Paper - Review

DOI: 10.1038/s41591-019-0450-2

Abstract

Preterm birth
→ about 10% worldwide
→ significant disparities ← among (population within countries & women)

∴ Report (community resource)
← which includes "omics" data from 12K samples

∴ 1⃣ Lower → Lactobacillus crispatus
2⃣ Higher → BVAB1, Sneathia amnii, TM7-H1

Introduction

Preterm birth (PTB)
← at less than 37 weeks of gestation
→ (second most common cause) ← of neonatal death
→ (most common cause) ← of infant mortality ← in high-income economies
→ NO ❌ effective strategies → for (predicting & preventing) PTB

Microbe-induced inflammation
← from 1⃣ urinary tract infection 2⃣ sexually transmitted infections
→ can be a cause of PTB

Homogeneous Lactobacillus-dominated microbiome
→ considered (the hallmark of health) ← in the female reproductive tract

Vaginal microbiome
← is with high species diversity
→ risk for (acquisition & transmission) of 1⃣ sexually transmitted infections 2⃣ PTB 3⃣ Pelvic inflammatory disease

❗Composition of (vaginal microbiome) → significant (population-specific impact) on PTB risk
❗ Lactobacillus crispatus → with lower risk of PTB

Report
1⃣ (a community resource) samples ← collected longitudinally during 1572 pregnancies of women ← from diverse ancestries
2⃣ an analysis (← of the longitudinal, comprehensive, multi-omic profiling) of vaginal samples from 45 women

Results

The Multi-Omic Microbiome Study: Pregnancy Initiative

Longitudinal iHMP study & Mutli-omic microbiome study
← total 1572 pregnancies
→ 2 comprehensive datasets (← of integrated microbiome & host functional properties)
1⃣ MOMS-PI PTB ← case-control study of 45 women predominantly of African ancestry & 90 case-matched women who delivered at term
2⃣ MOMS-PI Term Birth ← ethnically diverse retrospective cohort study of 90 women
∴ Selection 12039 samples from 597 pregnancies
1⃣ 16S rRNA taxonomic profiles ← from (6452 pregnant sample) & (2753 neonate samples)
2⃣ Meta-genome profiles ← from (930 pregnant & 146 neonates)
3⃣ Meta-transcriptome profiles ← from (297 pregnant)
4⃣ Cytokine profiles ← from (1223 pregnant) & (173 neonates)
5⃣ Lipid profiles ← from (63 pregnant)]

Vaginal microbiome profiles show PTB-associated trends

Comprehensive multi-omic profiling (← of vaginal samples) (← in the MOMS-PI PTB study)

Cohort
← which predominantly comprised women of African ancestry
← median annual income of (less than $20000)
← an average of 26 years

Microbial profiles (← of the first vaginal samples ) collected at enrollment
← were generated by 16S rRNA taxonomic analysis

Women
← who went on to deliver at term
→ were more likely to exhibit L. cripatus predominance

Overall diversity was increased
← from women who would go on to experience PTB

12 taxa showed (a significant difference in abundance)
← between the PTB vs. TB group
1⃣ L. crispatus was greatly reduced in PTB samples
2⃣ (BVAB1 & Sneathia amnii) → were more abundant in PTB samples
3⃣ (Megasphaera type 1) & (TM7-H1) → were significantly increased in PTB samples
∴ Association of TM7-H1 with PTB

Early prediction (← of risk for PTB) is critical
← for the development of new strategies for (prevention & intervention)
∴ 16S rRNA data (← from collected at 24 weeks of gestation or earlier)

❗ Examined longitudinal trends (← of key taxa)
← using a GAMM (Generalized additive mixed effect model)
← incorporating (BMI & pH & ethnicity & preterm status)
→ Many taxa identified as associated with PTB
e.g. 1⃣ S. amnii 2⃣ Prevotella cluster 2 3⃣ BVAB1 3⃣ P amnii

Women ← who delivered preterm
→ experienced (large decrease) during pregnancy
→ in 1⃣ S. amnii 2⃣ BVAB1 3⃣ P. amnii 4⃣ Gardnerella vaginalis 5⃣ TM7-H1 6⃣ Atopobium vaginae

❗ Stratification (← of the longitudinal GAMM analysis) (← by ancestry)

