Discovery of Muscle Atrophy Gene Regulatory Network Using

Download Report

Transcript Discovery of Muscle Atrophy Gene Regulatory Network Using

Caltech Wold Lab
MicroRNA Target
Prediction Using Muscle
Atrophy Genes As Models
Qing Yuan
Mentors:
Dr. Barbara Wold
Diane Trout
Brandon King
Gilberto
Hernandez, M.D.
MicroRNA and MicroRNA
Target Prediction Programs
A. What are microRNAs?
B. What biological function or functions do they
perform?
C. With which biomolecules do they interact?
D. How do microRNA target detection programs
predict mRNA/target interaction?
E. What information do microRNA target
detection programs provide?
MicroRNAs:
Gene Regulation at the Post-transcriptional
Level
MicroRNAs are small (17 to 25 nt.) RNA molecules which regulate
gene expression by degrading mRNAs of certain genes or
interfering with translational machinery of mRNAs.
mRNA Degradation
mRNA Suppression
RISC - RNA induced silencing complex
UTR - untranslated region of an mRNA
Images from Bartel. (2004) Cell, Vol 116: 281-297
MicroRNA Target Prediction Programs rely
on MicroRNA Targeting Promiscuity
microRNA1
microRNA1
mRNA 3’ UTR
microRNA1
microRNA2
microRNA2
microRNA
1. Multiple
2.
One microRNA
microRNAs
can
bind
can target
bind
to theto3’the
UTR
3’ UTR
of an
Known
previously unknown
mRNA.
of
one mRNA.
target
mRNA1 3’ UTR
microRNA1
mRNA3 3’ UTR
microRNA1
mRNA2 3’ UTR
microRNA1
3. A single microRNA can have many distinct
mRNA targets.
MicroInspector: MicroRNA Target Prediction
Using Databases of Known MicroRNAs
3’
1
2
Region of High Complementarity
TGACGTA
AUUGCAU
5’
mRNA
microRNA
3’
Predicted Structure
of mRNA:microRNA
Complex
mRNA
microRNA
U
Output
mRNA 3’UTR
miR-A
miR-B miR-B
miR-C
5’
miR-A
miR-A miR-A
miR-B
miR-B
miR-C miR-C
Biological Interest: Muscle Atrophy
Causes:
Prolonged disuse
Microgravity
Disease
Result:
Upregulation of muscle protein
degradation genes, such as
MuRF-1 and MAFbx (ubiquitin
ligases)
--> Loss of muscle mass
NASA
Evidence of MicroRNA Involvement in
Transcriptional Regulation of Muscle Differentiation
A Model
miR-1/133 clusters
MicroRNAs regulate genes which
MEF2 - muscle-related
miR-133
miR-1
makefactor
muscle.
transcription
HDAC4 - inhibitor of
muscle differentiation
SRF
SRF - Could
myoblast microRNAs regulate
proliferation
which destroy muscle?
MyoD - myogenic
differentiation
prolif.
Myoblast
HDAC4
genes
MEF2
MyoD
diff.
Myotube
See Chen et al. 2006 for more information
Potential MicroRNA Involvement in Muscle
Degradation
miR-1/133 clusters
miR-133
miR-1
SRF
HDAC4
MuRF-1
MAFbx
MyoD
MEF2
prolif.
Myoblast
diff.
Myotube
microRNAregulated?
Could microRNA
target detection
programs be used to
identify the
microRNAs regulating
MuRF-1 and MAFbx?
Chen et al. 2006
migo: Identifying Genes with Multiple
Common microRNA Binding Sites
• Created by Diane Trout from
the Caltech Wold Lab
• Identification of microRNA
binding sites by known
microRNAs for multiple
genes
• Visualization of binding site
profile for genes using
TreeView
List of Genes:
Gene1
Gene2
Gene3
miRBase
…
miRNA 1
miRNA 3
miRNA 5
miRNA 2
…
miRNA 1
miRNA 2
miRNA 3
Gene 1
3
1
10
Gene 2
5
2
0
Gene 3
17
5
2
Microinspector
microRNA target detection
XClust
2-D hierarchical clustering
TreeView
2-D hierarchical clustering
Microinspector - Tabler Lab
XClust - Eisen Lab
TreeView - Eisen Lab
migo Visualization Problem
• Linkage analysis:
how subtrees are combined
single, average and complete
• XClust bug
identical entries not grouped together
immediately
problem avoided by using complete linkage
analysis instead of average linkage analysis
• Alternative: PyCluster
offers different types of linkage analysis; user
can avoid the bug associated with average
linkage analysis
migo Screenshot: Results Viewed Using TreeView
list of microRNAs
which target one or
multiple mRNA
transcript submitted
Gene Info (retrieved
by migo from NCBI)
migo mRNA:microRNA
binding profile
migo Screenshot: Results Viewed Using TreeView
microLink - A First Addition To Migo
MuRF1
•
It allows the user to
use genes for
positive or negative
control to select or
exclude microRNA
candidates.
•
It allows the user to
visually inspect
results and select
strong microRNA
candidates with
ease.
MAFbx
miRs
How is migo different from MicroInspector?
a list of microRNAs
and their putative
binding sites
a mRNA sequence
or a Gene ID
Disadvantage: High Number of False Positives
a list of Gene IDs
migo
a list of
microRNAs
shared between
every two genes
Helps the user select microRNA candidates
microLink Screenshot
Figure 1: a microLink
analysis for a list of genes
(GUI)
• Center is the target gene
which the user wants to
examine.
• On the peripheral are the
other genes also submitted.
• Thickness of each line
connecting every two genes
reflect the number of
microRNAs they have in
common.
microLink Screenshot (Cont.)
Figure 3: a microLink analysis in
text format (right)
Figure 4: binding site positions
and mRNA:microRNA interaction
free energy (below)
Future Work
• Complete the graphical user interface
• Improve the visualization scheme for migo
• Implement migo’s own microRNA target detection
procedures
Acknowledgments
• Caltech Wold Lab
For your guidance and
encouragement
• NIH/NSF
For supporting our summer
program
• SoCalBSI2006 faculty and
staff
For your guidance,
encouragement, subways,
brownies and much much
more…