National Science Foundation news release

Decoding of a variety of plant genomes could accelerate due to two complementary methods that remove from
analysis vast stretches of DNA that do not contain genes.

The approaches, applied jointly in efforts to determine the gene sequences in maize, are described in the Dec. 19 issue of the journal Science.  The evaluation of these methods and the assembly of the resulting sequences were undertaken by two groups led by researchers from The Institute for Genomic Research (TIGR) in Rockville, Md., and Cold Spring Harbor Laboratory in New York.

The research was funded by the National Science Foundation's (NSF) Plant Genome Research Program.

Only about a quarter of the maize genome codes for genes, and these are found in small clusters scattered through a mixture of non-coding DNA and transposons (mobile DNA segments).  Two different methods tested by the TIGR group successfully captured parts of the maize genome containing genes.  The gene-sequences are of most interest because they provide the specific blueprint for an organism's development, structure and physiology.

With so much non-gene sequence to deal with, it has not been feasible to sequence and assemble the whole maize genome with current technologies.  Thus, it is a major shortcut to capture only the portion of the maize sequence containing its genes without having to sequence the entire genome. 

"Collecting the maize genes for sequencing is like panning for gold," said Jane Silverthorne, program director for NSF's plant genome program.  "Just as gold can be separated from the surrounding rock because it is denser, maize genes can be separated from the surrounding DNA by their chemical and sequence properties."

The first method tested, called methylation filtration, removes sequences that contain a chemical modification (methylation) found on most of the repeated sequences and transposons, leaving behind the proverbial gold of genes.  It was developed by a team led by Robert Martienssen and W. Richard McCombie at Cold Spring Harbor Laboratory.

The second method, developed by researchers at the University of Georgia, removes the repeated sequences by separating the DNA into "high-copy," gene-poor segments and "low-copy," gene-rich segments.

Led by Cathy Whitelaw, the research team at TIGR compared sequences obtained by the two methods.  About one fourth of the genes in each collection matched known gene sequences.  About 35 percent of the genes were represented in both collections.

Each method was found to enrich for distinct but complementary regions of maize's 10-chromosome genome.  Combined, the methods could cut the amount of sequencing necessary to find all of the maize genes to about one-fourth of what it would take to sequence the entire genome.

As both methods yielded short stretches of sequence, a major challenge was to reassemble these into complete genes.  To do this, the Cold Spring Harbor group lined up the sequence pieces from maize along the rice genome sequence, a deep draft of which was completed in 2002 by an international consortium.  The researchers then reassembled selected sets of sequence fragments into complete genes.  This approach will be an important part of assembling the short pieces of DNA yielded by the two enrichments methods into complete gene clusters.

According to Silverthorne, "Together, these findings suggest that scientists could be able to sift out the approximately 450 million base pairs of DNA containing the genes from the maize genome and then reassemble the sequence.  Such a comprehensive genomic resource would provide growers and breeders a wealth of tools to improve maize, as well as other cereal crops."

Other collaborators in the study included the Donald Danforth Plant Science Center and Orion Genomics, LLC, both of St. Louis, Missouri.

Less is more: New technology captures gene-rich DNA segments
Sequencing key regions speeds genome research in corn and other important crop species

Cold Spring Harbor Laboratory news release

Obtaining genome sequence information frequently leads to breakthroughs in the study of a particular organism. Bringing agriculturally important plant species into the genomic age is therefore an important goal. However, because they are typically larger or much larger than the 3-billion letter human DNA sequence and have a high proportion of so-called repetitive DNA that is difficult to sequence and contains few coding regions or genes, the genomes of many plants--including most agriculturally important species--have posed significant challenges to researchers interested in crop improvement, plant molecular biology, or genome evolution. A new study by Cold Spring Harbor Laboratory researchers is a significant step toward overcoming those challenges.

By applying a method they recently developed that captures gene-rich regions and excludes the vast majority of repetitive, gene-poor DNA, Cold Spring Harbor Laboratory researchers have now achieved a dramatic shortcut to sequencing the genes of corn. The approach should provide a similar boost to the sequencing and comparative analysis of other genomes in a wide variety of biological, biomedical, and biotechnological settings.

The study, led by Cold Spring Harbor Laboratory scientists W. Richard McCombie and Robert Martienssen, is published in the December 19 issue of Science along with a related study carried out by researchers at The Institute for Genomic Research in Rockville, Maryland. A key method used in both studies, called methylation filtration, was developed in 1999 by McCombie and Martienssen's groups through work funded by the U.S. Department of Agriculture.

Methylation filtration relies on the observation that the DNA of repetitive, gene-poor regions in the corn genome (and other plant genomes) is modified by a process called methylation, whose study has been pioneered in part by Martienssen's group. Methylation filtration takes advantage of this observation to preferentially capture the unmethylated, gene-rich regions of the corn genome for subsequent analysis. Indeed, the new study demonstrates that methylation filtration removes 93% of repetitive, gene-poor DNA. As a result, the researchers were able to focus their efforts on the sequencing and analysis of the gene-rich regions of the corn genome.

