REVIEWS
|fo)rna processing and modifications
f
Mechanistic insights into precursor messenger RNA splicing by the spliceosome
Yigong Shi'-2
Abstract | Precursor messenger RNA (pre-mRNA) splicing is an essential step in the flow of information from DN A to protein in all eukaryotes. Research over the past four decades has molecularly delineated the splicing pathway, including characterization of the detailed splicing reaction, definition of the spliceosome and identification of its components, and biochemical analysis of the various splicing complexes and their regulation. Structural information is central to mechanistic understanding of pre-mRNA splicing by the spliceosome. X-ray crystallography of the spliceosomal components and subcomplexes is complemented by electron microscopy of the intact spliceosome. In this Review, I discuss recent atomic-resolution structures of the intact spliceosome at different stages of the splicing cycle. These structures have provided considerable mechanistic insight into pre-mRNA splicing and have corroborated and explained a large body of genetic and biochemical data. Together, the structural data have proved that the spliceosome is a protein-directed metalloribozyme.
Intron lariat
A lasso-like intron structure formed by the first transesterifi cation of the splicing cycle.
Branch point sequence
(BPS). Also known as a branch site. A conserved intron sequence, which contains an invariant adenosine nucleotide that serves as the nucleophile for the branching reaction.
'Beijing Advanced Innovation Center for Structural Biology, School of Life Sciences, Tsinghua University, Beijing 100084, China. 2Institute of Biology, Westlake Institute for Advanced Study, Shilongshan Road No. 18, Xihu District, Hangzhou 310064, Zhejiang Province, China. shi-lab@tsinghua. edu. cn
doi:10.1058/nrm.2017.86 Published online 27 Sep 2017
Pre-mRNA splicing was discovered 40 years ago independently by Phillip Sharp1 and Richard Roberts2. Patient-derived antibodies were found to react with five uridine-rich small nuclear RNAs (snRNAs), known as Ul, U2, U4, U5 andU6 snRNAs, and with seven proteins of 12-35 kDa (REF. 3), which together with the snRNAs form small nuclear ribonucleoproteins (snRNPs). The 5' end sequences of Ul snRNA were recognized to be complementary to the 5' splice site (5'SS)4,5, and the snRNPs were partially purified6. The purified Ul snRNP specifically bound to the 5'SS in vitro7, and the depletion of Ul snRNP inhibited in vitro splicing8"10. Subsequently, the involvement of multiple snRNPs in pre-mRNA splicing was confirmed11"16.
Development of the splicing assay enabled the delineation of the splicing reaction17"20. ATP and magnesium (Mg2+) were found to be indispensable for the in vitro splicing of adenovirus pre-mRNA18'21, and in vitro splicing of the mutant p-globin pre-mRNA derived from individuals with p-thalassaemia was identical to its splicing in vivo22. Formation of an intron lariat through linkage between an adenosine nucleotide in the branch point sequence (BPS) and the guanine nucleotide at the 5' end of the 5'SS was biochemically confirmed23"25. Each cycle of pre-mRNA splicing entails two SN2-type transesterification steps. Step I (commonly known as branching) results in a free 5' exon and an
intron lariat-3' exon intermediate, and step II (exon ligation) leads to the joining of the 5' exon with the 3'exon (FIG. la).
A general two-metal catalytic mechanism was proposed for pre-mRNA splicing, in which the snRNA was thought to coordinate the catalytic metal ions26. This hypothesis was supported by experiments using sulfur-substituted pre-mRNA27. Sulfur substitution of the non-bridging phosphoryl oxygen of nucleotide U80 in U6 snRNA rendered the yeast spliceosome catalytically inactive; the step I reaction was rescued by thiophilic ions such as manganese (Mil2*) in the U80(Sp)-reconstituted but not the U80(_Rp)-reconstituted spliceosome, identifying U80 as a specific coordinating ligand of the catalytic metal28 (FIG. 1 b). Using a similar strategy, U6 snRNA was shown to catalyse both steps of the reaction by positioning divalent metals29 (FIG. 1 b). Structural comparison between the group IIC intron core and the known features of U6 snRNA enabled prediction of the molecular structure of the spliceosomal active site30,31. Consistent with the two-metal catalysis model26'31, all identified U6 metal-coordinating ligands correspond to those in the group II self-splicing introns29. These similarities suggest that the group II intron and the spliceosome share the same evolutionary origin, with the former being the ancestral type.
NATURE REVIEWS I MOLECULAR CELL BIOLOGY
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
VOLUME 18 I NOVEMBER 2017 I 655
REVIEWS
SN2-type transesterification
Transesterification in which one ester bond is broken and another ester bond is formed synchronously; named for substitution nucleophilic and bi-molecular transesterification.
Group II self-splicing introns
A large class of self-splicing catalytic RNAs found in bacteria, fungi and plants.
Pre-mRNA splicing is executed by a large ribo-nucleoprotein complex known as the spliceosome32"34. Assembly of the spliceosome depends strictly on the presence of the 5'SS, BPS and 3'SS, which are conserved sequence elements of introns34"38. Analysis of the in vitro splicing reaction by non-denaturing Polyacrylamide gel electrophoresis enabled the identification
and molecular characterization of distinct spliceo-somal complexes38-44 (FIG. 1 c). The pre-catalytic B complex is converted into an activated spliceosome (Bact complex), which no longer contains the Ul or U4 snRNP. Through additional remodelling, the Bact complex matures into a catalytically activated spliceosome (B* complex), in which the branching reaction occurs.
___ 3'-O.H
V     BPS /A...
Intron lariat
Pre-mRNA    5'SS      BPS 3'SS
7 1=1 GU-A-AGM
5 exon intron     3' exon
E complex I    V \ Prp5-ATP
UAP56-ATP 1^-*-^
A complex
(pre-spliceosome) r Prp28-ATP
B complex (pre-catalyic spliceosome]
P complex (post-splicing complex)
C* complex (step 2 catalytically activated)
C complex catalytic step 1 spliceosome)
[NTR
Bact complex (activated spliceosome
aU4
■»N0||NTC) Prp2-ATP l"NTRl B* complex
(catalytically activated spliceosome)
656 I NOVEMBER 2017 I VOLUME 18 www.nature.com/nrm
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
REVIEWS
MS2-tagged
RNA prepared for affinity purification using a technique based on the interactions of the MS2 bacteriophage coat protein with a stem-loop RNA structure from the phage genome.
The resulting catalytic step I spliceosome (C complex) is remodelled into a step II catalytically activated spliceosome (C* complex), where the exon ligation occurs (FIG. 1 c). The post-catalytic spliceosome (P complex) sheds the ligated exon to become the intron lariat spliceosome (ILS complex), which releases the intron lariat and disassembles into individual snRNPs and associated proteins.
The five snRNPs constitute the core components of the various spliceosomal complexes. With the exception of U6, each of the snRNAs contains a trimethylated guanine nucleotide (m3G) at the 5' end, and a methylated adenine nucleotide (m6A) is present in U2, U4 and U6 snRNAs45. Use of monoclonal antibodies against m3G and m6A improved the purification of the individual snRNPs and identification of their protein components46. In Saccharomyces cerevisiae, a central spliceosomal component, pre-mRNA processing factor 8 (Prp8), was shown to stably associate with U5 snRNA47, and the Prp8 gene of S. cerevisiae was shown to be required for pre-mRNA splicing48. In addition, the formation of the spliceosome and execution of splicing require many other proteins, particularly those of the yeast nineteen complex (NTC, known as the PRP19-CDC5L complex in mammals) and NTC-related (NTR) complex49'50.
Determination of the spliceosomal RNA map — the spatial placement of, and interactions between, the snRNAs and pre-mRNA — revealed that Ul and U2 snRNPs interact with the 5'SS and the BPS, respectively12, and that U2 snRNA is required for this interaction51'52. U5 snRNA appeared to select the 5'SS because a single nucleotide change in loop I of U5 snRNA allowed the selection of a cryptic 5'SS53; loop I specifically recognizes both 5' and 3' exon sequences54. The 5'end sequences of U2 snRNA form a duplex with 3'end sequences of U6 snRNA as part of the U4-U6 duplex in the tri-snRNP55. Crosslinking experiments identified
Figure 1 | The pre-mRNA splicing reactions and the splicing cycle, a | Two steps of transesterification take place during pre-mRNAsplicing. In step 1 (branching), the 2'-OH of the branch point sequence (BPS) adenine nucleotide attacks the phosphate of the guanine nucleotide at the 5' end of the 5' splice site (5'SS). In step 2 (ligation), the 3'-OH of the 3' end nucleotide of the 5' exon attacks the phosphate of the 5' end nucleotide of the 3' exon. b | Coordination of the catalytic metal ions before and after the first step of transesterification. The upper and lower panels represent the B"H and C complexes, respectively. The two metals, designated as Ml and M2, are bound mainly by phosphates from U6 small nuclear RNA. In the first step of transesterification, M2 activates the nucleophile, whereas Ml stabilizes the leaving group. In the second step of transesterification, Ml activates the nucleophile, whereas M2 stabilizes the leaving group, c | Assembly and activation of the yeast spliceosome and the complete splicing-reaction cycle. The 5'SS, BPS and 3'SS are first recognized by the Ul small nuclear ribonucleoprotein (snRNP), splicing factor 1 (SFl; also known as branchpoint-bridging protein) and U2AF, respectively, forming an early spliceosome (known as the E complex). SFl is displaced by the U2 snRNP to form the pre-spliceosome (A complex), which associates with the U4/U6.U5 tri-snRNP to assemble into the pre-catalytic spliceosome (B complex). The B complex represents the first fully assembled spliceosome. There are at least six additional distinct spliceosome complexes: B"ct, B*, C, C*, P and the intron lariat spliceosome (ILS). Each complex has a unique composition, and conversion between complexes is driven by highly conserved RNA-dependent ATPase/helicases (in bold). Notably, a spliceosomal complex can have distinct conformational states, which may also differ in composition. For example, the B and ILS complexes each have at least two distinct conformations.
base-pairing interactions between pre-mRNA and snRNA and between snRNAs56 57. The ACAGA box of U6 snRNA forms a duplex with the 5'SS, and U5 and U6 snRNAs are both located in the splicing active site58"61. These and subsequent investigations have established a two-dimensional RNA map in the spliceosome44'62.
The conversion between the different spliceosomal complexes is driven by eight highly conserved, RNA-dependent ATPase/helicases44'63"65 (FIG. 1 c). Prp5, UAP56, Prp28 and Brr2 function in earlier steps leading to the formation of the Bact complex. Unwinding of the U4-U6 duplex, which precedes the formation of the active spliceosome, is executed by Brr2 in yeast66 or by the helicase U5-200kD (also known as SNRNP200) in humans67. By contrast, the DEAH-box ATPase/ helicases Prp2, Prpl6 and Prp22 have a crucial role during the catalytic cycle. Prp2, which associates with the spliceosome before and throughout step I of the splicing reaction68'69, activates the spliceosome through ATP hydrolysis70'71. Prpl6 is indispensable for the step II reaction, which requires ATP binding and/or hydrolysis72'73. Prp22 is required for mRNA release74. Through these and subsequent studies, Brr2, Prp2 and Prpl6 were established to be responsible for the B-to-Bact, Bact-to-B* and C-to-C* transitions, respectively (FIG. 1 c). Prpl6 and Prp22 are also responsible for the selection of alternative BPSs and 3'SSs, respectively75'76.
