High-Tech Choir Master

Elaine Mardis can make DNA sequencers sing, generating genome data that shed light on evolution and disease.

By | January 1, 2012

BILL SAWALICH/BARLOW PRODUCTIONSElaine Mardis was in the right place at the right time. During her senior year as a zoology major at Oklahoma University, Mardis found herself at loose ends. “It wasn’t readily apparent to me what to do next,” she says. “Then I took a biochemistry class—and the instructor was one Bruce A. Roe. He started teaching us about this incredible world of molecular biology and it really opened my eyes. Very quickly, fruit fly genetics—which is what I was doing for my honors thesis—began to seem pretty mundane and boring, compared to working with DNA and enzymes at a molecular level.”

Mardis graduated in 1984, and Roe convinced her to apply for graduate school—and to join his lab. “He was transitioning away from being a tRNA guy to becoming more active in DNA sequencing. I didn’t realize it at the time, but Bruce had learned DNA sequencing directly from Fred Sanger.” By the time Mardis had finished her coursework and was ready to knuckle down in the lab, a company called Applied Biosystems came out with the first commercially available DNA sequencer that used fluorescence instead of radiolabeled nucleotides. “Bruce, by a variety of surreptitious means that weren’t entirely clear, found $100,000, flew out to California, and brought back the second instrument that Applied Biosystems produced.” Mardis jumped in with both feet. “Some of the first projects I worked on involved exploring how we could transition from radiolabeled to fluorescent sequencing and get the best-quality results. I’ve been doing technology development ever since.”

“It makes me crazy to listen to people say sequencing is so cheap. Sure, generating data is cheap. But then what do you do with it? That’s where it gets expensive.”

“To me, timing is everything,” says Mardis. “I’ve never had a great big Master Plan checklist as I go through life. I just sort of stumble from one opportunity to the next.” Those opportunities have resulted in some of the most stunning achievements in modern molecular biology, including the determination and analysis of the first animal genome sequence, that of C. elegans; the human genome reference sequence; and the first complete sequence of cancer cells and nonmalignant cells from the same patient. Here she talks about better therapies through sequencing, the trouble with SNPs, and prepping for her second-degree black belt.


Seeking resolution. As a thesis project, Mardis put her mastery of technology to the test by sequencing a plasmid isolated from Mycobacterium leprae. “This was a very high GC-content plasmid, so it required some pretty tricky modifications to the enzymology and chemistry of the sequencing reaction to read through those GCs.” When she finished in 1989, however, she discovered that few other laboratories were working on sequencing technology. “Most were focused on physical mapping—piecing together YAC [yeast artificial chromosome] clones and such. But I wanted nucleotide resolution. So I took an R&D job at Bio-Rad. That was basically a good plan, as it turns out”—because in 1993, fellow Roe trainee Rick Wilson tapped her to join him at the Genome Institute at Washington University in Saint Louis, where she currently serves as codirector—as well as director of technology development. “Rick and Bob Waterston were getting ready to scale up and take on sequencing the C. elegans genome and they needed a group that would focus on technology development. That’s what I had been doing since I started graduate school, so for me it was like a comfortable old shoe—but with a much larger playground to run around in.”

Good to the last nucleotide. “Our center, along with the Sanger Institute, really did a great job on the worm genome, when all was said and done. I feel a little arrogant saying that, but it really is one of the best-sequenced model organism genomes. Ask anybody in the C. elegans community. It was the quantum leap that was needed for that community to expand and move forward in its biological studies. The other thing that’s great about the worm is that it’s been completed down to the last nucleotide: it’s completely finished from one end to the other. The worm genome did throw us a few curveballs: it’s got regions with very long stretches of GC, as well as highly conserved tandem repeats, that even with the chemistries we had available to us in 1998 were very difficult to get through. But we rose to the challenge and produced a very high-quality reference sequence.”

