Transcript Comprehensive Analysis of DNA Copy Number Variations and

COMPREHENSIVE ANALYSIS OF
DNA COPY NUMBER VARIATIONS
AND
GENE EXPRESSION IN OSTEOSARCOMA
Nalan Gokgoz, Atta Goudarzi, Cheryl Wolting
Jay S. Wunder and Irene L. Andrulis
Lunenfeld-Tanenbaum Research Institute, Mount Sinai Hospital
Toronto, ON, Canada.
Connective Tissue Oncology Society Meeting
November 1, 2013
Identification and Characterization of Molecular Alterations
in Osteosarcoma
High resolution approaches to identify genes and pathways predictive of outcome in OS
 Gene expression profiling by Microarray Analysis
 Interrogation of biological pathways and networks
 Identification of the most relevant biological pathways for list of
discriminative genes by Ingenuity Pathway Analysis
 Identification of the significant effectors and organizing networks in
OS metastasis by Dynamo (Taylor and Chuang)
 Investigation of Copy Number Changes by Illumina SNP array technology
 Detection and characterization of alterations
 Analysis and visualization by Genome Studio and GAP
 Identification of significant recurrent targets by GISTIC
PATIENT COHORT
High-grade Intramedullary
63 patients
A
No Metastasis at Diagnosis
46 patients
B
Metastasis at Diagnosis
17 patients
A1
No Metastasis 4 years post Dx.
(29 patients)
A2
Metastasis within 4 years Dx.
(17 patients)
MICROARRAY ANALYSIS
Outcome of the Patients Presenting with “no Metastases”
No Mets.
Mets. within
4 yrs post Dx.
4 yrs post Dx.
No Metastases 4 years post Dx (A1)
vs
Metastases within 4 years Dx (A2)
18981 cDNAs
T-statistic
p<0.001
(BrB Array Tools)
n=53 genes
for tumor classification/clustering
Statistical validation by Leave-One Out cross-validation method
Molecular validation by Real-Time Analysis
Ingenuity Pathway Analysis
Upstream Regulators
Upstream regulators that may be
responsible for observed
increase/decrease in expression
Networks
Molecule Set
Interactions and
Relationships
between molecules in
set
Gene A
Gene C
.
.
Gene X
Gene Y
Pathways with which
molecules in set are
associated
Functions with
which molecules in
set are associated
Functions
Pathways
Summary of IPA in OS metastasis
Networks
cell morphology, organization, hematopoiesis
Pathways
Rac/Rho, actin cytoskeleton
Functions
hematopoiesis, cell movement
Regulators
Fas, Fos, SP1, SREBF1
Signaling by Rho Family GTPases
A1 – No mets
A2 – Mets in 4 yrs
Lower expression in A2
Higher expression in A2
GENETIC NETWORKS in OS METASTASIS
Significant Networks
• Transport
• Translation
• Signaling
•
•
•
The PRKCε, RASGPR3 and GNB2 networks
differentially activated
DLG2 network
differentially organized
The PRKCε, RASGPR3 and GNB2 networks are potential effectors of DLG2
Protein Kinase C Epsilon and Genetic Networks in Osteosarcoma Metastasis; A Goudarzi, N Gokgoz, M Gill, D Pinnaduwage, D. Merico,
J.S Wunder and IL Andrulis, Cancer, 2013, 5, 372-403
Osteosarcoma and Copy Number Alterations
 Illumina 610-Quad Whole-genome genotyping
beadchip
 Coverage includes >14,000 CNV regions and
550K evenly spaced TagSNPs from HapMap
data
 High Resolution: Spacing 2.7 kb
 Includes markers in the unSPNable Genome
 Allows detection of SNPs, Copy Number
Variation and Genotype
 Reference Genotype: Canonical genotype
clusters (200 HapMap DNA genotype data)
 44 Osteosarcoma Tumor DNA
 25 of them with matched blood DNA
Validation by Real Time PCR
Complexity of OS Tumour Genome
(Analysis by Genome Studio)
Allele Frequency
BB
Blood DNA
LogR Ratio
Tumour DNA
AB
AA
Genome Alteration Print (GAP) Analysis
OS-2550_Chromosome 3
Recurrent Copy Number Gains in OS identified by GISTIC
(Genome Identification of Significant Targets in Cancer)
*
*
CDK4
MDM2
COL12A1
COL9A1
AF086303
q value
*
PPFIBP1 COL4A1
FGFR1OP2COL4A2
LIG4
MYR8
COPS3
NCORI
PMP22
*Same family genes
Recurrent Copy Number Losses in OS identified by GISTIC
*
q value
*
*
*Same family genes
11q14.1 deletion in a matched tumor-blood DNA
DLG2
Implication of DLG2 as a tumour suppressor in cancer
 One of the most disorganized genetic networks in metastatic OS
tumours.
