Is It Time for a Metagenomic Basis of Therapeutics?

Henry J. Haiser and Peter J. Turnbaugh

Science 336, 1253 (2012);
DOI: 10.1126/science.1224396

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

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

Downloaded from www.sciencemag.org on June 9, 2012

Permission to republish or repurpose articles or portions of articles can be obtained by
following the guidelines here.

The following resources related to this article are available online at

www.sciencemag.org (this information is current as of June 9, 2012 ):

Updated information and services, including high-resolution figures, can be found in the online
version of this article at:
http://www.sciencemag.org/content/336/6086/1253.full.html
A list of selected additional articles on the Science Web sites related to this article can be
found at:
http://www.sciencemag.org/content/336/6086/1253.full.html#related
This article cites 13 articles, 10 of which can be accessed free:
http://www.sciencemag.org/content/336/6086/1253.full.html#ref-list-1
This article has been cited by 1 articles hosted by HighWire Press; see:
http://www.sciencemag.org/content/336/6086/1253.full.html#related-urls
This article appears in the following subject collections:
Microbiology
http://www.sciencemag.org/cgi/collection/microbio

Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the
American Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. Copyright
2012 by the American Association for the Advancement of Science; all rights reserved. The title Science is a
registered trademark of AAAS.
 SPECIALSECTION
birth to an infant or to a mother just before or References and Notes 19. W. H. Pope et al., PLoS ONE 6, e16329 (2011).
immediately after parturition. The goal may be 1. E. K. Costello, K. Stagaman, L. Dethlefsen, 20. J. Qin et al., Nature 464, 59 (2010).
B. J. M. Bohannan, D. A. Relman, Science 336, 1255 (2012). 21. M. Arumugam et al., Nature 473, 174 (2011).
to achieve benefits in the context of a diet or diets 2. J. K. Nicholson et al., Science 336, 1262 (2012). 22. P. Yilmaz et al., Nat. Biotechnol. 29, 415 (2011).
consumed by a given human population or by a 3. L. V. Hooper, D. R. Littman, A. J. Macpherson, 23. A. L. Goodman et al., Cell Host Microbe 6, 279 (2009).
broad range of populations. How will expanding Science 336, 1268 (2012). 24. A. L. Goodman et al., Proc. Natl. Acad. Sci. U.S.A. 108,
knowledge of diet-microbiota interactions affect 4. K. P. Lemon, G. C. Armitage, M. A. Fischbach, 6252 (2011).
Sci. Transl. Med. 4, 137rv5 (2012). 25. J. G. Caporaso et al., Genome Biol. 12, R50 (2011).
claims for the health benefits of foods in con- 5. E. Holmes et al., Sci. Transl. Med. 4, 137rv6 (2012). 26. J. G. Caporaso et al., Nat. Methods 7, 335 (2010).
sumer populations with different gut microbial 6. R. Blumberg, F. Powrie, Sci. Transl. Med. 4, 137rv7 27. A. Benezra, J. Destefano, J. I. Gordon, Proc. Natl. Acad.
community configurations and on strategies for (2012). Sci. U.S.A. 109, 6378 (2012).
7. J. H. Hehemann et al., Nature 464, 908 (2010). 28. A. K. Hawkins, K. C. ODoherty, BMC Med. Genomics 4,
differentiation of food products? The result could 8. T. Yatsunenko et al., Nature (2012). 72 (2011).
herald a new epoch of precision nutrition but 9. P. de Kruif, Microbe Hunters (Harcourt Brace, New York, 29. A. L. McGuire et al., Genome Res. 18, 1861 (2008).
will also put increasing focus on foods as drugs 1926). 30. R. R. Sharp, J.-P. Achkar, M. A. Brinich, R. M. Farrell,
for disease prevention as well as treatment. The 10. H. L. McClellan, S. J. Miller, P. E. Hartmann, Am. J. Gastroenterol. 104, 809 (2009).
Nutr. Res. Rev. 21, 97 (2008). 31. P. Berg, D. Baltimore, S. Brenner, R. O. Roblin III,
effects of this focus on the strategic plans and 11. A. I. Kendall, J. Biol. Chem. 6, 499 (1909). M. F. Singer, Proc. Natl. Acad. Sci. U.S.A. 72, 1981 (1975).
investments of food, dairy, and pharmaceutical 12. J. C. Torrey, J. Med. Res. 39, 415 (1919).
32. P. Berg, M. F. Singer, Proc. Natl. Acad. Sci. U.S.A. 92,
companies will be important to follow. Thus, 13. A. I. Kendall, Science 42, 209 (1915).
9011 (1995).
with so much opportunity and hope associated 14. R. Freter, in Human Intestinal Microflora in Health and
33. E. Metchnikoff, The Prolongation of Life: Optimistic