Reading cancer. In 2006, Mardis and her colleagues undertook a project to sequence the full genome of a cancer patient. “The project just got roundly thrashed by the review committee, so we did not get a fundable score.” One of the most common objections to the endeavor was that it was not business as usual. “ ‘Couldn’t you just do the same thing by focusing on individual genes? What are you going to find by sequencing the whole genome?’ Here’s a great example: in one of our patients, an entire gene—which encodes a DNA methyltransferase called DNMT3a—was removed by a 1.5-megabase deletion. That’s a mutation we were never going to find by PCR”—because the gene was not there to be amplified. “So you’re only going to get part of the picture if you go in using more traditional, focused methods.” The group eventually did get funded for the project. “By the time we went back in and resubmitted, we had already published the paper. So it was like, ‘Oh, OK. Maybe this can work.’ ”

Prescription for sequencing. By sequencing the genomes of additional patients with acute myeloid leukemia (AML), Mardis and her colleagues hope they can help doctors choose the best possible avenues of treatment. “In AML, one of the challenges is that most patients present with normal-looking chromosomes. Without visible translocations, there’s no way to predict the best method of treatment. We showed in our 2010 New England Journal of Medicine paper that mutations in DNMT3a are prognostic for patients who will have the worst outcome. You could imagine, in the next one or two years, seeing every patient presenting with AML being typed for their mutation spectrum in this gene. And if they do contain a DNMT3a mutation, they may actually go straight into more aggressive treatment: chemotherapy to induce remission, followed by bone marrow transplant.” At the same time, sequencing could save patients from unnecessary treatments and direct them toward more effective, targeted therapies. One woman the team sequenced had been slated for a bone marrow transplant. But sequencing revealed that she had suffered a subtle insertion that generated a type of gene fusion that can be effectively treated by standard chemotherapy followed by a chemical called all-trans retinoic acid, or ATRA. “It’s not the clinical standard of care to get ATRA—unless you have the common translocation,” says Mardis. This patient’s mutation had the same effect as that translocation, but would not have been picked up by conventional chromosomal analyses. “The coda is that we generated some smaller probes that could detect this insertional event and we went back and identified two additional patients that had essentially the exact same mutation. It was not just a one-off—it happens in other patients. So sequencing patients can impact clinical care, especially for those patients where it’s not clear how best to treat them.”


“Sequencing patients can impact clinical care, especially for those patients where it’s not clear how best to treat them.”

History lesson. I think people don’t stop to really appreciate what’s gone on before. If I give a talk about cancer genomics, I’ll point out that even though we now have this incredibly sensitive ‘microscope’—a sequencer that can tell us everything that’s gone wrong in the genome of a cancer patient—people recognized that cancer is a disease of the genome back in the ’70s by looking under a light microscope at cancer chromosomes. Their microscope wasn’t as sensitive as ours is. But it was still telling them basically the same thing.

Corporate learning. “Most scientists are crappy managers—because they never have much exposure to management practices or how to handle difficult situations, like when two people in the lab aren’t getting along. Being in the business environment at Bio-Rad introduced me to a lot of management concepts and practices. The other thing that was really important in the business setting was getting a basic understanding of quality assurance/quality control: how to make sure that the product going out to the customer is uniform from one batch to the next. That’s actually one of the first things I took on when I moved to Saint Louis. As we scaled up, it became very clear that we needed to move beyond this early practice where our technicians all made their own sequencing mixes every morning. My time at Bio-Rad taught me how to set up bulk processes: making enough mixes for an entire day or an entire week, doing shelf-life testing, things like that. It sounds really mundane, but it’s absolutely essential for a large operation.”

Hidden costs. “It makes me crazy to listen to people say sequencing is so cheap. Sure, generating data is cheap. But then what do you do with it? That’s where it gets expensive. ‘The $1,000 genome’ is just this throwaway phrase people use and I don’t think they stop to think about what’s involved. When we looked back after we analyzed the first tumor/normal pair, published in Nature in 2008, we figured that genome—not just generating the data, but coming up with the analytical approach, which nobody had ever done before—probably cost $1.6M. If the cost of analysis doesn’t fall over time, we’re never going to get to clinical reality for a lot of these tests.”

