|Table of Contents|
Table of Contents
List of Tables
List of Figures
Chapter 1. Introduction
Chapter 2. Materials and methods
Chapter 3. Identification of intron 5 RNA contact sites on CBP2 protein
Chapter 4. Mutational analysis of the N-terminus of CBP2
Chapter 5. Summary and perspectives
List of references
RNA-PROTEIN INTERACTIONS OF A MITOCHONDRIAL
GROUP I INTRON IN Saccharomyces cerevisiae
HYMAVATHI K. TIRUPATI
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1998
I thank Dr. Alfred S. Lewin, my mentor and supervisor, for his constant support in both academic and personal matters. He gave me immense freedom in the conduct of experiments and stood by me during difficult times. I knew I could always count on him. I would also like to thank all the past and present colleagues in the lab, who have made life in the lab enjoyable and friendly.
I would like to thank the members of my advisory committee, Drs. Bert
Flanegan, Henry Baker, and Phil Laipis, for the various constructive suggestions they have made during the course of the investigation and for taking the time to review various documents, including this one. I also appreciate the input from all the other faculty and students in the department, which has helped in my personal growth. I also thank the other personnel in the Department, especially Joyce Conners and Brad Moore, for their efficient help.
My special thanks go to Chandu, my husband, who has stood by me during all times and made this all possible.
TABLE OF CONTENTS
LIST OF TABLES. .. .....................v
LIST OF FIGURES .. ....................vi
ABSTRACT .. ........ ..............vii
1 INTRODUCTION. ...................
General Introduction .. .. ..................
Group I Introns. .. ..................4
Protein Facilitated Splicing ... ..............1
Two-component System of Cbp2 and Intron 5 RNA ... .. ....16 Major Objective of Dissertation .. ..............19
2 MATERIALS AND METHODS. .. ............21
Over-expression and Purification of Cbp2 .. .........21
In vitro Transcription. .. ... .............23
UV-crosslinking and Generation of Peptides. .........23
Site-directed Mutagenesis .. ...............27
In vitro Splicing Assay. ... .............29
Partial Proteolysis of Cbp2. .. ..............30
Equilibrium Binding Analysis. .. .............30
3 IDENTIFICATION OF INTRON 5 RNA CONTACT SITES ON CBP2
PROTEIN .. .... .................32
Introduction. .. ...................32
Results. .. .....................35
Discussion .. .. ...................46
4 MUTATIONAL ANALYSIS OF THE N-TERMINUS OF CBP2 . 55
Introduction. .. ...................55
Results. .. .....................58
Discussion. .. ....................100
5 SUMMARY AND PERSPECTIVES. ......... ..112
LIST OF REFERENCES. ............... ...130
BIOGRAPHICAL SKETCH. .. ...............149
LIST OF TABLES
2-1 Oligonucleotides used for mutagenesis of Cbp2 . . . . 28
4-1 Description of Cbp2 mutants . . . . . . . 59
4-2 Rate measurements for wild-type and mutant Cbp2 . . . 76 4-3 Dissociation constants of Cbp2 mutants . . . . 81
LIST OF FIGURES
1-1 Proposed secondary structure of yeast apocytochrome b intron 5 RNA. 6 3-1 Optimization of UV-dosage for crosslinking ... .......... .37
3-2 Chemical cleavage of Cbp2-intron 5 RNA complexes ... ...... 40
3-3 Confirmation of the crosslink site in the N-terminus of Cbp2. .. 44 3-4 Summary of UV-crosslinking results .... ............ .47
4-1 Western analysis of Cbp2 mutants .... ............. ..61
4-2 Functional analysis of deletion (aal7-aa28) and triple aromatic
mutants ......... ....................... .64
4-3 Partial proteolytic profiles of deletion (aal 7-aa28) and triple aromatic
mutants ......... ....................... .66
4-4 Time course of splicing for wild-type and mutant Cbp2 ......... 69
4-5 Splicing rates of wild-type and mutant Cbp2 ... .......... .72
4-6 Double filter-binding assay of wild-type and mutant Cbp2 ....... 77 4-7 UV-crosslinking of wild-type and mutant Cbp2 to intron 5 RNA 83 4-8 Effect of mutant proteins on wild-type Cbp2-mediated splicing 86 4-9 Effect of increasing concentrations of wild-type Cbp2 on proteinmediated splicing ....... ................... ..90
4-10 Effect of tRNA addition on wild-type Cbp2-mediated splicing . 93 4-11 Effect of mutant proteins on wild-type Cbp2-mediated splicing at
low total protein to RNA ratios ..... .............. .97
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
RNA-PROTEIN INTERACTIONS OF A MITOCHONDRIAL
GROUP I INTRON IN Saccharomyces cerevisiae
Hymavathi K. Tirupati
Chairperson: Alfred S. Lewin
Major Department: Molecular Genetics and Microbiology
Group I introns and associated proteins represent simple but valuable systems for understanding more complex RNP systems such as the spliceosomes or ribosomes, which employ multiple RNA-protein interactions. The terminal intron of yeast cytochrome b pre-mRNA (a group I intron) requires the nuclear protein Cbp2 for splicing in vivo. However, in vitro, this intron can be made to either selfsplice or undergo protein-facilitated splicing by altering the Mg2+ concentration. Since catalysis is intrinsic to RNA, the protein is believed to promote RNA folding at secondary and tertiary structure levels, leading to the formation of a catalytically competent intron. Therefore, this two-component system provides a model for understanding the role of proteins in promoting RNA folding.
The present study was aimed at identifying critical RNA binding sites on Cbp2 and gaining insights into Cbp2-intron 5 RNA interactions. 32p-labeled intron 5 RNA was UV-crosslinked to Cbp2, and the crosslink sites identified by chemical cleavage and label transfer. These experiments demonstrated that the termini of Cbp2 contain important RNA binding sites. A 12-amino acid region (aa17-28) in the N-terminal contact site (rich in basic and aromatic residues) was targeted for mutagenesis, and mutant proteins characterized for RNA binding and stimulation of splicing. Mutations in this region resulted in no, partial, and complete loss of function, demonstrating the importance of this N-terminal region for stimulation of RNA splicing. These studies also led to the finding that Cbp2 stimulates splicing only in a narrow range of concentrations, with higher concentrations being inhibitory. Addition of a non-specific competitor tRNA attenuated this inhibition, demonstrating the non-specific RNA binding ability of Cbp2.
The current study has identified an important RNA binding region (aal7-28aa) in the N-terminus of Cbp2. A tyrosine at position 21 is indispensable while three charged residues at positions 20, 22 and 24 are important for Cbp2 function. An offshoot of the current study is the identification of an RNA chaperone function for Cbp2. Cbp2 appears to engage in both non-specific and specific interactions with intron 5 RNA. At inhibitory concentrations of the protein, however, nonspecific interactions may predominate and preclude the formation of specific contacts that promote catalysis. This newly characterized chaperone function of Cbp2 may be important for promoting correct intron 5 RNA folding in vivo.
Ribozymes, or catalytic RNA molecules, have been shown to catalyze reactions at phosphorus centers with RNA or DNA as the substrate (Cech, 1987; Herschlag and Cech, 1990; Robertson and Joyce, 1990; Forster and Altman, 1990). These reactions include transesterification or hydrolysis of phosphate diesters or phosphate monoesters. For instance, ribozymes have been shown to catalyze the polymerization of RNA monomer strands (Been and Cech, 1988), replication of RNA strands (Green and Szostak, 1992), and hydrolysis at internal processing sites (Decatur et al., 1995; Einvik et al., 1997; Jabri et al., 1997). RNA catalysts can also act on substrates other than nucleic acids. In vitro selected RNA molecules have been shown to interact with amino acids such as arginine, phenylalanine, tryptophan and valine (Majerfeld and Yarus, 1994; Yarus, 1991; Zinnen and Yarus, 1995). Arginine-binding RNA motifs (Tan et al., 1993) and aromatic side chains (Valegard et al., 1994) have been found to be important for protein-RNA interactions. Although aliphatic-RNA interactions have been frequently neglected, their avidity and specificity seem sufficient for a biological role. Also, the guanosine binding site of group I introns has been shown to bind arginine,
suggesting that the proto-ribosome might be related to group I introns (Yarus, 1991).
RNA catalysis is not limited to ribozymes alone. It is also involved in two
important steps of gene expression: mRNA processing and protein synthesis. In mRNA splicing, small nuclear RNAs recognize the reaction sites (Guthrie, 1991; Steitz, 1992) and may participate directly in the chemical steps (Madhani and Guthrie, 1992; McPheeters and Abelson, 1992), while the associated proteins may facilitate proper folding of these RNA molecules. It is also possible that, in addition to participating in mRNA splicing, RNAs can catalyze a variety of posttranscriptional RNA modifications. For instance, a calcium-metalloribozyme has been shown to efficiently catalyze a self-capping reaction with free GDP, yielding the same 5'-capped structure as that formed by protein guanylyltransferase (Huang and Yams, 1997a). This ribozyme was subsequently shown to possess selfdecapping and pyrophosphatase activities (Huang and Yarus, 1997b), adding to the growing repertoire of the catalytic capabilities of RNA.
The catalytic activities of RNA may also include facilitation of protein synthesis. Following extensive digestion of proteins from the prokaryotic ribosome, the large rRNA was shown to support the peptidyl transferase reaction, suggesting that RNA may be the catalytic component while ribosomal proteins may serve a scaffold function (Noller et al., 1992). Recently, in vitro selected RNA molecules with a peptidyl transferase-like motif have been shown to bind a
puromycin analog (a high-affinity ligand of ribosomal peptidyl transferase), in the absence of protein (Welch et al., 1997). Together, these results strengthen the hypothesis that peptidyl transfer originated in an RNA world.
The catalytic role of RNA in translation of mRNAs appears to be versatile. A ribozyme derived from the group I intron of Tetrahymena thermophila was shown to catalyze the hydrolysis of an aminoacyl ester bond (which involves a carbon center), suggesting that the first aminoacyl tRNA synthetase could have been an RNA molecule (Piccirilli et al., 1992). Furthermore, an RNA molecule identified by in vitro selection has been shown to rapidly aminoacylate its 2'(3') terminus when provided with phenylalanyl-adenosine monophosphate (Illangasekare et al., 1995; 1997). Thus, RNA can accelerate the same aminoacyl group transfer catalyzed by protein aminoacyl-tRNA synthetases.
The ongoing discovery of the versatile properties of catalytic RNA has lent
more credence to the theories of a prehistoric "RNA world". This pre-biotic world might have been populated by life forms that stored genetic information in RNA and employed RNA catalysts prior to the advent of ribosomal protein synthesis (Visser, 1984; Benner et al., 1989). It has also been proposed that the functions of these ancient catalytic RNAs may have been modulated by low molecular weight effectors related to antibiotics (Davies, 1990, Davies et al., 1992). Antibiotics have been shown to inhibit translation by the prokaryotic ribosome (Moazed and Noller, 1987; Powers and Noller, 1994; Yamada et al., 1978), either inhibit selfsplicing (von Ahsen et al., 1991; Davies et al., 1992; von Ahsen and Noller, 1993)
or promote oligomerization (Wank and Schroeder, 1996) of group I introns, and inhibit the self-cleavage reaction of the human hepatitis delta virus ribozyme (Rogers et al., 1996). Parallels between the inhibition of group I intron splicing and the protection of bacterial rRNAs by antibiotics also raises the possibility that group I intron splicing and tRNA selection by ribosomes involve similar RNA structural motifs.
Group I Introns
Group I introns are abundant in mitochondrial RNA of fungi and plants (Palmer and Logsdon, 1991). Coding regions for group I introns are also found in the nuclear genomes of other lower eukaryotes (rRNA genes of Tetrahymena), chloroplast DNAs, bacteriophages, and in several tRNA genes of eubacteria (Cech, 1988; Michel and Westhof, 1990; Palmer and Logsdon, 1991; Reinhold-Hurek and Shub, 1992). Some of these group I introns can self-splice. The ability to selfsplice is related to the highly conserved secondary and tertiary structures of these introns (Burke, 1988; Cech, 1988; Michel and Westhof, 1990). Although group I introns have relatively little sequence similarity, all share a series of short conserved elements designated P, Q, R and S. The secondary structure common to group I introns was first proposed by Michel et al. (1982) and Davies et al. (1982) based on comparative sequence analysis. This eventually led to the development of a three dimensional model of the catalytic core by Michel and Westhof (1990). The basic features of this model have been confirmed by mutational analysis, photochemical crosslinking and chemical modification studies employing several
affinity cleavage reagents (Pyle et al., 1992; Wang and Cech, 1992). On the basis of these studies, Cech et al. (1994) proposed a revised two-dimensional secondary structure for group I introns that represents more accurately the domain organization and orientation of helices within the intron, the coaxial stacking of certain helices, and the proximity of key nucleotides in three-dimensional space. Based on these revisions, the secondary structure of the fifth intron of the COB gene of Saccharomyces cerevisiae (used in the current study) is shown in Figure 11. The folded structure of group I introns consists of two co-axially stacked helices, P5-P4-P6 and P7-P3-P8, that form a cleft to enclose a third helical domain, P1P2, which contains the 5' splice site. The 5' and 3' splice sites are stabilized by P 1 and P 10 interactions respectively. P 1 is formed by base pairing between the 5' exon and the internal guide sequence (IGS), whereas P 10 is formed by base pairing between the 3' exon and the IGS (not shown in Figure 1-1). P9, a 2 base pair helix near P7 also contributes to the formation of 3' splice site.
