|
By mining the riches of the human genome, Merck scientists are gaining new ideas about how some diseases are caused and how they should be treated.
Dr. Stephen Friend and his Merck affiliated colleagues in Kirkland, Wash., and West Point, Pa., are at the cutting edge of this field. Dr. Friend's group has pioneered the use of tiny "DNA chips" to study tens of thousands of genes at one time. Their work reveals how our cells function at the deepest, most fundamental, level and their findings are helping to move medicine into a new era, when a wider range of drugs will become available and genetic tests will help doctors choose the best treatment for a particular patient.
The story of how DNA chips were developed and turned into a revolutionary tool for medical research begins with the birth of genomics-the study of life's blueprint, the genome.
High-Speed Gene Chase
For many years scientists have known that genes influence health and that our genes are encoded in DNA, the famous twisted double-strand of molecules that programs all our cells. But until recently, the tools were so slow and clunky that it was very hard to study more than a single gene at a time, and it could take decades to track down just one trouble-making gene.
It wasn't until the Human Genome Project (HGP) began in 1990 that researchers could begin hunting genes at high speed. New instruments made it possible to "read" a lot of DNA very quickly. Suddenly there was an avalanche of data about genes and where they lay within the huge stretches (about 3 billion "building blocks" long) of DNA. Best of all, a lot of this data was posted for free on the Internet. This fueled an explosion in genomic research, and the launch of many new technologies to support it.
Of all the new tools developed, DNA microarrays have caused the most excitement. They were invented in the mid 90s, when someone had the idea of lining up many genes (actually, little bits of genes) across a microscope slide. People call them "microarrays," "gene chips," or "DNA chips" because they contain an array of DNA, and because they look a bit like the silicon chips that run computers. But DNA chips are really miniature laboratories. Tens of thousands of reactions can take place on the surface of each chip. The bits of DNA on the slide act like magnets, sticking to any complementary strands of DNA that they come in contact with.
Microarrays have completely changed the way genes are studied.
Researchers can take a DNA sample from a patient's tumor, for example, add some bright fluorescent "chemical tags" to it, and drop it onto the chip. The genes on the chip will pull matching genes out of the sample. The tags light up the spots where a lot of strands have matched up, and researchers can see what genes were activated, or turned on, in the tumor sample. The chips can now hold so many genes, that someone can actually scan the entire human genome with just a couple of chips.
That's important, because different genes are turned on or off in cells throughout the body, depending on what that cell is doing. Scientists want to know, "What's the difference between a healthy cell, and one with cancer? What genes are turned off in the cancer cell, but turned on in the normal cell?" DNA microarrays offer us a "peek under the hood" of the cell. We can see all the genes that are activated when the cell starts to grow and divide, change, sicken, or die.
A Powerful, but Complicated Tool
It sounds easy. Just compare the DNA chip patterns in healthy cells to the pattern seen in a sick person's cells, and the bad genes should jump out at you.
Unfortunately, it's not simple to do this. Because there are 10s of thousands of genes on each chip, and researchers may need to do 100s or 1000s of these studies to get reliable results, they end up with a huge amount of information to process and analyze. The effect is bewildering unless you have good software and databases on hand.
Fortunately, groundbreaking research from groups like Dr. Friend's helped to turn the chips into reliable tools. At Rosetta Inpharmatics (now a subsidiary of Merck) , which he co-founded, Dr. Friend and his colleagues created some of the best software available to analyze data from DNA microarray experiments. They pioneered a sophisticated approach to doing these studies that has become a model for the rest of the industry.
Thanks to their contributions, the field of microarrays as a whole is advancing very quickly. Researchers can now more easily build upon each other's findings, and we will all reap the benefits of this research sooner.
Signs From the Cell
By using microarrays, researchers have started to tease out what they call "molecular signatures" from cells. These are distinct patterns reflecting which genes are activated, and at what relative level, in a particular cell. This is invaluable information. Scientists know that you can't understand everything that's going on in a cell just by looking at it under the microscope. For example, two patients can have cancer cells that look exactly alike, but one patient gets better when given a particular medicine, while the second patient gains nothing at all from the treatment. A look at the molecular signatures (or gene expression profiles) of those cells, might show that, on a genetic level, these cells actually are very different. Studies like this have already been published, and some researchers think gene expression studies will show us that there are more different types of cancer than previously thought.
"One of the biggest things that has happened in microarray research is the introduction of molecular profiling, or forecasting," says Dr. Friend. "Before, microarrays were mainly used to find genes. But a few years ago, we started using them to find these molecular signatures, which can be used to forecast, or predict, drug response or disease progression."
In a groundbreaking study published in a major scientific journal early this year, Dr. Friend's group and researchers at the Netherlands Cancer Institute used DNA chips to study the gene expression signatures of breast tumors from 117 women. They made a startling discovery that could make a big difference in how some women with breast cancer are treated.
Typically, breast cancer patients undergo some type of surgery and radiation therapy. That's usually enough to keep the disease in check, but doctors have learned that some women need follow-up drug therapy or the cancer will come back. Deciding which women need that additional drug treatment is difficult because the tumors can look very similar, but behave quite differently. As a result, doctors tend to be cautious and (routinely) prescribe the follow-up drugs. Unfortunately, like many cancer therapies, these drugs can have serious side effects. So, thanks to this cautious approach some additional lives are saved, but many women are unnecessarily living with debilitating drug side effects and the fear and anxiety of not knowing whether the tumor had more serious implications.
"When the drugs are so toxic, the decision not to treat a patient is just as important as the decision to treat," says Dr. Alan Sachs, Dr. Friend's associate at Merck Research Laboratories. The researchers wanted to know if gene expression signatures could help predict which patients need the extra therapy. This study used banked tumors from women that had been diagnosed and treated for cancer several years ago, so the researchers could check whether or not the women's tumors had turned out to be the aggressive kind.
The results were remarkable. Not only were the gene expression signatures useful, they were far more accurate than any of the other tests available now. Doctors and researchers around the world were greatly encouraged by this news. "The effect of better selection [of patients for therapy] is potentially staggering..." one researcher wrote in an editorial. He added that, "If findings such as those by Friend and colleagues can be generalized, we can likewise hope that cancer treatment will be vastly improved by better predicting the response of individual tumors to therapy." [Footnote. Caldas, C, and Aparicio, SAJ, "The molecular outlook," Nature, January 31, 2000; 484-5.]
Like any promising research, the study needs to be confirmed before it can lead to changes in medical practice.
It's unlikely that DNA microarrays will themselves be used as diagnostic tests. Other technologies are being developed that give similar answers, and are easier to use in a doctor's office. But microarrays are definitely one of the most popular technologies in all of genomic research, and they will definitely provide more promising leads like Dr. Friend's findings. "As a research tool, DNA microarrays are amazing," says Dr. Sachs. "They clearly have a bright future."
|
