The Major Spliceosome
Introns are recognized and excised by a large molecular machine called the spliceosome that is composed of five small nuclear ribonucleoproteins (snRNPs) and additional non-snRNP protein components.26 Each of the five snRNPs contains one small nuclear RNA (snRNA) and a number of protein components. The U2-type snRNPs are U1, U2, U4, U5, and U6, the latter three of which form a U4/U6.U5 tri-snRNP. The spliceosome is formed by the sequential interactions of the snRNAs and spliceosomal proteins with the pre-mRNA substrate and with one another. In U2-type introns, the 5′ss is initially recognized by the U1 snRNP, while the BPS, PPT, and 3′ss are recognized by the protein factors SF1, U2AF65 and U2AF35, respectively, together forming the spliceosomal commitment (or E) complex. During the formation of the pre-spliceosome, or A complex, U2 snRNP replaces SF1 at the BPS. At later stages, the U4/U6.U5 tri-snRNP stably associates with the spliceosome (B complex). Rearrangements in RNA and protein interactions lead to the formation of the catalytically active spliceosome (B* complex) that catalyzes the first transesterification reaction, followed by the second transesterification in the C complex (Figure 3).
Figure 3. Spliceosome assembly. The interactions of the spliceosomal snRNPs and some selected non-snRNP protein complexes at various stages of spliceosome assembly (complexes E, A, B*, and C) are depicted schematically for both the U2- and U12-dependent spliceosomes. The Prp19/CDC5 complex is indicated by ‘19C’. Its association with the U12-dependent spliceosome is inferred from the major spliceosome and is therefore indicated with a question mark. (Adapted with permission from Ref 27. Copyright 2003 Macmillan Publishers Ltd)
Download figure to PowerPoint
Recognition of U12-Type Introns by the Minor Spliceosome
The U12-dependent spliceosome contains four specific snRNPs, U11, U12, U4atac, and U6atac, each of which contains a specific snRNA component that is equivalent to but distinct from its U2-type counterpart, i.e., U1, U2, U4, and U6, respectively. U5 snRNP is shared between the two spliceosomes. Although the sequences of the snRNAs with equivalent function are quite divergent in the two spliceosomes, they share a common overall secondary structure (Figure 4). U4atac, U6atac, and U5 associate into a tri-snRNP, similar to the major U4/U6.U5, and the protein composition of the major and minor tri-snRNPs appears to be very similar, if not identical.28 Moreover, it has also been shown that equivalent stem-loop structures of U4 and U4atac function as binding platforms for proteins that are required for the tri-snRNP formation.28,29 In contrast, while the major U1 and U2 snRNPs are distinct snRNPS, their counterparts are present in the nucleus as a U11/U12 di-snRNP.30,31 Interestingly, only seven protein components have thus far been reported to be specific to the minor spliceosomes (Table 1), and they are all located in the U11/U12 di-snRNP.32,33 U11/U12 also lacks all the U1-specific proteins and some U2-associated proteins (Table 1), making it the most divergent component of the minor spliceosome, in comparison to its counterparts in the major spliceosome.
Figure 4. The predicted secondary structures of the human spliceosomal snRNAs. The binding sites for Sm proteins are shaded in gray, and the sequences interacting with the 5′ss or BPS in cyan. Sequences involved in various U2/U6 or U12/U6atac interactions are indicated by green (helix I), purple (helix II), and yellow shading (helix III), similar to Figure 5. Nucleotide modifications are omitted. (Structures are based on data originally published in Ref 34 for U1, U2, and U5, Ref 18 for U11, Ref 35 for U12, and Ref 29 for U4, U6, U4atac, and U6atac). The locations and identities of the Taybi-Linder syndrome or microcephalic osteodysplastic primordial dwarfism type I (TALS/MOPD1) mutations in the U4atac snRNA are from Ref 36 and are indicated in red.
