THE VAGINAL MICROBIOTA IN HEALTH AND DISEASE

JACQUES RAVEL INSTITUTE FOR GENOME SCIENCES UNIVERSITY OF MARYLAND SCHOOL OF MEDICINE

Let’s Talk MPT

CAMI Health/Initiative for MPTs/Public Health Institute

December 10, 2020 DISCLOSURES

I am the co-founder of LUCA Biologics, a biotechnology company focusing on translating microbiome research into live biotherapeutics drugs for women's health.

CST I CST V CST III VAGINAL MICROBIOTA

• In reproductive age women, Lactobacillus spp. are often characteristic of an optimal vaginal microbiota

• Lactobacillus spp. are lactic acid producers and acidify the vagina to pH < 4

• An acidic environment is thought to restrict the growth of non-indigenous microbes

• Changing environment throughout a women’s lifespan - microbiota changes accordingly CST I CST V CST III THE VAGINAL MICROBIOTA THROUGH THE LIFESPAN

Pre- Birth Menarche Reproductive Years Menopause menarche

CST I CST VLactobacillus CST III Estrogen ESTROGEN AND EPITHELIAL CELL MATURATION

Estrogen Lactobacillus Glycogen Sugars Estrogen THE VAGINAL MICROBIOTA THROUGH THE LIFESPAN

Pre- Birth Menarche Reproductive Years Menopause menarche

Lactobacillus

Estrogen VAGINAL MICROBIOTA COMPOSITION AND STRUCTURE

020406080 100 % taxon abundance Well-powered study of 400 asymptomatic women, equally representing 4 ethnic groups

CST II III I V IV Five major community state types that difer in their L. iners L. crispatus L. gasseri microbial composition and abundance L. jensenii Prevotella Megasphaera Community state type IV lacks signifcant number of Sneathia Atopobium Lactobacillus - higher diversity - (non optimal) Dialister Lachnospira Anaerococcus CST V Peptoniphilus Eggerthella Rhodobaca Anaerotruncus Ureaplasma Mycoplasma CST IV Aerococcus CST II Parvimonas Corynebacterium Veillonella L.vaginalis Community groups

CST I CST III Shannon Diversity Index 01234

Ravel et al. The vaginal microbiome of reproductive age women. PNAS. 2011. 108 Suppl 1, 4680–4687. COMMUNITY STATE TYPES

• Community State Types (CST) aford reducing the high dimensionality of microbiota compositional data

• CSTs represents sets of microbes, which together can be associated with a specifc outcomes

• CSTs carry biological information that can be interpreted

• CST-level analysis complements phylotype-level analysis which can be difcult to interpret as the context of the microbiota composition is not considered

• However, current CST classifcation is difcult to compare study to study, is highly sensitive to adding or removing samples, and CST are broadly defned

• Develop a tool to reproducibly classify vaginal microbiota into better defned CSTs VAGINAL MICROBIOTA COMPOSITION AND STRUCTURE

Community state type classifcation based on over 13,000 samples (16S rRNA gene V3-V4 regions)

Expanded high resolution classifcation and CST hierarchy

I II III IV V

I-A I-B II III-A III-B IV-A IV-B IV-C V

I-A I-B II III-A III-B IV-A IV-B IV-C0 IV-C1 IV-C2 IV-C3 IV-C4 V

Streptococcus Bifdobacterium Staphylococcus

VALENCIA: VAginaL community state typE Nearest CentroId clAssifer github.com/ravel-lab/VALENCIA

France M et al. Microbiome. 2020 FREQUENCY OF COMMUNITY TYPES IN ETHNIC GROUPS

Asian Caucasian

Hispanic Black IMPROVED RESOLUTION IN CST ASSIGNMENTS

Asian Caucasian

Hispanic Black THE DYNAMICS OF THE VAGINAL MICROBIOTA

100 100 160 women L crispatus L iners e (%) e (%) L iners 80

c Streptococcus c 80 L otu4 Corynebacterium L gasseri Finegoldia 60 60 L vaginalis L crispatus Daily sampling e abundan e abundan v v ti

ti Staphylococcus

a L crispatus 1 40 a 40 el el r Peptoniphilus L jensenii r Streptococcus Anaerococcus ype ype t 20 t 20 L gasseri ylo ylo h

h 10-week study

P L jensenii P 0 0 10 7 10 Nugent Score 7 3 Nugent Score 0 3 7 0 6 pH 7 5 6 pH 4 5 4

0 1 2345 6789 10 weekseeks 0 1 2345 6789 10 weeks

100 100 Atopobium vaginae Gardnerella vaginalis Gardnerella vaginalis BVAB2 e (%) 80 Others 80 c e (%) BVAB3 c Bifidobacteriaceae Megasphaera sp type 1 Atopobium vaginae Others BVAB2 60 60 Mobiluncus Lactobacillus iners e abundan

candidate division TM7 v e abundan

ti Aerococcus christensenii v a ti Parvimonas micra Megasphaera sp type 1 el a Peptoniphilus r 40 el 40 Lactobacillus crispatus r Bifidobacteriaceae

ype Bifidobacterium bifidum t

ype Anaerococcus tetradius t Lactobacillus jensenii Prevotella genogroup 1 ylo h ylo 20 20 P

h P Finegoldia magna Anaerococcus 0 0 10 10 7 7 Nugent Score Nugent Score 3 3 0 0 7 7 6 6 pH pH 5 5 4

