• Relative stabilities of conserved and non-conserved secondary structure in OB-fold proteins

Although many proteins fold in a cooperative two-state transition, other proteins are less cooperatively organized, making it possible to characterize partially folded intermediate states. These partially folded states may offer the best chance to understand protein-folding mechanisms. What accounts for the ability of some proteins to fold to incomplete structures under certain conditions? Our hypothesis is that these partially folded states reflect how the protein structures developed during evolution.

The OB-fold is a diverse structure superfamily based on a five-stranded beta-barrel motif that is often supplemented with additional non-conserved secondary structures. Previous deletion mutagenesis and NMR hydrogen exchange studies of three OB-fold proteins (SN, LysN, CspA) showed that structural stabilities were larger for sites within the conserved beta-barrels than sites in the non-conserved segments of these proteins. To find the physical basis of this behavior we examined a database of 80 representative domain structures currently classified as OB-folds in the SCOP taxonomy. Residue-specific values were obtained for the number of C-alpha to C-alpha distance contacts, sequence hydrophobicities, crystallographic B-factors, and theoretical B-factors calculated from a Gaussian Network Model. All four parameters point to a larger average flexibility for the non-conserved structures compared to the conserved beta-barrels. The theoretical B-factors and contact densities show the highest sensitivity. The figure in the adjacent panel shows the subset of 40 OB-fold domains for which X-ray structures are availale color-coded according to the theoretical B-factors. The color ramp runs from blue (small B-factor, rigid) to red (large B-factor, flexible). For most of the OB-fold structures residues in the beta-barrel are blue, while residues in the non-conserved accessory structures tend towards the red regions of the spectrum.

These results suggest a model of protein structure evolution in which novel structural features develop at the periphery of conserved motifs. Core residues are more resistant to structural changes during evolution since their substitution would disrupt a larger number of interactions. Similar factors probably account for the differences in stability to unfolding between conserved and non-conserved structures.

Guardino, K.M., Sheftic, S.R., Slattery, R.E., & Alexandrescu, A.T. (2009) “Hallmarks of conserved and non-conserved structures in the OB-fold superfamily”. Int. J. Mol. Sci. 10, 2412-2430. (open access, free paper - click here)

• Structure of micelle-bound amylin.

Amylin is a hormone made by the pancreas that regulates food metabolism.  Amongst its actions is to induce a feeling of fullness after meals, by binding to receptors in the neurons of the brain's hypothalamus.  In addition to it's normal role, amylin is important in the development of type 2 (adult-onset) diabetes - a disease that affects over 100 million people worldwide and costs more than $130 billion a year to treat in the US alone. In the majority (95%) of patients with type 2 diabetes, amylin aggregates into fibrillar amyloid deposits in the pancreas, The deposits form in the extracellular spaces between the beta-cells that make both amylin and insulin.  The fibrils, or probably smaller soluble aggregates destroy the beta-cells as type 2 diabetes progresses.  The toxic effects of amylin are thought to be exerted through membrane-bound aggregates that form holes in the beta-cell membranes leading to cell death.  In order to better understand these processes we used nuclear magnetic resonance (NMR) to determine the structure of amylin in a membrane-like environment.  In the presence of SDS micelles (a membrane mimic amenable to NMR studies) amylin folds into an alpha-helix structure.  The first 17 residues are partially inserted in the micelle, away from solvent.  The segment between amino acids 20-29 that has the strongest inclination to form aggregates is more mobile and is placed at the interface between the micelle and solvent (NMR structures on the left hand side).  Our structural model suggests that the positioning and flexibility of the amyloidogenic segment 20-29 promotes the aggregation of amylin on the surfaces of membranes.  Aggregation of the 20-29 segment could lead to a repositioning of the first 17 amino acids, from the surface to the interior of the membrane leading to an annular pore structure that is toxic to cells (hypothetical model on the right hand side).  Our studies of amylin are still at an early stage.  Future goals include to look at the hormone bound to more realistic models of cellular membranes and to structurally characterize the amylin aggregates responsible for cell toxicity.

Patil, S.M., Xu, S., Sheftic, S. & Alexandrescu, A. (2009) “Dynamic alpha-helix structure of micelle-bound human amylin”. J. Biol. Chem. 284, 11982-11991.

• Charges as modulators of protein structure stability.

Electrostatic interactions play important roles in protein folding, molecular recognition, and catalytic activity. A better understanding of these forces is important for developing improved potential energy functions, for use in applications such as protein structure prediction and molecular design We obtained a complete set of ionization constants for the GCN4p leucine zipper coiled coil from NMR pH tritrations. This is the first complete set of pKa values of a protein in both its native and denatured states, and serves as a benchmark for testing structure-based electrostatic calculations. In collaboration with the Garcia-Moreno lab at Johns Hopkins, continuum electrostatic calculations were done starting from the GCN4p X-ray structure, to test pKa values and their contributions to protein stability against data from NMR and equilibrium unfolding experiments. The structure-based calculations (dotted red curve) matched the experimentally determined stability profile (blue points), except at the extremes of pH (arrows) where the contributions of charge interactions to stability are greatly overestimated.  Part of the reason for the discrepancy may be that the structure or dynamics of the protein change at the extremes of pH and that current theoretical modeling of these changes are inadequate. We are currently using NMR to characterize the structural changes that occur in proteins with changing pH.