"This study establishes that methylation filtration, combined with other simple techniques, can be used to successfully recover and properly assemble complete gene sequences from genomes that are otherwise extraordinarily difficult to decipher," says McCombie. "Moreover, both studies involved large-scale tests that validated our initial estimates regarding how well the procedure would work. Perhaps most importantly, we've shown that after gene-enriched draft DNA sequences are obtained, they can be converted into the complete sequence of the corn genes by using the related, but much smaller rice genome sequence as a guide. We believe that taking this short-cut approach has brought us a very close to a final sequence map of the biologically important regions of the corn genome at a fraction of the cost of other approaches," adds McCombie.

The rice genome, which is about 1/6 the size of the corn genome, is being sequenced as part of an international consortium funded in the United States by the National Science Foundation and the U.S. Department of Agriculture. Corn is the most important agricultural crop in the U.S. Because the genome structures of wheat, oats, barley, and many other crops are quite similar to that of corn, the approaches outlined by the new study provide the means to bring investigations of all of these important crops into the genomics era.

The study was funded by the National Science Foundation Plant Genome Research Program. Dr. Jane Silverthorne, Director of NSF's Plant Genome Research Program, says, "The success of this project highlights the importance of virtual center projects in bringing together the expertise required to tackle large complex problems in genomics."

Scientists discover way to streamline analysis of maize genome

Rockville, Maryland
TIGR news release

Combination of Two Techniques Can Help Identify "Gene Islands" in the Key Crop

Like tiny islands in a vast sea, the gene clusters in maize are separated by wide - and extremely difficult to decipher - expanses of highly-repetitive DNA. This complex structure has greatly complicated efforts to sequence the genome of maize, which is one of the world's most important crops.

In an effort to streamline the way that researchers identify and sequence the DNA in those gene-rich islands, scientists at The Institute for Genomic Research and collaborators have discovered that two different approaches to identifying the non-repetitive regions of the genome together provide a complementary and cost-effective alternative to sequencing the entire genomes of complex plants.

In a paper published in the December 19th issue of the journal Science, the researchers found that two independent gene-enrichment techniques - methylation filtering and High-C₀t selection - target somewhat distinct but overlapping regions of the genome and therefore could be used together to help identify nearly all of the genes in maize as well as their genomic structures.

This finding is significant because the maize genome, which includes about 2.5 billion base pairs of DNA, is about 20 times larger than the first plant genome to be deciphered, Arabidopsis thaliana, and nearly six times larger than the rice genome. The reason that the maize genome is so large is that approximately 80% consists of families of nearly identical repetitive sequences. The gene-containing sequences are concentrated in the remaining 20% of the genome.

The challenge for genomic researchers is to explore the gene-rich islands without having to negotiate through the sea of highly-repetitive DNA surrounding them. In the Science study, researchers reported on two "filtration" techniques that separate the gene-rich regions from the gene-poor ones, providing about a four-fold reduction in the amount of sequencing necessary to find all of the maize genes.

"A combination of these techniques may be an excellent method for sequencing maize as well as other large and complex plant genomes at a cost far lower than current approaches," says Cathy A.Whitelaw, the TIGR researcher who led the maize analysis project and is the first author of the Science paper.

The major collaborators for the study were the Donald Danforth Plant Science Center in St. Louis, MO.; the University of Georgia's genetics department in Athens, GA; and Orion Genomics, in St. Louis. The project was sponsored by the National Science Foundation's Plant Genome Research Program.

"The success of this project highlights the importance of virtual center projects in bringing together the expertise required to tackle large complex problems in genomics," says Jane Silverthorne, who leads the NSF's plant genome program.

TIGR Investigator John Quackenbush, the paper's senior author, says, "Maize is the single largest food crop in the United States, so developing strategies to decode its complex genome is a high priority. More importantly, the techniques that we have developed will be useful in the analysis of many other crops such as soybean whose genomes are also highly repetitive."

The two filtration techniques - methylation filtering and High-C₀t selection - are not new, but this was the first time that they were tested together on a major scale, in this case with a combined total of about 93 million DNA base pairs from the maize genome.

The methylation filtering technique excludes hyper-methylated DNA sequences (a characteristic of highly-repetitive DNA) by means of bacterial restriction systems that cleave those areas of the genome. The technique was first developed by scientists at Cold Spring Harbor Laboratory.

The High-C₀t selection technique, developed by researchers at the University of Georgia's genetics department, excludes highly-repetitive DNA sequences by using a different method that separates DNA segments into "low-copy" (High C₀t) and "high-copy" (Low C₀t) sequences, which correspond roughly to gene-rich and gene-poor sections of the genome, respectively.

