Cell, Vol. 63, 293-302,
October
19, 1990, Copyright
0 1990 by Cell Press
Targeted snRNP Depletion Reveals an Additional Role for Mammalian Ul snRNP in Spliceosome Assembly Silvia M. L. Barabino, Benjamin J. Blencowe, Ursula Ryder, Brian S. Sproat, and Angus I. Lamond European Molecular Biology Laboratory Meyerhofstrasse 1 Postfach 102209 D6900-Heidelberg Federal Republic of Germany
Summary HeLa cell nuclear splicing extracts have been prepared that are specifically and efficiently depleted of Ul, U2, or U4/U6 snRNPs by antlsense affinity chromatography using biotinylated 2’-OMe RNA oligonucleotides. Removal of each snRNP particle prevents premRNA splicing but arrests spliceosome formation at different stages of assembly. Mixing extracts depleted for different snRNP particles restores formation of functional splicing complexes. Specific binding of factors to the 3’ splice site region is still detected in snRNP-depleted extracts. Depletion of Ul snRNP impairs stable binding of U2 snRNP to the pre-mRNA branch site. This role of Ul snRNP in promoting stable prespllcing complex formation is independent of the Ul snRNA-5’ splice site interaction. Introduction The availability of cellular extracts that can accurately splice exogenously added pre-mRNA substrates has allowed a detailed analysis of the composition and function of the splicing apparatus. In vitro studies, using extracts derived from both mammalian and yeast cells, have shown that spliceosomes are comprised of multiple small nuclear ribonucleoprotein (snRNP) particles, specifically Ul, U2, U5, and U4/U6 snRNPs (for recent reviews, see Steitz et al., 1988; Guthrie and Patterson, 1988; Lamond et al., 1990). In addition, a number of protein factors that may act independently of snRNPs have been reported to be spliceosome components. These include hnRNP proteins (Choi et al., 1986) snRNP accessory factors (Kramer, 1988; Ruskin et al., 1988; Zamore and Green, 1989) intron binding proteins (Tazi et al., 1986; Gerke and Steitz, 1986; Garcia-Blanc0 et al., 1989) and other proteins of unknown function (Fu and Maniatis, 1990). Recent studies in Saccharomyces cerevisiae have also identified specific proteins that associate with splicing complexes (Chang et al., 1988; Whittaker et al., 1990). The assembly of functional spliceosomes involves a stepwise pathway of snRNP (and probably protein) binding to pre-mRNA substrates (Frendewey and Keller, 1985; Grabowski and Sharp, 1986; Pikielny et al., 1986; Konarska and Sharp, 1987; Bindereif and Green, 1987). Several types of experiments have shown that snRNPs are essential for splicing activity in vitro. For example,
sequence-specific inactivation of Ul, U2, and U4/U6 snRNPs has been demonstrated by cleavage with RNAase H in the presence of complementary DNA oligonucleotides (Kramer et al., 1984; Krainer and Maniatis, 1985; Black et al., 1985; Black and Steitz, 1986; Berget and Robberson, 1988) and more recently by antisense inhibition using 2’-OMe RNA oligonucleotides (Barabino et al., 1989; Blencowe et al., 1989; Lamond et al., 1989). Analysis of splicing complex assembly has demonstrated that inactivation of separate snRNP particles blocks spliceosome formation at different stages and that U2 snRNP in particular has at least two functional domains that are required for distinct steps in the assembly pathway (Frendewey et al., 1987; Chabot and Steitz, 1987; Zillmann et al., 1988; Lamond et al., 1989; Hamm et al., 1989). Current models envisage formation of a presplicing complex, containing Ul and U2 snRNPs interacting with the 5’splice site and branch site, respectively, followed by subsequent binding of U4, U5, and U6 snRNAs, probably in the form of a U4/U5/U6 triple snRNP particle (Konarska and Sharp, 1987; Cheng and Abelson, 1987). The formation of a functional spliceosome, however, almost certainly involves a large number of separate interactions and the assembly of multiple forms of presplicing complexes. A major goal of current studies is, therefore, to unravel the detailed snRNP-snRNP and snRNP-pre-mRNA interactions that lie at the heart of the splicing mechanism. In this paper we report the preparation of HeLa cell nuclear splicing extracts that have been efficiently and specifically depleted for Ul, U2, or U4/U6 snRNPs. This