Download figure to PowerPoint
Figure 5. RNA–RNA interactions in the catalytic cores of the minor and major spliceosomes. Interactions between snRNAs and the 5′ss or BPS are indicated by cyan shading. U2/U6 or U12/U6atac interactions are indicated by green (helix I), purple (helix II), and yellow shading (helix III), as in Figure 4. The minor spliceosome structure is based on data published in Ref 59. The U12/U6atac helix III structure is controversial as it is not conserved in plants,67 but mutations in U12 snRNA that weaken this structure reduce the splicing activity in mammals.35
Download figure to PowerPoint
Table 1. Proteins of the U1, U2, and U11/U12 snRNPs
|12S U1||17S U2||18S U11/U12||Functions (with Selected References)|
|Sm proteins1||Sm proteins1||Sm proteins1||snRNP core components26|
|U1-A (SNRPA)|| || ||Structural; RNA-binding37|
|U1-C (SNRPC)|| || ||5′ss recognition37|
|U1-70K (SNRNP70)|| || ||Structural; SR protein interactions37,38|
| ||U2A′ (SNRPA1)|| ||Structural; RNA-binding|
| ||U2B′′ (SNRPB2)|| ||Structural; RNA-binding|
| ||SF3a complex2|| ||BPS binding39|
| ||SF3b complex2||SF3b complex2||BPS binding33,39|
| || ||20K (ZMAT5)||Unknown; homology to U1C32|
| || ||25K (SNRNP25)||Unknown|
| || ||31K (ZCRB1)||Unknown; RNA-binding40|
| || ||35K (SNRNP35)||SR protein interactions, homology to U1-70K33,41,42|
| || ||48K (SNRNP48)||5′ss recognition43|
| || ||59K (PDCD7)||Structural, binds 48K and 65K43,44|
| || ||65K (RNPC3)||Structural, binds U12 snRNA44|
| || ||Urp3 (ZRSR2)||3′ss recognition45|
| ||hPrp434||hPrp43 (DHX15)|| |
| || ||Y Box-13 (YBX1)|| |
The overall assembly pathways of the two spliceosomes are similar, and the main difference is the absence of a separate commitment complex in the minor spliceosome. Instead, the preformed U11/U12 di-snRNP binds the intron as a unit, and the 5′ss and BPS are recognized in a cooperative manner within the A complex, although U11/5′ss basepairing still precedes the formation of stable U12/BPS basepairing.47 The initial basepairing interactions at the 5′ss are also different in the two spliceosomes: in contrast to the U1, the U11 snRNA does not basepair across the exon– intron boundary or even with the first three nucleotides of the intron.2,48 Instead, these nucleotides are recognized specifically by the U11-48K protein, which likely also stabilizes the U11/5′ss helix.43,49 Surprisingly, even though U11-48K does not share significant sequence homology with the U1-specific protein U1-C, both appear to stabilize the binding of their respective snRNAs in a similar manner through their zinc finger domains.37,49
Interaction of the U12 snRNA with the BPS is similar to the U2/BPS basepairing in the major spliceosome, resulting in the exclusion of the branch point (BP) adenosine from the U12/BPS helix.50,51 There seems to be flexibility in the choice of the BP adenosine, as either one of the two A residues present in the consensus BPS can be used as a BP, depending on the intron in question.52 BPS recognition is likely stabilized by proteins of the SF3b complex, which is present in both spliceosomes (Table 1), and is known to bind to the BPS in the major spliceosome, with the protein factor p14 shielding the BP adenosine from premature activation.39,53,54 U12-type introns do not have PPTs, and U2AF is not required for the recognition of U12-type introns.55 However, a U11/U12 di-snRNP component related to U2AF35, Urp,32 is required for A complex formation and 3′ss recognition.45 Urp is not specific to the minor spliceosome, but its function is different than in the major spliceosome, where it apparently displaces U2AF from the 3′ss after the first catalytic step.45
In the major spliceosome the U1 and U2 snRNPs recognize the 5′ss and BPS independently, and non-snRNP proteins are required for their association to each other.56 In contrast, the U12-type 5′ss and BPS are linked through the internal components of the U11/U12 di-snRNP already at the earliest phase of intron recognition. The protein factors U11-59K and U11/U12-65K interact at the interface of the two snRNPs.44 U11-59K further interacts with the U11-48K protein,43 while U11/U12-65K directly binds to the 3′ terminal stem loop of U12 snRNA.44 Owing to the compact structure of the di-snRNP,57 further protein and/or RNA interactions are likely to contribute to the association of the two snRNPs. Furthermore, the binding sites for the 5′ss and BPS must be close to each other within the U11/U12 structure, as the very 5′ end of U12 snRNA can be cross-linked to the pre-mRNA 2 nt upstream of the 5′ss already in the A complex, suggesting that the BPS and 5′ss must be within 50 Å of one another.58 Thus, the 5′ end of U12 is already close to the position required for the formation of the catalytic core of the spliceosome.
