RESEARCH

Research Interests: GTPases of the Ras superfamily Protein:Protein Interactions Protein X-ray Crystallography Solvent Mapping of Protein Surfaces Structure Based Ligand Design

Research in the Mattos Lab focuses on understanding the rules that govern the recognition and assembly of macromolecular complexes in the cell. It is clear that macromolecular interactions are central to the proper functioning, regulation and specificity of any cellular process, for example, signaling, transport, and replication. We are particularly interested in the protein-protein interactions that allow these assemblies to form in a specific manner. We are also interested in how small ligands are able to mediate or interfere with these interactions.

The biological system studied in this lab involves a group of closely related members of the Ras superfamily of GTPases. The protein-protein interactions mediating signal transduction pathways in which these GTPases are involved result in diverse and highly spefic biological outcomes, despite the fact that they have extremely similar protein architecture and funtion through a common mechanism of action. This is a system in which multiple protein-protein and protein-ligand interactions are at the heart of important signal transduction pathways within the cell.

The tools used to study the principles underlying specific macromolecular recognition include protein crystallography and computational biophysics. Molecular biology is used to engineer relevant constructs of the interacting proteins of interest. At the same time, we strive to give meaning to our results in the context of the cellular environment. One of the novel approaches used in this lab is the multiple solvent crystal structures (MSCS) method , developed initially using elastase as a model system.

Ras Superfamily of GTPases

GTPases are key proteins in many critical biological processes. They encompass a large group of enzymes that bind GTP and undergo a conformational change as the GTP is hydrolyzed to GDP in the presence of a bound Mg2+ ion. The hydrolysis of GTP and release of GDP are precisely regulated by other proteins, so that the GTPases can be thought of as molecular clocks for the timing and specificity of events that take place within the cell. The conformational switch mechanism involves the cycling of the GTPase between the active GTP-bound and the inactive GDP-bound states. The intrinsic GTPase activity of these enzymes is very low and the release of GDP is slow once hydrolysis occurs. The concentration of GTP in the cell is much greater than that of GDP, so that once GDP is released GTP takes its place and a new cycle begins. The relative inertness of the GTPases is a key element allowing the duration of the active GTP-bound form to be precisely controlled. This control is exerted by protein-protein interactions. Upstream signals in the cascade promote the binding of guanine nucleotide exchange factor (GEF) to its target site on the GTPase, facilitating the release of GDP. A GTP molecule from the cytoplasm then binds to the GTPase, resulting in the active conformation which can interact with a downstream effector, mainly through what is known as the effector region on the GTPase. The GTPase remains in the active GTP-bound state until it hydrolyses GTP to GDP. This deactivating mechanism is enhanced when the GTPase interacts with its specific GTPase activating protein (GAP).

There are five general groups of GTPases: the heterotrimeric G-proteins (involved in hormonal and sensory signals), the initiation and elongation factors (involved in ribosomal protein synthesis), the SRP/SR family (which translocate peptides into the ER), the tubulins and cytoskeletal motor GTPases, and the large group of monomeric GTPases known as the Ras superfamily (involved in signal transduction cascades and motility). Through the GTPases, nature has taken advantage of a highly effective mechanism for controlling the precise location, timing and specificity of signals to regulate a wide variety of cellular functions. The focus in this lab is on the Ras superfamily. The particular GTPases that are currently being studied in our lab are Ral, Ras, Rap, and Rin. We are pursuing the crystal structures of Ral and its complexes with downstream targets. We have also begun the solvent mapping experiments, using the MSCS method, on Ras and Ral.

The catalytic domain of Ras consists of a six stranded beta-sheets (b1-b6), five alpha-helices (a1-a5) and ten connecting loops (L1-L10), spanning 166 amino acid residues. There are two switch regions that undergo conformational changes upon GTP hydrolysis. Switch I includes the L2 residues 30-38 and is the effector binding region through which GAP interacts to activate GTP hydrolysis. The Switch II region consists of L4 and part of a2, spanning residues 60-72. The guanine nucleotide is in a pocket composed primarily of L1, L2, L4, L8 and L10. L1, L8 and L10 have the same structure and interactions with the nucleotide in the GTP and GDP-bound forms. The switch regions interact directly with the gamma-phosphate of GTP and are considerably altered upon loss of those interactions, giving rise to the conformational change that turns "off" the molecular switch.

The Multiple Solvent Crystal Structures Method

A flexible experimental approach has been developed to map the binding surface of any crystalline macromolecule. Crystals of the protein grown in aqueous mother liquor are crosslinked with gluteraldehyde and transferred to a variety of organic solvents, whose molecules are representative of functional groups that might appear in a larger ligand. The structure of the protein is solved in each of the solvents separately, revealing the binding pockets capable of accommodating each type of molecule. The sites where the solvent molecules bind to the protein are thus identified directly. All the structures are then compared based on a protein backbone least squares superposition.

