The National Institute for Computational Sciences

The Ties That Bind

Researchers Explore How Certain Proteins with Roles in HIV and Cancer Change Form and Latch Onto Cell Membranes

by Scott Gibson

The class of proteins that bind to membranes in the body is key to many of life’s vital functions. And not surprisingly, more than half of all modern medicinal drugs are directed at manipulating those proteins.

Getting a good grasp of the structure and function of membrane proteins has posed a persistent challenge to the field of structural biology. Yet, successes in the quest are essential to the advancement of biomedical research, including understanding the conformational, or structural, changes proteins undergo when they move from the cell fluid and anchor to the cell membranes.

Proteins can play either pernicious or positive roles in the dynamics of disease. Some proteins that anchor to cell membranes promote the development of HIV (human immunodeficiency virus), while some proteins thwart the growth of cancer, for example.

Researchers have found that combining experimental neutron-scattering methods with molecular simulations can tell them a lot about the conformational changes in proteins and how they bind to cell membranes. Hirsh Nanda of the National Institute of Standards and Technology (NIST) is leading a multi-institution research team engaged in exactly that endeavor.

Through the Extreme Science and Engineering Discovery Environment (XSEDE), Nanda’s team has received allocations of computing time on Kraken, one of the most powerful supercomputers in academia, along with other high-performance-computing (HPC) resources offered by the National Institute for Computational Sciences (NICS), which manages Kraken for the National Science Foundation.

“The HPC resources accelerate the ability to obtain molecular details and knowledge of the forces that govern protein assemblies on cell membranes,” Nanda says.

Investigating the Structure and Function of a Cancer Suppressor

Nanda explains that PTEN is the second most-mutated protein found in cancer cells, indicating its crucial role in preventing unchecked cell proliferation. Initially expressed in the cytosol, or intracellular fluid, PTEN functions at the cellular membrane. So, the subtle interactions of PTEN protein with the cell membrane determine whether the protein lays anchor and stays, or goes away; its absence can result in tumor growth, he adds.

PTEN protein’s role as a tumor suppressor has made it the subject of research aimed at understanding what happens when the protein undergoes conformational changes and what interactions occur at the interface with the cell membrane.

In earlier research the team established the reference structure of the PTEN protein bound to the lipid membrane — a hydrophobic, or water-repelling, portion of the cell. Their investigation brought to light important details regarding how the protein anchors itself by electrostatic means: that is, negatively charged lipids in the membrane were attracted to positively charged regions of the protein that had not previously been identified as important for lipid binding.

The center of the membrane is a hydrophobic region that separates from the aqueous rest of the cell, and it is this separation that actually forms the membrane.

“Besides electrostatic interactions, hydrophobic residues in the protein can actually dive down into that hydrophobic center of the membrane as another means of binding to the membrane,” he says. “Furthermore, the PTEN protein has an amino acid tail that is able to wrap around and block the face of the protein that binds the membrane. So, one idea is that this tail regulates whether the protein comes on or off the membrane by coming around and blocking the membrane-binding interface.”

If the PTEN protein can be kept from binding to the cell membrane by an amino acid tail, thus preventing the protein from performing its cancer-suppressor role, then the question arises as to what controls the tail.

The answer could be a chemical modification called phosphorylation, Nanda explains. Phosphorylation refers to the altering of a protein’s function and activity by the addition of a phosphate, an inorganic chemical.

“The cell has different enzymes that phosphorylate proteins, and some of those enzymes might phosphorylate the amino acid tail of PTEN, which could then as a result come down and block the membrane-binding face of the protein. We want to study to see if that’s what happens,” Nanda says.

The details of the research of Nanda’s team on PTEN protein were recently published in the Journal of Structural Biology (v180, pp394–408z) and featured as a research highlight in the Nature magazine publication Nature Chemical Biology (2013 v9 pp 9).

HIV-1 Gag Proteins: HIV Assemblers

Nanda’s research team is now investigating the interplay of molecular interactions that anchor what’s known as the HIV-1 Gag (group-specific antigen) protein to the cell membrane. The Gag protein is the product of the Gag gene, which expresses the basic physical infrastructure for the HIV virus.

“In an infected host cell, thousands of Gag proteins come together on the membrane surface to form an assembly site for new viruses,” Nanda says. “The Gag protein is also responsible for bringing the virus’ genetic material into the assembly.”

Nanda explains that the research team had previously realized that the structure of Gag protein before it associates with the membrane was dramatically different from its membrane-bound state. This, he adds, raised the question of how Gag undergoes large conformational changes that appear to be necessary for proper virus assembly.

HIV-Nef Protein: Agent for Evasion of the Immune System

Nanda’s team will also begin pilot simulations for the HIV-Nef (negative regulatory factor), a protein expressed by primate lentiviruses, which are characterized by long incubation periods.

