the dynamics of the microbiome: the human vagina · pdf filethe human vagina the dynamics of...

12
Jacques Ravel Institute for Genome Sciences Dept. of Microbiology and Immunology University of Maryland School of Medicine Human Microbiome Science:Vision for the Future - July 24, 2013 The Dynamics of the Microbiome: The Human Vagina The dynamics of microbial communities can be defined as temporal and spatial changes in community structure or function Definitions Dynamics

Upload: lytu

Post on 28-Mar-2018

215 views

Category:

Documents


2 download

TRANSCRIPT

Jacques RavelInstitute for Genome Sciences

Dept. of Microbiology and ImmunologyUniversity of Maryland School of Medicine

Human Microbiome Science: Vision for the Future - July 24, 2013

The Dynamics of the Microbiome:

The Human Vagina

The dynamics of microbial communities can

be defined as temporal and spatial changes in

community structure or function

DefinitionsDynamics

• In 1997 Grimm and Wissel catalogued 167 definitions of community

stability in the field of ecology

DefinitionsStability/Instability

V. Grimm and C. Wissel. Babel, or the ecological stability discussions: an inventory and analysis of terminology and a guide for avoiding confusion. Oecologia, 109:323–334, 1997.

• Stability quantifies the extend to which a community “stays the same”

over a long period of time (depending on the time scale) or in the face

of some disturbances.

resistance (or constancy) – staying essentially unchanged over time

resilience – returning to the reference state after a disturbance

persistence – persistence of an ecological state through time.

• Microbiota compositional change (16S rRNA gene survey) - Relative abundance

• Microbiota compositional change (species specific qPCR - pan-bacterial qPCR)

• Estimates of diversity (evenness and richness, relative entropy)

• Metagenomic characterization (microbiome: genes/genomes content, SNP)

• Functional analysis of metagenomic data (metabolic potential)

• Metabolomics (functional output of a community)

• Metatranscriptomics/metaproteomics (function of a community)

• Culture/physical measurement (i.e., pH), microscopy

DefinitionsMeasures used to study stability/instability

Communities can appear stable given one metrics and unstable given another one.

McNulty, N. P., Yatsunenko, T., Hsiao, A., Faith, J. J., Muegge, B. D., Goodman, A. L., et al. (2011). The impact of a consortium of fermented milk strains on the gut microbiome of gnotobiotic mice and monozygotic twins. Science translational medicine, 3(106), 106ra106.

• A disturbance may be defined as an event that disrupts community structure and/or

simultaneously changes resources, substrate availability, or the physical environment.

DefinitionsDisturbance/Resilience

Putman, R.J. (1994). Community Ecology (Chapman and Hall, New York).Ma, Z. S., Geng, J. W., Abdo, Z., & Forney, L. J. (2012). A Bird’s Eye View of Microbial Community Dynamics. Microbial Ecology Theory: Current Perspectives. Norwich, UK: Horizon Scientific Press, 57–70.

• A disturbance can also be defined as a killing, displacement, or damage to one or more

populations in a community that directly or indirectly creates opportunities for new

individuals to become established.

• Resilience is the amount of disturbance that an ecosystem can withstand without

changing its self-organizing processes.

• Communities with fundamental differences in species composition and structure will

differ in resilience.

Inte

nsity

Frequency or Duration

DisturbedEcosystem

Ecosystem disturbances/resilience

Increased Risk to Disease?Disease itself? Cause/effects?

Disturbances to the microbiome are often imposed by human actions:- hygiene- diet- behaviors- antibiotics

But also include: - natural states (menstruation)- hormonal variations - immunological changes

What are the drivers of change and resilience in a microbial community?