The rRNA intron of Tetrahynena thermophila (considered to be the prototype group I intron) has been demonstrated to undergo autocatalytic splicing by a two step trans-esterification mechanism (Cech, 1990). The same mechanism has been documented for a number of other group I introns (Garriga and Lambowitz, 1984; Garriga and Lambowitz, 1986). The first step comprises a nucleophilic attack by guanosine at the 5' splice site, resulting in a free 3' OH group on the upstream exon and guanosine addition to the intron. In the second step, the free 3' OH group on the exon attacks the phosphodiester bond at the 3' splice site, leading to
Figure 1-1. Proposed secondary structure of yeast apocytochrome b intron 5 RNA. The coaxially stacked helices, P4-P6 and P3-P7, along with the joining regions J3/4, J4/6, J6/7, and J8/7, constitute the catalytic core while P1-P2 forms the substrate domain. The peripheral element P7. I-P7.1 a is a signature element of subgroup IA introns. Lower and upper case letters represent the exons and the intron respectively. Short horizontal and vertical lines represent hydrogen bonds. Dotted lines represent long-range interactions. Sequences for LI and L8 are not shown.
a u a c u u a u u a-31
Li A a 710
200 nt u a A AA U
A g U U AA
5- U A-U A-U
a A A c U-A U-A
u A A c pg1jCC U-A p9
AUAAA c A A 220 a U-A A-U
A U c A G u U-A U-A
AU-A A u A U g A-U U-A
A-U u U A u +1 U-A GACG-C
A-U uU A U A-C G 738 G A
P5 U-A a 1 UA U U A
A-U g 1 u- U U G:
CGu a-U A A U-A
AUA a 1u-A A A C 444
GUA-AUC a ca c c agcauc-C A C C-G
AAC-G U UAUAUA U-A P7
U-A 305 A-U G
U-A -AU-AGAAAG AG uA
P4 P2 U-A A ii 1 1 1 1 I-A G L7.1la
C-C P2 CCCUUUC UG
A-U A C A U A
A-UG A A
350 C-GC A .............................. ............... ............. U u 7l
AUA-U A C P. 71
C-U U A
P6 C-C AA-...... .........;.................. ;......................
AC-GCL6 A A U-A
U-A U-A G-C P3
A-U L2 EA
P6a U-A A-UCCAA-U C-CAU-A U-A .....U-A
37UAC- P2a U-A U-A
L6a C A U-C 660 A-U
UCU-A 250 C-C A-U
A-U C-C A-U P
U-A C-U U-A P
P6b A-U C-G A-U
U-A L u UUU-A 500
U U-A A-U A-U
AUAA U AAU
excision of the intron and ligation of the exons. An exogenous guanosine nucleotide is used as the nucleophile in the first step while a universally conserved 3'-terminal guanosine residue of the intron is employed in the second step (Cech, 1990). These steps are chemically the reverse of each other, with the bound exogenous guanosine nucleophile in the first step equivalent to the 3'-terminal guanosine leaving group in the second step. This led to the proposal that a single guanosine-binding site was used in both steps (Inoue et al., 1986). Subsequent studies indicated that the rate constant of the chemical step is the same with exogenous guanosine bound to L-21Scal ribozyme (a model system for first step of splicing) and with the intramolecular guanosine residue of the L-21 G414 ribozyme (a model system for second step of splicing) (Mei and Herschlag, 1996). These results support the previously proposed single guanosine-binding site model, and further suggest that the orientation of the bound guanosine and the overall active site structure is the same in both steps of the splicing reaction.
Oligonucleotide substrate binding to the Tetrahymena ribozyme was found to be stronger than predicted for a simple duplex interaction with the IGS, suggesting that tertiary interactions in addition to base pairing stabilized the bound substrate (Herschlag and Cech, 1990; Pyle et al., 1990). These tertiary interactions were shown to involve specific 2'-OH groups on the substrate and IGS, as well as the GU wobble pair at the 5' cleavage site (Pyle and Cech, 1991; Pyle et al., 1994; Strobel and Cech, 1993, Strobel and Cech, 1995). These and other kinetic studies led to a 2-step model for substrate binding, wherein the substrate forms a duplex
(P1) with the IGS to give an 'open complex', followed by docking of the P1 duplex into tertiary interactions to give a closed complex (Herschlag, 1992). Thus, P 1 docking represents a tertiary-folding event in which a single duplex adopts its tertiary structure in the context of an otherwise fully folded ribozyme. P 1 docking was characterized further by isolating the open complex as a thermodynamically stable species using a site-specific modification and high Na+ ion concentrations (Narlikar and Herschlag, 1996). These authors proposed that P1 docking is entropically driven, and is possibly accompanied by a release of bound water molecules.
It is interesting to note that group I introns do not contain specific functional groups that are typically employed in the catalysis of protein-enzymes. Instead, they depend on divalent cations for chemistry and certain other functions such as structural stabilization of folded RNA and substrate binding. For instance, the Tetrahymena group I intron requires Mg2+ or Mn2+ ions for catalysis (Grosshans and Cech, 1989), while cations like Ca2+ can only promote RNA folding and substrate binding (Pyle et al., 1990). A two metal-ion mechanism has been proposed for group I introns and other catalytic RNAs (Steitz and Steitz, 1993). In this mechanism, one metal ion activates the 3'-OH of the guanosine factor which initiates the first step of group I intron splicing. The second one coordinates and stabilizes the oxyanion leaving group, that is, the 3'-OH of uridine created at the end of 5' exon which initiates the second step of group I intron splicing. These metal ions act as Lewis acids and stabilize the expected pentacovalent transition
states. In case of group I introns, the mirror symmetry Of two Mg2 ions in the catalytic center reflects the identical chemistry of the two transesterification reactions that effect splicing. The role of RNA in catalysis is to position the two metal ions and properly orient the substrates. Evidence for the involvement of two Mg 2+ions in the chemical step of group I intron splicing was provided by McConnell et al. (1997). Testudy alosoe htasingle Mg 2+ion increases the rate of RNA substrate binding while one or more Mg 2+ions reduce the rate of dissociation of substrate. Evidence for stabilization of the leaving group by a second Mg2 ion was subsequently provided by Weinstein et al. (1997). Studies on the crystal structure of the P4-P6 domain of Tetrahymena documented the first detailed view of metal-binding motifs in a structurally complex RNA (Cate et al., 1996). Three unique metal binding sites have been found in the major groove, two of which are occupied by fully hydrated magnesium ions in the native RNA (Cate and Doudna, 1996). It is interesting to note that the tandem GU wobble base pairs, which comprise two of these three metal binding sites, are also abundant and conserved in the ribosomal RNAs. These sites, upon metal binding, might facilitate higher order folding of ribosomal RNAs or their association with ribosomal proteins.
Group I introns have been classified into four major subgroups, IA through ID, based on distinctive structural and sequence features (Michel and Westhof, 1990). For example, group IA introns contain two extra base pairings, P7. 1 /P7. 1la or P7. 1/P7.2, between P3 and P7 while several other group IB and IC introns
including the Tetrahymena rRNA intron possess an extended RNA structure, P5abc, that is essential for catalysis (Joyce et al., 1989). Recent crystallographic studies revealed that P5abc stabilizes P4-P6, the major domain of the catalytic core, via two key interactions (Cate et al., 1996). The first one includes an adenosine-rich bulge which docks in the minor groove of the P4 helix while the second interaction takes place between a GAAA tetraloop and the minor groove of its conserved 11-nucleotide receptor. In addition to base-specific hydrogen bonding and base stacking, pairs of interdigitated riboses (ribose zippers) further stabilize these long-range interactions in the P4-P6 domain. Other group I introns lacking P5abc possess additional RNA structures such as a long, peripheral extension of the P9 stem, denoted P9.1 (Wallweber et al., 1997) or protein factors (Mohr et al., 1994) that bind and stabilize the intron active structure. For example, the mitochondrial large ribosomal intron of N. crassa that lacks this extended domain absolutely requires a protein factor Cyt- 18 for its activity, both in vitro and in vivo. The protein binds at the junction of P4-P6 stacked helices and facilitates correct geometry in this region (Saldanha et al., 1996). Cyt-18 could also replace the P5abc domain of Tetrahymena (Mohr et al., 1994). Thus an RNA-binding protein can provide substantial binding energy to stabilize the catalytic structure of the intron, obviating the requirement for an additional but important RNA element.
Protein Facilitated Splicing
Several mitochondrial transcripts in yeast employ protein factors to facilitate splicing of group I and group II introns, although some of them can self splice in
vitro. Protein-facilitated splicing has also been documented in other fungi such as Neurospora (Akins and Lambowitz, 1987; Saldanha et al., 1993) and Aspergillus (Ho et al., 1997). Some of these protein factors, termed maturases, are encoded within the intervening sequences (Carignani et al., 1983; Lamb et al., 1983; Lazowska et al., 1989). The reading frames encoding these maturases are in frame with the upstream exons. A proteolytic cleavage downstream of the 5' splice site generates the active form of maturase, presumably enabling a feedback mechanism of regulation. All yeast maturases (encoded by group I introns such as cob-I2 and 13, and group II introns like cox]-I and -I2) primarily function in splicing the intron that encodes them. However, the cob-I4 maturase enables splicing of both cob-I4 and another closely related group I intron, cox1-I4 (Burke, 1988; Lambowitz and Perlman, 1990).
The maturases encoded by group I introns are structurally related to sitespecific endonucleases that confer mobility (Bell-Pedersen et al., 1990; Perlman and Butow, 1989). It is well established that group I introns can transpose sitespecifically to intron-less alleles of the same gene after cleavage of the target DNA by an intron-encoded DNA endonuclease (Belfort and Perlman, 1995; Byrk and Mueller, 1996). This conservative process, known as intron homing (Dujon, 1989), is highly specific because of the large recognition sites (15-35 bp) of homing endonucleases (Byrk and Mueller, 1996). Group I introns are thought to have become mobile following the acquisition of open reading frames (ORFs) that encode specific DNA endonucleases (Dujon, 1989; Lambowitz and Bellfort,
1993). Evidence for this came from the demonstration of autonomous mobility of an ORF, independent of the entire intronic sequence, in the mitochondria of Podospora anserina (Sellem and Belcour, 1997). The mitochondrial nadl-i4 intron of Podospora contains one (monorfic) or two (biorfic) ORFs, according to the origin of the strain (Cummings et al., 1988; Sellem et al., 1996). The nadl-i4orfl, recently acquired by the Podospora nadl-i4 intron (Sellem et al., 1996), appears to remain as a mobile entity, as it could be efficiently transferred from a biorfic intron to its monoorfic counterpart, independent of the core intron sequence (Sellem and Belcour, 1997).
Reverse splicing coupled with reverse transcription and recombination may serve as an alternative mechanism for intron mobility (Cech, 1985; Sharp, 1985; Woodson and Cech, 1989). A related mechanism has been documented in the homing of group II introns of yeast mitochondria (e.g., a12 intron of coxl) (Zimmerly et al., 1995; Yang et al., 1996) and Lactobacillus lactis (e.g., LtrB intron) (Matsuura et al., 1997). The insertion of these introns has been shown to occur by reverse transcription of unspliced precursor RNA at a break in doublestrand DNA caused by an endonuclease activity. This DNA endonuclease activity is associated with RNP particles containing the excised intron RNA that cleaves the sense strand of the recipient DNA by reverse splicing and the intron-encoded reverse transcriptase protein that cleaves the anti-sense strand.
Integration of an intron into foreign RNA (instead of DNA) by reverse splicing, followed by reverse transcription and recombination, could also lead to its
transposition. Reverse splicing into RNA has been demonstrated in vitro for group I introns such as Tetrahymena rRNA intron (Woodson and Cech, 1989) and the 23S rRNA intron from Chlamydomonas reinhardtii chloroplast (Thompson and Herrin, 1994). Recently, RNA-dependent integration of the Tetrahymena group I intron into the 23S rRNA has been demonstrated in E. coli (Roman and Woodson, 1998). The process of reverse splicing into RNA, unlike homing of group I and group II introns, does not require intron-encoded proteins. However, stable transposition into the genome would presumably require reverse transcriptase activity in the host (Belfort and Perlman, 1995). This activity could be provided by "indigenous" group II introns (Kennell et al., 1993) or retroelements present in many cell types (Eickbush, 1994). Importantly, reverse splicing appears to be significantly less sequence-specific than homing endonucleases and could therefore expand the repertoire of intron-containing sites (Cech, 1985; Roman and Woodson, 1998).
