RNA-Protein Interactions of a Kink

Florida State University Libraries

Electronic Theses, Treatises and Dissertations The Graduate School

2005 Biophysical and Biochemical Investigation of an Archaeal Box C/D SRNP: RNA- Protein Interactions of a Kink Turn RNA within the Functional Enzyme Terrie Luong Moore

Follow this and additional works at the FSU Digital Library. For more information, please contact [email protected] THE FLORIDA STATE UNIVERSITY

COLLEGE OF ARTS AND SCIENCES

BIOPHYSICAL AND BIOCHEMICAL INVESTIGATION OF

AN ARCHAEAL BOX C/D SRNP: RNA-PROTEIN

INTERACTIONS OF A KINK TURN RNA WITHIN THE

FUNCTIONAL ENZYME

TERRIE LUONG MOORE

A Dissertation submitted to the Department of Chemistry and Biochemistry in partial fulfillment of the requirements for the degree of Doctor of Philosophy

Degree Awarded: Summer Semester, 2005

The members of the Committee approve the dissertation of Terrie Luong Moore defended on May 3, 2005.

______Hong Li Professor Directing Dissertation

______Lloyd M. Epstein Outside Committee Member

______Timothy M. Logan Committee Member

______John G. Dorsey Committee Member

Approved:

______Naresh Dalal, Chair, Department of Chemistry and Biochemistry

The Office of Graduate Studies has verified and approved the above named committee members.

I dedicate this work to my husband, Michael, for being there every step of the way and always giving me hope about the future. Without your love and support, this work would not have been possible.

iii

ACKNOWLEDGEMENTS

I would like to thank my mentor, Dr. Hong Li, first and foremost for taking me into her lab and allowing me to grow intellectually. I would like to thank all of my committee members, Dr. Timothy Logan, Dr. Lloyd Epstein, and Dr. John Dorsey, for their knowledge and guidance during my graduate career. I am very thankful for my husband, Michael Moore, who believed in me even when I didn’t and kept me going even when the waters were rough. I want to thank all my family for their love and support. They have always told me that I could do whatever I wanted to do. I want to thank all the members of the Li lab especially Rumana Rashid and Sri Oruganti for all the coffee and laughs and for keeping me sane. Thanks to Laura Jane Phelps, Meredith Newby Lambert, and Kersten Schroeder for helping me through the rough times. I would like to thank Soma for all his help with crystal diffraction and the multiple discussions on many other topics in the lunch room. Thank-you to Dr. Jack J. Skalicky for his NMR help and career advice. I would like to thank Marcia O. Fenley for her help on the electrostatics work. I would like to thank Daphne Williams and her little man for making me smile even when I didn’t want to. I would also like to thank (Kitty) Kat Phipps for sharing my goofy sense of humor. And last, but definitely not last, thanks to Christina Wimberly, my best friend, for helping me relax when I needed it the most and understanding things that nongraduate people would normally not understand. My time here has taught me so much about science and about life. I will miss all the friends I made here and I will miss Tallahassee a lot more than I ever imagined. Good luck in your future endeavors and may our paths cross again. I would also like to acknowledge the funding from the NIH and AHA agencies for the box C/D project.

TABLE OF CONTENTS

List of Tables vii List of Figures viii List of Abbreviations x Abstract xii

INTRODUCTION 1

Noncoding RNAs 1 Small Nucleolar Ribonucleoprotein Particles 2 History of snRNAS 6 Archaeal Homologs of Small Nucleolar RNAs 8 Location of SnoRNA genes 9 Significance of Nucleotide Modifications 10 Box C/D SnoRNPs 13

CRYSTAL STRUCTURE AND MOLECULAR DETAILS OF A BOX C/D RNA-L7AE COMPLEX 19

Introduction 19 Results and Discussion 23 Conclusion 52 Materials and Methods 54

BOX C’/D’ RNA-L7AE CRYSTALLIZATION AND PRELIMINARY X-RAY DIFFRACTION DATA 58

Introduction 58 Results and Discussion 59 Conclusion 64 Materials and Methods 66

DESIGN AND CRYSTALLIZATION OF AN ENTIRE BOX C/D SNORNP 68

Introduction 68 Results and Discussion 68 Conclusion 78 Materials and Methods 78

CONCLUSIONS 80

REFERENCES 82

BIOGRAPHICAL SKETCH 90

LIST OF TABLES

1. Data Collection, Phasing and Refinement Statistics 25 2. Crystallization Trials for L7Ae-box C'/D' RNA Co-crystal 61 3. Optimization of L7Ae-Box C'/D' 26mer Crystal 62 4. Crystallization Trials for Box C/D sRNP Supercomplex 75 5. Optimization of Supercomplex Crystal with Constructs S4, S5, and S6 76

vii

LIST OF FIGURES

1. Ribonucleoprotein Particles 3 2. Classes of snoRNAs 5 3. U3 snoRNA Secondary Structure 7 4. Modifications in the E. Coli and yeast rRNA 11 5. RNA Sugar Conformers 12 6. Models of Box C/D snoRNP and sRNP Assembly 15 7. Crystal Structure of AF Fibrillarin-Nop5p Complex 16 8. Representations of Protein-Kink turn Complexes 21 9. Kink Turn RNA Sequences 22 10. RNA Constructs for Structural Characterization with L7Ae 26 11. EMSA of Box C/D RNA Constructs with L7Ae 27 12. L7Ae-box C/D RNA Crystal 27 13. Electron Density Map of L7Ae-box C/D RNA Complex 28 14. Crystal Arrangement of L7Ae-box C/D RNA 25mer Crystal 30 15. L7Ae Structure and Alignment 31 16. Surface Electrostatic Potentials of AF L7Ae, human 15.5 kD, and HM L7Ae 33 17. Structure of Box C/D 25mer RNA 34 18. Structural Features of Box C/D RNA Kink Turn 36 19. Structural Comparison of Three K-turn RNAs Bound with the 15.5 kD/L7Ae protein 38 20. Comparison of Electrostatic Surface Potentials of Three Kink Turn RNAs 40 21. Comparison of K-turn Orientations with Respect to Their Bound Proteins 41 22. Superimposed Complexes of 15.5kD-U4 snRNA and L7Ae-box C/D RNA 43 23. Overall Crystal Structure of the L7Ae-box C/D RNA Complex 44 24. Comparison of L7Ae-box C/D RNA Contacts with the 15.5 kD-U4 5‘-stem RNA and HM L7Ae-KT-15 Complexes 46 25. Base Pairing that Gives Rise to Imino Resonances 48 26. 1D spectra of L7Ae in the Absence and Presence of Box C/D RNA 50

viii 27. 1D NMR Spectra of Box C/D 21mer RNA Titration with 15N-L7Ae 51 28. Box C’/D’ RNA Constructs for Crystallization with L7Ae 60 29. EMSA of Box C’/D’ Constructs with L7Ae 60 30. Pictures of L7Ae-box C’/D’ 26mer RNA Crystals 63 31. 1D Spectra of box C’/D’ 20mer RNA with L7Ae 65 32. Native Constructs for Supercomplex Crystallization 69 33. Oligonucleotide Box C/D RNA Constructs 71 34. EMSA of Supercomplex RNA Constructs with L7Ae and Fibrillarin- Nop5p Complex 73

35. Box C/D sRNP Crystal with Construct S4 77

LIST OF ABBREVIATIONS

AF: Archaeoglobus fulgidus A: adenosine ATP: adenosine triphosphate C: Cytidine CS: canonical stem CNS: Crystallography NMR System DNA: Deoxyribonucleic acid EDTA: Ethylene diamine tetra acetate EM: Electron microscopy EMSA: electrophoretic mobility shift assay FPLC: fast performance liquid chromatography G: guanosine monophosphate HM: Haloarcula marismortue HPLC: High-performance liquid chromatography Kt or K-turn: kink turn MAD: multiple wavelength anomalous diffraction miRNA: micro RNA MJ: Methanococcus jannaschii mRNA: messenger RNA MRP: mitochondrial RNA processing NAIM: nucleotide analog interference modification ncRNA: noncoding RNA NCS: noncannonical stem PBE: Poisson-Boltzmann equation PEI: polyethyleneimine PEG: polyethylene gylcol R-M enzyme: restriction modification enzyme

x RNA: ribonucleic acid RNase: ribonuclease RNP: ribonucleoprotein particle rRNA: ribosomal RNA SAM: S-adenosyl-l-methionine SBP: SECIS binding protein SECIS: selenocysteine insertion sequence scRNPs: small cytoplasmic RNPs siRNAs: small interfering RNAs snoRNA: small nucleolar RNA snoRNPs: small nucleolar RNP snRNA: small nuclear RNA snRNP: small nuclear RNP sRNA: small RNA sRNP: small RNP SRP: signal recognition particle TBE: tris-Borate EDTA Tt: Thermus thermophilus U: uridine UTP: uridine triphosphate UV: ultraviolet electromagnetic radiation WC: Watson-Crick base pair

ABSTRACT

Box C/D snoRNPs catylze the specific 2’O-methylation of rRNA in important regions the ribosome, although the role of the modifications is unclear. Eukaryotic box C/D snoRNPs consists of a box C/D RNA and four proteins, Fibrillarin, Nop56, Nop58, and 15.5kD. Archaeal homologs are simiplified containing three proteins, L7Ae, Nop5p, and Fibrillarin with a box C/D RNA. The box C/D sequences are proposed to form the recently recognized kink turn structure which is found in many types of RNA. Most are associated with proteins and protein binding may nucleate the assembly of other proteins onto the RNA. Dissecting the structure and biochemical properties of box C/D snoRNPs may not only help in understanding the function of the modifications, but may also give insight into the role of kink turn RNAs in RNP assembly. An archaeal box C/D RNA embedded within the intron of pre-tRNATrp from Archaeglobus Fulgidus(AF) that guides two modifications in the tRNA was used as the model for the investigation of three complexes: L7Ae-box C/D RNA, L7Ae-box C’/D’ RNA, and the entire box C/D sRNP. Extensive crystallization trials resulted in crystals for each complex. A crystal structure of the box C/D RNA-L7Ae complex was determined to 2.7Å and shows the box C/D sequences do form a kink turn. Detailed structural comparisons of the AF L7Ae-box C/D RNA complex with previously determined crystal structures of L7Ae homologs in complex with functionally distinct kink turn RNAs revealed a conserved RNA-protein interface suggesting a conformational “adaptability” of the kink turn RNAs in binding L7Ae homologs. NMR characterization of L7Ae-box C/D RNA and L7Ae-box C’/D’ RNA interactions suggests a structural change in the RNAs upon binding L7Ae and the RNAs may be dynamic structures that do not form stable kink turns alone. The underlying differences in primary and secondary structures of the kink turns may lead to different tertiary structures and dynamic behavior in kink turn RNAs that may confer specificity of L7Ae homologs for different kink turn RNAs. These analyses provide a

xii structural basis for interpreting the functional roles of the box C/D sequences in directing specific assembly of box C/D sRNPs.

xiii

CHAPTER 1 INTRODUCTION

Noncoding RNAs (ncRNAs)

The importance of RNA in cellular functions has only been recently recognized. Since the early 1950’s the role of RNA within the cell has been overshadowed by the “central dogma” of molecular biology. The “dogma” describes a general pathway of passing genetic information from DNA to RNA which is then translated into protein products. RNA was considered an accessory in the process of creating functional proteins and the ‘one gene-one protein’ theory prevailed. For prokaryotes whose genomes are composed mostly of tightly packed protein coding sequences, this theory is mostly true. In E. coli, noncoding RNA makes up about 0.5% of the total number of genes (protein and RNA) and about 0.2% of the transcriptional output (Mattick, 2003). This is not the case for higher organisms. It was discovered in the 1970’s that higher organisms have intervening sequences (introns) between protein coding sequences (exons) which are spliced out of mRNAs (Gilbert, 1978). This complicated the ‘one gene-one protein’ theory, but it was quickly assumed that the RNA introns were nonfunctional. At that time, RNA was considered to have only three flavors: rRNA, tRNA, and mRNA all of which function in making proteins. This describes a very minimal role for RNA within the cell. If we look at the human genome, the number of coding sequences (30-40,000 genes) makes up only about 2% of the transcriptional output leaving about 98% of transcribed RNA as non protein coding RNA (Gilbert, 1978). The numbers alone suggest a more complicated role of RNA in cellular functions otherwise making most of the transcriptional output very wasteful.

1 It is now widely known that RNA has multiple functions within the cell. Types of RNA include the three mentioned above, along with snRNAs (small nuclear), snoRNAs (small nucleolar), miRNAs (micro), siRNAs (small interfering), and many others. Many of these RNAs interact with proteins to form functional complexes responsible for steps in gene expression, gene imprinting, translational repression, immune responses, and more. These complexes are called ribonucleoprotein particles.

Small Nucleolar Ribonucleoprotein Particles (snoRNPs)

Ribonucleoprotein particles (RNPs) are complexes composed of protein and RNA that have critical functions within the cell. Perhaps the most well known RNP is the ribosome which is responsible for catalyzing peptide bond formation. Eukaryotic ribosomes are composed of over 70 proteins and four rRNAs making it a large RNP. There are many other small RNPs (sRNPs) comprised of one or more proteins and one small RNA molecule (less than 300 nucleotides). These complexes exist in multiple compartments of the eukaryotic cell and carry out various fundamental steps of gene expression. Small RNPs of the cytoplasm (scRNPs) such as the signal recognition particle (SRP) are responsible for protein translocation. Small nuclear ribonucleoprotein particles (snRNPs) are RNPs found in the nucleus of eukaryotic cells and can be divided into two families. The snRNPs found in the nucleoplasm are responsible for processing of pre-messenger RNAs (pre-mRNA) for export to the cytoplasm. The most extensively studied snRNP is the major spliceosome, which catalyzes the removal of intervening sequences from pre-mRNAs. Many of these RNPs are currently being studied and some of their assembled structures are shown in figure 1. The second family of snRNPs, called snoRNPs (small nucleolar ribonucleoprotein particles), is found in the nucleolus. The subnuclear compartment is responsible for the biogenesis of ribosomes and thus, snoRNPs are implicated in this process (Maxwell and Fournier, 1995). In the nucleolus of eukaryotic cells, ribosomal DNA is transcribed into precursor transcripts. These pre-ribosomal RNAs (pre-rRNAs) need to be processed before export to the cytoplasm. These processes include specific nucleotide modifications, cleavage of precursor rRNAs, and assembly of rRNAs into large and small ribosomal subunits (Yu, et

3 al., 1999). There are nearly 200 snoRNAs utilized in the modification and cleavage events that occur during ribosome biogenesis (Tollervey and Kiss, 1997; Lafontaine and Tollervey, 1998; Weinstein and Steitz, 1999). Most snoRNPs can be categorized into two major classes defined by conserved sequence elements of the snoRNAs (Figure 2). Each type of snoRNA associates with specific proteins essential for their snoRNP functions. The box H/ACA snoRNAs form an evolutionary conserved ‘hairpin-hinge-hairpin- tail’ structure. The single-stranded hinge region and the tail region of the RNA contain two conserved sequences termed boxes H and ACA, respectively. Box H/ACA snoRNAs also contain one or two antisense elements that direct site specific conversion of uridine to pseudouridine in rRNA. The box C/D snoRNAs contain two conserved sequence elements termed box C and box D near the 5’ and 3’ ends, respectively. The box elements come together to form a predicted ‘kink turn’ structure. The box C/D snoRNAs, similar to box H/ACA snoRNAs, also contain one or two antisense elements that guide methylation of the ribose 2’- hydroxyl in rRNA. SnoRNPs not only modify over 200 specific sites in rRNA, but also direct 2’-O- methylation or pseudouridylation in spliceosomal RNAs, transfer RNAs, and some messenger RNAs. SnoRNAs have been shown to guide U5 and U6 snRNA modifications (Tycowski, et al., 1998; Ganot, et al., 1999; Jady and Kiss, 2001). There is also a recently discovered subset of snoRNAs localized in the cajal bodies called scaRNAs (small cajal body RNAs) that are predicted to guide modification of the U1, U2, and U4 snRNAs (Darzacq, et al., 2002). Messenger RNAs (mRNAs) specific in the human brain and trypanosome are predicted targets of some guide snoRNAs (Cavaille, et al., 2000; Liang, et al., 2002). In Archaea, some guide sRNAs are specific for 2’-O- methlylations in tRNA (Omer, et al., 2000; Dennis et al., 2001; Clouet d'Orval, et al., 2001). SnoRNAs are also implicated with the cleavage of pre-rRNA although these processing events are less understood than the modification events (Yu, et. al., 1999). Most of the snoRNAs that function in the cleavage of rRNAs share some sequence homology with the snoRNAs involved in rRNA modification. There is one snoRNA involved in the cleavage of precursor rRNA that does not fall into the two major classes,

4 5 the endonuclease MRP (mitochondrial RNA processing or 7-2) RNA (Figure 2). This RNA has a similar secondary structure to RNase P and makes up its own minor class of snoRNAs (Bertrand and Fournier, 2004). These modification and processing functions mediated by snoRNPs may be important regulatory events in the gene expression pathway.

