Protein domains vs. structure domains - an example.

1

• Protein domains vs. structure domains - an
  example.

2
Genome of the week

• Nanoarchaeum equitans - archaea
• Hyperthermophile
• Diverged early in evolution from other archaea
• New kingdom of archaea?
• Obligate symbiont with Ignicoccus
• Smallest completely sequenced genome
• lt500kB
• Genome reduction observed in symbionts (Schmidt)
• Is N. equitans a primitive archaea or is the
  genome undergoing reductive evolution?

3

• N. equitans lacks genes necessary for many
  aspects of central metabolism.
• Cant make lipids, vitamins, amino acids, etc.
• Parasite, not symbiont? First archaea
• Genome is quite compact.
• 95 of genome codes for genes. 552.
• Not primitive.
• Has complete set of information pathway and cell
  cycle genes found in archaea.
• No longer undergoing reductive evolution.
• Normally would find pseudogenes - not found.

4
Protein complexes - why?

• Proteins often function as large, multi-subunit
  complexes.
• ribosomes
• Can get clues about the function of a protein by
  knowing what other proteins it contacts.

5
Proteinprotein interactions

• Genetic approach
• Yeast 2-hybrid
• Biochemical approach
• Co-immunoprecipitation
• Fusion protein affinity chromatography
• Cell-biology
• FRET - fluorescence resonance energy transfer
• Computational
• Rosetta Stone
• Co-regulation
• Phylogenetic analysis

6
Yeast 2-hybrid approach

• Based on the fact some transcriptional activators
  have separable DNA binding (BD) and
  transcriptional activation domains (AD).
• GAL4, LexA
• Protein you are studying Bait
• Fused to the DNA binding domain of GAL4
• Protein(s) you are screening Fish or Prey
• Fused to the activation domain of GAL4
• Transform Bait and Fish plasmids into yeast,
  measure the expression of a reporter gene.
• Usually a gene can be selected for when
  expressed.

7
Image from http//www.bioteach.ubc.ca/MolecularBi
ology/AYeastTwoHybridAssay/
8
Yeast 2-hybrid on a genome wide scale

• Clone every gene in your genome into both the
  bait and fish vectors.
• Systematically screen each gene for interactions.
• Mate individual yeast strains.
• Many false positives.

9
Interactome

• Term to define all of the protein interactions
  that take place in the cell.
• Book example - predicting human interactions.
• Based on data that only 10 of the measured
  interactions are physiological

10
Yeast 2-hybrid

• False-positives
• Some baits are sticky leading to non-functional
  interactions
• False negatives
• Binding not tight enough to detect interaction
• Fusion proteins often do not fold correctly
• Works best when comparing two proteins suspected
  of interacting
• Bacterial 2-hybrid systems

11
Co-immunoprecipitation

• Using an antibody to isolate and purify a protein
  from a whole cell lysate.
• Normally you will only purify the protein the
  antibody recognizes.
• Any additional proteins that co-purify are
  candidates for interacting proteins.

12
Hirano et al, 1997 Cell, Vol 89, 511-521, 16 May
1997
13
Fusion protein affinity chromatography

• Express the protein of interest as a fusion
  protein.
• 6-8X His residues
• Glutathione S-transferase (GST)
• Other tags
• Bind and purify the protein of interest
• Poly His residues will bind Ni2
• GST will bind glutathione

14
Image from Sigma-Aldrich
15
Fusion proteins - identifying interactions.

• In vivo - express fusion protein in vivo
• Purify complexes from the cell
• In vitro - overexpress protein in vitro
• Bind fusion protein to a column and run whole
  cell lysate through the column. Identify
  proteins that stick to the fusion protein.

16
Difficulties when using biochemical approaches

• Stability of proteinprotein interactions.
• Many are not stable enough to survive
  purification.
• Is the fusion protein functional?
• Many times fusions will not be functional.
• Quality of the antibody.
• Is it good enough to precipitate enough protein
  for analysis?

17
Computational methods

• Rosetta stone analysis
• Search for proteins that are separate in one
  organism but are fused into one protein in
  another organism.

18
Computational methods

• Co-expression
• Genes that are in operons are often functionally
  linked. (not always true).
• Determine if the structure of an operon is
  conserved, indicating co-expression.
• Candidates for interaction.
• Not a great method.

19
Phylogenetic analysis

• Search for the presence of a protein in all
  organisms.
• Determine the distribution.
• Identify other proteins that also show this
  distribution.
• Functionally interact? Physically?

20
PLEX

• Protein Link EXplorer.
• Uses phylogenetic profiles to predict possible
  associations.