The Human Microbiome

INTRODUCTION TO THE HUMAN MICROBIOME

PROJECT 

In 2007, an NIH Roadmap for Medical Research Project called the Human Microbiome Project

(HMP) was initiated (http://nihroadmap.nih.gov/hmp/). The overarching goal of the HMP is

to develop tools and resources for characterization of the human microbiota and to relate

this microbiota to human health and disease. The HMP is leveraging the constantly advancing

sequencing and bioinformatics technologies to address the following broad goals:

· Determining whether individuals share a core human microbiome.

· Understanding whether changes in the human microbiome can be correlated with changes

in human health and disease.

· Developing the new technological and bioinformatics tools needed to support these goals.

· Addressing the ethical, legal, and social implications raised by human microbiome

research.

The HMP is a multiphase project that began with a “jumpstart phase” that involved four

genome sequencing centers: the Baylor College of Medicine Human Genome Sequencing

Center, the Broad Institute, the J. Craig Venter Institute, and the Genome Center at

Washington University. The goals of the jumpstart phase have been to sequence 900

reference genomes to provide a catalog of genomes for metagenomic studies, to sample at

least 300 healthy adults between 18 and 40 years of age at five body sites, and to develop

sequencing and analysis protocols for the samples derived from human subjects (88). The

second phase of the HMP includes human microbiome studies that target particular disease

states. While the human microbiome includes bacteria, viruses, and small eukaryotes, such

as fungi, this chapter focuses on the bacterial members of the microbiome.

TECHNIQUES FOR THE STUDY OF THE HUMAN

MICROBIOME Back to top

Early studies of the human microbiome relied on culture-dependent methods; however, it is

now known that the majority of microorganisms from the human body cannot be cultured in

vitro. Most current techniques for characterization of a metagenomic sample are polymerase

chain reaction (PCR) based and target the highly conserved bacterial 16S ribosomal small

subunit RNA. Portions of the gene can be amplified and fingerprinted by using electrophoretic

techniques, such as terminal restriction fragment length polymorphism (TRFLP) and

denaturing gradient gel electrophoresis. Full-length 16S rRNA genes or segments of these

genes can be amplified prior to microarray analyses or DNA sequencing studies. 16S rRNA

gene sequences that are ≥97% identical are considered to be within the same species, while

those that are ≥95% identical are within the same genus. Currently, metagenomic samples

are most often analyzed by sequencing of 16S rRNA gene or gene fragment amplicons by

direct whole-genome shotgun (WGS) sequencing (5, 27, 33).

Culture-dependent and culture-independent surveys have shown that the human body is

home to only four predominant phyla (Table 1). The Firmicutes and Actinobacteria are the

most highly represented phyla when all body sites are considered. In a recent study, four

phyla comprised 92.3% of bacterial DNA sequences analyzed from multiple human sources,

including hair, oral cavity, skin, and gastrointestinal tract (21). The predominant phyla vary

by anatomical site, and presumably the host milieu has a crucial role in shaping the

composition of microbial communities at each site. Recent efforts have been aimed at

expanding the DNA sequence representation within each phylum so that more

comprehensive phylogenetic assessments can be performed in the future (109).

The Oral Microbiome

The oral cavity includes various ecologic niches, such as saliva, gingival crevices, the tongue

surfaces, and the posterior pharynx, and is colonized with hundreds of species of bacteria.

Estimates range from 500 to 700 or more different species (82, 83). The dental plaque

biofilm found on the surface of the teeth and in the subgingival pocket represents a complex

assemblage of microbes (1, 82), and such structured microbial communities reside in

intimate contact with host tissues. Health-associated microbial communities may protect

against infection, but the oral cavity also harbors organisms that are implicated in both local

and systemic diseases, including periodontal diseases (92), endocarditis (10), and aspiration

pneumonia (98). The relative preponderance of health- or disease-associated microbes

combined with human genetic susceptibilities may ultimately account for different disease

phenotypes.

Periodontal disease includes conditions of oral inflammation and oral infections that may be

associated with the composition of tooth-borne microbiomes in adults. It is a condition with

various degrees of inflammation and is considered to be an infectious disease. Socransky et

al. have proposed that periodontitis is the result of complexes or consortia of pathogens

(100). The so-called red complex, containing Porphyromonas gingivalis, Tannerella

forsythia, and Treponema denticola, is the pathogenic group that seems to be most strongly

associated with disease (49, 100). Periodontitis is associated with systemic disease, such as

coronary heart and cerebrovascular diseases (26). Antibody responses to oral bacteria

suggest that immune responses to specific microbial components may contribute to

conditions of chronic inflammation (8, 9). Progressive periodontitis in pregnant women has

been reported to increase the risk of severely adverse pregnancy outcomes (40, 80). These

associations underscore the significance of the oral microbiome to systemic health status and

predisposition to specific diseases.

