from AIMS and publicly available datasets used in this study, based on i) ‘Mother-Infant Strain Sharing’, ii) ‘BLlongum (mother or 1 month, %) ¿other bifidobacteria (mother or 1 month, %)’ and iii) if either i) or ii) are true. Significance between groups was calculated using Wilcoxon’s signed-rank tests: * p < 0.05, ** p < 0.01, ns = not significant. (B) Strength of priority effect quantified using the method by Vannette & Fukami (2014) and estimated for maternal factors identified in this study as proxies for priority effects. Negative values indicate that priority effects are in place for a specific proxy.

DOI: 10.7717/peerj.18602/supp-15

Abundance of shared and non-shared BLlongum in the gut of mother

The boxplots summarize the relative abundances of BLlongum in the gut of mothers based on strain sharing occurrence (yes = light blue, no = yellow) from AIMS and publicly available datasets used in this study. Shared B. longum strains were not significantly (= ns) more abundant than non-shared strains in the gut of mothers (p-value = 0.08507).

DOI: 10.7717/peerj.18602/supp-16

Overview summary statistics of mothers and their infants from AIMS and public datasets used in this study

DOI: 10.7717/peerj.18602/supp-17

List of B. longum subspecies-specific marker genes found in the MetaPhlAn-B.infantis database (Ennis et al., 2024) and their prevalence among B. longum MAGs used in this study

DOI: 10.7717/peerj.18602/supp-18

Overview of the high-quality B. longum metagenome-assembled genomes (MAGs) selected for this study

DOI: 10.7717/peerj.18602/supp-19

Overview Bifidobacterium spp. strain tracking output (StrainPhlAn4)

DOI: 10.7717/peerj.18602/supp-20

GH-encoding gene annotation per MAG

DOI: 10.7717/peerj.18602/supp-21

Overview of transporter-encoding reference sequences used for characterizing B. longum HMG-utilization profiles

This modified version of Table S1 from Arzamasov & Osterman (2022) shows an overview of transporters used for characterizing B. longum HMG-utilization profiles and the reference sequences employed for protein homology searches.

DOI: 10.7717/peerj.18602/supp-22

GH/transporter-encoding gene copy number per MAG used in this study

DOI: 10.7717/peerj.18602/supp-23

Results of pairwise Wilcoxon signed-rank tests comparing the gene copy number of GH/transporter-encoding genes among HMG-utilization clusters

DOI: 10.7717/peerj.18602/supp-24

Output of the functional enrichment analysis between MAGs belonging to the BLlongum-HMGcomplex and BLlongum-HMGsimple clusters

DOI: 10.7717/peerj.18602/supp-25

Output of the functional enrichment analysis between BLlongum and BLinfantis subspecies

DOI: 10.7717/peerj.18602/supp-26

Output of the metabolic enrichment analysis performed between BLlongum and BLinfantis subspecies

DOI: 10.7717/peerj.18602/supp-27

Overview of the collinearity analysis output performed to assist feature selection for our LASSO models

DOI: 10.7717/peerj.18602/supp-28

Overview LASSO feature selection

DOI: 10.7717/peerj.18602/supp-29

Overview important coefficients per LASSO regression model

DOI: 10.7717/peerj.18602/supp-30

AIMS gut microbiome data

Matrix containing the relative abundances of bacteria and archaea species detected in AIMS gut samples

DOI: 10.7717/peerj.18602/supp-31

BLlongum maternal abundance and vertical transmission events

The table contains BLlongum abundance and information of vertical transmission events for all maternal samples available. This data was used to make Fig. S16.

DOI: 10.7717/peerj.18602/supp-32

Overview of BLlongum abundances at one and six months and ’Early arrival’ groups

The table reports BLlongum abundances at one and six months of ages. For each infant, we also report information on the maternal factors identified in this study as proxies for priority effects: (i) mother/infant strain sharing, (ii) BLlongum (mother or 1mnt, %) ¿other bifidobacteria (mother or 1mnt, %) and (iii) instances where either (i) or (ii) or both are true. This data was used to make Figure S15.

DOI: 10.7717/peerj.18602/supp-33

Results of Wilcoxon signed-rank tests comparing BLlongum abundances between ’Early Arrival of BLlongum: Yes’ and ’Early Arrival of BLlongum: No’ groups

DOI: 10.7717/peerj.18602/supp-34

Additional Information and Declarations

Competing Interests

The authors declare there are no competing interests.

Author Contributions

Nicholas Pucci conceived and designed the experiments, performed the experiments, analyzed the data, prepared figures and/or tables, authored or reviewed drafts of the article, and approved the final draft.

Joanne Ujčič-Voortman conceived and designed the experiments, authored or reviewed drafts of the article, resources, Project administration, and approved the final draft.

Arnoud P. Verhoeff conceived and designed the experiments, authored or reviewed drafts of the article, resources, Project administration, Funding acquisition, and approved the final draft.

Daniel R. Mende conceived and designed the experiments, performed the experiments, analyzed the data, prepared figures and/or tables, authored or reviewed drafts of the article, funding acquisition, and approved the final draft.

Human Ethics

The following information was supplied relating to ethical approvals (i.e., approving body and any reference numbers):

Methods for sampling and questionnaires received approval by the Medical Ethical Examination Commission of the Amsterdam University Medical Center (METC Amsterdam UMC, Reference number: NL64399.018.17). All participants have provided written consent. Written consent for infants was provided by parents after birth.

Field Study Permissions

The following information was supplied relating to field study approvals (i.e., approving body and any reference numbers):

Methods for sampling and questionnaires received approval by the Medical Ethical Examination Commission of the Amsterdam University Medical Center (METC Amsterdam UMC, Reference number: NL64399.018.17).

Data Availability

The following information was supplied regarding data availability:

The dataset supporting the conclusions of this article is available both at the European Nucleotide Archive (ENA) PRJEB66728, and NCBI: 1050518, PRJEB66728.

Funding

This project was supported by the GGD Amsterdam, the University of Amsterdam - Research Priority Area Personal Microbiome Health (RPA-PMH), the Stichting Orale Biologie and the Dutch Research Council (NWO) - MetaHealth project (NWA.1389.20.080). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

MIT

Your institution may have Open Access funds available for qualifying authors. See if you qualify

Publish for free

Comment on Articles or Preprints and we'll waive your author fee
Learn more

Five new journals in Chemistry

Free to publish • Peer-reviewed • From PeerJ
Find out more