28 AUGUST 2015 • VOL 349 ISSUE 6251 929SCIENCE sciencemag.org

it will reduce the probability of release to

a level that is acceptably low. This prob-

ability must be defined on a case-by-case

basis. The analyses necessary to confidently

predict the efficacy of confinement strate-

gies for gene drive systems are in a nascent

form. Therefore, any proposal to use one

rather than multiple forms of confinement

requires even greater scrutiny and extensive

deliberation between regulatory authorities

and scientists.

3) Organisms carrying gene drive con-

structs that could spread if the reproduc-

tively capable life stages were to escape in

transit should not be distributed to other in-

stitutions until formal biosafety guidelines

are established. Whenever possible, labora-

tories should instead send DNA constructs

or information sufficient to reconstruct the

gene drive. Protocols for distributing ma-

terials should be established in discussion

with the wider research community and

other relevant stakeholders.

Broadly inclusive and ongoing discus-

sions among diverse groups concerning safe-

guards, transparency, proper use, and public

involvement should inform expert bodies as

they develop formal research guidelines for

gene drive research in the laboratory and

potential transitions to open field trials. We

applaud the U.S. National Academy of Sci-

ences for committing to provide recommen-

dations for responsible gene drive research

( 15). By recommending strong safeguards

and encouraging discussion of this technol-

ogy, we hope to build a foundation of pub-

lic trust for potential future applications in

public health, sustainable agriculture, and

ecological conservation. ■

REFERENCES AND NOTES

1. C.-H. Chen et al., Science 316, 597 (2007). 2. O. S. Akbari et al., Curr. Biol. 23, 671 (2013). 3. Y.-S. Chan, D. A. Naujoks, D. S. Huen, S. Russell, Genetics

188, 33 (2011). 4. K. M. Esvelt, A. L. Smidler, F. Catteruccia, G. M. Church, eLife

2014, e03401 (2014). 5. K. A. Oye et al., Science 345, 626 (2014). 6. V. M. Gantz, E. Bier, Science 348, 442 (2015). 7. A. Burt, Proc. R. Soc. London Ser. B 270, 921 (2003). 8. R. D. Henkel et al, Appl. Biosaf. 18, 171 (2012). 9. J. E. DiCarlo et al, bioRxiv 013896 (2015). 10. X. Ren et al., Proc. Natl. Acad. Sci. U.S.A. 110, 19012 (2013). 11. S. J. Gratz et al., Genetics 196, 961 (2014). 12. F. Port, H.-M. Chen, T. Lee, S. L. Bullock, Proc. Natl. Acad.

Sci. U.S.A. 111, E2967 (2014). 13. F. Port et al, G3 (Bethesda) 5, 1493 (2015). 14. S. Kondo, R. Ueda, Genetics 195, 715 (2013). 15. National Research Council, Gene Drive Research in Non-

Human Organisms: Recommendations for Responsible Conduct (DELS-BLS-15-06, National Academy of Sciences, Washington, DC, 2015); http://bit.ly/CurrProjects-regul.

ACKNOWLEDGMENTS

The authors are grateful for conversations with T. Wu, J. Lunshof, and A. Birnbaum. V.M.G., E.B., G.M.C., and K.M.E. are inventors on relevant provisional and nonprovisional patents filed by the University of California and Harvard University.

Published online 30 July 2015

10.1126/science.aac7932

The immune system in the intestine

is highly adapted to resist invading

pathogens while residing peacefully

with the abundant and diverse com-

mensal bacteria that colonize the

gastrointestinal tract. In turn, bac-

terial signals shape immunity in the intes-

tine, promoting intestinal homeostasis in

part by inducing and expanding specialized

regulatory T (Treg

) cells that prevent aberrant

inflammatory responses to self and environ-

mental stimuli ( 1). On pages 989 and 993 of

this issue, Ohnmacht et al. ( 2) and Sefik et

al. ( 3), respectively, report the development

and function of a subpopulation of Treg

cells

found primarily in the large intestine, and

characterized by expression of the nuclear

hormone receptor retinoic acid receptor-

related orphan receptor γt (RORγt). This is

surprising because RORγt classically pro-

motes the differentiation of T helper 17

(TH17) cells, a population associated with

tissue inflammation in many inflammatory

diseases ( 4). Both studies show that microbi-

ota-derived signals induce the expression of

RORγt in Treg

cells that control intestinal in-

flammation (see the figure). These findings

highlight the diversity of colonic Treg

cells,

their complex transcriptional programs, and

their important role in the intestine.

Treg

cells express the forkhead transcrip-

tion factor Foxp3, which promotes their dif-

ferentiation, maintenance, and function ( 5).

Alongside anti-inflammatory functions, they

control nonimmunological processes in-

cluding tissue repair and metabolism in the

parenchyma ( 6). Treg

cells also adapt to envi-

ronmental cues by expressing canonical ef-

fector T cell–associated transcription factors

to control pathogenic immune responses ( 7).

