UNDERSTANDING CANCER,
GENETICALLY SPEAKING

Daniel A. Notterman, MD

BY MARYANN BRINLEY

Spend just a little time in the trenches with cancer researchers and you’ll pick up an exhilarating level of excitement. Like explorers who have discovered a brave new world and are anxious to look over the horizon, they are supercharged by what they see using gene expression microarray analysis. While comprehensive cures for cancer are still miles away, genetically speaking, our understanding of what makes cells grow wildly in cancerous chaos is close at hand.

Connecting a gene to a disease used to take years of laboratory work and analysis. Now, it happens nearly overnight "cheaply, rapidly and fairly accurately," says Daniel A. Notterman, MD, University Professor and chair of pediatrics at UMDNJ-Robert Wood Johnson Medical School (RWJMS) and a scientist working in the molecular biology of cancer. The way we prevent and treat cancer today "is not ineffective but it is empiric and in 20 years, it’s going to seem as quaint as the treatment of infectious diseases did at the turn of the century." To exploit the new advances in technology, The Cancer Institute of New Jersey (CINJ), directed by William N. Hait, MD, PhD, RWJMS professor of medicine and pharmacology, has developed a Core Microarray Laboratory, with state-of-the-art equipment to examine the function of thousands of genes simultaneously. Notterman is the director of this core facility.

Within our 100,000 human genes, and how each is turned on and off (known as "expressed" in scientific circles), are the keys to clearing up cancer confusion. You’ll see the word "expression" often, as well as other novel phrases, if you are trying to learn more about cancer. Scientists have fashioned a distinct vocabulary linked specifically to their fast-paced crusade. (See sidebar, "Words Cancer Researchers Use.") The very system of classifying cancers—often by where in the body they begin or superficial features—is evolving because the underlying molecular signals being transmitted by individual cancer cells leave clear, and sometimes surprising, genetic fingerprints. What used to look like clear-cut cases of leukemia, lymphoma, breast, lung or colon cancer, simply aren’t what they seem to be on the surface at all.

"Look at this," Notterman says, demonstrating how a few clicks on the National Cancer Biotech-nology Center Web site—a repository of genetic information gathered by scientists all over the world—will offer him nearly instantaneous feedback.

Rows and rows of combinations of the letters of the genetic alphabet, and the four basic chemicals they represent, pop up in a stream. A (adenine), C (cytosine), G (guanine) and T (thymine) are the ingredients of every organism’s DNA or genetic information. It looks like secret cryptographic writing but 99.9 percent of the code for all human beings is identical. What makes us so individually unique, and perhaps more or less susceptible to diseases or environmental onslaughts, is the way our genes are sequenced and expressed.

"See. There are thousands of them," says Notterman, who has been studying the genetic activity of colorectal cancer. Pointing to the computer in his office and scrolling down, he explains, "Let’s say I discover a gene in my work and it has a particular sequence. I want to see what this gene is and if there are any similar matches to it. I type it in, highlight it and then, I use BLAST down here on the screen," which searches the database in Washington, DC (at the National Human Genome Research Institute). "These programs help biologists catalogue, track and compare the genes they study."

Once predicted to take 40 years, the $3 billion international race to identify all human genes began in 1990 and was just completed in April 2003—27 years sooner than expected. It’s been a mad dash and sometimes, the human genome, comprised of three billion genetic letters (strings of those four letters ACGT), looks a little like a book with no word breaks, paragraphs or punctuation. Yet, tools have emerged to help scientists understand what the code is saying. For instance, gene chips, miniature glass or silicon slides imprinted with grids of single stranded samples of DNA (up to tens of thousands), are used. Also known as cDNA microarrays or biochips, they let researchers measure the normal and abnormal activity in a gene by taking advantage of one of DNA’s natural tendencies: hybridization. Because your DNA is double-stranded, when one half is separated from the other, molecules will always try to re-unite. This technology relies on a complicated, multi-step process but, to simplify it, you might try picturing the twisting pairs of a single, stringy DNA fragment. Now, imagine pulling the strands apart. In a gene expression microarray lab, extracted strands are colored with fluorescent dye and then washed over gene chips already imprinted with samples of a patient’s DNA or RNA (colored a different hue), as well as chips of particular diseases (washed in yet another color). Within hours, strips of cancerous DNA bind naturally to their complementary partners
embedded on the chips and create colorful patterns of data which computer software programs collect and illustrate. The level of genetic activity—all those on and off signals—can be seen in the brilliance or dullness of the colors. Notterman says that linking cancer to specific genetic activity involves "huge pieces of data so we have to use computers to turn the specific signals into colored pictures that help us identify patterns and specific cancer signatures."