Women (← of African ancestry) (← who delivered preterm experienced) (significant decreases) in the prevalence of 1⃣ A. vaginae 2⃣ BVAB1 3⃣ G. vaginalis 4⃣ P. amnii 5⃣ S. amnii 6⃣ TM7-H1

Women (← of African ancestry) (← who delivered at term)
→ decreases in 1⃣ A. vaginae 2⃣ G. vaginalis
→ increase in 1⃣ L. iners

∴ Dynamic changes (← in the vaginal microbiome) ← in pregnancy
∴ Dynamics (← of vaginal microbiome) differ by ancestry

(Functional & metabolic) potentials
← of the microbial communities
→ largely conserved
→ exhibited (a much higher proportional metabolic potential & transcriptional activity) (← of UDP-N-acetyl-D-glucosamine)

Microbe associated molecular patterns
← produced by the vaginal microbiota
→ modulate host cytokine levels & impact urogenital health

Metagenomic assembly of reference genomes of bacterial taxa associated with PTB

MGS data
← generated with (Pacific BioSciences & Illumina sequencing) technologies
→ to generated the first genome of (TM7-H1 & BVAB1)

BVAB1
← with genome of up to 1.45 Mb
→ is classified to (Lachnospiraceae Clostridiales)
→ is NOT ❌ closely related → to any other known bacterium

TM7H1
→ encodes a putative α-amylase
→ is predicted to be able to utilize glycogen
→ lacks ❌ (de novo bio-synthetic capabilities) for essential amino acids

∴ We identified (243 & 421) metabolic reactions ← in (TM7-H1 & BVAB1)

Both organisms are predicted → to have the ability
→ to produce (pyruvate & acetate & L-lactate & propionate)

Bacterial taxa associated with PTB in metagenomic and metatranscriptomic data

(95% of MGS reads) & (30% of MTS reads)
→ were identified as human
∴ Only small fraction remaining unmapped

Abundance of taxa
← from the 16S rRNA assay
← with the relative proportional abundance of (metagenomic and metatranscriptomic data)

Paired MGS & MTS data ← 41 preterm & 81 term controls
∴ Analyzed (a single time point per participant) & Used (a global scaling approach) to normalize → to all gene in the 56 taxa

G. vaginalis gene
← analyzed using MGS data
→ were significantly higher ↑ in the term cohorts
→ Overall relative (transcriptional rate) (← of G. vaginalis) was higher ↑ (← in preterm samples)

Previous studies suggested that
→ PTB risk differs with different G. vaginalis clades
→ PTB risk differs with (transcriptional activity) (← of G. vaginalis)
❓ Mechanisms (← which impact replication & transcription) (← of different strains of G. vaginalis)
❓how (these mechanisms ↑) affect women's (reproductive health & pregnancy)

Differences (← between the PTB & TB samples)
→ reflect the elevated (proportional abundance) (← of these taxa)
❓ How vaginal microbiome → may (mediate & cause) to (pathology & PTB)

Host cytokine expression in PTB

Pass. Not my business

Cross-study comparisons of the vaginal microbiome and preterm birth

Limited correlation
← between (vaginal microbiome & PTB)
← in African cohorts

Compared distribution (← of distinct candidates taxa)
← for PTB risk across four studies (← of vaginal microbiome)

Non-trivial differences
← in every technical aspect (← of study design)
e.g. 1⃣ sample collection 2⃣ DNA extraction 3⃣ PCR primers 4⃣ sequencing platform 5⃣ data quality
→ which precluded (an integrated analysis) (← of these datasets)

PTB
→ is a syndrome (← with many underlying causes)
→ Microbes (← from vagina) → 1⃣ likely plays a causative role in PTB 2⃣ likely does NOT ❌ play a contributing role in all PTB
∴ Considerable differences ← in PTB case cohorts

(Examined ONLY ∃! the spontaneous PTB cases) & (Rematched controls 1:2)
Found (partial support) → for candidate taxa
← identified in the present study

Discussion

Generate → (omics data)
← (the largest & most comprehensive) multi-omic studies
→ showed (a signature of PTB) in African ancestry

Taxa (← associated with PTB)
→ tended to decrease in abundance in the vaginal environment
∴ (Microbiome composition) (← early in pregnancy)
→ may be most useful in the prediction (← of adverse outcomes)

---