History of SnoRNAs

The first snoRNAs were isolated from mammalian cells in the early 1970’s along with splicing snRNAs (Maxwell and Fournier, 1995). These RNAs were rich in uridine and thus were given U-designations that are still used today (Bertrand and Fournier, 2004). The first snoRNA to be characterized was the U3 which is required for cleavage of pre-rRNA transcripts (Kass, et al., 1990) and appears to ubiquitous in eukaryotes. Conserved box sequences A-D and similar secondary structures were disclosed by sequence analysis of the first few U3 snoRNAs isolated (Figure 3) (Wise and Weiner, 1980; Parker and Steitz, 1987; Porter, et al., 1988; Jeppesen, et al., 1988; Mereau, et al., 1997; Samarsky, et al., 1998; Wormsley, et al., 2001). Development of antibodies specific for the nucleolar protein fibrillarin in the late 1980’s allowed for the identification of more snoRNAs (Maxwell and Fournier, 1995). Analyses of these new snoRNAs identified box C and D sequences initiating the box C/D classification of one of the two families of snoRNAs (Tyc and Steitz, 1989). The other family of snoRNAs, the box H/ACA family, was not recognized until 1996 using comparative sequencing of small nuclear RNAs in yeast (Balakin, et al, 1996; Ganot, et al., 1997). The majority of snoRNAs identified are from human and yeast cells, but snoRNAs have been shown to exist throughout eukarya including many metazoan organisms, plants, and protists. Homologs of snoRNAs have also been found in archaea (Omer, et al., 2000), but not bacteria.

Archaeal Homologs of Small Nucleolar RNAs

Eukaryotic ribosomes were found to contain ten times more sites of 2’-O-methylation and pseudouridylation than bacterial ribosomes (Lafontaine and Tollervey, 1998). E. coli ribosomal RNAs contain four methylated riboses and ten pseudouridines compared with ~50 of each in Saccharomyces cerevisiae and ~100 of each in humans (Maden, 1990). Bacteria use different protein enzymes that recognize the sequence and/or structure of the target site and usually modify several similar sites (Lafontaine and Tollervey, 1998). Instead of using a different protein enzyme for each site, eukaryotes use common protein machinery with different snoRNAs to target a large number of nucleotide modifications. Many snoRNAs and several nucleolar proteins are involved in the modification of eukaryotic rRNA. Some archaeal ribosomes were also observed to contain a large number of ribose methylations such as Sulfolobus solfataricus which has 67 sites (Noon, 1998). Although archaea lack a nucleolus, their genomes encode several homologs of the protein machinery used in eukaryotic box C/D snoRNP systems and thus a similar method of ribose methylation was predicted for archaea. The first homologs of snoRNAs in archaea were discovered in 2000 using immunoprecipitation with homologous box C/D proteins from archaea and then cloning of the associated RNAs (Omer, et al., 2000; Gaspin, et al., 2000). Since archaea lack a nucleus, these RNAs are termed sRNAs and the resulting particles sRNPs. These sRNAs were determined to be box C/D RNAS and exist in both phyla of archaea, Crenarchaeota and Euryarchaeota. The box C/D RNAs in eukarya and archaea are very similar in box sequences and general structure (Bertrand and Fournier, 2004). In fact, archaeal box C/D sRNAs have been shown to properly assemble and guide methylation when introduced into Xenopus oocytes (Speckmann, et al, 2002). There are several differences observed between archaeal and eukaryl box C/D RNAs. Archaeal box C/D RNAs tend to be shorter with an average length of 50-60 nts while eukaryl snoRNAs are in the size range of 75-100 nts (Omer, et al., 2000). The majority of archaeal sRNAs are able to direct ribose methylation using both D and D’ boxes, compared to only 20% of eukaryl snoRNAs being used as double guides (Omer, et al., 2000). Archaeal box C/D RNAs

8 have also been found within the introns of tRNAs whereas eukaryl box C/D RNAs have not (Omer, et al., 2000; Gaspin, et al., 2000, Clouet d'Orval, et al, 2001; Tang, et al., 2002). Recently, homologs of box H/ACA snoRNAs have also been found in archaea. There are fewer occurrences of pseudouridines in archaea than in eukaryotes and thus far only four box H/ACA sRNAs have been identified from Archaeoglobus fulgidus (Tang, et al., 2002). It is predicted that these sRNAs assemble with proteins similar to those of the eukaryotic box H/ACA particle. Archaeal box H/ACA RNAs can vary significantly from the eukaryl RNAs such as lacking the H box and containing a single hairpin with the ACA box or even a three guided motif.

Location of SnoRNA genes

SnoRNA genes have been found both within introns and as independent transcription units. The majority of vertebrate snoRNAs are embedded within introns of protein genes that function in nucleolar, ribosome structure, or protein synthesis activities (Maxwell and Fournier, 1995). Although a few snoRNAs such as U3, U8, U13, and MRP snoRNAs are encoded as independent genes (Maxwell and Fournier, 1995). Some snoRNAs are transcribed in the intergenic regions of nontranslatable mRNA-like transcripts (Bertrand and Fournier, 2004). The recently discovered brain specific box C/D snoRNAs are also coded in large tandem repeats (Cavaille, et al., 2000). In yeast, the opposite is true where most snoRNAs are coded as genes that produce snoRNAs only. These genes can be monocistronic and polycistronic which code from two to seven snoRNAs (Bertrand and Fournier, 2004). There is a small minority of yeast snoRNA genes encoded within introns. In archaea, the situation is similar to vertebrates where most known or predicted box C/D sRNAs exist in intronic regions (Bertrand and Fournier, 2004). They are also known to exist within tRNA introns and in transcribed spacers of rRNA which are encoded as half-molecules joined after cleavage of pre-rRNA (Clouet d'Orval, et al., 2001; Tang, et al., 2002).

Significance of Nucleotide Modifications

The large number of modified nucleotides in eukaryotic ribosomes compared with bacterial ribosomes suggests functional or structural importance for 2’-O-methylation and psuedouridine modifications. The sites of modification are confined to the most conserved and functionally important regions of the ribosome including the PTC (peptidyl transferase center), A, P, and E sites, polypeptide exit tunnel, and subunit bridges (Decatur and Fournier, 2002) (Figure 4 ). Most modifications are in the core of the RNA mass and oriented towards the subunit faces while absent from regions dominated by ribosomal proteins such as the surfaces and outer edges of the interface domains (Ban, et al., 2000; Schluenzen, et al., 2000; Wimberly, et al., 2000; Yusupov, et al., 2001). Individual modifications have not been shown to be essential for ribosome function (Ni, et al., 1997; Green, et al, 1999; Khaitovich and Mankin , 1999;Lovgren and Wikstrom, 2002), but strong negative growth effects in yeast occur when modifications are blocked globally (Tollervey, et al., 1993; Zebarjadian, et al., 1999). Thus even though the positions of modified rRNA nucleotides are known, their roles in ribosome function are still unclear. Modifications may serve to stabilize rRNA structure, yield unique sites for RNA-RNA or RNA-protein interactions, or improve the activity of rRNA within the assembled ribosome. Both types of modifications are known to change the folding properties of RNA by stabilizing local RNA structure. 2’-O-methylation was found to constrain sugar residues into the more rigid C3’- endo conformation (figure 5) which decreases steric interactions between the methyl group and the base (Davis, 1998). Methylation of the ribose also blocks an H-bonding potential and protects against hydrolysis (Bertrand and Fournier, 2004). Conversion to a pseudouridine allows for an additional hydrogen bond potential at the N-1 position, which may contribute to a more rigid RNA structure (Yu, et al., 1999). Methylated riboses and pseudouridines may also stabilize RNA by increasing base stacking (Durant, et al., 1999; Kawai, et al., 1992). Incorporation of these modifications has been shown to elevate melting temperatures of short RNAs (Zmudzka, et al., 1970; Durant, et al., 1999). The number of box C/D sRNA genes in extreme

12 thermophiles, correlates with the optimal growth temperature (Omer, et al., 2000; Noon, et al., 1998). Furthermore, the number of 2’-O-methylated nucleotides was shown to increase in the archaeon S. solfactarius 16S rRNA with higher growth temperatures (Noon, et al., 1998). Small nuclear RNAs involved in the spliceosome also contain many modified nucleotides. Most are concentrated in regions involved in intermolecular RNA-RNA interactions or in regions known to fold into alternative secondary structures (Massenet, et al., 1998). Notably, placement of a conserved pseudouridine in eukaryotic U2 snRNA across from the intronic AA dinucleotide has been shown to flip out the branch site A extrahelically and positioning it for nucleophilic attack (Newby and Greenbaum, 2001; 2002). Most evidence suggests modifications function as structural stabilizers in RNA. However, the act of modifying may be as important as the modification itself. The snoRNPs could affect RNA folding and/or RNP assembly, directly or indirectly, during the process of modification. It has been suggested that snoRNPs function as molecular chaperones in the complicated processes of ribosome folding and assembly (Maxwell and Fournier, 1995; Bachellerie, et al., 1998; Steitz and Tycowski, 1995). Evidence for this includes the U2 snRNP where modified nucleotides were shown to be required for U2 snRNP assembly (Yu, et al., 1998). The U14 snoRNP has also been implicated in rRNA folding by potentially binding two distant segments of 18S rRNA (Liang et al., 1995). Interaction with multiple complementary snoRNAs may alter the RNA structure and help coordinate spliceosome and ribosome organization to form functionally mature RNPs. Thus, investigation of the structure, assembly, and biochemistry of snoRNPs is important to understanding their roles in the modification and organization of distinct RNPs in the gene expression pathway.

Box C/D SnoRNPs

All snoRNAs are associated with specific proteins to produce a functional snoRNP. A eukaryotic box C/D snoRNP is comprised of a box C/D RNA with four core proteins. The proteins identified in human are Fibrillarin (nucleolar protein 1 or Nop1p in yeast), Nop58p, Nop56p, and 15.5kD (small nuclear protein 13 or snu13p in yeast). Archaeal

13 box C/D sRNPs are composed of a more simplified core of 3 proteins, Fibrillarin, Nop5p, and L7Ae, interacting with a box C/D RNA. The models of assembly for eukarya and archaea are shown in Figure 6. Box C/D RNAs All box C/D RNAs contain at least one set of the characteristic sequences box C and box D. The consensus sequences for C and D are RUGAUGA (where R is any purine) and CUGA, respectively. Many box C/D RNAs also contain a second pair of usually imperfect C/D sequences designated C’ and D’. These snoRNAs have terminal and internal stems that juxtapose box C with box D and box C’ with box D’, respectively (Yu, et al., 1999). The box C/D motif is predicted to fold into a distinctive helix-asymmetric loop-helix structure called a ‘kink-turn’ (Klein, et al., 2001). The conserved box elements are crucial for direct and indirect binding of specific box C/D snoRNP proteins (Peculis and Steitz, 1994; Baserga, et al., 1991), accumulation of snoRNAs (Weinstein and Steitz, 1999), and cellular localization and processing of snoRNAs (Maxwell and Fournier, 1995; Weinstein and Steitz, 1999). Each box C/D RNA also contains one or two guide elements positioned 5’ to the box D and/or D’ motifs. The guides consist of 9- 21 nucleotides capable of forming canonical basepairs with its rRNA target. The site of methylation is located specifically five nucleotides upstream of the D and/or D’ box on the target pre-ribosomal RNA. 15.5kD (L7Ae) The smallest protein that associates with a eukaryotic box C/D snoRNA is 15.5kD. This protein was first isolated from purified U4/U6·U5 tri-snRNP particles of the major spliceosome (Stevens and Abelson, 1999) and found to be associated with the 5’ stem of the U4 snRNA (Nottrott, et al. 1999). 15.5kD was predicted to be a box C/D snoRNP protein based on the sequence similarities of box C/D RNA and the 5’ stem of the U4 snRNA. 15.5kD was determined to bind to several yeast box C/D RNAs specifically and was also co-purified with the U3 box C/D snoRNP (Watkins, et al., 2000). Thus it seems 15.5kD has a dual function in RNA processing as a spliceosomal protein and a box C/D snoRNP protein. In archaea, the spliceosome does not exist and a direct homolog for 15.5kD was elusive until sequence analysis showed that the ribosomal protein L7Ae was60% homologous to 15.5kD (Rozhdestvensky, et al., 2003). L7Ae was determined to

14 15 be a box C/D snoRNP protein by in vitro methylation assays (Omer, et al., 2001) demonstrating a dual function for L7Ae as a ribosomal protein and a box C/D snoRNP protein in archaea. Furthermore, mutation of the proposed tandem GA pairs in box C/D RNA disrupted in vitro interactions between box C/D RNAs and the 15.5 kD/L7Ae protein ( Huang, et al., 1992; Kuhn, et al., 2002; Watkins, et al., 2002; Rashid, et al., 2003; Tran, et al., 2003) as well as in vivo functioning of box C/D RNAs ( Huang, et al., 1992; Samarsky and Fournier, 1998). Fibrillarin Fibrillarin was the first protein to be identified as a box C/D snoRNP protein. It has been shown to be essential for growth in yeast (Tollervey, et al. 1991) and a mutation that diminishes the levels of fibrillarin leads to impairment of pre-rRNA methylation (Tollervey, et al. 1993). Several studies have shown fibrillarin associated with box C/D RNA (Tyc and Steitz, 1989; Watkins, et, al., 1996) and a crystal structure of an archaeal fibrillarin homolog from Methanococcus jannaschii (MJ) (Wang, et al., 2000) revealed a classic Rossmann-fold similar to those in a number of S-adenosyl-L-methionine (AdoMet)-dependent methyltransferases (Wang, et al., 2000). A more recent crystal structure of a fibrillarin/Nop5p complex from AF also shows that a SAM molecule was attached at the active site in fibrillarin (Aittaleb, et al., 2003) Figure 7. Thus, evidence implicates fibrillarin as the catalytic subunit of box C/D snoRNPs. Archaeal homologs of fibrillarin do exist but many are smaller than their eukaryotic counterparts and lack the glycine arginine rich (GAR) domain. Nop58p and Nop56p (Nop5p) Nop58p and Nop56p were first discovered in 1997 in yeast using a screen for lethality in combination with mutant alleles of Fibrillarin (Gautier, et al., 1997). Nop58p was isolated independently in 1998 using a screen for nucleolar antigens and named Nop5p (Wu, et al., 1998). Yeast Nop58p and Nop56p share 45% amino acid sequence homology. In archaea, only one homolog of Nop58p/Nop56p has been identified, suggesting a gene duplication event since the divergence from archaea. Most archaeal Nop5p proteins are smaller than their eukaryotic counterparts. Both Nop58p and Nop56p are essential nucleolar proteins that associate with box C/D RNAs (Gautier, et al., 1997; Lafontaine and Tollervey, 1999; 2000). They belong to a family of proteins that include

17 the pre-mRNA splicing factor Prp31. Prp31 was found to be associated with the U4/U6·U5 trisnRNP by coimmunoprecipitation and was proposed to stabilize interactions in the tri-snRNP and the pre-spliceosome (Weidenhammer, et al., 1996). Thus, Nop58p/Nop56p may also function in stabilizing the box C/D snoRNP structure. Archaeal Assembly Formation of a mature and functional box C/D sRNP occurs as a sequence of events. The box C/D sequences are first bound specifically by L7Ae. This initial interaction has been shown by in-vitro binding assays to be mandatory for the subsequent association of fibrillarin and Nop5p proteins (Watkins, et al., 2000; Rashid, et al., 2003). This suggests that the box C/D RNA undergoes a conformational change upon binding of L7Ae and thus generates binding sites for the other proteins (Granneman, et al., 2002; Kuhn, et al, 2002). It is predicted that the other two core proteins, fibrillarin and Nop5p, bind the L7Ae-box C/D RNA complex as a preformed heteroprotein complex. A crystal structure of the fibrillarin-Nop5p complex from Archaeoglobus fulgidus (AF) shows fibrillarin interacts with the N-terminal domain of Nop5p to form a heterodimer (Aittaleb, et al., 2003). The heterodimers are then further associated to form a homodimer through interaction of the coiled-coil domains at the C-terminal end of the Nop5p proteins. This association of fibrillarin and Nop5p is also predicted for the eukaryotic homologs where fibrillarin is associated with Nop56p or Nop58p and the two Nop proteins interact using their C-terminal ends. A reconstituted complex of L7Ae, fibrillarin, Nop5p and a box C/D RNA has been shown to properly methylate (Kuhn, et al., 2002; Rozhdestvensky, et al., 2003; Rashid, et al., 2003) and thus forms the minimal functional box C/D sRNP. Investigation of snoRNPs is a current research topic of interest. Efforts have produced a growing list of snoRNAs, the identification of important RNA sequence elements and proteins involved, and the localization and trafficking of snoRNPs. Little information exists on the structural and biochemical aspects of snoRNP systems. Molecular details will not only provide information about the assembly and function of snoRNP systems, but may give insight into ribosome biogenesis as well. Knowledge of protein-protein, protein-RNA, and RNA-RNA interactions in box C/D snoRNPs may also provide global information about RNPs and their proposed functions. This work herein presents a structural and biochemical investigation of box C/D snoRNPs.