Assembly of the Catalytic Core is Similar in the Two Spliceosomes
The formation of the catalytically active spliceosome is thought to follow a pathway similar to that of the major spliceosome. After initial recognition of the splice sites by the U11/U12 di-snRNP, the U4atac/U6atac.U5 tri-snRNP associates with pre-spliceosome to form complex B59 (Figure 3). A terminal stem-loop structure in the U6atac snRNA (possibly together with U4atac snRNA sequences) contains a signal that directs the U4atac/U6atac.U5 tri-snRNP to the minor and not to the major pre-spliceosome.60 A large number of structural rearrangements convert the pre-catalytic spliceosome (complex B) to the catalytic configuration (complexes B* and C) in a manner similar to the major spliceosome. U6atac snRNA replaces U11 at the 5′ss, U4atac/U6atac structure is unwound, and U12 and U6atac basepair with each other to form the ‘catalytic core’ structure in which the reactive A residue at the BP and the 5′ss are juxtaposed for the first step of the catalysis.59,61–63 During this process both the U11 and U4atac snRNAs are released from the spliceosome. In the major spliceosome, specific helicases drive forward spliceosome assembly and the transesterification reactions.64 There is virtually no data showing whether helicases have similar activities in the minor spliceosome. However, given that most proteins are shared28,32 and no helicases specific to the minor spliceosome have been identified, it is likely that the helicase activities are also the same during the assembly of the two spliceosomes.
Although the formation of the catalytic core appears to be somewhat more flexible in the U12-dependent spliceosome,61 the structure and function of the core are likely to be highly similar. Indeed, both spliceosomes employ the same two-step transesterification mechanism for intron removal,51 and U12-dependent splicing is even supported by the modified U6atac snRNAs in which the functional domain has been replaced by that of U6 snRNA.65 The U12/U6atac interactions in the catalytic core of the U12-dependent spliceosome resemble those in the major spliceosome, although the helix II structure present in the U2/U6 complex cannot be formed in the minor spliceosome (Ref 59; Figure 5). Nevertheless, the similarity of snRNA structural domains suggests that, like the major spliceosome,66 the minor spliceosome is likely to use RNA-based catalysis.
Removal of U12-Type Introns Is Slow
Both intron types are spliced co-transcriptionally within the nucleus (Ref 68 and Box 1) and might be expected to display similar kinetics. However, the removal of U12-type introns appears to be significantly slower. Early in vitro splicing experiments documented a splicing rate for U12-type introns that was three- to fivefold slower than that of U2-type introns.47,51,69 Similar observations were also made in in vivo experiments, where approximately twofold higher levels of unspliced U12-type introns were detected in the steady-state transcript pools isolated from insect and mammalian cells.69–71 This is consistent with the observation that co-transcriptional splicing of U12-type introns is at least twofold slower than that of U2-type introns.68 Thus far, the reason for the lower efficiency remains unknown. Minor spliceosome snRNPs are ca 100-fold less abundant than the major snRNPs,31,59 and this could underlie the observed kinetic differences. However, the observation that a further 10-fold reduction in the levels of U4atac snRNA has no apparent effect on the efficiency of endogenous U12-dependent splicing argues against this simplistic explanation.70 The slower rate could also be related to the kinetic effects caused by the less flexible recognition phase of U12-type introns or by the inability to form some of the structures present in the catalytic core of the major spliceosome. Regardless of the underlying mechanism, Patel et al.71 have suggested that the slower rate of splicing could constitute a rate-limiting mechanism for the expression of genes containing U12-type introns. In this model, the transcripts containing unspliced U12-type introns would get trapped in the nucleus where they could be targeted by nuclear surveillance mechanisms.72
NUCLEAR LOCALIZATION OF THE U12-DEPENDENT SPLICEOSOME
A provocative hypothesis by König et al. suggested that the U12-dependent spliceosome is located in the cytoplasm and that the splicing of U12-type introns also takes place there.73 This suggestion was based on in situ hybridization studies, mostly with zebrafish, as well as cell fractionation and reverse transcription polymerase chain reaction (RT-PCR) analyses. The publication contradicted many of the earlier studies on minor spliceosome component localization and function (See Ref 74 and references therein) and spurred a lively debate on the localization of the U12-dependent spliceosome. Subsequent publications failed to reproduce the key findings of the paper and instead demonstrated nuclear localization for U12-type spliceosome snRNP and protein components in mammalian cells and tissues,75 nuclear splicing of U12-type introns in Xenopus oocytes,76 and co-transcriptional splicing of U12-type introns.68 Together, these subsequent studies provide firm evidence for the nuclear localization of the splicing of U12-type introns.