Both Ohnmacht et al. and Sefik et al. found

that in mice, a large fraction of intestinal Treg

cells express RORγt. These cells were distinct

from colonic Treg

cells that express the tran-

scription factor GATA3 and are poised to

respond to the cytokine interleukin (IL)–33

after tissue damage ( 8, 9). However, RORγt-

expressing Treg

cells had an activated pheno-

type similar to that of GATA3-expressing Treg

cells, and bore markers related to Treg

cells

residing in lymphoid and non-lymphoid tis-

sues ( 6). Strikingly, the microbiota was an

absolute requirement for the induction and

maintenance of RORγt-expressing Treg

cells

in these animals. This Treg

cell population

was markedly reduced in germ-free mice,

and colonization with a diverse microbiota

or consortia of symbionts was sufficient for

the induction of RORγt-expressing Treg

cells.

Sefik et al. went further and recolonized

germ-free mice with 22 different bacterial

species, and showed that a number of them

(not belonging to any specific phylum or ge-

nus) elicited RORγt-expressing Treg

cells at

comparable frequencies to a diverse micro-

biota. Short-chain fatty acids, which are com-

mon bacterial metabolites, can selectively

expand intestinal Treg

cells ( 10). Ohnmacht et

al. could increase RORγt-expressing Treg

cells

by feeding mice a diet rich in the short-chain

fatty acid butyrate.

Which signals promote RORγt expression

in Treg

cells? The TH17-favoring cytokines

IL-6 and IL-23 were required for accumu-

lation of RORγt-expressing Treg

cells, which

raises the question of what tips the bal-

ance toward these T cells rather than TH17

cells. The vitamin A metabolite retinoic

acid promotes Treg

cell generation in vivo

and RORγt-expressing Treg

cells in vitro ( 11,

12). Consistent with this, Ohnmacht et al.

show that vitamin A metabolism influences

the differentiation equilibrium by favoring

the development of RORγt-expressing Treg

cells in vivo. Although both Treg

cells and

TH17 cells express RORγt, analysis of all the

transcripts expressed by each population re-

vealed marked differences, suggesting that

the transcriptional footprint of RORγt is

context-dependent in different T cells.

What is the function of RORγt-expressing

Microbiota RORgulates intestinal suppressor T cells

By Ahmed N. Hegazy 1, 2 and Fiona Powrie 1, 2

Gut microbes influence the balance of regulatory T cell subtypes to control inflammation

MICROBIOME

“These studies…are an important stepping stone to deciphering the complex dynamics of different tissue-resident T

reg cell subsets…”

Published by AAAS

on

Aug

ust 2

7, 2

015

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Aug

ust 2

7, 2

015

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

INSIGHTS | PERSPECTIVES

930 28 AUGUST 2015 • VOL 349 ISSUE 6251 sciencemag.org SCIENCE

Treg

cells in health and disease? Ohnmacht et

al. and Sefik et al. addressed this question

by conditional deletion of the Rorc gene in T

cells expressing Foxp3, producing mice spe-

cifically deficient in RORγt-expressing Treg

cells. The results were, however, ambiguous,

perhaps reflecting differences in experimen-

tal models, animal housing, or the indige-

nous microbiota. Ohnmacht et al. observed a

pronounced increase in type 2 cytokines un-

der steady-state conditions, with consequent

resistance to helminth infection. Inflam-

matory outcomes differed according to the

chemically induced colitis model used. In

oxazolone-induced colitis (a type 2 cytokine–

driven model), mice developed severe and

lethal colitis accompanied by an increase in

type 2 cytokines, whereas no alteration in

pathology or TH1 and T

H17 cell responses was

observed in dextran sulfate sodium–induced

colitis (a type 1 and type 17 cytokine–depen-

dent model). Sefik et al. chemically blocked

RORγt function and found that colonic Treg

cell frequency decreased, and interferon-γ

and IL-17 production by effector T cells

increased, under steady-state conditions.

These mice developed severe colitis in an-

other chemically induced (trinitrobenzene-

sulfonic acid) colitis model.

The findings of Ohnmacht et al. and Se-

fik et al. show that microbe-induced expres-

sion of RORγt by Treg

cells contributes to

the control of intestinal inflammation. But

what is the role of distinct colonic Treg

cell

subsets (RORγt and GATA3) in intestinal ho-

meostasis? Is there functional redundancy

or division of labor? It may be that RORγt-

expressing Treg

cells, which produce in-

creased amounts of cytotoxic T lymphocyte

antigen 4 (CTLA4) and IL-10 (both of which

dampen immune responses), decrease in-

flammation, whereas GATA3-expressing Treg

cells, which produce the tissue-remodeling

factor amphiregulin and respond to the

alarm signal (“alarmin”) IL-33, mediate tis-

sue repair. It is certainly possible that locally

produced inflammatory and tissue-derived

factors might activate or antagonize differ-

ent Treg

cell subpopulations to coordinate the

anti-inflammatory and healing processes.

