Alternative Splicing

AGRY 600 2014

What exactly are we talking about?

Constitutive splicing = all possible introns removed

Alternative splicing (AS): a single gene can encode many messages depending on how the initial transcript is spliced

~95% of human genes and >60% of plant genes

~15% of human genetic disease are errors in AS

Some molecular mechanisms governing alternative splicing have to be highly coordinated

Exon skipping/inclusion

Alternative 3’ splice site

Alternative 5’ splice site

Mutually exclusive exons

Intron retention

Human Plants

>40% 8%

18.4% 15.5%

7.9% 7.5%

~29% ~29%

<5% 40%

Difficulties making these estimations

Intron retention probably overestimated because many IR events have very low sequence coverage

Most cases of IR are predicted to undergo nonsense-mediated decay (more in a minute) but do not.

These transcripts may be stored for later use (e.g., Marsilea) or to prevent export to the cytoplasm (neurogenesis)

Exon skipping/inclusion

Alternative 3’ splice site

Alternative 5’ splice site

Mutually exclusive exons

Intron retention

Human Plants

>40% 8%

18.4% 15.5%

7.9% 7.5%

~29% ~29%

<5% 40%

Hutchinson-Gilford Progeria Syndrome in humans (premature aging from alternative splicing of nuclear lamin A gene)

Sticky rice

Firmness of strawberries

Increased detection of alternative splicing has come from much deeper sequencing of

transcriptomes• Detection of low abundance products• Identifying reads specific to alternate splicing

isoforms, especially when multiple isoforms can arise from the same gene

• BUT…

Challenges to using short-read sequencing technology to study alternative splicing

• Detecting low abundance products can be difficult if there are very HIGH levels of some transcripts– Particularly a problem in plants, so normalized sequencing

libraries are essential• Reads specific to alternate splicing isoforms can be found

but they are typically rare and need to be found in sufficient depth to be believable

• Inability to read across multiple splice sites, making quantification very difficult

• Higher error rates• Computational challenges (including that software

designed for animal data does not transfer well to plants and vice versa

Stable interaction confirms accuracy of splice site choice

Splicing is cotranscriptionalAs a result, subject to chromatin configuration changes, presence/absence of other proteins on the DNA

Histone modification patterns, nucleosome positioning and DNA methylation patterns impact alternative splicing

Mutations that alter these things change alternative splicing patterns

Transcripts that are elongated at a rapid rate are more prone to exon skipping than transcripts elongated at slower rates, possibly because there is more time for correct positioning of the spliceosome

Stable interaction confirms accuracy of splice site choice

Large number of splicing factors— “spliceosome”Whole genome sequences facilitate description of putative plant orthologs of spliceosome core components, so some parts of machinery the same (and cis-acting 5’, 3’ and “branch point A” signals are similar)

However, most animal introns cannot be processed in plants and vice-versa

In plants, at least 5 ribonucleoproteins, but many (>300) Ser/Arg-rich proteins (“SR Proteins”) that show differential expression during development

Suggests development specific regulation of AS, as observed in animals and yeast

Many are duplicate factors (whole genome duplications, etc.), so many mutations in SR proteins are non-lethal, and divergence of function is seen frequently

To understand regulation of AS, need to integrate the impacts of:• Chromatin landscape changes (DNA methylation, nucleosome positioning, histone

modification)• RNA structural features (secondary structures and folding)• Splicing regulatory sequences (binding sites of splicing factors)

Variations on sequencing:

Chromatin landscape Bisulfite sequencing (turns unmethylated C to Uin cDNA unmethylated C read as T, 5meC read as C)DNAseI-seq (treat chromatin with DNAseI to identify open chromatin, size select map to genome and cluster to see hypersensitive sites)Micrococcal-nuclease sequencing (treat chromatin with MNase to remove DNA not wrapped around nucleosomes recover DNA that IS around nucleosomes, sequence and map to genome)Chromatin-immunoprecipitation sequencing

Pre-mRNA/mRNA structure* Selective 2’-hydroxyl acylation analyzed by primer extension (SHAPE) sequencing PARS sequencing Double-stranded RNA sequencing Single-stranded RNA sequencing Fragmentation sequencing

Targets of RBPs RIP sequencing HITS-CLIP Individual-nucleotide-resolution UV-cross-linking and affinity purification Photoactivatable-ribonucleoside- enhanced CLIP

Consequences of Alternative Splicing

• Creating novel protein isoforms• Changing mRNA stability by causing Nonsense Mediated

Decay• Changing mRNA stability by changing either production of or

interaction with miRNA

Adjacent 3’ Splice sites (“tandem acceptors”, NAG,NAG!)