A second more recent interest is to look at the roles charges play in the conformational transitions of amyloidogenic proteins. We are investigating the electrostatic properties of amyloidogenic proteins in their monomeric, fibrillar, and micelle-bound states.

Matousek, W.M., Ciani, B., Fitch, C.A., Garcia-Moreno E., B., Kammerer, R.A. & Alexandrescu, A.T. (2007) "Electrostatic contributions to the stability of the GCN4 leucine zipper structure". J. Mol. Biol. 374, 206-219.

• Conserved folding of proteins that share a similar structure motif

The OB-fold (oligonucleotide/oligosaccharide-binding fold) occurs in more than 80 unrelated proteins. It consists of a 5-stranded beta barrel formed from two- and three-stranded beta-sheets. We have studied the folding of three OB-fold proteins that share no detectable sequence homology: SN (staphylococcal nuclease), LysN (anticodon binding domain of Lys-tRNA synthetase) and CspA (Cold shock protein A). Mutagenesis and hydrogen exchange experiments show that the structural motif conserved between the three proteins is more resistant to unfolding than structural elements that are not conserved between the proteins. This is shown in the figure on the right where the site-specific stabilities to unfolding events that expose amide hydrogens to exchange are mapped on the structures (blue is more stable, green is less stable, white is undetermined). In fact, misfolding of the three proteins also appears to be conserved. The first three strands of beta-sheet show the highest propensity for structure under denaturing conditions, and the acid denatured forms of all three proteins aggregate through the mispairing of this partially formed unfulfilled structure. We are currently exploring the structural determinants of conserved motifs in the OB-fold.

Alexandrescu, A. T., Gittis, A., Abeygunawardana, C., & Shortle, D. (1995) “NMR structure of a stable "OB-fold" sub-domain isolated from staphylococcal nuclease”. J. Mol. Biol. 250, 134-143.

Alexandrescu, A.T., Jaravine, V.A., Dames, S.A. & Lamour, F.P. (1999) "NMR hydrogen exchange of the OB-fold protein LysN as a function of denaturant: The most conserved elements of structure are the most stable to unfolding". J. Mol. Biol. 289, 1041-1054.

Jaravine, V.A., Rathgeb-Szabo, K., & Alexandrescu, A.T. (2000) "Microscopic stability of cold shock protein A examined by NMR native state hydrogen exchange as a function of urea and trimethylamine N-oxide". Protein Science 9, 290-301.

Alexandrescu, A.T., & Rathgeb-Szabo, K. (1999) "An NMR investigation of solution aggregation reactions preceding the misassembly of acid denatured cold shock protein A into fibrils". J. Mol. Biol. 291, 1191-1206.

Alexandrescu, A.T., Jaravine, V.A., & Lamour, F.P. (2000) "NMR evidence for progressive stabilization of native-like structure upon aggregation of acid denatured LysN". J. Mol. Biol. 295, 239-255.

Watson, E., Matousek, W.M., Irimies, E.L. & Alexandrescu, A.T. (2007) “Partially folded states of staphylococcal nuclease highlight the conserved structural hierarchy of OB-fold proteins”, Biochemistry 46, 9484-9494.

• Development of a method to study hydrogen exchange in amyloid fibrils

Amyloid fibrils are protein deposits involved in a number of neurodegenerative diseases such as Alzheimer's. Once thought by experts to be inert and impossible to study by solution NMR, we wanted to investigate the flexibility of fibrils and their interactions with solvent. We developed a method to look at exchange of amide hydrogens from the fibrils with solvent. Hydrogens exposed to the surface should exchange easily, those caught up in structure will be protected. In the method, amyloid fibrils are exposed to D2O, and exchange is quenched by flash freezing. The fibrils are solubilized and dissociated in the aprotic solvent DMSO, which enables the exchange history of the fibrils to be read out indirectly from the denatured state of the protein. This method has recently come into wide use; the publication from our lab preceded other reports by one year. In our original paper we proposed that exchange from fibrils of the E.coli protein CspA occurs predominantly through dissociation of protein from the fibrils. This hypothesis was initially met with much resistance, but similar mechanisms have recently been hailed "molecular recycling".

Alexandrescu, A.T. (2001) "An NMR-based quenched hydrogen exchange investigation of model amyloid fibrils formed by the protein CspA". Pac. Symp. Biocomput. 6, 67-78.

