Arthur EdwardsBroad Summer Research Program in Genomics
Cancer Program08/06/07
Genome-wide miRNA Expression Analysis in Lymphoma
miRNAs
Lymphoma
Why Lymphoma?
Source: American Cancer Society, 2006.
Men38,670
Women32,610
Deaths
Cases
Men10,370
Women9,360
Lymphoma
Hodgkin’slymphoma
Non-Hodgkin’slymphoma
7,350 patients(est. 2007)
56,390 patients(est. 2007)
Adapted from J.L Peck, SI 2007
09/01/06 – Stays at MGH
07/23/07 – Plays at Fenway
About 50% not so happy ending
Naive B-cells
AntigenMemory B-cells
Plasma cellsCentroblasts
Centrocytes
Diffuse Large B-Cell Lymphoma (DLBCL)
•DLBCL is a result of abnormal B-cell development.
•Mediastinal Large B-Cell Lymphoma (MLBCL) – Subtype of DLBCL with worse prognosis than DLBCL.
•Significant clinical and genetic heterogeneity.
Challenges to the Lymphoma Field
1. Lack of understanding of its molecular mechanisms
2. Lack of diagnostic / prognostic tools
3. Lack of effective therapeutic drugs
miRNAs
The start of something new: lin-4 and let-7
found in C.Elegans
Cell DeathCell proliferation
Differentiation
Physiological Responses
Expression Profiles Contain
Diagnostic Information
miRNAs as regulators of gene expression and biology
Some miRNAs play functional
roles in tumorgenesis particularly in
lymphoma“miR-17 cluster
cooperates with c-myc in lymphomagenesis”
Lu et. al. (2005) – Nature He et. al. (2005) – Nature Lee et. al. (1993) – Cell Reinhart et. al. (2000) - Nature
Controls n = Cell Line (HeLa) 9Cell Line (MCF-7) 9Cell Line (MOLT-4) 6Blank Controls 22Lymphoma Samples 242
288
Approach: genome-wide miRNA expression analysis
Diagnosis n = Lymphoma Cell Line 34Paired DLBCL 18Primary MLBCL 39Primary DLBCL 139Normal B-Cell 12
242
Overview of miRNA Profiling
Fig 1a. Plot of cell line controls and their relative expression values.
Quality Control of Technical Procedure
Fig 1b. Cell line controls cluster.
Molt4
MCF-7
HeLa
Plate 1
Plate 2
Plate 3
Code Category
Fig 2. Clustering of Normal B-Cell Samples
miRNA Expression Profiles Differentiate B-Cell Development
Naïve
Centroblast
Centrocyte
Memory
Code Category121
miRNA Expression Differentiates Primary Lymphoma Samples
Primary DLBCL
Primary MLBCL
Normal B-Cell
Naïve
Centroblast
Memory
Centrocyte
miR-17 cluster
Fig 3. Primary Lymphoma Samples Clustered
Code Category
miR-150
Non-Random Association of Patient Outcome
Fig 4a. DLBCL vs. MLBCL
MLBCL
DLBCL
Failure
Cure
No Data
Code Category1 2
Cluster 1
Cluster 2
Failure 20 20 40
Cure 8 27 35
28 47 75
p=0.018 Fisher’s Exact Test
Fig 4b. Correlation between Cure and Failure
Fig 5. Lymphoma Cell Lines Clustered
Lymphoma Cell Line Comparison
Hodgkin’s Lymphoma
DLBCL
Mantle Cell Lymphoma
MLBCL
miR-155
Conclusions
Classifying Primary Lymphoma Samples.
CB / CC – like lymphoma.
Naïve / memory - like lymphoma
Another class
miRNAs and prognosis.
miRNAs may give clues on molecular signatures in prognosis of lymphoma patients.
Lymphoma Cell Lines
Hodgkin’s Lymphoma have a different miRNA expression pattern
B-Cell Development
Two distinct patterns of miRNA expression.
CB / CC vs. Naïve / Memory
• Unraveling the functional aspect of miRNAs in B-Cell development and in lymphomagenesis.
• More comprehensive data analysis with additional clinical information.
• Integrative data analysis (Affy data, miRNA data, SNP data and sequence data)
Future Studies
Acknowledgements• Broad Members
– Jun Lu– Hao Zhang– Stefano Monti– Alper Uzun– Yuliya Kodysh– Arthur Liberzon– Xiaohui Xie– Jill Mesirov– Todd Golub
• Broad Summer Research Program in Genomics– Shawna Young– Bruce Birren– Maura Silverstein– Scott Breiding
• DFCI / Harvard Medical– Kunihiko Takeyama– Margaret Shipp
MIR WAY