In addition to maturases, several nuclear-encoded proteins essential for splicing of mitochondrial introns have been identified in yeast and Neurospora by screening cytochrome-deficient strains and by isolating nuclear suppressors of splicing mutants (Burke, 1988; Lambowitz and Perlman, 1990). In Neurospora, the products of three nuclear genes, cyt-18, cyt-19 and cyt-4 have been implicated in splicing the mitochondrial large rRNA intron and several other mitochondrial group I introns. In contrast, most of the yeast proteins facilitate splicing of a single intron. For instance, the product of MRS 1 in yeast appears to be specific to the
group I intron, cob-bl3 (Kreike et al., 1987, Kreike et al., 1986), although an intron-encoded maturase is also required for excision of this intron (Holl et al., 1985). Proteins facilitating the splicing of group I introns may exhibit additional biological functions as documented in the case of certain aminoacyl-tRNA synthetases. The yeast NAM2 gene, for example, encodes mitochondrial leucyltRNA synthetase which also facilitates the splicing of group I introns, cob-I4 and cox-I4 (Labouesse et al., 1987; Herbert et al., 1988). Similarly, the Cyt-18 protein of Neurospora which is responsible for the splicing of several group I introns also happens to be the mitochondrial tyrosyl tRNA synthetase (Akins and Lambowitz, 1987). These synthetases and other pre-existing RNA-binding proteins may have evolved to recognize sequences or structures in group I introns that resemble their normal cellular targets (Lambowitz and Perlman, 1990; Caprara et al., 1996).
Nuclear-encoded proteins also appear to be important for group II intron
splicing. Two genetically identified proteins that are likely to function directly in group II intron splicing are MRS2 and MSS 116 (Wiesenberger et al., 1992; Seraphin et al., 1989). MRS2 functions in splicing of all four yeast group II introns (coxl-I1, -12, -I5y and cob-Il 1), and is relatively specific for these introns, while MSS 116 is involved in splicing group II introns (cox]-I1 and cob-I1) and also some group I introns. However, both MRS2 and MSS 116 appear to have some additional function besides splicing, as gene disruptions result in a respiratory-deficient phenotype in yeast strains whose mtDNA contains no introns (Wiesenberger et al., 1992; Seraphin et al., 1989)
Two-Component System of Cbp2 and Intron 5 RNA
The yeast cytochrome b gene contains five introns, of which the terminal intron is a group IA intron (bI5). In some yeast strains, the gene has only two introns, with the terminal intron designated b12. The processing of this group I intron in vivo was demonstrated to be dependent on a protein factor designated Cbp2 (McGraw and Tzagoloff, 1983). The nuclear gene encoding Cbp2 was identified by complementation of cytochrome b mutants (defective in the excision of the terminal intron) with a yeast genomic library. This analysis identified an 1890 nucleotide-long ORF encoding a basic protein of 74 kDa. Deletion analysis revealed that the entire ORF was essential for complementation of the cbp2 mutants. Later, a mitochondrial revertant was shown to contain a precise deletion of the terminal intron of cytochrome b gene, demonstrating that neither Cbp2 nor the intron itself is required for growth on non-fermentable carbon sources (Hill et al., 1984). In addition to these findings, Cbp2 has been shown to be important in the splicing of the 0o intron of large ribosomal RNA (Shaw and Lewin, 1997).
The terminal intron (intron 5) of cytochrome b can self-splice in vitro at high concentrations of Mg2+ (Gampel and Tzagoloff, 1987; Partono and Lewin, 1988), whereas Cbp2 is essential to enable splicing at physiological concentrations of Mg2+ (Gampel et al., 1989). Although this group IA intron possesses the conserved secondary and tertiary structures found in all group I introns, it varies in important ways from the prototype, the Tetrahymena rRNA intron. The fifth intron of cytochrome b is about 738 nucleotides long, making structural probing
harder compared to the rRNA intron of Tetrahymena (-400 nucleotides long). The internal guide sequence (IGS) that establishes the substrate specificity starts 220 nucleotides downstream from the 5' splice junction, rather than the usual 14-20 nucleotides described for other group I introns. The intron is AU-rich, requiring higher levels of Mg2+ for stabilization of the active structure unlike the GC-rich Tetrahymena group I intron. It also possesses additional RNA structures like the P7.1 stem loop that are not found in Tetrahymena. Hence, intron 5 RNA, with its structural differences from the Tetrahymena rRNA intron, offers an opportunity to gain further insights into the mechanism of splicing of group IA introns.
The fifth intron of COB pre-mRNA is also devoid of the peripheral RNA
element, P5abc, that is important for the catalysis of Tetrahymena rRNA intron. It is therefore conceivable that Cbp2 compensates for this RNA structure and stabilizes its RNA partner by contributing substantial binding energy in a manner similar to the Cyt- 18 protein of Neurospora. UV-crosslinking, chemical and enzymatic modification studies indicate that Cbp2 contacts intron 5 RNA at multiple sites in the catalytic core (P4) and peripheral RNA elements such as exon 5, IGS, L2, L6 and stimulates the formation of the catalytically active structure (Shaw and Lewin, 1995; Weeks and Cech, 1995). Based on these and kinetic studies, Weeks and Cech (1996) proposed that Cbp2 serves as a tertiary structure capture protein. However, Cbp2 also induces the formation of RNA secondary structure, in addition to the stabilization of tertiary structure (Shaw and Lewin, 1995; Shaw et al., 1996). In addition, chemical modification studies (Shaw and
Lewin, manuscript in preparation) show that Cbp2 binds to intron 5 RNA even in the absence of Mg 2+and nucleates the formation of the catalytic core by stabilizing the P4/P6 domain. Thus, Cbp2 appears to be involved in a dynamic process of stabilizing RNA structure both at the secondary and tertiary structure levels, stimulating the formation of the catalytically active RNA structure.
Weeks and Cech (1995a; 1995b) provided a kinetic framework for both Cbp2mediated and self-splicing reactions of intron 5 RNA. At low Mg 2+ levels (5 mM), the self-splicing reaction is estimated to be 3 orders of magnitude slower than the protein-facilitated reaction. At near saturating concentrations of Mg 2+(40 mM), the protein-independent reaction is still 8-fold slower, indicating that high levels of the cation cannot completely compensate for Cbp2 function. The self-splicing reaction is always slower than the protein- facilitated reaction, since it has to proceed through two additional transitions compared to the latter. The first step involves a transition from secondary structure to an intermediate state that is efficiently promoted by Mg2+ However, self-splicing must still overcome a second barrier which is the transition from the intermediate to an active enzyme state that finally gives rise to products. The kinetics of Cbp2-mediated splicing, on the other hand, include two significant steps, namely, guanosine binding to the Cbp2-active intron 5 RNA complex followed by efficient conversion of this ternary complex to products. Studies on phosphorothioate substitution at the 5' splice site and pH profiles indicate that at physiological pH the self-splicing reaction is limited by chemistry while the Cbp2-facilitated reaction is limited by a
conformational step (Weeks and Cech, 1995a). These studies indicate that Cbp2 binding compensates for at least two structural defects while increasing the rate of chemistry.
Main Objective of Dissertation
The availability of a two-component in vitro system to study autocatalytic and protein- facilitated splicing offers the advantage of studying RNA catalysis in isolation or in combination with the RNA-binding protein simply by varying the Mg concentration. Insights obtained from the analyses of this one protein-one RNA system will aid in understanding more complex systems like the spliceosomes involved in nuclear pre-mRNA splicing or ribosomes involved in protein synthesis, all of which employ multiple protein and RNA components.
Studies so far have focused on mapping the Cbp2 contact sites on intron 5 RNA and the kinetics of splicing in the presence and absence of the protein. Little is known about the structure of Cbp2 protein or its interaction with intron 5 RNA from the protein point of view. In order to understand the role of Cbp2 in stimulation of splicing, it is important to determine the functional groups on the protein that intimately contact RNA and facilitate catalysis. Once the contact sites are identified, the actual mechanism of interaction between Cbp2 and intron 5 RNA can be investigated further. Therefore, one of the main aims of the current project was to identify potential intron 5 RNA binding regions of Cbp2 using the technique of UV-crosslinking and label transfer. Following the identification of major contact sites, site-directed mutagenesis was employed to confirm the
importance of various amino acid residues in these sites for interaction with intron
5 RNA and enable facilitation of splicing.
UTV-crosslinking identified two major RNA contact sites in the termini of Cbp2 with the N-terminal site comprising the first 37 amino acid residues. The deletion of a potential RNA binding motif (aal 7-SSSRYRYKFNM-aa28) in the Nterminal contact site abolished splicing activity, showing that this region was likely to be critical for Cbp2 function. Single and cluster mutagenesis of various residues in this region yielded a variety of mutants with no, partial, or complete loss of Cbp2 function. The characterization of these mutants and the significance of various amino acid residues in question are discussed in Chapter 4.
An offshoot of the current study was the identification of an RNA chaperone function for Cbp2. The studies reported in Chapter 4 show that Cbp2 has a nonspecific or generalized RNA binding capability besides its specific RNA binding component. Drawing parallels with studies on other RNA chaperones (Chapter 5), this study adds a new perspective on how the non-specific RNA binding activity of Cbp2 might play a critical role in the facilitation of intron 5 RNA splicing in vivo.
MATERIALS AND METHODS
Over-Expression and Purification of CBP2
In vitro studies of CBP2-bI5 RNA interactions were done with CBP2 protein purified to near homogeneity after over-expression in E. coli. Two versions of the protein, 6x histidine-tagged and native, were employed for different studies. Expression Clones
His-tagged Cbp2. Plasmid pET15b-CBP2 was constructed by cloning the
NdeI-ClaI fragment carrying the CBP2 cDNA from pET3a-CBP2, downstream of the 6x histidine tag in the T7 expression vector, pET15b (Novagen). The histidine tag adds an additional 20 amino acid residues at the N-terminus of Cbp2 protein. The plasmid was transformed into JM IO09(DE3) strain of E. coli for overexpression. This strain carries the T7 RNA polymerase gene driven by lac-uv5 promoter on a lambda lysogen and enables the induction of Cbp2 in the presence of IPTG.
Native Cbp2. This form of Cbp2 protein was over-expressed from pET3aCBP2 plasmid in BL2 I1(DE3), another E.coli strain carrying the T7 RNA polymerase gene. This plasmid was constructed by introducing an NdeI site at the start codon of CBP2 gene by PCR-mutagenesis and cloning the NdeI-SnaBI 21
fragment between the NdeI-BamHI sites of pET3a expression vector (Studier et al., 1990). The Cbp2 protein expressed from this construct is 20 amino acid residues shorter than the his-tagged version. Induction of Cbp2
An overnight culture of bacteria carrying the Cbp2 expression plasmid was
used to inoculate a large volume of LB/ampicillin medium at 1:100 dilution. The cultures were grown at 370C until they reached an A550 of 0.35 and the expression of Cbp2 was induced with 0.4-1 mM IPTG for 1-3 hours. Cells were pelleted after addition of 17 ig/ml PMSF, washed with 20 mM Tris, pH 7.5, 50 mM NaCI, snap frozen in a dry ice/ethanol bath and stored at -700 C until purification. Purification of Cbp2
His-tagged Cbp2. The protein was purified on Ni-NTA Superflow (Qiagen) column, adapting the protocol of Weeks and Cech (1995). This purification system is based on the high affinity of histidine residues for nickel ions immobilized on nitrilotriacetate resin. The contaminant proteins can be efficiently removed at low levels of imidazole (a competitor), while the his-tagged protein can be specifically eluted at slightly higher concentrations of imidazole.
The bacterial pellet was resuspended in 10 ml of column buffer (50 mM
HEPES, pH 7.6, 700 mM NaCl, 1mM imidazole, 17.5 ug/ml PMSF) and lysed by two passages through a French pressure cell at 18000 p.s.i. The lysate was cleared by centrifugation at 35000 rpm for 30 minutes in a Beckman Ti 42.1 rotor. The
supernatant was loaded on a 2 ml Ni-NTA Superflow column pre-equilibrated with 10 volumes of column buffer. The column was washed with 10 volumes of column buffer (lmM imidazole) followed by 7.5 volumes of wash buffer (20 mM imidazole). Cbp2 protein was then eluted with 7.5 volumes each of column buffers containing 80 mM and 200 mM imidazole. The fractions containing Cbp2 (detected by SDS-polyacrylamide gel electrophoresis) were pooled and dialyzed twice against 1 liter each of 10 mM Tris, pH 7.5, ImM EDTA, 20% glycerol and once with a liter of 10 mM Tris, pH 7.5, 1 mM EDTA, 50% glycerol and stored at 700 C after rapid freezing in a dry ice/ethanol bath.
