Proteomics: Interaction Proteomics

1
Proteomics Interaction Proteomics

• Yao-Te Huang
• Nov 2, 2009

2
Introduction

• Protein interactions and functions are intimately
  related.
• The structure of a protein influences its
  function by determining the other molecules with
  which it can interact and the consequences of
  those interactions.

3
Introduction (contd.)

• Experimental methods available to detect protein
  interactions vary in their level of resolution.
• These observations can be classified into four
  levels (a) atomic scale, (b) binary
  interactions, (c) complex interactions, and (d)
  cellular scale.

4
Introduction (contd.)

• Atomic-scale methods
• showing the precise structural relationships
  between interacting atoms and residues
• The highest resolution methods e.g., X-ray
  crystallography and NMR
• Not yet applied to study protein interactions in
  a high-throughput manner.

5
Introduction (contd.)

• Binary-interaction methods
• Methods to detect interactions between pairs of
  proteins
• Do not reveal the precise chemical nature of the
  interactions but simply report such interactions
  take place
• The major high-throughput technology the yeast
  two-hybrid system

6
Introduction (contd.)

• Complex-interaction methods
• Methods to detect interactions between multiple
  proteins that form complexes.
• Do not reveal the precise chemical nature of the
  interactions but simply report that such
  interactions take place.
• The major high-throughput technology systematic
  affinity purification followed by mass
  spectrometry

7
Introduction (contd.)

• Cellular-scale methods
• Methods to determine where proteins are localized
  (e.g., immunofluorescence).
• It may be possible to determine the function of a
  protein directly from its localization.

8
COIB (2001), 12334-339
9
Principles of protein-protein interaction analysis

• These small-scale analysis methods are also
  useful in proteomics because the large-scale
  methods tend to produce a significant number of
  false positives.
• They include (a) genetic methods, (b)
  bioinformatic methods, (c) Affinity-based
  biochemical methods, and (d) Physical methods.

10
Genetic methods

• Classical genetics can be used to investigate
  protein interactions by combining different
  mutations in the same cell or organism and
  observing the resulting phenotype.
• Suppressor mutation A secondary mutation that
  can correct the phenotype of a primary mutation.

11
Suppressor mutation
12
Synthetic lethal effect
13
Bioinformatic methods

• (A) The domain fusion method (or Rosetta stone
  method)
• The sequence of protein X (a single-domain
  protein from genome 1) is used as a similarity
  search query on genome 2. This identifies any
  single-domain proteins related to protein X and
  also any multi-domain proteins, which we can
  define as protein X-Y.
• As part of the same protein, domain X and Y are
  likely to be functionally related.

14
The domain fusion method (or Rosetta stone method)

• The sequence of domain Y can then be used to
  identify single-domain orthologs in genome 1.
• Thus, Gene Y, formerly an orphan with no known
  function, becomes annotated due to its
  association with Gene X. The two proteins are
  also likely to interact.
• The sequence of protein X-Y may also identify
  further domain fusions, such as protein Y-Z. This
  links three proteins into a functional group and
  possibly identifies an interacting complex.

15
The domain fusion method (or Rosetta stone method)
16
Bioinformatic methods

• (B) The phylogenetic profile
• It describes the pattern of presence or absence
  of a particular protein across a set of organisms
  whose genomes have been sequenced. If two
  proteins have the same phylogenetic profile (that
  is, the same pattern of presence or absence) in
  all surveyed genomes, it is inferred that the two
  proteins have a functional link.
• A proteins phylogenetic profile is a nearly
  unique characterization of its pattern of
  distribution among genomes. Hence any two
  proteins having identical or similar phylogenetic
  profiles are likely to be engaged in a common
  pathway or complex.

17
YPL207W clusters with the ribosomal proteins and
can be assigned a function in protein synthesis.
When homology is present, the elements are shaped
on a gradient from light red (low level of
identity) to dark red (high level of identity)
18
Affinity-based biochemical methods

• (A) Affinity chromatography can be used to trap
  interacting proteins. If protein X is immobilized
  on Sepharose beads (e.g., using specific
  antibodies), then proteins (and other molecules)
  interacting with protein X can be captured from a
  cell lysate passed through the column. After
  washing away unbound proteins, the bound proteins
  can be eluted, separated by SDS-PAGE and analyzed
  by mass spectrometry.

19
Affinity chromatography followed by SDS-PAGE
Mass spectrometry
20
Immunoprecipitation

• The addition of antibodies specific for protein X
  to a cell lysate will result in the precipitation
  of the antibody-antigen complex.
• The technique is usually carried out with
  polyclonal antisera.
• The precipitated complexes are separated from the
  cell lysate by centrifugation, washed and then
  fractionated by SDS-PAGE, and the bound proteins
  can be identified by mass spectrometry.