was achieved by an antisense affinity depletion strategy (AAD), using biotinylated oligonucleotides made of 2’-OMe RNA. Characterization of splicing complex assembly in such snRNP-depleted extracts reveals new information concerning snRNP function during spliceosome formation. Results Preparation of snRNP-Depleted HeLa Cell Nuclear Extracts Biotinylated, antisense 2’-CMe RNA oligonucleotides were prepared complementary to specific regions of Ul, U2, and U6 snRNAs (Table 1). For optimal depletion efficiency during the preparation of HeLa cell nuclear extracts, it was important to use oligonucleotides containing 2’-OMe guanosine and not 2’-OMe inosine as used previously (Barabino et al., 1989; Blencowe et al., 1989; Lamond et al., 1989). We note that with 2’-OMe guanosine-containing oligonucleotides an increased level of specific snRNP binding is accompanied by a small increase in nonspecific binding (our unpublished data). A protocol for preparing HeLa cell nuclear extracts efficiently depleted for targeted snRNP particles is described in detail in Experimental Procedures and outlined schematically in Figure 8. Optimal conditions of oligonucleotide concentration, time of incubation, etc., differed for the individual snRNP particles and were determined empirically in each case.
Cell 294
Table
1. Sequence
of 2’-OMe
Oligoribonucleotides
Oligonucleotide Reference
Complementary snRNA
Complementary Nucleotides
Sequence
69 114 115 75
Ul u2 u2 U6
1-13 1-14 31-45 62-101
GCCAGGUAAGUAU GGCCGAGAAGCGAU GAACAGAUACUACAC AUGGAACGCUUCACGAAUUU
All oligonucleotides are made of 2’-OMe RNA. Biotinylated oligonucleotides base pairing deoxycytidines as described by Sproat et al. (1969).
HeLa cell nuclear extracts specifically depleted for Ul, U2, or U4/U6 snRNPs were prepared using this AAD strategy. As a control a mock-depleted extract not exposed to antisense oligonucleotide was prepared in parallel. The efficiency of both snRNA and snRNP depletion for each extract was determined by gel electrophoresis and Northern hybridization analyses (Figures 1 and 2). Figures 1A and 1B show the total RNA present in the control (lanes 1) and in each depleted nuclear extract (lanes 2-4) detected by ethidium bromide staining (A) and by Northern hybridization (B). The most efficient snRNA removal was reproducibly observed for Ul and U4 (Figures 1A and lB, lanes 2 and 3). In the case of U6 and U2, a low level of snRNAs consistently remained after depletion (Figures 1A and lB, lanes 2 and 4). Only minor changes in the relative levels of nontargeted RNA species were observed. The snRNP particles in each extract were analyzed by nondenaturing gel electrophoresis (Figure 2) and detected by Northern hybridization using riboprobes for either Ul (A), Ui and U2 (B), U4 (C), or U5 (D) snRNAs. This shows essentially complete removal of the targeted snRNP particle from the corresponding depleted extract with little or no change in the levels of nontargeted snRNPs in each case. Using this assay we also detected no change in the structures of nontargeted snRNP particles, with the exception that U5 snRNA no longer migrates in the position of the U4/UW6 triple snRNP in the UWUG-depleted extract (Figure 20, lane 2). The pattern of U5 is not, however, altered in either the U2- or Ul-depleted extracts (Figure 2D, lanes 3 and 4). This strongly supports the existence of a stable complex containing U5 together with U4/U6 snRNPs, as previously reported (Konarska and Sharp, 1987; Cheng and Abelson, 1987; Lossky et al., 1987; Black and Pinto, 1989). The level of snRNP depletion detected in Figure 2 is significantly higher than the apparent level of removal of U6 and U2 snRNAs (cf. Figure lB, lanes 2 and 4, with Figure 2). The nondepleted U2 snRNA detected by Northern hybridization may therefore result from a subfraction of U2 not assembled into snRNP particles or represent a crosshybridizing, nonfunctional U2 species. The nondepleted U6 snRNA must correspond to a fraction of U6 not assembled into U4/U6 snRNP as apparently quantitative removal of U4 was obtained with the UG-specific antisense 2’-OMe RNA oligonucleotide. From the data presented in Figures 1 and 2 we conclude that the AAD protocol results in extracts efficiently and specifically depleted of targeted snRNP particles.