Downloaded from www.sciencemag.org on June 9, 2012

Disease (Academic Press, New York, 1983), pp. 3354.
with the gut microbiome, it would be wise to ad- 15. R. W. Schaedler, R. Dubos, R. Costello, J. Exp. Med. 122,
Studies, P. C. Mitchell, Ed. (Putnam, New York, 1908),
dress these issues now rather than later, through a p. 169.
59 (1965).
process that takes lessons from the group of 16. A. Lee, J. Gordon, C. J. Lee, R. Dubos, J. Exp. Med. 133,
Acknowledgments: Work cited from the authors lab was
individuals who, illustrating the precautionary 339 (1971).
supported by grants from the NIH, the Bill and Melinda
17. D. C. Savage, R. Dubos, R. W. Schaedler, J. Exp. Med.
principle, gathered together in 1975 to discuss 127, 67 (1968).
Gates Foundation, and the Crohns and Colitis Foundation of
recombinant DNA research and its implications America. The author declares no competing interests.
18. R. Dubos, R. W. Schaedler, R. Costello, P. Hoet,
(31, 32). J. Exp. Med. 122, 67 (1965). 10.1126/science.1224686

drugs for which this contribution likely plays an

PERSPECTIVE
important role) (3), members of the gut micro-
biota can also influence xenobiotic metabolism by
Is It Time for a Metagenomic Basis altering host gene expression (4) and producing
compounds that interfere with metabolism out-
of Therapeutics? side of the gut (Fig. 1) (5). Importantly, these
interactions are reciprocal, as exposure to xe-
nobiotics, especially antibiotics, can affect the
Henry J. Haiser and Peter J. Turnbaugh* structure of our indigenous microbial commu-
nities (6, 7).
The trillions of microbes associated with the human body are a key part of a comprehensive The discovery of antibiotics changed the way
view of pharmacology. A mechanistic understanding of how the gut microbiota directly and we have viewed infectious disease over the past
indirectly affects drug metabolism is beginning to emerge. century, yet we are just beginning to assess the
unintended collateral damage that antibiotics
he human microbiota, the vast number The past century has provided a wealth of often impart on the symbiotic microorganisms

T of microbes that live within and upon us,

is an important but largely underexplored
component of therapeutics, prompting efforts to
information in the field of pharmacology, elu-
cidating the rates of absorption, distribution,
metabolism, and excretion for hundreds of xeno-
living in our gastrointestinal tract (8). Repeated
exposure to broad-spectrum antibiotics can dra-
matically alter the composition of human gut mi-
complement pharmacological studies with de- biotics (compounds foreign to a living organism, crobial communities, which can persist for long
tailed analyses of our microbial communities. including therapeutic drugs, antibiotics, and diet- periods of time (6, 7). Antibiotic exposure has
Recent studies suggest that opportunities to de- derived bioactive compounds) (1). Combining also been linked to marked changes in the meta-
velop new diagnostics and treatments may arise knowledge of environmental risk factors and hu- bolic output of the intestinal tract; for example,
through a mechanistic understanding of our res- man genetics has further advanced our under- almost 88% of the ~2000 detectable metabolites
ident microbial communities abilities to extend standing of the basis for interindividual variations derived from fecal samples were present in al-
human metabolism and an exploration of their in drug bioavailability, toxicity, and efficacy (2). tered amounts after antibiotic treatment in mice
impacts on human health and predisposition to However, the existing clinical and genetic bio- (9). The functional consequences of such shifts
disease. Indeed, our emerging view of the hu- markers often fail to explain a large part of the are not always obvious; however, it is clear that
man microbiomethe aggregate genomes of our variation in patient response; for example, even antibiotic treatment can render individuals more
microbiota and the diverse metabolic activities in the case of the well-studied anticoagulant war- susceptible to infection (10). This is exempli-
that they encodehas the potential to revolution- farin, up to 50% of the variation is still un- fied by Clostridium difficile, a microbe that can
ize the way we view modern therapeutics. explained (2). cause severe diarrheal illness with potentially
A neglected but critical component of xeno- life-threatening complications following a reduc-
biotic metabolism is the influence of the trillions tion in diversity of the gut microbiota after anti-
Faculty of Arts and Sciences Center for Systems Biology, of microorganisms inhabiting our gastrointestinal biotic treatment (11).
Harvard University, Cambridge, MA, USA. tract. In addition to merely expanding the set of Although we are still in the early stages of in-
*To whom correspondence should be addressed. E-mail: human-associated enzymes that can directly mod- vestigating the impact of xenobiotics on the compo-
pturnbaugh@fas.harvard.edu ify xenobiotics (there are more than 30 known sition and function of gut microbial communities,