When researchers analyzed the composition of Simple Sequence Repeats - short, repetitive segments of two, three or four DNA bases - recovered from the two techniques, they were able to show that the filtration methods targeted different regions of the maize genome.

An analysis of "genetic markers" - sequences related to the maize genetic map - reinforced that conclusion and further indicated that these methods do not have significant biases, as newly-sequenced regions are evenly distributed across the 10 maize chromosomes.

"While both of these methods increase the rate of gene identification from maize genomic sequence, our analysis implies that they have biases; this suggests that both methods are required to ensure comprehensive coverage of the maize gene space," says W. Brad Barbazuk, Ph.D., senior bioinformatics specialist at the Danforth Center.

TIGR's President, Claire M. Fraser, calls the maize study is an important step in tackling the genomes of complex plants: "Not only has this project given us a preview of the structure of the maize genome, it also has helped us find a rapid and cost-effective alternative to sequencing the entire genome."

The Institute for Genomic Research (TIGR) is a not-for-profit research institute based in Rockville, Maryland. TIGR, which sequenced the first complete genome of a free-living organism in 1995, has been at the forefront of the genomic revolution since the institute was founded in 1992. TIGR conducts research involving the structural, functional, and comparative analysis of genomes and gene products in viruses, bacteria, archaea, and eukaryotes.

Danforth Center maize genome pilot sequencing project results in six-fold reduction of effective size of maize genome

St. Louis, Missouri
December 19, 2003

Initial Results From NSF-Funded Project May Serve As A Cost Effective Model For Sequencing Large Complex Genomes

As reported in the December 19, 2003 issue of Science magazine, the Maize Genomics Consortium, led by scientists at the Donald Danforth Plant Science Center, has evaluated and validated a gene-enrichment strategy for genome sequencing resulting in a six-fold reduction of the effective size of the Zea mays (maize or corn) genome while creating a four-fold increase in the gene identification rate when compared to standard whole-genome sequencing methods.

The Maize Genomics Consortium, consisting of The Donald Danforth Plant Science Center, The Institute for Genomic Research (TIGR), Purdue University, and Orion Genomics, was awarded a two-year, $6 million plant genome grant on September 20, 2002 by the National Science Foundation (NSF) to develop and evaluate high-throughput and robust strategies to isolate and sequence maize genes. The two gene-enrichment methods used in the research published in Science are methyl-filtration and high-Cot selection.

According to Karel R. Schubert, Ph.D., principal investigator and vice president of technology management and science administration, and W. Brad Barbazuk, Ph.D., senior bioinformatics specialist and assistant domain member, both at the Donald Danforth Plant Science Center, the overall goal of the pilot sequencing project in maize is to derive an effective strategy to sequence the maize genome. To meet this goal, the Maize Genomics Consortium will generate approximately 800,000 total sequence reads using the methyl-filtration and high-Cot methods, with the results published in Science describing the analysis of the first 200,000 sequence reads.

It is a challenging effort to sequence the maize genome, as its size and structure preclude using the standard whole-genome methods for sequence analysis and alignment. At about 2 to 3 billion base pairs, the maize genome is estimated to be 20 times larger than Arabidopsis, the first plant genome to be completely sequenced. However, maize probably has only twice as many genes as Arabidopsis. The rest of the maize genome is made up of a large amount of highly repetitive DNA including many mobile DNA elements. Unlike Arabidopsis genes, the maize genes are not spaced evenly throughout the genome but instead are clustered in "islands" floating in a large "sea" of repeat-sequence DNA.

To sequence these "islands", the Maize Consortium employed two methods for gene-enrichment, methyl-filtration and high-Cot selection. The methyl-filtration method was developed at Cold Spring Harbor Laboratory in Long Island, New York, and has been exclusively licensed to St. Louis-based Orion Genomics. This method is based on the finding that highly repetitive DNA is modified (methylated) while genes are largely free of such modification. The well-established high-Cot selection method was applied at Purdue University and exploits the fact that gene sequences are in relatively low abundance compared with the large amount of repeated non-genic sequences. These methods target overlapping, but non-identical fractions of the genome that are highly enriched for genes sequences.

The Donald Danforth Plant Science Center is a not-for-profit research institution that was founded in 1998 as the product of a unique and innovative alliance joining the University of Illinois at Urbana-Champaign, the Missouri Botanical Garden, the University of Missouri-Columbia, Monsanto Company, Purdue University, and Washington University in St. Louis. The mission of the Danforth Center is to increase understanding of basic plant biology; to apply new knowledge for the benefit of human nutrition and health and to improve the sustainability of agriculture worldwide; to facilitate the rapid development and commercialization of promising technologies and products; and to contribute to the education and training of graduate and postdoctoral students, scientists, and technicians from around the world.