Electron microscopy of spliceosomes
The highly dynamic nature of the spliceosome makes its crystallization practically impossible. X-ray-based investigations have been limited to individual components and subcomplexes of the spliceosome. Representative studies include the crystal structures of Ul snRNP77"79, U2 snRNP subcomplexes80"84, U4 snRNP core85, U6 snRNP subcomplexes86'87, a large fragment of Prp8 (REF. 88), Brr2 (REFS 89,90) and the splicing factor 3b (SF3b) complex91. In contrast to X-ray crystallography, electron microscopy performs better with complexes of larger molecular mass and is less dependent on conformational or compositional homogeneity. These advantages have made electron microscopy the only viable approach for structural determination of the intact spliceosome. For electron microscopy studies, spliceosomes were assembled in vitro on MS2-tagged pre-mRNA and successfully purified92'93. In the two decades before 2015, electron microscopy-based investigations have uncovered structural features of the spliceosome at various stages of the splicing reaction94. The mammalian spliceosome structures were elucidated for the A (REFS 95,96), B (REFS 97-99), Bact (REF. 1 00), C (REFS 101,102) and P (REF. 103) complexes. The yeast structures include those of the B (REF. 104), Bact (REF. 104), C(REFS 104,105) and ILS (REF. 106) complexes. These electron microscopy studies have moderate resolution, with the highest being around 20-29 A, and only reveal general features of the spliceosome. The cryo-electron microscopy (cryo-EM) structure of the U4/U6.U5 tri-snRNP from S. cerevisiae, reported at 5.9 A resolution107, allows identification of individual protein and RNA components but reveals few features for atomic modelling.
NATURE REVIEWS I MOLECULAR CELL BIOLOGY VOLUME 18 I NOVEMBER 2017 I 657
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
REVIEWS
Lariat junction
The three-way junction in which a 2' to 5' linkage is formed between the ribose 2'-OH of an adenosine nucleotide in the intron branch site and the phosphate of a guanine nucleotide at the 5' end of the intron 5' splice site.
Mechanistic understanding of pre-mRNA splicing has taken a quantum leap through recent cryo-EM studies of the spliceosome at atomic and near-atomic resolutions108"119 (Supplementary information SI (table)). The overall design principles of the spliceosome were revealed by the structure of the ILS complex from Schizosaccharomyces pombe at an average resolution of 3.6 A108 109 (FIG. 2a) and confirmed by subsequent structures of the S. cerevisiae spliceosome at different stages of the splicing cycle113"118. These principles govern organization of the splicing active site, coordination of the catalytic metal ions, recognition of the pre-mRNA sequences, assembly of the snRNPs and distribution of the snRNA elements, and the central role of protein components in shaping the splicing active site and the overall spliceosome architecture. We now have an atomic view of the spliceosomal RNA map, which includes three snRNAs (U2, U5 and U6) and the pre-mRNA (FIG. 2b,c). The bulk of the yeast splicing cycle has been visualized in atomic detail through structural elucidation of the fission or budding yeast Bact complex113118, C complex114115, C* complex116117 and ILS complex108 109 (FIG. 2a). In the following discussion, the focus is on the mechanistic information revealed by these structures.
Organization of the splicing active site
The splicing active site comprises the intramolecular stem-loop (ISL) of U6 snRNA, helix I of the U2-U6 duplex, the associated Mg2+ ions and loop I of U5 snRNA (FIGS 2b, 3a). The central element is a stretch of 35 nucleotides from U6 snRNA, including the ISL and the 10 preceding nucleotides that form helix I. Despite known information on the base-pairing specifics, atomic modelling of these 35 nucleotides proved to be the most challenging task in building the entire spliceosome108 109. The final detailed structure of these nucleotides is distinct from any other known RNA element (FIG. 3b). The backbone of these 35 nucleotides zigzags in a rectangular box that measures approximately 20 x 25 x 50 A3, making five 180-degree turns. The close contacts among the RNA nucleotides are likely stabilized in part by two metal ions that were interpreted to be magnesium108109. These two structural metals are buried in the centre of the ISL and presumably neutralize the negative charges of the RNA phosphates (FIG. 3c). The overall conformation of the splicing active site observed in the S. pombe ILS complex108 109, including the exact configuration of the ISL, remains largely unchanged in subsequent structures of the various S. cerevisiae spliceosomal complexes113"118.
Helix I is interrupted by an unpaired bulge of three nucleotides in U2 snRNA55 (FIG. 2c). Within helix I, the nucleobases in the last three nucleotides AGC of U6 snRNA also make non-Watson-Crick hydrogen bonds to three nucleotides within U6 snRNA, forming a characteristic triple helix known as the catalytic triplex120 (FIGS 2c, 3d). Loop I, which anchors both 5' and 3' exons54, is placed in close proximity to the ISL. The ligated exon is already released in the S. pombe ILS complex108109. By contrast, the 5' exon remains bound to
loop I in the structures of the S. cerevisiae Bact, C and C* complexes113"118, with three nucleotides from the 3' end of the 5' exon forming a duplex with loop I54.
Coordination of the catalytic metal ions
The two catalytic metal ions, known as Ml and M2, are coordinated by the backbone phosphates of U6 snRNA and the pre-mRNA29 (FIG. 1 b). The 3'-OH of the nucleotide at the 3' end of the 5' exon is stabilized by Ml as the leaving group in the first transesterification and activated by the same metal as the nucleophile in the second transesterification. The 2'-OH of an adenine nucleotide in the BPS is activated by M2 as the nucleophile in the first transesterification. Following the first reaction, the lariat junction is dislocated from the centre of the active site by approximately 20 A to make room for the intron-3' exon sequences. The leaving group in the second reaction — the 3'-OH of the nucleotide at the 3' end of the intron — is stabilized by M2. Notably, the displacement distance of 20 A for the pre-mRNA intron lariat is conserved in the group II intron lariat121.
In the activated but catalytically dormant Bact complex113, two putative catalytic metals are already loaded onto the active site, but neither is properly coordinated (FIG. 3e). In the C complex114, Ml is coordinated by the phosphates of G78 and U80 of U6 snRNA, the 3'-OH of the 3' end nucleotide of the 5' exon, and the phosphate of the 5' end nucleotide of the 5'SS. M2 is bound by the phosphates of A59 and U80 of U6 snRNA but is approximately 6 A away from the nucleophile. In the C* complex116, metal coordination is very similar to that of the C complex114, except that Ml is only coordinated by three ligands and separated from the fourth one — the phosphate of the 5' end nucleotide of the 5'SS — by a distance of approximately 21A (FIG. 3e). Coordination of M2 remains the same as that in the C complex114. In the S. pombe ILS complex108109, M1 is bound by G48, G66 and U68, which correspond to G60, G78 and U80, respectively, of the S. cerevisiae U6 snRNA. M2 is recognized by A47, G48 and U68 (A59, G60 and U80 in S. cerevisiae).
Consistent with previous analysis28'29, U80 of U6 snRNA has a central role in coordinating both Ml and M2 in the two transesterification steps. Coordination of Ml is fairly static because the ion is both a stabilizer and an activator for the same group — the 3'-OH of the nucleotide at the 3' end of the 5' exon, which serves as the leaving group in the first transesterification and the nucleophile in the second transesterification. The catalytic metals are presumably Mg2+ in these cryo-EM structures, in part because the pattern of metal coordination is consistent with that for Mg2+. In addition, the Mg2+ assignment is supported by structural comparison with the group II intron109 and the phosphoro-thioate substitution experiments in the spliceosome28'29. Nonetheless, the reported local resolutions at the active site are insufficient to conclusively identify the metals.
Prp8 and its three catalytic motifs
A central component of U5 snRNP, Prp8 (Spp42 in S. pombe), is the largest and most highly conserved spliceosomal protein122. Prp8 comprises seven domains:
658| NOVEMBER 2017 I VOLUME 18
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved
www.nature.com/nrm
REVIEWS
a ILS complex (S. pombe)
U2snRNP
U2 snRNA
Leal
Msll
U2 Sm ring
(U6 snRNA )
U5 snRNP
U5 snRNA Spp42/Prp8 Cwfl0/Snull4 U5 Sm ring Cwfl7
NTC
Cwf8/Prpl9
Cdc5/Cefl
Cwf7/Snr309
Cwf3/Syfl
Cwf4/Clfl
Cwfl/Prp46
Cwf2/Cwc2
Cwfl5/Cwcl5
U6 snRNA
U5 snRNA
Intron lariat
U2 snRNA
NTR
Prp45
Cwfll
Cwfl9
Cwf5/Ecm2
Cwfl4/Bud31
C complex
C* complex
ILScomplex
1W
Splicing factors
Prpl7 Cypl
Helix II
-Active site Catalytic centre Helix I
Loop I
u
30A
Stem III
'aaaacguuucacgaaacugc-I I I I • I I • I I I    I I I I I I
-5'-
U5 snRNA
Stem I
U6 snRNA
c -u -A-5'-g-
cacuuuggucaaauugaaac
3'-u -A-u -g -u -a u,
Loop I \
g ( ■A c
u ■ g ■ «u ■ u ■ u ■ c ■ u ■
uuuguaaggug uuugacuguau
g c c g ■A c
a50 ■A ■A g ■A
Stem I
a
; a cu
u • c
g -c
Intron lariat
a a a u
u u
V uU
a
ISL f
60 c
g u -c -c -
c
agaagau uag GK 3'I I I III
s'-uuuauacua^au
g • a70
g-c u
ioA
£ Triplex
a g u u u
u
I
n°u„
acugcg
A.,
cua a gu
«.........lb
gucuaugaugug     |a helix helix
auugagagaaa----3
I I I I I I I ucgguuuuccguuu cucucuua 5' Helix II
U2 snRNA
Figure 2 | Atomic structures of the spliceosome. a | The structure of the intron lariat spliceosome (ILS) of Schizosaccharomyces pombe at an overall resolution of 3.6 A (REFS 108,109) is shown in the left panel. Colour coding for the various spliceosomal subcomplexes and components is indicated. The structures of the U4/U6.U5 tri-small nuclear ribonucleoproteins110'111, the B"ct complex113'118, the C complex114'115 and the C* complex116'117 of Saccharomyces cerevisiae are shown in smaller panels on the right, along with the structure of the S. pombe ILS complex for comparison. The orientation and the scale of the spliceosome, as defined by U 5 small nuclear RNA (snRNA), are maintained exactly the same for all the complexes, b | Spatial arrangement of the RNA elements from the S. pombe ILS complex108 109. U2 snRNA, U5 snRNA, U6 snRNA and the intron lariat are coloured blue, orange, green and purple, respectively. This colour scheme is preserved in all structural figures throughout this Review unless otherwise indicated. The metal ions in the active site are shown as red spheres, c | The observed base-pairing specifics of the RNA map of the S. pombe ILS complex. Non-standard Watson-Crick interactions are marked by black dots. The 5' exon sequences (shown in grey) are absent in the ILS complex, but their interactions with loop I of U5 snRNA in other spliceosomal complexes are nonetheless indicated in this diagram. All essential features of the RNA map observed in the S. pombe ILS complex were confirmed by subsequent structures of the B"ct, C and C* complexes from S. cerevisiae113-118. Part b is reproduced with permission from REE 109.AAAS.