each
(5’ to 3’)
have four tandem
Length
Site of Biotinylation
13 14 15 20
3’ 3’ 3 5’
biotin residues
linked through
additional
non-
-
Characterization of Splicing Activity and Complex Assembly in snRNP-Depleted Extracts The snRNP-depleted and control extracts described above were assayed for their ability to splice exogenously added pre-mRNA and to assemble snRNP-pre-mRNA complexes (Figures 3 and 4). Figure 3A shows that all of the snRNPdepleted extracts are unable to splice a pre-mRNA substrate containing the first intron of the adenovirus major late transcript (lanes l-3), although the same substrate is spliced by the mock-depleted control extract (lane C). However, complementation experiments show that splicing activity can be restored by mixing together any pairwise combination of extracts depleted for different snRNPs (lanes 4-6). An identical result was obtained using a rabbit 8-globin pre-mRNA substrate (data not shown). The restoration of splicing activity by complementation indicates that the lack of splicing activity in each depleted
u5 U6
1234 U5 U6
12 Figure 1. Depletion Nuclear Extracts
34 of Ul,
U2, and U4/U6
snRNPs
from
HeLa Cell
Analysis of RNA recovered from snRNP-depleted nuclear extracts. RNA is shown separated on a 10% polyacrylamide-urea gel, detected by ethidium staining (A) and by Northern hybridization with snRNAspecific riboprobes (B). Lanes 1 show RNA recovered from a mockdepleted nuclear extract. Lanes 2, 3. and 4 show RNA recovered from nuclear extracts depleted with oligonucleotides complementary to U6, Ul, and U2 snRNAs, respectively.
$tsense
Affinity
Depletion
A
of Mammalian
B u2
C
D
U2+Ul I
October
19, 1990, Copyright
0 1990 by Cell Press
Targeted snRNP Depletion Reveals an Additional Role for Mammalian Ul snRNP in Spliceosome Assembly Silvia M. L. Barabino, Benjamin J. Blencowe, Ursula Ryder, Brian S. Sproat, and Angus I. Lamond European Molecular Biology Laboratory Meyerhofstrasse 1 Postfach 102209 D6900-Heidelberg Federal Republic of Germany
Summary HeLa cell nuclear splicing extracts have been prepared that are specifically and efficiently depleted of Ul, U2, or U4/U6 snRNPs by antlsense affinity chromatography using biotinylated 2’-OMe RNA oligonucleotides. Removal of each snRNP particle prevents premRNA splicing but arrests spliceosome formation at different stages of assembly. Mixing extracts depleted for different snRNP particles restores formation of functional splicing complexes. Specific binding of factors to the 3’ splice site region is still detected in snRNP-depleted extracts. Depletion of Ul snRNP impairs stable binding of U2 snRNP to the pre-mRNA branch site. This role of Ul snRNP in promoting stable prespllcing complex formation is independent of the Ul snRNA-5’ splice site interaction. Introduction The availability of cellular extracts that can accurately splice exogenously added pre-mRNA substrates has allowed a detailed analysis of the composition and function of the splicing apparatus. In vitro studies, using extracts derived from both mammalian and yeast cells, have shown that spliceosomes are comprised of multiple small nuclear ribonucleoprotein (snRNP) particles, specifically Ul, U2, U5, and U4/U6 snRNPs (for recent reviews, see Steitz et al., 1988; Guthrie and Patterson, 1988; Lamond et al., 1990). In addition, a number of protein factors that may act independently of snRNPs have been reported to be spliceosome components. These include hnRNP proteins (Choi et al., 1986) snRNP accessory factors (Kramer, 1988; Ruskin et al., 1988; Zamore and Green, 1989) intron binding proteins (Tazi et al., 1986; Gerke and Steitz, 1986; Garcia-Blanc0 et al., 1989) and other proteins of unknown function (Fu and Maniatis, 1990). Recent studies in Saccharomyces cerevisiae have also identified specific proteins that associate with splicing complexes (Chang et al., 1988; Whittaker et al., 1990). The assembly of functional spliceosomes involves a stepwise pathway of snRNP (and probably protein) binding to pre-mRNA substrates (Frendewey and Keller, 1985; Grabowski and Sharp, 1986; Pikielny et al., 1986; Konarska and Sharp, 1987; Bindereif and Green, 1987). Several types of experiments have shown that snRNPs are essential for splicing activity in vitro. For example,