• Formation of individual hydrogen bonds during folding of an alpha-helix

A technique developed by Stefan Grzesiek to detect scalar coupling mediated through hydrogen bonds was used to follow alpha-helix formation in the S-peptide of ribonuclease A. The results show that hydrogen bonds are not uniformly populated as alpha-helical structure forms. The hydrogen-bonds follow a stability gradient that decreases from the center to the ends of the helix. The results are roughly consistent with predictions from the Lifson-Roig theory of coil-helix transitions.

Jaravine, V.A., Alexandrescu A. T., & Grzesiek, S. (2001) "Observation of the closing of individual hydrogen bonds during TFE-induced helix formation in a peptide". Protein Science 10, 943-950.

• NMR structure of a homotrimeric coiled coil.

We determined the first NMR structure of a homotrimer. Homooligomers pose a particular problem for NMR because the monomer chains are magnetically equivalent. Additional approaches are thus needed to distinguish NOEs within a chain from those between chains. To solve the NMR structure of the coiled coil trimer from matrilin-1 we first obtained a highly defined structure for the monomers. We then used a self-consistent strategy to assign NOEs between chains, which were identified from isotope-filtered NOE experiments.

Dames, S.A., Wiltscheck, R., Kammerer, R.A., Engel, J., & Alexandrescu, A.T. (1998) "NMR structure of a parallel homotrimeric coiled coil". Nature Struct. Biol. 5, 687-691.

• Residual dipolar couplings in denatured proteins

We showed how dipolar couplings can arise from localized anisotropic structure in denatured proteins, that dipolar couplings are biased towards the most anisotropic structures in the denatured state ensemble, and examined how dynamics and changes in the distributions of conformers in the denatured state ensemble affect couplings.

Alexandrescu, A.T., & Kammerer, R.A. (2003) “Structure and disorder in the ribonuclease S-peptide probed by NMR residual dipolar couplings”. Protein Sci. 12, 2132-2140.

Sallum, C.O., Martel, D.M., Fournier, R.S., Matousek, W.M & Alexandrescu, A.T. (2005) “Sensitivity of NMR residual dipolar couplings to perturbations in folded and unfolded staphylococcal nuclease”. Biochemistry 44, 6392-6403.

• Dynamics in denatured and partially folded states

15N relaxation measurements have been used to investigate dynamics in a number of equilibrium folding intermediates. The figure shows four forms of the enzyme staphylococcal nuclease with different levels of cooperatively stabilized structure. SN-T is the nuclease in a ternary complex with Ca2+ and the inhibitor pdTp, which stabilize the protein. SN is the wild type protein. SN-OB is a folded fragment of the nuclease that is missing 1/3 of the chain from the C-terminus (see below). D131D is a highly denatured fragment of nuclease with near-random-coil NMR spectra. Blue colors indicate low order parameters (related to the amplitude of motion) and rigid structure, red colors indicate large order parameters and flexibility. As the stability of the native state stability decreases, the main-chain is subject to increasingly larger amplitude motions and the dynamic properties of the chain become more heterogeneous. These results paint a picture of unfolding as a fragmentation (or shattering) of the native state structure, and of folding as an accretion of structure in which the mobility of the main chain is frozen-out and initially separate structural elements become increasingly interdependent in order to achieve the maximum cooperativity and stability of the native state.

Alexandrescu, A.T., Jahnke, W., Wiltscheck, R., & Blommers, M.J.J (1996) "Accretion of structure in staphylococcal nuclease: An 15N NMR relaxation study". J. Mol. Biol. 260, 570-587.

• NMR structure of a sub-domain from a cooperatively folded protein.

Staphylococcal nuclease folds with the typical all-or none cooperativity of a single domain protein. In the presence of the two rare global suppressor mutations V66L and G88V (analogous mutations occur for thermostable nucleases found in nature) a 1-103 fragment of nuclease remains folded in the absence of the last 1/3 (55 residues) of the sequence (purple segment). The portion of the structure that remains folded corresponds to the conserved OB-fold motif found in a variety of protein structures. The C-terminal 1/3 of the sequence is certainly not dispensable since it contains part of the enzyme's active site. The fragment has a 1000-fold lower nuclease activity than the wild type, but still classifies as an enzyme (reaction rate is 10^12 fold higher with the fragment). The fact that part of the polypeptide chain of this single-domain protein can be deleted challenges two-state models of protein folding. It also suggests mechanisms by which protein structures can evolve by assimilating pre-existing structural motifs; sacrificing the independence of sub-domains in order to achieve the maximum cooperativity and stability of the new integrated structure.

Alexandrescu, A. T., Gittis, A., Abeygunawardana, C., & Shortle, D. (1995) “NMR structure of a stable "OB-fold" sub-domain isolated from staphylococcal nuclease”. J. Mol. Biol. 250, 134-143.