Two new intestinal microbiome initiatives involving large amounts of sequencing plan to focus on how microbes in the human gut contribute to metabolic and bowel disease.
In Europe, Metagenomics of the Human Intestinal Tract, or MetaHIT, a four-year consortium recently funded with €11.4 million ($17.9 million) from the European Commission, will seek to link these microbes to obesity and inflammatory bowel disease.
Late last month, the Crohn’s and Colitis Foundation of America announced the Gut Microbiome Initiative, a three-year, multi-million study led by researchers at Washington University to help uncover the role of gut microbes in IBD.
MetaHIT involves 13 research institutions and companies from France, Denmark, Germany, Spain, Italy, the Netherlands, the UK, and China. The €11.4 million EC funding, which was granted through the 7th Framework Program and started in January, covers only part of the project’s anticipated €20 million overall budget. Participants will contribute the remainder. MetaHIT had its formal “kick-off” meeting in April and plans to meet again this fall to discuss its progress.
Sequencing for various aspects of the project will be provided by Genoscope, the French National Sequencing Center; the Wellcome Trust Sanger Institute; the University of Southern Denmark in Odense; and the Beijing Genomics Institute.
MetaHIT’s first aim is to generate a catalog of bacterial genes and genomes that can be found in the human gut in heath and disease, to be used as a reference in comparative studies, according to Dusko Ehrlich, who coordinates the project and heads microbial genetics at the Institut National de la Recherche Agronomique in Joy en Josas in France.
The catalog will contain information from shotgun metagenomic sequencing as well as microbial reference genomes, generated both by MetaHIT and by other projects, such as the US Human Microbiome Project and research groups in Japan.
For MetaHIT’s metagenomic reference-sequencing project, the researchers plan to choose eight individuals: two obese, two with IBD, and four controls. “That is a challenge because we know that in some people, a few [microbial] species are quite dominant,” Ehrlich explained. In addition, they want to pick those individuals who cover a wide spectrum of gut microbes, he said.
To select samples with diverse microbiomes and few dominant species, the scientists will initially screen 40 samples from a cohort of obese individuals and 40 samples from an IBD cohort using 16S RNA microarrays developed by researchers at Wageningen University in the Netherlands.
Based on the results, they will select eight samples for metagenomic sequencing, although “if we get the impression that we can get better coverage of the diversity by looking at a few more normal or a few more obese [or IBD samples], we will do that,” according to Ehrlich.
Genoscope, which is based in Evry Cedex, will be in charge of sequencing the samples. The sequencing center, which participated in the Human Genome Project, has lately been focusing on environmental genomics projects, according to its website. Ehrlich said Genoscope’s original plan was to use dideoxy Sanger sequencing, and MetaHIT has funding to generate approximately 1.5 gigabases worth of sequence data using that technology.
But Genoscope is currently evaluating 454’s sequencing technology and might switch over to that platform, Ehrlich said, provided the researchers “convince themselves that the sequence quality will be adequate, and then convince everybody else in the consortium that it is better to spend the money that way.”
Jean Weissenbach, who heads Genoscope, did not respond to several requests for comment before deadline.
The second part of the catalog will consist of microbial reference genomes, of which MetaHIT currently plans to generate approximately 100. “We have not made any rigid decisions about what technologies to use for that,” said Julian Parkhill, head of pathogen genomics at the Sanger Institute, which is responsible for reference genome sequencing.
“It is very important to sample humans living under different conditions to understand how varied our microbial ecology is.”
Initially, the Sanger scientists will “almost certainly be using long paired-end 454 libraries” with 4-kilobase inserts, the goal being to generate draft assemblies rather than finished genomes, Parkhill told In Sequence last week.
The researchers are also currently testing paired-end sequencing with Illumina’s Genome Analyzer and ABI’s SOLiD platform for microbial genome sequencing. “All the data points to the fact, at the moment, that we should get reasonably good assemblies if we have long paired ends with Illumina’s technology, so we may well move on to that,” Parkhill said.
He and his colleagues have not yet chosen which microbial strains to sequence, and will coordinate closely with the National Institute of Health’s Human Microbiome Project, which plans to generate approximately 900 microbial genomes (see In Sequence 5/20/2008).
“The intention is to choose some initial cultured strains,” Parkhill said. “As the rest of MetaHIT gets into gear, they will start identifying interesting species and strains through the biological efforts, [and] strains identified by the other groups will be fed back [to us for sequencing].”
Some of the reference strains will be unculturable, Parkhill said, and as part of the project his group will explore methods for sequencing those species. These methods will include laser microdissection and other techniques for isolating bacteria, as well as chromosome amplification methods.
Parkhill’s group has approximately €1 million in total funding for the reference genomes project, which includes the cost of sequencing as well as maintaining strains and generating annotation pipelines.
