Hunting for hydrogen
ORNL neutron and supercomputing facilities illuminate biomedical research
Given that approximately half of the atoms in the human body are hydrogen atoms, you might think that locating this abundant element would be easy. Not so, says ORNL's Paul Langan, director of the Biology and Soft Matter Division.
"Hydrogen is by far the most common element found in biological systems, but it is also the smallest element," Langan says. "It is very difficult to study because it is so light and mobile."
Hydrogen's elusive nature can be attributed partly to it having only a single electron, which helps form critically important hydrogen bonds. This electron is shuffled around with the hydrogen atom in almost all biochemical reactions. As a result hydrogen is practically invisible to techniques such as x-ray and electron scattering that rely on interactions with electrons.
"Most of the other atoms in biology like carbon, oxygen and phosphorus have tons of electrons, so they interact with x-rays easily," Langan says. "X-ray diffraction, an important technique in biology, can provide the skeleton or general shape of a protein, DNA or whatever biological molecule you're looking at. But the skeleton doesn't show you where the hydrogen atoms are. That's a big problem because those hydrogen atoms are involved in the chemistry that underpins the biology."
Researchers who want to study these biochemical reactions can now turn to neutrons, which, unlike x-rays, can more easily detect the presence of hydrogen. Neutron diffraction—a technique first pioneered at ORNL in the 1940s—is now helping biomedical researchers tackle today's research challenges in areas such as drug design and antibiotic resistance.
Neutrons to the rescue
Neutrons can be found in ample supply at ORNL, where they are produced at the Spallation Neutron Source and the High Flux Isotope Reactor. These two US Department of Energy user facilities welcome scientists from around the world to conduct studies in physics, chemistry, biology, engineering, materials science and other fields. As three new instruments dedicated to the study of crystals of molecules such as proteins come online over the course of the next year, Langan anticipates that ORNL's neutron facilities will become a new resource for the biomedical community, opening up new lines of research.
"We only have one beamline in North America that allows us to do this type of biological study," Langan says. "We're going to quadruple the capacity for neutron protein crystallography in the US over the next year." The term beamline refers to how neutrons are used once they are produced: Beams of neutrons are guided toward different instruments designed to study certain characteristics or properties of a sample. Once neutrons reach the sample, they bounce off in different directions, creating a pattern that researchers analyze to determine the sample's structure and other properties.
The three new instruments—named TOPAZ, IMAGINE and MaNDi—will offer scientists unprecedented tools for studying the structure of molecules, in particular large biological molecules such as proteins. For instance, improvements in instrument design will allow researchers to study very small samples—a major benefit to those who prepare protein crystals for use in neutron studies.
Growing protein crystal samples in the lab is an arduous and time-consuming process, and some proteins simply cannot form large crystals. The ability to analyze smaller crystals with neutrons will open up new avenues of research that were once off-limits.
In addition, the comprehensive suite of instruments at ORNL allows researchers to look at a problem from multiple angles: for example, studying the dynamics of a system in one experiment and then determining its structure at a different instrument. Neutrons are nondestructive, meaning that a single sample can be used multiple times because its properties are not altered by interactions with neutrons.
"Biological structures are continually moving and can also be incredibly complex," Langan says. "The capabilities we have at ORNL in neutron science create a virtual microscope, which enables researchers to conduct experiments at different length and time scales. Crystallography can provide atomic-level information, whereas beamlines for small angle scattering, reflectometry and imaging can provide information at larger length scales.
"Tying the results together using highperformance computing capabilities at Oak Ridge allows us to focus in and out on a particular structure or system. When we view these biological structures through our virtual microscope, they come to life when they are animated by information from dynamic neutron spectroscopy beamlines and computer modeling."
Although neutron studies can stand on their own merits, they work best in tandem with other techniques, such as x-ray analysis. In a recent study, Langan and his ORNL colleague Andrey Kovalevsky were part of a multi-institutional research team that used neutrons from the Institut Laue-Langevin in France to refine their understanding of an enzyme that had been studied with x-rays for 20 years. The team looked at the interactions between HIV protease, a protein produced by the HIV virus, and an antiviral drug commonly used to block the virus' replication.
"We found that the hydrogen atoms thought to enhance the binding of the drug are not actually responsible for its enhanced binding," Langan says. "When this drug interaction was examined using x-rays, they didn't show where the hydrogen atoms were, so assumptions were made about the presence of hydrogen bonds. We now know these assumptions are wrong. Our neutron structure provides us with extra information that will be used to retune or redesign that drug into being more effective.
