MICROBIOME
Robust variation in infant gut microbiome assembly
across a spectrum of lifestyles
Matthew R. Olm^1 †, Dylan Dahan^1 †, Matthew M. Carter^1 , Bryan D. Merrill^1 , Feiqiao B. Yu^2 , Sunit Jain^2 ,
Xiandong Meng^3 , Surya Tripathi^4 , Hannah Wastyk^1 , Norma Neff^2 , Susan Holmes1,5,
Erica D. Sonnenburg^1 , Aashish R. Jha^6 , Justin L. Sonnenburg1,2*
Infant microbiome assembly has been intensely studied in infants from industrialized nations, but little is
known about this process in nonindustrialized populations. We deeply sequenced infant stool samples
from the Hadza hunter-gatherers of Tanzania and analyzed them in a global meta-analysis. Infant
microbiomes develop along lifestyle-associated trajectories, with more than 20% of genomes detected
in the Hadza infant gut representing novel species. Industrialized infants—even those who are breastfed—
have microbiomes characterized by a paucity ofBifidobacterium infantisand gene cassettes involved
in human milk utilization. Strains within lifestyle-associated taxonomic groups are shared between mother-
infant dyads, consistent with early life inheritance of lifestyle-shaped microbiomes. The population-
specific differences in infant microbiome composition and function underscore the importance of
studying microbiomes from people outside of wealthy, industrialized nations.
T
he human gut microbiome undergoes a
complex process of assembly beginning
immediately after birth ( 1 ). New microbes
attempting to engraft within this com-
munity often depend upon niches estab-
lished by previous colonizing species and thus
the final adult microbiome composition may
be contingent upon the species acquired early
in life. The microbiome assembly process of
infants living in industrialized nations is well
studied and tends to follow a series of char-
acterized steps that lead to the low-diversity
gut microbiome composition characteristic
of industrialized adults ( 2 ). The microbiome
assembly process that occurs in infants living
nonindustrialized lifestyles (which results in
the characteristically diverse adult microbiomes
of nonindustrialized adults) ( 3 ) is largely un-
known ( 4 ). Of particular interest are the fol-
lowing: the timing at which the microbiomes
of infants from different lifestyles diverge, the
microbes and functions that are characteristic
of infants from different lifestyles, and whether
there are differences in the taxa that are ver-
tically transmitted from mothers to infants,
whichseedthemicrobiomeassemblyprocess.
To address these questions we performed
metagenomic sequencing on infant fecal samples
from the Hadza, a group of modern hunter-
gatherers in sub-Saharan Africa ( 5 , 6 ). The Hadza
inhabit seminomadic bush camps of ~5 to 30
people, exhibit a moderate level of commu-
nity child rearing within these camps ( 7 ), and
are breastfed early in life and weaned onto a
diet of baobab powder and premasticated
meat at ~2 to 3 years of age ( 8 , 9 ). In this study
we (i) curated and analyzed a global dataset of
1900 16S rRNA sequencing samples of healthy
infant fecal samples from 18 populations (in-
cluding 62 Hadza infant samples) ( 2 , 3 , 5 , 10 – 14 )
to contextualize the Hadza infant microbiome
(figs. S1 and S2), and (ii) performed deep meta-
genomic sequencing on 39 Hadza infant fecal
samples and corresponding maternal fecal sam-
ples for 23 infants in order to assess subspecies
variation, functional potential, and patterns of
vertical transmission (tables S1 and S2).
A UniFrac ordination created from all 1900
16S rRNA sequencing samples revealed age
and lifestyle to be strongly associated with the
first and second axes of variation, respectively
(Fig. 1A) (EnvFit;n = 1900;R^2 =0.43and0.50;
P = 0.001 and 0.001). Comparing populations
that practice different lifestyles within the
same country demonstrates that shared life-
style affects microbiota composition more than
geographic proximity (Fig. 1A, right panel, and
fig. S3). The microbiome of infants living in-
dustrialized lifestyles diverges from others
within the first 6 months of life, whereas the
microbiomes of infants living transitional
versus nonindustrialized lifestyles diverge at
~30 months of life (Fig. 1B). DNA extraction
methods, differences in feeding practices, or
other study-specific aspects may contribute
to some of the variation in data. Intermedi-
ate trajectories are exhibited by populations
on the boundaries of industrialized or non-
industrialized lifestyles (Fig. 1B, dashed lines),
highlighting the apparent sensitivity of infant
microbiota development to lifestyle-related
factors.
