Drilling Deeper

Researchers once thought they could crack the code on biodiversity if only the entire sequence of human DNA could be charted.

But in February, 13 years and more than $1 billion after beginning such an effort, the multinational human genome project wrapped up its final analysis with a surprising finding: People have relatively few genes. Too few, it seems, to explain why humans are that much different from insects or even plants.

"When the Human Genome Project was started, it was estimated that we would have somewhere between a quarter-million and a half-million genes in the human genome," explains Dr. Jon Klein, a UofL researcher and director of the university's Core Proteomics Laboratory.

"As we began to map the genome, we found that there were fewer and fewer genes-only 32,000 to 34,000. It's very, very hard for scientists, looking at only 34,000 genes with tremendous similarity across a variety of organisms, to fathom where our complexity originates. If the genome is the road map, it's a pretty sparse road map."

Genes, it turns out, explain only part of the mystery. The rest can be chalked up to proteins like insulin, hemoglobin and serotonin, which help carry out our bodies' most basic functions at the cellular level.

Each gene is responsible for the production of multiple proteins, but no one really knows for sure how many proteins the human body generates, Klein says. Estimates range from 500,000 to 2 million. And when those proteins start to go awry, disease sets in.

"The genome is a recipe for a human being, and sometimes it works very well," says Klein, a professor of internal medicine who holds Ph.D.s in microbiology and immunology.

"But there are a lot of things that can happen when you put that recipe to work. Sometimes you've prepared an outstanding dish. Sometimes you've prepared a terrible dish because of the way the recipe was interpreted.

"It's not necessarily easy to tell why the failures occur from looking at the structure of the genome. It will help -- there are many, many genetic diseases that we're discovering at a strong pace -- but there also are diseases where the genes appear normal and the proteins are grossly abnormal.

"Ultimately, it is the proteins that do the work in cells, so if we are going to study abnormalities in the human body, we need to understand the accumulation of abnormal proteins that cause the mechanisms of disease."

Pinning down these proteins can, however, be a bit challenging. For one thing, it's simply not possible to map all the proteins in the human body -- the human proteome -- because these proteins vary by tissue type and are in a constant state of flux.

"Let's imagine," Klein says, "that I could take a protein snapshot of a human heart, and it would give me a list of probably 10,000 proteins, as well as the level at which those proteins are expressed -- how much is there.

"If I then put that person on a track and have him run a lap, the next protein snapshot would be very, very different. So, proteins are dynamic, as opposed to genes, which are static. This means we're always trying to study a moving target. There is no one proteome that we can measure."

Other problems facing researchers are the technological limitations of protein analysis techniques, which currently are not sensitive enough to detect as much as 70 percent of the body's proteins.

"There are a lot of proteins that are abundant, and those tend to get in the way of the less abundant proteins," Klein says. "We may have, in theory, 10,000 proteins in any given cell type, but our current technology may see only 3,000 or 4,000 of them. So there are always more that we're not able to detect."

Klein says, however, that new technology soon will be able to "drill deeper" into the human proteome.

In the meantime, UofL's $1.3 million Core Proteomics Laboratory is continuing to advance the state of the art. The facility was the first of its kind on a college campus when it opened in the fall of 1997, and it remains one of the most sophisticated proteomics labs in the country.

Currently, the lab employs a process called two-dimensional gel electrophoresis to examine proteins. This process separates proteins from tissue samples based on the proteins' sizes and electrical charges and provides information about their molecular weights and physical characteristics.

The proteins also are stained, producing a pattern of spots much like a star map, so technicians can pick them out of the gel substrate for analysis on a mass spectrometer. The device, overseen by William Pierce of the university's Mass Spectrometry Core Laboratory, then is able to identify the protein -- if it has been cataloged.

"Once we're able to identify the protein," Klein notes, "we can quantify it under different conditions and tell whether the level of a protein has increased or decreased in a given subject. It's basically a comparative analysis between the tissue of a healthy person and a diseased person."

Klein cited as an example of the process one recent study on hypertension.

"When we examined the protein expression in the kidneys of a subject with hypertension, we found a big decrease in a vasodilator-a protein that causes relaxation of the blood vessel. That's the sort of thing proteomics can address-we get right to the levers that move the cell around and cause diseases."

One unique feature of UofL's Core Proteomics Lab is its extensive use of sophisticated robotics to process tissue, resulting in a three-fold productivity increase.

"Our high-throughput robotics gives us capabilities no one else has outside of a good-sized drug company," Klein says. "In fact, drug companies have on occasion come here to see whether they want to set up a lab like this. "

The drug companies are interested in proteomics for good reason, Klein notes.

"We hope proteomics will help us understand disease mechanisms a lot faster and shorten the time to develop new therapies. Most major pharmaceutical companies have put a lot of investment into proteomics laboratories on the promise that they will be able to find drug targets. They're looking for proteins that we can change or interrupt somehow to cure disease."

One alternative to drug regimens is protein therapy, which also stands to benefit considerably from proteomics research.

"With protein therapy, you introduce the finished product, or protein, directly into the patient," Klein explains. "Insulin is the oldest example of this therapy, but we hope to find many more."

Current proteomic research at UofL includes a $675,000 Veteran's Administration study to determine what causes the death of common white blood cells.

"When these cells don't die on schedule, they cause a lot of multi-organ damage," explains Klein, the study's principal investigator. "It's a big clinical problem because they just accumulate, and their byproducts are very toxic."

The lab also is collaborating with Dr. David Gozal, a UofL pediatrician who has been researching sleep apnea in children as part of a $1 million grant from the National Institutes of Health.

For Klein, who recently was named one of the most influential people in genomics by the readers of Genome Technology magazine, the opportunity to help researchers like Gozal makes his job "more fun than I ever could have imagined."

"It's incredibly rewarding to work every day with a new technology that people find useful," Klein adds. "A lot of people with really good ideas, who are studying important problems, want to collaborate with us. I feel very fortunate -- like a kid in a candy shop."