Take two individuals, both diagnosed with leukemia and given a 50 percent chance of survival. Each one has leukemia cancer cells, which may look the same before microarray analysis, but these cells will appear distinctly different in the fluorescently colored readouts. This is why scientists are beginning to comprehend how two people, with the same cancer, will experience separate outcomes. As Notterman says, we can also note "what molecules are being abnormally expressed in these tumors to help them on their evil way."

"You can easily foresee a clinical oncologist having a set of special chips in her office and using them for diagnosis," says Notterman. (No one is predicting when this will happen because, as the race to decode the human genome illustrated, forecasts can certainly be off the mark.) Gene chip-based information may also be used to tailor the treatment to an individual’s DNA profile. This new science, pharmacogenomics, is just emerging but is already having a practical impact on some cancer patients. Right now, in a common form of childhood leukemia, doctors can screen kids, looking for deficiencies in a protein called TMPT which is critical to the success of chemo-therapy. Pharmacogenomic clinical trials are also being conducted to test cancer patients genetically for a liver enzyme that is needed for 30 different classes of drugs to work.

For the future, the hope is that drugs will be able to target only the wayward genes, leaving normal cells unfazed. Gene-based remedies may stop cancer before it starts at the cellular level in a preventive strike. At the very least, for example, if your genes are lighting up with colon cancer signals early on, you’ll certainly schedule routine colonoscopies, adopt dietary defenses or simply take aspirin daily because it appears to decrease polyp formation which can lead to cancer. Alerted to lung cancer susceptibility, you could find serious motivation to quit smoking. Researchers also suspect that it is possible to turn off, or silence, a mutant gene and the "antisense" drugs which will do this are in the investigative pipeline.

words cancer researchers use expressed: in genetic research circles, this verb refers to whether a gene is active or inactive, on or off. microarray: a glass or silicon slide printed with DNA molecules gene chip: another word for microarray DNA: deoxyribonucleic acid, the molecule carrying hereditary information RNA: ribonucleic acid, molecules which carry genetic code information and are involved in protein formation bioinformatics: using computers to understand biology better through-put: volume pharmacogenomics: the study of how genes affect your body's responses to drugs

"The approach will be earth-shaking as we get more and more into it," says Hait. "We are discovering that single nucleotide changes (single nucleotide polymorphisms or SNPs) in critical genetic regions can lead to surprising results. One might envision the day when at birth, we are given our unique set of SNPs that define our medical predispositions."

All these new analytical techniques started to emerge about six years ago, according to Notterman, at the same time as the worldwide effort to identify all human genes. "The synergy of these two developments has empowered both clinical and basic scientists in cancer to more fully understand the disturbed signaling at work that underlies cancer."

This synergistic wave even sent Notterman back to school after 12 years of being a pediatrician. "I was always a little restless just being a clinician. I do love it and when anyone asks me what I am, I have to say: pediatrician. Unfortunately, I came through medical school when all this molecular biology was just being discovered." In 1992, at a time when he was fearful about missing the most interesting developments in modern medicine, Notterman took a sabbatical from his position as director of the Division of Pediatric Critical Care at New York Hospital’s Cornell Medical Center to "go back" to being a student. "I really wanted to be a scientist but I needed an opportunity and someone willing to take me into a lab."

That someone was Arnold Levine, PhD, a molecular biologist at Princeton University and renowned researcher who discovered the p53 gene which is implicated in breast, lung, colon, prostate, bladder and cervical cancers. "Arnie Levine didn’t need me in his lab," Notterman explains. "Yet, he is the kind of scientist who understands that the real obligation of science is to train more scientists. He even paid me as a post-doctoral fellow and encouraged me to think of colon cancer."

Former president of Rockefeller University and currently a professor of pediatrics and biochemistry at RWJMS as well as a researcher affiliated with CINJ, Levine recalls, "We discovered p53, which acts like a checkpoint in suppressing tumors or unregulated cell growth, in 1979, but it took 10 years for us to understand how important it is. By 1989, we had found that p53 or whatever it was, was behaving like a tumor suppressor and that a virus could inactivate it."

In the next four years, Levine and his team received "the surprise of our lives when we found that 50 to 55 percent of all human cancers have mutations of this gene. That’s when elation really came forward. I don’t think there is any feeling you can have that is comparable." A p53 network, normally turned off, is activated when cells are under stress or damaged. Writing in the journal, Nature, with fellow researchers Bert Vogelstein and David Lane, Levine explains, "Just as a car’s brakes regulate its speed, properly functioning tumor-suppressor genes act as brakes to the cycle of cell growth, DNA replication and division into new cells. When these genes fail to function properly, uncontrolled growth—a defining feature of cancer—ensues."