• NAGNAG• Alternative use of one or the other adds or deletes one amino

acid to a protein (typically a polar amino acid)• In human and mouse, 20% of alternative splicing events that

preserve reading frame are because of NAGNAG AS, and 25% of these (so 5% overall) are tissue-specific alternatives. – This tissue-specific subgroup are highly conserved across mammalian species

• In Arabidopsis, ~7000 introns in >5000 genes have tandem acceptor sites (mostly in DNA binding proteins and SR proteins involved in splicing) and these are regulated in organ-specific and stress-specific ways– Arabidopsis U1-35K alternative use of a tandem acceptor ±Gln124. The

isoform lacking Gln shows altered binding affinity for SR proteins and U11 snRNA (Lorkovic et al., 2005)

– AS at a tandem acceptor site of ZINC-INDUCED FACILITATOR-LIKE1 produces two mRNA isoforms. One encodes a plasma membrane protein that functions in auxin-regulated processes. The second isoform encodes a truncated protein that localizes to the tonoplast membrane and functions in drought tolerance (Remy et al., 2013)

Alternatively spliced transcripts can encode proteins with entirely different functions and expression patterns

Alternatively spliced trancripts where one version is sensitive to a miRNA and the other is not

AS and Nonsense Mediated Decay

mRNA degradation when translation termination is perturbed. eRF1 and eRF3 are proteins that move into ribosome to recognize a stop codon. Interact with poly(A) binding protein

Disrupting that interaction recruits mRNA degradation machinery. So, a premature stop codon too far upstream.

UPF1, UPF2, and UPF3 are conserved core proteins of the NMD pathway, and their depletion results in the inactivation of NMD. They recruit mRNA decay enzymes to prematurely terminating mRNAs.

AS and Nonsense Mediated Decay

In mutants where NMD is defective…

• ~18% of all intron coding genes are regulated by NMD enriched to 25% of the alternatively spliced genes

• SO… some alternatively spliced genes are sensitive to NMD and some are not– Among those that are not, the majority of those that retain an intron

(the largest class in plants), possibly WHY they are the largest class in plants

– Genes with other versions of AS are sensitive to NMD

• Interestingly, some key splicing factors are themselves controlled by alternative splicing– AtSR34a (and a dozen others) , AFC2 (a kinase that phosphorylates

splicing factors), polypyrimidine tract binding proteins, etc.

Spliceosomes are complicated multiprotein

machines(hundreds of proteins)

AU2AF

Exon 1

U1snRNP

RS70K

RSSF2

U2AF35RS

SF1

Exon 2

SRINTRON is the unit of recognition mechanism. Complex forms through stabilized protein interactions across the intron

SRSR

Intron Definition

Exon

U1snRNP

RS70K

RSSF2A

U2AFU2AF35

RS

SF1

SR

SRSR

Exon Definition(Cote, Univ. of Ottawa)

Two models for splice site recognitionIntron/Exon boundaries are recognized in pairs through interaction with multiple proteins either across exon or intron

EXON is the unit of recognition mechanism. Complex forms through stabilized protein interactions across the exon. Excises out the flanking introns

Differential size distributions of exons (~50 to 300 nt) vs. introns (<100-100,000 nt)

• SR protein - preferentially binds to exon sequences - mark the 5’ & 3’ splicing sites in conjunction w/ U1 & U2 during transcription

• hnRNP - heterogenous nuclear ribonucleoproteins (twice the diameter of nucleosome) - consists at least eight different proteins - compacts introns, thereby masking cryptic splicing sites - preferentially binds to introns, but also bind to exons, although less frequently

Why are exons preferentially recognized?

GUGU AG

GUGU AG ESE ISS

U2AFU1

U1

On rate favored because interaction is stabilizing

Off rate favored because no stabilizing interaction or it is blocked

cryptic 5’ ss

U2AFU1SR

U1

hnRNP

Cryptic splice sites exist adjacent to strong hnRNP binding sites (splicing inhibitors)

2. SR proteins enhances this interactionESE: exonic

splicing enhancer

Regulation of alternative splicing involves the specific stabilization or destabilization of splice site recognition

Stabilization: exon inclusion

Destabilization: exon skipping

Alternative splicing alters miRNA interactions

Regulation of Alternative Splicing• Not enough information in GU and AG alone (too short)

• Most work is on cis-acting sequence elements in animals, best understood for control of exon skipping (Chasin, 2007; Chen and Manley, 2009; Barash et al, 2010).

