DNA/Protein structure-function analysis and prediction Protein Structure Determination: X-ray Diffraction (Titia Sixma, NKI) Electron Microscopy/Diffraction NMR Spectroscopy (Lorna Smith, Oxford) Other Spectroscopic methods 17 Jan 2006

Spectroscopy: The whole spectrum 17 Jan 2006

DNA/Protein structure-function analysis and prediction Protein Structure Determination: X-ray Diffraction Electron Microscopy/Diffraction NMR Spectroscopy Other Spectroscopic methods 17 Jan 2006

Structure determination method X-ray crystallography Crystallization Purified protein Phase problem Crystal X-ray Diffraction Electron density Biological interpretation 3D structure 17 Jan 2006

Protein crystals Regular arrays of protein molecules ‘Wet’: 20-80% solvent Few crystal contacts Protein crystals contain active protein Enzyme turnover Ligand binding 17 Jan 2006 Example of crystal packing

Examples of crystal packing Acetylcholinesterase ~68% solvent 2 Glycoprotein I ~90% solvent (extremely high!) 17 Jan 2006

Flexible and heterogeneous!! Problematic proteins Multiple domains Similarly, floppy ends may hamper crystallization: change construct Membrane proteins Glycoproteins Flexible Lipid bilayer hydrophilic hydrophobic Flexible and heterogeneous!! 17 Jan 2006

Experimental set-up Options for wavelength: monochromatic, polychromatic variable wavelength Liq.N2 gas stream X-ray source detector goniometer beam stop 17 Jan 2006

Increasing resolution Diffraction image Diffuse scattering (from the fibre loop) Water ring Direct beam Beam stop Increasing resolution reciprocal lattice (this case hexagonal) Reflections (h,k,l) with I(h,k,l) 17 Jan 2006

The rules for diffraction: Bragg’s law Scattered X-rays reinforce each other only when Bragg’s law holds: Bragg’s law: 2dhkl sin q = nl 17 Jan 2006

Phase Problem If phases hkl and structure factor F(hkl) known: compute the electron density (x,y,z) In the electron density build the atomic 3D model However, the phases hkl are unknown ! 17 Jan 2006

How important are these phases ?? FH, H FK, K Fourier transform photo’s of Karle (top left) and Hauptman (top right) (two crystallography pioneers) Combine amplitudes FK with phase aH and inverse-fourier transform Combine amplitudes FH with phase aK and inverse-fourier transform (Taken from: Randy J. Read) FH, K FK, H 17 Jan 2006

How can we solve the Phase Problem ? Direct Methods small molecules and small proteins needs atomic resolution data (d < 1.2 Å) ! Difference method using heavy atoms multiple isomorphous replacement (MIR) anomalous scattering (AS) combinations (SIRAS,MIRAS) Difference method using variable wavelength multiple-wavelength anomalous diffraction (MAD) Using a homologous structure molecular replacement 17 Jan 2006

Building a protein model Find structural elements: -helices, -strands Fit amino-acid sequence 17 Jan 2006

Building a protein model Find structural elements: -helices, -strands Fit amino-acid sequence 17 Jan 2006

Effects of resolution on electron density Note: map calculated with perfect phases 17 Jan 2006

Effects of resolution on electron density Note: map calculated with perfect phases 17 Jan 2006

Effects of resolution on electron density Note: map calculated with perfect phases 17 Jan 2006

Effects of resolution on electron density Note: map calculated with perfect phases 17 Jan 2006

Refinement process Bad phases  poor electron density map  errors in the protein model Interpretation of the electron density map  improved model  improved phases  improved map  even better model … iterative process of refinement 17 Jan 2006

Validation Free R-factor (cross validation) Number of parameters/ observations Ramachandran plot Chemically likely (WhatCheck) Hydrophobic inside, hydrophilic outside Binding sites of ligands, metals, ions Hydrogen-bonds satisfied Chemistry in order Final B-factor values 17 Jan 2006

DNA/Protein structure-function analysis and prediction Protein Structure Determination: X-ray Diffraction Electron Microscopy/Diffraction NMR Spectroscopy Other Spectroscopic methods 17 Jan 2006

Electron microscopy Single particle Low resolution, not really atomic Less purity of protein, more transient state analysis Two-dimensional crystals Suited to membrane proteins Fibres Acetylcholine receptor Muscles, kinesins and tubulin Preserve protein by Negative stain (envelope only) Freezing in vitreous ice (Cryo-EM, true density maps) High resolution possible but difficult to achieve For large complexes: Combine with X-ray models 17 Jan 2006

Electron diffraction from 2D crystals: Nicotinic Acetylcholine Receptor G-protein coupled receptors Unwin et al, 2005 17 Jan 2006

Electron diffraction: near atomic resolution Structure of the alpha beta tubulin dimer by electron crystallography. Nogales E, Wolf SG, Downing KH. Nature 1998 391 199-203 17 Jan 2006

Single particle Electron Microscopy Select particles Sort into classes Average Reconstruct 3D image Single particle EM methods The electron beam gives so little damage that single particles can be imaged, but to obtain more detail(higher resolution) the damage is too large, and an average over many such particles is still necessary. However, this means that these do not necessarily need to be in a crystal, and that can be a huge advantage. Images are taken of individual molecules lying on a grid. To minimize radiation damage the exposure is very brief and the contrast between the image and the background is minimal. In this method particles are selected and sorted into different classes. Each class is then averaged to improve the signal to noise and the averaged images are used to reconstruct a 3D images. The method depends on the molecule falling on the grids in a fairly random way. Molecules that preferentially fall in one or two orientations need to be treated specially by tilting the grid. 17 Jan 2006

