By Julia Karow
This story was originally published December 17
A team of scientists and engineers at the Industrial Technology Research Institute in Hsinchu, Taiwan, is working on a microchip that integrates chemical reaction wells with an optical sensing system. The chip, they predict, will eventually allow them to sequence millions of single DNA strands in parallel in real time.
The ITRI-funded group, about 30 people strong and called "Cracker," is currently looking for investors and strategic partners in order to spin out into a company and speed up the development of its technology. Last month, the team joined the Archon X Prize for Genomics as a competitor (see In Sequence 11/17/2009).
Cracker is headed by Chung-Fan Chiou, former chief technology officer and president of Phalanx Biotech Group, a microarray company and ITRI spin-off. At the request of ITRI president Johnsee Lee, Chiou several years ago started to look into new technical approaches to DNA sequencing.
After almost giving up because many competing groups had already patented potential sequencing methods, Chiou said, he and several of his colleagues about two years ago came up with the concept of a chip that places DNA sequencing reactions in close proximity to an integrated optical detection system. While current microscope- and camera-based sequencing systems are akin to "watching stars using a telescope," he explained, using Cracker's method, "you just bring the stars close to the eye."
At the heart of the patent-pending technology, which is still in early development, is a three-tiered chip — called the sTOP chip for "sequencing on top of a photodiode" — consisting of a layer of nanowells, a filtering layer, and a photo-sensing layer.
The nanowells — each 100 nanometers or less in diameter — are surrounded at the bottom by a light source of high intensity. Light emanates from the sides of each nanowell, illuminating a small volume near the bottom inside, thus exciting single fluorescently labeled nucleotides that are incorporated into a strand of DNA by a polymerase.
These emit light that travels through the 2-micrometer-thick filter tier, which blocks out the excitation light, and is converted into an electric signal by a photodiode in the tier underneath. The properties of this signal are then "decoded" by an integrated circuit adjacent to each diode, "hence the nucleotide incorporated is identified," according to a description of the technology on the group's website.
Labeled nucleotides in solution may produce background signals, but the sTOP chip limits this by illuminating only a small volume of the well. Also, signals from free nucleotides will be much shorter than from those that are being incorporated into DNA.
The chip will be about a square inch in size and carry up to millions of nanowells. It will be portable, have a low manufacturing cost, and a long shelf-life. "No sophisticated detection apparatus like microscopes are needed, as the light signal is immediately processed to a digital output at the chip level itself," according to the description. The chip "breaks away the boundaries of classical-optics resolution, and rather shifts the throughput limits to the enzyme's intrinsic properties."
Though the group is initially focusing on DNA sequencing as its main future application, the chip could also be developed for other applications that involve single-molecule real-time fluorescence.
Next Step: Chip Assembly
So far, the Cracker team has performed theoretical analyses and simulations, and has worked on the chip components, "and it seems that everything works how we predict," according to Chiou. "The next phase we have to move to is to integrate the different parts" and assemble the chip, he said. The group is working with the Taiwan Semiconductor Manufacturing Corporation, which his providing it with some prototype chips.
[ pagebreak ]
Work on the chip itself is expected to take about another two years — though Chiou said this timeline could be shortened given additional resources — after which the researchers could begin to use it to sequence DNA. Once the chip is ready, it will be easy to scale it up in size, and "once we can put the chemistry on, the path will be very fast," Chiou predicted.
In parallel to the chip design, the group has been working on the sequencing chemistry. Due to pending patents, it is currently not revealing any details about the chemistry beyond the information provided on its website, which describes a single-molecule sequencing-by-synthesis method that uses DNA polymerase and nucleotides labeled with different fluorophores. Through the X Prize, the team has also made contacts with experts in enzymology and biochemistry who might become partners in the future, Chiou said.
The goal is to sequence circular DNA molecules 2 kilobases in length, according to Hubert Renauld, a senior scientist at Cracker and the group's secretary.
For a sequencing run, the chip will be plugged into a notebook-sized station, following sample preparation in a separate station, he said. Because of the low expected manufacturing costs, the chip could be disposable, according to Renauld, though "in view of environmental concerns," the group intends to make it reusable.
In principle, Cracker's approach bears some similarities to Pacific Biosciences' sequencing technology — both monitor single fluorescently labeled nucleotides incorporated by a DNA polymerase in real time, and both confine the observation volume, so no more than one label is monitored at a time.
The main difference between the two appears to be the optical detection system, which in Cracker's case requires no lenses or separate camera. Its chip design, according to Chiou, will allow the group to achieve a greater throughput than camera-based sequencing systems.
But there will likely be challenges. According to John Nelson, principal scientist in the molecular and cellular biology laboratory at GE Global Research in Niskayuna, NY, "producing enough individually addressable wells in one chip to enable a throughput required for acquiring an entire genome's worth of sequence will be technically challenging."
He also said that it might be difficult to cool the optical system if it is in close proximity to the reaction wells. The nanowell layer that contains the enzyme and reagents "will need to be warm in order to function, and imaging systems typically use a cooled camera to reduce electronic noise," he said. "But in their case, [the optical system] is right next to the enzyme, so the optics can't be cooled."
Nonetheless, he said, Cracker's device might reduce the cost of a sequencing system significantly. "If they are indeed able to eliminate the expensive imaging system by incorporating that functionality into the chip itself, that would reduce the upfront capital expenditure that is typically associated with many sequencing platforms."
The Taiwanese team is not the only one developing an integrated chip-based sequencing system, though. Ion Torrent Systems, for example, a startup of 454 co-founder Jonathan Rothberg, is also working on a sequencer that will combine disposable chips, a chip reader, and fluidics, and is predicted to be able to sequence millions of DNA strands; it will not, however, rely on optical imaging.
According to the abstract of a recent grant that Ion Torrent won under the National Human Genome Research Institute's $1,000 Genome program, the company has already designed and developed a semiconductor sensor, called an "Ion Torrent Chip," to "directly detect polymerization of DNA without the need for any intermediate enzymatic reactions, chemiluminescence, fluorescence, optics, optical imaging, or other constraints of having to detect light or use unnatural reagents."