Transcript the protein (or proteins)

PROTEOMICS: LARGE-SCALE PROTEIN IDENTIFICATION
& ANALYSIS
2D-gel electrophoresis & mass spectrometry
1. Isolate proteins from tissue (organism, condition…) of interest
2. Electrophoretic separation of proteins by charge (isoelectric
focusing) and by size (SDS-PAGE)
3. Peptide fragmentation of individual protein
(with proteases eg trypsin)
Fig. 6.10
see Fig. 6.11
4. Determine precise peptide mass by MALDI-TOF (matrix-assisted
laser desorption ionization – time of flight) mass spectrometry
5. Compare aa sequences
to genomic data to correlate
protein with its gene
Fig. 6.12
STRUCTURAL PROTEOMICS
- large-scale determination of protein structures
Start with gene of interest
– cloning, expression, purification of protein
X-ray crystallography
- crystallize protein & X-ray diffraction analysis
NMR spectroscopy
- for small proteins or domains (in solution)
Protein data bank: www.rcsb.org/pdb/
Nov 2000 = 13,750 structures
Nov 2001 = 16,550
Nov 2006 = 40,132
Nov 2008 = 54,559
Nov 2009 = 61,418
Nov 2011 = 76,814
Analysis of protein sequences and structures
How similar are hemoglobin a-chain and b-chain?
hemoglobin
tetramer
Identification of
- protein motifs, catalytic centres…
- binding to ligands, drugs
- interaction with other macromolecules
- relatedness to other proteins (homology modelling)
clues from protein sequence/structure about biological function
For example:
EMBL-EBI (European Bioinformatics Institute) www.ebi.ac.uk/Tools/ppsearch/
ExPASy (Swiss Institute of Bioinformatics) www.expasy.org/
www.ncbi.nlm.nih.gov/structure
How to find proteins that interact with protein of interest?
1. Phage display
- generate phage library producing collection of fusion proteins
between phage coat protein & “ test protein” from genome of interest
- hybrid protein will be “displayed”
on outer surface of phage
- then screen library to find ones having expressed protein which
interacts with “test” protein of interest
Fig. 6.14
2. Yeast 2-hybrid system
Background info about transcription in eukaryotes
Transcription
factors have 2
domains
Activator
domain
DNA
binding
domain
RNA pol II
gene
mRNA
Regulatory
cis-element
If TF domains physically
separated, no transcription
But if “bait” & “prey” interact to bring TF
domains close together, then transcription
occurs
mRNA
Determining protein-protein interactions using yeast 2-hybrid system
- use separate vectors to prepare
[1] “bait” fused to DNA binding domain of a yeast transcription factor
[2] shotgun library of possible “prey” fused to activation domain of yeast TF
- fuse (1) to gene for protein X = “bait”
“prey” – generate library where (2) is fused to random
coding sequences from organism of interest (eg. human)
Fig. 6.15
- co-transform yeast cells (which lack this transcription factor TF)
Fig. 6.15
if protein X and “prey” (from library) interact, the 2 domains of
yeast TF will be close together (& functional), so activate reporter gene
eg if use lacZ reporter gene – blue colour of yeast colony
3. Affinity column chromatography
B
- protein B (“bait”) attached on column to “fish out”
the protein (or proteins) which specifically bind to it
B
If bait protein does not interact
directly with protein(s) in a
complex, they may not be isolated
Fig.6.18
or use co-immunoprecipitation
(p.182)
Fig.6.17
4. Computational approaches to predict protein-protein interactions
10 different organisms
If 2 genes are functionally-related,
expect them to be co-inherited…
… and may be physically close in
genome (as well as co-inherited)
Do genes 1 and 2 (orange & green)
pass this test?
Gene fusion/fission (2 short genes in some
organisms vs. one long gene in others)
- premise that composite (naturally-fused)
proteins have direct physical interaction (or
functional association)
Harrington FEBS Lett. 582:1251, 2008
3. Bioinformatics approach to predict protein-protein interactions
- search for one large gene in organism X vs. two
separate smaller genes in Y
his10
his2
in E.coli
in yeast
“ HIS2 ”
Fig.6.19
Search of complete genomes of E.coli, Haemophilus, Methanococcus & yeast
- found 215 cases of “fused vs. split” state
Enright et al. Nature 402: 86, 1999
Yeast protein-protein interaction map (from experimental data)
- lines connecting dots represent
known protein- protein interactions
2002 colour-coded for
2001
biological function
Fig. 6.20
red dots = essential proteins (so knockout is lethal)
green = non-lethal; orange = slow growth; yellow = unknown effect
“Protein interaction network in Huntington’s disease”
- triplet repeat (CAG) expansion disease (p.510)
Figure 2. Protein Interaction Network for Huntington’s Disease Comprehensive PPI network for htt [huntingtin protein] Y2H interactors [35 bait and 51 prey
proteins & verified in pull down assays], red diamonds; previously published interactors, blue squares; interactors identified from databases HRPD, MINT,
and BIND, bridging any two proteins in the extended network, green triangles Htt interactors previously reported and recapitulated in our screens.
Goehler et al. Mol Cell 15:853, 2004
Marcotte Nature 402:83, 1999
www.yeastgenome.org
Papers related to questions on 2d mid-term test
They double-checked
some by RT-PCR
RNA-seq
microarray profiling
Hawrylycz et al. Nature Sept. 2012
“… a small number of high-quality, clinically unremarkable brains profiled with
DNA microarrays for quantitative gene-level transcriptome coverage”
“… RNA sequencing methods, which were costprohibitive and technologically immature when
the project was initiated, hold great promise for
elucidating …[brain] transcriptional regulation in
the future.”
Dopamine
pathway genes
Different parts
of brain
Transcriptional profiles of multi-gene family members
(in different parts of human brain)
Actin genes
different genes
different parts of brain
Hawrylycz et al. Nature Sept. 2012
Example of highly polymorphic human gene…
Factor VIII (F8) blood clotting gene
3 of its introns have microsatellites which
differ in copy number among individuals
TB: Tibeto-Burman
AA: Austro-Asiatic
Allele frequency for
microsatellite in intron 1
Number of tandem repeats (GTGTGT…) in
intron 1 varies among populations in India
Saha Haemophilia 17: e928, 2011
“The finished human genome-assemblies comprise several
hundred un-sequenced euchromatic gaps, which may be rich in
long polypurine/polypyrimidine stretches.
“In this study, we sequenced 99% of all three [such] unfinished
gaps on human chr 20..”
Barcode tags used in gene overexpression experiments
- similar in design to gene deletion
experiments discussed in Topic 7
- to track whether having too much protein X
is lethal under certain growth conditions…
Douglas et al. G3:Genes Genomes
Genet 2:1279-89, Oct. 2012