Arabidopsis Intron Splice-Site Exceptions


[line]

This page gives information on exceptions to Arabidopsis splice-site consensus sequences and was kindly supplied by Gordon Simpson, The John Innes Centre, Norwich, United Kingdom. Also available, on separate TAIR pages, are tables of Arabidopsis Intron Splice-Site Consensus Sequences and Mutations.

Gordon is collating and updating Arabidopsis exceptions to the GT-AG rule and would be grateful if you could inform him of any that you find.

UPDATED ON 9 OCTOBER, 1997

The splice site sequences of the vast majority of introns in Arabidopsis mRNA genes fit the canonical GT-AG - U2-dependent consensus borders (where GT refers to the terminal residues of the intron at the 5' splice site and AG refers to the terminal residues of the intron at the 3' splice site). However, there are rare examples of introns possessing distinct non-canonical splice site sequences. If an Arabidopsis gene you have sequenced does not fit the U2-dependent GT-AG class you may have found something novel and important - the following information is a pointer to exceptions.

Two intron classes processed by two classes of spliceosome
Introns are excised from pre-mRNA by a dynamic multicomponent complex called the spliceosome. It has recently emerged that there are at least two compositionally and functionally distinct classes of spliceosome: the major class processes U2-dependent introns, while the recently identified minor class processes non-canonical U12-dependent introns. This terminology derives from UsnRNA components (U2 and U12 snRNA) that are specific components of each spliceosome class.

Background
The importance of the identity of the terminal intron residues
Until very recently almost all research into pre-mRNA splicing had centred on U2-dependent introns possessing the canonical GT-AG splice sites - this is not surprising, since these are far and away the most frequently occuring intron class. The terminal G residues of such introns play a key role in the splicing reaction. When either G is mutated, splicing to these sites is blocked. However, splicing can be restored to some extent if particular mutations are made at both terminal positions: A-C and A-A terminal residues work the best. This suppression of mutant defect is indicative of either a direct non-Watson-Crick interaction between the terminal residues taking place within the spliceosome or the action of an unknown spliceosomal component(s) sensitive to the identity of both terminal residues.

Previous reference to AT-AC class introns and unified view of two mechanistic classes
Recent work has led to the rapid characterisation of the processing of a class of introns with non-canonical splice sites - widely referred to as AT-AC class - that possess AT-AC borders in place of the more typical GT-AG and which, in addition, possess sequence distinctions in the wider splice site and branch site consensus sequences. These introns are processed by the U12-containing minor class spliceosome, but the overall reaction mechanism resembles that of the major class. Further characterisation has now revealed that the AT-AC terminology is misleading - rare, but naturally occuring introns with divergent terminal residues have been identified. The specificity in their processing is governed by the wider splice site and branch site consensus sequences and not by the identity of the terminal residues.

Table 1: Consensus sequences for Intron classes

5' splice site branch site 3' splice site
Mammalian
U2-dependent		 GURAGU UNCURAC (Yn) YAG

GC


A
C
U12-dependent GUAUCCUU UCCUUAAC YCCAG

A
C

A
G

Table 1. The consensus sequences derived for splice site and branch sites in U2 and U12-dependent introns. Alternative terminal residues that have been found in naturally occuring animal genes are included.

Analysis of the small number of U12-dependent introns described indicates that the branch site may normally be closer (9-19nt) to the 3' splice site than in U2-dependent introns (typically 18-40nt), but exceptions may exist.

Non canonical Arabidopsis introns identified so far
Most introns in Arabidopsis are U2-dependent and possess GT-AG borders. Some U2-dependent introns possess the sequence GC-AG - this is the most frequently identified non-canonical site in Arabidopsis. Introns bearing the properties of the non-canonical U12-dependent class, that possess a variety of terminal residues, are also present in Arabidopsis.

Table 2: Candidate Arabidopsis U12-dependent introns
Termini Intron Accession
A-A intron 7 [Wu et al. (1996) Nat Genet 14, 383-384]
A-C intron 6 U53856

intron 14 D45415

intron 14 U76670
G-G intron 9 U76670

intron 9 D45415

Further non-canonical splice sites?
While U2-dependent and U12-dependent introns are processed in distinct spliceosomes, the overall reaction mechanism is similar. In contrast, the processing of an intron with non-canonical splice sites in the HAC1 gene of yeast proceeds via a novel mechanism. HAC1 is involved in the Unfolded Protein Response (UPR) of yeast. Splicing is initiated upon activation of Ire1p, a transmembrane kinase that lies in the ER and/or inner nuclear membrane. Ire1p is also a site specific endoribonuclease that cleaves HAC1 mRNA specifically at both splice junctions. Cleaved ends can be ligated by tRNA ligase (Sidrauski C, Walter (1997) Cell 90: 1031-1039). The distribution of this kind of pre-mRNA processing in nature is presently unknown. We are currently investigating the alternative splicing of an Arabidopsis intron that does not appear to conform either to the U2 or U12-dependent classes (Macknight, R et al (1997) Cell 89, 737-745).

Take-Home Message
The splice sites of the introns in Arabidopsis mRNA genes typically conform to the GT-AG consensus of U2-dependent introns. However, on rare occasions they may not and currently, a GC at the 5' splice site is the most common exception. If an intron you are characterising does not fit this consensus you should check for the consensus sequences of the U12-dependent class: the 5'splice site sequence appears to be highly conserved. If the sequence of the intron differs from the properties of both these classes you may have something novel and interesting. I intend to collate and update Arabidopsis exceptions to the GT-AG rule and would be grateful if you would inform me of any that you find: gordon.simpson@bbsrc.ac.uK.

Further Information
Pierre Rouzé and Søren Brunak, authors of NetPlantGene, a program that identifies Arabidopsis introns, are also collecting information on introns with non-canonical splice sites in order to develop a program that will identify them in Arabidopsis gene sequences as well.

Up to date information of the U2 and U12-dependent classes of introns is available in Dietrich, RC., Incorvaia, R and Padgett RA (1997) Molecular Cell in press.

I have recently reviewed splicing in higher plants: Simpson GG and Filipowicz W (1996) Splicing of precursors to mRNA in higher plants: mechanism, regulation and sub-nuclear organization of the spliceosomal machinery. Plant Mol Biol, 32, 1-41. I am currently studying the role of RNA processing in regulating Arabidopsis flowering time in Caroline Dean's lab at the John Innes Centre.

Acknowledgments
Thanks to Pierre Rouze and Rick Padgett for helpful information.

Return to TAIR Home Page | Questions/Suggestions