CHAPTER 2 CRYSTAL STRUCTURE AND MOLECULAR DETAILS OF A BOX C/D RNA-L7AE COMPLEX

Introduction

Kink-Turn Motif A recently recognized RNA secondary structure has been coined by Moore and Steitz as the kink turn motif (k-turn or kt) (Klein, et al., 2001). The box C and D sequences are predicted to form a kink turn structure. The motif consists of ∼15 nucleotides forming a two-stranded, helix-internal loop-helix structure (Klein, et al., 2001; Vidovic, et al., 2000, Winkler, et al., 2001). The internal loop is asymmetric and flanked by two sheared GA pairs on one side termed the Noncanonical stem (NCS) or Stem II and two GC pairs on the other side termed the Canonical stem (CS) or Stem I. The asymmetrical loop usually consists of 3 unpaired nucleotides on one strand with the nucleotide closest to the GA pairs flipped out of the loop and protruding into solution. This motif is characterized by a ‘kink’ in the phosphate backbone that causes a sharp turn (∼120°) in the RNA helix. The consensus sequence for this motif is shown in figure 9. The existing crystal structures of kink-turns show four principal surface features: a widened major groove, a flattened minor groove, a severely kinked backbone with a protruded nucleotide, and several exposed base planes (Klein, et al., 2001). Sequence comparison of box C/D RNA and U4 snRNA lead to the hypothesis that box C/D RNA forms a kink turn motif similar to that of U4 stem II in complex with 15.5kD ( Watkins, et al., 2000). The k-turn motif is common in many different types of RNA. It appears six times in the 23S rRNA of Haloarcula marismortue and two in the Thermus thermophilus 16S rRNA (Klein, et al., 2001). Five of the 23S rRNA kink-turns make significant contacts with nine ribosomal proteins including L7Ae, which interacts

19 with Kt 15. The motif is also found in the U4 snRNA and the L30e mRNA (Vidovic, et al., 2000; Mao, et al., 1999). Figure 8 shows some kink turn-protein complexes. Kink turns are also predicted for T and S boxes of bacterial antitermination systems (Winkler, et al, 2001), L10 mRNA, RNase MRP, and U3 box B/C snoRNA (Lafontaine and Tollervey, 1999, Watkins, et al., 2000; Klein, et al., 2001). Although proteins have not yet been implicated in the binding of every predicted kink-turn, surface features of the motif make it well suited to serve as a nucleation site for assembly of ribonucleoprotein particles. Specificity of L7Ae and Its Homologs L7Ae and its homologs have been implicated in the binding of several different kink turns or proposed kink turns: Kt 15 of the large ribosomal subunit (Ban, et al., 2000), U4 snRNA (Vidovic, et al., 2000), box B/C snoRNAs (Watkins, et al., 2000; 42), selenocysteine insertion sequence (SECIS) RNA (Allmang, et al., 2002), archaeal box H/ACA sRNAs (Rozhdestvensky, et al., 2003) and box C/D snoRNAs (Watkins, et al., 2000; Kuhn, et al., 2002; Szewczak, et al., 2002). Figure 9 shows the sequences for each of these RNAs. All the kink turn RNAs have similar but not identical primary structures that fall within the consensus sequence. All have two sheared GA pairs, except Kt 15 of the 23S rRNA which has a GUA base triple in lieu of the second GA pair. The protruded U is another common feature with the exception of the SECIS RNA which has an extended loop structure that does not follow the consensus sequence for K-turns, but the protein (SECIS binding protein or SBP) that binds to this RNA has the same RNA binding domain as L7Ae (Allmang, et al., 2002). The box H/ACA ends in a terminal loop unlike any of the other kink turn RNAs. All of these RNAs share a purine adjacent to the protruded uridine except for the SECIS RNA. Box C/D RNA, box B/C RNA, and SECIS RNA also have a conserved UU pair adjacent to the GA pairs in common. In Kt- 15 and U4 snRNA this UU pair is replaced by a GC pair and in box H/ACA it is replaced by a GU pair. The similar features of these kink turn RNAs suggest synonymous roles in the recognition of L7Ae and its homologs, while distinct features suggest variations in recognition. The difference in recognition is especially true for Kt-15 of the 23S rRNA which not only interacts with L7Ae, but also with L15e simultaneously (Klein, et al., 2001). This suggests that the interaction of L7Ae with box C/D RNA may be slightly

22 different than with Kt 15 of the ribosome. L7Ae was only recently implicated in archaeal box H/ACA sRNPs, which would be a third functional role in archaeal RNP assembly (Rozhdestvensky, et al., 2003). Multiple roles in the assembly of distinct RNPs suggest a regulatory function for L7Ae in coordinating multiple RNP biogenesis (Watkins, et al., 2000; Kuhn, et al., 2002), but it is unclear how L7Ae and homologs select a specific kink turn RNA during multiple RNP assembly. Following the initial interaction of the L7Ae with their cognate k-turn RNAs, accessory proteins bind to the L7Ae-RNA complexes to produce functional mature RNPs. These subsequently assembled proteins in different RNPs vary significantly in composition. For example, binding of 15.5 kD to U4/U6 snRNAs in human enables snRNP proteins such as the 61 kD protein and the 20/60/90 kD complex to bind to the U4/U6 snRNP (Nottrott, et al., 2002). Association of L7Ae with the 23S rRNA may enable the organization of other large subunit proteins such as L15e. In the assembly of box C/D sRNPs, binding of L7Ae to box C/D RNA initiates subsequent assembly of other sRNP proteins that include fibrillarin and Nop5p in archaea or Nop56/58p in eukarya (Watkins, et al., 2002; Cahill, et al., 2002; Rashid, et al., 2003; Tran, et al., 2003). The structure and dynamics of a particular k-turn RNA, combined with the diversified binding activities of L7Ae, may determine the unique tertiary structure of each core complex and therefore drives the specificity for subsequently assembled proteins in these RNPs. Thus, comparing the three-dimensional structures of different k- turn RNAs can provide insights on how L7Ae initiates and coordinates different assembly events. The only co-crystal structures reported previously are of 15.5 kD bound with the 5’-stem loop of U4 snRNA (Vidovic, et al., 2000) and L7Ae from Haloarcula marismortui (HM) bound with KT-15 of 23S rRNA (Ban, et al., 2000).

In this manuscript, the co-crystal structure of L7Ae and a box C/D RNA at 2.7 Å resolution and characterization of this interaction by solution NMR will be described. A recombinant form of the L7Ae from Archaeoglobus fulgidus (AF) and a synthetic 25mer RNA containing conserved box C/D sequences was used for crystallization. We have also determined a crystal structure of the complex of AF L7Ae bound with a mutant box C/D RNA where the UU mismatch pair was replaced by a GC Watson-Crick pair (U5G/U21C mutant) that occurs at the same position within the U4 snRNA. Structural

23 differences and similarities between the L7Ae-box C/D RNA complex and previously determined 15.5 kD/L7Ae-K-turn RNA complexes and a new L7Ae-box C’/D’-like terminal loop K-turn RNA complex (T. Hamma and A. Ferre-D'Amare, personal communication) will also be reported.

Results and Discussions

Structure Determination A number of RNA constructs containing the consensus box C/D sequences were used for crystallization of the AF L7Ae-box C/D RNA complex (Figure 10). L7Ae was combined with 23-25mer oligos containing a GAAA tetraloop that closed at the end of the CS. The constructs were tested for binding to L7Ae using electrophoretic mobility shift assays (Figure 11). All of the constructs were shown to bind L7Ae. RNA and L7Ae of varying molar ratios were combined and subjected to crystallization trials. Only the complex of L7Ae and a 25mer RNA at a 1:1.2 (protein:RNA) ratio yielded diffractable crystals. The structure was determined by the multiple wavelength anomalous diffraction (MAD) method from a three-wavelength data set collected on a single crystal containing three 5-Br-uridines (Figure 12). The final structure of the complex was refined to 2.7 Å with excellent crystallographic residual and stereochemistry values (Table 1). The electron density map around the box C/D RNA is shown in figure 13. Five mutant RNAs were also subjected to crystallization trials. These mutants are shown in Figure 10. Mutants 1 and 4 were based on sequences from the U4 snRNA and Kt 15 of 23S rRNA, respectively. Mutants 2, 3, and 5 were based on mutation data that did not abolish binding of L7Ae. Only the Mutant 1 (U5G/U21C) complex produced crystals. The Mutant 1-L7Ae complex crystallized in the same space group with nearly identical unit cell dimensions and its structure was refined to 3.2 Å by using the wild-type complex as the starting structure model. Arrangement of L7Ae-RNA Complex in the Crystal The cubic co-crystal of L7Ae and a box C/D RNA contains two complexes per asymmetric unit. The two L7Ae-RNA complexes in the asymmetric unit are related by a C2 symmetry, stacking head-to-head on the CSs of the two box C/D RNAs (Figure 14).

Table 1. Data Collection, Phasing and Refinement Statistics* Data Collection Statistics + † Data Sets Wave- Resolution Measured Unique Rsym / Comp- Diffraction Ratios length (Å) Reflections reflections (%) leteness <σ(I)> λ1 λ2 λ3 (Å) λ1 (edge) 0.9193 50.0-2.7 704959 16140 0.084 91.5 99.9 0.037 0.032 0.033 (2.8-2.7) (0.727) (7.8) (100.0) λ2 (peak) 0.9202 50.0-2.8 453789 14768 0.082 68.9 99.9 0.027 0.021 (2.9-2.8) (0.665) (7.2) (100.0) λ3 (remote) 0.9062 50.0-3.0 369199 12047 0.090 63.3 100.0 0.033 (3.1-3.0) (0.481) (9.9) (100.0) Refinement Statistics Phasing Statistics

¶ ¶ Resolution Rcryst Rfree #. of non- RMSD RMSD %Phi-Psi in core Wavelength λ1 λ2 λ3 (Å) (%) (%) hydrogen bond angle region (disallowed) atoms (Å) (°) Phasing Power$ 1.5 0.9 1.2 iso ‡ R Cullis 0.73 0.634 0.757 50-2.7 23.5 26.0 2851 0.007 2.1 A: 86.6% (0%) (2.9-2.7) (37.6) (39.6) B: 90.1% (0%) Figure of merits 0.67 2/1 2/1 * † 2 2 2 Values in parenthesis are those for the highest resolution shell. Diffraction ratios are fined as Δ F / F , where Δ F are taken between the current and

+ i i i the reference wavelength (λ ) for dispersive ratios and between matched Bijvoet pairs for anomalous ratios. R = I − I hkl)( / I , where I hkl)( is 3 sym ∑hkl hkl)( ∑hkl hkl)( ¶ the ith measured diffraction intensity and I is the mean of the intensity for the miller index (hkl). R = F (hkl) − F (hkl /) F (hkl) ; R hkl)( cryst ∑hkl obs calc ∑hkl obs free

2 2 $ th iφ = Rcryst for a test set of reflections (10%) not included in refinement. Phasing Power for the i wavelength = Fλ − Fλ P()φ Fλ e Δ+ FH − Fλ dφ , where P(φ ) i ref ∫ ( ref i ) φ

‡ iso is the experimental phase probability distribution. R cullis = F ± F − F / F − F . ∑hkl λi λref H ∑hkl λi λref

25 26

28 The RNA molecules are involved in all three of the major packing interactions between the two complexes. The GAAA tetraloop from one chain (x, y, z) interacts with the minor groove of the NCS in the symmetry mate (-x, -y+1, z), forming hydrogen bonds using the ribose hydroxyl groups of the A nucleotides. This is similar to the previously characterized A-minor interactions in group I intron RNA (Soukup et al., 2002) and ribosomal RNA structures (Nissen, et al., 2001). The total buried solvent accessible surface of this interaction is 514.9 Å2, comparable to a typical protein-protein crystal packing interface (Carugo and Argos, 1997). Another RNA-RNA mediated crystal packing interaction involves base stacking of six adenosine nucleotides from two GAAA tetraloops (Figure 14), resulting in a strong interaction through delocalized electrons (Shana, et al., 1999). The third packing interface involves protein-RNA interactions. The C- and the N-terminal helices of L7Ae rest in the minor groove of the continuous helix formed between the two RNA chains (Figure 14), resulting in a non-specific protein- RNA interface of a large buried solvent accessible surface (995.5 Å2). This packing surface is substantial in comparison to the 1232 Å2 total buried surface at the L7Ae-kink turn RNA interface itself (see below) which could contribute to non-specific RNA binding activity as previously observed (Kuhn, et al., 2002). However it’s implication in L7Ae function either as a ribosomal protein or sRNP protein is unclear. This analysis indicates use of both the GAAA tetraloop and the optimal oligo length is critical in obtaining the well-packed crystals. L7Ae Structure The AF L7Ae protein is a mixed α/β protein with a central four-stranded β-sheet sandwiched by a total of six α-helices on both sides (Figure 15). This structural fold has been observed in its homologs, 15.5 kD protein (Vidovic, et al., 2000), HM L7Ae (Ban, et al., 2000), and yeast ribosomal protein L30 (Mao, et al., 1999). Note that helix α3 in

AF L7Ae corresponds to the tighter 310 helix, η3, of the 15.5 kD structure (Vidovic, et al., 2000), hence AF L7Ae has six instead of five helices. The AF L7Ae structure superimposes well with each of these protein structures (Figure 15). The root mean square deviation (rmsd) values between Cα atoms of AF L7Ae and those of 15.5 kD, HM L7Ae, and the yeast L30 protein (RNA bound form) are 0.9 Å (112 Cα atoms), 0.67 Å (116 Cα atoms) and 1.47 Å (79 Cα atoms) respectively. Recently two other crystal

31 structures of L7Ae from Methanococcus jannaschii (MJ) were determined, one bound with a terminal stem loop K-turn RNA (T. Hamma & A. Ferre-D’Amare, personal communication) and one free of bound RNA (B. Brown II, personal communication). The rmsd values between AF L7Ae Cα atoms and those of the MJ L7Ae bound with the terminal loop K-turn RNA and those of the free MJ L7Ae are 0.65 Å (100 Cα atoms) and 0.61 Å (111 Cα atoms) respectively. The three-dimensional homology search server, DALI (http://www2.ebi.ac.uk/dali/), also matched the L7Ae structure with domain 3 of human eukaryotic release factor Erf1 (Song, et al., 2000) with a 1.6 Å RMSD for a total of 76 Cα atoms. This comparison showed that the L7Ae structure is well conserved and it undergoes minimal conformational changes upon association with K-turn RNAs. We also compared the surface electrostatic potentials of AF L7Ae, human 15.5kD, and HM L7Ae. The potentials were calculated by solving the nonlinear Poisson- Boltzmann equation. All three proteins exhibit an RNA-binding surface of mixed charge polarities consisting of a large region of neutral potential separating a small negatively charged patch and a large positively charged patch (Figure16). The distinct positive patch contrasts a large negative region on the opposite side of the proteins, resulting in a favorable dipole moment for interaction with the kink turn RNAs. This general electrostatic feature at the interface of all three L7Ae homologs suggests a conserved functional role of electrostatic force in recognition of kink turn motifs by these proteins.