NATURE REVIEWS I MOLECULAR CELL BIOLOGY VOLUME 18 I NOVEMBER 2017 I 659
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
REVIEWS
C complex U6 snRNA
C* complex U6 snRNA
Figure 3 | Organization of the splicing active site and the catalytic metal ions, a | Structure of the splicing active site in the Schizosaccharomyces pombe intron lariat spliceosome (ILS)108,109. In addition to the intramolecular stem-loop (ISL) and the associated metals (spheres), the active site also includes helix I of the U2-U6 duplex and loop I of U5 small nuclear RNA (snRNA), which recognizes the 5' exon. b | Structure of the ISL and the preceding 10 nucleotides of U6 snRNAfrom the S. pombe ILS complex108109. The overall conformation is preserved in subsequent structures of the B"H, C, and C* complexes from Saccharomyces cerevisiae113-117. c | Two structural metal ions in the active site of the S. pombe ILS complex stabilize the RNA fold by neutralizing negative charges of the phosphodiester backbone108 109. The phosphorus atoms are highlighted as large orange spheres. The location and coordination of these two metals are highly conserved in the other S. cerevisiae spliceosomal complexes113-118, d | A close-up view of the catalytic
triplex in the active site of the S. pombe ILS complex. Hydrogen bonds are represented by red dashed lines, e | In the S. cerevisiae B"ct complex113, two metals are bound in the active site but neither is properly coordinated. Only one metal ion, which may correspond to M2, is shown. For comparison, the overall orientation of the active site in panels a-d is approximately the same. In the S. cerevisiae C complex114, Ml is coordinated by C78 and U80 of U6 snRNA, the 3' end nucleotide of the 5' exon, and the 5' end nucleotide of the 5'SS. M2 is recognized byA59 and U80 of U6snRNA but is located 6.2 Aaway from the nucleophile. In the S. cerevisiae C* complex116, metal coordination is similar to that in the C complex, except that Ml is no longer bound by the 5' end nucleotide of the 5'SS. In the S. pombe ILS complex108109, Ml is bound bythe nucleotides C48, C66 and U68 of the U6snRNA, which correspond toC60, C78 and U80, respectively, in S. cerevisiae. M2 is recognized byA47, C48 and U68, which correspond to A59,C60and U80 in S. cerevisiae.
an amino-terminal domain (N-domain), reverse transcriptase (RT) fingers/palm, thumb/X, linker, endo-nuclease-like, RNase H-like and Jabl/MPN (FIG. 4a). The core of Prp8 and Spp42, also known as the large domain88, contains four domains: RT fingers/palm, thumb/X, linker and endonuclease-like. The core has a compact structure and, except for a few surface elements, maintains a rigid conformation throughout spliceosome assembly, activation and catalysis. Relative to the core, the N-domain is rotated by approximately 30 degrees during activation of the spliceosome110-112. The RNase H-like domain adopts quite different positions in the different spliceosomal complexes114. The Jabl/MPN domain, which regulates the function of Brr2 through direct interactions89, is highly mobile and has only been structurally captured with reasonable detail in the U4/U6.U5 tri-snRNP110'111 and the Bact complex113118.
Prp8 and Spp42 harbour a positively charged cavity that functions as a central scaffold for splicing. The N-domain is chiefly responsible for the recognition of U5 snRNA108 109 123. The relative orientation of the N-domain with respect to the core remains unchanged
in the Bact, C, C* and ILS complexes. This arrangement yields a positively charged scaffold between the N-domain and the core; the scaffold accommodates the splicing active site108109 (FIG. 4b). Notably, the catalytic cavity is different from that predicted on the basis of the structure of the isolated Prp8 core88.
Three Prp8 structural motifs — the Switch loop, the 1585 loop and the p-finger — have important roles during splicing. The Switch loop stabilizes 5' exon binding by interacting with the splicing factor Cwc21, which directly recognizes the 5' exon (FIG. 4c). Compared with its position in the tri-snRNP110111 or the ILS complex108109, where the 5' exon is no longer bound, the Switch loop is flipped by approximately 180 degrees in the Bact, C or C* complex to stabilize the bound 5' exon113"119. The 1585 loop may directly promote the splicing reaction124. In the Bact complex113, the 1585 loop interacts with helix la of the U2-U6 duplex and contributes to pre-mRNA binding by closely associating with the elements of Prpll and Cwc24 that recognize the 5' end guanine nucleotide of the 5'SS (FIG. 4d, left panel). In the C* complex116117, the 1585 loop directly contacts the ISL and the lariat junction, with its tip
660 I NOVEMBER 2017 I VOLUME 18 www.nature.com/nrm
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
REVIEWS
Switch loop (1,402-1,439)
N-domain I   RT fingers/palm
127885A886
1585 loop (1,585-1,598)
.ß-finger (1,860-1,874)
Prp8
m  ^(Thumb/X(   ( [Linker (( ^(        ase I^RNase H-like(   (jabl/MPN (|Q
1,251 I      1,375A1,376       1,649 1,653   1,824 1,839       2,092   2,150 2,396
1,257 RNaseH-like
Thumb/X
Jabl/MPN
RT fingers/palm
1585 loop
Switch loop
b Prp8 Pre-mRNA        U2 snRNA
Electrostatic surface potential Negative Positive
d Bact complex
B* complex
e C* complex
r 4 s~t
Lariat junction
U2 snRNA
Lariat junction
Figure 4 | Prp8 and its catalytic motifs, a | Structure of Prp8 from Saccharomyces cerevisiae. The seven domains of Prp8 are colour-coded. The core comprises four domains (reverse transcriptase (RT) fingers/palm, thumb/X, linker and endonuclease-like) and maintains a rigid conformation throughout spliceosome assembly, activation and catalysis, b | A positively charged catalytic cavity is located between the amino-terminal (N)-domain and the core108 109, c | The Switch loop stabilizes the bound 5' exon sequences through direct interactions with Cwc21. d | The 1585 loop has a prominent role during the splicing reaction. In the B"H complex113, the 1585 loop interacts with the U2-U6 small nuclear RNA (snRNA) duplex and stabilizes the 5' end nucleotide of the 5'SS through direct interactions with Cwc24 and Prpll (left panel). In the B* complex116, the 1585 loop binds directly to the intramolecular stem-loop of U6 (not shown) and the lariat junction, with its tip close to the 5' exon and the catalytic Mg2t ions (pink dots; right panel), e | In the C* complex, the [3-finger interacts with Prpl7 near the lariat junction116. The [3-finger and the 1585 loop are located on two sides of the lariat junction.
NATURE REVIEWS I MOLECULAR CELL BIOLOGY VOLUME 18 I NOVEMBER 2017 I 661
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
REVIEWS
HEAT repeats
Repetitive motifs frequently found in signalling proteins; each HEAT repeat comprises a pair of anti-parallel a-helices, linked by a short loop.
close to the 5' exon and the catalytic metals (FIG. 4cl, right panel). The third catalytic motif is the p-finger125 from the RNase H-like domain. In the C* complex116117, the p-finger interacts with the step II splicing factor Prpl7, and the lariat junction is sandwiched between the p-finger and the 1585 loop (FIG. 4e). This arrangement may stabilize the active site conformation for exon ligation. The p-finger remains unengaged in the Bact complex113 and contacts the intron-U2 duplex in the C complex114115.
Recognition of the intron sequences
The 5'SS and the BPS are recognized through duplex formation with U6 and U2 snRNAs, respectively, in the structures of the various yeast spliceosomes108'109'113-118. In all cases, five consecutive nucleotides of the intron, including the last three nucleotides UGU from the 5'SS, form Watson-Crick interactions with the nucleotides AUACA of U6 snRNA (FIG. 2c). Notably, only three nucleotides of the U6 snRNA ACAGA box are directly involved in 5'SS recognition. This base-pairing arrangement is likely necessary for both catalytic steps of splicing. In step I, the guanine nucleotide at the 5' end of the 5'SS is flipped out of base-pairing register (FIG. 5a, top left), thereby disallowing duplex formation between the GUA nucleotides of the 5'SS and U6 snRNA. In step II, delivery of the 3'SS-3' exon sequences into the active site requires the dislocation of the lariat junction away from the centre of the active site, which requires that the U6 sequences between those recognizing the 5'SS and those forming helix I are sufficiently flexible.
The first three nucleotides of the 5'SS (GUA) are differentially recognized by surrounding proteins in different spliceosomal complexes. In the S. cerevisiae Bact complex113, the guanine nucleotide is recognized by three hydrogen bonds from Cwc24 and Prpll, with its base sandwiched between the aromatic side chains of Tyrl55 and Phel61 of Cwc24 (FIG. 5a, top right). The uridine nucleotide also interacts with Cwc24 through hydrogen bonds, whereas the adenine nucleotide stacks against A51 of U6 snRNA. In the C complex114115, the guanine nucleotide is bonded to the BPS, with its ribose accepting a hydrogen bond from Arg4 of Yju2 (also known as Cwcl6; FIG. 5a, bottom left). The uridine nucleotide interacts with U2 snRNA and Arg4 of Yju2 through hydrogen bonds, and the adenine nucleotide remains stacked with A51 of U6 snRNA. In the C* complex116117, the guanine nucleotide is hosted by the 1585 loop and the RNase H-like domain of Prp8, whereas the uridine and adenine nucleotides are close to Prp8 and U6 snRNA (FIG. 5a, bottom right).
In the Bact, C, C* and ILS complexes108 109 113"118, the consensus BPS UACUAAC and the preceding nucleotides are similarly recognized through duplex formation with U2 snRNA. In the Bact complex113118, the intron-U2 duplex is held in place mainly by the SF3b proteins Hshl55 and Rds3 and the SF3a protein Prpl 1 and to a lesser extent by Cusl and the SF3b protein Ysf3 (also known as RDS3 complex subunit 10). The protein-RNA interactions are mostly hydrogen bonds directed at the RNA phosphate backbone and stabilized by the positive
electrostatic surface potential of the proteins. Duplex formation between the BPS UACUaAC (lower case marks the adenosine nucleophile) and the U2 snRNA sequence GUAGUA leaves the adenosine nucleophile unpaired and flipped out of register (FIG. 5b). The adenine base is sandwiched by two residues from Hshl55; the ribose, the base and the phosphate each accept a hydrogen bond from Hshl55 (FIG. 5c). This pattern of recognition is analogous to that of the 5' end guanine nucleotide in the 5'SS114. The catalytic dormancy of the Bact complex is ensured by the structural feature that the nucleophile for the first transesterification — the ribose 2'-OH — is separated from the scissile bond by approximately 49 A, with the two splicing factors Cwc24 and Prpll located in between113118.
The proteins that hold the intron-U2 duplex in the Bact complex are replaced in the C complex by Prp8, Isyl and the splicing factors Cwc25 and Yju2 (REFS 114,115). Owing to the formation of the intron lariat, the adenosine nucleophile is caged by three surrounding intron nucleotides and the GU nucleotides at the 5' end of the 5'SS (FIG. 5d). In the C* complex116117, the intron-U2 duplex is recognized by Cefl, Prp8 and the splicing factor Prpl7 (also known as Cdc40). Unlike in the Bact or C complex, in the C* complex, the adenosine nucleophile is no longer recognized by surrounding elements. The lariat junction is located close to the RNase H-like domain of Prp8. In the S. pombe ILS complex108 109, the adenine base is uncoordinated by the protein components; the lariat junction contacts Cwfl9, a homo-logue of the S. cerevisiae debranching-enzyme cofactor Drnl (REF. 126).
