Structural proteomics

1
Structural proteomics

• Handouts. Proteomics section from book already
  assigned.

2
What is structural proteomics/genomics?

• High-throughput determination of the 3D structure
  of proteins
• Goal to be able to determine or predict the
  structure of every protein.
• Direct determination - X-ray crystallography and
  nuclear magentic resonance (NMR).
• Prediction
• Comparative modeling -
• Threading/Fold recognition
• Ab initio

3
Why structural proteomics?

• To study proteins in their active conformation.
• Study proteindrug interactions
• Protein engineering
• Proteins that show little or no similarity at the
  primary sequence level can have strikingly
  similar structures.

4
An example

• FtsZ - protein required for cell division in
  prokaryotes, mitochondria, and chloroplasts.
• Tubulin - structural component of microtubules -
  important for intracellular trafficking and cell
  division.
• FtsZ and Tubulin have limited sequence similarity
  and would not be identified as homologous
  proteins by sequence analysis.

5
FtsZ and tubulin have little similarity at the
amino acid sequence level
Burns, R., Nature 391121-123 Picture from E.
Nogales
6
Are FtsZ and tubulin homologous?

• Yes! Proteins that have conserved secondary
  structure can be derived from a common ancestor
  even if the primary sequence has diverged to the
  point that no similarity is detected.

7
Current state of structural proteomics

• As of Feb. 2002 - 16,500 structures
• Only 1600 non-redundant structures
• To identify all possible folds - predicted
  another 16,000 novel sequences needed for 90
  coverage.
• Of the 2300 structures deposited in 2000, only
  11 contained previously unidentified folds.
• Overall goal - directly solve enough structures
  directly to be able to computationally model all
  future proteins.

8
Protein domains - structure

• clearly recognizable portion of a protein that
  folds into a defined structure
• Doesnt have to be the same as the domains we
  have been investigating with CDD.
• RbsB proteins as an example.

9
Main secondary structure elements

• a-helix - right handed helical structure
• b-sheet - composed of two or more b-strands,
  conformation is more zig-zag than helical.

10
Images from http//www-structure.llnl.gov/Xray/tut
orial/protein_structure.htm http//www.expasy.org/
swissmod/course/text/chapter1.htm
11
Folds/motifs - tertiary structure

• How these secondary structure elements come
  together to form structure.
• Helix-turn-helix
• Determining the structure of nearly all folds is
  the goal of structural biology

12
Quaternary structure

• Refers to the structure formed by more than one
  polypeptide.
• Many proteins function as complexes - best to
  know the structure of the complex rather than
  each individual
• Proteins may have different conformations when in
  a complex vs. alone.

13
X-Ray Crystallography

• Make crystals of your protein
• 0.3-1.0mm in size
• Proteins must be in an ordered, repeating
  pattern.
• X-ray beam is aimed at crystal and data is
  collected.
• Structure is determined from the diffraction
  data.

14
Image from http//www-structure.llnl.gov/Xray/101i
ndex.html
15
Schmid, M. Trends in Microbiolgy, 10s27-s31.
16
X-ray crystallography

• Protein must crystallize.
• Need large amounts (good expression)
• Soluble (many proteins arent, membrane
  proteins).
• Need to have access to an X-ray beam.
• Solving the structure is computationally
  intensive.
• Time - can take several months to years to solve
  a structure
• Efforts to shorten this time are underway to make
  this technique high-throughput.

17
Nuclear Magnetic Resonance Spectroscopy (NMR)

• Can perform in solution.
• No need for crystallization
• Can only analyze proteins that are lt300aa.
• Many proteins are much larger.
• Cant analyze multi-subunit complexes
• Proteins must be stable.

18
Structure modeling

• Comparative modeling
• Modeling the structure of a protein that has a
  high degree of sequence identity with a protein
  of known structure
• Must be gt30 identity to have reliable structure
• Threading/fold recognition
• Uses known fold structures to predict folds in
  primary sequence.
• Ab initio
• Predicting structure from primary sequence data
• Usually not as robust, computationally intensive

19
Structure of the ribosome

• Ribosome - made up of 3 RNA molecules and over 50
  proteins.
• Structure of the 70S ribosome solved by combining
  several models of the individual 30S and 50S
  subunits

20
Bacterial ribosome

• Ribosome is a 2.3 MDa complex
• 50S and 30S subunits
• 54 proteins and 3 RNAs (23S, 16S, and 5S RNAs)
• Can account for 50 of cell mass during rapid
  growth.
• Major target for many antibiotics.
• Ribosome is a ribozyme!

Ramakrishnan (2002) Cell. 108557-572