"Our cells contain a complex network of biochemical reactions, with different enzymes facilitating different stages of those reactions. X-ray crystallography can show you the shape of those molecules and how they might bind to substrates or small molecules, but to understand the chemistry in biology, to understand how the catalytic mechanism occurs, you need neutrons."
Neutrons are also a necessity for John Katsaras, an ORNL researcher who holds a joint appointment with Canada's National Research Council. Katsaras uses both neutron and x-ray techniques to study cell membranes, the boundary layers around cells that are involved in numerous biological processes. Biologists have studied membranes for decades, but many mysteries remain about their complex structure and function.
"We have experiments that we would like to do to resolve research questions that have been ongoing for 40 years," Katsaras says. "Neutrons are almost uniquely positioned to do this."
One question that Katsaras and collaborators would like to tackle with neutron scattering is the existence of "lipid rafts" in living cells, a topic of much debate in membrane biology. Membranes consist of two layers of fatty molecules called lipids, which are interspersed with other molecules, mainly proteins.
Toward the end of the 1970s, researchers began to suspect that instead of forming an even pattern throughout the membrane, certain areas of the lipid bilayers would clump together and separate from the surrounding material. Scientists also began to understand that these segregated areas, or rafts, could play a role in a range of cellular functions, including drug uptake and interactions with pathogens. But the theorized rafts remained frustratingly invisible.
"We can study these rafts in carefully controlled model systems, but the question of their existence in living cells is like a biological black hole," Katsaras says. "It has been shown through biochemical means that rafts probably exist, but no one has ever seen them or characterized them.
"Why is that so? Is it because they don't exist, or is it the fact they're very small? Or is it the fact they're transient? They could be appearing and disappearing all over the place, but if you're only looking at one spot with optical techniques, you may not see them."
Neutron scattering can be used to probe a relatively large amount of material, allowing scientists to draw conclusions about the bulk material instead of a single location. Katsaras plans on using neutrons to detect lipid rafts in living cells by exploiting the biological contrast in the rafts and the surrounding areas.
"Everything in life is contrast," Katsaras says. "If your clothes and your face were the exact same color, then I couldn't tell where your face was. You need contrast. X-rays are relatively poor in contrast when it comes to biological materials. Neutrons, on the other hand, can be very good, especially when you induce contrast by exchanging hydrogen for deuterium."
Deuterium is an isotope of hydrogen that neutron scientists use as a labeling tool. By substituting deuterium for certain hydrogen atoms, researchers can more easily track the position of the labeled atoms because neutrons "see" the difference between the two isotopes as a black and white contrast.
Making supercomputers sweat
Katsaras is among a fleet of ORNL researchers developing ways to integrate the force of the lab's neutron facilities with its computing clout. ORNL has long held leadership roles in supercomputing and neutron science, but only in recent years have the two fields started to build upon the other's successes.
Powerful supercomputers including ORNL's Titan—ranked the world's second fastest as of June 2013—are capable of running simulations that require mindboggling amounts of calculations. But real-world data is still needed to keep the complex simulations based in reality, Katsaras says.
"Computational simulations need to be validated by experimental data," he says. "If there is no validation, because these systems are so complicated, it is very difficult to know if you're on the right track."
One of Katsaras's proposed projects would unite complementary powers of neutron scattering and simulation to study an entire vesicle, a molecule that mimics the composition of an asymmetric cell membrane. One such vesicle is about 50 nanometers in diameter, about 2,000 times smaller than the average diameter of a human hair.
"That may seem very small, but it would make Titan sweat," Katsaras says. "It's about 60 million atoms, whereas most people are working with systems that are a hundred thousand atoms. That is enormous."
The fine-grained simulation of a whole vesicle will help researchers better understand how cell membranes are structured and how they interact with other molecules within and outside the cell.
"Membranes' structure may dictate how drug action happens," Katsaras says. "This could lead to an understanding of how a drug interacts with a membrane or how it communicates. You need both experiment and simulation. The simulation, in this case, gives us the finer, atomistic detail—if we could trust it. Experiments give us a broader perspective and validation of the simulation. Basically, it's a system of checks and balances."
Through a National Institutes of Health consortium in partnership with Lawrence Berkeley National Laboratory, work is already under way at ORNL to develop the computational tools needed to understand data from neutron crystallography studies.
Langan also envisions creating a biomedical neutron technology research center at ORNL that would further integrate the lab's supercomputing and neutron capabilities, as well as offer training and assistance to biomedical researchers.
"ORNL's emerging capabilities in neutron science are providing the biomedical research community with unprecedented opportunities to solve problems once considered intractable," Langan says. "It's a truly exciting time." —Morgan McCorkle