Insulin Study Sheds Light On Physics Of Crystal Growth; University Of Houston Analysis May Impact Medicine, Computer Chip Research

In a new study focusing on the physics of crystal growth in the protein insulin, University of Houston chemical engineer Peter Vekilov and his colleagues have determined why insulin does not form a certain defect, called step bunching, as most other substances do as they crystallize. The paper was published electronically in the June 6, 2003, online edition of the journal Physical Review Letters (Phys. Rev. Lett. 90, 225503 (2003)).

Typically, crystals are grown from molecules of a substance that are in a liquid solution. As the molecules in the solution assemble themselves to form a crystal, they build up in layers much like a brick mason lays down bricks. But crystals don't grow in neat, flat layers. Unlike a sensible bricklayer, nature starts new layers before the underlying one is finished, Vekilov says.

As crystals grow, new molecules add on to the edge of a layer, called a "step." Molecules tend to aggregate and pile up unevenly from one spot on the step, building up in clumps in a process called step bunching. This results in crystals that are generally undesirable for applications and research purposes.

Insulin, however, does not form step bunches as it crystallizes, which makes it a very interesting system to study, says Vekilov, an associate professor of chemical engineering at UH.

"Step bunching is generally bad," Vekilov says. "There are whole meetings dedicated to step bunching. It's a very major problem. In growing semiconductor crystals used to make computer chips, step bunching is always bad.

"Step bunching is one of the reasons we're still using silicon in computer chips," Vekilov says, noting that compound materials such as gallium arsenide offer advantages over silicon but crystals with the desired properties needed for next-generation computer chips are difficult to grow.

"Step bunching doesn't occur in insulin crystals, and in this paper we explored why in order to better understand the physics involved and the process in general," Vekilov says. Co-authors on the paper are UH chemical engineering research associates Olga Gliko and Ilya Reviakine.

Vekilov describes the way molecules aggregate in step bunches on crystals as "competing for supply." He has determined that in insulin, the steps, or edges, of a particular crystal layer do not "compete" in this way for molecules.

"What makes them compete for supply? It's a combination of the rate of transport of the molecules and the manner in which they get into the step," he says. "The molecules first land on the surface of the layer, and then get into the step. The mechanism is very complex for this competition, but in the case of insulin, all these factors combine in such a way that the steps do not compete for supply. The second most important thing is that the steps are generated almost regularly, they do not generate in bursts."

In other words, the insulin molecules tend to fall into place in a fairly regular, orderly manner, resulting in relatively level steps.

While Vekilov's research has focused on trying to limit step bunching in protein crystals, he says understanding the physics of how step bunching occurs - or doesn't occur in the case of insulin - can help scientists by increasing their knowledge of crystal systems and improving crystal-growing methods. For example, he is investigating the crystallization of mutant hemoglobin - an oxygen-carrying protein in the blood - that underlies sickle cell anemia. Vekilov's hemoglobin crystals were recently featured on the cover of the journal Science.

Step bunching can lead to defects in crystals that can cause adverse consequences if those crystals then make their way into lasers. As the step bunches pile higher, they also pile up faster because they are reaching areas of greater concentration of molecules in the solution. Really tall bunches bend over and close in front of the step bunch, Vekilov says, trapping a droplet of solution within the crystal.

Such crystals are the bane of researchers like those at the National Ignition Facility at Lawrence Livermore National Laboratory who have devoted a great deal of time and effort to perfecting their crystal-growing abilities. They need large crystals of potassium dihydrogen phosphate, or KDP, to incorporate into each of the 192 lasers that make up the world's most powerful system of laser beams. The facility is used for nuclear fusion and nuclear weapons research.

Vekilov, who also has worked with KDP crystals in his lab, has investigated ways to control step bunching in this substance.

"At Lawrence Livermore they shine all 192 lasers on a little target to study the process of nuclear fusion, and to ensure the safety of the country's nuclear weapons stockpile," he says. "Each of those lasers needs KDP crystals. If there is a little drop of solution in a crystal, and you shine these powerful lasers through the crystal, it will overheat at the droplet and it will crack. You then have to turn everything off and stop all work in order to replace that one crystal. You don't need to completely eliminate step bunching, but you need to keep it small" in order to not form droplets.

Growing good crystals of proteins, such as insulin, is a difficult and time consuming process, and is the first step in determining the structure of important biological molecules, says Dr. Kurt Krause, a medical doctor and UH professor of biochemistry who studies infectious diseases.

"We're looking for some of the same things in our tuberculosis protein crystals as a jeweler is in his gemstones - clarity and color, among other things," he says.

In a process called X-ray crystallography, a beam of X-rays hits a crystallized molecule and then bends, or refracts, in a certain way as it travels through the crystal, creating a pattern as it exits. From those patterns, scientists can then determine the 3-D structure of complex biological molecules, allowing researchers like Vekilov and Krause to better understand the function of the molecules and ultimately leading better drugs and treatments for disease.