Edward Leof, Ph.D., has stalked the TGF-β receptor for decades. Recently, three things happened that rapidly accelerated his basic research toward producing results that could help a group of people who have never before been offered therapy. They are people who have idiopathic pulmonary fibrosis (IPF)—a diagnosis that comes with the devastating news that they have between two and five years to live. The three fortuitous events were Dr. Leof's attendance at his first Mayo seminar, a love match that brought an Australian to Minnesota and the timely arrival of the first clinical research fellow in Dr. Leof's lab.

Developing the Yeast Model of TGF-β Signaling

The protein TGF-β, which stands for transforming growth factor beta, received its name in the late 70s when investigators were looking for assays that could mimic cancer growth. They discovered that TGF-β stimulated the growth of normal fibroblasts in soft agar and caused morphologic changes that make them look like cancer cells. However, fibroblasts are poor models for cancer because most cancers are of epithelial origin.

In the mid 80s, the Mayo lab in which Dr. Leof was a post-doctoral fellow, discovered a more robust assay by adding TGF-β to epithelial cells in culture. The cells first showed inhibited growth, then lost the TGF-β growth inhibitory response, and, finally, tended toward growth stimulation as they progressed toward being transformed into tumor cells.

"We then made two assumptions," says Dr. Leof. "First, there had to be something unique about the signaling pathways in epithelial vs. fibroblast cells that resulted in such a distinct growth response. And second, it had to be Smad-independent."

The Smad family of proteins is the primary identified mediator of TGF-β; action. When the receptor is activated, they are the first proteins to be stimulated. They then migrate to the nucleus and modulate gene transcription. Families of Smads have been found to be turned on in almost every cell type that responds to TGF-β.

"But since they're all being turned on, and we're getting totally different biologies, it couldn't be these guys," says Dr. Leof. "It had to be independent of them."

So Dr. Leof's challenge was to find out what it was about the difference between a fibroblast and an epithelial cell that changed TGF-β signaling. To do that, he needed a genetic system—finding it was the first of the three fortuitous events.

"I was sitting in the audience at the very first seminar that I went to after returning to Mayo in 1992," explains Dr. Leof. "The topic was estrogen-progesterone function and the speaker was studying it by putting estrogen or progesterone receptors in yeast. It didn't take much to recognize that trying yeast as a genetic system for exploring TGF-β signaling was worth a shot. We later found out that yeast don't have Smad proteins, so it's an ideal model for us because, if we could get a signal in yeast it would confirm TGF-β's independence from the Smads."

The modest little brewer's yeast (Saccharomyces cerevisiae) cell has a smaller number of genes but a similar number of proteins to humans. The proteins bind together and control the organism's functions in much the same way that they do in us. Geneticists like yeast because it can grow as a haploid, unlike mammalian cells, which are diploid—contain pairs of chromosomes, one from mom and one from dad.

For the next three years, Dr. Leof worked on developing the yeast model for TGF-β signaling. Armed with a supply of brewer's yeast, Dr. Leof's lab conducted experiments that showed that the yeast cells respond identically to mammalian cells.

"There are two major TGF-β receptors, Type I and Type II and you need both for normal signaling," explains Dr. Leof. "When we put both of them into yeast, we got TGF-β signaling. But we are not yeast geneticists and consequent experiments were disappointing—until Mark Wilkes started working with us."

Geography Matters

As luck would have it, lab technician Mark Wilkes, met his wife while they were students at a university in Wilkes's hometown in Australia. His then girlfriend, who was in a study abroad program, hailed from Minnesota—fortuitously in terms of this project because that's what brought Wilkes to Dr. Leof's lab. Wilkes, who was studying biochemistry and molecular biology had read about Dr. Leof's work and asked to volunteer in his lab while his girlfriend finished her degree at a Minnesota college.

Wilkes returned to Australia to finish university. Upon graduation, Oxford University in England accepted his application for a Ph.D., however, Wilkes decided to take a year off before beginning. That was five years ago. Ever since he has been Dr. Leof's star lab technician.

"I was unbelievably lucky to find him," says Dr. Leof. "He's as good as, or better than any post-doc in any lab. He's creative, he reads the literature, and he's technically adept. This project was a team effort but it would never have happened without him."

When Wilkes returned to the lab, he began working on the yeast project. When he hit upon the idea of growing them in liquid culture rather than on agar plates, he noticed an unusual clumping-type growth on those expressing the receptors.

"Being the good Aussie that he is, he went to the brewing industry literature and found out that this was a process called flocculation," says Dr. Leof. "Subsequently we found out that yeast geneticists had discovered an analogous phenotype on agar plates by depriving yeast of nitrogen and causing them to burrow into the agar—presumably in search of nitrogen. But Mark was able to induce this invasive phenotype under normal growth conditions with the TGF-β receptors."

The brewer's yeast genome was sequenced in 1996. It was the first organism whose genetic material is enclosed in a cell nucleus to be sequenced and yeast geneticists have since worked out pathways in yeast genes that are similar to those of mammals. So the next course of action was to knock out the genes individually and find the culprit.

Wilkes had a breakthrough when he knocked out the sterile 20 (STE20) gene and found no invasive growth. That indicated that STE20 was the involved gene in yeast, but did it mean that TGF-β stimulates the same way in mammalian cells?

STE20 is similar to a family of proteins in mammals called P21 activated kinase (PAK) proteins. The PAK family in mammalians, and STE20 in yeast, are involved in many processes downstream of the receptors. They mediate functions such as migration, cell motility, stress response, and cell-type specificity.

"With fibroblasts, we found that only PAK2 was turned on—and PAKs of any flavor had never been shown to be turned on before, so we had a new target to TGF-β," says Dr. Leof. "But what really got us going, was with epithelial cells, none of the PAKS were turned on—so that meant there was cell-type specificity."

