Pulsed-Field Electrophoresis Enhances Genome Effort

When the technique of pulsed-field gel electrophoresis (PFGE) was first described by David C. Schwartz and Charles R. Cantor almost a decade ago (Cell, 37:67-75, 1984), many molecular biologists recognized its potential immediately. A flurry of innovation in instrumentation based on the idea followed in the next half-dozen years. Despite this interest and activity, only in the past few years has PFGE secured its place as a mature, integral laboratory tool in molecular genetics generally and the Human Genome Project particularly. What Schwartz and Cantor demonstrated in 1984 was that, by alternately applying two electric fields at an angle across the gel instead of using one constant field, larger DNA could be resolved than ever before--much larger. With conventional electrophoresis in a constant field, DNA fragments above a certain size co-migrate in a broad band and do not separate properly. "The key advantage to PFGE is that you can separate bigger pieces of DNA than you could before," says Gilbert Chu, an assistant professor in medical oncology at Stanford University Medical Center. "Practically speaking, the upper size limit has been extended from 20,000 base pairs [with conventional electrophoresis] up to 5 million base pairs. It's a gigantic jump." Katheleen Gardiner, a molecular biologist at the Eleanor Roosevelt Institute for Cancer Research in Denver, adds perspective to these figures. "It takes the human genome initiative from being a thought experiment to being something that people can actually go out and do," she says. "Previously, if you had two genes that were 100,000 base pairs apart, they might as well have been on different chromosomes, as far as analysis on standard gels would go." As a result of its ability to handle larger pieces of DNA, PFGE technology has become an indispensable tool in many labs, key to some of the fastest-changing fields of molecular biology. A number of instruments offering variations on the basic concept are now commercially available. "It's opened up this molecular genetic front," says Maynard V. Olson, a professor of genetics at Washington University in St. Louis. "It's had a big impact on work on microorganisms, where the genomes are of sufficiently low complexity that you can do direct, whole-genome analysis. It's encouraged the development of microbial cloning systems for larger DNA molecules, [such as] the yeast artificial chromosome [YAC] systems. And, finally, there is the direct analysis of DNA from higher organisms--[although] it's not enough just to separate the digested genome. You have to do additional indirect assays to find particular fragments of interest. "In a few years," Olson adds, "PFGE has gone from being fairly esoteric technology, where a laboratory would have one pulsed-field gel apparatus set up over in the corner, to being routine technology. Many, many labs now have five or 10 of these things running all the time." While Cantor's name is, perhaps, more closely identified with the human genome initiative, both as a researcher and an administrator, PFGE, one of the main research tools of the effort, appears to have originated with David Schwartz. "The grapevine," says Chu, "is that David Schwartz started thinking about these issues [earlier], and, then, when he moved over to Cantor's lab, he brought the problem with him. The germ of the idea started with him." Schwartz, now a professor of chemistry at New York University, explains the pulsed-field effect he and Cantor discovered as being conceptually straightforward. Essentially, larger DNA molecules, migrating through a gel under the influence of an electric field, are not able to reorient themselves to resume migration as quickly as are smaller DNA molecules, when the angle of the electric field is changed. This difference results in a greater ability to separate larger molecules. "DNA molecules are deformable [their shapes can be changed], and, by changing the direction of the electric field, you can change their shape," Schwartz says. "Big molecules change shape at a different rate than small molecules do, and the way they move through the gel reflects the shape change." Chu has developed an effective computer model of PFGE in which the molecules are seen as deformable "bags" moving through the gel. The bags elongate along the electric field lines to migrate quickly and then must reorient and elongate along new lines when the field is switched. "Larger molecules spend more time in slowly migrating transition shapes and less time in a rapidly migrating, fully oriented state," he writes (Methods, 1:129-42, 1990). "Selective retardation of larger molecules then leads to resolution in the