Transcript CST-LINCS data and analyses.ppsx

QUANTITATIVE POSTTRANSLATIONAL SIGNATURES ACROSS MULTIPLE
SIGNALING SPACES IN 54 LUNG CANCER CELL LINES (LCCLS) AND 6
CANCER LINES (CCLS) TREATED WITH TYROSINE KINASE INHIBITORS (TKIS):
The CST-LINCS Data
Nicolas Fernandez1, Peter Hornbeck2, Avi Ma’ayan1, Bin Zhang2, Klarisa Rikova2,
and Mark Grimes3
1BD2K-LINCS
DCIC, Department of Pharmacology and Systems Therapeutics, Icahn School of Medicine
at Mount Sinai, New York, NY 10029 2Cell Signaling Technology (CST), 3University of Montana
DCIC grant number: U54-HL127624-01, -02
Initially described at the “PostTranslational Regulation of Cell
Signaling” Meeting,
Salk Institute for Biological Studies
August 5, 2016
© 2015 Cell Signaling Technology, Inc.
The data presented in this talk are from experiments performed
at CST by Klarisa Rikova and the MS Discovery Group, and
analyzed by members of the BD2K LINCS DCIC consortium
funded by NIH U54-HL127624-01,-02.
GOALS:
Identify phospho (p), acetyl (ac) and methyl (me) PTMs that
cocluster and may be functionally linked in lung cancer cell
lines
Identify and analyze differences between SCLC and NSCLC
PTM signatures
Identify p[ST], acK, meK, and meR sites that may be
regulated downstream from pY signaling in CCLs with known
driver mutations
Note that only the pTyr (pY), pSer (pS), and pThr (pT)
results are (briefly) analyzed here.
However, the downloads will include all the
modification types sampled in the TKI experiments:
p[STY], acK, meK, and meR.
The LCCL data will be made available closer to the
time of its publication
Two sorts of MS2 TMT studies on CCLs:
I.
“Lung Cancer Cell Line” (LCCL) study of 45 lung cancer cell
lines:
• 12 SCLC and
• 33 NSCLC
I.
“Drug Treament” (TKI) study:
• 6 cancer cell lines with known disease drivers: 5 NSCLC,
one gastric carcinoma
• Three TKIs
• Treatment times 1-24 hrs
OUTLINE
Overview of TMT methodology and strategies
Introducing “Clustergramer” for interactive clustering of the CSTLINCS datasets
Accessing downloads of quantitative phospho, acetyl, and methyl
datasets from PSP.
Tandem Mass Tag (TMT) overview:
Sequential purification of different PTMs prior to TMT 6-plex analysis
Proteolysis
TMT label
Mix
6 isobaric
tags
Normal
mix
Cell Line
Cell Line 11
Treatment
Cell Line 22
Treatment
Cell Line 33
Treatment
Cell Line 44
Treatment
Cell Line 55
Treatment
Sequential Immunoaffinity Purification
LC/MS/MS
Quant Values
Log2-Ratios
•
•
TMT quant values  log2 [exp values/ref value].
Apply centering and normalization procedures similar to those
used to graph DNA microarray data where appropriate.
I. Clustergram dynamic visualization of CST-LINCS “Cell Line” study
click cell line to
order by cell line
click to cluster
by mutation
order by rows
& columns
Gene
search
sliders
click URL to open pY or p[ST] Clustergram for Cell Line study :
https://maayanlab.github.io/cst_drug_treatment/lung_CL_phos_ratios_Y
https://maayanlab.github.io/cst_drug_treatment/lung_CL_phos_ratios_ST
I. Subset(s) of ser/thr pSites in lung cancer cell lines with
KRAS mutations appear to be differentially regulated
KRAS CLUSTER
NSCL
C
A cluster of pS/T sites that appear to be consistently
upregulated in NSCLC cell lines with KRAS mutations
I. RB1 clusters of pS/T sites tend to segregate with
histological type...
