Joseph A. Bedell

Sorghum genome sequencing by methylation filtration.

Sorghum bicolor is a close relative of maize and is a staple crop in Africa and much of the developing world because of its superior tolerance of arid growth conditions. We have generated sequence from the hypomethylated portion of the sorghum genome by applying methylation filtration (MF) technology. The evidence suggests that 96% of the genes have been sequence tagged, with an average coverage of 65% across their length. Remarkably, this level of gene discovery was accomplished after generating a raw coverage of less than 300 megabases of the 735-megabase genome. MF preferentially captures exons and introns, promoters, microRNAs, and simple sequence repeats, and minimizes interspersed repeats, thus providing a robust view of the functional parts of the genome. The sorghum MF sequence set is beneficial to research on sorghum and is also a powerful resource for comparative genomics among the grasses and across the entire plant kingdom. Thousands of hypothetical gene predictions in rice and Arabidopsis are supported by the sorghum dataset, and genomic similarities highlight evolutionarily conserved regions that will lead to a better understanding of rice and Arabidopsis.

MaskerAid : a performance enhancement to RepeatMasker

UNLABELLED Identifying and masking repetitive elements is usually the first step when analyzing vertebrate genomic sequence. Current repeat identification software is sensitive but slow, creating a costly bottleneck in large-scale analyses. We have developed MaskerAid, a software enhancement to RepeatMasker that increased the speed of masking more than 30-fold at the most sensitive setting. AVAILABILITY On request from the authors (see http://sapiens.wustl.edu/MaskerAid). CONTACT maskeraid@watson.wustl.edu

Identification of Novel High-Frequency DNA Methylation Changes in Breast Cancer

Recent data have revealed that epigenetic alterations, including DNA methylation and chromatin structure changes, are among the earliest molecular abnormalities to occur during tumorigenesis. The inherent thermodynamic stability of cytosine methylation and the apparent high specificity of the alterations for disease may accelerate the development of powerful molecular diagnostics for cancer. We report a genome-wide analysis of DNA methylation alterations in breast cancer. The approach efficiently identified a large collection of novel differentially DNA methylated loci (∼200), a subset of which was independently validated across a panel of over 230 clinical samples. The differential cytosine methylation events were independent of patient age, tumor stage, estrogen receptor status or family history of breast cancer. The power of the global approach for discovery is underscored by the identification of a single differentially methylated locus, associated with the GHSR gene, capable of distinguishing infiltrating ductal breast carcinoma from normal and benign breast tissues with a sensitivity and specificity of 90% and 96%, respectively. Notably, the frequency of these molecular abnormalities in breast tumors substantially exceeds the frequency of any other single genetic or epigenetic change reported to date. The discovery of over 50 novel DNA methylation-based biomarkers of breast cancer may provide new routes for development of DNA methylation-based diagnostics and prognostics, as well as reveal epigenetically regulated mechanism involved in breast tumorigenesis.

MethylMapper: a method for high-throughput, multilocus bisulfite sequence analysis and reporting