Nanda explains that similar to HIV-Gag proteins, HIV-Nef protein has a compact structure when in the cytosol but may adopt an extended conformation on the membrane.

“The extending of the HIV-Nef protein on the cell membrane may be a trigger for HIV-Nef to interact with other host-cell proteins and realize its function of helping cells infected with HIV evade immune-system detection,” Nanda says. “What we’re not sure of now is if membrane-binding alone is sufficient for HIV-Nef to extend itself or if there are some other biochemical factors that might cause that extension.”

Neutron Scattering and Molecular Simulations

Nanda’s team uses a special neutron-scattering technique called reflectivity. In this technique neutrons reflect off of a thin film or surface at low grazing angles, much in the same way as light would. A thin film such as a membrane will reflect neutrons differently depending on whether proteins are bound to it or not. Neutrons that reflect off of the water, protein and lipid layers modulate, or adjust, the reflected intensity. Nanda says that careful analysis of the data reveals an envelope structure of protein on the membrane surface. A real cell membrane has many different proteins associated with it, making interpretation of the neutron results difficult, he adds. Therefore, a simplified engineered model cell membrane has been developed at NIST that is especially suited for neutron-reflectivity measurements.

“Using neutron-scattering techniques, we found that simultaneous interactions between the cell membrane and viral genetic material were necessary for correct Gag structure,” he says. “Now, using molecular simulations we hope to further understand the molecular detail of this conformational change. Such insights might provide new targets for therapeutics that stop viral assembly, an important stage of the viral lifecycle.”

Scaling Up Simulations

Dwayne John — HPC consultant from the NICS User Assistance group and co-principal investigator on Nanda’s research team — shared with Nanda the types of complex problems the Kraken supercomputer is able to solve, and Nanda became very interested in learning how to scale up his team’s simulations on the system.

“Kraken is a large cluster, offering more than 100,000 processor cores,” Nanda says. “To take advantage of this — even though a single simulation ran efficiently on only a few hundred cores — we set up multiple simulations in parallel, varying the initial protein configurations. This allowed for an enhanced sampling of our systems and enabled us to acquire meaningful data from our simulations.”

To rigorously interpret the neutron-scattering results, the researchers performed all-atom molecular dynamics simulations on Kraken using the CHARMM force fields and NAMD software package.

CHARMM (acronym for Chemistry at HARvard Macromolecular Mechanics) refers to the set of force fields — that is, form and parameters of mathematic functions — widely used in molecular dynamics, as well as the simulation and analysis program associated with the force fields. NAMD is a parallel molecular-dynamics code designed for high-performance simulation of large biomolecular systems.

A Clearer View of Protein Function

“Atomic simulations provide a detailed picture of the interactions between proteins and lipid membranes at a resolution beyond the neutron measurements,” Nanda says. “A close agreement between the simulations and the experimental data provided confidence in the computer-simulation models of protein–membrane structure and dynamics. For PTEN the simulations revealed the specific regions of the protein that are responsible for cell membrane anchoring. In addition, simulations showed that PTEN can self-regulate its own membrane binding through its own C-terminal tail (the end of an amino acid chain).”

Nanda explains that neutron scattering is capable of providing structural information of membrane proteins in their native lipid membrane environment. He adds, however, that these techniques resolve only an envelope structure of the protein and membrane. Using molecular simulations to augment these structural investigations provides in atomistic detail the interactions and mechanisms important for protein function.

“The combined use of neutron scattering and molecular simulations results in a synergy that could be advantageous to the entire community of researchers studying membrane-associated proteins,” Nanda says. “We believe that partnerships between institutions that provide neutron-scattering resources and those that provide HPC resources would streamline the combining of the techniques and have a significant impact on membrane–protein research.”


HIV-Gag Proteins:

Datta, S. A. K., Heinrich, F., Raghunandan, S., Krueger, S., Curtis, J. E., Rein, A., & Nanda, H. (2011). HIV-1 Gag Extension: Conformational Changes Require Simultaneous Interaction with Membrane and Nucleic Acid. Journal of Molecular Biology, 406(2), 205–214. doi:10.1016/j.jmb.2010.11.051

Nanda, H., Datta, S. A. K., Heinrich, F., Losche, M., Rein, A., Krueger, S., & Curtis, J. E. (2010). Electrostatic interactions and binding orientation of HIV-1 matrix studied by neutron reflectivity. Biophysical Journal, 99(8), 2516–2524. doi:10.1016/j.bpj.2010.07.062

HIV-Nef Protein:

Kent, M. S., Murton, J. K., Sasaki, D. Y., Satija, S., Akgun, B., Nanda, H., et al. (2010). Neutron reflectometry study of the conformation of HIV Nef bound to lipid membranes. Biophysical Journal, 99(6), 1940–1948. doi:10.1016/j.bpj.2010.07.016