The Nih Human Microbiome Project
The Nih Human Microbiome Project
The Human Microbiome Project (HMP), funded as an initiative of the NIH Roadmap for Biomedical Research (http://
nihroadmap.nih.gov), is a multi-component community resource. The goals of the HMP are: (1) to take advantage of new,
high-throughput technologies to characterize the human microbiome more fully by studying samples from multiple body
sites from each of at least 250 normal volunteers; (2) to determine whether there are associations between changes in the
microbiome and health/disease by studying several different medical conditions; and (3) to provide both a standardized
data resource and new technological approaches to enable such studies to be undertaken broadly in the scientific com-
munity. The ethical, legal, and social implications of such research are being systematically studied as well. The ultimate
objective of the HMP is to demonstrate that there are opportunities to improve human health through monitoring or
manipulation of the human microbiome. The history and implementation of this new program are described here.
It has been known for some time that the human body is inhabited et al. 1984; Woese and Olsen 1986; Giovannoni et al. 1990;
by at least 10 times more bacteria than the number of human cells Schmidt et al. 1991; Dymock et al. 1996). Much of our un-
in the body, and that the majority of those bacteria are found in the derstanding of the human microbiome comes from culture-based
human gastrointestinal tract (Savage 1977). Throughout the his- approaches using the 16S rRNA technology. However, it is esti-
tory of microbiology, most human studies have focused on the mated that as much as 20% to 60% of the human-associated
disease-causing organisms found on or in people; fewer studies microbiome, depending on body site, is uncultivable (Pei et al.
have examined the benefits of the resident bacteria. As noted in 2004; Verhelst et al. 2004; Zhou et al. 2004; Aas et al. 2005; Bik et al.
reviews by Relman and Falkow (2001) and Relman (2002), the 2006), which has likely resulted in an underestimation of its di-
endogenous flora of the human body are poorly understood. Fol- versity.
lowing the publication of the human genome sequence in 2001 More recently, studies have been published that describe the
(International Human Genome Sequencing Consortium 2001; human microbiome in different biological states using the 16S
Venter et al. 2001), Julian Davies argued that although completing rRNA gene sequencing technique. For example, studies of the gut
the human genome sequence was a crowning achievement in microbiome at the 16S rRNA gene level have revealed a significant
biology, it would be incomplete until the synergistic activities be- diversity in the flora of individuals (Eckburg et al. 2005), have
tween humans and microbes living in and on them are understood shown differences in the flora of obese versus lean donors (Ley
(Davies 2001). Relman and Falkow (2001) called for a second et al. 2006), and have followed the evolution of the microbiome in
human genome project that would entail a comprehensive in- infants (Palmer et al. 2007). Studies also have used the 16S rRNA
ventory of microbial genes and genomes at the four major sites of gene as a metagenomic marker of the microbiome in the oral cavity
microbial colonization in the human body: mouth, gut, vagina, (Faveri et al. 2008), vagina (Hyman et al. 2005), and skin (Gao et al.
and skin. Relman (2002) envisioned that the characterization of 2007).
the microbiome would be accomplished through random shotgun Although enormously important in helping scientists define
sequencing procedures, targeted large-insert clone sequencing, evolutionary relationships among bacteria, there are limitations to
and assessments of intra- and inter-individual variation by using the 16S rRNA gene sequencing approach that have prompted more
high-density microarrays. This approach, coupled with a study recent studies to examine the complexity of environmental sam-
of host genome-wide expression analysis, would yield major ples by sequencing genomic libraries made from DNA extracted
insights into the role of the endogenous flora in health and directly from the mixed sample (Handelsman et al. 1998). This ap-
disease. proach is called metagenomics and was initially applied in sev-
eral studies of environmental microbial communities (Handelsman
2004; Tyson et al. 2004; Tringe and Rubin 2005; Nealson and Venter
Scientific background 2007).
