Transcript PPT - NIH LINCS Program

Accelerated Protein Signaling Signatures:
Highly Multiplexed Assays to Monitor Perturbations of
Serine/Threonine Phosphosignaling
Jacob D. Jaffe1, Michael MacCoss2
1Broad
Institute, Proteomics Platform, Cambridge MA
2Department of Genetics, University of Washington, Seattle WA
Phospho-signaling
q
Gene Expression
• q is large (hopefully)
• Phospho-signaling is inaccessible through expression profiling
• Phospho-signaling can be acute or sustained
Phosphoproteomics: current developments
Phosphosite.org database (CST)
Non-redundant sites:
97,222
Non-redundant proteins:
13,384
Sites curated from literature:
94,031
Sites using site-specific (SS)
methods:
10,006
Sites using only discoverymode MS (MS) methods:
86,378
Sites using both SS and MS
methods:
4,773
• There are a lot of phosphosites! ( > # genes)
• How can we study these systematically?
Interrogation of extant CMAP Data
ATP-competitive
kinase inhibitors
Perturbations
staurosporine:MCF7
sanguinarine:MCF7
sanguinarine:HL60
cardiovascular
agents
PPAR
agonists
PO4
• No DNA/RNA involved
• Kinases and phosphatases are the key
regulators
• Therefore, perturbagens that modulate
kinase/phosphatase expression or activity
should have effects on phosphosignaling
Kinase/phosphatase genes
TK inhibitors
tyrphostin AG1478:MCF7
tyrphostin AG825:MCF7
gefitinib:HL60
imatinib:MCF7
imatinib:PC3
digitoxigenin:HL60
digitoxigenin:MCF7
digitoxigenin:PC3
digoxigenin:HL60
digoxigenin:MCF7
digoxin:HL60
digoxin:MCF7
helveticoside:HL60
helveticoside:MCF7
helveticoside:PC3
lanatoside C:HL60
lanatoside C:MCF7
HDAC Inhibitors
trichostatin A:PC3
trichostatin A:MCF7
valproic acid:MCF7
valproic acid:HL60
valproic acid:PC3
valproic acid:ssMCF7
valproic acid:SKMEL5
Step 1: Discovery and learning
2.8x ↑
Digest
Control Tx
‘Light’ Cells
Tx 1
Fractionate
‘Medium’ Cells
Tx 2
‘Heavy’ Cells
Mix
Phosphopeptide
Enrichment
2.4x ↓
Mass Spectrometry
• Cells are colored by isotopic labels (i.e., 13C, 15N, but not radioactive)
• Generic technology enriches all phosphopeptides
• However, most phosphosites are Ser or Thr and NOT Tyr
• Ser/Thr phosphorylation is low hanging fruit
• Mass Spec provides both identification AND quantification
We propose to do for phosphosignaling what the
Broad LINCS group has done for gene expression
A.
Cell Lines/Conditions
B.
Example Coherent Cluster
5
C.
Select
Representative
Member(s)
Cluster avg.
4
log2 fold change
Phosphosites
Phosposites
3
Expert
Criteria
2
1
0
-1
-2
-3
-4
-5
Conditions
• Natural synergy between projects
• Exploit existing robust methods
FNHM(pS)QQGPR
LLWIDA(pT)AGGNK
...
Signal
Step 2: Equivalent of L1000 – the “P100”
Assay time
•
•
•
•
•
Use synthetic peptide internal standards for better quantification and proof of ID
LOD/LOQ /copies per cell
When you want to guarantee you measure it each and every time!
Next-gen instruments will make this even more selective
May enable us to skip phosphopeptide enrichment altogether
What should we see?
Protein
Copy
#/cell
1,000
1,000
1,000
10,000
10,000
10,000
100,000
100,000
100,000
Phosphorylation
Stoichiometry
1%
10%
50%
1%
10%
50%
1%
10%
50%
# cells req to
see phospho
(250 amol)
1.51E+07
1.51E+06
3.01E+05
1.51E+06
1.51E+05
3.01E+04
1.51E+05
1.51E+04
3.01E+03
# cells req to
see protein
(250 amol)
1.51E+05
1.51E+05
1.51E+05
1.51E+04
1.51E+04
1.51E+04
1.51E+03
1.51E+03
1.51E+03
• Assays will be constructed such that we will always monitor the phospho- and nonphospho-states of the site as well as a different peptide to serve a surrogate for total
protein levels.
End result
• ~100-plex phosphosite MRM-MS assay
– 60-90 minutes/sample
– $100-200/sample
• Reduced representation suitable for signature generation
• Requirements compatible will low cell numbers or tissue
samples
• Absolute stoichiometry on every site, every time
Step 3: Standardize and Disseminate
LINCS Repository
Other public databases
LINCS Member Labs
• Standard software platform (MacCoss Lab, U. Wash.)
• Cross-laboratory reproducibility
Call for nominations!
• Perturbations
– Exploit extant CMAP data
– Look for kinase and phosphatase modulators
– Can be small molecule, shRNA, or “other”
• Systems
– Relevant cell lines / disease models
– Should cover “signaling space”
• Cancer signaling
• Immune Signaling
• Cell cycle
Acknowledgements
• LINCS Program and Program Officers
– U01 CA164186-01/Jaffe
• MacCoss Lab, Univ. of Washington
– Brendan MacLean
• Broad Institute Proteomics Platform
– Philipp Mertins
– Steve Carr
• Broad Institute LINCS Centers
– Todd Golub
– Aravind Subramanian