Though the funding was originally calculated based on the estimated cost of sequencing 100 reference genomes, since the cost of sequencing continues to decline, “it is very likely that over the course of the next two or three years we will be able to make that funding stretch to do more” than the original 100 strains, Parkhill said.
Once the catalog of genes and genomes is established, MetaHIT researchers plans to profile the intestinal microbiomes of approximately 400 individuals, 200 each for an obesity and an IBD study, using both microarrays and high-throughput short-read sequencing.
In the obesity study, the researchers want to compare microbial communities of individuals with normal weight, ventral obesity, and subcutaneous obesity. For the IBD study, they plan to compare gut microbiomes between patients with Crohn’s disease, ulcerative colitis, and controls, and intend to follow changes in their microbiomes over time.
The researchers will likely use custom-designed NimbleGen arrays for the array part of the project “because they have the highest density of features, and [NimbleGen] can change the design quite flexibly,” which would allow the scientists to add more features to the array if the catalog of reference genes expands.
In parallel, they will also profile the samples using short-read technologies, which would enable them to “see not only things that you know — things that you have put on an array — but also [to] discover many things that you don’t know,” he said. The original plan was to use Illumina’s Genome Analyzer, but “maybe we will use [ABI’s] SOLiD as well,” he said.
If short-read sequencing reveals that certain samples are not well covered in the reference catalog, Genoscope will sequence them in greater depth, he said.
Jun Wang, a professor at both the University of Southern Denmark in Odense and at the Beijing Genomics Institute in Shenzhen, will oversee the sequencing-based profiling.
“It is tag-based profiling work, so we will use mapping technology to map all the short reads to the reference genome [and] to evaluate the profiling [regarding] the composition of the metagenome,” Wang told In Sequence by e-mail last week. “We could also use the data to do de novo assembly, but [this] is not the major purpose of my part of the project.”
Both USD and BGI will generate sequence data for the profiles, according to Wang. BGI currently has 18 Illumina Genome Analyzers and two SOLiD instruments, while the University of Southern Denmark owns one Genome Analyzer, he said.
By the end of the year, MetaHIT plans to have the initial gene catalog from metagenomic sequencing as well as several microbial genomes completed, and hopes to have several sample profiles done as well, according to Ehrlich. Peer Bork’s group at the European Molecular Biology Laboratory in Heidelberg will focus on the bioinformatics analysis of the data.
CCFA’s Gut Microbiome
Another project that aims to study intestinal microbes involved in disease is the Gut Microbiome Initiative, launched late last month by the Crohn’s and Colitis Foundation of America.
The foundation is funding the research of Jeff Gordon, director of Washington University School of Medicine’s Center for Genome Sciences, with approximately $1 million this year, a CCFA spokesperson told In Sequence last week. The project also involves collaborators at the University of Colorado at Boulder.
In total, the CCFA wants to invest several million dollars in the three-year initiative, according to the spokesperson.
Gordon told In Sequence this week that his group has been studying relationships between gut microbes and their hosts for several years, most recently working with 454’s sequencing technology.
Although the NIH-funded Human Microbiome Project will focus on disease in this and other areas as well, most of its grants will only be awarded next year.
In the meantime, CCFA will “jump-start research in areas related to gut microbes and IBD,” Gordon said, focusing initially on ways how to best study and analyze gut microbiomes.
The CCFA knows that “it’s important [that] some basic studies of healthy populations be forthcoming so that we better understand the variations that exist within an individual and between individuals, before we try to correlate particular microbial community structures with a disease state,” Gordon said.
Under the CCFA’s grant, he and his colleagues plan to develop new experimental designs, techniques, and computational methods “that will be valuable to researchers who are interested in IBD,” he said.
For example, he and his colleagues will develop new computational methods to compare large datasets of 16S ribosomal RNA. “The amount of 16S ribosomal RNA sequences generated has accelerated dramatically, both in terms of the method and speed of generation of these sequences but also the size of the datasets,” he said. “There are a lot of challenges about how to deal with these large datasets.”
Last month, for example, his group published a study online in Science in which they compared 16S ribosomal RNA sequences from the human gut with those of 59 other mammalian species.
Gordon’s group also wants to develop tools to compare data from shotgun metagenomic studies. Most recently, his group has characterized the microbiomes of identical twins and their mothers, a study Gordon plans to publish soon. “One of the functions CCFA would like us to address is, what is the degree of variation in the microbial gene composition of communities, and what are the reference controls that should be used?”
Both Gordon and MetaHIT’s Ehrlich agree that there is room for several initiatives focusing on gut microbiomes. “There is so much to do,” Ehrlich said.
“I think that these efforts are complementary, not competitive,” said Gordon, who is also an advisor to MetaHIT. “It is very important to sample humans living under different conditions to understand how varied our microbial ecology is.”