Native Cbp2. This version of Cbp2 protein was isolated according to the 4step purification protocol described by Shaw and Lewin (1995).
In Vitro Transcription
pSPI5 plasmid DNA purified by CsCl gradients (Maniatis et al., 1989) was
linearized with SmaI and used for in vitro transcription with T7 RNA polymerase (Partono and Lewin, 1988). The transcripts contain the entire intron 5 RNA sequence and the flanking exon sequences. The transcripts were internally labeled using Ca-32P UTP and/or a-32P ATP (ICN).
UV-Crosslinking and Generation of Peptides UV-Crosslinking
Cbp2-RNA complexes were generated according to the UV-crosslinking
technique of Zamore and Green 1989. 32-labeled intron 5 RNA transcripts were
technique of Zamore and Green, 1989). P-labeled intron 5 RNA transcripts were
incubated at room temperature or 370 C for 30 minutes with a molar excess of histagged Cbp2 (7 fold over RNA) or native Cbp2 (21 fold) in a low salt buffer (50 mM Tris, pH 7.5, 10 mM MgCl2, 50 mM NH4CI) containing excess tRNA (nonspecific competitor). Each sample was split into several aliquots, 10 p1 each, in a 96-well microtiter plate (Falcon) placed on ice (in a petridish) and exposed to 600 mJ of UV radiation in a UV-Stratalinker (Stratagene). The aliquots of each sample were pooled into a 1.5 ml Eppendorf tube and treated with 0.32 ug/ml of RNAse A and 100 units of RNAse TI (Boehringer Mannheim) at 370 C for 2 hours to remove uncrosslinked RNA. The samples were resolved by electrophoresis on a 10% SDS-polyacrylamide gel (Laemmli, 1970) and the band corresponding to Cbp2 excised after Coommassie blue staining. The Cbp2 thus purified includes both the crosslinked and un-crosslinked forms of the protein. Generation of Peptides
The purified gel fragments were incubated with chemical cleavage reagents such as hydroxylamine (NH20H) and 2-nitro-5-thiocyanobenzoate (NTCB) and the resulting peptides were resolved on high percentage tris-tricine gels.
Cleavage. The gel pieces were washed four times with distilled water over a period of 20 minutes, placed into appropriate cleavage buffer and thoroughly macerated with a Kontes Eppendorf pestle. The slurry so obtained was completely covered with the cleavage buffer and incubated overnight at appropriate temperature.
Chemical cleavage of proteins with hydroxylamine generates relatively large
peptides due to the infrequency of Asn-Gly bonds. The asparaginyl side chain has a tendency to form a cyclic imide that is susceptible to nucleophilic attack by hydroxylamine (Bornstein, 1977). The cyclization is more favored in the context of a smaller amino acid like glycine resulting in increased susceptibility of AsnGly bonds. Hydroxylamine (NH2OH) cleavage of Cbp2 was performed by overnight incubation of Cbp2 containing gel pieces in 2.4 M guanidine-HC1, 2M hydroxylamine buffer, pH 9, at room temperature, as described above. LiOH was used to neutralize the guanidine-HCI and hydroxylamine-HCl due to increased solubility of LiCl compared to NaCI. The process of gel purification contributes to partial denaturation of the protein while the presence of guanidine-HC1, a strong solvent, enhances the exposure of the Asn-Gly bonds to the nucleophile. Chain cleavage occurs in the presence of alkaline hydroxylamine liberating a new aminoterminal amino acid.
Cleavage with 2-nitro-5-thiocyanobenzoate (NTCB) is a two-step process.
First, the thiol groups on cysteine residues of denatured proteins are modified to SCN groups by NTCB (Jacobson et al., 1973; Degani and Patchornik, 1974), followed by cleavage at the amino group of the modified cysteine by exposure to alkaline pH conditions. Gel purified Cbp2 protein was incubated in 2.4 M guanidine-HC1, 5 mM DTT, 1 mM EDTA, 0.2 M tris acetate, pH 8, buffer at 370 C for 2 hours in order to denature the protein and reduce the disulfide bonds to SH groups. A 10 fold excess of NTCB (50 mM) over the total thiol was added to the
gel slurry, and the incubation was continued for half hour at the same temperature to effect modification of the SH groups to SCN groups. The slurry was filtered through a 0.22 LM low protein-binding, cellulose acetate spin column (CorningCostar), and washed once with distilled water. The slurry was later transferred to a
1.5 ml Eppendorf tube and incubated overnight in 2.4 M guanidine-HCl, pH 9, cleavage buffer at 370 C.
Extraction of peptides. After cleavage, the slurry was filtered through a
Costar column, washed once with distilled water and incubated overnight at 370C in the extraction buffer (0.1% SDS, 50 mM Tris pH 8.8, 0.1 mM EDTA and 0.2 M ammonium bicarbonate). On the third day, the gel slurry was heated at 850 C for 5 minutes and rapidly filtered through a Costar column to recover soluble peptides. The slurry was further incubated with 0.5% SDS, 10 mM Tris, pH 8, for 20 minutes at room temperature and filtered to extract the residual peptides in the gel.
Acetone precipitation and electrophoresis. The filtrates containing the
peptides were pooled, dried in a Speed-Vac (Savant), resuspended in water and precipitated overnight at -20o C with 9 volumes of acidified acetone. Peptides were pelleted at 12000 rpm in a microcentrifuge (Eppendorf) for 20 min and resuspended in 15 1l of SDS gel-loading buffer. The samples were dried in a Speed-Vac to remove the residual acetone, brought to a final volume of 40 g1 with water, resolved on 15% (for hydroxylamine ) or 16.5% (for NTCB) tris-tricine gels (Schagger and von Jagow, 1987), along with 14C-labeled low molecular weight peptide markers (Amersham Corporation), and autoradiographed.
The N-terminal RNA contact site on Cbp2 (identified by the UV-crosslinking strategy) was subjected to site-directed mutagenesis to identify key residues for Cbp2 function. Mutations were designed to either delete the region of interest (aal7 SSSRYRYKF aa25) or make point mutations that do not severely perturb the conformation of the protein. The Xbal-BamH1 fragment encoding the first 97 codons of CBP2 from pET15b-CBP2 was sub-cloned into the M13mpl 19 vector and used as the template for oligonucleotide-directed mutagenesis. Mutagenesis Scheme
Single stranded DNA isolated from the phage clone mentioned above was used as a template for mutagenesis. The oligonucleotides used for mutagenesis are shown in Table 2-1. The double primer method of Zoller and Smith (1984) was employed to introduce mutations into the CBP2 segment cloned into M13mpl9 vector. Briefly, 10 pmoles each of the kinased mutagenic oligonucleotide and the universal M13 primer were annealed to 0.5 pmoles of single-stranded DNA template by incubation at 650 C for 30 minutes, 370 C for 20 minutes and room temperature for 5 minutes. The annealed complexes were extended and ligated overnight at 150 C using 4 units of Klenow DNA polymerase (Promega) and 2.5 units of T4 DNA ligase (BRL) to form a gapped heteroduplex. The reaction was diluted 200-fold, transformed into competent TG 1 cells (Amersham) and overlaid
Table 2-1. Oligonucleotides used for mutagenesis of CBP2
Name Length Sequence (5' to 3') Description
AL 260 19 TAGCAAGCCCAATAGGAAC Universal primer (position
1822-1804 in antisense
strand of M13mp 19)
AL 261 17 TAAACGCTTGCTTACAG Sequencing primer to
verify mutations (position
790-774 in antisense
strand of CBP2)
AL 262 40 TCTCCATGTTGAACAAATACA Changes RYRYKF to
AGTAAAGAGAGGAACTGCC LYLYLF at aa 20,22,24
AL 263 33 CACCTGATGTGTGATATTGCCC Deletion of aa 17-28 of
AL 264 42 GATATTCTCCATGTTCAACTTC Changes RYRYKF to
AACCTTAAACGAGAGGAACT RLRLKL at aa 21,23,25
AL 290 30 GTTGAACTTATACCTTAAACG Changes Y to L at aa 21
AL 291 30 CTCCATGTTGAACTTTAACCTG Changes Y to L at aa 23
AL 292* 36 GAACTTATACCTGTAACGACC Deletion of SSS at aa
AL 293 31 GATATTCTCCATGTTCAACTTA Changes F to L at aa 25
* AL292 also changes the codon usage for glycine at position 16 from GGC to GGU
with molten agar to allow formation of plaques. Mutants were identified by plaque lift hybridization with y-32P-labeled mutagenic oligonucleotide as the probe. The resulting mutants were plaque purified once and the single-stranded DNA sequenced. Double-stranded DNA was prepared by mini-prep protocols (Maniatis et al., 1989) from the mutant TG1 clones. The CBP2 segment carrying the mutation of interest was then recloned into pETI 5b-CBP2 expression vector, and sequenced using Sequenase 2.0 kits (Amershami).
In Vitro Splicing Assay
ihe act:i ity of various mutant Cbp2 proteins was determined by an in vitro spl.iciig assay (Partono and Lewin, 1988). 32P-labeled intron 5 RNA transcripts were incubated with wild-type or mutant Cbp2 proteins in 5 mM MgCl2, 50 mM Nt1CI, 50 mM tris-HC1, pH 7.5, buffer, in the presence of 5 mM DTT and 2 units of RNAsin RNase inhibitor (Promega), at 370 C for 10 minutes. Splicing was initiated with 0.2 mM GTP (Pharmacia) and the reactions allowed to continue for varying lengths of time. Reactions were terminated by the addition of equal volumes of 90% formamide, 25 mM EDTA or ethanol precipitated and resuspended in the formamide buffer. Reaction products were resolved on 4% polyacrylamide-8M urea gels and autoradiographed. Splicing Competition Assays
These assays were done essentially as described above but in the presence of a constant amount of the wild-type protein and increasing concentrations of mutant
Cbp2 proteins. The deletion mutant lacking amino acids 17-28 and the triple aromatic mutant with Y21, Y23 and F25 residues converted to leucine were employed to compete with the wild-type protein in splicing assays. As a control, the concentration of wild-type Cbp2 was increased to the same level of total Cbp2 protein (wild-type + mutant) used in the above reactions but in the absence of mutant proteins. Spliced products were resolved on denaturing gels and quantitated using Phosphorlmager (Molecular Dynamics).
Partial Proteolysis of Cbp2
The conformation of mutant Cbp2 proteins was determined by comparing the partial proteolytic profiles of wild-type and mutant Cbp2. Partial proteolysis was done by incubating 0.5-1 ug of the wild-type (native or heat-denatured) or the mutant Cbp2 protein with trypsin, at protease:Cbp2 ratios of 1:50 and 1:100 (w/w), for 1 hour at room temperature and the peptides resolved on 12% SDS-PAGE gels. The peptide profiles were detected by Western blotting performed according to Towbin et al. (1979), with a Cbp2-specific polyclonal antibody (a generous gift of Dr. Alexander Tzagaloff).
Equilibrium Binding Analysis
The affinity of wild-type and mutant Cbp2 proteins for b15 RNA was
determined by the double-filter binding assay (Wong and Lohman, 1993) with the exception that a charged nylon membrane (Hybond N+ from Amersham) was used in place of DEAE. This method involves filtration of protein-RNA mixtures through a sandwich of two membranes, a nitrocellulose filter on top and a nylon
membrane at the bottom, in a 96-well dot-blot apparatus. The protein-RNA complexes are retained on the nitrocellulose while the free RNA is trapped by the nylon membrane. The fraction [RNA bound] can be calculated as follows:
[RNA]bound = [RNA]total (CNc (Y CNY) / (CNC + CNY)
where, CNC and CNY correspond to the Phosphorlmager counts retained on the nitrocellulose and nylon filters respectively. The parameter refers to the RNA retained nonspecifically on nitrocellulose and is empirically derived from RNA bound in the absence of the protein (G = CNC / CNY at [protein] = 0).
Protein-RNA complexes were generated by incubating a low concentration (16 pM) of 32p-labeled intron 5 RNA with increasing concentrations (0-4000 pM) of wild-type or mutant Cbp2 in 5 mM MgCI2, 5 mM DTT, 50 mM NH4CI, 50 mM tris-HC1, pH 7.5, buffer at 370C for 30 minutes. The reactions exhibit equilibrium binding by 30 minutes. Duplicate reactions were filtered through a pre-soaked BA 85 nitrocellulose membrane (Schleicher and Schuell) overlaid on a pre-wetted Hybond N+ nylon membrane, in a 96-well dot-blot apparatus (Bio-Rad). The filters were washed four times with low salt buffer and the radioactivity retained on both the membranes was quantitated using a Phosphorlmager (Molecular Dynamics). The fraction [RNA bound] was calculated and Kd of the mutants determined by the Cbp2 concentration needed for half maximal RNA binding.
