Tools For Protein Folding

If proteins don’t fold properly, they don’t work properly. Problems with protein folding underlie conditions such as Alzheimer’s disease and type 2 diabetes. So researchers are always trying to refine the methods they use to study the phenomenon. Two groups have recently done just that.

To study protein folding, it is often important to control whether a protein is folded. Adding denaturing agents and adjusting the temperature, pressure, or pH are easy ways to do this, but these approaches can produce undesired changes in the protein, so scientists have come up with other ways to manipulate folding.

One way of controlling protein folding without fussing with solution conditions is to use mutually exclusive proteins. Stewart N. Loh of the State University of New York Upstate Medical University, in Syracuse, and coworkers originally developed such pairings as biosensors, but they can also be used to study protein folding.

These constructs consist of a guest protein inserted in the middle of a host protein. The proteins are selected so the distance between the ends of the folded guest protein is three to four times longer than the corresponding distance in the host protein. This size difference imposes a topological constraint so that only one of the proteins can fold at a time. If the stabilities of the two proteins are similar enough, small changes in conditions can tip the balance and switch which protein ends up folded.

Using stopped-flow fluorescence measurements, Hongbin Li and Qing Peng of the University of British Columbia have now captured direct evidence of this tug-of-war (J. Am. Chem. Soc. 2009, 131, 13347). They use a construct with a protein called GB1-L5 as the host and a mutant of an immunoglobulin domain from the muscle protein titin as the guest. The researchers engineer a fluorescent tryptophan residue into the host protein, and thus the fluorescence measurements reflect the folding state of the host.

Which protein prevails in the power struggle depends on the relative thermodynamic stability, the kinetic folding rate, and the mechanical driving force of the two proteins. For Li and Peng’s particular pair, the host protein has the kinetic advantage and folds first, but then the guest protein, which is thermodynamically more stable, starts to fold, and the resulting mechanical strain triggers the unfolding of the host protein. Because the fluorescence does not return to its original level, they believe that the system reaches an equilibrium in which both conformations—folded guest with unfolded host, and unfolded guest with folded host—are present.

Li and Peng verified the coexistence of multiple populations using single-molecule atomic force microscopy. Each conformation has a signature response when it is pulled with an AFM tip. In addition to the populations of folded and unfolded proteins they expected to see, Li and Peng also observed rare instances where both proteins were folded at the same time. “It’s possible that both conformations could coexist, but they would most likely be in a highly strained state,” Li says.

“We can introduce other factors to affect the tug-of-war consequence,” Li says. “For example, the host protein we’re using binds an immunoglobulin-G antibody with very high affinity. Preliminary data show that IgG binding will stabilize the host,” and that makes the tug-of-war favor that protein.

Mutually exclusive proteins provide one system for manipulating protein folding, but the geometric requirements limit their broad applicability. A more general and flexible way to control protein folding would allow researchers to study a wider variety of proteins.

James U. Bowie and coworkers at the University of California, Los Angeles, achieve such a general method with “steric traps.” Instead of using insertion proteins, the researchers use the high-affinity binding pair biotin and streptavidin to control the folding of individual proteins (J. Am. Chem. Soc. 2009, 131, 13914).

They label a target protein by attaching biotin molecules to two cysteines that are close together in space and easily accessible when the protein folds. Streptavidin can bind to one of the biotins without steric hindrance, but a second streptavidin can bind to the other biotin only if the target protein unfolds. The binding affinity of the second streptavidin is a measure of the target protein’s stability—it binds more tightly to proteins whose folded form is less stable.

Because the biotin-streptavidin interaction is so strong that even stable proteins unfold, Bowie and coworkers adjust the strength of the steric trap by tinkering with the streptavidin binding affinity. “The problem with streptavidin is that it binds so tightly that the off-rate is very slow,” Bowie says. “If we use mutants, we can make the off-rate quite reasonable. Or, by using wild-type streptavidin, we can essentially trap the protein unfolded and keep it there for long periods of time. The method has that flexibility.”

For the method to work, the target protein must have two cysteines that are close to one another when the protein folds. This requirement doesn’t limit the proteins the researchers can study because they can engineer cysteines into a protein wherever they want them. This level of control means that they can even target a specific domain within a multidomain protein, whereas mutually exclusive proteins are typically restricted to single-domain proteins.

Bowie plans to use the steric trap to study membrane proteins. “We’re really hoping to use this to drive unfolding of membrane proteins so we can begin to study the unfolded state of proteins in natural bilayers,” he says. The results are promising so far. “In a simple model system,” Bowie says, “it definitely works in detergent, and it appears to be working in lipid bilayers.”