In the Bact, C and C* complexes113-118, five consecutive nucleotides downstream of the 5'SS closely associate with the NTR component Cwc2 (FIG. 5e). Cwc2 may help orient the 5'SS into the active site through interactions with downstream intron sequences127. In the Bact complex113118, the pyrimidine-rich sequences downstream of the BPS are recognized by Hshl55, Rds3 and the retention and splicing (RES) complex (FIG. 5f). Human HSH155 (also known as SF3B1) is frequently mutated in several malignancies128. Most of the mutated residues are at the HEAT repeats 3 and 4of HSH155, which bind to the pyrimidine-rich sequences but not to the BPS. The mutations that abolish recognition of the BPS are presumably lethal; the milder mutations that sabotage binding of the pyrimidine-rich sequences might compromise the splicing fidelity by promoting selection of alternative BPSs, thus contributing to oncogenesis129130. In the C complex114115, these pyrimidine-rich sequences are recognized by the fingers/palm, thumb/X and linker domains of Prp8.
Recognition of the exon sequences
In the Bact, C and C* complexes113-118, the 5' exon is anchored identically in the active site, with three consecutive nucleotides (modelled as A 3A 2G,X) at the 3' end forming a duplex with U96U97U98 of the U5 snRNA loop I (Supplementary information S2 (figure)). The phosphodiester backbone of the three 5' exon nucleotides is in close contact with the N-domain of
662 I NOVEMBER 2017 I VOLUME 18
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved
www.nature.com/nrm
REVIEWS
Prp8 in all three complexes. The N-terminal three residues Glu3-Arg4-Lys5 of Yju2 in the C complex114115 and Lysl637 of the Prp8 linker domain in the C* complex116117 interact directly with the backbone of these three nucleotides. Seven preceding nucleotides are also recognized similarly in all three spliceosomal complexes.
Nucleotides -10 to -8 of the exon are bound between Cwc21 and the Prp8 N-domain, with Cwc21 being oriented by the Prp8 Switch loop; nucleotides -7 to -4 interact with the N-domain and the linker domain of Prp8 (Supplementary information S2 (figure)). In the C and C* complexes114"117, nucleotides -13 to -11 are close
Bact complex
Bact complex
_Cwc24f\ Fl61„*-*
Prpu \ !<:u;;^-;:^-'''\
U2 - BPS duplex
C
LAPmll
U2 snRNA
pre-mRNA
Snul7
Figure 5 | Recognition of the intron sequences, a | Recognition of the 5' splice site (5'SS). In all structurally characterized spliceosomal complexes, the UCU nucleotides of the 5'SS and two ensuing intron nucleotides form a duplex with U6 small nuclear RNA (snRNA). In the B"H complex113, the guanine nucleotide at the 5' end of the 5'SS is flipped out of base-pairing register and recognized by Cwc24 and Prpll (upper panels). After branching, the guanine nucleotide and the ensuing nucleotides are bound by Yju2 and by snRNAs in the C complex and by the 1585 loop and RNase H-like domain of Prp8 in the C* complex (lower panels), b | Duplex formation between the branch point sequence (BPS) and U2 snRNA. Two bases of the intron are unpaired. Similartothe 5' end guanine nucleotide of the 5'SS,
Budl3
the adenosine nucleophile of the BPS is flipped out of base-pairing register with the BPS-U2 duplex. c| Recognition of the adenosine nucleophile of the BPS in the B"ct complex113. The adenine nucleophile interacts closely with residues of Hshl55. d | Recognition of the adenosine nucleophile of the BPS in the C complex114. The base of the adenosine nucleophile is surrounded by intron nucleotides and the CU nucleotides at the 5' end of the 5'SS. e | Recognition of intron sequences downstream of the 5'SS by Cwc2 in the B"ct complex113. These structural features are preserved in the C and C* complexes114,116, f | In the B"ct complex, the pyrimidine-rich sequences downstream of the BPS are recognized by Hshl55, Rds3 and the retention and splicing complex proteins Pmll, Snul7 and Budl3 (REF. 11 3).
NATURE REVIEWS I MOLECULAR CELL BIOLOGY
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
VOLUME 18 I NOVEMBER 2017 I 663
REVIEWS
HAT repeats
Repetitive motifs implicated in RNA processing; each HAT repeat comprises two a-helices.
to Cwc21, Cwc22 and the Prp8 N-domain. In contrast to the observed recognition of the 5'SS and the 5' exon, how the 3'SS and the 3' exon are recognized and selected during splicing remains to be structurally characterized. U6 snRNA is thought to be involved in recognizing the 3'SS through duplex formation131"133, and the spacing between the BPS and 3'SS matters in this process134 135. The step II factor Slu7 may have an essential role in the selection of the 3'SS136"139.
The core of the spliceosome is rigid
Despite frequent remodelling of the spliceosome during splicing, the structural core around the active site in the Bact, C, C* and ILS complexes is rigid. This core comprises U5 and U6 snRNA, a portion of U2 snRNA and at least 17 associated proteins. The entire U6 snRNA, the bulk of U5 snRNA and a stretch of 30 consecutive nucleotides of the 5' end of U2 snRNA are perfectly aligned in these spliceosomal complexes (FIG. 6a). This portion of U2 snRNA forms helices I and II with U6 snRNA. Unlike U5 snRNA, which forms several internal stems and loops, U6 snRNA adopts an extended conformation that is maintained by spliceosomal proteins. These proteins include nine of U5 snRNP (Spp42 (S. pombe) or Prp8 (S. cerevisiae), Cwfl0/Snull4 and the heptameric Sm complex), two of NTC (Cdc5/Cefl and Syf2) and six of NTR (Cwfl/Prp46, Cwf2/Cwc2, Cwf5/Ecm2, Cwfl4/Bud31, Cwfl5/Cwcl5 and Prp45) (FIG. 6b). The structurally resolved portions of these 17 proteins remain largely static in the Bact, C, C* and ILS complexes, which, given the overall compositional and conformational variability of the spliceosome, is unanticipated and has important ramifications for understanding the mechanism of splicing.
Of the five snRNPs, only U5 is well represented in the core and serves as the scaffold in all spliceosomal complexes. This structural core appears to be nucleated from a rigid centre that is already present in the U4/U6.U5 tri-snRNP and includes the bulk of U5 snRNA, the N-domain of Spp42/Prp8 and Cwfl0/Snull4 (REF. 110). Notably, components of the NTC and NTR heavily contribute to the core by closely interacting with U5 snRNP and among themselves. These proteins have an essential role in shaping the splicing active site. Among all three snRNAs in these spliceosomal complexes, the only mobile elements are the U2 sequences that are downstream of nucleotide 30 (FIG. 6c). A well-documented function of this mobile U2 element is to bind and deliver the BPS to the active site. During the Bac,-to-B* transition, the duplex between this element and the BPS is translocated by nearly 50 A, thereby bringing the nucleophile into close proximity with the leaving group for the first-step transesterification. During the C-to-C* transition, the lariat junction, which remains duplexed with this element, is moved away from the active site by approximately 20 A to accommodate the 3'SS. Notably, the 5' and 3' ends of all three snRNAs are placed away from the active site, usually anchored by spliceosomal proteins109. For example, the 3' end sequences of U2 and U6 are bound by the Sm heptamer and by Syf2, respectively.
Distinct classes of NTC and NTR proteins
The spliceosome must maintain structural stability around the active site while allowing sufficient flexibility so that pre-mRNAs of varying lengths and sequences can be bound, spliced and released140. The stable yet flexible organization of the spliceosome is ensured by myriad inter-molecular interactions. Among the 37 proteins and four RNA molecules of the S. pombe ILS complex, there are more than 100 pairs of bi-molecular interactions108109. For example, each of 14 spliceosomal components of the S. pombe ILS complex interacts with five or more other components. Spp42 interacts with 12 other spliceosomal components and is the only protein that simultaneously binds to all three snRNAs and the pre-mRNA.
Two classes of components help to ensure the structural integrity and functional flexibility of the spliceosome: the rope-like proteins, which glue multiple components together, and the spring-like proteins, which allow large conformational shifts. All these proteins are components of the NTC and NTR, which are recruited to the Bact complex following the dissociation of Ul and U4 snRNPs50 104 141. The rope-like proteins are exemplified by the NTR protein Prp45 and the NTC protein Cdc5/Cefl (Supplementary information S3 (figure)). Prp45 spans more than 150 A and interacts with at least nine spliceosomal components. Extended sequences in each of these rope-like proteins are likely intrinsically disordered in isolation and assume a specific conformation only upon association with other spliceosomal components. The spring-like proteins are represented by the NTC proteins Syfl and Syf3 (also known as Cwf4)/Clfl (Supplementary information S3 (figure)), which contain multiple HAT repeats. The inherent elasticity of these superhelical proteins may enable the spliceosome to bend and stretch in order to facilitate the splicing reaction. In the C-to-C* transition, U2 snRNP is translocated by 60-100 A, resulting in alteration of the overall appearance of the spliceosome. Together, these proteins and the Prpl9 tetramer constitute a binding platform for other spliceosomal components (Supplementary information S3 (figure)).
The roles of splicing factors
Splicing factors regulate the splicing reaction at least in part by stabilizing the conformation of the active site. In the Bact complex113118142, the splicing factor Cwc24 helps maintain catalytic dormancy by insulating the 5' end nucleotide of the 5'SS, whereas Cwc21 stabilizes the binding of the 5' exon143. The role of Cwc21 is preserved in the C and C* complexes and likely in the B* and P complexes as well. The step I factors Cwc25 and Yju2 are both located in the centre of the active site of the C complex114115 (FIG. 7a). The N-terminal sequences of both Cwc25 and Yju2 are inserted into the active site; Cwc25 interacts with the intron-U2 duplex, ISL, helix I and a p-sheet domain of Yju2; Yju2 binds to the 5' exon, a portion of the intron-U2 duplex and both the N-domain and the core of Prp8. An a-helix from Yju2 intercalates between the intron-U2 duplex and the al and a2 helices of Isyl. These interactions place the nucleophile close to the scissile phosphodiester bond144"146.
664 I NOVEMBER 2017 I VOLUME 18
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved
www.nature.com/nrm
REVIEWS
Helix II
Helix I U2 snRNA
I Bact complex | C* complex I C complex    ^ "-S complex (S. pombe)
I Bact complex I C complex I C* complex I ILS complex (S. pombe)
Bact complex
ILS complex (S. pombe)
BU5snRNP	■ U2 snRNA
Bntc	>U5 snRNA
NTR	>U6 snRNA
Figure 6 | A shared rigid core of the various spliceosomal complexes, a | The overall conformation of U5 and U6 small nuclear RNAs (snRNAs) and part of U2 snRNA remains largely the same in the B"ct, C, C* and intron lariat spliceosome (ILS) complexes, as shown here in an overlay of these RNA elements. For clarity, only the RNA elements from the C complex are colour-coded; for each of the other three complexes, the RNA elements are shown in the same colour, b | Approximately 20 protein and snRNA components constitute a rigid core of the spliceosome. The structures and relative orientations of these components remain largely unchanged in the B"H, C, C* and ILS complexes. Shown is a comparison of these components between the Saccharomyces cerevisiae B"ct and C complexes (left panels) and the Schizosaccharomyces pombe ILS complex (right panel). The components include ten U5 snRNP components (U5 snRNA, SPP42 (S. pombe)/ Prp8 (S. cerevisiae), Cwfl0/Snull4and the heptameric Sm ring), U6 snRNA, a portion of U2 snRNA, the NTC proteins Cdc5/Cefl and Syf2, and six NTR proteins (Prp45, Cwf5/Ecm2, Cwfl/Prp46, Cwf2/Cwc2, Cwfl5/Cwcl5 and Cwfl4/ Bud31). c | The U2 snRNA sequences downstream of nucleotide 30 are mobile during the two steps of transesterification. Although the 5' end 30 nucleotides of U2 snRNAs are well aligned, the spatial placements of the ensuing nucleotides are vastly different in the yeast B"H, C, C* and ILS complexes.