IDENTIFICATION OF INTRON 5 RNA CONTACT SITES ON CBP2 PROTEIN
Induction of crosslinks by ultraviolet light in nucleic acid-protein complexes has been a valuable tool for probing structural aspects of protein-DNA/RNA interactions. Ultraviolet (UV) photolysis provides a useful approach to determine the contact points between nucleic acid and protein, as it produces zero-length crosslinks in contrast to chemical crosslinking agents. The latter interpose spacers of varying lengths at the interface and hence are less appropriate to probe intimate contacts at the interface of protein-nucleic acid complexes. A free radical mechanism has been proposed to explain the process of UV-crosslinking of amino acids to nucleic acid bases (Shetlar, 1980). Photoexcitation of a nucleic acid base followed by abstraction of a hydrogen atom from a favorably positioned amino acid residue generates a purinyl or pyrimidinyl radical which recombines with the corresponding radical on the proximate amino acid residue. Such a zero-length recombination event requires an amino acid to be present in extremely close proximity to an excited base. Studies on the bacteriophage fd gene 5 protein, a single-stranded DNA-binding protein suggest that the amino acid and the base 32
must also be present in a relatively specific topological arrangement to achieve photochemical crosslinking (Williams and Konigsberg, 1991). For instance, the amino acids Tyr-26 and Phe-73 of the bacteriophage fd gene 5 protein could not be crosslinked in a gene 5-ssDNA complex (Paradiso et al., 1979; Paradiso and Konigsberg, 1982), although 'H nuclear magnetic resonance data suggest that these two residues form part of the DNA-binding domain of the protein (King and Coleman, 1988. The extent of photocrosslinking also depends on the intrinsic structure of the nucleic acid or protein. Among the nucleotides, thymine and uridine appear to be the most photoreactive, yielding greatest extent of crosslinking to proteins. On the other hand, in principle, any of the 20 amino acids found in proteins can be crosslinked to nucleic acids by UV-irradiation (Williams and Kongsberg, 1991).
Photochemical crosslinking has been adapted to detect protein bound to specific sites on double-stranded DNA using 32P-labeled, site-specific probes (Safer et al., 1988). This method permits transfer of 32P from specific phosphodiester bonds to amino acid residues at the interface upon photocrosslinking (Williams and Konigsberg, 1991). We have employed a similar method to detect intron 5 RNA binding sites on Cbp2 protein. We synthesized intron 5 RNA transcripts (internally labeled with a-32P UTP), UV-crosslinked it to purified Cbp2 under conditions that favor specific complex formation, and detected the crosslinked Cbp2-RNA complexes on SDS-polyacrylamide gels by autoradiography as the protein became indirectly labeled upon photocrosslinking.
Various biochemical methods have been employed by several groups to identify the crosslinked peptides and amino acid residues at the interface of protein-nucleic acid complexes. The most common approach has been to digest the crosslinked complex with trypsin, rapidly isolate the peptides using anion-exchange HPLC, detect the peptides by their absorbance at 220 nm or 254 nm, and identify the crosslinked fragments by Cerenkov counting of the resultant fractions (Merrill et al., 1984; Merrill et al., 1988 and Shamoo et al., 1988). In case of proteins with known primary structure, the crosslinked amino acid residues have been identified by amino acid analysis following acid hydrolysis. For example, the crosslink site in the bacteriophage fd gene 5 protein was identified as cysteine-33 by this method (Paradiso et al, 1979). However, this may not represent a general approach as it depends on the ability of acid hydrolysis to regenerate the free amino acid from the crosslinked adduct. In most instances, the crosslinked amino acid was identified by amino acid sequencing, based on the following principle. A gas or liquid phase sequencer cannot extract the phenylthiazolinone derivative of the crosslinked amino acid from the polybrene-coated support disk and therefore leaves a hole in the sequence at the crosslinked position. Thus, the site of crosslinking is determined by the absence of an identifiable phenylthiohydantion derivative in the peptide sequence. Using this approach, the contact sites in E. coli SSB (Merrill et al., 1984) and Al hnRNP (Merrill et al., 1988) proteins crosslinked to 32P-labeled
(dT)8 oligonucleotides were identified. In the case ofE. coli SSB, the site of crosslinking was further confirmed by solid-phase sequencing which employed a
sufficiently polar solvent such as trifluoroacetic acid to extract the 32P-labeled phenylthiohydantion derivative of the crosslinked amino acid.
In addition to the above biochemical methods, gel electrophoresis is a simple but powerful analytical technique to resolve complex mixtures of peptides and identify the indirectly labeled, crosslinked peptides by autoradiography. In the studies reported in this chapter, we have analyzed the UV-crosslinked, Cbp2-32P labeled intron 5 RNA complexes by digesting the protein-RNA complexes with non-enzymatic cleavage agents and resolving the resultant peptides by one dimensional tris-tricine gel electrophoresis. We have mapped the major crosslink sites of intron 5 RNA to the N- and C-termini of Cbp2.
Optimization of UV-Crosslinking Conditions
One of the technical hurdles in biochemical characterization of crosslinked complexes is isolation of sufficient amounts of the protein-RNA complexes, relatively free from other species. As the yield of the product depends on the extent of crosslinking, reaction conditions must be optimized to maximize the crosslinking efficiency. To that end, the dosage of UV-radiation employed for crosslinking Cbp2 protein to intron 5 RNA was titrated, holding other conditions constant. 32P-labeled intron 5 transcripts were incubated with native Cbp2 under low salt conditions (5 mM MgCl2, 50 mM NH4C1) without GTP. Cbp2 binds to intron 5 RNA under these conditions and induces formation of the catalytic RNA conformation (Shaw and Lewin, 1995). The Cbp2-RNA complexes generated
were UV-crosslinked in the presence of excess tRNA (added as a non specific competitor) by the technique of Zamore and Green (1989), as described in Materials and Methods. The samples were irradiated at an increasing UV-dosage ranging from 100 to 950 mJ. As a control, intron 5 RNA was irradiated with a noncognate protein, BSA, at the highest UV-dosage (950 mJ) employed in the experiment. All samples were extensively treated with RNAse A and RNAse TI to remove uncrosslinked RNA, and the protein-RNA complexes were resolved on SDS-polyacrylamide gels. The gel was autoradiographed (Figure 3-1) and also quantitated using Phosphorlmager. No crosslinked complex was observed in the presence of BSA (lane 1) even at a high dosage of UV-radiation, demonstrating the specificity of Cbp2-intron 5 RNA interaction. The extent of crosslinking of intron 5 RNA to Cbp2 increased by about 2-fold at a UV-dosage of 600 mJ (lane 7) compared to that at 100mJ (lane 2) and almost remained the same at higher doses (lanes 8 and 9). Thus a UV-dosage of 600 mJ was chosen as the lowest UV-dosage which yielded optimal complex formation. Identification of Cbp2 Peptides that Contact Intron 5 RNA
The UV-crosslinking technique standardized above was successfully employed to identify the RNA contact sites on Cbp2 protein. 32P-labeled intron 5 RNA transcripts were crosslinked to his-tagged Cbp2 under low salt conditions at a UVdosage of 600 mJ as described above. The Cbp2-RNA complexes generated were purified on SDS-polyacrylamide gels. The gel fragments were then soaked in different chemical cleavage reagents like hydroxylamine (NH2OH) and 2-nitro-5-
Figure 3-1. Optimization of UV-dosage for crosslinking. 32P -labeled intron 5 RNA was incubated with Cbp2 or BSA under low salt conditions as described in Materials and Methods. Samples were irradiated with UV-doses ranging from 100 to 950 mJ, extensively RNAase-treated, resolved on 10% SDS-polyacrylamide gels, and crosslinked complexes detected by autoradiography. UV-irradiation was done at 950 mJ for BSA (lane 1), and for Cbp2 at 100 mJ (lane 2), 200 mJ (lane 3), 300 mJ (lane 4), 400 mJ (lane 5), 500 mJ (lane 6), 600 mJ (lane 7), 800 mJ (lane 8), and 950 mJ (lane 9).
UV-dosage (mJ) Ln
BSA +--------------Cbp2 -+ + + + + + + +
12 34 56 78 9
thiocyanobenzoate (NTCB) to generate peptides. The peptides were separated on high percentage tris-tricine gels (Schagger and Von Jagow, 1987), and the crosslinked peptides that retained the label were identified by autoradiography (Figure 3-2).
NH20H cleaves proteins at asparaginyl-glycyl peptide bonds (Bornstein and Balian, 1977) and would yield three large peptides (15.1, 24, 37.1 kDa) in a complete digest of Cbp2. Cleavage of the crosslinked Cbp2-RNA complex (Figure 3-2, panel A) showed that the 24 kDa amino terminal and the 15.2 kDa carboxy terminal fragments of Cbp2 strongly crosslinked with intron 5 RNA, whereas the large central 37.1 kDa fragment displayed only a very weak signal. The fact that only two of the three peptides were strongly labeled suggests that the terminal fragments of Cbp2 might comprise important RNA binding domains. The weak signal retained by the central 37.1 kDa peptide suggests that other minor contact sites may be distributed throughout the length of protein. These sites of interaction may also contribute to the stabilization of the active intron structure, although the termini may be absolutely essential for the activity.
NTCB is specific to amino groups of cysteines (Jacobson et al., 1973; Degani and Patchornik, 1974). NTCB Cleavage of Cbp2 would produce 9 peptides ranging from 0.17 to 29.5 kDa if the reaction proceeded to completion. However, for several reasons, only a partial digestion of the protein could be achieved. Incomplete cleavage results from p-elimination and/or incomplete modification due to the reversible nature of the cyanylation reaction (Degani and Patchornik,
Figure 3-2. Chemical cleavage of Cbp2-intron 5 RNA complexes. His-tagged Cbp2 was crosslinked to 32P-labeled intron 5 RNA under low salt conditions in the absence of GTP, extensively RNAse treated, gel purified on 10% SDSpolyacrylamide gels, and digested in-gel with hydroxylamine (panel A) and NTCB (panel B). Following cleavage, peptides were extracted as described in Materials and Methods, separated on 15% (hydroxylamine) or 16.5% (NTCB) tris-tricine gels, dried and autoradiographed. Molecular weights of strongly crosslinked peptides are indicated by arrows, with asterisks representing partial cleavage products. Panel C. 0.2 mM GTP was added to the crosslinking reaction mixture and incubated at 370C for 30 minutes. The reaction products were ethanol precipitated, resolved on 4% polyacrylamide-8M urea gels and autoradiographed. Lane 1 shows intron 5 RNA alone incubated in the reaction mixture. Lanes 2 and 3 show the spliced products of intron 5 RNA incubated with native and his-tagged Cbp2 repectively. The input RNA and the spliced products are schematically represented on the left.
NH OH NTCB
-4-19.3 kDa* 24 kDa-P 15 kDa-- -4-7 kDa
C.1 2 3
1974). While complete denaturation of the protein is essential to obtain cleavage at the internal sites, Cbp2, a relatively large protein (73.4 kDa), appears to be somewhat refractile even to the strong denaturation conditions employed in these cleavage reactions. Cleavage of crosslinked Cbp2-RNA complexes with NTCB (Figure 3-2, panel B) generated several peptides that retained the label. Among the various indirectly labeled peptides, the 7.0 kDa N-terminal peptide and its corresponding partials (indicated by asterisks) of sizes 8.7 and 19.3 kDa could be readily identified. This further supports the finding that the N-terminus comprises an important RNA binding domain. However, the 5.6 kDa C-terminal fragment generated by NTCB (identified by silver staining, data not shown) did not retain the label, suggesting that the extreme C-terminal region may not be important for Cbp2-RNA interactions. The putative C-terminal contact site identified by NI20H cleavage may therefore be located upstream of this 5.6 kDa C-terminal fragment.
In order to demonstrate that the conditions employed for UV-crosslinking
promote the formation of active Cbp2-intron 5 RNA complexes, the reaction mixes were incubated with 0.2 mM GTP at 370 C for 30 minutes and the products were resolved on denaturing gels and autoradiographed. The results are shown in Figure 3-2, panel C. Reaction mixes containing either his-tagged (lane 3) or native Cbp2 (lane 2) protein clearly demonstrated splicing, while the RNA alone (lane 1) could not splice under similar low salt conditions.