Cryo EM reconstruction: Tail of bacteriophage T4 17 Jan 2006 Leiman et al. Cell. 2004 Aug 20;118(4):419-29

DNA/Protein structure-function analysis and prediction Protein Structure Determination: X-ray Diffraction Electron Microscopy/Diffraction NMR Spectroscopy Other Spectroscopic methods 17 Jan 2006

1D NMR spectrum of hen lysozyme (129 residues) Too much overlap in 1D:  2D 1H-1H 1H-15N 1H-13C 17 Jan 2006

Step 1: Identification of amino acid spin systems H O b R g Amino acid spin system 17 Jan 2006

2D COSY spectrum of peptide in D2O Val ab Val bg Thr ab Thr bg 17 Jan 2006

Step 2: Sequential assignment H(i)-NH(i+1) NOE NH(i)-NH(i+1) NOE R H H O C a f N C N C C a N H H O C b H NH(i)-NH(i+1) NOE H H a H(i)-NH(i+1) NOE C g 17 Jan 2006

2D NOESY spectrum Gly Val Gly Leu Thr Ser Peptide sequence (N-terminal NH not observed) Arg-Gly-Asp-Val-Asn-Ser-Leu-Phe-Asp-Thr-Gly Gly Val Gly Leu Thr Ser Phe Asn Asp Asp 17 Jan 2006

Nitroreductase Dimer: 217 residues Too much overlap in 2D:  3D 1H-1H-15N 1H-13C-15N 17 Jan 2006

Structural information from NMR: NOEs For macromolecules such as proteins: Initial build up of NOE intensity  1/r6 Between protons that are < 5Å apart 17 Jan 2006

NOEs in a-helices NH-NH(i,i+1) 2.8Å aH-NH(i,i+3) 3.4Å R NH-NH(i,i+1) 2.8Å aH-NH(i,i+3) 3.4Å aH-bH(i,i+3) 2.5-4.4Å Cytochrome c552 3D 1H-15N NOESY aH-NH(i,i+3) resides 38-47 form a-helix 17 Jan 2006

NOEs in b-strands aH-NH(i,j) 2.3Å aH-NH(i,j) 3.2Å 17 Jan 2006

Long(er) range NOEs Provide information about Packing of amino acid side chains Fold of the protein NOEs observed for CaH of Trp 28 in hen lysozyme 17 Jan 2006

Spin-spin coupling constants Fine structure in COSY cross peaks For proteins 3J(HN.Ha) useful: Probes main chain f torsion angle Hen lysozyme: N C a H O b R g f 2 3 4 5 6 7 8 9 10 20 40 60 80 100 120 140 Coupling constant (Hz) Residue number 17 Jan 2006

Structural information from NMR 3. Hydrogen exchange rates Dissolve protein in D2O and record series of spectra Backbone NH (1H) exchange with water (2H) Slow exchange for NH groups In hydrogen bonds Buried in core of protein After 20 mins After 68 hours 1H-15N HSQC spectra for SPH15 in D2O 17 Jan 2006

Structural information from NMR NOEs 1Ha - 2Hb 1.8-2.5Å (strong) 2Ha - 40HN 1.8-5.0Å (weak) 3He - 88Hg2 1.8-5.0Å (weak) 3Hd - 55Hb 1.8-3.5Å (medium) Spin-spin coupling constants 1C-2N-2Ca-2C -120+/-40° 4C-5N-5Ca-5C -60+/- 30° 5C-6N-6Ca-6C -60+/- 30° Hydrogen exchange rates 10HN-6CO 1.3-2.3Å 10N-6CO 2.3-3.3Å 11HN-7CO 1.3-2.3Å 11N-7CO 2.3-3.3Å 17 Jan 2006

NMR structure determination: hen lysozyme 129 residues ~1000 heavy atoms ~800 protons NMR data set 1632 distance restraints 110 torsion restraints 60 H-bond restraints 80 structures calculated 30 low energy structures used 2000 4000 6000 8000 1 10 4 1.2 10 10 20 30 40 50 60 70 Total energy Structure number 17 Jan 2006

Solution Structure Ensemble Disorder in NMR ensemble lack of data ? or protein dynamics ? 17 Jan 2006

DNA/Protein structure-function analysis and prediction Protein Structure Determination: X-ray Diffraction Electron Microscopy/Diffraction NMR Spectroscopy Other Spectroscopic methods 17 Jan 2006

Ultrafast Protein Spectroscopy Structure-sensitive technique State of protein and substrate Redox state Protonation Elektronen Follow reactions in real-time Why ultrafast spectroscopy? Molecular movement: time scales of 10 fs – 1 ps (10-15 – 10-12 s) O  H stretching frequency: ~3500 cm-1 (~ 10-14 s) Transfer or movement of proton, electron or C-atom 17 Jan 2006

Excited state difference spectra of Chl and Pheo keto ester Keto and ester C=O in Chlorophyll a, Pheophytin a are redox and environmental probes 17 Jan 2006

Why is vibrational spectroscopy sensitive to structure? ω1’ ω2’ X – C – O – H O – X O ω1 ω2 17 Jan 2006

H-bond response during the photocycle weaker stronger broken Replace Glu with Gln which donates weaker H-bond 17 Jan 2006

X-ray crystallography Electron microscopy / diffraction Summary Method Pro Con X-ray crystallography High resolution No size limits Easy addition of ligands Crystals required Phase problem Electron microscopy / diffraction Single particles possible Well suited to membrane proteins Labour intensive High resolution difficult NMR spectroscopy In solution Dynamic information possible Resolution variable Limited size (< ~30kDa) Other spectroscopy (Very) high time resolution (ps!) Very sensitive Technically extremely complex Limited applicability Limited information 17 Jan 2006