WD40 domain
Also known as p-propeller or WD repeats. A type of circular solenoid protein domain that consists of 5-8 short structural motifs of approximately 40 amino acids each, often terminating in a tryptophan-aspartic acid (WD) dipeptide.
The step II splicing factors Prpl7, Prpl8 and Slu7 facilitate exon ligation139147"149. In the C* complex116117, both Prp 17 and Prp 18 are intimately involved in the formation of the active site (FIG. 7b). Prpl8 directly interacts with the RNase H-like domain of Prp8. Although the RNase H-like domain remains bound to the intron-U2 duplex in the C-to-C* transition, its location is markedly shifted because of spliceosomal remodelling. Consequently, the p-finger is placed close to the lariat junction and directly
contacts Prpl7. The WD40 domain of Prpl7, which is positioned between the BPS-U2 duplex and the 5'SS-U6 duplex, is similarly shifted during the C-to-C* transition.
RNP remodelling by the ATPase/helicases
The available structures provide insights into how the RNA-dependent ATPase/helicases remodel the spliceosome. In the yeast U4/U6.U5 tri-snRNP110111, the ATPase/helicase Brr2 constitutes one arm of the
NATURE REVIEWS I MOLECULAR CELL BIOLOGY
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
VOLUME 18 I NOVEMBER 2017 I 665
REVIEWS
Cwc25
Pre-mRNA
Pre-mRNA
U2sn
Cwclö N terminus
Isyl U6 snRNA
Prpl7
'jf ? Prpl8
~|Prpl8, Slu7 l ) JPrpl6-ATP
complex
ILS complex
_--r^3 Prp22 "ATP
Figure 7 | Step I and II splicing factors facilitate transesterifications by participating in the formation of the active site, a | The step I splicing factors Cwc25 and Yju2 support the conformation of the active site in the C complex114115. The conformation of the active site is supported also by Isyl. b | The step II splicing factors Prpl7, Prpl8 and Slu7 (not shown) are located near the active site in the C* complex116117, c | A mechanistic view of the two steps of transesterification in a cycle of pre-mRNAsplicing. This model is based on the high-resolution cryo-electron microscopy structures of the yeast spliceosome and a summary of prior biochemical studies44'108,109,113-118. The three ATPases/helicases Prp2, Prpl6 and Prp22 (in bold text) have essential roles in the conversion of B"ct to B* C to C* and P to ILS, respectively. The step I factors Cwc25 and Yju2 stabilize the active site conformation forthe first transesterification reaction, whereas the step II factors Prpl8 and Slu7 facilitate the second transesterification reaction.
666 I NOVEMBER 2017 I VOLUME 18 www.nature.com/nrm
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
REVIEWS
Sm ring
Also known as heptameric Sm complex. A doughnut-shaped structure formed by seven distinct Sm proteins, with single-stranded RNA bound within the hollow centre.
triangular-shaped complex, with a heptameric Sm ring binding the 3' end sequences of U4 snRNA at the far end of the arm. The Jabl/MPN domain of Prp8 binds to the other end of Brr2 and connects it to the centre of the complex90. The single-stranded RNA sequences between the 3'-stem-loop and stem I of the U4-U6 duplex are placed into the active site of the N-terminal helicase domain of Brr2 through a channel formed by its RecAl, RecA2, ratchet and WH domains111. In the B-to-Bact transition, ATP binding and hydrolysis by Brr2 presumably generates a pulling force on the U4-U6 duplex, thereby dissociating the surrounding proteins and unwinding the U4-U6 duplex66 150 151. Intriguingly, in human tri-snRNP112, the position of Brr2 differs from that in yeast tri-snRNP by as much as 200 A. This arrangement in yeast likely causes constant binding by Sadl, which functions to stabilize the tri-snRNP152.
The ATPase/helicase Prp2 is located at the periphery of the Bact complex113118. Prp2 interacts with the super-helical protein Hshl55 and is placed near the RES complex, which binds to the single-stranded intron sequences downstream of the BPS153 1 54. Prp2 presumably associates with the intron sequences downstream of those bound by the RES complex155"158. The adenosine nucleophile, together with the neighbouring intron-U2 duplex, is surrounded by proteins of the SF3b complex. In the Bact-to-B* transition, Prp2 may pull the 3' end sequences of the intron, thereby dissociating the RES complex and the SF3b complex142156 157. These changes likely result in the exposure of the nucleophile in the BPS144 159. In the C complex115, a lobe of density was assigned to the ATPase/ helicase Prpl6, and another lobe was interpreted to be Brr2. The last ordered nucleotide at the 3' end of the intron is more than 60 A away from the RNA entry site of Prpl6. In the C-to-C* transition, Prpl6 would presumably pull the intron sequences from the 3' end towards the BPS, thereby dissociating the step I factors Cwc25 and Yju2 (REFS 75,147,160,161). In the C* complex117, an ATPase/helicase tentatively assigned as Prp22 associates with the linker and RT fingers/palm domains of Prp8. In the P-to-ILS transition, Prp22 is required to dissociate the ligated exon in the 3' to 5' direction162 163. Notably, the ATPases/helicases in the Bact, C and C* complexes are located in the peripheral regions of the spliceosome and display inadequate density for atomic modelling. Consequently, the proposed functional mechanisms of these ATPase/helicases are speculative and await further experimental investigation.
Comparison with a self-splicing intron
Pre-mRNA splicing is thought to share the same catalytic mechanism as that of the group IIB self-splicing introns26'31'164'165, both involving two catalytic metal ions and an intron lariat intermediate. Although the overall structure of the spliceosome is very different from that of the group IIB introns, they do share many common features at the active site. Structural comparison between the S. pombe ILS complex and the group IIB intron P.U.LSUI2 from the brown alga Pylaiella littor-alis121 reveals similar structural features between the ISL of U6 snRNA and the Domain V of the group IIB
intron, with the metal-coordinating U68 of the ILS complex replaced by the nucleotide A574 in the group IIB intron108 109 (Supplementary information S4 (figure)). The BPS-U2 duplex of the ILS complex is analogous to Domain VI of the group IIB intron, with the adenosine nucleophile in the BPS corresponding to A615 in P.U.LSUI2 (REFS 108,109). The catalytic triad of nucleotides A47-G48-C49 in the spliceosome corresponds to A551-G552-C553 in the group IIB intron (Supplementary information S4 (figure)). Finally, the formation of the catalytic triplex in the spliceosome is nearly identical to that in the self-splicing intron.
U4/U6.U5 tri-snRNP and assembly of the snRNPs
Cryo-EM structures of the S. cerevisiae U4/U6.U5 tri-snRNP at near-atomic resolution provide insights into spliceosome assembly and activation110111. Within the tri-snRNP, the ISL is sequestered through duplex formation with U4 snRNA, and the U4-U6 duplex is coated by Prp8 and the tri-snRNP-specific components Prp3, Prp6, Prp31 and Snul3 (also known as 13kDa ribonucleoprotein-associated protein). In particular, the U6 nucleotides that coordinate the catalytic metals during the splicing reaction are protected by Prp3. In the tri-snRNP structure reported in one study110, the 5'SS already forms a duplex with the AC AG A box of U6, and four nucleotides at the 3' end of the exon are anchored to loop I. This observation is supported by prior biochemical characterizations57166. By contrast, in the tri-snRNP structure reported in another study111, the 5' sequences of U6 snRNA occupy the same general location as that occupied by the pre-mRNA110. This portion of U6 forms a previously unknown short internal stem-loop, and a few consecutive nucleotides are anchored to loop I of U5 snRNA. The cryo-EM density maps in both studies have comparable overall quality. However, these two studies used different purification methods, and the tri-snRNP samples appear to represent different states of the tri-snRNP. Comparison of the two structures suggests a model in which the 5' sequences of U6 snRNA may serve as a decoy to compete with pre-mRNA binding in an early tri-snRNP. In this model, the 5' sequences of U6 snRNA may be bound to loop I of U5 snRNA in the tri-snRNP, and the recruitment of pre-mRNA may competitively displace the 5' sequences of U6 snRNA away from loop I.
Conclusions and future perspective
During the past 2 years, the emergence of a number of atomic structures of the spliceosome by single-particle cryo-EM analysis has fundamentally advanced our mechanistic understanding of pre-mRNA splicing. The cryo-EM structure of the ILS complex from S. pombe enabled atomic modelling of an intact spliceosome with unprecedented structural clarity and revealed several principles that govern the function of the spliceosome and the splicing reaction108 109. This structure provided the first atomic view of the splicing active site and has enabled us to analyse the fine structural features that support the two steps of transesterification. The structure of the ISL of U6 snRNA is novel, and its coordination of catalytic metals by specific nucleotides confirms
NATURE REVIEWS I MOLECULAR CELL BIOLOGY VOLUME 18 I NOVEMBER 2017 I 667
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
REVIEWS
a decades-old hypothesis26 and subsequent biochemical investigations27-29,31. Unlike the mammalian ribosome, in which RNA accounts for the bulk of the molecular mass, the three snRNAs (U2, U5 and U6) of the functional human spliceosome represent no more than 5% of its total molecular mass. It is the spliceosomal protein components that organize the splicing active site, deliver the reactive moieties into the active site, and regulate the splicing process. Thus, the spliceosome is a protein-directed metalloribozyme.
The organizational principles observed in the S. pombe ILS complex108 109 are remarkably conserved in the subsequently elucidated structures of the S. cerevisiae spliceosome113"118. U5 snRNP serves as the central pillar of the entire spliceosome, Prp8 anchors the splicing active site, and there is a rigid core in the spliceosome of at least 20 components. The available structural information of the yeast spliceosome enables the proposition of a mechanistic model for a pre-mRNA splicing cycle (FIG. 7c): the Bact complex contains a well-formed active site but cannot yet catalyse the first transesterification owing to separation of the two reactive moieties; such a structural organization is presumably suited for selection of alternative BPSs. Driven by Prp2 and its co-activator Spp2, approximately a dozen proteins in the SF3a, SF3b and RES complexes are dissociated from the Bact complex142157, and a few splicing factors are recruited to form the B* complex, where the first transesterification occurs. The resulting C complex, which contains a free 5' exon and an intron lariat-3' exon, is stabilized by the splicing factors Cwc25 and Yju2 and the NTC proteins Isyl and Cefl. The stable C complex may enable selection of alternative 3'SS. The second transesterification requires translocation of the lariat junction to vacate
space for the 3'SS-3' exon sequence, which is driven by the ATPase/helicase Prpl6 and proceeds in the C* complex. The post-catalytic P complex contains a ligated exon and a free intron lariat. The exon is released by Prp22, resulting in the ILS complex; the intron lariat is dissociated by the ATPase/helicase Prp43 with help from NTC-related protein 1 (Ntrl; also known as Spp382) andNtr2(REFS 167,168).
Intact spliceosomes for cryo-EM studies were either directly purified from cell nuclei or assembled in vitro using synthetic pre-mRNA (Supplementary information SI (table)). The former approach more faithfully reflects cellular circumstances, whereas the latter may yield more homogeneous materials. Notably, the Bact, C and C* structures of the S. cerevisiae spliceosome were obtained from the same EM data set, which was collected on the same sample derived from endogenous yeast nuclear extract113114116. In this case, the distribution of the various spliceosomal complexes may represent that of yeast cells. Intriguingly, the Bact and C complexes were the two most abundant spliceosomal species in the sample, suggesting that they represent rate-limiting regulatory steps in the splicing reaction. The selection of alternative BPSs and 3'SSs may crucially depend on the Bact and C complexes, respectively.
