(Page 2 of 2)

Uncharted Waters

Bacteria use a host of methods to foil antibiotics. The bacterium S. aureus produces an enzyme, ß-lactamase, which specifically degrades penicillin and its analogues through hydrolytic cleavage of the penicillin ß-lactam ring. This "target modification" mechanism is also used by S. aureus and other bacteria against different classes of antibiotics, such erythromycin and chloramphenicol. There are a host of other mechanisms bacteria employ to evade antibiotics, such as encoding enzymes that sequester antibiotics by binding to them, undergoing small point mutations in the molecular targets that lower a drug’s affinity, overproducing a drug’s substrate in the cell, and efflux pumps.

Efflux pumps and transporters are perhaps two of the most difficult structures to study, though, because—like all other membrane proteins—they are notoriously hard to solve. Less than one half of one percent of the structures contained in the Brookhaven National Laboratory Protein Data Bank are of integral membrane proteins, despite the fact that over a third of all proteins in the body are in the membrane.

The difficulty with solving membrane proteins begins with obtaining them. Producing enough protein to work with can be insurmountable. A crystallographer might need several milligrams of protein to start with, but since most channels and transporters constitute such a small percentage of the cellular composition, getting enough material to work with becomes a challenge. "You just can’t grow that many cells," says Chang, who adds that all the membrane structures solved before 1998 are from naturally abundant sources.

Even assuming success in producing sufficient quantities, the proteins must be purified in their native state, which entails purifying them in the presence of detergents so that their hydrophobic membrane-spanning region can be surrounded by the hydrophobic ends of detergent molecules in micelle-like formations. The conditions are highly sensitive to such variables as detergent type, concentration, pH, and ionic strength. Worse, these conditions are almost always unique, demanding a lengthy trial and error search of an unknown biochemical landscape for that exact novel solution in which the protein-detergent complexes will be soluble and stable. "You have to draw on your own experience," says Chang, "You have to have quite a lot of it, actually."

Then a whole separate set of novel conditions must be worked out for growing the crystals, described by Chang as, "crystallizing out of soap." Several tricks can be employed, such as using antibody fragments, cubic lipid phases, or non-detergent systems. But, says Chang, these techniques are highly system specific and may only work with one or two proteins. Crystallizing presents another large problem in that the costs of the detergents can easily add up to hundreds of thousands of dollars.

Finally, any crystals that are grown can be difficult to work with because of their solvent content. Most water-soluble protein crystals have a protein-solvent ratio as high as 3:4—almost as hard or as dense as salt crystals. Membrane protein crystals, however, have a much higher solvent content because of the added bulk of the detergent micelles. As much as 88 percent of the membrane protein crystal may be a mixture of solvent and detergent, which makes the crystals unusually fragile under an x-ray beam. Chang calls it "shooting through Jell-O."

Every part of the process must come together for the science to work. No protein, no crystals; no crystals, no structure. Chang’s solution follows Thomas Edison’s—try as many different targets and conditions as necessary, build on these experiences, and go with those that work. "It’s really still a new field," says Chang. "There’s almost no literature on how to crystallize membrane proteins."

Once he has crystals, Chang will then make data collection trips to synchrotrons at the Stanford Synchrotron Radiation Laboratory and the Advanced Light Source at UC Berkeley. Synchrotrons are radiation sources that use x-ray radiation produced by electrons moving close to the speed of light in particle accelerator storage rings. These x-rays are intense, collimated, and tunable to various wavelengths, which make synchrotrons particularly useful for examining protein crystals. Chang, for example, will bombard his frozen membrane protein crystals with synchrotron x-rays and collect diffraction patterns for a variety of crystals. He will spend months analyzing and refining this data, and eventually—hopefully—solve a novel structure of a membrane complex involved in antibiotic and chemotherapy drug resistance.

Chang Honored at White House

Chang came to TSRI after spending three years as a postdoctoral fellow in Douglas Rees’s laboratory at the California Institute of Technology. In October, just over a year after he arrived, Chang was named one of the recipients of the Fifth Annual Presidential Early Career Awards for Scientists and Engineers, the highest honor bestowed by the United States government on young professionals at the outset of their careers. He spent a day in Washington, D.C., touring the National Institutes of Health and standing for a presentation ceremony at the White House.

Chang takes his award out of a file in his desk drawer and shows it to me—a simple design with nice lettering and William Jefferson Clinton’s bold signature across the bottom. It is still in the presidential blue folder it came in, and after a few moments, he flips the cover closed and slips it back in the file drawer. "I haven’t had time to get it framed," he explains.

1 | 2 |

Protein crystals, such as those of insulin pictured here, are notoriously difficult to make. "You have to draw on your own experience," says Chang. "You have to have quite a lot of it, actually." (Photo courtesy of NASA).