Protein Structure

1
Protein Structure

• Tertiary structure determination
• Evaluation
• Comparison

Supplemental primer on protein structure
http//www.ncbi.nlm.nih.gov/About/primer/molecular
mod.html
BIO520 Bioinformatics Jim Lund
2
Protein Structure3D Determination

• X-ray crystallography
• requires crystals
• NMR methods (Nobel prize 2002, Kurt Wüthrich)
• lt30kD

3
X-ray crystallography

• Purify protein
• Crystallize protein
• Test an array of conditions to determine which
  give good quality crystals.
• X-ray diffraction
• Compute atom locations (C, N, O, S)
• Relax structure computationally (energy
  minimization)

4
X-ray crystallography
5
NMR Structure Determination

• Isotopically label protein
• NMR NOEs, J-coupling constants measurements give
  structural constraints
• Compute structures using constraints.
• Relax structure computationally (energy
  minimization)

6
Applications of NMR

• Degree of folding (folded, partially folded,
  unfolded) and kind of fold (primarily
  alpha-helix or beta-sheet) of a protein
  independent from a biological test of activity.
• Optimization of conditions such as pH,
  temperature, salt concentration etc. for best
  solubility, folding, least aggregation etc in
  short time.
• With 15N-labelling (not radioactive!) counting of
  folded/unfolded residues becomes possible this
  might help to determine domain borders more
  accurately when cloning protein domains.
• Binding studies are possible with and without
  isotope labeling.
• Ligand conformations can be determined.
• Reaction products or intermediates can be
  investigated.
• Folding and unfolding studies for slow kinetics
  (ms - sec - min) are possible, e.g. with
  H/D-exchange experiments or stopped-flow
  deuterium-pulse-experiments.
• Dynamics-investigations yield information on the
  degree of oligomerization (monomer/dimer/...) as
  well as mobility of active sites of the protein.

7
Structure Evaluation

• Torsion angles in disallowed Ramachandran plot
  regions
• Maximize H-bonding
• Minimize exposed hydrophobic side chains
• Maximize exposed polar/charged side chains
• Holes-gtbad
• Minimize distorted H-bonds, distorted helix
  angles
• Low R-factor (X-ray structure), low backbone RMSD
  (NMR)

8
Structure Evaluation Programs

• DSSP
• Identify secondary struct, solvent accessibility
  (buried or surface aa)
• PROCHECK
• Calculates torsion angles, bond angles,
  interatomic distances
• VADAR (Volume, Area, Dihedral Angle Reporter)
• Classifies secondary struct, salt bridges
• Solvent accessibility (buried or surface aa),
  buried charge

9
Protein Structure Prediction

• Homology modeling
• SwissModel, SDSC1, CPH Models
• Threading
• SAMt99, 3D-PSSM, FUGUE
• Ab initio
• GAMESS, Rosetta (Baker lab)
• EVA Evaluation of automatic structure prediction
  servers (http//cubic.bioc.columbia.edu/eva/)

10
Homology modeling

• Align query protein to known structure.
• Replace known with query.
• Backbone insertions/deletions.
• Side chains.
• Refine model using energy minimization.
• SwissModel, SDSC1, CPH Models

11
Threading (Fold Recognition)

• Align sequences, one has known structure
• thread backbone of test sequence into known
  structure.
• minimize energies
• predicts 3D structure
• Sequence of unknown structure
• thread through all possible folds

12
Threading Services

• Generate a .pdb from a thread of a gt30 similar
  homolog
• http//www.expasy.ch/swissmod/SWISS-MODEL.html
• Search for threads in proteins 0-25 similar to
  your query (structural analogs without homology)
• http//www.embl-heidelberg.de/predictprotein/predi
  ctprotein.html

13
Molecular Mechanics

• Practical, rapid
• Apply to many problems
• thousands of atoms
• different atom types
• proteins, organics, nucleic acids, drugs
• dynamic kinetics and thermodynamics
• interactions
• DOCKING

14
Ab initio prediction

• Two components
• Devising a scoring (ie, energy) function that
  can distinguish between correct (native or
  native-like) structures from incorrect ones.
• A search method to explore the conformational
  space.

15
Molecular Mechanics Assumptions

• Nuclei, electrons lumped into atoms
• All atoms are spherical
• Atoms interact on springs, via classical
  potentials
• Interactions
• determine spatial distribution
• determine energies of states

16
Energy Calculations
Imagehttp//cmm.cit.nih.gov/modeling/guide_docume
nts/molecular_mechanics_document.html
17
Energy
Energy StretchBendTorsionNon-bonded
Charge van der Waals Dipole...
18
Simple Calculations
Ek(r-r0)2 Bond stretching
19
Non-bonded Energies
E?? -Aij Bij ?? qiqj i j rij6 rij12
i j rij
20
Geometry/Computation

• Software packages for molecular dynamics
• SYBL
• AMBER
• CHARMM

Used for Ab initio Energy Minimization Docking
21
Molecular mechanics Tweaking

• Single-site mutation
• Effect on protein active site, eg
• Energy Minimization of Residue Conformers

ESyPred3D "Poor Man's Version"
22
Docking

• Energy minimize bound molecules
• combinatorial optimization
• Important in drug design/modelling

QSAR/Superposition Combinatorial Chemistry
23
Docking Programs

• RELIWE
• www.gmd.de/SCAI/area-bioinf.html
• DOCK
• www.cmpharm.ucsf.edu/kuntz/dock.html
• AutoDock
• www.scripps.edu/pub/olson-web/doc/autodock/

24
Structure Neighboring

• Recognize protein 3D similarities w/o sequence
  similarity
• restricted to known 3D
• Heirarchical databases
• CATH
• Class(C), Architecture(A), Topology(T) and
  Homologous superfamily (H)
• Computational
• SCOP
• Structural Classification of Proteins
• hand annotated
• Structure similarity searches

25
Structure comparison searches

• Similar groups of secondary structure
• DALI
• Distance matrices
• CE, VAST (Vector Alignment Search Tool)
• Vector representation of secondary structure
• VAST
• Identify pairs of secondary structure elements
  with similar type, orientation, connection, then
  extend these matches.
• NCBI tool
• Can view aligned structures with Cn3D