16S rDNA sequencing and other techniques have been used to evaluate the oral microbiome.

These methods include denaturing gradient gel electrophoresis (63, 64, 91), TRFLP

(50, 51, 94, 95), and checkerboard DNA-DNA hybridization, where 45 DNA samples can be

queried against 30 to 40 DNA probes (101). The last technique was used by Socransky and

coworkers to examine microbial communities in supragingival (44) and subgingival (100)

plaque. In both studies, distinct complexes were identified by principal component and

correspondence analyses and were assigned to color groups. The supragingival plaque

samples from 187 subjects (4,475 samples total) clustered into six groups (Fig. 1), including

the aforementioned pathogen-associated red complex. In the subgingival study, a similar

clustering was revealed, though the blue complex, composed of Actinomyces species, was

not observed in these samples.

The Gastric Microbiome

The low pH and rapid peristalsis in the stomach suppress persistent colonization by many

bacteria. The stomach and small intestine each are thought to contain about 100 culturable

organisms per milliliter, but the organismal counts can increase to 105 per ml following a

meal (66). The best-studied and most dominant member of the stomach microbiota

is Helicobacter pylori (22). Culture-dependent methods have revealed other genera, such

as Lactobacillus, Streptococcus, and Staphylococcus, as well as members of

the Enterobacteriaceae,in the stomach (2, 103), although a metagenomic analysis of gastric

biopsy specimens revealed far more diversity (128 phylotypes) than had been appreciated

previously by culture-based approaches (11).

The Small Intestine Microbiome

Like the stomach, the small intestine is colonized by relatively low numbers of bacteria,

especially in proximal regions, such as the duodenum and jejunum. Bile in the small intestine

inhibits bacterial colonization. The organisms found in the small intestine are usually

lactobacilli, enterococci, other gram-positive aerobes, and facultative anaerobes (97, 111).

This portion of the GI tract has been assessed mainly by culture-dependent methods; no

metagenomics studies of the small intestine have been reported. A recent quantitative-PCRbased

study described the differences in microbial composition between ileostomy specimens

and intact small intestinal tissue (45). More distally in the ileum, the microbial composition

becomes more complex and approaches that of the colon in terms of species richness and

the nature of predominant bacterial genera.

The Colon Microbiome

The human colon is colonized by 1011 to 1012 bacteria per gram; this number represents at

least 800 bacterial species and nine phyla (and one archaeal phylum), the majority of which

are obligate anaerobes of the Bacteroidetes and Firmicutes (7, 29). Some studies have

suggested that 15,000 to 36,000 species are present in the intestine (32, 87).

Bacteria from the distal gut are critical to host nutrition and may play key roles in health and

disease. Microbe-derived carbohydrate fermentation by-products, such as short-chain fatty

acids (SCFAs) like butyrate, acetate, and propionate, provide 10% or more of the body’s

metabolic requirements (65). Butyrate, produced by the clostridial clusters IV and XIVa, is

the primary energy source of the colonic epithelium, and this SCFA has been reported to

possess anticancer features (6, 47, 73). SCFAs also may play a role in preventing ulcerative

colitis (19, 76) and infection by pathogens such as Salmonella enterica serovar Typhimurium

(59). Colonic bacteria further contribute to nutrition by synthesizing amino acids (74) and

vitamins (e.g., K, B12, biotin, folic acid, and pantothenate) (3, 46).

Studies in germfree mice have indicated that the gut microbiota is important for the

maturation and function of the mucosal immunity (72). Intestinal epithelial cells are in direct

contact with the lumen and are involved in signaling to host innate and adaptive mucosal

immune responses. Interactions between commensal ligands and the Toll-like receptors are

critical for maintenance of epithelial homeostasis in the gut (90). The microbiota directs

production of pathogen-specific mucosal immunoglobulin A (67). Bacteroides

fragilis produces a polysaccharide that can correct T-cell deficiencies and T helper 1 and T

helper 2 imbalances that lead to immune maturation (72). Some Lactobacillus species reduce

cytokine responses to lipopolysaccharide, resulting in decreased inflammation (86).