21
Immunoprecipitation
22
GST pulldown

• The protein X is expressed as a fusion to GST.
  After mixing the fusion protein with a cell
  lysate and allowing complexes to form,
  glutathione-coated beads are added to capture the
  GST part of the fusion. The beads are recovered
  by centrifugation, washed and the recovered
  proteins fractionated and identified by mass
  spectrometry.

23
GST pulldown
24
Crosslinking

• Interacting proteins can be identified by
  crosslinking. A labeled crosslinker is added to
  protein X in vitro and the cell lysate is added
  so that interactions can occur. If the crosslink
  is activated at this stage, interacting proteins
  become covalently attached to the bait. After
  purification, the crosslink can be cleaved and
  the interacting proteins separated by 2D SDS-PAGE.

25
Crosslinking (contd.)
26
Physical methods

• High-resolution methods (e.g., X-ray
  crystallography NMR) providing data about the
  relative spacing of atoms of interacting
  molecules.
• Low-resolution methods e.g., electron
  crystallography and electron tomography.

27
FRET (Fluorescence Resonance Energy Transfer)
28
FRET
FRET is the energy transfer that occurs when two
fluorophores are close together, and one of
fluorophores (the donor) has an emission
spectrum that overlaps the excitation spectrum
(absorption spectrum) of the other fluorophoe
(the acceptor).
29
Basic Theory of FRET
kT(r) (QD?2)(1/?Dr6)(9000 In10)(1/128?5NAn4)(?F
D(?)?A(?) ?4d ?) (1/?D)(R0/r)6
where R0 is the Förster distance r
is the distance between the donor and the acceptor
J(?), the so-called overlap integral ?FD(?)?A(?)
?4d ?
E 1/(1(r/R0)6)
E IA/(IDIA)
where E is the efficiency of the energy
transfer IA the fluorescence intensity of the
acceptor ID the fluorescence intensity of the
donor
30
FRET
31
FRET
R0 is the Förster distance
32
FRET distance-dependent
R0 is the Förster distance
r is the distance between the donor and the
acceptor
E is the efficiency of the energy transfer
FD the fluorescence intensity of the donor in
the absence of the acceptor FDA the fluorescence
intensity of the donor in the presence of the
acceptor
Note when rR0, E0.5
33
Library-based methods for the global analysis of
binary interactions

• Standard cDNA expression libraries
• Phage display method
• The yeast two-hybrid system

34
Standard cDNA expression libraries

• Expression libraries are usually screened with
  labeled antibodies. In place of antibodies, other
  proteins can be used as probes. For example,
  labeled calmodulin has been used to screen for
  calmodulin-binding proteins.
• Low throughput
• Does not provide the native conditions for the
  folding of all proteins, so a significant number
  of interactions would not be detected.

35
Phage display method (1)
M13 (a filamentous phage containing ss-DNA
encased in a protein coat) contains five coat
proteins, two of which are gVIIIp (gene VIII
protein) and gIIIp (gene III protein).
36
Phage display method (2)
37
Phage display method (2) contd.
38
The phage display method
39
The yeast two-hybrid system

• Transcription factors generally comprise two
  functionally independent domains, one for DNA
  binding and one for transcriptional activation.
  These do not have to be covalently joined
  together, but can be assembled to form a dimeric
  protein. This principle is exploited to identify
  protein interactions. Bait proteins are expressed
  in one yeast strain as a fusion with a
  DNA-binding domain and candidate prey proteins
  are expressed in another strain as fusions with a
  transactivation domain. When the two strains are
  mated, functional transcription factors are
  assembled only if the bait and prey interact.
  This can be detected by including a reporter gene
  activated by the hybrid transcription factor.

40
The yeast two-hybrid yeast
41
Limitations of the yeast two-hybrid system

• First, where independent groups have carried out
  similar, large-scale studies, the degree of
  overlap in the reported interactions is very low
  (10-15). This suggest either that the screens
  were not comprehensive or that even minor
  differences in experimental conditions could
  influence the types of interactions that are
  detected.

42
Limitations of the yeast two-hybrid system

• Secondly, a significant number of
  well-characterized interactions are not detected
  in the large-scale screens, suggesting there is a
  high level of false negatives.
• Thirdly, a significant number of interactions
  that are detected in large-scale screens appear
  spurious when investigated in more detail,
  suggesting there is also high level of false
  positives.

43
A variant of the yeast two-hybrid system
44
Protein interaction maps
Node proteins or protein complexes are treated
as nodes. Edge (or link) interactions between
them. Some proteins serve as hubs for very large
numbers of interactions.
45
Binary interaction map including 1200 interacting
proteins in yeast
Trends in Cell biology (2001), 11 102-106
46
A simplified version in which yeast proteins have
been clustered according to their function
47
(No Transcript)