The 6x histidine tag adds an additional 20 amino acid residues (about 2 kDa) to the N-terminus of the his-tagged Cbp2 protein. Therefore, the indirectly labeled N-terminal peptide and the partials obtained with his-tagged Cbp2 would migrate slower in the gels than their non-tagged counterparts. This difference in electrophoretic mobility was used as an analytical tool to confirm the assignment of the crosslink site to the N-terminal fragment (Figure 3-3, panels A and B). The NH20H and NTCB digestion patterns of the crosslinked complexes using either his-tagged or native Cbp2 were compared on the same gel. The results for NH2OH cleavage reactions run on 15% tris-tricine gels are shown in Figure 3-3, panel A. The N-terminal peptide of the his-tagged protein (lane 2) obtained by NH2OH cleavage migrated at 24 kDa level while the non-tagged peptide (lane 1) migrated faster (at 21.6 kDa level). The C-terminal peptides derived from both versions exhibited similar mobilities, since the his-tag is present only at the N-terminus. The corresponding NTCB digests of both versions of crosslinked Cbp2 are shown in Figure 3-3, panel B. The N-terminal peptide (7kDa) and the corresponding 8.7 and 19.3kDa partials (shown by asterisks) of his-tagged Cbp2 (lane 2) were shifted up in the 16.5 % tris-tricine gels compared to those generated from the native Cbp2 (lane 1). No other peptides shifted in the gel relative to the non-tagged version. These experiments showing differential mobility clearly demonstrate that the N-terminal crosslink site corresponds to the extreme N-terminal fragment. The minimal N-terminal peptide that crosslinked with intron 5 RNA was the 4.6 kDa peptide of the native Cbp2 that corresponds to the first 37 residues of the protein.
Figure 3-3. Confirmation of the crosslink site in the N-terminus of Cbp2.
Native and his-tagged Cbp2 proteins were crosslinked to 32P-labeled intron 5 RNA and digested with hydroxylamine (panel A) and NTCB (panel B) as described in the legend to Figure 3-2. Lanes 1 and 2 represent the cleavage patterns of native and his-tagged Cbp2, respectively. Molecular weights of strongly crosslinked peptides are indicated by arrows, with asterisks representing partial cleavage products. Note the slower migration of N-terminal derived fragments in the histagged Cbp2 lanes.
21.6 kDa--O- -4-24 kDa
16.9 kDa--O- 1 1 -4-19.3 kDa*
It -4-8.7 kDa*
*6.3 kDa--po- -4-7.0 kDa
4.6 kDa---lo1 2
Further analysis of the identified contact sites (Chapter 4) was restricted to the identified N-terminal fragment. However, certain conclusions can be drawn about the C-terminal fragment from the experiments described above. The 15 kDa Cterminal peptide generated by NH2OH cleavage showed strong crosslinking with intron 5 RNA (Figure 3-2, panel A). But the 5.6 kDa C-terminal fragment generated by NTCB (Figure 3-2, panel B) did not retain the label in crosslinking experiments. These results suggest that the 29.5 kDa penultimate C-terminal peptide of Cbp2 (aa 502-aa582) generated by NTCB has a potential RNA binding site.
The summary of findings from the UV-crosslinking experiments are shown in Figure 3-4. The map shows the two strong RNA binding regions (hashed boxes) that have been identified by these experiments, with one site being in the first 37 amino acids of the N-terminus and the other in a distally located C-terminal region (aa 502-aa 582). The digestion sites of NH2OH and NTCB on Cbp2 are also indicated in Figure 3-4. Further analysis of the importance of amino acid residues in the N-terminal fragment was carried out using site-directed mutagenesis (Chapter 4).
UV-crosslinking is a powerful tool to identify RNA contact sites on a protein, especially when the primary structure and homology searches do not afford any clues about critical functional elements of the protein. Cbp2, required for the
Figure 3-4. Summary of UV-crosslinking results. The peptide map represents the cleavage sites of NTCB (top arrows) and hyroxylamine (bottom arrows) on the 630-aa long Cbp2 protein. From the UV-crosslinking and chemical cleavage results shown in Figures 3-2 and 3-3, the intron 5 RNA crosslink sites on Cbp2 were mapped to the first 37 amino acids at the N-terminus and aa502-aa582 in the C-terminus (hashed boxes).
co 11 tC o0
CO LnC14 o coNTCB
__ _ _ __ _ _ __ _ _ I1l 630
1 -M VNWQTLFMVSLRRQG SS SRYRYKFNMENITHQVFPRC-37
splicing of intron 5 RNA, fits into this profile of proteins. Nothing is known so far about the RNA contact sites on the protein while extensive studies are available on the RNA component of the system. In the studies reported here, UV-crosslinking was employed to determine the contact points of intron 5 RNA on Cbp2 in an attempt to investigate the functional interactions between this group I intron and its protein co-factor. The photocrosslinking studies were performed under conditions that favored the formation of stable protein-RNA complexes. The Mg levels (10 mM) employed in these experiments were sufficient to promote 90-95% RNA binding to Cbp2 in equilibrium filter binding assays (data not shown), but splicing of RNA would not occur under these conditions due to the absence of guanosine (nucleophile). However, control experiments show that addition of GTP to the reaction facilitates splicing, demonstrating that the crosslinking conditions enable the formation of productive Cbp2-RNA complexes (Figure 3-2, panel C).
Prior to determining the sites of crosslinking, it is essential to establish the specificity of crosslinking between the RNA and the protein in question. In control experiments, intron 5 RNA failed to crosslink to the non-cognate protein BSA (Figure 3-1), showing that photocrosslinking (zero-length crosslinking) can occur only between functionally interacting species. Also, Cbp2 proteins carrying mutations in the N-terminal domain showed reduced or no crosslinking with intron
5 RNA compared to the wild-type protein (discussed in Chapter 4). Furthermore, Cbp2 does not facilitate splicing of intron 4 of COB pre-mRNA (Lewin, unpublished observation), a group I intron that also requires a protein co-factor in
vivo (Lamb et al., 1983; Banroques et al., 1986). Cbp2 protein is specific to intron 5 RNA in its splicing-enhancing function. This corroborates the authenticity of the crosslinking results, as UV-light crosslinks an amino acid to its neighboring nucleic acid base only when present in a specific orientation (Williams and Konigsberg, 1991). The results of various experiments described above strongly suggest that the crosslinking conditions employed permit the structural probing of specific, functional interactions between Cbp2 and intron5 RNA.
Precise identification of the residues that participate in photocrosslinking can be accomplished by amino acid analysis or amino acid sequencing, as described earlier. These conventional biochemical techniques, however, require crosslinking efficiencies of 20% or more. Unfortunately, low crosslinking efficiencies (less than 10%) were obtained under our conditions. Therefore, the RNA employed for some of the crosslinking experiments was double labeled with c 32P-UTP and a32PATP to increase the specific activity of RNA and enable detection of crosslinked peptides. There are several possibilities for the low yields of crosslinked product obtained in our system. Cbp2 and intron 5 RNA may have inherently poor tendencies to crosslink in spite of appreciable complex formation. On the other hand, exposure to UV-light could be causing significant photoinactivation of Cbp2. Photodamage of protein was reported to be a problem by other groups (Gott et al., 1991; Tanner et al., 1988). Finally, though UV-crosslinking indicates sites of protein-RNA contact, most of the affinity between protein and RNA may be attributable to contact sites that are not crosslinked under the conditions employed.
The low level of crosslinked product obtained in our system proved to be a potential problem for further biochemical characterization of crosslinked complexes. One of the ways to overcome this problem would be to enhance the photosensitivity of RNA using modified residues like 5-bromouridine or 5iodouracil and using a monochromatic laser instead of a broad spectrum ultraviolet light. For instance, the amino acid residue Tyr 85 of R17 bacteriophage coat protein was shown to be crosslinked with singly BrU-substituted hairpin RNA 1 (Gott et al., 1991) using a monochromatic XeCI excimer laser (308 nm) that yielded crosslinking levels exceeding 50%. Substitution of 5-iodouracil for uracil in the binding site for bacteriophage R 17 coat protein improved crosslinking levels to 80% in less than 5 minutes of irradiation (Willis et al., 1993). However, incorporation of 5-BrU into intron 5 RNA significantly reduced its autocatalytic activity, indicating that the structure of the ribozyme was perturbed (data not shown). Also, our attempts to crosslink Cbp2 protein with 5-BrU-RNA using a broad spectrum UV-source (Stratalinker) did not significantly enhance the extent of crosslinking. Since we did not have ready access to a monochromatic laser, we opted to employ alternative analytical methods which accommodate low crosslinking efficiencies.
A popular approach to analyze crosslinked complexes obtained at low yields is to employ extremely sensitive analytical techniques like Matrix-assisted, laser desorption/ionization-time of flight (MALDI-TOF) mass spectrometry (Karas and
Hillenkamp, 1988; Beavis and Chait, 1990 and Hillenkamp et al., 1991) and ladder sequencing (Chait et al., 1993), which typically require 10-15% crosslinking efficiency. Elaborate attempts were made to identify the crosslinked peptides/amino acid residues using this approach in collaboration with the Protein Core facility of the Interdisciplinary Center for Biotechnology Research (ICBR), University of Florida. Unfortunately, all attempts failed due to various technical problems including difficulty in removal of SDS, Coommassie Blue and other gelderived contaminants from the peptides. Consequently, the indirectly labeled Cbp2 peptides (crosslinked with intron 5 RNA) were identified by size separation on tris-tricine gels followed by Phosphorlmager analysis (Figures 3-2 and 3-3). Non-enzymatic cleavage agents were used instead of proteases to avoid interference from protease-derived peptides resulting from self-cleavage. Of several chemical cleavage reagents, hydroxylamine was chosen for the initial analysis of crosslinked complexes as it generates only a few large peptides in Cbp2 that can be readily identified on high percentage tris-tricine gels. NTCB was selected as a secondary reagent to further narrow down the contact points on Cbp2.
Chemical cleavage of the protein-RNA complexes with NH2OH and NTCB
showed that the termini of Cbp2 comprise important RNA binding domains, while several stretches in the central core of the protein may contribute to the overall stabilization of protein-RNA interactions (Figures 3-4). Experiments with histagged and native versions of Cbp2 (Figure 3-3) unambiguously demonstrate that the first 37 amino acids in the N-terminus of Cbp2 constitute a strong RNA contact
site. A second site may be located in the C-terminus between residues 502 and 582 of the protein. A similar architecture of RNA binding domains has been demonstrated with the mitochondrial tyrosyl-tRNA synthetase protein (Cyt- 18) of Neurospora which is essential for splicing several mitochondrial group I introns in addition to aminoacylation of tRNATyr (Akins and Lambowitz, 1987). The regions required for splicing are distributed throughout the Cyt-18 protein, as it binds the precursor RNA and facilitates formation of the catalytic RNA structure. These regions overlap with the stretches required for synthetase activity but are not identical to them. However, the principal RNA binding regions include a small, idiosyncratic N-terminal domain significantly absent in bacterial tyrosyl-tRNA synthetases (Cherniack et al., 1990) and a C-terminal tRNA-binding domain required for both splicing and synthetase activities (Kittle et al., 1991).
A comparison of Cyt 18 binding sites in N. crassa mt LSU and ND1 introns with that in N. crassa mt tRNATyr has revealed a remarkable three-dimensional overlap between the tRNA and the catalytic core of group I introns, suggesting an evolutionary relationship between group I introns and tRNA and perhaps the evolution of RNA splicing factors from cellular RNA-binding proteins (Caprara et al., 1996). Adaptation of a synthetase to facilitate group I intron splicing appears to be a relatively recent evolutionary improvement as it has been reported in only one other closely related fungus, Podospora anserina (Cherniack et al., 1990; Lambowitz and Perlman, 1990 and Kamper et al., 1992).
Cbp2 protein does not possess any sequence homology with Cyt 18 of N. crassa or Ytsl protein of P. anserina. However, the latter two share three blocks of amino acids required for splicing, of which one corresponds to the idiosyncratic Nterminal domain of Cyt- 18, while the other two are located in the putative Cterminal tRNA binding domain. Although Cbp2 is not a bifunctional protein, the structural similarities between the catalytic core of group I introns and tRNA may point to a common mechanism of recognition of a conserved tRNA-like structural motif in its cognate intron, intron 5 RNA.
MUTATIONAL ANALYSIS OF THE N-TERMINUS OF CBP2
UV-crosslinking studies (Chapter 3) show that the N- and C-termini of Cbp2 intimately contact intron 5 RNA. Now the challenge is to identify which amino acids in these regions are important for Cbp2 function. Unlike most other RNAbinding proteins, Cbp2 does not contain any well-characterized RNA binding motifs like the RGG box, RNP or KH motif (Burd and Dreyfuss, 1994). However, Cbp2 is rich in basic and hydrophobic amino acids, which is also a characteristic of double stranded RNA-binding proteins (St Johnston et al., 1992; Gatignol et al., 1993). The only protein with which Cbp2 displays any homology is its counterpart in S. douglasii (Li et al., 1996). However, the extremely high identity (87%) between the two homologs does not lend itself to the identification of potential RNA binding domains by sequence alignments. Therefore, the only viable alternative is to identify the important residues by biochemical or genetic methods. The low crosslinking efficiencies obtained in our system ruled out the use of biochemical methods such as anion-exchange HPLC or mass spectrometry to identify the critical residues. Therefore, site-directed mutagenesis was used to identify critical residues for Cbp2 function.