We identified five microbial coabundance
groups (CAGs) ( 15 , 16 ) in our dataset, which
together account for an average of 93.8% of themicrobiota composition per sample (Fig. 1C
and fig. S4). TheBifidobacterium-Streptococcus
CAG dominates infants from all lifestyles in
earlylife(0to6months),andovertimethis
CAG yields to theBacteroides-Ruminocccocus
gnavusCAG in industrialized infants and the
Prevotella-FaecalibacteriumCAG in infants
living transitional or nonindustrialized life-
styles (Fig. 1C). Lifestyle-related differences
in dominant CAGs become more pronounced
over time and mirror taxonomic trade-offs ob-
served in late infancy ( 17 ) and adulthood ( 5 ).
We next used our deep metagenomic se-
quencing data to assess microbiome-encoded
functional differences between lifestyles. Broad
lifestyle and age associated differences are seen
in the overall functional capacity of the infant
microbiomes (Fig. 2A), consistent with 16S rRNA
amplicon-based analysis (Fig. 1A). Hadza infant
metagenomes were assembled and binned into
metagenome-assembled genomes (MAGs) repre-
senting 745 species, 175 (23.4%) of which rep-
resent novel species compared to the Unified
Human Gastrointestinal Genome (UHGG) col-
lection ( 18 ) (table S3). Novel species were re-
covered from diverse phylogenetic groups (fig.
S5A); 88.6% (n = 155) were recovered from
multiple Hadza samples (fig. S5B) and their
genome quality was observed to be similar to
that of genomes in the UHGG (fig. S5C). To
assess prevalence through read mapping,
MAGs were integrated with genomes recov-
ered from Hadza adults ( 19 ) and public
genomes from the human gut ( 18 ) into a
comprehensive database of 5755 species-
representative genomes. Overall, 23.4% of
microbial species detected in the Hadza infants
represent novel species (table S4). These data
support that—similar to the adult Hadza gut—
the Hadza infant gut contains extensive pre-
viously uncharacterized diversity.
The taxonomic specificity afforded by meta-
genomic sequencing allowed us to identify
particular species that are depleted or en-
riched in infants living industrialized versus
nonindustrialized lifestyles. Identified among
the infants in this analysis were 310 VANISH
(Volatile and/or Negatively associated in In-
dustrialized Societies of Humans) and 12
BloSSUM (Bloom or Selected in Societies of
Urbanization/Modernization) species (table S5
and fig. S6). Comparison against a large database
of microbial species from nonhuman habitats
( 20 ) revealed that no VANISH and only one
BloSSUM species match genomes recovered
outsideofthedigestivetractorindustrialwaste-
water, whereas 21 VANISH and three BloSSUM
species match microbes recovered from non-
human animal feces (table S6). VANISH spe-
cies are more numerous and abundant than
BloSSUM (fig. S7), and 63 VANISH species are
effectively extinct (never detected) in infants
living industrialized or transitional lifestyles.
Many VANISH species (45.2%; 140 of 310) areRESEARCH
Olmet al., Science 376 , 1220–1223 (2022) 10 June 2022 1of4
(^1) Department of Microbiology and Immunology, Stanford
University School of Medicine, Stanford, CA, USA.^2 Chan
Zuckerberg Biohub, San Francisco, CA, USA.^3 ChEM-H
Institute, Stanford University, Stanford, CA 94305, USA.
(^4) Department of Plant and Microbial Biology, University of
California, Berkeley, Berkeley, CA, USA.^5 Department of
Statistics, Stanford University, Stanford, CA, USA.^6 Genetic
Heritage Group, Program in Biology, New York University
Abu Dhabi, Abu Dhabi,United Arab Emirates.
*Corresponding author. [email protected]
†These authors contributed equally to this work.