Despite the recent structural advances108"118, important questions remain to be answered. The second transesterification step, particularly the recognition and delivery to the active site of the 3'SS-3' exon sequences, remains poorly understood, and exactly how the eight conserved ATPases/helicases remodel the spliceosomal complexes remains enigmatic. Importantly, how alternative splicing occurs and how alternative BPSs and 3'SSs are selected are yet to be structurally explained.
1. Berget, S. M., Moore, C. & Sharp, P. A. Spliced segments at the 5' terminus of adenovirus 2 late mRNA. Proc. Natl Acad. Sci. USA 74, 3171-3175 (1977).
2. Chow, L. T., Celinas, R. E., Broker, T. R. & Roberts, R. J. An amazing sequence arrangement at the 5' ends of adenovirus 2 messenger RNA. Cell 12, 1-8 (1977).
3. Lerner, M. R. & Steitz, J. A. Antibodies to small nuclear RNAs complexed with proteins are produced by patients with systemic lupus erythematosus. Proc. Natl Acad. Sci. USA 76, 5495-5499 (1979).
4. Lerner, M. R., Boyle, J. A., Mount, S. M., Wolin, S. L. & Steitz, J. A. Are snRNPs involved in splicing? Nature 283, 220-224 (1980).
5. Rogers, J. & Wall, R. A mechanism for RNA splicing. Proc. Natl Acad. Sci. USA 77, 1877-1879 (1980).
6. Hinterberger, M., Pettersson, I. & Steitz, J. A. Isolation of small nuclear ribonucleoproteins containing U1, 112, 114, 115, and 116 RNAs. J. Biol. Chem. 258, 2604-2613 (1983).
7. Mount, S. M., Pettersson, I., Hinterberger, M., Karmas, A. & Steitz, J. A. The Ul small nuclear RNA-protein complex selectively binds a 5' splice site in vitro. Cell 33, 509-518(1983).
8. Padgett, R. A., Mount, S. M., Steitz, J. A. & Sharp, P. A. Splicing of messenger RNA precursors is inhibited by antisera to small nuclear ribonucleoprotein. Ce»35, 101-107 (1983).
9. Yang, V. W., Lerner, M. R., Steitz, J. A. & Flint, S. J. A small nuclear ribonucleoprotein is required for splicing of adenoviral early RNA sequences. Proc. Natl Acad. Sci. USA 78, 1371-1375 (1981).
10. DiMaria, P. R., Kaltwasser, C. & Coldenberg, C. J. Partial purification and properties of a pre-mRNA splicing activity. J. Biol. Chem. 260, 1096-1102 (1985).
11.   Kramer, A., Keller, W., Appel, B. & Luhrmann, R. The 5' terminus of the RNA moiety of U1 small nuclear ribonucleoprotein particles is required for the splicing of messenger RNA precursors. Cell 38, 299-307 (1984).
1 2.  Black, D. L, Chabot, B. & Steitz, J. A. 112 as well as U1 small nuclear ribonucleoproteins are involved in premessenger RNA splicing. Ce»42, 737-750 (1985).
1 3.  Krainer, A. R. & Maniatis, T. Multiple factors including the small nuclear ribonucleoproteins U1 and U2 are necessary for pre-mRNA splicing in vitro. Cell 42, 725-736(1985).
1 4.  Berget, S. M. & Robberson, B. L. U1, 112, and U4/U6 small nuclear ribonucleoproteins are required for in vitro splicing but not polyadenylation. Cell 46, 691-696 (1986).
1 5.  Crabowski, P. J. & Sharp, P. A. Affinity
chromatography of splicing complexes: U2, U5, and U4 + U6 small nuclear ribonucleoprotein particles in the spliceosome. Science 233, 1 294-1 299 (1 986).
1 6.  Pikielny, C. W. & Rosbash, M. Specific small nuclear RNAs are associated with yeast spliceosomes. Cell 45, 869-877 (1986).
1 7.  Coldenberg, C. J. & Hauser, S. D. Accurate and efficient in vitro splicing of purified precursor RNAs specified by early region 2 of the adenovirus 2 genome. Nucleic Acids Res. 11, 1337-1348(1983).
1 8.  Hernandez, N. & Keller, W. Splicing of in vitro
synthesized messenger RNA precursors in HeLa cell extracts. Cell 35, 89-99 (1983).
1 9.  Kole, R. & Weissman, S. M. Accurate in vitro splicing ofhuman beta-globin RNA. Nucleic Acids Res. 10, 5429-5445 (1982).
20.  Padgett, R. A., Hardy, S. F. & Sharp, P. A. Splicing of adenovirus RNA in a cell-free transcription system. Proc. Natl Acad. Sci. USA 80, 5230-5234 (1983).
21. Hardy, S. F, Crabowski, P. J., Padgett, R. A.
& Sharp, P. A. Cofactor requirements of splicing of purified messenger RNA precursors. Nature 308, 375-377 (1984).
22. Krainer, A. R., Maniatis, T, Ruskin, B. & Green, M. R. Normal and mutant human beta-globin pre-mRNAs are faithfully and efficiently spliced in vitro. Cell 36, 993-1005 (1984).
23. Crabowski, P. J., Padgett, R. A. & Sharp, P. A. Messenger RNA splicing in vitro: an excised intervening sequence and a potential intermediate. Ce»37, 415-427 (1984).
24. Padgett, R. A., Konarska, M. M., Crabowski, P. J., Hardy, S. F. & Sharp, P. A. Lariat RNAs as intermediates and products in the splicing of messenger RNA precursors. Science 225, 898-903 (1984).
25. Ruskin, B., Krainer, A. R., Maniatis, T. & Green, M. R. Excision of an intact intron as a novel lariat structure during pre-mRNA splicing in vitro. Cell 38, 317-331 (1984).
26. Steitz, T. A. & Steitz, J. A. A general two-metal-ion mechanism for catalytic RNA. Proc. Natl Acad. Sci. USA 90, 6498-6502 (1993).
27. Sontheimer, E. J., Sun, S. & Piccirilli, J. A. Metal ion catalysis during splicing of premessenger RNA. Nature 388, 801 -805 (1997).
28. Yean, S. L, Wuenschell, G., Termini, J. & Lin, R. J. Metal-ion coordination by U6 small nuclear RNA contributes to catalysis in the spliceosome. Nature 408, 881 -884 (2000).
29. Fica, S. M. etal. RNA catalyses nuclear pre-mRNA splicing. Nature 503, 229-234 (2013).
30. Toor, N., Keating, K. S., Taylor, S. D. & Pyle, A. M. Crystal structure of a self-spliced group II intron. Science 320, 77-82 (2008).
668 I NOVEMBER 2017 I VOLUME 18 www.nature.com/nrm
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
REVIEWS
31. Keating, K. S., Toor, N., Perlman, P. S. & Pyle, A. M. A structural analysis of the group II intron active site and implications for the spliceosome. RNA 16, 1-9 (2010).
32. Brody, E. & Abelson, J. The "spliceosome": yeast pre-messenger RNA associates with a 40S complex in a splicing-dependent reaction. Science 228, 963-967 (1985).
33. Grabowski, P. J., Seiler, S. R. & Sharp, P. A.
A multicomponent complex is involved in the splicing of messenger RNA precursors. Cell 42, 345-353 (1985).
34. Frendewey, D. & Keller, W. Stepwise assembly of a pre-mRNA splicing complex requires U-snRNPs and specific intron sequences. Cell 42, 355-367 (1985).
35. Aebi, M., Hornig, H., Padgett, R. A., Reiser, J.
& Weissmann, C. Sequence requirements for splicing of higher eukaryotic nuclear pre-mRNA. Cell 47, 555-565 (1986).
36. Vijayraghavan, U. etal. Mutations in conserved intron sequences affect multiple steps in the yeast splicing pathway, particularly assembly of the spliceosome. EMBO J. 5, 1683-1695(1986).
37. Newman, A. J., Lin, R. J., Cheng, S. C. & Abelson, J. Molecular consequences of specific intron mutations on yeast mRNA splicing in vivo and in vitro. Cell 42, 335-344(1985).
38. Lamond, A. I., Konarska, M. M. & Sharp, P. A. A mutational analysis of spliceosome assembly: evidence for splice site collaboration during spliceosome formation. Genes Dev. 1, 532-543 (1987).
39. Konarska, M. M. & Sharp, P. A. Electrophoretic separation of complexes involved in the splicing
of precursors to mRNAs. Ce//46, 845-855 (1986).
40. Pikielny, C. W., Rymond, B. C. & Rosbash, M. Electrophoresis of ribonucleoproteins reveals an ordered assembly pathway of yeast splicing complexes. Nature 324, 341-345 (1986).
41. Konarska, M. M. & Sharp, P. A. Interactions between small nuclear ribonucleoprotein particles in formation ofspliceosomes. Ce//49, 763-774 (1987).
42. Cheng, S. C. & Abelson, J. Spliceosome assembly in yeast. Genes Dev. 1, 1014-1027(1987).
43. Bindereif, A. & Green, M. R. An ordered pathway of snRNP binding during mammalian pre-mRNA splicing complex assembly. EMBOJ. 6, 241 5-2424 (1987).
44. Wahl, M. C, Will, C. L. & Luhrmann, R. The spliceosome: design principles of a dynamic RNP machine. Cell 1 36, 701-718(2009).
45. Guthrie, C. & Patterson, B. Spliceosomal snRNAs. Annu. Rev. Genet. 22, 387-419 (1988).
46. Bringmann, P. & Luhrmann, R. Purification of the individual snRNPs U1, U2, U5 and U4/U6 from HeLa cells and characterization of their protein constituents. EMBOJ. 5, 3509-3516(1986).
47. Lossky, M., Anderson, G. J., Jackson, S. P. & Beggs, J. Identification of a yeast snRNP protein and detection of snRNP-snRNP interactions. Ce//51, 1019-1026 (1987).
48. Jackson, S. P., Lossky, M. & Beggs, J. D. Cloning of the RNA8 gene of Saccharomyces cerevisiae, detection of the RNA8 protein, and demonstration that it is essential for nuclear pre-mRNA splicing. Mol. Cell. Biol. 8, 1067-1075 (1988).
49. Tarn, W. Y. et al. Functional association of essential splicing factor(s) with PRP19 in a protein complex. EMBOJ. 13, 2421-2431 (1994).
50. Chan, S. P., Kao, D. I., Tsai, W. Y. & Cheng, S. C. The Prpl 9p-associated complex in spliceosome activation. Science 302, 279-282 (2003).
51. Chabot, B. & Steitz, J. A. Multiple interactions between the splicing substrate and small nuclear ribonucleoproteins in spliceosomes. Mol. Cell. Biol. 7, 281-293 (1987).
52. Parker, R., Siliciano, P. G. & Guthrie, C. Recognition of the TACTAAC box during mRNA splicing in yeast involves base pairing to the U2-like snRNA. Cell 49, 229-239 (1987).
53. Newman, A. & Norman, C. Mutations in yeast U5 snRNA alter the specificity of 5' splice-site cleavage. Ce//65, 115-123 (1991).