The concept of the human microbiome was first suggested by
Joshua Lederberg, who coined the term microbiome, to signify
the ecological community of commensal, symbiotic, and path- Initiation of the HMP
ogenic microorganisms that literally share our body space
(Lederberg and McCray 2001). Initial efforts to determine the The early studies examining the microbiome stimulated interest in
numbers of microbes in a community and their phylogenetic re- undertaking a large-scale investigation of the human intestinal
lationships comprised analyzing the relatively well-conserved 16S microbiome. An international meeting was held in Paris in No-
rRNA genes in mixtures of organisms (Woese and Fox 1977; Stahl vember 2005 to discuss such an effort. This meeting, hosted by the
French National Institute for Agricultural Research (INRA) and
1
chaired by Dusko Ehrlich, led to the recommendation that a Hu-
A complete list of authors and affiliations appears at the end of man Intestinal Metagenome Initiative (HIMI) be undertaken to
the paper, before the Acknowledgments section. See also, http://
nihroadmap.nih.gov/hmp/members.asp. define more completely the human intestinal microbiome in
2
Corresponding author. health and disease. The meeting attendees also recommended that
E-mail [email protected]; fax (301) 480-2770. an International Metagenome Consortium be formed to bring
Article published online before print. Article and publication date are at
http://www.genome.org/cgi/doi/10.1101/gr.096651.109. Freely available together common efforts from around the world to accomplish
online through the Genome Research Open Access option. the goals of the HIMI (http://human-microbiome.org).
complexity at a body site. A major part of these efforts was de- project, again being informed of the special privacy issues associ-
velopment of specialized software such as for chimera checking. ated with the deposition of data in a public database. High-
Many of these programs are now available as free downloads throughput sequencing will be used to produce microbiome
through the individual Jumpstart sequencing center websites ac- (bacterial, viral, and fungal) sequence data (both bacterial and
cessible through http://hmpdacc.org/. fungal rRNA genes and bacterial, viral, and fungal whole genomes)
of samples from normal and diseased individuals. At the same
time, in many of these studies, complementary molecular ap-
Second phase
proaches will be used to measure microbial gene expression and
In the next phase of the HMP, the participants will continue the host genotypes. Data analysis will be performed using existing or
work started in the Jumpstart phase to generate reference databases new computational methodologies, to define the potential re-
and to develop new technology, and then extend the effort with lationships between changes in the human microbiomes and the
a set of demonstration projects to try to determine changes in the health conditions under study. Benchmarking to compare tech-
microbiome correlated to specific disease states. nical and analytical pipelines from the demonstration projects is
currently underway, using the mock community approach
HMP sequencing centers
mentioned above.
The four centers that participated in the Jumpstart phase will
continue to develop and use next-gen methods in the next phase Examples of HMP Demonstration Projects
to sequence at least 400 more bacterial genomes to complete the
goal of adding 900 new bacterial reference genomes to the se- The HMP is supporting a very diverse set of research projects
quence databases. (An additional 100 new bacterial genomes are among the 15 funded grants. Rather than attempt to summarize
being sequenced by the European MetaHit Project; thus, a total of each research project, we give a few examples of projects illus-
at least 1000 new genome sequences from human commensal trating approaches being taken in different health conditions:
microorganisms will be added to the public repository.) The centers Skin: Martin Blaser, PI. Evaluation of the Cutaneous Microbiome
will also sequence the genomes of viruses and eukaryotic microbes in Psoriasis. Psoriasis, a chronic disease involving the immune
found in the human microbiome and perform 16S rRNA gene and system, affecting more than 7.5 million people in the United
whole-genome shotgun sequencing to characterize the micro- States, appears on the skin, usually in the form of thick, red, scaly
biomes of the 250 donors sampled in the Jumpstart phase at ad- patches. Its cause is unknown. The goal of this study is to ex-
ditional time points. All of these data will be used to study the amine how changes in the normal cutaneous microbiome may
diversity of the microbiome at each body site and to attempt to contribute to the disease. The skin microbiome of 75 donors
define a core microbiome at each site, at the highest phylogenetic with and without psoriasis will be examined at several taxo-
resolution that the data allow (most likely at or below the family nomic levels. Additionally, the research will seek to examine
level). Conservation of gene function across the sites also will be whether the immunosuppressive agents used to treat psoriasis
examined. The HMP centers plan to publish manuscripts de- alter the microbiome.