A ligand binding site has the characteristic of binding almost any type of organic molecule or functional group in a small area defined by some degree of exposure of hydrophobic residues. For instance, the crystal structure of elastase was solved in seven different organic solvents. Molecules of all of these solvents appear in the S1 subsite. In general, a series of structures in about six different organic solvents suffices to locate the major binding regions on the protein surface unambiguously. The chemical properties and orientation of the small organic molecules within the site can be used to extract information about specificity. The different subsites occupied by organic solvent molecules within the vicinity of the natural binding site can then be targeted with characteristic functional groups belonging to a larger inhibitor, providing increased specificity for the desired target.

The power of the MSCS method goes beyond the simple location of binding sites. The work on elastase has delineated important components particular to binding sites, such as plasticity, apolar interactions within the binding pocket and the role of displaceable water molecules. It has resulted in a detailed analysis and classification of the way water molecules interact on the surface of proteins, surving a variety of different functional roles. We are currently obtaining the multiple solvent crystal structures of Ras.

Structures Solved by the Mattos Lab

Ras

Room temperature crystal structures of crosslinked H-Ras bound to GMPPNP were solved in 50% 2,2,2-trifluoroethanol, 60% 1,6-hexanediol, and 50% isopropanol. The disordered switch II region of Ras is ordered in the presence of 2,2,2-trifluoroethanol or 1,6-hexanediol. The overall backbone conformation of switch II in these organic solvents is the same as in the Ras-GMPPNP complexes with RalGDS, PI3 kinase, and RasGAP, indicating a biologically relevant form. Key polar interactions that stabilize the ordered switch are enhanced in the presence of hydrophobic cosolvents. These results suggest that hydrophobic solvents can be used in general to order short biologically relevant segments of disordered regions in protein crystals by favoring H-bonding interactions between atoms that are highly solvated and mobile in aqueous solution.

Buhrman G, de Serrano V, Mattos C., "Organic Solvents Order the Dynamic Switch II in Ras Crystals.", Structure (Camb)., 11(7):747-751 (2003).

Ral

RalA is a GTPase with effectors such as Sec5 and Exo84 in the exocyst complex and RalBP1, a GAP for Rho proteins. We report the crystal structures of Ral-GppNHp and Ral-GDP. Disordered switch I and switch II, located away from crystal contacts, are observed in one of the molecules in the asymmetric unit of the Ral-GppNHp structure. In the other molecule in the asymmetric unit, a second Mg²⁺ ion is bound to the GppNHp small gamma, Greek-phosphate in an environment in which switch I is pulled away from the nucleotide and switch II is found in a tight small beta, Greek turn. Clustering of conserved residues on the surface of Ral-GppNHp identifies two putative sites for protein-protein interaction. One site is adjacent to switch I. The other is modulated by switch II and is obstructed in Ral-GDP.

Nicely, N., Kosak, J., de Serrano, V., and Mattos, C., "Crystal structures of Ral-GppNHp and Ral-GDP reveal two binding sites that are also present in Ras and Rap", Structure, 12(11):2025-2036 (2004).

Cdc25B phosphatase

Cdc25B phosphatase, an important regulator of the cell cycle, forms an intramolecular disulfide bond in response to oxidation leading to reversible inactivation of phosphatase activity. We have obtained a crystallographic time course revealing the structural rearrangements that occur in the P-loop as the enzyme goes from its apo state, through the sulfenic (Cys-SO^-) intermediate, to the stable disulfide. We have also obtained the structures of the irreversibly oxidized sulfinic (Cys-SO2^-) and sulfonic (Cys-SO3^-) Cdc25B. The active site P-loop is found in three conformations. In the apoenzyme, the P-loop is in the active conformation. In the sulfenic intermediate, the P-loop partially obstructs the active site cysteine, poised to undergo the conformational changes that accompany disulfide bond formation. In the disulfide form, the P-loop is closed over the active site cysteine, resulting in an enzyme that is unable to bind substrate. The structural changes that occur in the sulfenic intermediate of Cdc25B are distinctly different from those seen in protein tyrosine phosphatase 1B where a five-membered sulfenyl amide ring is generated as the stable end product. This work elucidates the mechanism by which chemistry and structure are coupled in the regulation of Cdc25B by reactive oxygen species.

Buhrman, G., Parker, G., Sohn, J., Rudolph, J. and Mattos, C., "Structural mechanism of oxidative regulation of the phosphatase Cdc25B via an intramolecular disulfide bond", Biochemistry, 44:5307-5316 (2005).