DNA Sequencing: Today's Technology And Beyond

SUPPLIERS OF DNA SEQUENCERS, SYSTEMS, AND SOFTWARE American Bioanalytical Inc. Natick, MA Circle No. 101 on Reader Service Card Amersham Corp. Arlington Heights, IL Circle No. 102 on Reader Service Card Applied Biosystems Division of Perkin-Elmer Foster City, CA Circle No. 103 on Reader Service Card Beckman Instruments Inc. Fullerton, CA Circle No. 104 on Reader Service Card Biomed Instruments Inc. Fullerton, CA Circle No. 105 on Reader Service Card Biomedical Research & Development Labs Inc. Gaithersburg, MD Circle No. 106 on Reader Service Card Bio-Rad Laboratories Life Science Group Hercules, CA Circle No. 107 on Reader Service Card Boehringer Mannheim Corp. Indianapolis, IN Circle No. 108 on Reader Service Card C.B.S. Scientific Co. Del Mar, CA Circle No. 109 on Reader Service Card DNA ProScan Inc. Nashville, TN Circle No. 110 on Reader Service Card DNASTAR Inc. Madison, WI Circle No. 111 on Reader Service Card EG&G Berthold Nashua, NH Circle No. 112 on Reader Service Card Fisher Scientific Pittsburgh, PA Circle No. 113 on Reader Service Card Hoefer Scientific Instruments San Francisco, CA Circle No. 114 on Reader Service Card Inotech Biosystems Lansing, MI Circle No. 115 on Reader Service Card Integra BioSciences Inc. Woburn, MA Circle No. 116 on Reader Service Card IntellliGenetics Inc. Mountain View, CA Circle No. 117 on Reader Service Card Intermountain Scientific Corp. Kaysville, UT Circle No. 118 on Reader Service Card Jordan Scientific Co. Inc. Bloomington, IN Circle No. 119 on Reader Service Card Laboratory & Research Products New Haven, CT Circle No. 120 on Reader Service Card Li-Cor Inc. Lincoln, NE Circle No. 121 on Reader Service Card Life Technologies Inc. Gaithersburg, MD Circle No. 122 on Reader Service Card Marsh Biomedical Products Inc. Rochester, NY Circle No. 123 on Reader Service Card Midwest Scientific St. Louis, MO Circle No. 124 on Reader Service Card Molecular Biology Resources Inc. Milwaukee, WI Circle No. 125 on Reader Service Card Molecular Dynamics Sunnyvale, CA Circle No. 126 on Reader Service Card National Biosciences Inc. Plymouth, MN Circle No. 127 on Reader Service Card Owl Scientific Woburn, MA Circle No. 128 on Reader Service Card Perkin-Elmer Norwalk, CT Circle No. 129 on Reader Service Card Pharmacia Biotech Inc. Piscataway, NJ Circle No. 130 on Reader Service Card Promega Corp. Madison, WI Circle No. 131 on Reader Service Card Scanalytics Billerica, MA Circle No. 132 on Reader Service Card Stratagene La Jolla, CA Circle No. 133 on Reader Service Card Textco Inc. West Lebanon, NH Circle No. 134 on Reader Service Card Time Logic Inc. Moorpark, CA Circle No. 135 on Reader Service Card UVP Inc. Upland, CA Circle No. 136 on Reader Service Card Zymark Corp. Hopkinton, MA Circle No. 137 on Reader Service Card Since the basic chemistry for DNA sequencing was developed in the 1970s by Frederick Sanger of the Medical Research Council, Cambridge, England, this technology has become a ubiquitous tool in genetics, molecular-biology, and microbiology labs. But the fastest-growing application of DNA sequencing today is in clinical diagnostics. Also, rapid advances in DNA sequencing have pushed the technology beyond its conventional base in polyacrylamide slab gel electrophoresis. These new approaches involve capillary gel electrophoresis and even glass-plate DNA "chips" made with photolithography processes. The current market in automated DNA sequencers is a well- ordered one, dominated by a handful of manufacturers. The leading companies making these products all currently use Sanger's basic chemistry for producing nested sets of progressively longer DNA fragments. These sets of DNA fragments are tagged with fluorescent dyes and then separated by gel electrophoresis and automatically scanned for identification of base-pair sequences. "A DNA sequencer is nothing but an automated DNA electrophoresis detection and analysis system," explains Jack Ball, vice president of molecular biology systems at Pharmacia Biotech Inc. in Piscataway, N.J., which markets the ALF (automated laser fluorescence) DNA sequencer. From Ball's description, this may sound like a simple enough instrument, but sequencers are not cheap. Prices range from $55,000 for an aggressively priced system introduced this summer by Molecular Dynamics Corp. of Sunnyvale, Calif., to $110,000 for the market leader, made by Applied Biosystems Inc., a division of Perkin-Elmer Corp. in Foster City, Calif. Researchers interested in purchasing a DNA-sequencing system have a basic choice. They may select a device from Applied Biosystems, whose sequencer offers the highest throughput-- that is, the most number of bases that can be read in a day. Alternatively, they may choose one of several other manufacturers' systems, which cannot offer the same throughput but are lower priced and offer higher sensitivity as well as, in at least one case, longer read length-- referring to the base-pair length of DNA fragments. In fact, one system, available from Lincoln, Neb.