www.sciencemag.org SCIENCE VOL 336 8 JUNE 2012 1253

 The Gut Microbiota
we know even less about how the microbes living creasing toxicity. This was recently demonstrated Can the efficacy of a therapeutic agent be im-
in the gut influence the metabolism of therapeutic by monitoring pre- and postdose metabolite pro- proved through a mechanistic understanding of
drugs. A review of the pharmacological literature files in human participants who were given the how the indigenous microbiota interact with the
from the past few decades reveals several cases in drug acetaminophen. A microbial compound, compound? Administered as a prodrug, the che-
which a particular biotransformation (the chemi- p-cresol, was abundant in the predose metab- motherapeutic agent irinotecan becomes activated
cal alteration of a compound by an organism) is olite profiles of individuals who did not fully by enzymes in serum and tissue, where it inhibits
suspected to result directly from a reaction carried metabolize acetaminophen (5). Because the ex- topoisomerase I (an enzyme that regulates the
out by gut microbes (3). Typically, these studies cretion of both p-cresol and acetaminophen de- structure and replication of DNA) in rapidly
involve either monitoring drug metabolism during pends on the same O-sulfonation conjugation growing tumor cells. Before excretion, the drug
the ex vivo incubation of fecal samples with a giv- reaction in the liver, the results suggest that p- is converted in the liver to a nontoxic metabolite;
en drug or observing that antibiotic pretreatment cresolproducing members of the gut microbiota however, bacterial enzymes (b-glucuronidases)
abolishes the biotransformation, presumably due can indirectly inhibit the metabolism of this can reactivate this metabolite after it is released
to the widespread depletion of the gut microbiota. widely used xenobiotic. back into the intestine, resulting in severe diarrhea
The classical examples of direct microbial The gut microbiome also affects the metabo- that often prevents further increases in dosage.
biotransformations often consist of either the ac- lism of diet-derived bioactive compounds. A Although the administration of broad-spectrum
tivation or inactivation of a parent compound. metabolomics screen of human plasma revealed antimicrobials prevents this undesirable bio-
An example of the former comes from the anti- that the phosphatidylcholine metabolites choline, transformation of irinotecan, this approach also