Understanding the phenotypic contribution of epigenetic components is making DNA methylation pattern analysis more important in higher eukaryotic genomes as well as human disease. Bisulfite sequencing protocols report DNA methylation occupancy information as a positive assay output that allows methylation patterns to be elucidated from particular devel-opmental or disease states. Reported here is a new method for bisulfite sequencing project management, data analysis, and site-specific methylation test development that is designed for integration in high-throughput genomic and bioinformatics analyses.The phenotypic consequences of epigenetic alterations are being appreciated in nearly every biological system. In higher eukaryotes, DNA methylation patterns serve as a second genomic information code, an easily monitored proximate marker reflecting epigenetic cellular decisions. Most notably, in human cancer, abnormal cellular DNA methylation patterns can directly contribute to the mechanisms of tumorigenesis, most often through the induction of erroneous gene silencing (1). Recent work by many groups has demonstrated that DNA methylation abnormalities may be exploited for the development of powerful cancer diagnostic and prognostic tools (2).One tool in the arsenal of methylation monitoring is bisulfite sequencing (3). Bisulfite treatment of DNA causes the deamination of cytosine and conversion to thymine upon amplification and cloning. However, if the cytosine is methylated, it remains unchanged. In this way, subsequent sequencing of the amplified, mutagenized clones with conventional technology allows one to understand the methylation occupancy at each cytosine.Several groups have released publicly available tools for bisulfite sequencing experimental design or data analysis (4–7). Furthermore, online tools such as methBLAST (medgen.ugent.be/methblast) have become available for in silico bisulfite modification to aid in PCR primer design. However, the utility is limited in that they either only help with experimental design or with data analysis (i.e., pattern elucidation). For those that offer methylation pattern elucidation, they also require the experimenter to employ a locus-by-locus, one-gene-at-a-time approach. Given the revolution in automated sequencing and the capacity available at most genome centers, a combined design and analysis suite that affords investigators the opportunity to take advantage of a high-throughput, low-effort analysis procedure is missing. Most researchers frustratingly address this unmet need by performing each outlined step in the most labor-intensive but readily apparent manner. Typically, this involves a stepwise approach for each target, often aligning trace files by hand and manually performing the occupancy calculations. Moreover, as multiple-locus biomarkers are discovered and employed, validation of methylation patterns in a multi-locus manner from many samples will be necessary. Having an automated analysis capacity will become an even more pressing need.We have created an efficient package of PERL programs called MethylMapper that, when combined with a primer-picking program and BLASTN, simplifies the design and analysis of bisulfite experiments in a high-throughput environment. MethylMapper allows the data to self-organize to minimize mistakes and expedite simultaneous analyses of multiple loci. Furthermore, it makes data quality control as streamlined as possible. The package requires only PERL and NCBI-BLAST (www.ncbi.nlm.nih.gov/blast/download.shtml) and requires very little memory or CPU time. System requirements and CPU usage are dependent only upon the requirements of the BLASTN operation. The package is freely available for download at methylmapper.sourceforge.net.For demonstration purposes, we used MethylMapper to design primers and analyze bisulfite sequencing results for a genomic region spanning the second exon of the SLC4A3 gene on human chromosome 2 (Figure 1). Primer sequences for the analysis shown in Figure 1 are 5′-TGATTTGGGTAAGATTTTGGTTGTGAGTAG-3′ (forward) and 5′-CATCCCTAATAAACAAAACATAAAACT-3′ (reverse). Bisulfite conversion was performed with the EZ DNA Methylation Kit™ (Zymo Research, Orange, CA, USA) under the manufacturer’s recommended conditions. PCR amplification was performed under standard conditions that employed 0.2 pmol of primers, 40 ng of template in a 25-μL volume of water, to which 25 μL of FailSafe™ G 2× premix and 1 U of FailSafe Taq DNA Polymerase (EPICENTRE, Madison, WI, USA) were added. The cycling conditions employed a single 3-min incubation at 95°C, followed by 30 cycles of 95°C for 45 s, 52°C for 15 s, and 72°C for 30 s. Finally, a 10-min chase step was performed by incubating the sample at 72°C. The PCR product was purified from an agarose gel slice and cloned using the pCR2.1-TOPO TA Cloning

Survey sequencing of soybean elucidates the genome structure, composition and identifies novel repeats

In order to expand our knowledge of the soybean genome and to create a useful DNA repeat sequence database, over 24 000 DNA fragments from a soybean [Glycine max (L.) Merr.] cv. Williams 82 genomic shotgun library were sequenced. Additional sequences came from over 29 000 bacterial artificial chromosome (BAC) end sequences derived from a BstI library of the cv. Williams 82 genome. Analysis of these sequences identified 348 different DNA repeats, many of which appear to be novel. To extend the utility of the work, a pilot study was also conducted using methylation filtration to estimate the hypomethylated, soybean gene space. A comparison between 8366 sequences obtained from a filtered library and 23 788 from an unfiltered library indicate a gene-enrichment of ~3.2-fold in the hypomethylated sequences. Given the 1.1-Gb soybean genome, our analysis predicts a ~343-Mb hypomethylated, gene-rich space.