-based LI-COR Inc., can routinely read DNA fragments up to 800 and even 1,000 bases in length, compared with the 500 bases typically read by its competitors. But in the final analysis, all of the commercially available automated sequencers, and the supporting reaction chemistry, are quite similar. The current market in sequencers, dominated by a few major manufacturers, may begin to fragment, however, as new technologies--and the companies developing them--start breaking onto the sequencing scene. The Human Genome Project, with its ambitious goal of mapping the entire human genome in a matter of years, has increased demand for far faster, less cumbersome sequencing methods than are available at present. This has catalyzed the development of a number of new and unconventional approaches that could allow a single scientist to sequence up to 10 million bases per year. At this rate, the entire human genome, with its 3 billion bases, could be sequenced in a single year by just 300 workers. "Everything that spins off the Human Genome Project will [also] require confirmatory sequencing," says Ball. That is, sequences derived rapidly will have to be confirmed, and thus "the market for DNA sequencers will be absolutely huge," he predicts. New applications for DNA-sequencing technology, most notably in the clinical diagnostic arena, will also expand the market for sequencers, bringing new technologies and lower- priced instruments to commercialization, say many industry observers. Alternative approaches--such as capillary gel electrophoresis, hybridization, and time-of-flight mass spectrometry--may bear fruit before the end of the decade. But until then, the current technique of using slabs of polyacrylamide gel to separate DNA fragments will remain the method of choice. And, at least for now, the Applied Biosystems technology is predominant. Applied Biosystems claims to have about 85 percent of the market in automated DNA sequencers in the United States, leading Steve Lombardi, product manager for the AB 373 DNA sequencer, to boast, "We own the market." One of the main reasons for the company's market dominance is that it was the first to come out with a sequencer, introducing the AB 373 in June 1986. Another reason is throughput. With its unique four-dye technology, Applied Biosystems sequencers read all four nucleotides in a single lane. Other systems require four parallel sequencing reactions for a single DNA strand. This multiplexing approach quadruples Applied Biosystems' output over competing manufacturers, according to the company. Applied Biosystems, as well as Molecular Dynamics, also sells automated robotic systems that perform the pre- separation reaction chemistry developed by Sanger, which is still the basis for DNA-sequencing technology. "Our goal is to automate the molecular biology lab and all the different applications in that lab," says Lombardi. "We don't just sell sequencers anymore. We sell genetic analyzers." In addition to DNA sequencing, these systems can be used to perform DNA fingerprinting, mutation screening, linkage mapping, and microsatellite analysis. The chemistry used to prepare samples for the currently available automated sequencers is based on the tried-and- true method developed in the 1970s (F. Sanger, Proceedings of the National Academy of Sciences, 74:5463-7, 1977). In this reaction, DNA is fragmented with restriction enzymes into single-stranded lengths that readily pass through gels, with the preferred length about 500 bases. The single- stranded fragment is then mixed with DNA polymerase, and with free nucleotides. The DNA polymerase runs along the strand, "repairing" it by building a complementary strand. If left alone, the reaction would end when the entire double strand was completed. However, this would provide no information on the sequences within the strand. The trick used in the Sanger method is to force DNA polymerase to occasionally terminate the reaction at base number 1, then base number 2, number 3, and so on, resulting in sequential DNA strands up to 500 base pairs in length. To do this, a small amount of aberrant nucleotides, called dideoxynucleotides, is mixed in with the normal deoxynucleotides. These aberrant nucleotides lack the usual free 3*-OH, which is needed to continue the polymerization reaction. Whenever the DNA polymerase picks up an aberrant nucleotide and incorporates it into the complementary strand, the reaction halts. In the end, the test tube is filled