Scripps Florida Scientists Develop New Way to Speed Research on Estrogen Receptors

By Eric Sauter

Crystallography makes the invisible visible. By bouncing x-rays off crystals, scientists can determine the atomic architecture of complex molecules like proteins.

Crystallography starts with protein crystallization, an often-difficult first step that makes the entire process reminiscent of Steve Martin's advice on how to become a millionaire and never pay taxes: First, get a million dollars….

Creating crystallized versions of estrogen receptors is complicated and slow paced, resulting in just a few crystal structures a year.

Now, scientists at Scripps Florida, a division of The Scripps Research Institute, have developed a new way to accelerate the crystallization of steroid receptors.

In the process, they managed to uncover a unique structural conformation of an estrogen receptor that allows for selective suppression of inflammatory gene expression, a discovery that may one day prove important in the discovery and development of more effective, less side-effect-prone, cancer treatments.

The research, led by Kendall Nettles, a Scripps Florida scientist and assistant professor of cancer biology, was published in the April 4, 2008 (Volume 4, Number 4) of the journal Nature Chemical Biology.

Parallel Crystallization

The new process, which Nettles calls "parallel crystallization," allows scientists to test multiple compounds against the receptor proteins to see which ones produce crystals. With parallel crystallization, scientists can also simultaneously crystallize whole classes of compounds and identify subtle structural features that might not be apparent with individual structures.

"With conventional methods, you can only test one crystallization compound at a time because the protein—specifically, the ligand-binding domain on the estrogen receptor—misfolds when you try to produce the protein itself," said Nettles. "It's extremely low throughput."

Nettles described the new crystallization process as similar to getting salt crystals from seawater: "If you had a glass of salt water and let it evaporate you would get salt crystals. Basically, this is what we do with receptor proteins—except that we now can test hundreds of different compounds simultaneously. We think it will be easily transferable to other types of molecules."

Nettles, who joined the faculty of Scripps Florida in 2005, is focused on small molecule ligands (atoms or molecules that join with other molecules and result in a specific biological action) and how they control specific physiological outcomes through their interaction with nuclear receptors like estrogen receptors.

"Members of the nuclear receptor superfamily make excellent drug targets because they can be regulated by small molecule ligands, including steroids," he said. "But the drugs that specifically target the estrogen receptor in patients with breast cancer have some undesirable side effects in other tissues. For example, tamoxifen therapy increases the risk of a patient developing uterine cancer."

Estrogen is a natural hormone involved in breast cancer and osteoporosis. Estrogen receptor ligands (also known as selective estrogen receptor modulators or SERMs), in addition to treating those two diseases, are sought for the treatment of a variety of inflammatory and neurological conditions. Targeting estrogen receptor ligands can suppress what is known as the nuclear factor-kappa B (NF-kB) inflammatory cascade; NF-kB is a transcription factor found in mammalian cells that helps organisms respond to stress and infection.

"Steroid hormones like estrogen are anti-inflammatory," Nettles said. "For example, when estrogen levels rise during pregnancy, symptoms of inflammatory conditions like arthritis subside. Using our new method, we were able to identify NFkB selective ligand conformations that are selectively anti-inflammatory—that is, they don't activate normal genetic pathways for growth in breast and uterine tissue. That's very important when you're dealing with estrogen-driven cancers.""

Helix Changes

Such steroid receptor ligand binding domains have proven especially difficult to crystallize, partly because of misfolding and partly because of ligand binding domain instability, particularly in the most pharmacologically interesting ones. That instability makes crystallization exceedingly difficult.

The new approach devised by Nettles, and his colleagues at Scripps Florida and around the country, involves a relatively subtle change in the ligand binding domain structure, which in turn stabilizes the receptor.

The structure of receptor ligand binding domains is the same across the entire nuclear receptor superfamily. It consists of about 250 amino acids that form into twelve helical shapes, which fold into three layers of helices. That distinctive structure is subject to dramatically different shape changes, particularly in helix 12. Helix 12 covers the ligand binding cavity, plays a role in the binding process when exposed to agonists (activators) but gets shoved aside when exposed to antagonists.

"Helix 12 is the part of the protein that acts as a molecular switch," Nettles said. "When one compound binds to it, antagonists, for instance, the shape of the protein changes so that it does certain things—the binding process alters the shape of the protein. This shape change alters the receptors ability to turn genes on and off in the nucleus of the cell, ultimately controlling estrogen mediated cellular processes. "

One of Nettles' collaborators from the University of Illinois first identified a particular mutation in helix 12 which stabilized the domain in the active conformation.

"We looked at that structure, and decided to search for mutations that would stabilize helix 12 in an inactive conformation," he said. "We did that by adding a hydrogen bond to the protein surface, basically using a different amino acid that adds a charge to it. Our study shows that these crystallization challenges, which are common to steroid hormone receptors, can be circumvented by mutations that stabilize well-known receptor conformations."

Other authors of the study, Nfkb Selectivity of Estrogen Receptor Ligands Revealed By Comparative Crystallographic Analyses, include John B Bruning, German Gil and Jason Nowak of The Scripps Research Institute Florida; Sanjay K Sharma, Johnnie B Hahm and Geoffrey L Greene of the University of Chicago; Kristen Kulp of Lawrence Livermore National Laboratory; Richard B Hochberg of Yale University School of Medicine; Haibing Zhou, John A Katzenellenbogen and Benita S Katzenellenbogen of the University of Illinois; Younchang Kim and Andrzej Joachimiak of the Argonne National Laboratory.

The study was supported by the National Institutes of Health, the Ludwig Fund for Cancer Research, and the U.S. Department of Defense.

Send comments to: mikaono[at]scripps.edu

"Using our new method, we were able to identify NFkB selective ligand conformations that are selectively anti-inflammatory—that is, they don't activate normal genetic pathways for growth in breast and uterine tissue."

—Kendall Nettles