54. Newman, A. J. & Norman, C. U5 snRNA interacts with exon sequences at 5' and 3' splice sites. Cell 68, 743-754(1992).
55. Madhani, H. D. & Guthrie, C. A novel base-pairing interaction between U2 and U6 snRNAs suggests a mechanism for the catalytic activation of the spliceosome. Cell 71, 803-81 7 (1992).
56. Wassarman, D. A. & Steitz, J. A. Interactions of small nuclear RNAs with precursor messenger RNA during in vitro splicing. Science 257, 1918-1925 (1992).
Wyatt, J. R., Sontheimer, E. J. & Steitz, J. A. Site-specific cross-linking of mammalian U5 snRNP to the 5' splice site before the first step of pre-mRNA splicing. Genes Dev. 6, 2542-2553 (1992). Lesser, C. F. & Guthrie, C. Mutations in U6 snRNA that alter splice site specificity: implications for the active site. Science 262, 1982-1988 (1993). Sontheimer, E. J. & Steitz, J. A. The U5 and U6 small nuclear RNAs as active site components of the spliceosome. Science 262, 1989-1996 (1993). Kandels-Lewis, S. & Seraphin, B. Involvement of U6 snRNA in 5' splice site selection. Science 262, 2035-2039 (1993).
Newman, A. J., Teigelkamp, S. & Beggs, J. D. snRNA interactions at 5' and 3' splice sites monitored by photo activated crosslinking in yeast spliceosomes. RNA 1,968-980(1995).
Anokhina, M.etai. RNA structure analysis of human spliceosomes reveals a compact 3D arrangement of snRNAs at the catalytic core. EMBO J. 32, 2804-2818(2013).
Jankowsky, E. RNA helicases at work: binding and rearranging. Trends Biochem. Sci. 36, 19-29 (2011). Cordin, O., Hahn, D. & Beggs, J. D. Structure, function and regulation of spliceosomal RNA helicases. Curr. Opin. Cell Biol. 24, 431-438 (201 2). Staley, J. P. & Guthrie, C. Mechanical devices of the spliceosome: motors, clocks, springs, and things. Ce//92, 315-326 (1998).
Raghunathan, P. L. & Guthrie, C. RNA unwinding in U4/ U6 snRNPs requires ATP hydrolysis and the DEIH-box splicing factor Brr2. Curr. Biol. 8, 847-855 (1998). Laggerbauer, B., Achsel, T. & Luhrmann, R. The human U5-200kD DEXH-box protein unwinds U4/U6 RNA duplices in vitro. Proc. Natl Acad. Sci. USA 95, 4188-4192 (1998).
Chen, J. H. & Lin, R. J. The yeast PRP2 protein,
a putative RNA-dependent ATPase, shares extensive
sequence homology with two other pre-mRNA splicing
factors. Nucleic Acids Res. 18, 6447 (1990).
King, D. S. & Beggs, J. D. Interactions of PRP2 protein
with pre-mRNA splicing complexes in Saccharomyces
cerevisiae. Nucleic Acids Res. 18, 6559-6564
(1990).
Kim, S. H. & Lin, R. J. Spliceosome activation by PRP2 ATPase prior to the first transesterification reaction of pre-mRNA splicing. Mol. Cell. Biol. 16, 6810-6819 (1996).
Kim, S. H., Smith, J., Claude, A. & Lin, R. J. The purified yeast pre-mRNA splicing factor PRP2 is an RNA-dependent NTPase. EMBOJ. 11, 231 9-2326 (1992).
Burgess, S., Couto, J. R. & Guthrie, C. A putative ATP binding protein influences the fidelity of branchpoint recognition in yeast splicing. Cell 60, 705-717 (1990).
Schwer, B. & Guthrie, C. PRP1 6 is an RNA-dependent ATPase that interacts transiently with the spliceosome. Nature's^, 494-499 (1991). Company, M., Arenas, J. & Abelson, J. Requirement of the RNA helicase-like protein PRP22 for release of messenger RNA from spliceosomes. Nature 349, 487-493 (1991).
Semlow, D. R., Blanco, M. R., Walter, N. G.
& Staley, J. P. Spliceosomal DEAH-box ATPases
remodel pre-mRNA to activate alternative splice sites.
Cell 164, 985-998 (2016).
Semlow, D. R. & Staley, J. P. Staying on message:
ensuring fidelity in pre-mRNA splicing. Trends Biochem.
Sci. 37, 263-273 (2012).
Weber, G., Trowitzsch, S., Kastner, B., Luhrmann, R. & Wahl, M. C. Functional organization of the Sm core in the crystal structure of human Ul snRNP. EMBOJ. 29,4172-4184 (2010).
Pomeranz Krummel, D. A., Oubridge, C, Leung, A. K., Li, J. & Nagai, K. Crystal structure of human spliceosomal Ul snRNP at 5.5 A resolution. Nature 458, 475-480 (2009).
Kondo, Y., Oubridge, C, van Roon, A. M. & Nagai, K. Crystal structure of human Ul snRNP, a small nuclear ribonucleoprotein particle, reveals the mechanism of 5' splice site recognition. Elite http://dx.doi.org/10.7554/eLife.04986 (201 5). Price, S. R., Evans, P. R. & Nagai, K. Crystal structure of the spliceosomal U2B"-U2A protein complex bound to a fragment of U2 small nuclear RNA. Nature 394, 645-650 (1998).
Sickmier, E. A. etal. Structural basis for polypyrimidine tract recognition by the essential pre-mRNA splicing factor U2AF65. Mol. Cell 23, 49-59 (2006).
Lin, P. C. & Xu, R. M. Structure and assembly of the SF3a splicing factor complex of U2 snRNP. EMBO J. 31, 1579-1590(2012).
Jenkins, J. L, Agrawal, A. A., Gupta, A., Green, M. R. & Kielkopf, C. L. U2AF65 adapts to diverse pre-mRNA splice sites through conformational selection of specific and promiscuous RNA recognition motifs. Nucleic Acids Res. 41, 3859-3873 (201 3). Yoshida, H. etal. A novel 3' splice site recognition by the two zinc fingers in the U2AF small subunit. Genes Dev. 29, 1649-1660 (2015). Leung, A. K., Nagai, K. & Li, J. Structure of the spliceosomal U4 snRNP core domain and its implication for snRNP biogenesis. Nature 473, 536-539 (2011).
Zhou, L. etal. Crystal structures of the Lsm complex bound to the 3' end sequence of U6 small nuclear RNA. Nature 506, 116-120(2014). Montemayor, E. J. et al. Core structure of the U6 small nuclear ribonucleoprotein at 1.7-A resolution. Nat. Struct. Mol. Biol. 21, 544-551 (2014). Galej, W. P., Oubridge, C, Newman, A. J. & Nagai, K. Crystal structure of Prp8 reveals active site cavity of the spliceosome. Nature 493, 638-643 (2013). Mozaffari-Jovin, S. etal. Inhibition of RNA helicase Brr2 by the C-terminal tail of the spliceosomal protein Prp8. Science 341, 80-84 (201 3). Nguyen, T. H. etal. Structural basis of Brr2-Prp8 interactions and implications for U5 snRNP biogenesis and the spliceosome active site. Structure 21, 910-919 (2013).
Cretu, C. et al. Molecular architecture of SF3b and structural consequences of its cancer-related mutations. Mol. Ce//64, 307-31 9 (201 6). Zhou, Z., Sim, J., Griffith, J. & Reed, R. Purification and electron microscopic visualization of functional human spliceosomes. Proc. Natl Acad. Sci. USA 99, 12203-12207 (2002). Jurica, M. S., Licklider, L. J., Gygi, S. R., Grigorieff, N. & Moore, M. J. Purification and characterization of native spliceosomes suitable for three-dimensional structural analysis. RNA 8, 426-439 (2002).
Luhrmann, R. & Stark, H. Structural mapping of spliceosomes by electron microscopy. Curr. Opin. Struct. Biol. 19, 96-102 (2009). Behzadnia, N. etal. Composition and three-dimensional EM structure of double affinity-purified, human prespliceosomal A complexes. EMBO J. 26, 1 737-1 748 (2007). Furman, E. 6i Glitz, D. G. Purification of the spliceosome A-complex and its visualization by electron microscopy. J. Biol. Chem. 270, 15515-15522 (1995).
Boehringer, D. etal. Three-dimensional structure of a pre-catalytic human spliceosomal complex B. Nat. Struct. Mol. Biol. 11, 463-468 (2004). Wolf, E. et al. Exon, intron and splice site locations in the spliceosomal B complex. EMBOJ. 28, 2283-2292 (2009).
Deckert, J. et al. Protein composition and electron microscopy structure of affinity-purified human spliceosomal B complexes isolated under physiological conditions. Mol. Cell. Biol. 26, 5528-5543 (2006).
. Bessonov, S. et al. Characterization of purified human Bact spliceosomal complexes reveals compositional and morphological changes during spliceosome activation and first step catalysis. RNA 16, 2384-2403 (2010).
. Golas, M. M. etal. 3D cryo-EM structure of an active step I spliceosome and localization of its catalytic core. Mol. Cell 40, 927-938 (2010).
. Jurica, M. S., Sousa, D., Moore, M. J. & Grigorieff, N. Three-dimensional structure of C complex spliceosomes by electron microscopy. Nat. Struct. Mol. Biol. 11, 265-269 (2004).
. Ilagan, J. O., Chalkley, R. J., Burlingame, A. L. & Jurica, M. S. Rearrangements within human spliceosomes captured after exon ligation. RNA 1 9, 400-412 (2013).
. Fabrizio, P. et al. The evolutionarily conserved core design of the catalytic activation step of the yeast spliceosome. Mol. Cell 36, 593-608 (2009).
. Ohi, M. D., Ren, L, Wall, J. S., Gould, K. L. & Walz, T. Structural characterization of the fission yeast U5. U2/U6 spliceosome complex. Proc. Natl Acad. Sci. USA 104, 3195-3200 (2007).
. Chen, W. et al. Endogenous U2. U5. U6 snRNA complexes in S. pombe are intron lariat spliceosomes. RNA20, 308-320 (2014).
NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 18 | NOVEMBER 2017 | 669
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
REVIEWS
107. Nguyen, T. H. et al. The architecture of the spliceosomal U4/U6. U5 tri-snRNP. Nature 523, 47-52 (2015).
108. Yan, C. et al. Structure of a yeast spliceosome at 3.6-angstrom resolution. Science 349, 1182-1191 (2015).
109. Hang, J., Wan, R., Yan, C. & Shi, Y Structural basis of pre-mRNA splicing. Science 349, 1191 -1198 (201 5).
110. Wan, R. et al. The 3.8 A structure of the U4/U6. U5 tri-snRNP: insights into spliceosome assembly and catalysis. Science 351, 466-475 (201 6).
111. Nguyen, T H. etal. Cryo-EM structure of the yeast U4/U6. U5 tri-snRNP at 3.7 A resolution. Nature 530, 298-302 (2016).
11 2. Agafonov, D. E. et al. Molecular architecture of
the human U4/U6. U5 tri-snRNP Science 351,
1416-1420 (2016). 11 3. Yan, C, Wan, R., Bai, R., Huang, G. & Shi, Y. Structure
of a yeast activated spliceosome at 3.5 A resolution.
Science 353, 904-911 (2016). 11 4. Wan, R., Yan, C, Bai, R., Huang, G. & Shi, Y.
Structure of a yeast catalytic step I spliceosome at 3.4
A resolution. Science 353, 895-904 (201 6). 11 5. Galej, W. P. et al. Cryo-EM structure of the
spliceosome immediately after branching. Nature
537, 197-201 (2016). 11 6. Yan, C., Wan, R., Bai, R., Huang, G. & Shi, Y. Structure
of a yeast step II catalytically activated spliceosome.