Downloaded from www.sciencemag.org on June 9, 2012

inflammatory drug sulfasalazine, which is used trimethylamine N-oxide, and betaine serve as leaves individuals susceptible to other serious
to treat rheumatoid arthritis and inflammatory predictive biomarkers for the development of complications. A recent study addressed this
bowel disease. Sulfasalazine is subject to metab- cardiovascular disease (13). In this same study, issue by harnessing a chemical screen to iden-
olism by microbial azoreductases that split the germ-free mice revealed that the gut microbiota tify compounds that specifically inhibit bacterial
drug into two halves: 5-aminosalicylic acid, which is directly involved in producing these metabo- b-glucuronidase (14). Remarkably, the admin-
is pharmacologically active, and sulfapyridine, lites and that suppression of the gut microbiota istration of irinotecan with a b-glucuronidase in-
which is thought to cause the undesirable side (via antibiotic treatment) can prevent choline- hibitor alleviated the negative side effects in mice.
effects associated with this drug (3). The role of induced atherosclerosis. These findings highlight Advances in DNA sequencing, cell sorting,
the individual bacterial species involved and the the critical role of the gut microbiota in extending mass spectrometry, microfluidics, and computa-
molecular mechanisms underlying the biotrans- our own metabolism through modifying the mi- tional biology are contributing an array of new
formation are largely uncharacterized. Converse- cronutrient component of our diet and also serve tools to the study of human-associated microbial
ly, the cardioactive drug digoxin is susceptible to as a notable example of how the composition of communities. These approaches could lead to the
almost complete inactivation through the reduc- commensal microbial communities can directly first metagenomic basis of therapeutics that
tion of a single double bond in the lactone ring of affect disease. merges the vast knowledge of pharmacology
this compound. A common member of the gut
microbiota, Eggerthella lenta, can carry out this
reduction reaction in isolation (12). Activation of
a drug by the gut microbiota can be undesirable if Drug administration
the reaction occurs in an unintended body site. Active Prodrugs and
compounds inactive compounds
This is the case for levodopa, an orally admin-
istered drug used to treat Parkinsons disease by
ameliorating a dopamine imbalance. Levodopa
can cross the blood-brain barrier, after which it
depends on a host-catalyzed decarboxylation
reaction to convert it to dopamine in the central
nervous system (CNS). However, the gut micro- Direct interactions Gut microbiota Indirect interactions
biota can also perform this biotransformation (3);
having this modification take place in the gut
prevents dopamine from reaching the CNS, po-
Activation
tentially contributing to the variability in patient
responses to levodopa treatment. An integral part
of translating these observations into practical
recommendations that improve drug efficacy Inactivation
will be identifying which particular microbes, or Host gene
microbial consortia, carry out the relevant bio- expression Microbial
changes metabolites
transformations and unraveling the underlying Byproduct
molecular mechanisms. toxic to host
Our capacity to metabolize therapeutic drugs
is also affected indirectly by the gut microbiota.
Microbes can secrete metabolites that act as Altered host drug metabolism
substrates for host enzymes essential to process a Antimicrobial
given drug, thus titrating away their activity. By
diminishing the hosts capacity to metabolize Fig. 1. A microbial view of xenobiotic metabolism. The gut microbiota interacts with xenobiotics directly
drugs, the gut microbiota may indirectly affect by catalyzing various biotransformations. In turn, many xenobiotics inhibit microbial growth or cause cell
the desired pharmacological action, potentially damage. Indirect interactions have also been described, including microbial effects on the expression and
prolonging the time spent in circulation or in- activity of components of host xenobiotic metabolism.

1254 8 JUNE 2012 VOL 336 SCIENCE www.sciencemag.org

 SPECIALSECTION
with a quantitative understanding of key envi- understanding of the short- and long-term impact 6. C. Jernberg, S. Lfmark, C. Edlund, J. K. Jansson,
ronmental risk factors and their interactions with of xenobiotics on host and microbial physiology. Microbiology 156, 3216 (2010).
7. L. Dethlefsen, D. A. Relman, Proc. Natl. Acad. Sci. U.S.A.
our human and microbial genomes. To complete Furthermore, the detailed study of pharmaceuti- 108, 4554 (2011).
this picture, it is important to move beyond large- cally active compounds may be a tractable first 8. M. Blaser, Nature 476, 393 (2011).
ly DNA sequencingbased association studies step toward understanding the fundamental rules 9. L. C. Antunes et al., Antimicrob. Agents Chemother. 55,
toward a mechanistic understanding of how that govern the immense phylogenetic and meta- 1494 (2011).
10. I. Sekirov et al., Infect. Immun. 76, 4726
members of the gut microbiota, either in isola- bolic diversity of our microbial partners and how
(2008).
tion or through mutualistic interactions with each they influence our predisposition to and recovery 11. J. Y. Chang et al., J. Infect. Dis. 197, 435 (2008).
other, can transform each compound. This will from disease. 12. J. F. Dobkin, J. R. Saha, V. P. Butler Jr., H. C. Neu,
require multiple complementary top-down and J. Lindenbaum, Science 220, 325 (1983).
bottom-up approaches, including detailed in vitro References and Notes 13. Z. Wang et al., Nature 472, 57 (2011).
1. L. S. Goodman, L. L. Brunton, B. Chabner, B. C. Knollmann, 14. B. D. Wallace et al., Science 330, 831 (2010).
analyses of culturable microbes; studies in germ- Goodman & Gilman's Pharmacological Basis of
free and intentionally colonized animal models; Therapeutics (McGraw-Hill, New York, ed. 12, 2011). Acknowledgments: H.J.H. is supported by a postdoctoral
metagenomic surveys of patients before, during, 2. Q. Ma, A. Y. Lu, Pharmacol. Rev. 63, 437 (2011). fellowship from the Canadian Institutes of Health
and after treatment; and large-scale clinical trials. 3. T. Sousa et al., Int. J. Pharm. 363, 1 (2008). Research (MFE-112991). P.J.T. is supported by a grant
4. S. P. Claus et al., mBio 2, e00271-10 (2011). from the NIH (P50 GM068763).
These types of studies will likely lead to new 5. T. A. Clayton, D. Baker, J. C. Lindon, J. R. Everett,
microbial therapeutic targets, noninvasive bio- J. K. Nicholson, Proc. Natl. Acad. Sci. U.S.A. 106, 14728

Downloaded from www.sciencemag.org on June 9, 2012

markers for drug toxicity or efficacy, and a broader (2009). 10.1126/science.1224396

spatial patterns of diversity. Here, we explore

REVIEW
how community assembly theory can be used to
understand the human-associated microbiota and
The Application of Ecological Theory its role in health and disease.