Box C and D Sequences Form a K-turn Motif Box C and D nucleotides fold into the K-turn motif; a secondary structure first predicted by Watkins et al. based on their biochemical studies (Watkins, et al., 2000). The principle features of the K-turn motif include an asymmetric loop, a protruded nucleotide, two tandem GA pairs and a sharp bend in the phosphate backbone (Klein, et al., 2001). All conserved sequences in box C and D reside within the asymmetric loop and stem II regions of the K-turn (Figure17). The first U in box C (NRU18GAUGA) protrudes from the helix into solution. The next two downstream nucleotides, GA (NRUG19A20UGA), form two sheared GA base pairs with the last two nucleotides in box D (CUG6A7) respectively. The second U in box C (NRUGAU21GA) forms a UU mismatch with the U in box D (CU5GA) through a single imino-4-carbonyl hydrogen bond. The last G in box C (NRUGAUG22A) forms a Watson-Crick base-pair with the first C in box D (C4UGA)

34 and the last A in box C (NRUGAUGA23) also forms a Watson-Crick AU base pair in our structure. The sequences of the two bulged nucleotides preceding the protruded U in box C (N16R17UGAUGA) are not conserved; however they participate in stacking interactions with both stems I and II, contributing to formation of the distinct architecture of the K- turn. Notably, the unpaired purine nucleotide (NR17UGAUGA) adopts a syn conformation around its glycosidic bond in order to avoid clashing with the G nucleotide downstream (NRUG19AUGA) and to maximize a cross-strand adenine stack with the A nucleotide in box D (CUGA7) (see below). The same structural features for this nucleotide are also observed in the U4 and KT 15 RNAs. The phosphate backbone of the box C nucleotides is bent nearly 58 degrees, bringing the CS and NCS in close proximity, thus allowing a number of tertiary interactions between the two stems. Hydrogen bonding and base stacking interactions are the principle stabilization forces for the box C/D fold. Structural analysis performed by the 3DNA program (Lu, et al., 2000) showed that the nucleotides within the NCS have more extensive base-stacking interactions than those within CS, which is facilitated by the two tandem GA pairs and the unpaired purine nucleotide. As shown in Figure 18, a coaxial base stack consisting of U21, A20, A7, and G17 constitutes a central hydrophobic core for organizing the kink-turn RNA structure, a folding principle reminiscent of globular proteins. A network of hydrogen bonds formed between the centrally stacked nucleotides and those in the CS further stabilize the box C/D RNA fold (Figure 18). The A20 2’-hydroxyl group forms hydrogen bonds with C9 O4’ and G15 N2 and A7 N1 interacts with the O2’ of C16. In addition, there are five nucleotides, G6, A7, G17, U18, and G19 adopting the 2’-endo sugar puckering as opposed to the 3’-endo puckering seen in A-form RNAs. The transition from 3’-endo to 2’-endo provides maximum exposure of the major groove edges of these nucleotides for interacting with other bases and with L7Ae. Detailed Comparison of Three K-turns Bound to the L7Ae/15.5kD Protein These three cognate RNAs that bind to the L7Ae/15.5kD protein differ in their primary and secondary structures. In order to assess the effect of these differences on the tertiary structures of the three RNAs, we compared the tertiary structures of the protein- bound KT-15 and U4 5’-stem RNAs to that of box C/D RNA. Two methods were used to

35 36 align the common regions of the three RNAs. First, the sugar-phosphate backbone of the 15 nucleotides comprising the internal loops plus two base pairs in the canonical stem (Figure 19) was aligned (nucleotide 248 in KT-15 was excluded because it is part of the unique base-triple in KT-15). This alignment allowed us to directly compare the three K- turns without considering how they bind to the 15.5 kD/L7Ae protein. We found that the overall backbone root-mean-square differences (rmsds) are 1.2 Å and 1.9 Å between box C/D RNA and U4 5’ stem and KT-15 RNAs, respectively. Thus KT-15 RNA differs more in its overall tertiary structure than U4 5’ stem RNA from that of box C/D RNA, reflecting the difference in secondary structures between the KT-15 and the U4 5’ stem RNAs. To further characterize geometric differences among the K-turn RNAs, we defined the degree of bending (bending angle) for a K-turn RNA to be the angle between two vectors that originate from the phosphate atom of the protruded U and end at the phosphates of the fourth nucleotide either upstream or downstream of the protruded U. The bending angles for the box C/D RNA, KT-15, and U4 5’-stem RNA are 55º, 63º, and 45º respectively, indicating the U4 5’-stem RNA has the sharpest bending angle (Figure 19). The KT-15 RNA has the largest bending angle, which is likely due to the replacement of one GA pair by an A-U•G base triple. In order to assess how primary sequences influence the geometry of K-turns, we also determined a co-crystal structure of L7Ae bound with a mutant 25mer box C/D RNA fragment containing a GC pair (U5C/U21G). This mutation replaces the UU mismatch pair by a GC pair thus replicating the non-canonical stem (stem II) sequences of the U4 5’-stem RNA. The mutant RNA’s bending angle was 53º, a value not significantly different from that of the wild-type box C/D RNA. This suggests that L7Ae homologs may confer the shape of the RNA it binds making the RNA adapt to the protein. Bending angles of non crystallographic symmetry- related RNA copies of the wild-type box C/D RNA, of the U5C/U21G mutant, and of U4 RNA were measured to be 56º, 52º, and 48º respectively. Thus crystal packing appears to have minor effects on the bending angle and the observed differences seem to be an effect of individual complexes. However, it remains unclear whether these observed geometric differences in K-turn RNAs contribute to their binding specificities for different proteins.

The juxtaposed phosphate groups of the sharply bent backbone can develop a

37 38 strong electrostatic field in facilitating further protein assembly. We thus examined and compared the surface electrostatic potentials of the three K-turn RNAs bound to their cognate proteins. As shown in Figure 20, despite their subtle differences in bending angles, the three K-turn RNAs exhibited generally negative potentials on the surface near the sharply bent phosphate backbone, but away from 15.5 kD/L7Ae binding site. This electrostatic property could potentially attract other RNP proteins by exerting long-range interactions (Zacharias, et al., 1992; Fogolari, et al., 1997; Thomasson, et al., 1997; Tworowski and Safro, 2003), such that the 15.5 kD/L7Ae-K-turn complexes and other RNP proteins can form “close encounters” before the final binding. This interpretation is consistent with the fact that the sharply bent phosphate backbone is a major protein- binding site for five K-turns of the 23S rRNA (Klein, et al., 2001). The most interesting feature of the three RNAs is a positive patch on the back of each that interacts with a strictly conserved glutamate (Glu34 AF L7Ae) on bound proteins. In the second alignment method, we first aligned the Cα atoms of the bound 15.5 kD/L7Ae proteins in their respective protein-RNA complexes. We then applied the same transformation operations to the corresponding RNAs in order to compare the relative orientations of the K-turn RNAs with respect to their bound proteins. In this alignment, the overall backbone rmsds of the RNA molecules are 3.0 Å between box C/D and U4 5’- stem RNAs and 3.2 Å between box C/D and KT-15 RNAs. These alignment differences show that although box C/D RNA has an overall similar structure to the other K-turn RNAs, it differs in its orientation with respect to L7Ae than the other K-turns with respect to their cognate proteins. To further identify the regions in K-turn RNAs that are most different from one another with respect to their bound proteins, we computed averaged backbone rmsd as a function of residues and plotted these results in Figure 21 (in this calculation, the nucleotide 248 in the KT 15 involved in base triple was not included). As shown in Figure 21A, the three K-turn RNAs differ most in their stem regions, particularly in the CS. The best aligned region resides in the three nucleotides where the phosphate backbone kink occurs. In the box C/D consensus sequence, this corresponds to RUG nucleotides in box C (NRUGAUGA). This result highlights a common structure formed by the three nucleotides that is important for interacting with

40 41 the 15.5 kD/L7Ae protein. The same alignment procedure was also applied to the MJ L7Ae-terminal loop K-turn RNA complex (T. Hamma and A. Ferre-D’Amare, personal communication). We found that the C’/D’-like terminal loop K-turn RNA also superimposed well with the box C/D RNA except for the terminal loop that is absent in box C/D RNA (overall 0.6 Å rmsd using the sugar phosphate of the core nucleotides minus those in the canonical stem). The fact that the three K-turns superimpose well among themselves but deviate significantly when their bound proteins were aligned suggests that the K-turn moves as a single unit when it is bound to the homologous, yet different, proteins. This is illustrated in Figure 22 when the U4 5’-stem RNA and the box C/D RNA are compared when their bound proteins were aligned. The U4 5’ stem RNA structure is rotated more towards the loop connecting α1 and β1 than the box C/D RNA, leading to an average displacement of nearly 2.7 Å of both stems I and II. This displacement results in subtle differences in RNA-protein interactions between the box C/D and the U4 5’-stem RNAs as discussed below and may be explored further to determine whether they are specificity factors at the step of 15.5 kD/L7Ae binding and/or subsequent protein association. Protein-RNA Interactions in the L7Ae-box C/D RNA Complex L7Ae binds to the box C/D RNA structure around the severely bent phosphate backbone (figure 23), primarily contacting six RNA nucleotides, U5, G6, C16, G17, U18, and G19 (Figure 24, colored nucleotides), three of which (G17, U18 and G19) are those that have the most conserved positions with respect to the bound protein (see above section). Protein residues that are in close contact with box C/D RNA reside in α2, β1, and the loops joining β1-α2 and α4-β5. Because the overall mode of interaction has been previously described for the 15.5 kD-U4 RNA (Vidovic, et al., 2000) and the HM L7Ae- KT-15 (Klein, et al., 2000) complexes, we will focus on comparing features in box C/D RNA to those in the two previously known complexes. U18 (NRUAGUAG) protrudes from the center of the kink-turn into an electrostatically neutral surface on L7Ae, with its minor groove edge against the L7Ae surface (data not shown). Thus this nucleotide is primarily stabilized by hydrophobic interactions. Two hydrogen bonds are observed between U18 and L7Ae. The Nζ atom of Lys79 forms a hydrogen bond with the O4 atom in U18 and the carbonyl oxygen of

44 Glu54 forms a hydrogen bond with the N3 atom, which is specific for pyrimidines. Interestingly, Lys79 is not conserved among the L7Ae family of proteins. It is replaced by a glutamine in a number of L7Ae homologs including the HM L7Ae (sequence alignment not shown). This mode of interaction suggests that shape complementarity between L7Ae and U18 is the strongest specificity determinant for the protruded nucleotide. Therefore other pyrimidines can also be accommodated at this position. In agreement with this conclusion, gel mobility shift results showed that L7Ae bound equally well to the box C/D RNA when the protruded U was mutated to C (Kuhn, et al., 2002). However, this uridine is strictly conserved in box C of all archaeal box C/D RNAs, which raises the strong possibility that this nucleotide is required for additional interaction with other box C/D proteins, such as the Nop5p protein. The most highly conserved region of L7Ae is NExxK in α2, which interacts with the two guanosines forming the tandem GA pairs in a base-specific manner (Figure 24). Glu34 forms two hydrogen bonds with G19 (RUGAUGA), one with the N2 and one with the N1 atoms. The N2 atom of G19 also participates in formation of the sheared G19A7 base pair. The importance of the N2 hydrogen bonds was demonstrated experimentally in a nucleotide analog interference modification (NAIM) experiment where inosine substitution of this G nucleotide impaired snoRNP assembly (Szewczak, et al., 2002). In addition, the O6 atom of G19 is in close contact with two protein amide groups, those of Asn33 and Glu34. Different from what was observed in the U4 RNA-15.5 kD complex, the G43-equivalent nucleotide (G6 in box C/D RNA) does not form sequence-specific interactions with a conserved lysine residue (Lys44 in 15.5 kD and Lys37 in AF L7Ae). This is the result of the differing orientations that box C/D and U4 RNAs hold with respect to their bound proteins (Figure 22). However, the hydroxyl groups of G6 have close contact with residues in α2 (Figure 24). These observed specific interactions formed by the two G nucleotides in boxes C and D provide the structural basis for the previously observed mutagenesis data on the two G nucleotides (Watkins, et al., 1996; 2000; Kuhn, et al., 2002). The two A nucleotides of the GA pairs do not form base- specific interactions with L7Ae residues at all, yet their mutations also disrupted L7Ae- box C/D RNA interactions (Szewczak, et al.,2002; Kuhn, et al., 2002; Rashid, et al., 2003) This suggests that the base-pairing capability of A with G nucleotides is important

45 46 in maintaining RNA-protein interaction. Deletion of the two bulged nucleotides upstream of the protruding U resulted in reduced affinity of L7Ae for box C/D RNA (Kuhn, et al., 2002). These two nucleotides also participate in protein-K-turn RNA interaction. C16 (C16GUGAUGA) forms a hydrogen bond with Glu89 in the loop connecting α5 and β4 (Figure 22). G17 (CG17UGAUGA) forms hydrophobic interactions with a conserved nonpolar residue, Ile88, within the same L7Ae loop. The unique syn conformation of G17 is necessary as it enables G17 to extend the favorable hydrophobic interaction mediated by the co-axial purine stack within the K-turn to the conserved hydrophobic residue, Ile88, in L7Ae. This mode of protein-RNA interaction is consistent with the preference for a purine residue at this position in box C and is likely to be important in restricting the positioning of the K-turn with respect to L7Ae. The identity of this unpaired purine in box C/D motifs was shown experimentally to be important in maintaining binding affinities of Snu13p (yeast homolog of 15.5 kD) for the box C’/D and the box B/C motifs of a truncated U3 snoRNA (A42). Given the importance of the loop connecting α5 and β4 in interacting with the two unpaired nucleotides, it is surprising that the amino acids of this loop differ substantially between 15.5 kD and L7Ae (Figures 15 and 22). This difference could potentially contribute to the observed difference in U4 5’-stem RNA orientation with respect to the bound 15.5 kD protein. By inference, a difference in orientation would be expected in box C/D RNA bound to 15.5 kD rather than L7Ae. NMR Characterization of L7Ae-box C/D RNA Solution NMR was used to characterize the box C/D RNA and L7Ae interactions. A 21 mer box C/D RNA construct similar to the 25mer used for crystallization with two less GC base pairs was used for the experiments (Figure 10). The shorter construct was used to simplify the box C/D RNA NMR spectra. Changes to L7Ae and the box C/D 21mer RNA were monitored by comparing 1D spectra of the protein and RNA, bound and unbound. The spectra of L7Ae were collected using presaturation for water suppression. The imino region (9-15ppm) of the box C/D RNA spectra was observed to determine changes in the RNA structure due to the addition metals ions or L7Ae. This region of the NMR spectra contains resonances due to imino protons protected from exchange with the solvent due to base pairing (Figure 25) or protein interaction. 1D spectra for the RNA were obtained using the g11 echo pulse sequence which is a jump and return type of pulse sequence used for water suppression.

48 Comparison of the bound and unbound L7Ae shows little change in the protein structure (Figure 26). This is consistent with the alignment of the bound and unbound L7Ae structures seen in figure15. Changes in box C/D RNA upon complex formation with L7Ae were observed by monitoring the imino region of the NMR spectra as 15N- L7Ae was titrated into a 0.8mM 21mer box C/D RNA sample. The addition of L7Ae occurred in five steps to reach a 1:1 complex of RNA:protein. The spectrum of box C/D RNA alone shows three imino resonances (Figure 27), but as L7Ae is titrated into the box C/D RNA sample, more resonances appear and the initial imino resonances seem to have shifted. At the 1:1 molar ratio of RNA:L7Ae, there are at least 8 imino resonances and some shifted amide resonances from L7Ae in the 9-15ppm region. The shifted amide peaks were identified using a using a 15N-HSQC (data not shown). Two of the eight imino resonances can be accounted for in protein-RNA interactions. The other iminos are most likely due to basepairing within the RNA structure induced by complexation with L7Ae. The appearance and shifting of the imino resonances of box C/D RNA suggests a structural change in the RNA upon binding to L7Ae. This is in agreement with experiments showing that the L7Ae-box C/D RNA interaction is mandatory for the subsequent assembly of other sRNP proteins (Omer, et al., 2001; Rashid, et al., 2003). Thus, a conformational change in box C/D RNA caused by interaction with L7Ae may be necessary for Fibrillarin and Nop5p to assemble with the box C/D RNA to form a mature functional complex. Box C/D RNA was also titrated with monovalent and divalent metal ions to determine if the metal ions alone could change the conformation of box C/D RNA. A

0.8mM box C/D RNA sample was then titrated with both KCl and MgCl2 to final concentrations of 500mM and 50mM, respectively (data not shown). As stated above, the spectrum of the box C/D RNA alone has 3 imino peaks. Titration of the box C/D RNA with either metals did not result in any additional imino resonances or shifting of the original resonances. The spectra of the metal titrated box C/D RNA samples did not look similar to spectrum of the bound box C/D RNA, but looked very similar to the spectrum of the unbound RNA with the exception of peak broadening due to the high salt concentrations. There are at least eight imino resonances in the L7Ae bound box C/D RNA and only three peaks in the metal titrated box C/D RNAs. This suggests that metal

50 51 ions alone can not form and stabilize the box C/D RNA kink-turn structure. This is consistent with molecular dynamics simulations of the U4 snRNA kink turn demonstrating a significant degree of opening in the absence of 15.5kD (Cojocaru, et al., 2005). There have been several studies that show metal ions to be important for formation of a tightly kinked structure, although Goody et al., 2004 also demonstrated that there was a significant population of the more open structure even at relatively high concentrations of divalent metal ions (Matsumura, et al., 2003; Razga, et al., 2005). Thus formation and stabilization of a tightly kinked conformation may need protein factors such as L7Ae interaction with box C/D RNA.

