Today's Lab

Tom Sargent remembers the day a student in his lab forgot to add boiling chips to phenol before firing up the heater on the distillation apparatus, and the panicked shouting and tearing off of the lab coat, goggles, gloves, and shoes that ensued when the phenol superheated and boiled over. "Fortunately he wasn't hurt," said Sargent, now chief of the section on vertebrate development at the National Institute of Child and Human Development, "but what a mess." Then, there was the time he hooked up a sequencing gel to a 5000-volt paper electrophoresis power supply. Another near miss—fortunately no one was in the room when the gel vaporized, causing the glass plates to explode. These scenarios are thankfully a thing of the past. Technology and capitalism have together provided instruments and kits that have taken over tasks that back in the 1960s and '70s could occupy weeks to months, and as Sargent's stories show, were occasionally dangerous. The days of labeling your own nucleotides, distilling your own phenol, purifying your own enzymes, or even making your own buffers are long gone. Innovation, the commercialization of reagent production, and simple modern conveniences such as the microwave oven have transformed labs into more efficient and safer places to work. Personal Computing Conceived Bob Klevecz, professor of biology at City of Hope, remembers spending a year localizing chromosome breaks by eye on 10,000 metaphase chromosomes, and then spending days keypunching the data onto computer cards, only to have the cards sent airborne after a collision between a student carrying the box of cards and a door. In those days, computers were sequestered in sealed rooms, often in separate buildings, and had none of the user-friendliness of today's PCs. Forever changing the face of personal computing in the laboratory was the LINC, Laboratory Instrument Computer, so-called to emphasize what this computer's place in the lab was to be—a computer to run instruments. The LINC's designer, Wes Clark, who describes himself as a "would-have-been physicist who got interested in computers," recognized the need for computers in the lab, right in the lab. This was a totally new concept at the time; bringing computers into the lab was simply not done, it was not even being considered, Clark recollects. But, Clark said, "It needed to be done. It was clear that laboratories could use an instrument like a computer," he said. Joe Towler, Washington UniversityPre-Desktop Computing: In this 1964 photo, Severo Ornstein of Washington University, in front of a rack-mounted LINC, points out some of the features of the display screen to Donald Blough of Brown University. The display produced graphical/alpha numeric information. In the black-and-white cover photo, taken in 1965, a neurophysiologist examines data recorded in the laboratory using the LINC embedded with other lab equipment. Tall cabinets hold an FM tape recorder that had data recorded on it to transcribe from analog to digital form. Four reels were the LINC's digital units used to read and record data, forerunners to today's floppy disk. According to Jerry Cox, senior professor of computer science at Washington University in St. Louis, Clark's concept from the start was to design a laboratory machine that could be controlled by one investigator, with many of the same features of today's personal computers. The LINC was self-contained, and it allowed researchers to acquire data and compute without requiring other people or equipment. "It was ground-breaking," said Cox. The LINC and the PDP series, considered the direct descendant of the LINC, set a new paradigm. Cox recalls, "People began to see what you could do with a laboratory computer and they saw their competitors with a LINC or a descendent, and said, 'Gee, I better get with the program or I'll get left behind.'" The LINC's first real test came in 1963 when Clark and his team brought it to the National Institutes of Health. In the laboratory of Robert Livingston, then scientific director of the National Institutes of Mental Health and Neurological Diseases and Blindness, the LINC was hooked up to an array of electrodes implanted in the brain of the lab cat, Jasper, and recorded the brain's electrical activity in real time. A historical account of the event quotes Livingston as saying: "It was such a triumph that we danced a jig right there around the equipment. No human being had ever been able to see what we had just witnessed. It was as if we had an opportunity to ski down a virgin snow field of a previously undiscovered mountain."1 From Benchtop to Desktop Now doing science without computers is unimaginable; they have inserted themselves into almost every aspect of lab life—collecting, analyzing, and plotting data, writing papers and grants, searching and reading the literature. With the number and variety of databases available online, researchers can make discoveries without setting foot in a lab. Consider, for example, microarray experiments, arguably the premier molecular genetic technique of the last decade. Microarrays allow researchers to interrogate the activity of an organism's entire genome at once. And given the large size of genomes, these experiments generate massive amounts of data—a million pieces of data from a 20,000-element array. To make the best use of this wealth of data, microarray researchers, spearheaded by Pat Brown of Stanford University, have created Web-based microarray databases, which provide access to raw data, and in the case of the Stanford site, a battery of software tools to analyze it. Not simply egalitarian in purpose, these databases allow researchers to confirm and extend the results of others. Says Gavin Sherlock, director of Stanford's microarray database, "There simply is so much [data], to be able to have other people verify your analyses of the data, you really need to provide them with everything, so that they can make their own judgment calls about it, not simply trust you." The concept of conducting experiments wholly in silico provides the business model for the Rockville, Md.