By Monica Heger
This article was originally published Dec. 4.
Researchers at Massachusetts General Hospital, the Broad Institute, and Helicos BioSciences published a study last month that demonstrated the feasibility of using the Helicos single-molecule sequencing platform for direct sequencing of chromatin immunoprecipitated nucleic acid, or ChIP-seq analysis.
The study, published online last month in Nature Methods, showed that the company's HeliScope Genetic Analysis platform was able to obtain chromatin maps comparable to those generated from ChIP-seq studies on the Illumina Genome Analyzer.
The study follows another proof-of-principle paper that the company published in Nature in September that demonstrated the use of the platform for direct RNA sequencing (see In Sequence 9-29-2009).
According to the authors of last week's Nature Methods paper, single-molecule sequencing could allow scientists to use less starting DNA and perform fewer steps than ChiP-seq on the Illumina platform, while achieving comparable results with less bias.
"This is perhaps the first example of getting reliable ChIP-seq data from single-molecule sequencing," said Peter Park of Harvard's Center for Biomedical Informatics, who was not involved with the study. "I think this is a very promising platform. We have tried both [Helicos and Illumina] platforms and are doing more and more runs on Helicos," he added.
The work, led by Bradley Bernstein's group at Mass General and the Broad Institute, used the Helicos platform and only 50 picograms of starting DNA. By comparison, the Illumina GA requires 5 nanograms of starting DNA for ChIP-seq analysis, according to the paper.
The group did four different ChIP-seq analyses on mouse embryonic stem cells — three for chromatin modifications and one on a binding protein. They then compared the data to the Illumina system.
"We found that it works. With direct sequencing, we can get similar maps that we got from previous techniques," the Broad Institute's Alon Goren, lead author of the paper, told In Sequence.
The study indicates that the Helicos platform offers several improvements over Illumina and other second-generation sequencing technologies for ChIP-seq. First off, there are fewer steps, so the protocol is simpler, leaving less room for error. Also, because there is no amplification step, there is less bias in the end result.
To assess bias, the group compared the read density of the Helicos sequencing protocol to that of Illumina's and also to a control that served as a theoretical expected distribution. Illumina's procedure elicited a strong G-C bias, while the expected distribution was more or less even. The Helicos distribution was closer to the expected distribution, showing much less of a bias, the authors concluded. Goren said this is likely due to the procedure's fewer steps, and also the absence of an amplification step.
Specifically, the authors noted that in the Illumina data, they observed a "modest over-representation" of reads from regions of G-C content of around 40 percent to 65 percent, which they attributed to "bias introduced by PCR or cluster amplification." The Helicos data, meantime "had a relatively even distribution across 20-80 percent G+C content."
In addition, the researchers were able to use much less starting DNA with the Helicos system than with the Illumina — 50 picograms as opposed to 5 nanograms.
Park said that this aspect of the platform should make it particularly useful for stem cell research. With other methods and platforms, researchers have not been able to get enough cells to accurately sequence stem cell DNA. But "by being able to do ChIP-seq from a smaller number of cells, we can profile stem cells that could not otherwise be profiled," he said.
Ali Mortazavi, a Beckman Institute Fellow in genetics at Caltech, agreed that being able to do ChIP-seq from such a small sample is a huge advantage.
"Because there are many interesting biological samples that we cannot get a large number of cells from, there has been a lot of interest in trying to lower the amount of material you use in ChIP-seq," he said. In particular, for stem cells and other rare cells, it might take nearly a year to collect enough cells to do a ChIP-seq analysis, he added. "So, being able to lower the starting matter by an order of magnitude is pretty significant."
However, Mortazavi thinks that there are still a few problems that will inhibit the Helicos platform from being widely adopted. The machine's $999,000 price tag could be prohibitive for researchers who may not use it regularly, and also, the quality of the reads is still not on par with those produced from the Illumina platform, he said.
Goren said that the Helicos platform wasn't quite as sensitive as Illumina's in picking up small signals, so it might not be appropriate in all cases. For example, Illumina and other second-generation platforms may be more suited for providing details on repetitive regions, he said.
The next step, said Goren, is to improve the method so that even smaller DNA samples will yield reliable data. Also, in this study, the team began with around 10 million cells, which yielded about 3 nanograms of DNA. They then diluted that DNA down to 50 picograms. In future experiments, Goren said he wants to start with only around 20,000 to 50,000 cells and run a ChIP-seq analysis from the DNA he is able to extract from just those cells, which would be a more realistic test of the technique on truly rare cells.
While the study was a proof of principle, showing that ChIP-seq could be done on the Helicos platform, Park thought it was indicative of the direction of future sequencing technologies. "This is just a taste of the single molecule technologies to come," he said.