Conclusion

In Archaea, the large ribosomal subunit protein L7Ae plays a key functional role in box C/D sRNP assembly separate from its function in protein biogenesis. It interacts specifically with the conserved sequences of box C/D RNA as the leading step in box C/D sRNP assembly. Our structural studies show that, upon binding of the L7Ae protein, the box C/D sequences common to all known small ribonucleolar RNAs form a K-turn motif characterized by a core co-axial purine stack and a sharply bent phosphate backbone. The refined crystal structure of the AF L7Ae-box C/D RNA complex shows in atomic detail the molecular mechanism by which L7Ae recognizes the box C/D region of a s(no)RNA. The molecular details revealed by the crystal structure of the complex illustrate the importance of box C/D sequences in ribosomal RNA modification and processing, as mutations within the box C/D K-turn abolish assembly of the particles in both Archaea and Eucarya. The analogous protein-RNA interface formed between 15.5 kD/L7Ae with four different K-turn RNAs (Kt 15 of 23S rRNA, U4 snRNA, box C/D RNA, and box C’/D’ like RNA) suggests a conformational “adaptability” of K-turn RNAs in binding to the 15.5 kD/L7Ae protein. The underlying differences in primary and secondary structures of the K-turns may lead to different tertiary structures and dynamic behavior in K-turn RNAs prior to protein binding. A study using eletrophoretic mobility and time-resolved

52 fluorescence resonance energy transfer (FRET) shows that a K-turn RNA in solution is polymorphic with a tightly kinked conformation and a less kinked form in equilibrium, even at high metal ion concentrations (Goody, et al., 2004). Association of a protein with its cognate K-turn could drive the RNA conformation towards that of tightly kinked structure, such as those observed in the crystals structures. Molecular dynamics simulations of several kink turn RNAs, including the U4 snRNA kink turn, show many regions of flexibility that may affect RNA dynamics (Cojocaru, et al., 2005; Razga, et al., 2005). Such dynamic process could potentially confer the binding specificity of the 15.5 kD/L7Ae protein for different K-turns. It is well recognized that conformational dynamics is an important specificity determinant in protein-RNA interactions (Frankel and Smith, 1998; Mittermaier, et al., 1999; Showalter and Hall, 2002). The L7Ae-box C/D RNA crystal structure provides a platform previously not available to rationalize subsequent assembly of other s(no)RNP proteins. Different L7Ae- RNA complexes are further associated with different proteins to form functionally distinct RNPs. The structural and dynamic differences of the final L7Ae-RNA complexes may determine the specificity of the subsequently assembled proteins. The immediate environment of L7Ae-KT-15 RNA complex within the large ribosomal subunit is now known (Ban et al., 2000). L15e is associated with the L7Ae-KT-15 complex at the sharply bent RNA backbone and together with L7Ae, forms a tight pocket for the protruded uridine ( Ban et al., 2000). Currently there are no co-crystal structures of the L7Ae-box C/D RNA complex with other sRNP proteins. Previous structural studies on fibrillarin-Nop5p complex alone and biochemical studies on sRNP assembly have unambiguously identified the COOH domain of Nop5p as the primary site responsible for binding to the box C/D RNA bound with L7Ae (Aittaleb, et al., 2003; Rashid, et al., 2003). There is no structural homology between the COOH domain of Nop5p and L15e. Thus it is expected that the fibrillarin-Nop5p complex interacts differently with the RNA-L7Ae complex than L15e does. The L7Ae-box C/D RNA complex as observed in our crystal structure presents a number of RNA sites which can potentially interact with Nop5p: the sharply bent phosphate backbone of box C, the widened major grooves of stems I and II, and the flattened minor groove. Proteins that can interact with each of these surfaces in a kink turn RNA have been observed in the

53 HM 50S ribosome crystal structure (Ban et al., 2000). Thus it is conceivable that the COOH domain of Nop5p interacts with the bent phosphate backbone of box C including the major groove edge of the protruded U and stem II. Evidence for this includes: 1) Mutation of the protruded U to C in a box C/D RNA did not influence its binding to L7Ae but reduced the binding affinity of the fibrillarin-Nop5p complex (our unpublished results); 2) Mutation of stem II nucleotides in U14 snoRNA abolished association of Nop58p/Nop56p (Watkins, et al., 2002). Being tightly associated with fibrillarin, Nop5p thus functions as a bridge between the catalytic subunit of the methylation enzyme and the target RNA base paired with the antisense sequence of box C/D RNAs. In eukaryotic cells, how do the Nop56/58p proteins and the spliceosomal protein, 61 kD, as well as the 20/60/90 kD complex associate with their cognate RNA-15.5 kD complexes? Interestingly, the 61 kD protein (or Prp31p in yeast) has ~30% sequence identity to Nop56/Nop58p. This suggests that box C/D snoRNPs may share the same core complex with the U4 snRNP, at least up to the level of associations of the Nop56p/58p and the 61 kD proteins. As crystal structures of more completely assembled RNPs become available, it will be possible for us to test these hypotheses on RNP assembly initiated by the association the 15.5 kD/L7Ae protein with K-turn RNAs.

Materials and Methods

Expression and purification of L7Ae Two slightly different L7Ae proteins were cloned. The first clone, L7Ae-his, consists of the AF gene encoding the L7Ae homologue and a polyhistidine tag at the N- terminus of the protein. The second clone, L7Ae-thm is similar except it contains a thrombin site after the polyhistidine tag for removal of the tag. Both clones are in the vector pET13b and the plasmids were transformed into E. coli BL21 (DE3) cells. Both proteins were purified using similar protocols. Cells were grown in LB and the harvested cells were sonicated in buffer A (25 mM sodium phosphate pH 7.4, 5% glycerol, 1 M KCl). The sonicated cells were then heated for 15 minutes at 70°C and centrifuged. The cell supernatant was then treated with polyethyleneimine (PEI) followed by ammonium sulfate precipitation to remove nonspecific RNAs. The cell pellet was collected and

54 redissolved in buffer B (buffer A with 5mM Imidazole) and heated at 37°C for 30 minutes. The supernatant was loaded onto a Qiagen Ni-NTA column equilibrated with buffer B followed by a washing step with buffer C (buffer A with 25mM Imidazole). The L7Ae proteins were then eluted using buffer D (buffer A with 300mM Imidazole). The L7Ae-thm protein required some additional steps here to cleave the polyshistidine tag off of the protein. The L7Ae-thm Ni pool was exchanged into a cleavage buffer (20mM Tris pH 7.0, 100mM KCl) and then subjected to cleavage using thrombin (Haematologic Technologies) at 25°C overnight. Both the L7Ae-his and L7Ae-thm proteins were then loaded onto a Superdex S200 gel-filtation column (Amersham Pharmacia) in buffer E (20 mM Tris–HCl (pH 7.5), 5% (v/v) glycerol, 300mM KCl, 5 mM β-mercaptoethanol,0.5 mM EDTA). 15N-labeled L7Ae for NMR experiments was purified similarly, but cells were grown in M9 minimal media enriched with 15N- Ammonium Chloride. Using these protocols pure L7Ae proteins, with and without the polyhistidine tag, were obtained free of nonspecific RNA and ribonucleases.

RNA Preparation The consensus box C/D RNA sequences were used as a model for RNA constructs. The constructs contained the minimal box C/D motif enclosed by a GAAA tetraloop. Box C/D 21mer, 23mer, and 25mer RNAs were purchased from Dharmacon Research Inc. Both RNAs were deprotected using a buffer consisting of 100 mM acetic acid pH 3.8 (adjusted using TEMED) and desalted using a C18 reversed phase cartridge (Waters) according to Dharmacon protocols. Oligos were then lyophilized for use in multiple experiments. Gel mobility-shift experiments RNA oligos were kinased using T4 Polynucleotide Kinase (PNK) (New England Biolabs) in the presence of γ-32P-ATP (MP Biomedicals) at 37°C for 1 hour. The reactions were then phenol/chloroform extracted and mini Sephadex G-25 columns (Amersham Biosciences) were used to remove excess γ-32P-ATP. The specific activities of the kinased RNAs were obtained by scintillation count. Trace amounts of 32-P-labelled RNA were incubated with L7Ae in a binding buffer (25 mM Tris pH 8.0, 2.5 mM MgCl2, 300 mM KCl, and 500 μM yeast tRNA) for 30 minutes at 70°C and 30 minutes at 37°C. Binding reactions were resolved on 8%

55 nondenaturing polyacrylamide gels in 0.5X TBE running buffer at room temperature. Gels were frozen at -80°C and exposed to a phosphorimager screen. The intensities of the free and shifted RNA were visualized using the Storm system phosphorimager (Amersham Biosciences). Some gel mobility experiments were also run using cold RNA (minus yeast tRNA in binding buffer) and visualized using ethidium bromide staining and a low wavelength UV lamp. Crystallization and Structure Determination The RNA constructs were screened for crystallization with AF L7Ae and only the box C/D 25mer formed cubic crystals. L7Ae and the 25mer of varying molar ratios (1:1.1 – 1:1.4) were mixed to yield a final overall concentration of 15 mg/ml. Crystals were grown at 30°C in a mother liquor consisting of 100 mM magnesium acetate, 100 mM HEPES pH 7.5, and 28%-32% PEG 400. Crystal complexes that contained three 5-Br- uridines or the U5C/U21G mutant were grown similarly. The crystals belong to the space group of P23 with a = b = c = 120.58 Å. Each asymmetric unit contains two L7Ae-25mer complexes, which corresponds to a 61% solvent content. The structure was determined by the multiple wavelength anomalous diffraction (MAD) data collected from a single crystal containing three 5-Br-uridines. The SOLVE program (A43) was used to identify the six Br sites (three from each 25mer) and calculate the initial experimental phases. The six Br sites identified by SOLVE clearly clustered into two sets of three, which allowed us to obtain a non-crystallographic symmetry operation matrix. The non-crystallographic symmetry operation was employed to improve the initial experimental phases through density modification and averaging by the program DM (Cowtan, 1994). The density modified electron density map showed clear connectivity for both the protein and the RNA molecules (Figure 13). Protein and RNA models were initially traced and manually modified with program XtalView ( McRee, 1999). Structure refinements for the Br-containing, wild-type, and the U5C/U21G mutant complexes were carried out with the program CNS (Brünger, et al., 1998). No water molecules were modeled because of insufficient resolution. Data collection, experimental phasing, and the final refinement statistics are shown in Table 1. Surface Electrostatic Potential Calculation

56 The surface electrostatic potentials were obtained with a novel and accurate numerical solution of the three-dimensional nonlinear Poisson-Boltzmann equation (PBE) (Boschitsch , et al., 2004). The solute (RNA) was represented by a set of atomic charges embedded in a low dielectric medium (dielectric constant=2 to mimic solute polarizability), surrounded by a solvent, which is represented by a high dielectric region (dielectric constant=80). Physiological salt conditions were chosen, with an ionic strength of 0.1 M NaCl. No ion exclusion region was added to the surface of the molecule. The solute/solvent interface was defined as the solvent excluded surface generated by rolling a probe sphere with a 1.4 Å radius over the van der Waals surface of the solute. Parse- like atomic radii (H=1.0, C=1.7, N=1.5, O=1.4, and P=1.9, in Å) were employed to define the molecular surface (Sitkoff, et al., 1994). The grid spacing at the finest level was 0.2 Å. The total grid size was at least four times the largest dimension of the solute to enforce minimal outer boundary condition errors and charge neutrality errors. Hydrogen atoms were added to each RNA structure and the all-atom Amber94 partial charges were assigned to the center of each atom (Cornell, et al., 1995). The net charge of the K-turn RNA structures is -17.35e, -18.54e and -17.47e, for the U4 snRNA KT-15 and the box C/D RNA, respectively. NMR Titration Experiments

The smallest box C/D RNA construct was used for NMR experiments. All 1D spectra were recorded on a 600 MHz Varian spectrometer. NMR samples were dissolved in a buffer consisting of 20 mM Sodium phosphate pH 7.0, 100 mM KCl, and 50 μM

NaN3. 10% D2O was also added to each 600μL sample. Spectra were obtained for a 0.8mM box C/D 21mer and 0.5mM 15N-L7Ae alone using presaturation to suppress the water signal in these spectra. A 0.8mM box C/D 21mer sample was titrated in multiple 15 steps with N-L7Ae, KCl, or MgCl2. L7Ae was titrated in four additions to reach a final 1:1 complex. KCl and MgCl2 were titrated in 15 additions to reach final concentrations of 500mM and 50mM, respectively. 1D NMR spectra were obtained for each of the titration steps using the g11echo pulse sequence, which is a jump and return type of experiment used for water suppression.

CHAPTER 3 BOX C’/D’ RNA-L7AE CRYSTALLIZATION AND PRELIMINARY X-RAY DIFFRACTION DATA

Introduction

The Box C’/D’ Sequences Box C/D RNAs can sometimes contain two sets of box C/D sequences. The second set of sequences, termed box C’ and box D’, are usually less conserved (Balakin , et al., 1996; Kiss-Laszlo, et al., 1998). Comparison of the C’ and D’ sequences reveal they are well conserved in archaea, but much less conserved in eukarya (Kiss-Laszlo, et al., 1998; Lowe and Eddy, 1999). The box C’/D’ sequences are located in an internal region of the snoRNA and are normally separated by 3 - 9 nucleotides in eukarya and 1 - 5 nucleotides in archaea resulting in a terminal loop in place of the CS (Kiss-Laszlo, et al., 1998; Omer, et al., 2000). Recent studies suggest the human 15.5 kD exhibits weaker affinity towards the box C’/D’ motif of snoRNAs. 15.5kD was shown to bind only the box C/D motif and not the box C’/D’ motif, while the archaeal L7Ae has been shown to bind both sets of motifs (Szewczak, et al., 2002; Tran, et al., 2003; Rashid, et al., 2003). It is unclear why the homologs would have different affinities for the degenerate box C’/D’ motif. L7Ae may have a less strict specificity allowing for the recognition of a wider variety of kink turns and may be able to accommodate replacement of the CS with a terminal loop. This is consistent with the fact that L7Ae is involved in three distinct RNPs: ribosome (Kt 15 of 23S rRNA), box C/D sRNP, and box H/ACA sRNP. Investigation of the L7Ae-box C’/D’ RNA complex may give insight into why there is a difference in the binding specificities of 15.5kD and L7Ae for the degenerate box C’/D’ sequences on a molecular level. Here the crystallization and characterization of an L7Ae- box C’/D’ RNA complex will be reported.

58 Results and Discussion

Crystallization Four box C’/D’ RNA constructs were used for possible crystallization of an L7Ae-box C’/D’ RNA complex (Figure 29). These constructs were based on the box C’/D’ region of the AF pre-tRNATrp (shown in chapter 4) and consisted of 20-26 nucleotides ending in a terminal loop box C’/D’ motif. EMSA was used to test binding of L7Ae with all four constructs (Figure 30). The four constructs were subjected to multiple crystallization trials with RNA and L7Ae of varying molar ratios. Table 2 shows screens that were used for the trials along with any conditions that produced crystals. The box C’/D’ 20mer gave one rod cluster that was not reproducible. Preliminary data collected for this crystal at Brookhaven National Laboratory was low resolution and an entire data set was not obtained. The box C’/D’ 25mer also gave plate crystals, but were also not reproducible. Preliminary X-ray data also showed that the crystals were twinned. The box C’/D’ 26mer gave two crystal forms, plates and cubes. The plate crystals were not reproducible and the diffraction pattern indicated twinning of the crystals. The L7Ae-box C’/D’ 26mer RNA cube crystals were reproducible. They were grown at 30°C (Figure 31) and were unfortunately unstable at room temperature. The crystals dissolved upon manipulation and gave streaky diffraction patterns. Additives shown in Table 3 were systematically added to the mother liquor in an effort to stabilize the crystal. Co2+, spermine, dioxane, and acetone seemed to stabilize the crystal the best, but the diffraction pattern was still streaky. The crystals were also not as large and did not diffract to the same resolution as the original crystal. Pictures of some of the crystals are shown in Figure 31. There still seems to be an inherent instability with the co-crystal. Growth at room temperature also does not solve the unstable crystal form. A different type of crystal form grows at room temperature and more of the crystals grow, but none of the room temperature crystals diffract even with long periods of soaking in different cryo protectants. The instability of the crystal may only be overcome by including a crosslinking agent after the crystal is grown such as 1-ethyl-3-(3-dimethyl aminopropyl) carbodiimide (EDC) or glutaraldehyde.