pulsed field." Olson emphasizes a slightly different aspect of the process in describing PFGE--speed vs. acceleration. "Classical electrophoresis exploits differences in the steady-state electrophoretic mobility of DNA molecules or proteins," he says. "It involves the constant speed component of their migration. What PFGE plays on is differences in their acceleration." "It's sort of like a truck vs. a motorcycle," says Schwartz. "We're separating trucks here." A number of apparatus designs taking advantage of the pulsed-field effect emerged during the mid- and late 1980s. Some involve vertical gel boxes, some horizontal gel boxes. In some, the plane of the electric fields is parallel to the flat gel; in others, it is perpendicular to the plane of the gel. The electronics and controlling systems vary in sophistication. Each has its advantages and disadvantages, although not everyone views the differences among the devices as significant. "They're different widgets to do the same damned thing," says Schwartz, apparently unimpressed by the various interpretations of his pulsed-field discovery. "Some widgets work better than others, but the effects are the same." Eric Lai, an assistant professor of pharmacology at the University of North Carolina, Chapel Hill, also plays down the differences in competing systems. He suggests that some of the variation in design of PFGE instruments is at least partially motivated by a desire to avoid patent constraints. "I don't know how many people have tried to pour a vertical agarose gel," he says, "but it's not that simple. It requires a lot more work than a horizontal gel, where you melt your agarose in a microwave in three minutes, pour it, and walk away, go have a cup of coffee. When you come back, it's done. "If there were no patents," Lai observes, "I'm sure that all the companies that make the vertical gel boxes would make horizontal gel boxes. Let's put it that way." Still, there are admitted differences in the designs. One early method based on the ideas of Schwartz and Cantor was the orthogonal-field agarose gel electrophoresis (OFAGE) system developed by George F. Carle and Washington's Olson (Nucleic Acids Research, 12:5647-64, 1984). The electric fields are parallel to the gel and cross the gel lanes on the diagonal, at approximately 90 degrees to each other. As the fields are switched, the DNA zigzags down the lanes. One problem with the OFAGE design is that the fields are not uniform, so that only DNA in the center of the gel migrates straight down the gel. This makes lane-to-lane comparisons difficult or impossible. The field-inversion gel electrophoresis (FIGE) device expresses an elementary solution to the problem of nonuniform fields (G.F. Carle, M. Frank, and M.V. Olson, Science, 232:65-8, 1986). In the FIGE device, a single set of electrodes is situated with one at the top and one at the bottom of the gel. The field is regularly inverted during the run. The effect is of two electric fields at an angle of 180 degrees to each other, rather than at 90 degrees to each other, as in the OFAGE system. To achieve forward migration, the field strength or the pulse time is greater in one direction. FIGE devices are inexpensive and accurate for separating smaller DNA fragments and give straight lanes for comparison. Unfortunately, with larger DNA sizes, say, more than 600,000 to 700,000 base pairs, the molecules tend to co-migrate, thus failing to resolve into bands. Voltage adjustments can ameliorate this effect, but it remains a problem in separating larger DNA with FIGE. Transverse alternating-field electrophoresis (TAFE) represents another solution to the problem of nonuniform fields (K. Gardiner, W. Laas, and D. Patterson, Somatic Cell Molecular Genetics, 12:185-95, 1986). The angles and orientation of the electric fields to each other in TAFE are similar to those in OFAGE, but the plane of the gel, which stands vertically, is perpendicular to the plane of the fields. This causes the DNA to zigzag front to back, across the thickness of the gel. An advantage is that the resulting separations can easily be compared lane to lane, right to left. But a disadvantage is that the field varies from top to bottom, in both strength and angle. Rotating gel electrophoresis (RGE) features a single set of fixed electrodes and fields parallel with the gel, like OFAGE, but it is the gel that moves, rather than the fields--rotating back and forth at the switch times to subject the gel to a different field orientation (E.M. Southern, R. Anand, W.R.A. Brown, and D.S. Fletcher, Nucleic Acids Research, 15:5925-43, 1987). One advantage introduced by RGE is that the angle between the fields can be varied. A major disadvantage of the RGE system