•While key studies, most looked at either low number of time points (short

sampling period, long period between samples) or very few subjects

(sampled frequently) - Difficulties in generalizing the findings/conclusions

• Feasibility and cost have limited these studies

Issues with current study

•Compositional analysis often done at low taxonomic resolution (Phylum,

Class, Order)

• Principles of community dynamics are certainly body site specific and not

applicable across the spectrum of community types on the human body

The dynamics of the vaginal microbiota in reproductive age women

The Vaginal Community Space

A plot of principle component analysis (PCA) shows distribution of community states in 3-D space.

Graphics generated with inVUE (inVUE.sourceforge.net)

CST III(L. iners)

CST I(L. crispatus)

CST V(L. jensenii)

CST II(L. gasseri)

CST IV

• Five community state types (CST) that differ in their microbial composition and abundance

• Community state type IV lacks significant number of Lactobacillus - higher diversity - (IV-A and IV-B)

• Cross sectional study of 410 asymptomatic healthy women

Longitudinal studies of the vaginal microbiotaStudy design

160 women cohort - prospective longitudinal study

• Self-collect vaginal smears, pH and swabs daily for 10 weeks

• Three swabs a day: Amies, RNAlater and dry

• Daily diary and weekly clinic visits

• Clinical assessment at enrollment, week 5 and 10

• Determine bacterial community composition by 454 pyrosequencing of

barcoded V1-V3 16S rRNA gene of 50 women

Longitudinal profiles - stability/instability

0

20

40

60

80

100 Gardnerella vaginalisLactobacillus inersLactobacillus crispatusOthersAtopobium vaginaeBifidobacteriaceaeBVAB2Megasphaera sp type 2Veillonella montpellierensisPeptoniphilusBVAB3Parvimonas micraPeptostreptococcusStreptococcusAerococcus christenseniiFinegoldia magnaPeptoniphilus lacrimalisLactobacillus gasseriMegasphaera sp type 1Lactobacillus jenseniiStreptococcus anginosusBacillalesStreptococcus salivariusBVAB1Streptococcus agalactiaeMobiluncus curtisii

037

10Nugent Score

4567

pH

weeks0 1 2 3 4 5 6 7 8 9 10

Phyl

otyp

e re

lativ

e ab

unda

nce

(%)

0

20

40

60

80

100 Gardnerella vaginalisOthersMegasphaera sp type 2Atopobium vaginaeLactobacillus inersPrevotella genogroup 1Streptococcus agalactiaeAnaerococcusBVAB1BVAB2PeptoniphilusFinegoldia magnaBVAB3PrevotellaClostridialesPrevotella genogroup 2Anaerococcus tetradiusAerococcus christenseniiBacteriaPeptostreptococcusBacillalescandidate division TM7Streptococcus anginosusEnterococcus faecalisPrevotella biviaActinobacteria class

037

10Nugent Score

4567

pH

weeks0 1 2 3 4 5 6 7 8 9 10

Phyl

otyp

e re

lativ

e ab

unda

nce

(%)

0

20

40

60

80

100

L crispatusL inersL otu4L gasseriL vaginalisClostridiumL crispatus 1L jenseniiStreptococcusGemella

Nugent Score

pH

Phyl

otyp

e re

lativ

e ab

unda

nce

(%)

037

10

4567

weeks0 1 2 3 4 5 6 7 8 9 10

Nugent Score

pH

0

20

40

60

80

100

L inersStreptococcusCorynebacteriumFinegoldiaL crispatusStaphylococcusPeptoniphilusAnaerococcusL gasseriL jensenii

Phyl

otyp

e re

lativ

e ab

unda

nce

(%)

037

10

4567

weeks0 1 2 3 4 5 6 7 8 9 10

0

20

40

60

80

100

Gardnerella vaginalisOthersBifidobacteriaceaeAtopobium vaginaeBVAB2Lactobacillus inersAerococcus christenseniiMegasphaera sp type 1Lactobacillus crispatusBifidobacterium bifidumLactobacillus jensenii

037

10Nugent Score

4567

pH

weeks0 1 2 3 4 5 6 7 8 9 10

Phyl

otyp

e re

lativ

e ab

unda

nce

(%)