Site-directed mutagenesis is a powerful tool to study the structure-function relationship of single or combinations of amino acid residues in proteins. It has been classically used in probing various RNA-protein interactions. One of the many applications of this technique has been to dissect the individual functions of various amino acids in dual-function proteins. For instance, the MS2 (R17) bacteriophage coat protein which binds and encapsidates viral RNA also acts as a translational repressor of viral replicase by binding to an RNA hairpin in the RNA genome. LeCuyer et al., 1995, LeCuyer et al., 1996) successfully studied the RNA binding properties of MS2 coat protein independent of capsid assembly by isolating a Val75Glu; Ala8 1Gly double-mutant coat protein which had wild-typelike affinity and specificity for RNA, but was defective in capsid assembly.
Site-directed mutagenesis has also been employed to study the evolutionary
relationships of RNA-binding proteins in various species. For example, the second intron (bi2) of cyt b gene has two homologs in related Saccharomyces species that differ in their mobility. The S. capensis intron product is bifunctional, with both a DNA endonuclease and an RNA maturase function (Lazowska et al., 1992; Szczepanek and Lazowska, 1996), whereas the homologous S. cerevisiae intron product has only an RNA maturase function and is not mobile (Meunier et al., 1990; Lazowska et al., 1992). These two intron-encoded proteins differ by only four amino acid substitutions. Mutational analysis showed that replacement of two non-adjacent amino acids (Thr212Ala; Thr239His) in the S. cerevisiae maturase was necessary and sufficient for the acquisition of an endonuclease activity
promoting intron mobility (Szczepanek and Lazowska, 1996). Thus, the S. capensis protein could be considered more primitive in terms of mitochondrial group I intron-encoded protein evolution as the two activities have not yet diverged. The S. cerevisiae protein, on the other hand, lost the original function (intron mobility) and maintained the acquired one (RNA maturase function) during evolution.
Site-directed mutagenesis has been instrumental in delineating novel RNA binding motifs of proteins with little or no sequence homology to known RNA binding consensus sequences. The viral coat proteins of the plant viruses alfalfa mosaic virus (AMV) and tobacco streak virus (TSV) share little primary amino acid sequence identity (van Vloten-Doting, 1975; Reusken et al., 1995), but are functionally interchangeable in RNA binding (Zuidema and Jaspars, 1984) and initiation of infection (Gonsalves and Garnsey, 1975). The lysine-rich N-terminal RNA binding domain of the AMV coat protein lacks previously identified RNA binding motifs. Mutational analysis of this N-terminal region identified a single arginine whose specific side chain and position were crucial for RNA binding (Ansel McKinney et al., 1996). Consequently, protein sequence alignments between AMV, TSV, and other related viruses centered on this key arginine residue revealed a new RNA binding consensus sequence. This also explained in part why heterologous viral RNA-coat protein mixtures were infectious.
In the case of Cbp2, we employed site-directed mutagenesis to identify the
important residues (by either partial or complete loss of Cbp2 function) in the N-
terminus that crosslinked to intron 5 RNA. The C-terminal region that strongly crosslinked to RNA was not pursued further, as attempts to narrow down this region by double digestion of crosslinked complexes with NH2OH followed by cyanogen bromide were not feasible within the resolution of our gel system.
In order to identify targets for site-directed mutagenesis of Cbp2, the putative N-terminal RNA contact site on Cbp2 (spanning 37 residues) was scanned closely to allow the prediction of residues that might be important in RNA-protein interactions. This sequence, aal7 SSSRYRYKFNME aa28, has the following interesting features:
a. Charged residues alternate with aromatic residues. b. Polar residues flank the region of alternating charged and aromatic residues. While the charges could promote ionic interactions between protein and RNA, the aromatic residues could engage in stacking interactions. The stretch of serines could participate in hydrogen bonding interactions. These various possibilities led us to target this N-terminal region (aal7-aa28) for site-directed mutagenesis. The details of the analyses of all these mutants and a discussion of their implications are described in the following sections.
All mutant his-tagged Cbp2 proteins (Table 4-1) were purified by one-step
metal affinity chromatography as described in Materials and Methods. The serine deletion mutant and the Y23L mutant could not be successfully purified. The
Table 4-1. Description of Cbp2 mutants
Name of Mutant Mutant Description
Deletion mutant Deletion of aal7-aa28
Triple aromatic mutant Changes Y21, Y23, and F25 to L Triple charged mutant Changes R20, R22, and K24 to L
Y21IL mutant Changes Y21 to L
Y23L mutant Changes Y23 to L
F25L mutant -Changes F25 to L
Serine, deletion mutant Deletion Of S17, S18, and S19
former eluted in the wash fractions with other E. coli proteins, while the latter copurified with a nuclease activity and could not be assayed. The serine deletion appears to have altered the global conformation of Cbp2 (perhaps the topology of the histidine tag), resulting in poor binding to the nickel column. Independent attempts to purify the Y23L mutant with fresh reagents and columns still yielded preparations with high nuclease activity, raising the possibility that this mutation has conferred a nuclease function to Cbp2. This problem was not encountered in parallel preparations of any of the other mutants. Although these mutants appear to possess interesting properties, they were not characterized further.
The mutant Cbp2 proteins purified from E.coli were first analyzed by Western blotting with a Cbp2-specific polyclonal antibody to check for production of the full-length protein. Mutant proteins were separated on a 10% SDS-polyacrylamide gel and electroblotted to a nitrocellulose membrane at 15 volts, overnight. The membrane was first probed with Cbp2-specific primary antibody followed by secondary goat anti-rabbit antibody as described by Towbin et al. (1979). The Cbp2 bands were detected by chemiluminescence using ECL detection system (Amersham) (Figure 4-1, panel A and panel B). All mutant proteins (in both panels) except the deletion mutant (lacking aal 7-aa28) exhibited electrophoretic mobilities similar to that of wild-type Cbp2. The deletion mutant (aa 17-aa28) was shorter by about 1.3 kDa, as expected. These data confirm the synthesis of full length mutant Cbp2 proteins in E. coli. To test the effects of these mutations on splicing function, in vitro splicing assays were carried out. 32P-labeled intron 5
0 Cd > 0
-0 C;3 cd
(sz-zl) U014alaa C4
x4puio.ir aldiijL V-4
RNA transcripts were incubated with increasing concentrations of wild-type or mutant Cbp2 proteins under low salt splicing conditions. The reaction products were separated on 4% polyacrylamide-8M urea gels and autoradiographed (Figure 4-2). Increasing concentrations of deletion mutant (aal 7-aa28) (lanes 4 and 5) and triple aromatic mutant (lanes 6 and 7) failed to stimulate splicing of intron 5 RNA, while wild-type Cbp2 spliced normally at both concentrations tested (lanes 2 and 3). These results strongly suggest that the N-terminal residues (aal7-aa28) may comprise an important RNA binding domain essential for Cbp2 function. However, it is important to demonstrate that the loss of activity observed was not due to structural destabilization caused by these mutations.
Partial proteolysis is a useful technique to analyze the conformational states of proteins (Chang and Doi, 1993; Hay and Nicholson, 1993; Petersen et al., 1995; Ikeda et al., 1996; Liu et al., 1996). In this assay, deletion and triple aromatic mutant proteins (in native states) were incubated separately with trypsin under conditions that favored partial proteolysis as described in Materials and Methods. A control digest with native or heat-denatured wild-type Cbp2 was also done. The peptides were resolved on 12% SDS-PAGE gels and detected by Western blotting with a Cbp2-specific polyclonal antibody (Figure 4-3). The tryptic peptide profiles of native deletion (lanes 3 and 4, panel A) and triple aromatic (lanes 3 and 4, panel B) mutants closely resembled the pattern obtained with native wild-type Cbp2 (lanes 5 and 6 of both panels). In contrast, heat-treated wild-type samples (lanes 1 and 2 of both panels) exhibited a different pattern, showing aggregation of
Figure 4-2. Functional analysis of deletion (aa17-aa28) and triple aromatic mutants. 32p-labeled intron 5 RNA was incubated with increasing concentrations of wild-type and mutant Cbp2 under low salt splicing conditions at 370C for 1 hour, ethanol precipitated, resolved on a 4% polyacrylamide-8M urea gel, and autoradiographed. The precursor RNA and the products of splicing are schematically represented on the left side of the gel, with the ratio of protein to RNA shown on top of the lanes. Lane 1, precursor RNA alone; lanes 2 and 3, incubation with wild-type Cbp2; lanes 4 and 5, incubation with deletion (aal7aa28); lanes 6 and 7, incubation with triple aromatic mutant (Y21, Y23, F25 to L).
Protein: RNA 6
2 3 4 5 6 7
cis 4.) 4 0
0 0 J. .- 0 V) co cp
C-) u cq
u Cq u
u 00 u
o3 4-o a) q3
-cl u V)
ZdqD 4m 09:1
:)i Pwon aldlijL
ZdcD Im F os:l
zdqD IMF os:i
U014ala(l 001:1 co
ZdcD ImF osi: I pajnlPuaP-ILaH
denatured protein at the top of the gel. In addition, partial products corresponding to those obtained with native proteins were markedly absent in these digests. This is probably due to complete degradation of denatured protein molecules that were not present in aggregates. Thus, these results demonstrate that the mutations in Cbp2 did not alter the global conformation of these proteins.
Preliminary splicing experiments with triple charged mutant (R20, R22, K24 changed to leucine) and two point mutants, namely Y2 1 L (tyrosine to leucine change at position 2 1) and F25L (phenylalanine to leucine at position 25), showed varying degrees of activity. These mutants were characterized further by a series of time-course experiments that measured their initial rates of splicing (Figures 4-4 and 4-5). 32P-labeled transcripts were pre-incubated with each mutant protein at 37 0C for 10 minutes under low salt conditions and splicing was initiated by the addition of 0.2 mM GTP. The reactions were terminated at different times, resolved on denaturing polyacrylamide gels (Figure 4-4A & B) and autoradiographed. The F25L mutant (Figure 4-4A, right panel) stimulated splicing at levels comparable to wild-type Cbp2 (Figure 4-4A, left panel), while the triple charged and Y21IL mutants (Figure 4-413 left and right panels, respectively) showed drastic reduction in the extent of splicing. It is interesting to note that in the case of triple charged and Y2 I L mutants (compare Figures 4-4 A and 4-4B), the products of the first step of splicing (5' exon and the intron-3' exon) were barely detectable compared to the products of the second step (ligated exons and free intron), suggesting that the first step of splicing is rate limiting for these two
Figure 4-4. Time course of splicing for wild-type and mutant Cbp2. Radiolabeled intron 5 RNA was incubated with wild-type or mutant Cbp2 proteins under low salt splicing conditions. Reactions were terminated at indicated times, resolved on 4% polyacrylamide-8M urea gels, and autoradiographed. A schematic representation of the precursor RNA and spliced products are given on the left side of the figures. A. Shows the splicing time course for the wild-type Cbp2 (left panel) and F25L mutant (right panel). B. Shows the splicing time course for the triple charged mutant (left panel) and the Y21L mutant (right panel). A marker lane with wild-type Cbp2 was run on the right of each panel to indicate the location of spliced products.
Wt Cbp2 F25L
Time (min) 0 1 2 4 6 8 10 12 24 60 0 1 2 4 6 8 10 12 24 60
.. -..... .. .. -.... .. ......
.. .. . ... .. .... .0
Triple charged mutant Y21L mutant
Time (min) 0 1 2 4 6 8 10 12 24 60 0 1 2 4 6 8 10 12 24 60
Fgr t -A continued
Figure 4-5. Splicing rates of wild-type and mutant Cbp2. The gels shown in Figure 4-4 were quantitated using Phosphorlmager and the RNA fraction spliced calculated as the ratio of the sum of ligated exons and free 5' exon to the exon sequences present in the precursor RNA. The plots show RNA fraction spliced (+/-S.D) vs. time in min. A. Filled circles, Wt Cbp2; open circles, F25L mutant; filled squares, Triple charged (LYLYLF) mutant; open squares, Y2 1 L mutant. B. The plots for the last two mutants shown in A are plotted on an expanded scale to show initial rates of splicing. Filled circles, Triple charged (LYLYLF) mutant; open circles, Y2 1 L mutant.