is that the gel can be disturbed by the physical movement, making very short pulse times and the use of fragile, low-percentage agarose gels a problem. The development of the contour-clamped homogeneous electric field (CHEF) system represented a major improvement in PFGE (G. Chu, D. Vollrath, and R.W. Davis, Science, 234:1582-5, 1986). With this system, a hexagonal array of electrodes controlled by a resistor circuit creates highly uniform fields parallel to the gel plane. Reorientation angles can be varied electronically, so that the instrument is extremely adaptable. One result of work with the CHEF system was the discovery that angles between the fields of about 120 degrees generally resulted in the best separations. The CHEF system also allows good lane-to-lane comparisons. This system has been able to produce good resolution of DNA as large as 10 million base pairs. The predictable next step was to computerize the CHEF system, so that the variables of voltage, switch time, and angle could be fully controlled. The programmable autonomously controlled electrodes (PACE) system does this (S.M. Clark, E. Lai, B.W. Birren, and L. Hood, Science, 241:1203-5, 1988). "With the fully automated version," says Lai, "we assume you're going to use a certain percentage of agarose, a certain buffer--basically, a standard cocktail. All you have to do is put in the [DNA] size range you want to separate, and the program will tell you how long to run it and what switch time, voltage, and angle you'll need. If you want to get fancy and change some of those parameters, you can also do that, and the program will provide you with a correction factor." The commercial instrument that integrates the CHEF and PACE systems is the CHEF Mapper from Bio-Rad Laboratories of Richmond, Calif. The company says that, because the system relies on software rather than hardware for its flexibility, the user can simulate most of the systems described above--and many others likely to be conceived in the near future. "If you go to a meeting," Lai explains, "and somebody says, `You get much better data if you run it this way than if you run it the old way,' you can just go back to your lab and, the very next day, you're doing it." For some users, the CHEF Mapper system, while versatile--but also very expensive--is an example of overkill. "To my mind," says Gardiner, "it's like taking a battleship out duck hunting. It looks like it belongs in the cockpit of a 747." Schwartz, who originated PFGE, feels that the technology is reaching a logical limit in its sophistication. "All this smacks of maturity," he says. "The advances are sort of asymptotic, I would say. Things are leveling off." Not everyone thinks that PFGE has run its course. Olson, for one, anticipates important advances in using the technique for DNA sequencing. He notes that PFGE allows not only precise separation of larger molecules in an absolute sense--and molecules that differ by only a few percent of their length--but also separation over a larger range of molecular sizes, by a factor of roughly 10. If this aspect of PFGE technology could be extended to include DNA sequencing, he says, it would be a significant contribution. "In principle," Olson says, "with current DNA-sequencing gels resolving five to 500 bases--I'm speaking now of single-stranded DNA--it's conceivable that a pulsed-field gel machine might extend that to five to 5,000, something like that. I think that, with more advanced instrumentation, very fast switching, and, importantly, better theory, you'll see this same expansion of size range of gel systems pushed down into smaller sizes of molecules." Many people are working on this idea, says Olson, and they have had encouraging results. "If you go into any genome analysis laboratory doing large-scale physical mapping, you'll find PFGE apparatuses all over the place," says Olson. "If you go into a big DNA-sequencing lab right now, you won't find them. Perhaps in a few years you will, if all this is successful." NYU's Schwartz has an entirely different view of the future of not only his PFGE creation, but electrophoresis generally. "The next wave is no electrophoresis at all--not using gels and electric fields to manipulate and analyze molecules at all," he says. "Gel electrophoresis is cheap, it's easy to do, but it's too slow, too painful." If not electrophoresis, what then? "We're working on something," he says, "that sizes molecules without separating them." Schwartz is vague in describing the new technique, called optical contour maximization. Details, he says, will appear in a soon-to-be-published article in Nature.

Pulsed-Field Electrophoresis Enhances Genome Effort

Interested in reading more?

Become a Member of