0

20

40

60

80

100Atopobium vaginaeGardnerella vaginalisBVAB2BVAB3Megasphaera sp type 1OthersMobiluncuscandidate division TM7Parvimonas micraPeptoniphilusBifidobacteriaceaeAnaerococcus tetradiusPrevotella genogroup 1BacteriaFinegoldia magnaAnaerococcus

037

10Nugent Score

567

pH

weeks0 1 2 3 4 5 6 7 8 9 10

Phyl

otyp

e re

lativ

e ab

unda

nce

(%)

weeks

Modeling the dependence of the log of Jensen-Shannon divergence rate of change (estimate of stability) on the menstrual time (normalized time) - Bayesian Markov Chain Monte Carlo methods using linear mixed effect models.

Drivers of instability

Gajer et al. The temporal dynamics of the vaginal microbiota. Science Translational Medicine. 2012. 4(132): 132ra52.

Menstrual cycle (days)

log(6

DJS

6t)

0 7 14 21 28

��

���

���

���

���

��

100

���

200

���

Con

cent

ratio

n of

Pro

gest

eron

e [4

x] a

nd E

stra

diol

[pg/

ml]

Menstrual cycle (days)

log(6

DJS

6t)

0 7 14 21 28

��

���

���

���

���

��

100

���

200

���

Con

cent

ratio

n of

Pro

gest

eron

e [4

x] a

nd E

stra

diol

[pg/

ml]

Estrogen

Progesterone

The rate of community change is affected by time in the menstrual cycle (natural change) and sexual activities to some extend

Vaginal community dynamics

pH, Nugent score, microbial community state

0-3

4-6

7-10

Nugent Score

CST III(L. iners)

CST I(L. crispatus)

CST V(L. jensenii)

CST II(L. gasseri)

CST IV

4.0-4.5

4.6-5.0

5.1-5.5

>5.5

pH

CST III(L. iners)

CST I(L. crispatus)

CST V(L. jensenii)

CST II(L. gasseri)

CST IV

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

In these healthy, asymptomatic women, CST IV is associated with higher Nugent scores and higher pH

•At any given time, >25% of women are in a non-Lactobacillus dominated state.

CST IV: an normal state that carry risks?

• This state is associated with higher Nugent scores and higher pH

• High Nugent score is strongly associated with increased risk of

sexually transmitted infection acquisition and transmission, including

HIV, as well as preterm birth

• These women are asymptomatic and apparently healthy, but

potentially at increased risk of STI or other adverse outcomes

Wiesenfeld HC, et al. : Bacterial vaginosis is a strong predictor of Neisseria gonorrhoeae and Chlamydia trachomatis infection. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 2003, 36(5):663-668.Taha TE, et al. : Bacterial vaginosis and disturbances of vaginal flora: association with increased acquisition of HIV. AIDS (London, England) 1998, 12(13):1699-1706.Cohen, C. R. et al. (2012). Bacterial Vaginosis Associated with Increased Risk of Female-to-Male HIV-1 Transmission: A Prospective Cohort Analysis among African Couples. (H. Watts, Ed.)PLoS Medicine, 9(6), e1001251.

There are windows of higher risk that open and close on a temporal scale

CST IV: an healthy state that carry risks?

Vaginal community dynamics

Community stability/dynamics (frequency and duration of CST IV) might better

represent risks to disease

Low resilience = increased risk

Windows of higher risk that open and close on a temporal scale

Understand the molecular basis of the association between stability and susceptibility

Still, we do not understand the underlying causes for stability/dynamics (frequency and duration of CST IV)

Vaginal community dynamicsWindows of higher risk that open and close on a temporal scale

Stability and the community genome

Is there a correlation between the gene content of some of the species and community stability?