sequence-specific inactivation of Ul, U2, and U4/U6 snRNPs has been demonstrated by cleavage with RNAase H in the presence of complementary DNA oligonucleotides (Kramer et al., 1984; Krainer and Maniatis, 1985; Black et al., 1985; Black and Steitz, 1986; Berget and Robberson, 1988) and more recently by antisense inhibition using 2’-OMe RNA oligonucleotides (Barabino et al., 1989; Blencowe et al., 1989; Lamond et al., 1989). Analysis of splicing complex assembly has demonstrated that inactivation of separate snRNP particles blocks spliceosome formation at different stages and that U2 snRNP in particular has at least two functional domains that are required for distinct steps in the assembly pathway (Frendewey et al., 1987; Chabot and Steitz, 1987; Zillmann et al., 1988; Lamond et al., 1989; Hamm et al., 1989). Current models envisage formation of a presplicing complex, containing Ul and U2 snRNPs interacting with the 5’splice site and branch site, respectively, followed by subsequent binding of U4, U5, and U6 snRNAs, probably in the form of a U4/U5/U6 triple snRNP particle (Konarska and Sharp, 1987; Cheng and Abelson, 1987). The formation of a functional spliceosome, however, almost certainly involves a large number of separate interactions and the assembly of multiple forms of presplicing complexes. A major goal of current studies is, therefore, to unravel the detailed snRNP-snRNP and snRNP-pre-mRNA interactions that lie at the heart of the splicing mechanism. In this paper we report the preparation of HeLa cell nuclear splicing extracts that have been efficiently and specifically depleted for Ul, U2, or U4/U6 snRNPs. This was achieved by an antisense affinity depletion strategy (AAD), using biotinylated oligonucleotides made of 2’-OMe RNA. Characterization of splicing complex assembly in such snRNP-depleted extracts reveals new information concerning snRNP function during spliceosome formation. Results Preparation of snRNP-Depleted HeLa Cell Nuclear Extracts Biotinylated, antisense 2’-CMe RNA oligonucleotides were prepared complementary to specific regions of Ul, U2, and U6 snRNAs (Table 1). For optimal depletion efficiency during the preparation of HeLa cell nuclear extracts, it was important to use oligonucleotides containing 2’-OMe guanosine and not 2’-OMe inosine as used previously (Barabino et al., 1989; Blencowe et al., 1989; Lamond et al., 1989). We note that with 2’-OMe guanosine-containing oligonucleotides an increased level of specific snRNP binding is accompanied by a small increase in nonspecific binding (our unpublished data). A protocol for preparing HeLa cell nuclear extracts efficiently depleted for targeted snRNP particles is described in detail in Experimental Procedures and outlined schematically in Figure 8. Optimal conditions of oligonucleotide concentration, time of incubation, etc., differed for the individual snRNP particles and were determined empirically in each case.
Cell 294
Table
1. Sequence
of 2’-OMe
Oligoribonucleotides
Oligonucleotide Reference
Complementary snRNA
Complementary Nucleotides
Sequence
69 114 115 75
Ul u2 u2 U6
1-13 1-14 31-45 62-101
GCCAGGUAAGUAU GGCCGAGAAGCGAU GAACAGAUACUACAC AUGGAACGCUUCACGAAUUU
All oligonucleotides are made of 2’-OMe RNA. Biotinylated oligonucleotides base pairing deoxycytidines as described by Sproat et al. (1969).