-based company, Gene Logic. The company is creating microarray databases from a broad swath of tissues and organs, diseased and healthy, which it collects from collaborators at research institutes and universities around the world. Researchers who submit samples get access to the data from their own samples for free, which Robert Burrows, Gene Logic's director of corporate communications, points out creates the best kind of incentive for the researcher to provide high-quality samples. But the company puts a high price tag on its collection, with subscriptions costing millions of dollars per year. With most of their competitors migrating into the clinic, Burrows describes Gene Logic as the "last bastion of pure information play in the space. We think it's an extremely exciting, viable, and potentially profitable business. " At the Core Judy Singer-Sam, associate professor of biology at the Beckman Research Institute of the City of Hope, remembers spending months working out the sequence of a single 800-base-pair fragment. "It took lots of radio-activity, three people working into the wee hours of the night, running these long acrylamide gels and reading off the sequence by eye. Then came automated sequencers and core facilities to run them. "Now", Singer-Sam points out, "it takes one person less than a day." According to Ron Niece, a consultant for the Association of Biomolecular Resource Facilities (ABRF), the notion of shared instrumentation in core facilities was the direct result of automation. "In the '70s and '80s a lot of biochemistry was still being done in test tubes. But once the process was automated, the bar for chemical purity and delivery of small volumes went way up, and pretty soon it was a cyclical process—better equipment leads to better chemicals—and pretty soon the instruments were expensive." The ABFR, which was founded in 1988 to support laboratories that provide services for analyzing or synthesizing biomolecules, conducts periodic voluntary surveys to assess both the numbers and types of cores, and the quality of their work. They recorded a dramatic rise in cores in the late '80s; the number of cores responding to their surveys tripled between 1987 and 1989.2 The surveys also reflect research trends, with services such as amino acid analysis and HPLC becoming less popular with core customers, and new services including biosensors, chip technology, and real-time PCR, showing what technologies have captivated researchers. While cores were initiated as a cost-sharing measure, they also can save time in providing access to people trained to use the instrument. Niece explains, "If the instrument costs $100,000 and it's going to take you six months to get up and running, that's a very expensive experiment. But if you can walk down the hall or ship it to a specialized core, you are sharing the cost of the instrumentation and the cost of developing and maintaining the experts." In a 1999 survey to assess whether researchers' needs for instrumentation were being met, conducted by the science policy committee of the Federation of American Societies for Experimental Biology (FASEB), 84% of respondents said that, "shared equipment and core facilities at my institution are extremely important for my research."3 On a less optimistic note, only 5% felt that their institutions were able to adopt new technologies in a timely manner. NIH's National Center for Research Resources (NCRR) supports the core concept through its shared instrumentation grant program. This program, started in 1982, enables groups of NIH grantees to apply jointly for high-priced instrumentation. The original program covered instruments costing a minimum of $75,000 and a maximum of $250,000. Last year, a new version of this program, the High End Instrumentation Program, raised the ceiling to $2 million to bring on board instruments including high-resolution mass spectrometers and supercomputers. Research in the Fast Lane Another growth area is in the production of premade kits and reagents. Fueled by increases in the number and complexity of molecular biology techniques, the number of kits has skyrocketed in the last few years. Ellen Prediger, director of scientific communication at Austin, Texas-based Ambion, says her company has gone from offering one or two kits to more than 60 major kits and many more minor ones in the 12 years since the company's founding. Jerry Forrest, director of City of Hope's cloning lab, which stocks enzymes and kits for molecular biology techniques, sells $50,000 to $60,000 worth of reagents each month to a small research staff. This, he says, is up, way up, from the $500 to $1000 per month the lab sold when this started in the early '80s, an increase due principally to the large number of kits now available, according to Forrest. Some of the early kits were primitive by today's standards, says Chris Paddon, senior scientist at the biotech start-up, Xenoport in Santa Clara, Calif.—but he thinks that competition among manufacturers has led to better variety as well as quality. And it's not just a matter of convenience, though that certainly enters into the equation. Prediger points out that kits make a technique or even a whole field accessible to researchers outside their area of expertise and enable them to get data quickly for a paper or a grant. She said, "I remember just being able to publish a sequence. That was enough. That was such an incredible feat. Now you are expected to have expression pattern and genomic pattern. A lot more is expected and I think that is a big part of [the advantage of kits]." This sentiment is echoed by Paddon, whose group recently took up adenovirus for the first time. "We hadn't done this before, and to read the literature would have taken a long time. But if it's well explained in the handbook, you can start off and then branch off from there." The Down Side Today's lab is better, more efficient, and safer, largely due to a vigorous and inventive scientific supply industry, according to Sargent. The completion of the human genome project in record time exemplifies what can be done when the best and fastest tools are brought to bear on a problem. In fact, without automation and powerful computers, it might never have been attempted. Referring to the size of the human genome, Cox said, "If you want to deal with three billion of anything, you better get some help." Laura DeFrancesco (defrancesco1@earthlink.net) is a freelance science writer in Pasadena, Calif. References 1. S. Rosenfeld, "LINC: The genesis of a technological revolution," LINC's Twentieth Anniversary Seminar, NIH Report, Nov. 30, 1983. 2. R.L. Niece et al., "State-of-the-art biomolecular core facilities: A comprehensive survey," FASEB Journal, 5:2756-60, 1991; 6:792, 1992. 3. T.R. Zemlo et al., "Research equipment and resource requirements of NIH-supported investigators: An assessment of current conditions and recommendations for future funding and programs," FASEB, 2000.

Today's Lab

Interested in reading more?

Become a Member of