Toward an Understanding of the Ecological Processes Within Humans

The essential building blocks of community as-
sembly theory encompass the processes that
Human Microbiome create and shape diversity in local assemblages:
dispersal, diversification, environmental selection,
Elizabeth K. Costello,1 Keaton Stagaman,2 Les Dethlefsen,1,3 and ecological drift (6). In addition, coevolution
Brendan J. M. Bohannan,2 David A. Relman1,3,4* provides another lens through which to view the
human-microbial ecosystem (7), although in this
The human-microbial ecosystem plays a variety of important roles in human health and disease. review we focus on shorter-term dynamics at the
Each person can be viewed as an island-like patch of habitat occupied by microbial assemblages level of individual hosts.
formed by the fundamental processes of community ecology: dispersal, local diversification, Dispersal, or the movement of organisms
environmental selection, and ecological drift. Community assembly theory, and metacommunity across space, is a fundamental process by which
theory in particular, provides a framework for understanding the ecological dynamics of the human diversity accumulates in local microbial com-
microbiome, such as compositional variability within and between hosts. We explore three core munities. The concept put forth in the late 19th
scenarios of human microbiome assembly: development in infants, representing assembly in and early 20th centuries that everything is ev-
previously unoccupied habitats; recovery from antibiotics, representing assembly after disturbance; erywhere, but the environment selects had a
and invasion by pathogens, representing assembly in the context of invasive species. Judicious powerful impact on thinking about community
application of ecological theory may lead to improved strategies for restoring and maintaining the assembly (8), but a more recent appreciation of
microbiota and the crucial health-associated ecosystem services that it provides. microbial dispersal limitation suggests that this
conceptualization was overly simplistic. Think-
ach person is an assemblage of not only ciated with alterations in host-associated microbial ing in terms of dispersal leads to a view of the

E human cells but also many symbiotic spe-

cies. The abundant and diverse microbial
members of the assemblage play critical roles in
communities (the microbiota), including obe-
sity, malnutrition, and a variety of inflammatory
diseases of the skin, mouth, and intestinal tract.
human body as an island, a patch of habitat
that is continually sampling the pool of available
colonists. The list of available colonists may
the maintenance of human health by liberating Thus, the human body can be viewed as an eco- be influenced by microbial traitsthose affect-
nutrients and/or energy from otherwise inacces- system, and human health can be construed as a ing dispersal efficiency, transmission routes, and
sible dietary substrates, promoting differentiation product of ecosystem services delivered in part ex-host survivabilityand by patterns of host
of host tissues, stimulating the immune system, by the microbiota. contact and carriage, among other factors. Con-
and protecting the host from invasion by path- There is growing interest in the use of theo- trol of infectious disease transmission depends
ogens. A number of clinical disorders are asso- retical methods to study microbial community on accurate models of host-to-host microbial dis-
ecology in general and host-associated micro- persal (9), and these could guide investigations
1
Department of Microbiology and Immunology, Stanford Uni- biota in particular (1, 2). Recent discoveries of into the dissemination of the human microbiome.
versity School of Medicine, Stanford, CA 94305, USA. 2Institute unexpected variation in the composition of the Selection favors efficient dispersal in pathogens,
of Ecology and Evolution, University of Oregon, Eugene, OR microbiome of healthy individuals (35) high- but perhaps less so among beneficial bacteria,
97403, USA. 3Department of Medicine, Stanford University
School of Medicine, Stanford, CA 94305, USA. 4Veterans Affairs light the importance of identifying the processes because the host is harmed by the first and not by
Palo Alto Health Care System, Palo Alto, CA 94304, USA. that could possibly give rise to such variation. the second; for beneficial microbes, transmission
*To whom correspondence should be addressed. E-mail: Ecological theory seeks to explain and predict routes such as direct or close contact may be more
relman@stanford.edu observable phenomena, such as temporal and important. The density and spatial arrangement

www.sciencemag.org SCIENCE VOL 336 8 JUNE 2012 1255