Science 355, 149-155(2017). 11 7. Fica, S. M. et al. Structure of a spliceosome
remodelled for exon ligation. Nature 542, 377-380
(2017).
11 8. Rauhut, R. et al. Molecular architecture of the
Saccharomyces cerevisiae activated spliceosome.
Sc/e/7Ce353, 1399-1405(2016). 11 9. Bertram, K. et al. Cryo-EM structure of a human
spliceosome activated for step 2 of splicing. Nature
542, 318-323 (2017). 1 20. Fica, S. M., Mefford, M. A., Piccirilli, J. A.
& Staley, J. P. Evidence for a group II intron-like
catalytic triplex in the spliceosome. Nat. Struct. Mol.
Biol. 21, 464-471 (2014). 121. Robart, A. R., Chan, R. T, Peters, J. K.,
Rajashankar, K. R. & Toor, N. Crystal structure
of a eukaryotic group II intron lariat. Nature 514,
193-197 (2014). 1 22. Grainger, R. J. & Beggs, J. D. Prp8 protein: at the
heartofthe spliceosome. RNA 11, 533-557 (2005). 1 23. Turner, I. A., Norman, C. M., Churcher, M. J.
& Newman, A. J. Dissection of Prp8 protein defines
multiple interactions with crucial RNA sequences
in the catalytic core ofthe spliceosome. RNA 1 2,
375-386 (2006). 1 24. Galej, W. P., Nguyen, T H., Newman, A. J. & Nagai, K.
Structural studies of the spliceosome: zooming into
the heart of the machine. Curr. Opin. Struct. Biol. 25,
57-66 (2014). 1 25. Yang, K., Zhang, L, Xu, T, Heroux, A. & Zhao, R.
Crystal structure of the beta-finger domain of Prp8
reveals analogy to ribosomal proteins. Proc. Natl
Acad. Sci. USA 105, 13817-13822 (2008). 1 26. Carrey, S. M. etal. A homolog of lariat-debranching
enzyme modulates turnover of branched RNA.
RNA20, 1337-1348 (2014). 1 27. Rasche, N. etal. Cwc2 and its human homologue
RBM22 promote an active conformation of
the spliceosome catalytic centre. EMBO J. 31,
1591-1604 (2012). 1 28. Hahn, C. N. & Scott, H. S. Spliceosome mutations
in hematopoietic malignancies. Nat. Genet. 44, 9-10
(2012).
1 29. Darman, R. B. etal. Cancer-associated SF3B1 hotspot mutations induce cryptic 3' splice site selection through use of a different branch point. Cell Rep. 1 3, 1033-1045 (2015).
1 30. Alsafadi, S. etal. Cancer-associated SF3B1 mutations affect alternative splicing by promoting alternative branchpoint usage. Nat. Commun. 7, 1061 5 (201 6).
1 31. Jacquier, A. & Michel, F. Base-pairing interactions
involving the 5' and 3'-terminal nucleotides of group II
self-splicing introns. J. Mol. Biol. 213, 437-447 (1990).
1 32. Chanfreau, G. & Jacquier, A. Catalytic site components common to both splicing steps of a group II intron. Science 266, 1383-1387(1994).
1 33. Umen, J. G. & Guthrie, C. The second catalytic step of pre-mRNA splicing. RNA 1, 869-885 (1995).
1 34. Luukkonen, B. G. & Seraphin, B. The role of
branchpoint-3' splice site spacing and interaction between intron terminal nucleotides in 3' splice site selection in Saccharomyces cerevisiae. EMBO J. 16, 779-792 (1997).
1 35. Collins, C. A. & Guthrie, C. Genetic interactions between the 5' and 3' splice site consensus sequences and U6 snRNA during the second catalytic step of pre-mRNA splicing. RNA 7, 1845-1854 (2001).
1 36. Frank, D. & Guthrie, C. An essential splicing
factor, SLU7, mediates 3' splice site choice in yeast. Genes Dev. 6, 2112-2124 (1992).
1 37. Umen, J. G. & Guthrie, C. Prpl 6p, Slu7p, and Prp8p interact with the 3' splice site in two distinct stages during the second catalytic step of pre-mRNA splicing. RNA 1,584-597(1995).
1 38. Umen, J. G. & Guthrie, C. Mutagenesis of the yeast gene PRP8 reveals domains governing the specificity and fidelity of 3' splice site selection. Genetics 1 43, 723-739 (1996).
1 39. Chua, K. & Reed, R. The RNA splicing factor hSlu7 is required for correct 3' splice-site choice. Nature 402, 207-210 (1999).
1 40. Chen, W. & Moore, M. J. The spliceosome: disorder and dynamics defined. Curr. Opin. Struct. Biol. 24, 141-149 (2014).
1 41. Ohi, M. D. & Gould, K. L Characterization of
interactions among the Cef 1 p-Prpl 9p-associated splicing complex. RNA 8, 798-81 5 (2002).
1 42. Ohrt, T etal. Prp2-mediated protein rearrangements at the catalytic core of the spliceosome as revealed bydcFCCS. RNA 18, 1244-1256(2012).
1 43. Grainger, R. J., Barrass, J. D., Jacquier, A., Rain, J. C. & Beggs, J. D. Physical and genetic interactions of yeast Cwc21 p, an ortholog of human SRm300/ SRRM2, suggest a role at the catalytic center of the spliceosome. RNA 15, 2161-2173 (2009).
1 44. Warkocki, Z. et al. Reconstitution of both steps of Saccharomyces cerevisiae splicing with purified spliceosomal components. Nat. Struct. Mol. Biol. 16, 1237-1243 (2009).
1 45. Chiu, Y F. et al. Cwc25 is a novel splicing factor required after Prp2 and Yju2 to facilitate the first catalytic reaction. Mol. Cell. Biol. 29, 5671-5678 (2009).
1 46. Krishnan, R. etal. Biased Brownian ratcheting leads to pre-mRNA remodeling and capture prior to first-step splicing. Nat. Struct. Mol. Biol. 20, 1450-1457 (2013).
1 47. Ohrt, T etal. Molecular dissection of step 2 catalysis of yeast pre-mRNA splicing investigated in a purified system. RNA 19, 902-915 (2013).
1 48. James, S. A., Turner, W. & Schwer, B. How Slu7 and Prpl 8 cooperate in the second step of yeast pre-mRNA splicing. RNA 8, 1068-1077 (2002).
1 49. Zhang, X. & Schwer, B. Functional and physical interaction between the yeast splicing factors Slu7 and Prpl 8. Nucleic Acids Res. 25, 2146-21 52 (1997).
1 50. Santos, K. F. etal. Structural basis for functional cooperation between tandem helicase cassettes in Brr2-mediated remodeling of the spliceosome. Proc. Natl Acad. Sci. USA 109, 1 7418-1 7423 (201 2).
1 51. Hahn, D., Kudla, G., Tollervey, D. & Beggs, J. D. Brr2p-mediated conformational rearrangements in the spliceosome during activation and substrate repositioning. Genes Dev. 26, 2408-2421 (2012).
1 52. Huang, Y H., Chung, C. S., Kao, D. I., Kao, T C. & Cheng, S. C. Sadl counteracts Brr2-mediated dissociation of U4/U6. U5 in tri-snRNP homeostasis. Mol. Cell. Biol. 34, 210-220 (2014).
1 53. McPheeters, D. S. & Muhlenkamp, P. Spatial
organization of protein-RNA interactions in the branch
site-3' splice site region during pre-mRNA splicing in yeast. Mol. Cell. Biol. 23, 41 74-4186 (2003).
1 54. Schneider, C. et al. Dynamic contacts of U2, RES, Cwc25, Prp8 and Prp45 proteins with the pre-mRNA branch-site and 3' splice site during catalytic activation and step 1 catalysis in yeast spliceosomes. PLoS Genet. 11, el005539 (2015).
1 55. Edwalds-Gilbert, G., Kim, D. H., Silverman, E.
& Lin, R. J. Definition of a spliceosome interaction domain in yeast Prp2 ATPase. RNA 10, 210-220 (2004).
1 56. Liu, H. L. & Cheng, S. C. The interaction of Prp2 with a defined region of the intron is required for the first splicing reaction. Mol. Cell. Biol. 32, 5056-5066 (2012).
1 57. Warkocki, Z. etal. The G-patch protein Spp2 couples the spliceosome-stimulated ATPase activity of the DEAH-box protein Prp2 to catalytic activation of the spliceosome. Genes Dev. 29, 94-107 (2015).
1 58. Teigelkamp, S., McGarvey, M., Plumpton, M.
& Beggs, J. D. The splicing factor PRP2, a putative RNA helicase, interacts directly with pre-mRNA. EMBO J. 13, 888-897(1994).
1 59. Lardelli, R. M., Thompson, J. X., Yates, J. R. 3rd & Stevens, S. W. Release of SF3 from the intron branchpoint activates the first step of pre-mRNA splicing. RNA 16, 516-528 (2010).
1 60. Schwer, B. & Guthrie, C. A conformational
rearrangement in the spliceosome is dependent on PRP1 6 and ATP hydrolysis. EMBO J. 11, 5033-5039 (1992).
1 61. Tseng, C. K., Liu, H. L & Cheng, S. C. DEAH-box ATPase Prpl 6 has dual roles in remodeling ofthe spliceosome in catalytic steps. RNA 17, 145-154 (2011).
1 62. Schwer, B. A conformational rearrangement in the spliceosome sets the stage for Prp22-dependent mRNA release. Mol. Cell 30, 743-754 (2008).
1 63. Schwer, B. & Gross, C. H. Prp22, a DExH-box RNA helicase, plays two distinct roles in yeast pre-mRNA splicing. EMBOJ. 17, 2086-2094 (1998).
1 64. Sharp, P. A. On the origin of RNA splicing and introns. Ce//42, 397-400 (1985).
1 65. Cech, T R. The generality of self-splicing RNA: relationship to nuclear mRNA splicing. Cell 44, 207-210 (1986).
1 66. Maroney, P. A., Romfo, C. M. & Nilsen, T W.
Functional recognition of 5' splice site by U4/U6. U5 tri-snRNP defines a novel ATP-dependent step in early spliceosome assembly. Mol. Cell 6, 317-328 (2000).
1 67. Tsai, R. T et al. Spliceosome disassembly catalyzed by Prp43 and its associated components Ntrl and Ntr2. Genes Dev. 19, 2991-3003 (2005).
1 68. Fourmann, J. B., Tauchert, M. J., Ficner, R., Fabrizio, P. &Luhrmann, R. Regulation of Prp43-mediated disassembly of spliceosomes by its cofactors Ntrl and Ntr2. Nucleic Acids Res. 45, 4068-4080 (201 7).
Acknowledgements
The author apologizes to those colleagues whose important work is not cited in this review owing to space limitation or inadvertent omission. The author thanks R. Wan, R. Bai, J. Hang, X. Zhang, X. Zhan and C. Yan for preparation of the figures and suggestions on text and references. This work was supported by funds from the Beijing Advanced Innovation Center for Structural Biology at Tsinghua University, Beijing, China.
Competing interests statement
The author declares no competing interests.
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
SUPPLEMENTARY INFORMATION
See online article: Sl(table) | S2 (figure) | S3 (figure) | S4 (figure) ALL LINKS ARE ACTIVE IN THE ONLINE PDF
670| NOVEMBER 2017 | VOLUME 18
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
www.nature.com/nrm