Several groundbreaking gut microbiome studies have been reported in the literature during

the past several years. Eckburg et al. sampled six sites within the colon of three individuals

and performed 16S rDNA analysis of these samples plus fecal samples (29). Most of the

sequences represented uncultured or novel microorganisms. Of the sequences that could be

characterized, the majority were members of the Firmicutesand Bacteroidetes, with 95% of

the Firmicutes belonging to the class Clostridia. The compositions of the six mucosal sites

clustered together, but the mucosal microbiomes differed in composition from those found in

stool specimens. Since the fecal microbiome overlaps that of the colonic lumen but is not

identical, the authors suggested that the stool bacterial population is composed of a

combination of mucosa-associated bacteria and a nonadherent luminal population (29). The

distinction between microbiomes intimately associated with the intestinal mucosa and those

associated with fecal specimens is important to consider, because many studies rely on fecal

samples alone.

In another deep sequencing study, both 16S rDNA gene and WGS reads from fecal samples

of two individuals were analyzed (39). The communities represented

only Firmicutes and Actinobacteria plus one archaeon,Methanobacter

smithii. No Bacteroidetes were observed, in contrast to other studies. This key difference

may be due to the individuals sampled (e.g., diet or genotypes) or methodological details

(e.g., the PCR primers used). More than 50,000 open reading frames were predicted and

examined for enrichment of COG (clusters of orthologous groups) and KEGG (Kyoto

Encyclopedia of Genes and Genomes) pathways. Analysis of these pathways revealed

enzymes required for degradation of plant polysaccharides, production of SCFAs, vitamin

biosynthesis, methanogenesis, and degradation of toxic plant phenolics. These results are

consistent with premicrobiome studies examining the metabolic capabilities of the gut

microbiota.

Tap et al. identified 66 OTUs that are common to 50% or more of fecal samples from 17

healthy adult donors (105). This “core” microbiome was dominated by seven species (Fig. 3).

Fifty-seven of the OTUs belonged to the Firmicutes phylum, and seven belonged to

the Bacteroidetes. Forty-two percent of the OTUs could not be assigned to a species. The

authors also compared these OTUs to those discovered in other published gut metagenomics

analyses (29, 39, 62, 68). All 66 OTUs were detected at least once in each of the other

studies (Fig. 3, inset). The authors concluded that the core human intestinal microbiome is

composed of approximately 50 bacterial species.

The Genitourinary Microbiome

The Vaginal Microbiome

The vaginal microbiota plays an important role in preventing genital and urinary tract

infections. It is known that the composition of the vaginal microbiota varies with age, pH,

and hormonal levels (37). Lactobacilli are usually considered to be the most prevalent

organisms in healthy premenopausal women (57) and are considered protective for the host

by virtue of their presumed role in suppression of pathogen colonization. Such effects may

result from mucus adherence by lactobacilli, production of organic acids and reduction of

vaginal pH, and production of antimicrobial compounds that prevent pathogen proliferation

(12). In addition to the lactobacilli, the predominant culturable vaginal microbes

are Mobiluncus spp., Gardnerella vaginalis,Bacteroides spp., Prevotella spp., and Mycoplasma

hominis (48).

Most 16S rDNA-based metagenomics studies have demonstrated that the vaginal microbiota

are highly variable, and differences presumably depend on differences in sexual and hygienic

practices in addition to host genetics. Lactobacillus is the predominant member of the vaginal

community in most individuals, but in some cases, vaginal lactobacilli may be undetectable.

The microbiotas of eight healthy women with three different “grades” of vaginosis were

examined by sequencing the V1 to V3 16S region by the Sanger method (107). The biotas

associated with each of these grades were quite distinct: grade one individuals (healthy)

were almost exclusively colonized by Lactobacillus crispatus, Lactobacillus

gasseri, and Lactobacillus jensenii;grade two subjects had Lactobacillus iners, Atopobium

vaginae, Prevotella bivia, and Sneathia sanguinegens; and grade three subjects were

predominantly colonized with A. vaginae or Peptostreptococcus anaerobius. Several studies

have also associated altered vaginal microbiota with an increased risk for viral coinfection

(55, 77, 99).