A search through the literature in late 2001, next uncovered two relevant papers. One proposed that platelet-derived growth factor (PDGF), a protein that induces proliferation of lung fibroblasts, activates c-Abl. The other showed that PAK2 and c-Abl, a non-receptor tyrosine kinase that's known to be involved in chronic myelogenous leukemia (CML), interact and regulate each other's activity.

The pieces of the puzzle were beginning to fit together. If TGF-β activates PAK, and TGF-β, PAK and c-Abl all modulate cytoskeletal changes, does that mean that TGF-β activates Abl? It was the job of a new player on the scene to answer this important question.

Making the Clinical Connection

Craig Daniels, M.D., is a pulmonologist who, in July 2001, became the first pulmonary fellow to undertake a research fellowship in Dr. Leof's lab. His timing was impeccable.

"Ed was interested in me doing something with a clinical tie," says Dr. Daniels. "In April, a new paper showed that c-Abl was always turned on and, therefore, propagated diseases like CML (chronic myelogenous leukemia). The next month, a drug called Gleevec® (imatinib mesylate) made the cover of TIME magazine. They positioned it as a new bullet in the war against cancer, especially for CML, because it turns c-Abl off."

Dr. Daniels arrived in the lab on the cusp of the new revelations and, with Wilkes's help, dove into experiments that promised to answer the question, did TGF-β activate Abl?

"Mark taught me everything about laboratory research—how to pipette, how to do cell culture, how to run a gel—and within six months we had some fairly good results," says Dr. Daniels. "Our experiments showed that TGF-β begins activating c-Abl in 15 minutes, peaks its effect in 30 minutes and starts coming down after an hour."

So the answer was yes, TGF-β activates Abl. The next step was to add imatinib to the cells before adding TGF-β, to see if it inhibited the response. It did—by inhibiting the induction of c-Abl, it stopped gene expression and, importantly, stopped the structural transformation—at least in cell culture.

This welcome news prompted two more questions. Is the activation of TGF-β by c-Abl the key to TGF-β-induced fibrosis? And can imatinib prevent fibrotic changes in mammals?

Fibrous tissue forms in every organ system in the body and there are many diseases that could be targets for this potential new treatment. Since the Leof lab resides under the pulmonary umbrella and pulmonary fibrosis is such a devastating disease, the lab geared up to perform animal studies using the bleomycin mouse model for idiopathic pulmonary fibrosis (IPF).

After initiation of IPF, fibroblasts migrate to the lung and TGF-β causes them to proliferate. Eventually the collagen builds up causing a honeycombed appearance under the microscope—a transformed structure that impairs the patient's lung capacity. The disease usually causes death between two and five years after diagnosis. There are two major growth factors involved in this process—PDGF (platelet-derived growth factor) and TGF-β. As a post-doctoral student, Dr. Leof discovered that TGF-β can also induce PDGF. If they were going to stop the fibrotic process, the lab would have to find something that would inhibit TGF-β.

"Imatinib is recognized as the biggest success story for intelligent drug design," explains Dr. Leof. "It was initially synthesized with the purpose of being a PDGF receptor antagonist to inhibit PDGF. They recently discovered that it also got c-Abl and one other, the C-kit receptor, but no other receptors at reasonable doses. Having a drug that can inhibit c-Abl and with such a specific effect really moved things along. "

Since animal studies were new to the Leof lab, they recruited Andrew Limper, M.D., to help set them up. By late 2002 they had the exciting data—imatinib inhibited collagen production in mammalian lungs. Dr. Limper then negotiated with Novartis to begin a clinical trial that would determine imatinib's effect on pulmonary fibrosis in humans.

Dr. Daniels is now the primary investigator of the Gleevec® clinical trial currently in progress at Mayo. The trial is being run in collaboration with Dr. Joseph Lasky at Tulane University. Mayo filled its initial goal of 25 patients on the Rochester campus and has been approved for another 10 patients.

"Our Arizona and Jacksonville campuses will each enroll five to eight patients so Mayo will have about 50 of the 100 who will be enrolled nationwide," says Dr. Daniels who is looking forward to participating in another multi-center clinical trial involving an anti-TGF beta antibody.

Downstream for the Leof Lab

With the clinical trial launched, the Leof lab has turned to other models of fibrotic disease and continues to explore new basic questions.

In collaboration with a nephrologist from the University of California at Los Angeles, the lab has shown that imatinib inhibits kidney fibrosis in a rat fibrosis model of uretal obstruction. And they recently began experiments in collaboration with ophthalmologist, Jose Pulido, M.D., looking at the potential relation between TGFb, eye fibrosis and glaucoma. Since glaucoma occurs in four percent of the population, this could be a significant study. In addition, Dr. Leof is continuing animal studies on IPF. The goal now is to combine drug therapies that work on distinct pathways in an attempt to lower drug dosage, increase efficacy and decrease toxicity problems.

"We also want to follow up with the Abl story," says Dr. Leof. "The question now becomes, what is it about epithelial cells that allows us to couple PAK and Abl in the fibroblasts but not in the epithelial cells? We want to understand more about what's regulating Abl—what elements in the receptor are turning it on and which proteins are associating with it. I would also like to study how these discoveries might lead to markers for diagnosis and a target for treatment for cancer."

Dr. Leof looks forward to increasing his collaboration with yeast geneticists, David Katzmann, Ph.D., and Bruce Horazdovsky, Ph.D. He hopes to use the yeast model to develop a drug screen for small molecule inhibitors of TGF action.

"I've always been excited by simply seeing how biology works," says Dr. Leof. "I've been writing grants to justify basic science investigations for more than two decades. But to initiate a fundamental question and see it move up the food chain to where it might actually help people with this terrible disease is unbelievably exciting."