Researchers at Columbia University have developed a new stepwise sequencing-by-synthesis method that combines features of Sanger sequencing and existing next-generation SBS chemistries.
In an article published online in the Proceedings of the National Academy of Sciences this week, the scientists, led by Jingyue Ju and Nicholas Turro, used the hybrid approach to determine 32 bases of DNA on a chip, reading it with a four-color fluorescence scanner. With further improvements, especially in the sensitivity of fluorescent detection, they say the read length could potentially go up to more than 700 base pairs.
Like traditional Sanger sequencing, the new chemistry uses four types of fluorescently labeled dideoxy nucleotides to terminate a DNA strand and to read its last base. But unlike Sanger sequencing, which separates terminated DNA fragments of all sizes generated in a single reaction by electrophoresis — a process that is difficult to do in high-throughput — the new approach is stepwise and reads each base after it is added on a chip, potentially in a massively parallel manner.
“The idea is to take advantage of the Sanger dideoxy termination reaction, because it’s very robust and there are no enzymatic issues. That’s why Sanger sequencing has been successful for over 30 years,” said Ju, who is a professor of chemical engineering at Columbia University and head of DNA sequencing and chemical biology at Columbia’s Genome Center.
Polymerases that efficiently incorporate the dideoxy terminators are well known from Sanger sequencing and “can be readily used” for the new approach, he said, and modified polymerases can recognize the unlabeled reversible terminators “relatively easily.”
In each cycle, the researchers add a mixture of reversible terminators that are capped at the 3’-OH group with a cleavable azidomethyl group, and four fluorescent dideoxy nucleotides that carry fluorescent groups at the 5-position of the pyrimidines and the 7-position of the purines via a cleavable linker.
A small fraction of the DNA strands incorporates the fluorescent terminators, while the majority adds the reversible terminators. Following a synchronization step with an excess of reversible terminators to extend all DNA strands to the same length, the labeled strands are read by a four-color fluorescence scanner to determine which bases were added. A cleavage reaction then removes both the fluorescent labels and the reversible terminator blocking groups, and the next sequencing cycle starts.
The fraction of dideoxy terminators added, which increases from cycle to cycle, determines the strength of the fluorescent signal, and how many DNA strands are permanently terminated after each cycle. As a result, the total number of DNA strands that grow and can be read decreases with every cycle.
“In the first few cycles, because there are a lot of templates available, we use a very small percentage of dideoxy terminators, but later on, the percentage increases,” Ju told In Sequence this week.
Similar to other stepwise sequencing-by-synthesis chemistries, the Columbia approach allows researchers to read large numbers of DNA fragments in parallel on a high-density chip.
“The idea is to take advantage of the Sanger dideoxy termination reaction, because it’s very robust and there are no enzymatic issues.”
But the other technologies that use nucleotides with a cleavable label, such as Illumina’s and Helicos BioSciences’, leave a modification on the newly-incorporated base after the label is cleaved off, which might interfere with the ability of the polymerase to add nucleotides in subsequent cycles.
Using the Columbia researchers’ chemistry, DNA strands that carry a base modification from a cleaved label are no longer used in subsequent sequencing cycles, while DNA strands that have added a reversible terminator and continue to grow are “natural” after the capping group is removed.
“Therefore, there will be no adverse effect on the DNA polymerase for the incorporation of the next complementary nucleotide,” the researchers write in the PNAS paper.
They also point out that other SBS chemistries that use reversible terminators with both a capping group and a fluorescent label, like Illumina’s, require “the further improvement of the DNA polymerase that efficiently recognizes the modified nucleotides.”
But neither the “leftover” group on the base nor “inefficient enzymology” has been shown to be an issue with other sequencing chemistries, including Illumina’s and Helicos’, according to Harold Swerdlow, head of sequencing technology at the Wellcome Trust Sanger Institute. Until 2006, he was senior director of research at Solexa, which was acquired by Illumina last year.
He also remarked that the Columbia method is “more complicated” than both Illumina’s and Helicos’ SBS chemistries. “A key complication of the chemistry described here is the need to adjust the ratio of the fluorescent and 3’-blocked nucleotides during the course of the reaction, and to perform an extra ‘synchronization’ step after each cycle,” Swerdlow wrote in an e-mail message.
He said that while the approach of mixing reversible terminators and fluorescent dideoxy terminators is unique and novel, “the nucleotides themselves, the enzyme used to incorporate them, and the chemistry of the cleavage of both the linkers and the 3’ block are not novel” and are described in detail in a Solexa patent and patent application that are cited in the article.
Ju said that his group is currently working on extending the read length, mainly through “tweaking the chemistry.”
The next step is to push the read length to 50 bases, he said. This will be possible with existing CCD cameras that are sensitive enough to detect 1,000 fluorescent molecules. With further improvements of the process, “our goal is to reach, eventually, 100 base pairs,” he said.
The ultimate possible read length, the researchers write, depends on the number of DNA starting molecules in each spot, the reaction efficiency, and the detection sensitivity of the sequencing system. They claim that their approach “may have the potential to reach” read lengths of more than 700 base pairs, which are routinely reached in Sanger dideoxy sequencing, “especially with improvements in the sensitivity of the fluorescent detection system, where single molecules can be reliably detected.”
But according to Swerdlow, the new method’s read length will suffer from the same limitations as other stepwise sequencing-by-synthesis chemistries. Unlike in Sanger sequencing, where all terminated fragments are generated in a one-step biochemical reaction, stepwise SBS methods lose DNA molecules in each cycle due to incomplete nucleotide incorporation and incomplete deblocking; small quantities of contaminating species; DNA loss from the chip surface; DNA damage during chemical and imaging steps; and accumulation of background noise over several cycles, he said.
In terms of biological applications, the technology could “probably be used for similar purposes” as commercially available second-generation sequencing systems from Illumina and Applied Biosystems, according to Ju. He stressed that his publication was focused on the new sequencing chemistry but not yet an entire sequencing system.
In the PNAS study, the researchers used synthetic DNA that did not require amplification. To sequence biological DNA samples, a sample preparation method would be required. According to Ju, the chemistry is compatible with known clonal amplification methods, such as emulsion PCR.
Ju said that Columbia University has filed for patent protection of the method but had no information about any licensing agreements.
In 2006, Intelligent Bio-Systems, a startup company based in Waltham, Mass., exclusively licensed a different sequencing-by-synthesis chemistry from Columbia University that was developed by Ju’s lab and published in PNAS that year (see In Sequence 1/2/2007). That four-color SBS chemistry uses cleavable nucleotide reversible terminators that also carry fluorescent dyes, similar in principle to Illumina’s fluorescent reversible terminators. Last fall, IBS said it was planning to launch a sequencing platform this year (see In Sequence 10/23/2007).
IBS CEO Steven Gordon declined to say whether his company has also licensed the new chemistry from Ju’s lab. “We are considering the chemistry we are using in the instrument confidential,” he told In Sequence this week.