Table 2 Crystallization Trials for L7Ae-box C'/D' RNA Co-crystal Construct Screen Hits Crystal Condition Crystal rod cluster 0.1M KCl, 0.01M MgCL , Box C'/D' Screen (50), 2 (Not 0.05M Na Cac pH 6.0, 20mer Natrix Screen reproducible) 10% PEG 4000 (48) Crystal 0.01M MgCL , 0.05M Na Box C'/D' Screen (50), plates (Not 2 HEPES pH 7.0, 25% 25mer Natrix Screen reproducible) PEG MME 550 (48) 0.2M Li2SO4, 0.1M Bis- plates (Not Tris pH 6.5, 25% PEG reproducible) 3350 Box C'/D' Natrix Screen 0 24mer (48) Crystal 0.01M Mg(C H O ) , Box C'/D' Screen (50), plates (Not 2 3 2 2 0.05M Na Cac pH 6.0, 26mer Natrix Screen reproducible) 1.7M (NH )SO (48) 4 4 0.01M Mg(C H O ) , cubes 2 3 2 2 0.05M Na Cac pH 6.5, (Reproducible) 15% PEG 400

Table 3 Optimization of L7Ae-Box C'/D' 26mer Crystal

Metal Organic or Precipitant Additives Additives Detergents Buffers Polyamines 1-10 mM Na Cac Kit 1-4mM CaCl2 1-5% Hexandiol C12E9, C12E8 pH 5.1-7.4 spermine n-Dodecyl-ß-D- 1-10 mM maltoside, n-Decyl- 1-4mM Ca(C2H3O2)2 1-5% MPD ß-D-maltoside Na HEPES spermidine Sucrose 10-30 mM monolaurate, n- MgCl2 1-5% Dioxane Octanoylsucrose TrisHCl CYMAL®-6, 1-10 mM CYMAL®-5, Mg(C2H3O2)2 1-5% Acetone CYMAL®-3 1-10 mM Zn(C2H3O2)2 1-5% Ethanol TRITON® X-100 30-100mM KCl 3-5% PEG 4000 CTAB 1-10 mM CoCl2 3-5% PEG 8000 Deoxy BigChap 1-10 mM SrCl2 3-5% PEG3350 LDAO, DDAO ZWITTERGENT® 3-12, 1-10 mM ZWITTERGENT® BaCl2 3-5% PEG MME 550 3-10 Nonyl-ß-D- glucoside, n-Octyl- 1-10 mM ß-D-glucoside, n- CsCl2 3-5% Isopropanol Hexyl-ß-D-glucoside 1-s-Octyl-ß-D- thioglucoside, Heptyl-ß-D- 20-50mM KF thioglucoside HECAMEG C-HEGA-10, MEGA-8

63 Characterization by NMR Box C’/D’ RNA was characterized by NMR bound and unbound with L7Ae. The 20mer box C’/D’ RNA was used for NMR experiments. Changes in the box C’/D’ 20mer RNA were observed by following changes in the imino spectra of the RNA upon complex formation with L7Ae. The imino region as mentioned earlier shows imino protons that are protected from exchange due to interaction with the protein or formation of base pairs involving the imino protons (Figure 25). 1D spectra were obtained similarly to those obtained for the box C/D RNA. 1D spectra of the box C’/D’ 20mer RNA alone suggests the RNA is already structured without L7Ae. The spectrum of the unbound RNA shows about 8 sharp imino resonances. These iminos become less sharp and a few are shifted as more protein is added to the box C’/D’ 20mer RNA sample. The addition of 15N-L7Ae also seems to make two imino peaks disappear (denoted by a red star in figure 32). Changes in the imino resonances suggests that some change to the box C’/D’ RNA occurs with the addition of L7Ae, albeit minimal changes when compared to the structural changes in box C/D RNA upon interaction with L7Ae.

Conclusion

In Archaea, the box C’/D’ motif has nearly perfect consensus sequence of the box C/D motif, but its CS is replaced by a terminal (stem) loop of variable length (Omer et al., 2000). The fact that the protein-RNA interface of the AF L7Ae-box C/D RNA superimposes well with that of MJ L7Ae-box C’/D’-like terminal loop K-turn RNA (T. Hamma and A. Ferre-D'Amare, personal communication) suggests that the terminal loop has minimal effect on L7Ae-K-turn RNA interaction. This is consistent with a symmetric assembly model of L7Ae with box C/D sRNAs. On the contrary, the box C’/D’ motif in eukaryotic snoRNAs contains imperfect box C/D sequences in addition to similar structural difference in the CS as that in sRNAs (Kiss-Laszlo et al., 1998; Lowe and Eddy, 1999). In light of the important structural roles played by box C/D sequences in maintaining the K-turn structure for interacting with L7Ae, deviations in box C’/D’ led to reduced affinity of Snu13p for the two K-turn motifs in the truncated, but

65 functional, U3 snoRNA (Marmier-Gourrier et al., 2003). In the gel mobility shift and nucleotide analog interference modification (NAIM) study where 15.5 kD was found to preferentially associate with box C/D motif, the composite snoRNA contained a cytidine, instead of a purine, immediately 5’ to the protruded uridine within its box C’ (Szewczak et al., 2002). Thus this unpaired nucleotide in boxes C and C’ may have a more important role in protein-K-turn recognition than previously appreciated. The NMR experiments reported here also suggest that the box C’/D’ motif may be more structured than the box C/D motif alone and thus may not need L7Ae to stabilize the kink turn structure for subsequent binding of Fibrillarin and Nop5p. This is consistent with Szweczak, et al., 2002 where only one 15.5kD protein was needed for subsequent association of other box C/D snoRNP proteins. Differences in the L7Ae and 15.5kD proteins may also be a determinant for the decreased specificity of 15.5kD for the box C’/D’ sequences. As shown Figure 22, the loop between α5-β4 is significantly different in the two homologous proteins even though this loop makes significant contacts in both the U4 snRNA-15.5kD and box C/D RNA-L7Ae structures. Changes in this loop may alter the specificity of the homologous proteins for the same RNA. A systematic investigation on the functional roles by each of the box C/D consensus sequences, together with that of the key amino acids of 15.5 kD/L7Ae involved in RNA recognition, is necessary to understand the origin of 15.5 kD/L7Ae specificity for K-turn RNAs.

Materials and Methods

Expression and purification of L7Ae L7Ae was purified as described in chapter 2 of this manuscript. RNA Preparation All box C’/D’ RNA constructs were purchased from Dharmacon Research, Inc. and treated as described in chapter 2. Gel mobility-shift experiments EMSA experiments were performed similarly to those reported for the box C/D RNA constructs. Crystallization and Optimization

66 Crystal setups were performed using hanging vapor drops. The Crystal Screen or Natrix Screen consisting of 50 conditions or 48 conditions was used for multiple screenings of each construct. Both screens were purchased from Hampton Research. Trials were performed at room temperature and 30°C for all the constructs. Different molar ratios and final concentrations of the complex were screened. Trials consisted of adding 1.5ul of RNA-protein mixture with 1.5ul of the well condition. Optimization was slightly different where 1.0ul of RNA-protein mixture was mixed with 0.5ul of an additive and 1.5ul of the well condition. Addition of polyamines was directly to the RNA-protein mixture before mixing with well conditions. Crystal trays were then incubated at room temperature or 30° for several days before viewing under a microscope. Trays were viewed on a periodic basis, usually at 3 days, 1 week, 2 weeks, 1 month, and every month after that. NMR Experiments NMR data was collected similar to the titration of box C/D RNA as described in chapter 2. The smallest box C’/D’ RNA construct, the box C’/D’ 20mer, was used for the titration experiments.

CHAPTER 4 DESIGN AND CRYSTALLIZATION OF AN ENTIRE BOX C/D SNORNP

Introduction

2’-O-methylation of ribosomal RNA in eukaryotes and archaea are catalyzed by box C/D snoRNPs (sRNPs in archaea). In archaea, the minimal functional box C/D sRNP consists of a box C/D RNA and three proteins, L7Ae, Nop5p, and Fibrillarin. Although Fibrillarin is the required methyltransferase, the modification reaction is directed by the box C/D RNA. The box sequences have been shown to be important for protein binding and the antisense regions determines the reaction site on the target RNA. The stepwise assembly of box C/D sRNPs is central to the function of the holoenzyme. The assembly includes interaction of L7Ae initially followed by the subsequent association of Nop5p and Fibrillarin around a box C/D RNA that perfectly positions the target RNA and methyltransferase for action. Although recent crystal structures have been determined for the Nop5p-Fibrillarin (Aittaleb, et al., 2003) and L7Ae-box C/D RNA complexes (chapter 2), little is known about the holoenzyme as a whole and how the two previously determined structures interact. This chapter herein presents the crystallization study of the entire box C/D sRNP complex using the AF pre-tRNATrp as a model for RNA constructs in an effort to obtain its crystal structure.

Results and Discussion

RNA Design There are a total of 10 constructs used for possible crystallization of the entire box C/D complex or supercomplex. The three native constructs are shown in figure 33 and

include pre-tRNATrp, Intron II, and Intron I. The pre-tRNATrp from AF is an archaeal box C/D RNA that contains antisense guides with complementarity to the tRNA itself. The pre-tRNATrp construct includes the tRNA, bulge-helix-bulge (BHB), and box C/D domains. The Intron II construct includes the BHB and box C/D domains, but excludes the tRNA domain. Intron II is the smallest of the native constructs and only includes the box C/D RNA portion of the pre-tRNATrp. Two target strands were also designed according to basepairing prediction to the target sites (Omer, et al., 2000). There are also seven oligomer constructs shown in figure 34. Two of the oligomer constructs (S1 and S2) are two stranded box C/D RNAs that contain double helices in the guide region in an effort to stabilize the floppy guide regions. Four of the constructs (S3, S4, S5, and S6) are 3 stranded box C/D RNAS which use a target strand to stabilize the guide RNA. Construct S3 did not anneal properly and the subsequent 3 stranded constructs were subjected to structure analysis using Mfold (Zuker, 2003). Construts S5 and S6 are variations of S4 where S5 has an adenosine overhang and S6 has an extra GC pair. The S7 construct was designed to mimic circular box C/D RNAs found to exist in vivo from the Pyrococcus furiosus genome (Starostina, et al., 2004). Gel mobility-shift experiments Gel shift experiments were used to determine if each of the box C/D RNA constructs formed interactions with L7Ae and the Nop5p/fibrillarin complex. Some of these gel shift experiments are shown in figure 35. All three of the native constructs were shown to bind L7Ae and Nop5p/fibrillarin. Most of the oligomer complexes were also shown to bind to the three box C/D sRNP proteins. The first two oligomer constructs were able to bind to all three proteins despite the double helix in the guide regions. Only the gel for construct S2 is shown, since both constructs are similar. The straight helix in constructs S1 and S2 did not seem to affect the binding ability of L7Ae and the fibrillarin/Nop5p complex. Construct S3 was not able to bind either any of the proteins. This was due to the fact that this construct did not form the correct RNA structure for interaction with the box C/D sRNP proteins. Construct S4 was shown to interact with L7Ae and Nop5p/fibrillarin. Similar results were seen for constructs S5 and S6, but are not shown since the constructs are similar to S4.

Crystallization All ten constructs were subjected to multiple crystallization trials using screens shown in Table 4. The screens that were used were the Crystal Screen, Natrix Screen, Index Screen, and Peg Ion Screen. Different ratios for RNA:L7Ae:Nop5p/Fibrillarin were used in multiple screenings. There were many conditions that produced crystals using the pre-tRNATrp, Intron I, S1, S2, S4, S5, and S6 constructs. Most of these crystals did not grow immediately, but over longer periods of time consisting of 2 months to 1 year after the initial setup. This made regrowth of the same crystals difficult because of time constraints or unavailability of the same RNA and protein preps. The most heavily pursued crystal conditions were for constructs S4, S5, and S6. The optimization of several crystal conditions for these three constructs are shown in Table 5. Most of the conditions did not produce multiple crystals making it extremely difficult to determine the composition of the crystals. Only the S4 construct produced multiple diamond shaped crystals that allowed for the determination of the crystal composition. These crystals are shown in figure 36. Several crystals were harvested and washed extensively for SDS-PAGE which showed that all three box C/D sRNP proteins were present in the co-crystal. It was also assumed the box C/D RNA was present in the crystals since no interaction between L7Ae and the Nop5p/fibrillarin complex has been demonstrated. These crystals grew in about 4 months time and were not reproducible as of yet even with trial optimization of the crystal conditions. These crystals are also very hardy and stay intact even with vigorous manipulation, but they unfortunately do not diffract even with dehydration of the crystal by soaking in high salt concentrations overnight. This is common for co-crystals of large complexes. More investigation is needed to determine the methods for ordering the co-crystal for diffraction. There are also two other crystal types with the S5 and S6 constructs that have been heavily pursued with optimization additives shown in Table 5. Unfortunately no reproducible results have been obtained so far, but the options for optimization for the S4, S5, and S6 complexes have not been exhausted.

Table 4 Crystallization Trials for Box C/D sRNP Supercomplex Construct Screen Hits Crystal Condition

Crystal Screen (50), 0.2M KCl, 0.1M MgCl2, Natrix Screen (48), Cubes- Not 0.05M Na Cac pH 6.5, Pre-tRNATrp Index Screen (96) Reproducible 10% PEG 4000

Crystal Screen (50), 0.2M NH4C2H3O2, Natrix Screen (48), Cube Cluster 0.1M Tris HCL pH 7.5, Intron I Index Screen (96) - Not Pursued 45% MPD

0.05M NH4C2H3O2, 0.1M Bis-Tris pH 6.5, Messed up 30% pentaerythritol Cube - Not ethoxylate (15/4 Pursued EO/OH) Crystal Screen (50), Natrix Screen (48), Intron II Index Screen (96)

Plate Cluster - 0.05M MgCl2, 0.05M Crystal Screen (50), Not Na HEPES pH 7.0, Construct S1 Natrix Screen (48) Reproducible 15% PEG MME 550 Crystal Screen (50), Teardrop - Natrix Screen (48), Not 0.5M Li2SO4, 15% PEG Construct S2 PEG Ion Screen (48) Reproducible 8000

0.2M KCl, 0.1M MgCl2, Cube - Not 0.05M Na Cac pH 6.5, Pursued 10% PEG 4000 Crystal Screen (50), Construct S3 Natrix Screen (48) Crystal Screen (50), Natrix Screen (48), 0.01M MgCl2, 0.05M Index Screen (96), Tris HCl pH 7.5, 1.8M Construct S4 PEG Ion Screen (48) Diamonds (NH4)2SO4

0.05M CaCl2, 0.1M Bis- Cubes- Not Tris pH 6.5, 30% PEG Pursued 550 MME Cubes - Not Pursued 30% PEG 1500

0.2M NH4C2H3O2, Crystal Screen (50), 0.02M MgCl2, 0.05M Natrix Screen (48), Na HEPES pH 7.0, Construct S5 PEG Ion Screen (48) Small plates 10% PEG 8000

Crystal Screen (50), 0.025M MgCl2, 0.010 Natrix Screen (48), NaCac pH 6.5, 15% Construct S6 PEG Ion Screen (48) Small plates PEG 3350 Crystal Screen (50), Natrix Screen (48), Construct S7 Index Screen (96)

Table 6 Optimization of Supercomplex Crystal with Constructs S4, S5, and S6 Organic or Metal Additives Precipitant Additives 50-100mM KF Dioxane

1-10mM CdCl2 MPD

1-10mM CdI2 PEG 3350

1-10mM CaCl2 PEG 8000 10-100mM KI 10-50mM Kthiocyanate

1-10mM Mg(NO3)2

1-10mM CoCl2

1-10mM NiCI2

1-10mM CuSo4

1-10mM MnCl2

1-10mM Sr(NO3)2

1-10mM BaCl2

1-10mM CrCl3

1-10mM CsCl2

1-10mM Zn(C2H3O2)2

Conclusion The best leads so far are the S4, S5, and S6 constructs. More optimization options such as detergents, organic solvents, and precipitants could be investigated. Also more trials need to be repeated with construct 7 and different poly linkers in the middle.

Materials and Methods

Expression and purification of AF proteins Genes of AF Nop5p and Fibrillarin were cloned as an operon into the pET13b vector for co-purification. The Nop5p protein contains an N-terminal his-tag. Harvested cells grown in LB were sonicated in buffer A (25 mM sodium phosphate pH 7.4, 5% glycerol, 1 M KCl). Cells were then heated at 15 minutes at 70°C and centrifuged. The resulting supernatant was loaded onto a Qiagen Ni-NTA column equilibrated with buffer B (buffer A with 5mM Imidazole) followed by a washing step with buffer C (buffer A with 25mM Imidazole). The Fibrillarin-Nop5p complex was eluted using buffer D (buffer A with 300mM Imidazole). The Ni pool was then FPLC purified using a Superdex S200 gel filtation column (Amersham Pharmacia) in buffer E (20 mM Tris– HCl (pH 7.5), 5% (v/v) glycerol, 1.0M KCl, 5 mM β-mercaptoethanol,0.5 mM EDTA). Purification of AF L7Ae was described in the materials and methods section of chapter 2 in this manuscript. RNA Preparation Sequences for the native RNA constructs were cloned into pUC18 vectors individually. The T7 promoter and Sma I or Hind III sites were also cloned into the pUC18 vector for each construct. DNA for each construct was obtained using plasmid preps from Qiagen and linearized with the appropriate restriction enzyme (NEB). The linearized plasmid DNA was then used for in vitro transcription using T7 RNA polymerase and standard conditions. Large transcription preps were then performed and the resulting RNA was gel purified and ethanol precipitated. This protocol produced clean RNA for crystallization. Oligomer construct RNAs were obtained from Dharmacon and processed as described in chapter 2.