with a series of fragments ranging from one to 500 bases in length. This is the point at which the manufacturers' technologies diverge. AB utilizes four dyes, each of which is specific to a particular dideoxynucleotide. One dye would react only with dideoxyguanosine, another would react only with dideoxyadenosine, and so on, so that each base fluoresces at a characteristic wavelength, allowing all four bases to be read in the same lane of the gel. In the next step, the fragments are loaded manually onto a DNA sequencer gel, which separates the fragments into discrete bands that differ in length by one base. These fragments will then migrate down the gel at a rate proportionate to their length, with the shortest passing through the most quickly. The length is thus determined by its time of migration, and the base is determined by which of the four wavelengths it fluoresces at. The other manufacturers use a single-dye approach. At the chain-termination step, the DNA fragment is put into four separate reaction tubes, each one having only one of the four possible aberrant nucleotides. The four separate reactions are then loaded into four separate lanes on the gel, with each lane corresponding to a different nucleotide. The lower throughput obtained on these systems is simply a consequence of needing four separate lanes, instead of just one, to run one DNA strand. The single-dye approach has advantages, however, explains James Nelson, product manager for the Vistra line of sequencers from Molecular Dynamics: "Because it's a single- dye system, it's selected for much less expensive optics and laser systems." The Applied Biosystems technology uses a fairly complex argon laser, which is designed to produce excitation wavelengths matched to the four different dyes. As such, its output is not optimized for each individual dye's absorption maxima, says Nelson. The single-dye system used by Molecular Dynamics allows use of an inexpensive helium-neon laser optimized for the company's dye. The LI-COR system uses a simple infrared (IR) laser, the same type of laser used in compact disc players. According to Abraham Oommen, an applications scientist at LI-COR, the IR laser system is far more sensitive than other detectors, allowing less sample to be loaded. Plus, he notes, "most biological molecules don't fluoresce in the infrared, so background [interference] is negligible." Less interference translates to longer read lengths, with LI-COR claiming that researchers can routinely read out to 800 bases with "base calling," or the identification of the bases, accomplished automatically by the computer software. Longer read lengths mean that fewer subclones have to be generated, which can be a great cost and time advantage for university investigators. But LI-COR does not claim to be competitive with AB when it comes to throughput. "Massive sequencing facilities would not benefit from our machine," Oommen says. "We get out to 1,000 bases, but not with thousands of samples. So, if you want to process 100 samples a day, go with an Applied Biosystems machine." As is true of any lab equipment, price is always a powerful consideration for researchers considering the purchase of a sequencer. The $110,000 price tag for an Applied Biosystems machine is something only the most dedicated user or core facility can afford. For those on a tighter budget, Pharmacia offers a sequencer that lists at $90,000 but can usually be discounted into the $75,000 to $80,000 range, says Ball. "And that's a significant [reduction] in prices." Besides being a laser-excited, single-fluorescent dye system with computer-based, automated base calling, the ALF sequencer has a thermally jacketed gel plate, allowing precise temperature control of the separation process. Also, sequencing and data analysis are controlled by a 486 computer running under an OS/2 operating system. The OS/2 system was selected because it is truly multi-tasking, meaning that data from one run can be manipulated, stored, or printed with a completely different application while the computer is running the separation and data analysis for the current run. The LI-COR system, which also takes advantage of the multitasking capability of OS/2, is significantly less expensive than Applied Biosystems' sequencer, however. Priced at about $79,000, it has the advantage of longer read lengths and higher sensitivity over other systems. The Applied Biosystems and Molecular Dynamics sequencers run on a Macintosh-based system. And, as Molecular Dynamics' Nelson points out, "If you look out into the marketplace, all the software [for genomic analysis] has been written for the Mac. That's not because it's superior. It's just history." Apple aggressively marketed its Macintosh systems to colleges and