Following in Levine’s footsteps, thousands of researchers, including Notterman, have focused on this gene, the most commonly mutated one in cancer, including colorectal cancer. Because colon tumor tissue is available in all stages of the disease, "It has been an excellent model," Notterman says. "You can correlate the tissue to the patient’s outcome." Working with several other scientific teams on "Expression Profiling of Colorectal Cancer," Notterman is being funded through a $2,454,000 multi-institutional grant from the National Institutes of Health and National Cancer Institute.

Together, Levine and Notterman have also received $367,219 from the Breast Cancer Research Foundation to create tools which will test breast cancer cells for a protein called WISP-1. If you are diagnosed with breast cancer and WISP-1 turns up in your cellular makeup, then chemotherapy is more likely to fail. The goal of this work is not only to save a patient from unnecessary therapy but also to make these tumors more responsive to treatment. "I really love doing the science, analyzing the data, and struggling to get the answers," Levine says. "We now have a fairly complete parts manual and the challenges are to decipher the regulatory signals that control timing, positions and levels of activity of each gene."

Lured back into the world of pediatrics last year by Harold L. Paz, MD, Dean of RWJMS, Notterman administers, teaches, and treats sick children at the Bristol-Myers Squibb Children’s Hospital at Robert Wood Johnson University Hospital. He would like to get up to the seventh floor of the Clinical Academic Building (CAB), where the Cancer Institute’s new million dollar microarray facility is located, more often than his regular two mornings a week. "This is a very important priority for me. Because of my interest in gene expression studies, when I joined the faculty I was asked to take over the core expression lab." With funding from New Jersey Governor James McGreevey’s allocation, "we now operate a center where any scientist in the medical school, particularly the Cancer Institute, can send us samples for study."

On a Wednesday afternoon, a staff of six young investigators with varying academic backgrounds and goals—Emmanuel Selvanayagam, PhD, Hao Liu, PhD, Yvonne Wen, PhD, Jen Greenman, Wei Liu, PhD, and microarray facility manager Curtis Krier—are busy in three airy rooms on the top floor of the CAB in New Brunswick.

"I love to give tours," Krier says. "This is great stuff. Years ago when someone would ask a scientific question and do research, the tools they had were limited. This is so high tech and high through-put (high-volume) that it’s revolutionary. Everyone is invited to come up here and take a closer look."

"Here we have a set of 7,500 human genes all represented on these plates," Krier explains, holding out a small rectangular box with probe holes on top. "Each little well here holds a sample of a particular gene. We used to study one gene but what we couldn’t do is see how it affected others. Now if we turn one gene on, we can see how it changes hundreds of others. What I’m doing today is making a copy of a set for someone. You can imagine how great the margin of human error would be if I tried to do that by hand, individually pipetting 7,500 genes." Krier lets mechanical "picking wizards load up the chips and the plates. You just program the system and away you go. Like a big liquid handling robot, it’s really amazing and fun to watch, too."

Here’s the gel documentation room, the imaging system, the plate reader, various multi-systems of computer networks, and cell culture equipment.

There are two systems—a commercially produced Affymetrix system with gene chips holding more than 20,000 samples each, as well as the lab’s own version which produces and reads their homemade cDNA arrays. Chips can be customized to suit a researcher’s particular interests.

What can seem impersonal when described in futuristic descriptions of cancer research suddenly becomes very humanistically real. These cancers being studied come from real people. These researchers are really captivated.

Greenman is working with combinations of anti-cancer agents. "I had one assay that needed my attention every two hours," she laughs. "I’m glad I live close."

Wen has been concentrating on cancer growth factors. "The idea is to see what makes tumor growth speed up."

Hao Liu is a microbiologist as well as a computer scientist or what is now called a bioinformatics specialist. "Yes, that’s a good educational mix," he says. Try to visualize the reams of data you get when "signals from hundreds of thousands of genes are picked up. You have to keep track of this stuff and the machines are only as good as we train them to be."

Left to right: Hilah Gal, Hao Liu, PhD, Emmanuel Selvanayagam, PhD, Jennifer Greenman and Curtis Krier, Core Expression Array Facility Manager, all in the RWJMS microarray laboratory.

Standing behind Selvanayagam and Liu while they pore over the results of a microarray test, you can see how 22,384 genes have responded when fluorescently labeled RNA from abnormal cells tried to attach to their DNA.

"Each one of those tiny dots is a piece of a gene," Selvanayagam explains. "We’re looking for the signals. There you go. Zoom in so you can see more." The image becomes clearer and more colorful. What follows with the next click is a readout of the activity of each of several thousand genes. Some of these genes are really turned on to a high degree. Others are quiet, indicated by an NC. There are rows and rows of NC.

"NC means no change," he says.

Now, look in the palm of your hand or anywhere on your body, in fact. Then go back to the screen. The twinkling lights appear to be from outer space but they could be pieces of you in this brave new world of inner space.