– Both EXONIC splicing enhancers/silencers and INTRONIC splicing enhancers/silencers (Huelga et al., 2012)

– Often 5-10 nt sequences that occur in clusters and either enhance or suppress use of a splice site

– In some cases these are RNA 2° and even 3° structures rather than specific sequences, sometimes binding metabolites (“riboswitches”) sometimes acting as sensors of environmental change (temperature, salinity, etc,)

• One case in Drosophila with 95 alternatively spliced exons where competing RNA structures control mutually exclusive exons (May et al 2011)

Regulating AS allows response to environment quickly WITHOUT having to induce transcription

• Abiotic stress responses

– Arabidopsis HsfA2 (a heat shock factor) induced transcriptionally at 37°C AND a 31 bp intron left in the transcript (premature stop codon and NMD– “HsfA2-II”)

– At 42°C, a shorter alternatively spliced form is made (HsfA2-III) that encodes a protein that goes into the nucleus and binds its own promoter as a positive regulatory signal

– At 45°C, more HsfA2-III than HsfA2-II

– Rice DREB2B (heat and drought response) includes a 2nd exon with a premature stop codon. Upon high temperatures, this exon is removed and functional protein is produced.

Regulating AS allows response to environment quickly WITHOUT having to induce transcription

• Abiotic stress responses

– Genes encoding splicing factors are, themselves, alternatively spliced and stress signals affect both the phosphorylation and subcellular localization of Arabidopsis SR and splicing-related proteins

– AtSR30 isoform containing intact ORF increases at elevated temperatures and high light, while an unproductive isoform (premature stop codon) decreases at elevated temperatures (Filichkin et al., 2010).

– Dehydration stress and heat stress increase production of transcripts encoding full-length AtSR45a protein that contains a phosphorylation site required for function, relative to other splice variants that do not Gulledge et al., 2012).

Regulating AS allows response to environment quickly WITHOUT having to induce transcription

• Biotic stress responses

– Resistance responses are metabolically expensive

– Some indirect evidence that premature stop codons may be preferentially spliced out of resistance genes in response to virulence determinants on the pathogens.

Regulating AS during development

• Organ- or tissue-specific alternative splicing

– Ubiquitously expressed YUCCA4 transcript isoform (auxin biosynthesis) encodes a protein that localizes to the cytoplasm, but a second, flower-specific isoform codes for a protein localized at the endoplasmic reticulum (ER)

– Mentioned ZINC-INDUCED FACILITATOR-LIKE1 (ZIFL1) earlier leads to two H+-coupled K+ transport proteins with different tissue distribution and subcellular localization (Remy et al., 2013).

• Full-length ZIFL1.1 is targeted to the vacuolar membrane of root cells and influences cellular auxin efflux and polar auxin transport in roots.

• ZIFL1.3 transcript uses alternative 3′ splice site located two nucleotides downstream of the authentic 3′ splice site, causing a frameshift and premature stop codon that removes last two membrane-spanning domains and prevents localization to the plasma membrane of leaf stomatal guard cells

Regulating AS during development

• Many genes that control development are alternatively spliced

• Multiple isoforms of these genes detected AND there are developmental responses to mutations that disrupt Alternative Splicing:

– Flowering time

– Circadian Clock

– Differentiation of cell types

– Organ number

– Cell-cell interactions

– Endosperm-embryo interactions

Regulation of AS in response to gene duplication

• MANY different trans-acting factors (SR proteins and hnRNPs) and MANY interactions, still poorly understood

• Genes with higher rates of transcription elongation have more exon skipping (slower transcription better at allowing spliceosome to form properly)

• Large gene families tend NOT to be alternatively spliced (may be little in the way of new functions gained), but small gene families (even just duplicates) are alternatively spliced much more than singleton genes

Is Alternative Splicing Biologically Meaningful or just noise in the system?

• For most genes that undergo AS, one major splice form makes up >90% of the transcript for a gene

• Conservation of alternative splicing patterns is relatively rare (only about 10% from mouse to human) but some exon skipping forms can be conserved across fungi, plants and animals

• Rapid divergence in AS may be contributing to organogenesis and speciation events