In 2007, the Forney group published the results of a 16S rDNA survey of the midvaginal

microbiota from five healthy Caucasian women (112). The authors targeted the V1 to V5

region of the 16S rRNA gene and classified ca. 1,200 clones. The number and distribution of

phylotypes differed among the five subjects. Two females had exclusively or nearly

exclusively L. crispatus, one had L. iners as the predominant species, and one was

predominantly colonized by A. vaginae. The fifth subject had seven phylotypes,

predominantly L. inersbut also significant amounts of A. vaginae,

Megasphaera, and Leptotrichia. Like lactobacilli, Leptotrichia andAtopobium are lactic acid

producers. This suggests that different microbial communities may manifest shared

functions. A study of the vulval microbiota of four of the five women from the prior study was

published by the same group (15). While the communities in the labia majora were more

diverse than those from the vaginal samples, the general trend

of Lactobacillus predominance was reported for four of the five women. The overall

conclusion was that the vaginal and vulval microbiotas are highly variable. These results

have in general been confirmed by another group that sequenced full-length 16S rDNA

clones (2,000 reads per each of 20 healthy subjects) (54).

The Microbiome of the Upper Respiratory Tract

The healthy nares and nasopharynx contain streptococci, staphylococci,

corynebacteria, Moraxella spp.,Neisseria spp. (including N. meningitidis),

and Haemophilus (53). Viridans streptococci predominate, and organisms associated with

inner ear infections of children are also found (e.g., Streptococcus pneumoniae, Haemophilus

influenzae, and Moraxella catarrhalis), though the numbers of these organisms vary

significantly with age (56). The carriage rate for Staphylococcus aureus has been estimated

to be ca. 30%, with methicillin-resistant S. aureus representing 1.5% of the strains isolated

(41). The paranasal sinuses are normally sterile, but they can become infected because of

their close proximity to other sites that are colonized by bacteria.

The Skin Microbiome

Most skin microbes are gram-positive organisms, such as staphylococci, micrococci,

brevibacteria, propionibacteria, and corynebacteria (61, 93). Colonization by gram-negative

organisms was previously thought to be extremely rare (18), although Acinetobacter spp.

could sometimes be cultured. Metagenomics studies have changed these views. The skin

microbiota generally protects individuals against colonization by pathogens, but under

certain circumstances the microbiota may be pathogenic if these organisms can penetrate

the skin in a susceptible host. For example, Staphylococcus epidermidis is a common skin

colonizer. However, this species may cause infections in immunocompromised patients or

those with indwelling devices. Conversely, Pseudomonas fluorescens is thought to be a

protective skin organism, because it produces the polyketide antibiotic mupirocin (34), which

is active against gram-positive bacteria, including methicillin-resistant S. aureus (104).

A 16S rDNA survey of samples from the human volar forearm (six healthy subjects; 1,221

16S rDNA clones total) revealed that 94.6% of the sequences fell into three phyla —

Actinobacteria, Firmicutes, andProteobacteria—yet the total diversity represented 182

species-level OTUs (36). Chao estimation suggested that the communities were composed of

ca. 250 OTUs, and the results indicated that communities varied substantially between

subjects. Samples from the antecubital fossae of five healthy subjects (ca. 200 clones per

subject) revealed 113 OTUs belonging to six bacterial divisions and a predicted community

size of 130 OTUs (43). In contrast to the study by Gao et al. (36), this study revealed that

the Proteobacteria predominate the antecubital fossae of the subjects tested. It was

proposed that the differences might be representative of the different environments of the

two sites: the volar forearm is dry and hairy, while the antecubital fossa is sweaty and

hairless.

The differences in the composition of the skin microbiome based on microenvironment were

underscored in a 2009 report by Grice et al. (42). Twenty sites on 10 healthy volunteers

were sampled, and full-length 16S amplicons were sequenced. The skin sites included

sebaceous, moist, and dry sites. While a significant amount of variability was observed

between subjects, clustering was apparent by site type (Fig. 4). The sebaceous sites were

dominated by propionibacteria and staphylococci and the moist sites by corynebacteria, while

the dry sites contained a mixed population of bacteria. The volar forearm had the greatest

number of OTUs (44 OTUs), while the retroauricular crease was the least rich with 15.

Overall, the sebaceous sites were the least complex, and intrapersonal variation was

generally less than interpersonal variation. Temporal intrapersonal variation, based on a

second sampling of five subjects 4 to 6 months after the primary sampling, revealed

similarities for samples from the external auditory canal, inguinal crease, alar crease, and

naris. In contrast, considerable variation was observed during this time period for samples

taken from the popliteal fossa, volar forearm, and buttock.