universities, and much of the DNA software written by professors and students is Mac-based, he says. In light of this, Molecular Dynamics' sequencer allows the user to easily interface with the extensive repertoire of existing Mac-based programs by running on an Apple Quadra 610. But the real advantage of Molecular Dynamics' Vistra system is its comparatively low price, listing at just $54,900, half the cost of an AB 373. A robotic system, also available from Molecular Dynamics, that can be used for automating sequencing and PCR reactions may be added for $59,000. Thus, for the price of an AB 373, a lab can get both an automated sequencer and reaction system. Will this competitive price put Applied Biosystems at a market disadvantage? Lombardi contends that Applied Biosystems has no intention of dropping its prices to compete with the newcomer. "As the market leader, you don't really like to respond," he explains. "We just say that for $110,000, you're going to get a well-respected instrument that's going to do the work that needs to be done." Instead of competing head-to-head with Applied Biosystems, Molecular Dynamics is focusing on a different market. "The AB system has been targeted at core labs and genome sequencers with high throughput requirements," says Nelson. "We're focusing on a different market niche, where throughput is not as much an issue." Throughput is relatively immaterial to Lonnie Ingram, a professor in the microbiology and cell science department at University of Florida, Gainesville, who has a LI-COR sequencer in his lab. He confirms that the LI-COR will give a longer read length than an Applied Biosystems sequencer, and for his lab that's an advantage, since there are fewer subclones and less wasteful overlapping of sequences. His lab was doing all its sequencing manually before getting an automated system. One big advantage of going from a manual to an automated system was getting the radioactive isotopes used in manual processes out of the lab--and the radiation safety department off his back. A number of new sequencing approaches are under active development, with some of them possibly coming onto the market within the next few years. Chief among these technologies is multiple-capillary gel electrophoresis. Molecular Dynamics and Beckman Instruments Inc. of Fullerton, Calif., are both reportedly working on such systems, though neither company's representatives would comment on exactly what stage of development it has reached nor when the units might become available. Capillary gel electrophoresis differs from the conventional method in that the polyacrylamide, instead of being poured out as a slab, is pumped into a 100-micron (internal diameter) capillary. This narrow-bore capillary permits the use of a much higher field strength, allowing separations to be completed in about an hour, instead of the usual six hours or more. With currently available systems, only one capillary can be run at a time, so throughput is not very high. Under development, though, is a capillary array in which 24, 48, or even 96 capillaries can be bundled and run in parallel, allowing more sample lanes than a slab gel and a six-fold increase in throughput over the current Applied Biosystems technology. One of the drawbacks with such a system is that the price of the capillaries does not lend them to disposability. And once the polyacrylamide gel is cross-linked it cannot be scraped out as is done to recycle conventional gel plates. But Barry Karger, a chemist at the Barnett Institute of Northeastern University in Boston, says he is close to a solution: "What we're pushing pretty hard is replaceable matrices for capillaries." In this approach, the linear polyacrylamide is not cross-linked and can thus be blown in and then blown out of capillaries at the end of each run. In a commercial system, the capillary preparation steps could even be automated, he says. Better yet, unlike gel slabs, which require tedious manual loading of samples, the sample loading in a multiple-capillary system can be automated by simply dipping the capillaries into sample wells and letting the electric field draw in the DNA. When will a multiple-capillary system be available? Karger says that Molecular Dynamics has already debuted a 48- capillary system at a recent trade show. While the company didn't actually demonstrate the separation of DNA samples with this system, it showed that it has solved some of the design issues, such as how to scan the array of capillaries with the laser. Karger thinks it could be as soon as "a year or so" before such a system hits the market. Also just over the horizon is the hybridization approach, such as the one being developed at Affymetrix Inc. in Santa Clara, Calif. According to Robert