HeLa cell nuclear extracts specifically depleted for Ul, U2, or U4/U6 snRNPs were prepared using this AAD strategy. As a control a mock-depleted extract not exposed to antisense oligonucleotide was prepared in parallel. The efficiency of both snRNA and snRNP depletion for each extract was determined by gel electrophoresis and Northern hybridization analyses (Figures 1 and 2). Figures 1A and 1B show the total RNA present in the control (lanes 1) and in each depleted nuclear extract (lanes 2-4) detected by ethidium bromide staining (A) and by Northern hybridization (B). The most efficient snRNA removal was reproducibly observed for Ul and U4 (Figures 1A and lB, lanes 2 and 3). In the case of U6 and U2, a low level of snRNAs consistently remained after depletion (Figures 1A and lB, lanes 2 and 4). Only minor changes in the relative levels of nontargeted RNA species were observed. The snRNP particles in each extract were analyzed by nondenaturing gel electrophoresis (Figure 2) and detected by Northern hybridization using riboprobes for either Ul (A), Ui and U2 (B), U4 (C), or U5 (D) snRNAs. This shows essentially complete removal of the targeted snRNP particle from the corresponding depleted extract with little or no change in the levels of nontargeted snRNPs in each case. Using this assay we also detected no change in the structures of nontargeted snRNP particles, with the exception that U5 snRNA no longer migrates in the position of the U4/UW6 triple snRNP in the UWUG-depleted extract (Figure 20, lane 2). The pattern of U5 is not, however, altered in either the U2- or Ul-depleted extracts (Figure 2D, lanes 3 and 4). This strongly supports the existence of a stable complex containing U5 together with U4/U6 snRNPs, as previously reported (Konarska and Sharp, 1987; Cheng and Abelson, 1987; Lossky et al., 1987; Black and Pinto, 1989). The level of snRNP depletion detected in Figure 2 is significantly higher than the apparent level of removal of U6 and U2 snRNAs (cf. Figure lB, lanes 2 and 4, with Figure 2). The nondepleted U2 snRNA detected by Northern hybridization may therefore result from a subfraction of U2 not assembled into snRNP particles or represent a crosshybridizing, nonfunctional U2 species. The nondepleted U6 snRNA must correspond to a fraction of U6 not assembled into U4/U6 snRNP as apparently quantitative removal of U4 was obtained with the UG-specific antisense 2’-OMe RNA oligonucleotide. From the data presented in Figures 1 and 2 we conclude that the AAD protocol results in extracts efficiently and specifically depleted of targeted snRNP particles.
each
(5’ to 3’)
have four tandem
Length
Site of Biotinylation
13 14 15 20
3’ 3’ 3 5’
biotin residues
linked through
additional
non-
-
Characterization of Splicing Activity and Complex Assembly in snRNP-Depleted Extracts The snRNP-depleted and control extracts described above were assayed for their ability to splice exogenously added pre-mRNA and to assemble snRNP-pre-mRNA complexes (Figures 3 and 4). Figure 3A shows that all of the snRNPdepleted extracts are unable to splice a pre-mRNA substrate containing the first intron of the adenovirus major late transcript (lanes l-3), although the same substrate is spliced by the mock-depleted control extract (lane C). However, complementation experiments show that splicing activity can be restored by mixing together any pairwise combination of extracts depleted for different snRNPs (lanes 4-6). An identical result was obtained using a rabbit 8-globin pre-mRNA substrate (data not shown). The restoration of splicing activity by complementation indicates that the lack of splicing activity in each depleted
u5 U6
1234 U5 U6
12 Figure 1. Depletion Nuclear Extracts
34 of Ul,
U2, and U4/U6
snRNPs
from
HeLa Cell
Analysis of RNA recovered from snRNP-depleted nuclear extracts. RNA is shown separated on a 10% polyacrylamide-urea gel, detected by ethidium staining (A) and by Northern hybridization with snRNAspecific riboprobes (B). Lanes 1 show RNA recovered from a mockdepleted nuclear extract. Lanes 2, 3. and 4 show RNA recovered from nuclear extracts depleted with oligonucleotides complementary to U6, Ul, and U2 snRNAs, respectively.
$tsense
Affinity
Depletion
A
of Mammalian
B u2
C
D
U2+Ul I