Use metagenomics analysis of microbial communities

Establish community composition and assemble the genomes of community members

Perform comparative genome analysis with known genomes

Metagenomics analysis: L. iners genomicsLactobacillus_iners

Gardnerella_vaginalis

Lactobacillus_jensenii

Lactobacillus_crispatus

Atopobium_vaginae

Bifidobacterium_unclassified

Prevotella_bivia

Prevotella_timonensis

Lactobacillus_gasseri

Streptococcus_agalactiae

Prevotella_buccalis

Prevotella_amnii

Peptoniphilus_unclassified

Streptococcus_mitis

Anaerococcus_tetradius

Lactobacillus_brevis

Peptoniphilus_harei

Bacteroides_unclassified

Candidatus_Zinderia_unclassified

Dialister_microaerophilus

Finegoldia_magna

Corynebacterium_amycolatum

Anaerococcus_unclassified

Mobiluncus_mulieris

Peptostreptococcus_anaerobius

Candidatus_Carsonella_ruddii

Leuconostoc_unclassified

Lactobacillus_vaginalis

Ureaplasma_parvum

Mycoplasma_hominis

CH

ARM

198_V2445_020507C

HAR

M110_V 2

CH

ARM

144_V1C

HAR

M180_V2

CH

ARM

170_V2C

HAR

M154_V2

CH

ARM

170_V1414_050806C

HAR

M066_V 1

CH

ARM

066_V2H

MP40_W

6D2

CH

ARM

154_V 1H

MP40_W

5D1

HM

P40_W3D

6H

MP40_W

1D2

HM

P40_W3D

7C

HAR

M068_V 1

HM

P40_W4D

2C

HAR

M148_V 1

EM10W

6D7

VM077

HM

P10_W3D

7H

MP10_W

4D7

VM009

VM177

CH

ARM

140_V 2C

HAR

M148_V2

CH

ARM

140_V1EM

O9W

9D3

EM12W

4D3

CH

ARM

014_V 1VM

255401_030106AYAC

05W3D

7401_022506C

HAR

M150_V 1

CH

ARM

150_V2

L. iners and community stability

60.7

65.6

52.7

94.8

100

100

100 100

66.4

100

100100

100

100

82

100

100

83.556.6

64.8

74.1

82.8

88.5

85

100

100 87.1

89.1

100100

100

100

• Certain kinds of L. iners genomes appear associated with community resilience

to enter CST IV

Drivers community stability

• Does the “fragility” of a single keystone species drive the community into

dysbiotic states that carry risk to disease?

• We hypothesize that vaginal microbiome stability/dynamics (frequency and

duration of CST IV) might better represent estimates of risks to infection and

that stability/instability can be driven by the lack of resilience of certain

keystone Lactobacillus sp.

Gaps and Challenges

• The issue of cause or effects - Need for prospective longitudinal studies

• Better understand the drivers of dynamics/stability or instability in health (risk to disease) over

a women’s lifespan

• Develop predictive models of stability/instability/susceptibility that accounts for health practices

and behaviors.

• Need to evaluate the dynamics using different measures (metabolomics, metatranscriptomics

[host and microbes] and immunological)

• Expand the studies to dynamics of phages, viruses and fungi

• Understand what a shifting microbiome means functionally (role of microbiota and host).

• Translate this information in better diagnostics, prognostics, and treatments - moving toward

personalized medicine

Acknowledgments

Larry Forney

National Institutes of Health NIAID U01 - AI070921- R03-AI061131NIAID UH2 - AI083264 - U19 AI084044

Pawel Gajer

Sara KoenigStacey McCulle

Guoyun Bai

Li Fu

Rebecca Brotman

Zaid Abdo

Maria Schneider

Xia Zhou

Joyce Sakamoto

Melissa Nandy

Hongqiu YangXue Zhong

Bishoy Michael

Bing Ma

Doug Fadrosh

Anup MahurkarUrsel Shütte

Ligia Peralta Reshma Gorle

Matt Settles

Roxana Hickey

Patrik Bavoil

Mishka TerplanEsther Colinetti

Owen White

Jane Schwebke

Molly Flynn

Charles Rivers