Structural Biology and Protein Folding


Modern Structural Biology uses 3 experimental methods to determine medium and high resolution structures of macromolecules. These include X-ray crystallography, Nuclear Magnetic Resonance Spectroscopy (NMR) and Structural Electron Micrsocopy. Research in these areas makes extensive use of BioInformatics tools to evaluate protein superfamilies, select appropriate protein coding regions for cloning and over-expression and to analyze the results of a structure determination. Molecular biology methods are used to produce most of the proteins that are being studied. Department facilities for Structural Biology are excellent and include a new Raxis IV for cryo-crystallography and a Tecnai 200 kV Field Emission Gun (FEG) electron cryo-microscope.

X-ray crystallography

A map of the dNLP-core chaperone at a resolution of 1.5 Å highlights a well ordered ß-hairpin that may play a role in histone binding. Image courtesy of V.M.H. Namboodiri & C.W. Akey.


In this approach, macromolecules must be crystallized and this requires the over-expression and purification of milligram quantities of proteins and nucleic acids. Once a protein or complex is crystallized, its diffraction quality is evaluated at our in-house X-ray crystallography laboratory using either an R-axis II or IV camera system. Diffraction patterns may be collected from crystals at room temperature or at liquid nitrogen temperature. To achieve higher resolution and carry out Multi-wavelength anomalous dispersion (MAD) phasing, we generally take crystals to the National Synchrotron Light Source (Brookhaven, New York) or to other synchrotrons like CHESS (Cornell University, New York) or the Advanced Photon Source (Argonne National Laboratory, Illinois).

Members of the Department are particularly interested in the structure and function of enzymes and intramolecular proteases, chaperones, large macromoleculer complexes (such as chaperone-histone complexes, restriction endonuclease-DNA complexes etc.) DNA transcription factors, annexins and proteins that interact at membrane surfaces, lipoproteins, receptors, channel forming proteins and actin binding proteins.

For more detailed information on X-ray crystallographic projects see the following home pages:

Christopher Akey

David Atkinson

Graham Shipley


Nuclear Magnetic Resonance Spectroscopy (NMR)

This region of the 2D-NOESY NMR spectrum reveals that switching two amino acids in the sequence of the phage P22 ARC repressor results in a sheet to helix structural transition in the NllL/L12N mutant. Image courtesy of M.H.J. Cordes, R.T. Sauer, N.P. Walsh & C.J. McKnight.


NMR structures give a series of snapshots of the dynamic motions of macromolecules. Importantly, these macromolecules do not have to be crystallized. However, they must be stable and soluble to achieve the concentrations required for data collection. Proteins are generally over-expressed in a system that allows selective labeling with various isotopes to give more informative spectra. After peak assignments have been made, a geometrical analysis gives an ensemble of structures that do not violate the given constraints. Currently proteins up to ~30,000 may be used for de novo structural analysis, but this boundary is constantly being pushed upwards with new technology, as witnessed by a recent analysis of the GroEL-GroES complex (Nature, 2002; 418: 207-211).

Members of the Department are particularly interested in the structure and dynamics of headpiece domains, proteins which bind lipids and lipoprotein domains (please see NMR facility page).

For more detailed information on NMR projects see the following home pages:

Jim Hamilton

James McKnight

Assen Marintchev


Cryo-Electron Microscopy

The development of a practical method to prepare specimen frozen in a thin film of vitreous ice by Dr. J. Dubochet, opened the door to the structure determination of native specimen using the electron microscope (cryo-electron microscopy). This method of structural analysis can be applied to single particles (50 kDa and larger), viruses, helices and thin 2D crystals and thus is quite versatile. In particular, single particle methods have now achieved near atomic resolution 4.5-2.5 Å resolution with the advent of direct electron detectors, automated data acquisition and greatly improved software. In addition, large complexes can be studied by docking high resolution models determined by cryo-EM, X-ray crystallography and NMR within EM derived 3D density maps, to provide insights into the functions of these dynamic machines. Currently, the resolution that can be achieved depends upon a number of factors including specimen flexibility and possible damaging interactions with the air water interface. Our Structural EM facility includes a Tecnai 200 kV Field Emission Gun electron microscope with a 4K x 4K CMOS camera for medium resolution work on biological specimens and also allows us to screen suitable specimen for high resolution data collection at cryo-EM beamlines.