Another issue raised by these studies

is whether Treg

cell subsets are specialized

in controlling particular effector T cell re-

sponses. Ohnmacht et al. propose that

RORγt-expressing Treg

cells are critical in

controlling aberrant type 2 responses and

that deficiencies in these microbiota-driven

Treg

cells may contribute to increases in al-

lergic disease. However, Sefik et al. observed

control of TH1 and T

H17 cells by RORγt-

expressing Treg

cells. It seems highly likely

that the relative activity and function of

distinct colonic Treg

cell populations will be

highly context-dependent and influenced

by the microbiota. It is important to under-

stand the ontogeny of RORγt-expressing Treg

cells and GATA3-expressing Treg

cells and

whether they represent distinct lineages

or can interconvert. Inducible labeling and

tracking of Treg

cell subsets would provide

valuable insights into their interplay and sta-

bility under steady-state and inflammatory

conditions. It remains to be established why

only certain bacterial species induce RORγt

expression in Treg

cells and whether we can

identify similar Treg

cell subsets in humans

and manipulate them in vivo. The studies

by Ohnmacht et al. and Sefik et al. are an

important stepping stone to deciphering

the complex dynamics of different tissue-

resident Treg

cell subsets toward the under-

standing of how their dysregulation precipi-

tates human disease. ■

REFERENCES AND NOTES

1. Y. Belkaid, T. W. Hand, Cell 157, 121 (2014). 2. C. Ohnmacht et al., Science 349, 989 (2015). 3. E. Sefik et al., Science 349, 993 (2015). 4. T. Korn, E. Bettelli, M. Oukka, V. K. Kuchroo, Annu. Rev.

Immunol. 27, 485 (2009). 5. S. Z. Josefowicz, L.-F. Lu, A. Y. Rudensky, Annu. Rev.

Immunol. 30, 531 (2012). 6. D. Burzyn et al., Nat. Immunol. 14, 1007 (2013). 7. D. J. Campbell, M. A. Koch, Nat. Rev. Immunol. 11, 119

(2011). 8. E. A. Wohlfert et al., J. Clin. Invest. 121, 4503 (2011). 9. C. Schiering et al., Nature 513, 564 (2014). 10. P. M. Smith et al., Science 341, 569 (2013). 11. D. Mucida et al., Science 317, 256 (2007). 12. M. Lochner et al., J. Exp. Med. 205, 1381 (2008).

ACKNOWLEDGMENTS

A.N.H. was supported by a European Molecular Biology Organization fellowship (ALTF 116-2012) and currently is a Marie Curie fellow (FP7-PEOPLE-2012-IEF, Proposal 330621). F.P. is supported by a Wellcome Trust Investigator Award.

ILL

US

TR

AT

ION

: K

. S

UT

LIF

F/SCIENCE

1Kennedy Institute of Rheumatology, Nuf eld Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford, Headington, Oxford OX3 7FY, UK. 2Translational Gastroenterology Unit, Nuf eld Department of Clinical Medicine, Experimental Medicine Division, John Radclif e Hospital, University of Oxford, Oxford OX3 9DU, UK.E-mail: f ona.powrie@kennedy.ox.ac.uk 10.1126/science.aad0865

Intestinal epithelium

Reduced infammation

Blocks Blocks

Blocks

Intestinal epithelium

Lumen

Commensal microbiota

Macrophage

ROR t

Retinoicacid

GATA3

IL-6

IL-23IL-33

Treg

Treg

TH2 TH17 TH1

Metabolites Damage

IL-23

Fine-tuning intestinal homeostasis. Microbiota and tissue-derived factors regulate the balance between RORγt-

expressing and GATA3-expressing Treg

cells in the mouse intestine. The microbiota promotes RORγt expression in

intestinal Treg

cells through multiple factors including bacterial metabolites, retinoic acid, and cytokines. Tissue-

resident Treg

cells control effector T cell responses to promote intestinal homeostasis.

Published by AAAS

DOI: 10.1126/science.aad0865, 929 (2015);349 Science

Ahmed N. Hegazy and Fiona PowrieMicrobiota RORgulates intestinal suppressor T cells

This copy is for your personal, non-commercial use only.

clicking here.colleagues, clients, or customers by , you can order high-quality copies for yourIf you wish to distribute this article to others

  here.following the guidelines

can be obtained byPermission to republish or repurpose articles or portions of articles

  ): August 27, 2015 www.sciencemag.org (this information is current as of

The following resources related to this article are available online at

http://www.sciencemag.org/content/349/6251/929.full.htmlversion of this article at:

including high-resolution figures, can be found in the onlineUpdated information and services,

http://www.sciencemag.org/content/349/6251/929.full.html#relatedfound at:

can berelated to this article A list of selected additional articles on the Science Web sites

http://www.sciencemag.org/content/349/6251/929.full.html#ref-list-1, 5 of which can be accessed free:cites 12 articlesThis article

http://www.sciencemag.org/cgi/collection/microbioMicrobiology

subject collections:This article appears in the following

registered trademark of AAAS. is aScience2015 by the American Association for the Advancement of Science; all rights reserved. The title

CopyrightAmerican Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by theScience

on

Aug

ust 2

7, 2

015

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om