Gel mobility-shift experiments Native constructs for gel shift experiments were prepared by in vitro transcription in the presence of α-32P-CTP (MP Biomedicals) and then gel purified. Synthetic oligo construct RNAs were prepared as previously described in chapter 2. Gel shift experiments with both the native and oligomer constructs were performed similarly to those with L7Ae except the Nop5p-Fibrillarin complex was also added to the binding reactions. Crystallization There were four major types of screens used for crystallization trials for each of the constructs: Crystal Screen (50), Natrix Screen (48), Index Screen (96), and Peg Ion Screen (48) (Hampton Research). The trials were performed using hanging vapor diffusion drops. Trials were performed at room temperature and 30°C for all the constructs. The crystal mix was put together by adding different molar ratios of box C/D RNA:L7Ae:Nop5p/fibrillarin. Citrate salts were also sometimes added to the crystal mix to keep Nop5p/fibrillarin soluble. Crystal trays were set up and viewed similarly to those described in the chapter 3.

CHAPTER 5 CONCLUSION

Box C/D snoRNPs catylze the specific 2’O-methylation of rRNA in important regions the ribosome, although the role of the modifications is unclear. Dissecting the structure and biochemical properties of box C/D snoRNPs may not only help in understanding the function of the modifications, but may also give insight into the role of kink turn RNAs in RNP assembly. Our structural studies show that, upon binding of the L7Ae protein, the box C/D sequences common to all known small ribonucleolar RNAs form a K-turn motif characterized by a core co-axial purine stack and a sharply bent phosphate backbone. The analogous protein-RNA interface formed between 15.5 kD/L7Ae with four different K-turn RNAs (Kt 15 of 23S rRNA, U4 snRNA, box C/D RNA, and box C’/D’ like RNA) suggests a conformational “adaptability” of K-turn RNAs in binding to the 15.5 kD/L7Ae protein. The underlying differences in primary and secondary structures of the K-turns may lead to different tertiary structures and dynamic behavior in K-turn RNAs prior to protein binding and thus confer the binding specificity of the15.5 kD/L7Ae protein for different K-turns. NMR experiments suggest that the box C’/D’ motif may be more structured than the box C/D motif alone and thus may not need L7Ae to stabilize the kink turn structure. Differences in the L7Ae and 15.5kD proteins such as the α5-β4 loop, may also be a determinant for the decreased specificity of 15.5kD for the box C’/D’ sequences. The L7Ae-box C/D RNA crystal structure provides a platform previously not available to rationalize subsequent assembly of other s(no)RNP proteins. The structural and dynamic differences of the final L7Ae- RNA complexes may determine the specificity of the subsequently assembled proteins. The L7Ae-box C/D RNA complex as observed in our crystal structure presents a number of RNA sites which can potentially interact with Nop5p: the sharply bent phosphate backbone of box C, the widened major grooves of stems I and II, and the flattened minor

80 groove. In eukaryotic cells, how do the Nop56/58p proteins and the spliceosomal protein, 61 kD, as well as the 20/60/90 kD complex associate with their cognate RNA-15.5 kD complexes? Interestingly, the 61 kD protein (or Prp31p in yeast) has ~30% sequence identity to Nop56/Nop58p. This suggests that box C/D snoRNPs may share the same core complex with the U4 snRNP, at least up to the level of associations of the Nop56p/58p and the 61 kD proteins. As crystal structures of more completely assembled RNPs become available, it will be possible for us to test these hypotheses on RNP assembly initiated by the association the 15.5 kD/L7Ae protein with K-turn RNAs.

REFERENCES

Aittaleb M, Rashid R, Chen Q, Palmer JR, Daniels CJ, Li H. (2003) Structure and function of archaeal box C/D sRNP core proteins. Nat Struct Biol 10. 256-63. Allmang, C., Carbon, P., and Krol, A. (2002). The SBP2 and 15.5kD/Snu13p proteins share the same RNA binding domain: Idenitfication of SBP2 amino acids important to SECIS RNA binding. RNA 8, 1308-1318. Bachellerie JP, Nicoloso M, Qu LH, Michot B, Caizergues-Ferrer M, Cavaille J, Renalier MH. (1995) Novel intron-encoded small nucleolar RNAs with long sequence complementarities to mature rRNAs involved in ribosome biogenesis. Biochem Cell Biol 73, 835-43. Balakin AG, Smith L, Fournier MJ. (1996) The RNA world of the nucleolus: two major families of small RNAs defined by different box elements with related functions. Ban, N., Nissen, P., Hansen, J., Moore, P. B., and Steitz, T. A. (2000) The complete atomic structure of the large ribosomal subunit at 2.4 Å resolution. Science 280, 905-920. Baserga, S. J, Yang, X. D., and Steitz, J. A. (1991) An intact Box C sequence in the u3 snRNA is required for binding of fibrillarin, the protein component to the major family of nucleolar snRNPs. EMBO J 10, 2645-51. Batey RT, Rambo RP, Lucast L, Rha B, Doudna JA. (2000) Crystal structure of the ribonucleoprotein core of the signal recognition particle. Science 287, 1232-9. Bertrand, E. and Fournier, M. J. (2004) The snoRNPs and Related Machines: Ancient Devices that Mediate Maturation of rRNA and Other RNAs. The Nucleolus by Mark O.J. Olson. Kluwer Academic / Plenum Publishers. Boschitsch, A. H., and Fenley, M. O. (2004) A Hybrid Boundary Element and Finite Diffeence Method for Solving the Nonlinear Poisson-Boltzmann Equation. J. Comput. Chem. In press. Brünger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse-Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., and Warren, G. L. (1998). Crystallography & NMR system: A new software suite for macromolecular structure determination. Acta Crystallogr D Biol Crystallogr 54, 905-21. Cahill NM, Friend K, Speckmann W, Li ZH, Terns RM, Terns MP, Steitz JA. (2002) Site-specific cross-linking analyses reveal an asymmetric protein distribution for a box C/D snoRNP. EMBO J 21, 3816-28. Carugo O, Argos P. (1997) Protein-protein crystal-packing contacts. Protein Sci. 6, 2261- 3 Cavaille, J., Buiting, K., Kiefmann, M., Lalande, M., Brannan, C. I., Horsthemke, B., Bachellerie, J. P., Brosius, J., and Huttenhoffer, A. (2000) Identification of brain- specific and imprinted small nucleolar RNA genes exhibiting an unusual genomic orgainization. Proc Natl Acad Sci U S A 97, 14311-4.

82 Clouet d'Orval, B., Bortolin, M. L., Gaspin, C., and Bachellerie, J. P. (2001) Box C/D RNA guides for the ribose methylation of archaeal tRNAs. The tRNATrp intron guides the formation of two ribose-methylated nucleosides in the mature tRNATrp. Cojocaru V, Nottrott S, Klement R, Jovin TM. (2005) The snRNP 15.5K protein folds its cognate K-turn RNA: a combined theoretical and biochemical study. RNA 11:197-209. Cornell, W. D., Cieplak, P., Bayly, C. I., Gould, I. R., Merz, K. M., Ferguson, D. M., Spellmeyer, D. C., Fox, T., Caldwell, J. W., and Kollman, P. A. (1995). A second generation force field for the simulation of proteins, nucleic acids, and organic molecules. Journal of American Chemical Society 117, 5179-5197. Cowtan, K. (1994). Joint CCP4 and ESF-EACBM Newsletter on Protein Crystallography 31, 34-38. Darzacq X, Jady BE, Verheggen C, Kiss AM, Bertrand E, Kiss T. (2002) Cajal body- specific small nuclear RNAs: a novel class of 2'-O-methylation and pseudouridylation guide RNAs. EMBO J. 2002 Jun 3;21(11):2746-56. Davis D.R. Biophysical and conformational properties of modified nucleosides in RNA (nuclear magnetic resonance studies). In: Grosjean, H. and Benne, R. eds. (1998) Modification and Editing of RNA. Washington, DC, ASM Press, 85-102 Dennis, P. P., Omer, A., and Lowe, T. (2001) A guided tour: small RNA function in Archaea. Mol Microbiol 40, 509-19. Decatur WA, Fournier MJ. (2002) rRNA modifications and ribosome function. Trends Biochem. Sci. 27, 344-51. Durant, P. C. and Davis, D. R. (1999) Stabilization of the anticodon stem-loop of tRNALys, 3 by an A+-C basepair and by pseudouridine. J Mol Biol 285, 115-31. Fogolari, F., Elcock, A. H., Esposito, G., Viglino, P., Briggs, J. M., and McCammon, J. A. (1997). Electrostatic effects in homeodomain-DNA interactions. J Mol Biol 267, 368-81. Frankel, A. D. and Smith, C. A. (1998) Induced Folding in RNA-Protein Recognition: More than a Simple Molecular Handshake. Cell 92, 149-151. Ganot P, Caizergues-Ferrer M, Kiss T. (1997) The family of box ACA small nucleolar RNAs is defined by an evolutionarily conserved secondary structure and ubiquitous sequence elements essential for RNA accumulation. Genes Dev 11, 941-56. Ganot, P., Jady, B. E., Bortolin, M. L., Darzacq, X., and Kiss, T. (1999) Nucleolar factors direct the 2’-O-ribose methylation and pseudouridylation of U6 spliceosomal RNA. Mol Cell Biol 19, 6906-17. Gaspin ,C., Cavaille, J., Erauso, G., and Bachellerie, J.P.(2000) Archaeal homologs of eukaryotic methylation guide small nucleolar RNAs: lessons from the Pyrococcus genomes. J Mol Biol 297, 895-906. Gautier, T., Berges, T., Tollervey, D., and Hurt, E. (1997) Nucleolar KKE/D repeat proteins Nop56p and Nop58p interact with Nop1p and are required for ribosome biogenesis. Mol Cell Biol 17, 7088-98. Gilbert, W. (1978) Why genes in pieces? Nature 271, 501. Goody, T. A., Melcher, S. E., Norman, D. G., and Lilley, D. M. J. (2004). The kink-turn motif in RNA is dimorphic, and metal ion-dependent. RNA 10, 254-264.

83 Granneman, S., Pruijn, G.J.M., Horstman, W., van Venrooij, Lührmann R., and Watkins, N. J. (2002). The hU3-55K protein requires 15.5kD binding to the box B/C motif as well as flanking RNA elements for its association with the U3 small nucleolar RNA in vitro. J Biol Chem 277,48490-500 Green R, Noller HF. (1999) Reconstitution of functional 50S ribosomes from in vitro transcripts of Bacillus stearothermophilus 23S rRNA. Biochemistry 38, 1772-9. Huang, G. M., Jarmolowski, A., Struck, J. C., and Fournier, M. J. (1992). Accumulation of U14 small nuclear RNA in Saccharomyces cerevisiae requires box C, box D, and a 5', 3' terminal stem. Mol Cell Biol 12, 4456-63. Jady, B. E. and Kiss, T. (2001) A small nucleolar guide RNA funcions both in 2’-O- ribose methylation and pseudouridylation of the U5 spliceosomal RNA. EMBO J 20, 541-51. Jeppesen C, Stebbins-Boaz B, Gerbi SA. (1988) Nucleotide sequence determination and secondary structure of Xenopus U3 snRNA. Nucleic Acids Res 16, 2127-48 Joyce, G. F. and Orgel, L. E. (1999) Prospects for understanding the origin of the RNA world in The RNA World, Cold Spring Harbor Press, Cold Spring Harbor, NY. Jurica MS, Sousa D, Moore MJ, Grigorieff N. (2004) Three-dimensional structure of C complex spliceosomes by electron microscopy. Nat Struct Mol Biol 11, 265-9 Kass S, Tyc K, Steitz JA, Sollner-Webb B. (1990) The U3 small nucleolar ribonucleoprotein functions in the first step of preribosomal RNA processing. Cell 60:897-908. Kawai, G., Yamamoto, Y., Kamimura, T., Masegi, T., Sekine, M., Hata, T., Iimori, T., Watanabe, T., Miyazawa, T., and Yokoyama S. (1992 ) Conformational rigidity of specific pyrimidine residues in tRNA arises from posttranscriptional modifications that enhance steric interaction between the base and the 2'-hydroxyl group. Biochem 31, 1040-6 Khaitovich P, Mankin AS. (1999) Effect of antibiotics on large ribosomal subunit assembly reveals possible function of 5 S rRNA. J Mol Biol 291, 1025-34. Kiss-Laszlo, Z., Henry, Y., and Kiss, T. (1998) Sequence and structural elements of methylation guide snoRNAs essential for site-specific ribose methylation of pre- rRNA. EMBO J 17, 797-807. Klein, D. J., Schmeing, T. M., Moore, P. B., and Steiz, J. A. (2001) The kink-turn: a new RNA secondary structure motif. EMBO J 20, 4214-21. Kuhn, J. F., Tran E. J., and Maxwell E. S. (2002) Archaeal ribosomal protein L7 is a functional homolog of the eukaryotic 15.5kD/Snu13p snoRNP core protein. Nucleic Acids Res 30, 931-41. Lafontaine, D. L. and Tollervey, D. (1998) Birth of snoRNPs: the evolution of the modification-guide snoRNAs. Trends Biochem Sci 23, 383-8. Lafontaine, D. L. and Tollervey, D. (1999) Nop58p is a common component of the box C+D snoRNPs that is required for snoRNA stability. RNA 5, 455-67. Lafontaine, D. L. and Tollervey, D. (2000) Synthesis and assembly of the box C+D small nucleolar RNPs. Mol Cell Biol 20, 2650-9. Liang WQ, Fournier MJ. (1995) U14 base-pairs with 18S rRNA: a novel snoRNA interaction required for rRNA processing. Genes Dev 9, 2433-43.

84 Liang X. H., Xu Y. X., and Michaeli S. (2002) The spliced leader-associated RNA is a trypanosome-specific sn(o) RNA that has the potential to guide pseudouridine formation on the SL RNA. RNA 8, 237–246. Lovgren JM, Wikstrom PM. (2002) The rlmB gene is essential for formation of Gm2251 in 23S rRNA but not for ribosome maturation in Escherichia coli. J Bacteriol 183, 6957-60. Lowe, T. M., and Eddy, S. R. (1999). A computational screen for methylation guide snoRNAs in yeast. Science 283, 1168-71. Lu, X., Shakked, Z., and Olson, W. K. (2000). A-DNA Conformational Motifs in Ligand- bound Double Helices. Journal of Molecular Biology 300, 819-840 Maden BE. (1990) The numerous modified nucleotides in eukaryotic ribosomal RNA. Prog Nucleic Acid Res Mol Biol 39, 241-303. Mao, H., White, S.A., and Williamson, J.R., (1999) A novel loop-loop recognition motif in the yeast ribosomal protein L30 autoregulatory RNA complex. Nat Struct Biol 6, 1139-1147. Massenet, S., Mougin A., and Branlant, C. (1998) Posttranscriptional modifications in U small nuclear RNAs. In Modification and Editing of RNA editors Henri Grosjean and Rob Benne. Pp201-227, ASM Press, Washington D.C. Matsumura S, Ikawa Y, Inoue T. (2003) Biochemical characterization of the kink-turn RNA motif. Nucleic Acids Res. 31:5544-51. Mattick, J.S. (2003) Intron and noncoding RNAs: the hidden layer of eukaryotic complexity in Noncoding RNAs: Molecular Biology and Molecular Medicine, Kluwer Academic/Plenum Publishers. Maxwell, E. S. and Fournier, M. J. (1995) The small nucleolar RNAs. Annu Rev Biochem 64, 897-934. McRee, D. E. (1999). XtalView/Xfit--A versatile program for manipulating atomic coordinates and electron density. J Struct Biol 125, 156-65. Mereau A, Fournier R, Gregoire A, Mougin A, Fabrizio P, Luhrmann R, Branlant C. (1997) An in vivo and in vitro structure-function analysis of the Saccharomyces cerevisiae U3A snoRNP: protein-RNA contacts and base-pair interaction with the pre-ribosomal RNA. J Mol Biol 273, 552-71. Mittermaier, A., Varani, L., Muhandiram, D. R., Kay, L. E., and Varani, G. (1999). Changes in side-chain and backbone dynamics identify determinants of specificity in RNA recognition by human U1A protein. J Mol Biol 294, 967-79.Newby MI and Greenbaum NL. (2001) A conserved pseudouridine modification in eukaryotic U2 snRNA induces a change in branch-site architecture. RNA 7, 833- 45. Newby MI and Greenbaum NL. (2002) Sculpting of the spliceosomal branch site recognition motif by a conserved pseudouridine. Nat Struct Biol 9, 958-65. Ni J, Samarsky DA, Liu B, Ferbeyre G, Cedergren R, Fournier MJ. (1997) SnoRNAs as tools for RNA cleavage and modification. Nucleic Acids Symp Ser. 36, 61-3. Nissen P, Ippolito JA, Ban N, Moore PB, Steitz TA. (2001) RNA tertiary interactions in the large ribosomal subunit: the A-minor motif. Proc Natl Acad Sci U S A. 98:4899-903.