scribing broad analyses of the set of sequenced bacterial genomes Virome: Gregory Storch, PI. The Human Virome in Children and
and microbial metagenomic data generated from the collection of Its Relationship to Febrile Illness. An estimated 20 million visits
normal samples from 18 body sites in more than 250 individuals. per year to hospital emergency departments are because of fever
in children. The causes of a vast majority of these fevers remain
HMP Demonstration Projects undiagnosed. This project seeks to describe the human virome in
The Demonstration Projects aim to tackle the most important children and to investigate its relevance to febrile disease. It will
question of the HMP: whether changes in the microbiome can be also study the relationship of the patients immune systems to
related to human health and disease. Because of the short time the composition of their viromes. Next-gen sequencing tech-
frame of the HMP, the primary goal of these projects is to establish nologies will be used to examine the virome of blood, re-
a correlation between microbiome changes and health/disease, spiratory, and gastrointestinal samples from healthy, from
rather than demonstrate causation. If a project can successfully febrile, and from immunosuppressed children.
demonstrate correlation early in the timeline, work to begin to GI Tract: Claire Fraser-Liggett and Alan Shuldiner, PIs. The Thrifty
establish causation may be undertaken. But the HMP recognizes Microbiome: The Role of the Gut Microbiome in Obesity in the
that such studies may take years of work to complete and go be- Amish. Obesity is a major health problem in the United States.
yond the initial goals of the HMP. This project directly addresses the causes of obesity by testing the
Fifteen investigator-initiated projects have recently been Thrifty Microbiome Hypothesis, which poses that the gut
funded for an initial 1-yr pilot phase, during which each in- microbiome plays a key role in human energy homeostasis.
vestigator will have the opportunity to demonstrate the fea- Previous studies have indicated that a difference in the gut
sibility of his/her project (http://nihroadmap.nih.gov/hmp/ microbiome can be found in obese and lean adults. This study
fundedresearch.asp). At the end of the pilot phase, the subset of will perform a functional and genomic assessment of the gut
the projects that demonstrate the most promise, as judged by peer microbiome in donors whose genetics and phenotypic traits
review, will be scaled up and continued for three additional years. (weight, fat deposition, etc.) are carefully recorded. The Old
Collectively, the 15 pilot projects will study bacterial, fungal, and Order Amish population was chosen for this study because it is
viral changes related to various health conditions in microbiome genetically homogeneous and has already been characterized for
samples from consented human participants, including those that many of the traits being studied.
affect the skin, the nasopharynx, the oral cavity, the gastrointes- Vagina: Jacques Ravel and Larry Forney, PIs. The Microbial Ecol-
tinal tract, the genitourinary tract, and the blood. Individuals who ogy of Bacterial Vaginosis: A High-Resolution Longitudinal
provide the samples to be studied in the demonstration projects Metagenomic Analysis. Bacterial vaginosis (BV) arises in wo-
will provide consent analogous to those used for the Jumpstart men when the vaginal microbiome is disrupted. It is a common
condition that is very difficult to control. This project will test that have not previously been applied to cultivation of human
the hypothesis that vaginal microbiome dynamics and activities microbes; to trap large genomic DNA fragments by virtue of small
are indicators of risk of BV. The study will examine daily changes, amounts of known sequence within them; and to produce
over two menstrual cycles, in the vaginal microbiome of 200 microfluidic devices to capture large numbers of individual mi-
women and correlate them with occurrence of BV to better de- crobes and then amplify and sequence those that are recognized by
fine the syndrome and identify patterns that are predictive of BV. various criteria as being unknown or rare.
It will use 16S rRNA genes, metagenome and metatranscriptome
sequencing utilizing next-gen sequencing technologies, in Computational tools
combination with metabolomics, to assess the diversity of mi- Vast amounts of data will be generated by the HMP using next-gen
crobial species, genes, and functions of the microbiome associ- sequencing technologies and other high-throughput methods.
ated with these samples. Computational methods to process and analyze such data are in
Cancer of the GI Tract: Zhiheng Pei, PI. Foregut Microbiome in their infancy, and, in particular, objective measures and bench-
Development of Esophageal Adenocarcinoma. Esophageal ad- marks of their effectiveness have been lacking. Projects supported by
enocarcinoma (EA), the type of cancer linked to heartburn due to the HMP will address issues related to: genome assembly and gene-
gastroesophageal reflux diseases (GERD), is the fastest rising finding software for metagenomics data sets generated from new
malignancy in the United States. The recent increases cannot be sequencing technologies; characterization of biodiversity in sam-
explained by environmental or host factors. Initial research by ples; statistical models and simulations to compare different mea-
the PIs laboratory has shown that patients carrying particular sures of microbial diversity; and gene annotation tools for classifi-
types of microbiome are more likely to have the early stages of EA cation of protein families and functional prediction of HMP data.