SNP Sniping. “I’ve long been a proponent that there’s much more to genome diversity than SNPs. People are really enamored of their SNPs. Don’t ask me to explain why. I think SNPs were a good first target because they’re easy to find. But to a very large extent—and this is now being borne out by studies in patients with developmental delays or autism—it looks like many of these diseases are due to copy number variations, rearrangement events, and deletions or amplifications. So I just think it’s really important to sequence and look for structural variations in whole genomes, especially in cancer.”

Losing stock in science? “Right now, if you look at the major genomics companies—the number of people they employ and the instruments and reagents they produce—we’re probably one of the few US enterprises that’s still making something of value. And doing new science. But by downgrading the funding of NIH and not supporting science at its very essence, we’re damaging those manufacturers and their ability to sell products and sustain a realistic stock price. This worries me, because it doesn’t take long to lose your advantage by underfunding science and underfunding innovation.”

Knowledge is power. “The specter of full genome sequencing makes people uncomfortable because we do find out things about their genomes and in some cases their susceptibility to disease. For ApoE3 or ApoE4 sequencing, you have a couple of alleles that make you more susceptible to Alzheimer’s disease later in life. That’s probably not going to be so germane to somebody with stage IV cancer. But there’s a lot of hand wringing that goes on about this. This is one of the reasons we published a paper, in the JAMA April 20, 2011 issue, about a patient who had a novel Li-Fraumeni mutation that potentially impacted her three children.” Carriers of this type of mutation have a 90% lifetime risk of developing cancer. “That’s not an easy message. But at the end of the day you hope they’re thankful for having that information and that they did get tested. Maybe none of them inherited that mutation. But I think it’s better to know, and to be prepared, than to hide from knowledge.”


Up and at 'em. “I’ve always been an early riser. I can actually wake up at 4:30 in the morning anywhere on the planet. It’s like I’m hardwired, no matter what time zone I’m in.”

Taking a break. The Genome Institute has a taekwondo club. “I have my first-degree black belt and I’ve been poised on the edge of testing for my second-degree black belt for about six months now. For my next degree test, I’m going to break four boards with my heel, turn around and break another board with my fist and a last board with my other heel. That’s the plan, anyway.” But it isn’t all about splintering wood. “We do occasionally put on pads and kick and punch each other. So that’s fun. Especially for the people who get to beat up on me. I’m reasonably high up in the organization at the Institute, so it must be fun for them to throw me to the ground.”

For the love of food. “I just read this amazing book called Life on the Line by Grant Achatz. He talks about his life growing up and becoming one of the most amazing chefs on the planet—and then finding out, after getting lots of accolades and Beard awards, that he had stage IV tongue cancer”—a disease for which the standard therapy is removal of the lymph nodes along with most of the tongue. Achatz found an oncologist who used an alternative therapy regimen that allowed him to keep his tongue—although he did lose, and then slowly regain, his sense of taste. “I loved the book—and I am a foodie. Now I have a mission to make it to his restaurant in Chicago, Alinea. The AACR meeting is in Chicago next year and I was chosen to give a talk in the opening plenary session, which is a real honor. To celebrate, I’m going to try very, very hard to get a reservation. It’s something I have to experience at least once in this life.”

Add a Comment

Avatar of: You



Sign In with your LabX Media Group Passport to leave a comment

Not a member? Register Now!

LabX Media Group Passport Logo

Popular Now

  1. Broad Wins CRISPR Patent Interference Case
    Daily News Broad Wins CRISPR Patent Interference Case

    The USPTO’s Patent Trial and Appeal Board has ruled in favor of the Broad Institute of MIT and Harvard retaining intellectual property rights covered by its patents for CRISPR gene-editing technology.

  2. Henrietta Lacks’s Family Seeks Compensation
  3. Can Plants Learn to Associate Stimuli with Reward?
  4. Humans Never Stopped Evolving
    Features Humans Never Stopped Evolving

    The emergence of blood abnormalities, an adult ability to digest milk, and changes in our physical appearance point to the continued evolution of the human race.

Business Birmingham