Members of the Faculty are currently interested in the structure of bacterial adhesion pili, actin thin filaments, viral RNA-polymerase tubes and crystals, ribosome-protein translocation channel complexes from the ER and E. coli with associated components, the apoptosome, the nuclear pore complex, membrane protein receptors and LDL particles.

Representative surface and mesh density maps for the E. coli ribosome. Atomic models in standard stick representation (RNA in gold and proteins in blue).

For more detailed information on Structural EM projects see the following home pages:


Protein Folding

Olga Gursky

Exchangeable apolipoproteins are soluble protein components of lipoproteins that mediate cholesterol transport and metabolism and play crucial roles in the pathogenesis of atherosclerosis and several other major human disorders, including amyloidosis. Our research is focused on the structural adaptability of apolipoproteins to heterogeneous lipoproteins and to plasma. This distinct adaptability is key to apolipoproteins functions and is attributed to re-folding and re-packing of the constituent amphipathic alpha-helices. Our earlier CD spectroscopic and calorimetric analyses suggest that lipid binding by apolipoproteins is mediated via the molten globule-like state in plasma. In our on-going NIH-funded research, we analyze the smallest human apolipoprotein C-1 (6 kD) by using mutagenesis and spectroscopic techniques to1) dissect the apolipoprotein folding pathway, from partially folded monomeric to fully folded lipid-bound state; 2) test whether the helix-turn-helix motif forms an autonomous folding and functional unit in apolipoproteins; 3) determine the role of hydrophobic clustering for the nucleation and stability of the helical structure.

Helix-turn-helix autonomous folding unit in human apolipo-protein C-1 (right) inferred from the results of Proline scanning mutanegesis coupled with far-UV CD spectroscopy.
Helix-turn-helix autonomous folding unit in human apolipo-protein C-1 (right) inferred from the results of Proline scanning mutanegesis coupled with far-UV CD spectroscopy.

Our recent spectroscopic and electron microscopic studies of protein folding / unfolding on apoC-1:lipid complexes that model nascent discoidal high-density lipoproteins (HDL), as well as on-going studies of mature spherical HDL isolated from human plasma, reveal a novel mechanism of lipoprotein stabilization that is based on kinetics rather than thermodynamics. Our current work is aimed at identifying in detail the molecular origins of the high kinetic barriers that determine lipoprotein stability.

Haya Herscovitz – Chaperone-assisted protein folding

Folding and maturation of nascent proteins in the cell is dependent on the assistance of a diverse group of proteins termed molecular chaperones. These folding assistants bind to unfolded proteins, to stabilize them thus increasing their folding efficiency so that they attain their functional conformation in the overly-crowded environment occurring in the cell.

Our laboratory is interested in elucidating the details of chaperone-assisted folding of apolipoprotein B (apoB) whose correct folding involves its assembly with lipids to form very low density lipoprotein (VLDL), the precursors of low density lipoproteins (LDL). LDL is a major risk factor for developing heart disease and stroke. Therefore, it is essential to characterize the underlying mechanisms that regulate the secretion of VLDL. We have identified a number of ER-resident molecular chaperones (e.g. GRP94, ERp72, BiP, calreticulin and cyclophilin B) that interact with apoB during various stages of its maturation. We are characterizing these interactions and developing model systems to determine which of these chaperones are critical for the folding and secretion of apoB-VLDL and how do they perform their chaperone function.

In addition, we are using a proteomic approach to identify potentially novel proteins that interact with apoB and may thus be critically involved in its folding. Some might be unique to apoB and may therefore serve as targets for modulating the secretion of VLDL.

C. James McKnight

The McKnight laboratory is interested in the structure and function and folding of a variety of proteins. In particular we are interested in proteins of minimal length that are still able to form a fully native folded structure. For example, we use a combination of molecular biological, biophysical, and NMR methods to investigate the 35-residue subdomain from the headpiece domain of the F-actin-binding protein villin. The headpiece subdomain is one of the smallest folded proteins that is thermostable and monomeric, yet contains no disulfide bonds, and binds no metals. We hope to take advantage of the small size of the headpiece subdomain to bridge the gap between experimental and computational approaches to protein folding.

Backbone ribbon representation of the 35-residue subdomain from villin headpiece is shown in the image to the right.