than those that do not. The group will examine the 16S rRNA
genes and whole-genome sequence composition of the micro- Ethical legal and social implications (ELSI) of HMP research
biome in each stage of development of EA. If a significant asso-
Since its inception, the field of genomics has been characterized by
ciation of the changes in the microbiome during development
a component designed to anticipate and address the ethical, legal,
of EA can be shown, early diagnosis and treatment of the dis-
and social implications of genomics research at the same time that
ease may be possible through strategies that convert the disease-
the genomics research itself is being conducted. The findings of
related microbiome to the healthy microbiome.
these studies have in many cases been very influential in helping
Data Analysis and Coordination Center (DACC) to inform the way genomics research is carried out, while also in-
creasing the knowledge base regarding more distal applications of
The HMP Data Analysis and Coordination Center is an informatics
the science, including its broader societal impact. Although ELSI
resource that will make information about the HMP, including
research in the field of microbiology has a less well-established
results and conclusions, available to the scientific community. The
tradition, its importance in the field is increasingly being recog-
HMP DACC will also coordinate the development of data stan-
nized. Thus, an integral part of the HMP involves consideration of
dards and facilitate the analysis and deposition of data to the ap-
the potential ELSI issues that might be associated with microbiome
propriate public repositories (http://hmpdacc.org/). The HMP
research (Box 1). To date, five projects have been initiated as part of
DACC website contains the HMP Project Catalog of the reference
this component of the HMP. One of these projects will involve
strains. This searchable and sortable Project Catalog contains vari-
interviews with the individuals who donated the samples to be
ous types of information about each of the reference strains, in-
studied to develop the reference microbiome resource, as well as
cluding body site of isolation, sequencing status, and which center
with individuals who were asked to donate samples but declined,
is sequencing the strain. Links from the Project Catalog lead to
in order to explore general perceptions and attitudes in the public
more biological information and references for the strains.
about this new area of research. Other funded projects in this area
Technology development include studies to analyze how risk and benefit are conceptualized
in human microbiome research; to investigate patient perceptions
Producing a reference set of complete genome sequences is central
of bioengineered probiotics and clinical metagenomics; to ana-
to the HMP because such information is needed to interpret met-
lyze existing regulatory frameworks for the federal regulation of
agenomic sequence data. To sequence individual microbial ge-
nomes, it has generally been necessary to grow cultures of those
microbes; however, much of the microbiome cannot currently be
grown in culture. Therefore, the HMP is also supporting the de- Box 1. Ethical, legal, and social implications
velopment of new technologies that will allow the isolation of
many more purified microbial species so their genomes can be se- Unique challenges in recruiting research participants and obtaining
quenced, thereby expanding the reference collection. Among the informed consent (e.g., returning research results, dealing with
approaches supported are the development of methods to culture reporting requirements for communicable diseases)
Privacy issues (do different people have unique microbiome
previously uncultivable bacteria; to isolate single microbial cells; to signatures that can individually identify them?)
isolate, amplify, or clone unamplified or amplified DNA of whole New challenges to definitions of self, normality, abnormality,
genomes from individual cells at high fidelity and coverage; and to purity, and contagion
use tagging strategies to facilitate the enrichment of cells of a given Potential for individual or group stigmatization based on particular
microbiome characteristics
species to essential purity. Individual projects seek, for example, to Physical safety issues
collect the microbiome from distinct sites in the gut that would not Ethical implications of prenatal or neonatal manipulation of the human
routinely be sampled (e.g., tightly adherent to the intestinal mu- microbiome
cosa) and then enrich individual species by flow-sorting from Intellectual property issues
Ethical, legal, and regulatory issues in direct-to-consumer marketing of
those reduced-complexity samples; to reproduce in vitro growth
products based on microbiome research (e.g., probiotics)
conditions that mimic the microoxic environment in the gut and
probiotics; and to investigate the implications of research on the 10 countries. The IHMC is open to membership by any research
ancient and contemporary human microbiome for the social and project that is willing to agree to the IHMC principles (http://
ancestral identities of indigenous people. www.human-microbiome.org).