Transcript Yeobeyondgenome_final

Intronic Splicing Regulatory Elements
Genome-wide comparative
genomics
Array-based discovery from
neural differentiation of human ES
cells
C. Carson
J. Simon
Eric Van Nostrand, Tiffany Liang
Gene Yeo, Salk Institute
Nicole Coufal, Christian
Carson, Alysson Muotri,
Xiangdong Xu, Tiffany
Liang, Rusty Gage
Understand the rules to predict (cell-type, signal-induced
specific) alternative splicing
Constitutive Constitutive
Alternative
Constitutive Constitutive
ACEScan[+] exons on UCSC browser
Yeo, PNAS 05
NMDA receptor 1
Sorek, Sugnet, Yeo
Cell-type specific AS controlled by ciselements and trans-factors
UAGG
GGGGG
UC-rich; UCUU
UGCAUG
Identification Of Intronic Splicing Regulatory Elements
161,730 exons
24.2 Mb
exonic sequence
129 Mb
intronic sequence
1. Count conserved kmers and unconserved kmers
2. Derived a chi-square score for significantly enriched and
conserved kmers
3. Clustered the kmers into motif families
156 upstream motif
clusters
158 downstream motif
clusters
Example of a downstream motif cluster:
TGCATG ,TGCATGA, ATGCATG, CTGCATG, TGCATGC, TGCATGT, TGCATGG,
GTGCATG
What are the properties of these motifs?
•Positional biases?
•Near alternatively spliced exons?
•Expression biases?
•Overlap known elements?
I. ISREs have strong positional distribution biases
76% of downstream ISREs exhibit
significant position biases
80% of upstream ISREs exhibit
significant position biases
New protocol for computing conserved
positional bias, while controlling for background
conservation
II. ISREs are enriched near alternative exons
30% downstream ISREs
56% upstream ISREs
Some resemble known binding sites
FOX1,2
Enriched near muscle alternative exons (Ares)
Nova1,2
MBNL
CELF
PTB
III. Genes containing ISRE-proximal exons are tissue-specific
Expression biases (71% downstream ISREs, 81% upstream ISREs)
IV. ISREs overlap with Exonic Splicing Silencers
ESE
Fairbrother WG, Yeh RF, Sharp PA, Burge CB (2002)
ESS
Wang Z, Rolish ME, Yeo G, Tung V, Mawson M et al.
(2004)
ISE
Yeo G, Hoon S, Venkatesh B, Burge CB (2004)
ESE, ESS Zhang XH, Chasin LA (2004)
Like ESS, do ISREs affect Splice Site Choice ?
Competing 5’ss and 3’ss reporter (Wang et al, Mol Cell, July, 2006)
Downstream ISREs
suppress intron-proximal 5’ splice sites
Upstream ISREs
suppress intron-proximal 3’ splice sites
Applications of ISREs: (1) splicing arrays --tissue
specific alternative splicing?
included
muscle
skipped
included
skipped
brain
Enriched ISREs in
downstream introns
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
TGCATG
ACTAAC
TTGGTT
TGGTTT
GCATG
TCATTTT
TTTCAT
Enriched
Depleted
Sugnet et al. PLoS Comput Biol, 2006
Applications of ISREs: (2) predicting RNA binding sites
1.
Several proteins have been reported to affect their own alternative
splicing (e.g. hnRNP A1, SRP20, SC35, TIA1, TIAR2, FOX2, PTB)
2.
Evolutionarily conserved AS exons have high intronic conservation
flanking the exon, resulting in algorithms that perform genomic
predictions of alternative conserved exons
3.
Alternative conserved exons are enriched in genes encoding RNA
binding proteins and splicing factors
Applications of ISREs: (2) predicting RNA binding sites
Applications of ISREs: (2) predicting RNA binding sites
ISREs are likely functional
• ISREs identified in mammals via comparative genomics.
• ISREs have positional biases, are enriched in tissue-specific
genes, and overlap with ESS.
• ISREs alter splice site choice in vitro.
• Some ISREs resemble known sites of known alt splicing factors.
• A fraction of ISREs are proximal to alternative exons.
• ISREs can be utilized to analyze splicing-array data.
• ISREs can be utilized to identify autoregulated exons, and has
other implications.
Alternative splicing differences in human embryonic
stem cells versus neuronal progenitors
hESC
NP from hESC
1. Developed an algorithm to
detect AS from exon arrays
(REAP).
2. Verify AS events.
3. Identify ISREs proximal to AS
events.
4. Identify RNA binding protein.
Endogeneous NP
Exon arrays have probesets in every exon
Simple representation
Exon array,
alternative splicing,
gene expression
Tiling arrays
REAP predictions agree with
A) EST-verified alternative splicing
B) ACEScan[+] exons
Confirming isoforms by RT-PCR
Discovery of ISREs proximal to ES/NP alternative exons
ISREs in ES/NP ongoing…
• REAP algorithm designed for exon array based detection of AS
events
• REAP[+] exons correlate with EST-based and ACEScan[+]
exons
• ISREs identified specific for ES/NP AS events
• FOX1/2 may regulate ES/NP-specific AS events
Alternative splicing at the Crick-Jacobs Center, Salk Institute
Computational Modeling, Integration
Constitutive
Alternative Constitutive
Stem cells, early neuronal differentiation
Cis-elements
Association of RNA binding proteins
to elements
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.