 The PRKCε, RASGPR3 and GNB2 networks are potential effectors of
DLG2
 Tumour suppressor function of dlg2 in Drosophila
 Scribble complex (SCRIB, DLG1-4 and LGL1/2) deregulation in
Prostate Cancer
 DLG2 implicated in Wilms Tumour
DLG2 (discs, large homolog 2)
Channel associated protein of synapse 110
Chromosome 11q14
Member of the membrane-associated guanylate kinase (MAGUK) family.
• PDZ domains; interaction with signalling proteins at postsynaptic sites
• SH3 domains are found in proteins of signaling pathways regulating the
cytoskeleton and regulate the activity state of adaptor proteins and other
tyrosine kinases
• GuKinase Domain; catalyzes ATP-dependent phosphorylation of GMP to
GDP
Gene: 2 MB, 33 alternative spliced transcripts
Longest transcript :3.7KB, 26 Exons
• Expression site: Brain, hypothalamus
Relative Expression of DLG2 in OS tumours and cell-lines
0.350
0.300
DLG2/STAM2
0.250
0.200
0.150
0.100
0.050
0.000
OS Tumours
Deletion of DLG2 gene detected by SNP array
Work in Progress
 SiRNA Knockdown of the DLG2 Gene in U2OS
70% knockdown at 72 hours
140
120
100
80
60
40
20
0
48
72
96
 The effect of DLG2 knockdown in
 Cell viability and growth by XTT assay
 Migration by scratch assay

Sequencing of DLG2 gene for inactivating mutations
CONCLUSION
We identified a 53-gene expression signature that may predict outcome
of OS patients with localized tumours.
High-resolution approaches identified candidate pathways and networks that
may be biologically relevant in OS.
 Cell morphology and organization pathways may be involved in OS
metastasis.
A large number of chromosomal aberrations were detected in OS tumours by
SNP array technology.
The DLG2 gene that is deleted in 20 percent of the OS cases and belonging
to a significantly disorganized metastatic OS network and was chosen for
further functional analysis.
Further experiments will be performed to investigate the functional role of
DLG2 in cell growth, proliferation and migration.
Acknowledgement
I. Andrulis
S. Bull
J. Wunder
R. Parkes
Andrew Seto
Andrulis and Wunder Lab
R. Kandel
Mount Sinai Hospital
Orthopedic Surgeons
Hospital for Sick Children
D.Malkin
Vancouver General Hospital
C.Beauchamp
University of Washington
E.Conrad III
Royal Orthopedic Hospital
R.Grimer
Memorial Sloan-Kettering
J.Healey
Mayo Clinic
M.Rock/ L.Wold
During progression from tumour growth to
metastasis, specific integrin signals enable
cancer cells to detach from neighbouring
cells, re-orientate their polarity during
migration, and survive and proliferate in
foreign microenvironments. There is
increasing evidence that certain integrins
associate with receptor tyrosine kinases
(RTKs) to activate signalling pathways that
are necessary for tumour invasion and
metastasis. The effect of these integrins
might be especially important in cancer
cells that have activating mutations, or
amplifications, of the genes that encode
these RTKs.