--- Wt Cbp2
1.0 ---- Y21 L
0 5 10 15 20 25
0.08 -o- Y21L
.' 0.06 C-)
0 5 10 15 20 25
mutants. A quantitative representation of splicing for the wild-type and mutant proteins is shown in Figure 4-5A. A re-plot of the data corresponding to triple charged and Y2 IL mutants is shown on an expanded scale in Figure 4-5B3. Initial rates of splicing were calculated from the above plots for all mutants (Table 4-2). Figure 4-5A shows that the triple charged (filled squares) and the Y2 1 L (open squares) mutants exhibited extremely low activity compared to F25L mutant (open circles) or wild-type Cbp2 (filled circles). However, it is obvious from Figure 45B that both triple charged and Y2 1 L mutants accumulated spliced products with time. Initial rate measurements (Table 4-2) indicated that the rate of splicing of the F25L mutant, though appreciable, was slightly lower than that of the wild-type protein. In the case of the Y2 I L and triple charged mutants, the initial rate of splicing was lowered by -45-fold and -4O-fold, respectively, in comparison to wild-type Cbp2. Thus, tyrosine at position 21 is critical for activity while a phenylalanine at position 25 is dispensable. The charged residues at positions 20, 22 and 24 are also important for Cbp2 function.
To determine if lower splicing activity of the Cbp2 mutants corresponded to a reduction in overall affinity for intron 5 RNA, equilibrium binding assays were performed. 32P-labeled intron 5 RNA (16 pM) was incubated with increasing concentrations of wild-type or mutant Cbp2 (0-4000 pM) at 370C for 30 minutes and equal aliquots were filtered in duplicate through a double-filter consisting of nitrocellulose on top and charged nylon at the bottom, as described in Materials and Methods. Representative filter binding data are shown in Figure 4-6A. The
Table 4-2. Rate measurements for wild type and mutant Cbp2
Cbp2 protein RNA fraction spliced Initial rate of splicing at 60 minutes (fraction min'l)
Wild-type 0.94 0.22 0.189
F25L 0.68 0.09 0.162
LYLYLF 0.11 0.01 0.474 x 10-2
Y21L 0.05 0.0005 0.417 x 10-2
Figure 4-6. Double filter-binding assay of wild-type and mutant Cbp2. A. Radiolabeled RNA was incubated with increasing concentrations of wild-type and mutant proteins under low salt splicing conditions (without GTP) at 370C for 30 min. Equal aliquots of each reaction were filtered in duplicate through a sandwich consisting of nitrocellulose on top and charged nylon at the bottom, as described in Materials and Methods. RNA-binding patterns for the triple charged and F25L mutants and wild-type Cbp2 on nitrocellulose (left panel) and nylon (right panel) are shown. B. The dot blots shown in A. were quantitated using Phosphorlmager and the fraction RNA bound calculated as described in Materials and Methods. The fraction of RINA bound is plotted as a function of Cbp2 concentration. Open squares, Y2 1 L; filled squares, triple charged mutant; open circles, deletion (aalI7-aa28); filled circles, triple aromatic mutant; open triangles, F25L; filled triangles, wildtype Cbp2.
Nitrocellulose Nylon N+
Triple 0 0
charged 0 *#*#*0
0 0.6-c 0.4,
0.2 0 Deletion (aal7-28)
*-- Triple aromatic A-- Wild type & F25L
0.0 I I
0 1000 2000 3000 4000
Cbp2 (pM) Figure 4-6...continued
nitrocellulose membrane, which contains protein-RNA complexes (left panel) showed an increase in the bound RNA fraction with increasing Cbp2 concentrations, while the nylon membrane which contains free RNA (right panel) showed a corresponding decrease in the free RNA retained. Each filter binding experiment (in duplicate) was repeated at least two times and the RNA binding data (Figure 4-6B) used to calculate dissociation constants (kd) (Table 4-3).
The F25L mutant (open triangles), which showed wild-type-like splicing
activity, demonstrated RNA binding levels comparable to that of wild-type (filled triangles). The kd values of wild-type protein and F25L mutant were very similar (147 pM and 143 pM respectively). The triple charged (filled squares) and the Y21L mutants (open squares), which displayed partial splicing activity, showed reduced binding. These two mutants showed '-2.5-fold and '-5-fold increase in kd values, respectively, relative to wild-type Cbp2. The deletion mutant (open circles) showed slightly tighter binding (kd of 77 pM) than wild-type, whereas the triple aromatic mutant (filled circles) showed slightly lower binding levels (kd Of 184 pM) compared to wild-type Cbp2. Although the kd1 values are variable, the overall RNA binding profiles of mutants are similar to that of wild-type Cbp2 (hyperbolic). The similarity in binding isotherms for most of the mutant proteins confirms our conclusion that the amino acid changes have not significantly destabilized the higher order structure of protein. These results also suggest that the overall RNA binding pattern does not necessarily reflect the ability to stimulate splicing of intron 5 RNA.
Table 4-3. Dissociation constants of Cbp2 mutants
Cbp2 protein kd (pM)
Deletion (aal7-aa28) 77
Wild type Cbp2 147
F25L mutant 143
Triple aromatic mutant 184
(Y21, Y23, and F25 to L) Triple charged mutant 367
(R20, R22, and K24 to L) Y21L mutant 710
Since the putative contact sites in Cbp2 were initially identified by UVcrosslinking, the mutants were tested for their ability to crosslink to intron 5 RNA. 32P labeled RNA transcripts were UV-crosslinked to wild type or mutant Cbp2 protein, RNAase treated and resolved on 10% SDS-polyacrylamide gels. The crosslinked complexes were detected by autoradiography (Figure 4-7) and quantitated using PhosphorImager. RNA crosslinked in the absence of protein (lane 5) was almost completely degraded, without any detectable complexes. However, RNA crosslinked to wild type Cbp2 (lane 4) showed a prominent crosslinked species corresponding to the molecular weight of Cbp2. Crosslinking to triple charged (lane 3) and F25L (lane 1) mutant proteins was reduced by -50% and -53%, respectively, compared to wild type. The Y21L mutant (lane 2) showed extremely poor crosslinking to intron 5 RNA. The background bands seen below the level of Cbp2 (lanes 1 to 4) could be photodamaged Cbp2-RNA complexes or RNA-RNA crosslinks that were resistant to RNAase. Thus, mutations at positions R22, R24, K26, Y21 and F25 lowered the crosslinking efficiency of Cbp2. It is important to note that mutations targeting the above residues also affected splicing, with the exception of the F25L mutation. Thus, some of the residues in this region (aal7-aa28) that are important for Cbp2 function also appear to be involved in crosslinking Cbp2 to intron 5 RNA.
As partial proteolytic profiles (Figure 4-3) and filter binding curves (Figure 46B) of deletion (aal7-aa28) and triple aromatic mutants were similar to that of
Figure 4-7. UV-crosslinking of wild-type and mutant Cbp2 to intron 5 RNA. Wild-type or mutant Cbp2 was incubated with radiolabeled intron 5 RNA under low salt conditions in the presence of 20 ug/ml tRNA. The samples were then UVcrosslinked at 600 mJ, extensively RNAase treated, resolved on 10% SDS-PAGE gels, and autoradiographed as described in Materials and Methods. Lane 1, F25L mutant; lane 2, Y21L mutant; lane 3, triple charged (LYLYLF) mutant; lane 4, wild-type Cbp2; lane 5, radiolabeled RNA alone. The molecular weight position of Cbp2 is indicated by the arrow.
cq C4 74 kDa-01 2 3 4 5
wild-type, they were tested for their ability to compete with wild-type Cbp2 in splicing assays. Splicing reactions were set up at wvild-type Cbp2 to intron 5 RNA ratio of 7:1 in the presence of increasing concentrations of mutant proteins. Reaction products were resolved on denaturing gels and autoradiographed (Figure 4-8A). Wild-type Cbp2 alone (lane 1) spliced normally while the deletion (lane 2) or triple aromatic (lane 7) mutant alone was completely defective in splicing activity, as reported before. Addition of increasing concentrations of deletion (lanes 3 to 6) or triple aromatic (lanes 8 to 11) mutant protein to wild-type Cbp2mediated reactions showed a progressive inhibition of splicing (compare lane 1 with lanes 3 to 6 and 8 to 11). These results are quantitatively represented in Figure 4-8B3. The graphs show percentage splicing as a function of mutant:wildtype ratios, with the extent of splicing obtained in the presence of wild-type Cbp2 alone set to 100%. Results of the addition of deletion and triple aromatic mutants are shown in the left and right panels, respectively. Addition of a 3-fold excess of deletion mutant lowered splicing levels to 40% of control levels. A similar inhibition of splicing was observed with triple aromatic mutant, although at a higher ratio (9: 1) of mutant to wild-type Cbp2. Thus, Cbp2 mutants inhibited the protein-mediated splicing of intron 5 RNA when present in excess over the wildtype protein.
It is important to note that at the highest level of splicing inhibition (Figure 4813), total protein (mutant + wilId-type) to RNA ratio was 7 1: 1 for the deletion mutant and 140:1 for the triple aromatic mutant. It is, therefore, possible that the
Figure 4-8. Effect of mutant proteins on wild-type Cbp2-mediated splicing. Splicing reactions were performed at 7:1 wild-type Cbp2:RNA in the presence of increasing concentrations of deletion (aal 7-aa28) or triple aromatic mutant proteins. Reactions were resolved on 4% polyacrylamide-8M urea gels and autoradiographed. A. Lane 1, wild-type Cbp2 alone; lane 2, deletion mutant (aal7-aa28) alone; lanes 3-6, constant amount of wild-type + increasing concentrations of deletion mutant proteins; lane 7, triple aromatic mutant alone; lanes 8-11, constant amount of wild-type + increasing concentrations of triple aromatic mutant proteins. The ratio of mutant to wild-type Cbp2 in each reaction is indicated above the lanes. The precursor RNA and spliced products are schematically represented on the left of the gel. B. The splicing gels in A were quantitated using PhosphorImager and the % splicing (determined from ligated exons) plotted as a function of mutant:wildtype Cbp2, setting the extent of splicing with wild-type Cbp2 alone to 100%. The left and right panels show data corresponding to deletion (aal 7-aa28) and triple aromatic mutants, respectively.
Deletion Triple aromatic
Mutant: Wt Cbp2 ', 7 71 71' c: 00 67 6 6$
C) CD M cr, 00
12 3 4 56 78 910 1
Deletion (aal7-28) Triple aromatic mutant
60 6040o 40
0:1 0.6:1 3:1 6:1 9:1 0:1 1.8:1 9:1 18:1 27:1
Mutant: Wt-Cbp2 Mutant : Wt-Cbp2
inhibition observed in this experiment was a function of total protein concentration (mutant + wild-type Cbp2), rather than being a property of mutants. To test this possibility, splicing was performed with increasing concentrations of wild-type Cbp2 alone (Figures 4-9A and 4-9B3). Maximum splicing activity was obtained at a protein:RNA ratio of 7:1 (lane 4 in Figure 4-9A; expressed as 100% activity in Figure 4-9B3). However, splicing was severely inhibited at 28:1 (lane 6), and almost completely inhibited at higher ratios (lanes 7 and 8) of wild-type Cbp2:RNA. The splicing levels dropped to 21% (for 28:1) and -1% (for 56:1 and 112: 1) of the activity obtained at 7:1 ratio (Figure 4-9B3). Thus, wild-type Cbp2 appears to stimulate splicing only in a narrow range of protein:RNA ratios, with higher levels being inhibitory. In order to determine whether aggregation was the cause for the observed inhibition of activity, these experiments were repeated over a wide range of concentrations of wild-type Cbp2. Identical inhibition was observed whether the protein was titrated at lower (3.6-5 8 nM) or higher (up to 116 nM) ranges, suggesting that aggregation of protein was not a problem in these experiments (data not shown).
It is possible that the observed inhibition of splicing at higher concentrations of wild-type protein could be due to non-specific interactions of Cbp2 (a highly basic protein) with its RNA counterpart. This possibility was tested using the nonspecific competitor, tRNA, in partially inhibited splicing reactions. A titration of Cbp2 concentration in the absence of tRNA is shown in Figures 4- 1 OA (lanes 1 to 6) and 4-l1OB (left panel). The maximal splicing observed at 14:1 ratio (lane 3,
Figure 4-9. Effect of increasing concentrations of wild-type Cbp2 on proteinmediated splicing. Radiolabeled intron 5 RNA was pre-incubated with increasing concentrations of wild-type Cbp2 for 10 min at 370C. Splicing was initiated by the addition of 0.2 mM GTP and incubation continued at 37'C for 30 min. Spliced products were resolved on 4% polyactylamide-8M urea gels and autoradiographed. A. Lane 1, intron 5 RNA alone; lane 2, splicing at protein:RNA of 1:1; lane 3, 3.5:1; lane 4, 7:1; lane 5, 14:1; lane 6, 28:1; lane 7, 56:1; and lane 8, 112:1. B. PhosphorImager quantitation of the gel shown in A was used to plot the % splicing (ratio of the sum of ligated exons and 5' exon to unspliced precursor) as a function of wild type-Cbp2 to RNA ratios, setting the extent of splicing obtained at 7:1 to 100%.
Cbp2: RNA 5 t 4 N
1 2 3 4 5 6 7 8
1:1 3.5:1 7:1 14:1 28:1 56:1 112:1
Figure 4-9... continued