Lipshutz, a mathematician and director of bioinformatics at Affymetrix, the company is developing a "DNA chip," which takes advantage of the propensity of single-stranded DNA to complex with complementary strands of DNA. A set of probes, consisting of oligomers from eight to 20 nucleotides in length, can be generated that will adhere to the target DNA. In the Affymetrix approach, these oligomers are attached to glass plates using the process of photolithography, widely used in making computer chips and nano-machines. In this application, a 1.28 x 1.28 cm region of a glass surface is chemically prepared so that linker molecules, terminated with a photolabile compound, are attached. Next, using a photolithographic mask, a stripe of light is beamed across a quarter of the plate's width, thus removing the photolabile compounds in that area. Then, adenosine, for instance, can be covalently bound to the glass through the linker molecule. The free end of the nucleotide is capped with a photolabile group to prevent any further reaction until it is next exposed to light. The process is repeated until there are four stripes: one with adenosine (A), one with guanosine (G), one with cytosine (C), and one with thymine (T). Afterward, the mask is turned 90 degrees to form a cross hatch of stripes with A, G, C, and T attached to the first set of nucleotides so that all 16 possible nucleotide dimers are formed. Using successively finer stripes of light, the oligomers can be built up to a chain with a length up to 20 nucleotides, which are attached to the glass plate at known positions, says Lipshutz. "We can make a high-density chip with over 65,000 probes in 30 to 60 steps, taking about 10 minutes per step." And the company can make 16 identical DNA chips at the same time. Fluorescently labeled DNA strands can then be washed over the DNA chip, whereupon the DNA will hybridize with the complementary sequences in the probes. The fluorescence is determined by reading the chip with a confocal microscope linked to an automated image scanner. For sequencing, this approach could be extremely rapid, but there is still the "ambiguity problem" to overcome, explains Lipshutz. That is, DNA occasionally will form mismatches when complexing with the oligomers. Affymetrix is currently working out how to detect and deal with these mismatches. "It's still a ways off until people can do de novo sequencing," he says. But high-speed genome sequencing is not the company's first goal. Affymetrix would like to see the DNA chips applied to the rapid determination of mutations in HIV, for example. When the virus is challenged with a drug like AZT, which targets the reverse transcriptase (RT) enzyme, the virus develops resistance via mutations in the RT gene. What the DNA chip could provide is information on when that mutation has taken place so that an alternative drug regimen can begin. Lipshutz says the company hopes to commercialize the DNA chip technology by 1996. The fastest-growing application for traditional DNA sequencing today is in clinical diagnostics. Take, for example, the well-known oncogene p53--mutations of this gene play a role in a growing list of cancers. The p53 gene consists of 11 exons, and it encodes for a 53,000-dalton phosphoprotein that regulates cell proliferation. What's been found, says Pharmacia's Ball, is that there are dozens of different mutations that the p53 gene can undergo. For researchers to discriminate among these mutations, the p53 gene from a patient's tumor is sequenced to see whether it is an aggressive type or one that is more treatable. Another application is in diagnosing retinoblastoma (RB), a rare inherited eye tumor that can cause blindness but, if detected early, can be eradicated with laser surgery. To date, more than 300 mutations of the RB gene have been documented. The Eye Research Institute of Canada, affiliated with the Hospital for Sick Children in Toronto, is using a Pharmacia system to rapidly identify members of families that carry a mutation of the RB gene. This allows the researchers to exclude the unaffecte Despite the promise of newly emerging capillary- and chip- based technologies, users shouldn't expect that conventional sequencing will vanish any time soon, says Pharmacia staff scientist Anthony Murray. "I don't think the slab is dead," he asserts. Detectors capable of reading narrower lanes will allow higher lane densities to be read on conventional gel systems, which will extend the life span of the existing technology. "I think in the near term, the gel approach is going to be pretty hard to displace," he says. Scott Veggeberg is a freelance science writer based in Swarthmore, Pa.

DNA Sequencing: Today's Technology And Beyond

Interested in reading more?

Become a Member of