SCLC
RB1 CLUSTER
RB1
...SCLC are of neural origin, and many of the proteins whose
pSites cluster in SCLC play roles in neural cell biology
I. S/T pSites whose expression levels cluster in SCLC vs. NSCLC
SCLC CLUSTER
RB1
Can these differences between the level of of phosphorylation be explained by
differences in the amount of protein in the cell?
differences in the levels of of phosphorylation cannot be explained
(entirely) by differences in the amount of protein in the cell...
18
I. The S/T kinase signatures of heavily phosphorylated
sites in NSCLC differ from those in SCLC cell lines
NSCLC
SCLC
NSCL
C
SCLC
NSCL
C
SCLC
II. Drug treatment study:
TKIs, cancer cell line, and driver TK mutations
KRAS
PDGFRA
ALK-EML4
II. TKI study caveat:
complex patterns of expression of RTKs renders
interpretation of results challenging
expression levels are based
on z-scoring the gene
expression value across all
~1000 CCLE cell lines.
II. TKI study: kinetics of changes in tyrosine phosphorylation
URL to for pY Clustergram:
https://maayanlab.github.io/cst_drug_treatment/phos_ratios_all_treat_Tyrosine
II. TKI study: changes in S/T phosphorlyation following TK inhibition
II. TKI study
URL for pS/T Clustergram:
https://maayanlab.github.io/cst_drug_treatment/phos_ratios_all_treat_ST
II. S/T phosphorylation: three distinct kinetic response types(1–24 hrs)
indicate distinct S/T kinases involved in each kinetic type
Prot Types:
epigenetics
Total: N = 505
DNA-repair
GEFs/GAPs
transcription
Type I
hi > lo
Type II
lo > lo
Type III
hi > hi
**
**
*
ACCESS TO ANALYTICAL TOOLS & DATA:CLUSTERGRAMMER
Clustergram visualizations (b versions) of the CST-LINCS data, under
development by Nick Fernandez in the Ma’ayan lab, are available at:
LUNG CANCER CELL LINE STUDY:
Clustergram of pY PTMs:
https://maayanlab.github.io/cst_drug_treatment/lung_CL_phos_ratios_Y
Clustergram of p[ST] PTMs:
https://maayanlab.github.io/cst_drug_treatment/lung_CL_phos_ratios_ST
TKI DRUG TREATMENT STUDY
Clustergram of pY PTMs:
https://maayanlab.github.io/cst_drug_treatment/phos_ratios_all_treat_Tyrosine
Clustergram of p[ST] PTMs:
https://maayanlab.github.io/cst_drug_treatment/phos_ratios_all_treat_ST
ACCESS TO DATA: Quantitative drug-study CST-LINCS data will be
Available the week of Sept 1, 2016 in PSP & will be announced in the
“WHAT’s NEW” section of the homepage:
ACCESS TO DATA: The PSP download page and additional
datasets of interest to the BD2K community
LINCS-BD2K DCIC
LINCS-BD2K DCIC: DATA SCIENCE RESEARCH
COMPONENT: to integrate multiple signaling spaces (e.g.,
transcriptomics, proteomics) to identify regulatory nodes within
multidimensional networks.
CST: an External Data Science Research (eDSR) resource from
2014-2015. Project description: Analysis of Multiple Signaling
Spaces to Identify Potential Disease Drivers and Network
Connectivity in Lung Cancer Cell Lines.
• Specific Aim 1: Analyze Multiple PTMs from NSCLC and
SCLC cell lines to Identify multidimensional Network
connections and novel Drivers in Lung Cancer Cell Lines.
• Specific Aim 2: The Curation of CST-LINCS data and
Computational Access to Cancer TMT Datasets via
PhosphoSitePlus.
ACKNOWLEDGEMENTS
BD2K LINCS DCIC : U54-HL127624-01,-02
Mt Sinai LINCS DCIC group:
Nick Fernandez
Neil Clark
Avi Ma’ayan
CST LINCS DCIC eDSR Group:
Klarisa Rikova
Bin Zhang
Beth Murray
Vidhisha Nandhikonda
Mark Grimes, Univ Montana
Sean Beausoleil
Stephen Brinton
Michael Comb for his support of PSP