012 345 6789 10 012 345 6789 10 weeks weeks

100 Gardnerella vaginalis 100 Gardnerella vaginalis Others Lactobacillus iners Lactobacillus crispatus Megasphaera sp type 2 Others Atopobium vaginae 80 Lactobacillus iners 80 Atopobium vaginae e (%) e (%) c c Bifidobacteriaceae Prevotella genogroup 1 BVAB2 Megasphaera sp type 2 Anaerococcus 60 Veillonella montpellierensis 60 BVAB1 Peptoniphilus BVAB2 e abundan e abundan

BVAB3 v Peptoniphilus v ti

ti Parvimonas micra Finegoldia magna a a

Peptostreptococcus el 40 BVAB3 r el 40 r Streptococcus Prevotella Aerococcus christensenii Clostridiales ype ype Finegoldia magna t

t Prevotella genogroup 2 Peptoniphilus lacrimalis ylo Anaerococcus tetradius

ylo 20 20 Lactobacillus gasseri h

h Aerococcus christensenii P P Megasphaera sp type 1 Bacteria Lactobacillus jensenii Streptococcus anginosus 0 Bacillales 0 Bacillales candidate division TM7 Streptococcus salivarius Streptococcus anginosus BVAB1 0 1 2345 6789 10 Streptococcus agalactiae 0 1 2345 6789 10 Prevotella bivia Mobiluncus curtisii weeks weeks class 10 10 7 Nugent Score 7 Nugent Score 3 3 0 0 7 7 6 pH 6 pH 5 5 4 4 THE DYNAMICS OF THE VAGINAL MICROBIOTA

Stability (rate of change) of the microbiota on time in the menstrual cycle analysis

Estrogen

Progesterone

The rate of community change is afected by time in the menstrual cycle (natural change) NON OPTIMAL VAGINAL MICROBIOTA NON OPTIMAL VAGINAL MICROBIOTA

• The ‘‘clinical iceberg’’ concept of adverse health outcomes, applied to BV. With better molecular methods, we now appreciate that clinically evident BV, as diagnosed by a technique such as Amsel’s criteria (Amsel-BV), does not capture a high proportion of women diagnosed with BV by Nugent (Nugent-BV) or with molecular methods (Molecular- BV) that contributes to adverse sexual and reproductive health outcomes, including increased HIV risk.

• There is a need to standardize our evaluation of non optimal vaginal microbiota in clinical studies HOW TO STUDY THE MICROBIOME

Diferent questions – DiferentC approaches

Who is there? (microbiota) Metataxonomic – relies on the sequencing of a marker genes ubiquitous to all bacteria and archea,C or all eukaryotes. Survey of the composition and abundance of these microbes. The most robust gene is the 16S rRNA gene and 18S rRNA gene for eukaryotes.

CST I CST V CST III THE 16S RIBOSOMAL RNA GENE

The ribosome comprises protein and structural RNAs. The 16S rRNA is part of the small subunit of the ribosome of bacteria and archaea.

It has conservedC and variable C regions. The conserved regions are useful for designing PCR primers and the variable regions useful for and phylogeny. CST I CST V CST III It is like a bacterial ID. HOW TO STUDY THE MICROBIOME

Who is there? (microbiota) Metataxonomic – relies on the sequencing of a marker genes ubiquitous to all bacteria and archea,C or all eukaryotes. Survey of the composition and abundance of these microbes. The most robust gene is the 16S rRNA gene and 18S rRNA gene for eukaryotes.

Comparison to CST I CST V sequenceCST III databases Bioinformatics METATAXONOMICS - COMPOSITIONAL/STRUCTURE

The relative abundance of each bacterial taxa is estimated

Combined with quantitative PCR of the 16S rRNA gene, estimates of absolute abundance can be obtained for each bacterial taxa HOW TO STUDY THE MICROBIOME

Diferent questions – DiferentC approaches

What can they do? (Metagenome) Metagenomics determines the microbiome genes and genomes content by sequencing total genomic DNA C Access composition and abundance of the microbiome Understand the functional potential of a microbiome from reconstructed genomes of membersCST I of the microbiotaCST V CST III METAGENOMICS - COMPOSITION/FUNCTION

Cost of metagenomics is almost equivalent to that of 16S rRNA gene sequencing

Metagenomics afords compositional/structural profling microbiota

Metagenomics afords functional profling of the microbiome

Less biased approach to compositional profling HOW TO STUDY THE MICROBIOME

Diferent questions – DiferentC approaches

What are they doing? (and the host) Metatranscriptomics - Determine the microbiome expressed genes content by sequencing cDNA

Metaproteomics – Determine theC protein content (identity) of a microbiome/host using mass spectrometry Metabolomics – Determine the metabolitesCST I compositionCST V of a CST III microbiome using mass spectrometry. The output of microbial/ host metabolisms METAGENOMICS IDENTIFIES HIGH WITHIN-COMMUNITY

INTRASPECIES DIVERSITY virgo.igs.umaryland.edu

Comparative analysis of 1,507 vaginal metagenomes and vaginal bacteria genomes Number of non-redundant genes Number of non-redundant

Genomes Metagenomes Metagenomes containing L. crispatus & genomes from isolates

Ma et al.. Nature Communications 2020

Genomes from isolates METAGENOMICS / METATRANSCRIPTOMICS

Ma et al.. Nature Communications 2020 IN SUMMARY…

Studying the microbiome is answeringC …

Who’s there? Metataxonomic profling (compositional/structural changes)

What are they doing? Functional proC fling (functional changes) What does this all mean? Complex statistical and modeling analyses CST I CST V CST III (Establishing statistical support for observed changes) ACKNOWLEDGEMENTS

github.com/ravel-lab/VALENCIA Bing Ma Michael France Johanna Holm Pawel Gajer Vonetta Edwards Steve Smith

virgo.igs.umaryland.edu