85 Noon KR, Bruenger E, McCloskey JA. (1998) Posttranscriptional modifications in 16S and 23S rRNAs of the archaeal hyperthermophile Sulfolobus solfataricus. J Bacteriol 180, 2883-8. Nottrott, S., Hartmuth, K., Fabrizio, P., Urlaub, H., Vidovic, I., Ficner, R., and Luhrmann, R. (1999) Functional interaction of a novel 15.5kD [U4/U6.U5] tri- snRNP protein with the 5' stem-loop of U4 snRNA. EMBO J 18, 6119-33 Nottrott, S., Urlaub, H., and Luhrmann, R. (2002). Hierarchical, clustered protein interactions with U4/U6 snRNA: a biochemical role for U4/U6 proteins. Embo J 21, 5527-38. Omer, A. D., Lowe, T. M., Russell, A. G., Ebhardt, H., Eddy, S. R., and Dennis, P. P. (2000) Homologs of small nucleolar RNAs in Archaea. Science 288, 517-22. Omer, A. D and Dennis P. P. (2001) Archaeal RNP complexes containing methylation guide RNAs. Sixth Annual Meeting of the RNA society. Parker, KA and Steitz, JA. (1987) Structural analysis of the human U3 ribonucleoprotein particle reveal a conserved sequence available for base pairing with pre-rRNA. Mol Cell Biol 7, 2899-913 Peculis, B. A. and Steitz, J. A.(1994) Sequence and structural elements critical for U8 snRNP function in Xenopus oocytes are evolutionarily conserved. Genes Dev 8, 2241-55. Porter GL, Brennwald PJ, Holm KA, and Wise JA. (1988) The sequence of U3 from Schizosaccharomyces pombe suggests structural divergence of this snRNA between metazoans and unicellular eukaryotes. Nucleic Acids Res 16, 10131-52. Puglisi E. V. and Puglisi J. D. (1998) Nuclear magnetic resonance spectroscopy of RNA. In: RNA Structure and Function. Simons E. W. and Grunberg-Manago M., eds. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, 117-116. Rázga, F, Koča, J, Šponer, J and Leontis, N B (2005) Hinge-like motions in RNA Kink- turns: The role of the second A-minor motif and nominally unpaired bases. Biophysical J 2005 Feb. 18, epub. Rashid R, Aittaleb M, Chen Q, Spiegel K, Demeler B, Li H. (2003) Functional requirement for symmetric assembly of archaeal box C/D small ribonucleoprotein particles. J Mol Biol 333, 295-306. Rozhdestvensky TS, Tang TH, Tchirkova IV, Brosius J, Bachellerie JP, Huttenhofer A. (2003) Binding of L7Ae protein to the K-turn of archaeal snoRNAs: a shared RNA binding motif for C/D and H/ACA box snoRNAs in Archaea. Nucleic Acids Res 31, 869-77. Saenger, W. (1984) Principles of Nucleic Acid Structure. Springer Verlag, p336. Schluenzen F, Tocilj A, Zarivach R, Harms J, Gluehmann M, Janell D, Bashan A, Bartels H, Agmon I, Franceschi F, Yonath A. (2000) Structure of functionally activated small ribosomal subunit at 3.3 angstroms resolution. Cell 102, 615-23. Schmitt M. E., Bennett J. L., Dairaghi D. J., and Clayton D. A. (1993) Secondary structure of RNase MRP RNA as predicted by phylogenetic comparison. FASEB J 7, 208–213. Samarsky DA, Fournier MJ. (1998) Functional mapping of the U3 small nucleolar RNA from the yeast Saccharomyces cerevisiae. Mol Cell Biol 18, 3431-44. Shana O. Kelley and Jacqueline K. Barton (1999) Electron Transfer Between Bases in Double Helical DNA Science Jan 15 1999: 375-381.

86 Showalter, S. A., and Hall, K. B. (2002). A functional role for correlated motion in the N- terminal RNA-binding domain of human U1A protein. J Mol Biol 322, 533-42. Sitkoff, D., Sharp, K. A., and Honig, B. (1994). Accurate Calculation of Hydration Free Energies Using Macroscopic Solvent Models. Journal of Physical Chemistry 98, 1978-1988. Song, H., Mugnier, P., Das, A. K., Webb, H. M., Evans, D. R., Tuite, M. F., Hemmings, B. A., and Barford, D. (2000). The crystal structure of human eukaryotic release factor eRF1--mechanism of stop codon recognition and peptidyl-tRNA hydrolysis. Cell 100, 311-21. Soukup JK, Minakawa N, Matsuda A, Strobel SA (2002) Identification of A-minor tertiary interactions within a bacterial group I intron active site by 3- deazaadenosine interference mapping. Biochemistry 41:10426-38. Speckmann WA, Li ZH, Lowe TM, Eddy SR, Terns RM, Terns MP. (2002) Archaeal guide RNAs function in rRNA modification in the eukaryotic nucleus. Curr Biol 12, 199-203. Starostina NG, Marshburn S, Johnson LS, Eddy SR, Terns RM, Terns MP. (2004) Circular box C/D RNAs in Pyrococcus furiosus.Proc Natl Acad Sci U S A. 101:14097-101. Steitz JA, Tycowski KT. (1995) Small RNA chaperones for ribosome biogenesis. Science 270, 1626-7. Stevens, S. W. and Abelson, J. (1999) Purification of the yeast U4/U6.U5 small nuclear ribonucleoprotein particle and identification of its proteins. Proc Natl Acad Sci U S A 96, 7226-31. Szewczak, W. L. B., DeGregorio, S. J., Strobel, S. A., and Steitz, J. A. (2002) Exclusive Interaction of the 15.5kD Protein with the Terminal Box C/D motif of a Methylation Guide snoRNP. Chemistry and Biology, 9, 1095-1107. Tang TH, Bachellerie JP, Rozhdestvensky T, Bortolin ML, Huber H, Drungowski M, Elge T, Brosius J, Huttenhofer A. (2002) Identification of 86 candidates for small non-messenger RNAs from the archaeon Archaeoglobus fulgidus. PNAS 99, 7536-41. Tang TH, Rozhdestvensky TS, d'Orval BC, Bortolin ML, Huber H, Charpentier B, Branlant C, Bachellerie JP, Brosius J, Huttenhofer A. (2002) RNomics in Archaea reveals a further link between splicing of archaeal introns and rRNA processing. Nucleic Acids Res 30, 921-30. Thomasson, K., Ouporov, I., Baumgartner, T., Czlapinski, J., T, K., and Northrup, S. (1997). Free Energy of Nonspecific Binding of Cro Repressor Protein to DNA. Journal of Physical Chemistry B 101, 9127-9136. Tollervey, D., Lehtonen, H., Carmo-Fonseca, M., and Hurt, E.C. (1991) The small nucleolar RNP protein NOP1 (fibrillarin) is required for pre-rRNA processing in yeast. EMBO J 10, 573-83. Tollervey, D., Lehtonen, H., Jansen, R., Kern, H., and Hurt, E.C. (1993) Temperature- sensitive mutations demonstrate roles for yeast fibrillarin in pre-rRNA processing, pre-rRNA methylation, and ribosome assembly. Cell 72,443-57. Tollervey, D. and Kiss, T. (1997). Function and synthesis of small nucleolar RNAs. Curr Opin Cell Biol 9, 337-42.

87 Tran, E. J., Zhang, X., and Maxwell, E. S. (2003). Efficient RNA 2'-O-methylation requires juxtaposed and symmetrically assembled archaeal box C/D and C'/D' RNPs. Embo J 22, 3930-40Tyc K, Steitz JA. (1989) U3, U8 and U13 comprise a new class of mammalian snRNPs localized in the cell nucleolus. EMBO J 8, 3113-9. Tworowski D, Safro M. (2003) The long-range electrostatic interactions control tRNA- aminoacyl-tRNA synthetase complex formation. Protein Sci. 12, 1247-51. Winkler, W. C., Grundy, F. J., Murphy, B. A., and Henkin, T. M. (2001). The GA motif: An RNA element common to bacterial antitermination systems, rRNA, and eukaryotic RNAs. RNA 7, 1165-1172.

Tyc, K. and Steitz, J. A. (1989) U3, U8 and U13 comprise a new class of mammalian snRNPs localized in the cell nucleolus. EMBO J, 8, 3113-9. Tycowski, K. T., You, Z. H., Graham, P. J., and Steitz, J. A. (1998) Modification of U6 spliceosomal RNA is guided by other small RNAs. Mol Cell 2, 629-38. Vidovic, I., Nottrott, S., Hartmuth, K., Luhrmann, R., and Ficner, R. (2000) Crystal structure of the spliceosomal 15.5kD protein bound to a U4 snRNA fragment. Mol Cell 6, 1331-42. Wang, H., Boisvert, D., Kim, K. K., Kim R., and Kim S. H. (2000) Crystal structure of a fibrillarin homologue from Methanococcus jannaschii, a hyperthermophile, at 1.6 A resolution. EMBO J 19, 317-23. Watkins, N. J, Leverette, R. D., Xia, L., Andrews, M. T., and Maxwell, E. S. (1996) Elements essential for processing intronic U14 snoRNA are located at the termini of the mature snoRNA sequence and include conserved nucleotide boxes C and D. RNA 2, 118-33. Watkins, N. J., Segault, V., Charpentier, B., Nottrott, S., Fabrizio, P., Bachi, A., Wilm, M., Rosbash, M., Branlant, C., and Luhrmann, R. (2000) A common core RNP structure shared between the small nucleoar box C/D RNPs and the spliceosomal U4 snRNP. Cell 103, 457-66. Watkins, N. J., Dickmanns, A., and Luhrmann, R. (2002). Conserved stem II of the box C/D motif is essential for nucleolar localization and is required, along with the 15.5K protein, for the hierarchical assembly of the box C/D snoRNP. Mol Cell Biol 22, 8342-52. Weidenhammer, E. M., Singh, M., Ruiz-Noriega, M., and Woolford, J. L. Jr. (1996) The PRP31 gene encodes a novel protein required for pre-mRNA splicing in Saccharomyces cerevisiae. Nucleic Acids Res 24, 1164-70. Weinstein, L. B. and Steitz, J. A. (1999) Guided tours: from precursor snoRNA to functional snoRNP. Curr Opin Cell Biol 11, 378-84.

Wimberly BT, Brodersen DE, Clemons WM Jr, Morgan-Warren RJ, Carter AP, Vonrhein C, Hartsch T, Ramakrishnan V. (2000) Structure of the 30S ribosomal subunit. Nature 407, 327-39. Wise, JA and Weiner, AM. (1980) Dictyostelium small nuclear RNA D2 is homologous to rat nucleolar RNA U3 and is encoded by a dispersed multigene family. Cell 22, 109-18.

88 Wormsley S, Samarsky DA, Fournier MJ, Baserga SJ. (2001) An unexpected, conserved element of the U3 snoRNA is required for Mpp10p association. RNA. 7, 904-19. Wu, P., Brockenbrough, J. S., Metcalfe, A. C., Chen, S., and Aris J. P. (1998) Nop5p is a small nucleolar ribonucleoprotein component required for pre-18 S rRNA processing in yeast. J Biol Chem 273, 16453-63. Yu YT, Shu MD, Steitz JA. (1998) Modifications of U2 snRNA are required for snRNP assembly and pre-mRNA splicing. EMBO J 17, 5783-95. Yu, Y. T., Scharl, E. C., Smith, C. M., and Steitz, J. A. (1999) The growing world of small nuclear ribonucleoproteins in The RNA World, Cold Spring Harbor Press, Cold Spring Harbor, NY. Yusupov MM, Yusupova GZ, Baucom A, Lieberman K, Earnest TN, Cate JH, Noller HF. (2001) Crystal structure of the ribosome at 5.5 A resolution. Science 292, 883-96. Zacharias, M., Luty, B. A., Davis, M. E., and McCammon, J. A. (1992). Poisson- Boltzmann analysis of the lambda repressor-operator interaction. Biophys J 63, 1280-5 Zebarjadian Y, King T, Fournier MJ, Clarke L, Carbon J. (1999) Point mutations in yeast CBF5 can abolish in vivo pseudouridylation of rRNA. Mol Cell Biol 19, 7461-72. Zmudzka, B. and Shugar, D. (1970) Role of the 2'-hydroxyl in polynucleotide conformation. Poly 2'-O-methyluridylic acid. FEBS Lett 8, 52-54 Zuker, M. (2003) Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res. 31, 3406-15.

BIOGRAPHICAL SKETCH

EDUCATION Ph.D., Chemistry and Biochemistry, FSU, Tallahassee, FL 08/02 - 5/05 Dissertation: Biophysical and biochemical investigation of an archaeal box C/D sRNP: RNA- protein interactions of a kink turn RNA within the functional enzyme Advisor: Dr. Hong Li

M.S. Thesis, Chemistry and Biochemistry, FSU, Tallahassee, FL 08/98 - 05/02 Thesis: Exploring possible RNA-peptide interactions in the regulation of stp expression Advisor: Dr. Nancy Greenbaum

B.S., Chemistry, Randolph-Macon Woman’s College, Lynchburg, VA 08/94 - 05/98

RESEARCH EXPERIENCE Doctoral Research, FSU, Tallahassee, FL 08/02 - 5/04 Investigated assembly and structure of an archaeal box C/D sRNP. Determined X-ray crystal structure of one box C/D sRNP complex. Optimized crystallization schemes for 3 different co- crystals of box C/D sRNP complexes including design of multiple RNA constructs and proteins. Probed structural change upon RNA and protein interaction of 2 box C/D sRNP complexes using solution NMR. Developed molecular biology and biochemical methods for purification, analysis of assembly, and manipulation of RNA and protein components.

Masters Research, FSU, Tallahassee, FL 08/98 - 05/02 Analyzed the autoregulation of stp (suppressor of 3’ phosphatase polynucleotide kinase). Determined specificity of RNA-peptide system using biochemical methods including UV-Vis, circular dichroism (CD), and electrophoretic mobility shift assay (EMSA). Developed in-vitro transcription assays for 4 different RNA constructs and RNA/DNA purification protocols.

Undergraduate Research, Chemistry Research Intern, VMI, Lexington, VA 05/97 - 08/97 Advisor: Dr. Robert M. Granger, II 05/96 - 07/96 Examined Pt bio-organometallic compounds as possible chemotherapeutic agents. Synthesized and characterized 2 different Pt (IV) bio-organometallic compounds. Explored synthetic routes for Pt(dppz) compound and a dppz analog ligand. Used biochemical and spectroscopic techniques including UV-Vis, CD, and florescence to determine DNA binding mode of Pt (IV) compounds.

PROFESSIONAL EXPERIENCE Laboratory Technician, Chemistry and Biochemistry, FSU, Tallahassee, FL 05/02 - 08/02 Professor: Dr. Hong Li • Managed and trained 3 graduate students, 3 undergraduate students, and 1 postdoc in biochemical and molecular biology techniques • Conducted preliminary research on several new RNA-protein projects • Directed laboratory activities including purchasing, equipment upkeep, and research

TEACHING EXPERIENCE Teaching Fellow Supervisor, Chemistry and Biochemistry, FSU, Tallahassee, FL01/01 - 01/02

90 • Supervised and trained 10 general chemistry I and II teaching assistants per semester • Coordinated lab scheduling for general chemistry labs

Teaching Fellow, Chemistry and Biochemistry, FSU, Tallahassee, FL 08/98 - 01/01 • Oversaw and instructed classes of 20-28 students in general chemistry and advanced biochemistry labs • Received excellent student evaluations for all classes and teaching award for advanced biochemistry lab from Alpha Chi Sigma, Gamma Beta Chapter

AWARDS/HONORS Biophysical Society Travel Grant, Baltimore, MD (2004) Outstanding Senior Level Teaching Award, Alpha Chi Sigma, FSU (2001) Hoffman Teaching Scholarship, FSU (1998-2000) Chair, Exam Scheduling Committee, R-MWC (1995-1996) Dean’s Award, R-MWC (1994-1998) Commonwealth Centennial Citizenship Award, R-MWC (1994-1998)

SKILLS Technical • X-ray crystallography including crystallization, optimization, and structure determination of RNA-protein crystals • 1D and 2D High field nuclear magnetic resonance (NMR) spectroscopy (Varian) of RNA- protein complexes • Circular dichroism (CD), mass spectrometry (MS), UV-Vis, and fluorescence spectroscopy of nucleic acids and proteins • Chromatography: LC, HPLC, FPLC, GC, affinity, gel filtration, reverse phase, and ion exchange • Electrophoresis: denaturing, native, SDS-PAGE, large-scale PAGE purification, and electrophoretic mobility shift assay (EMSA) • Molecular biology: cloning, mutagenesis, PCR, transformation, cell culture, DNA purification, protein expression and purification, in-vitro transcription, RNA extraction and purification, and RNA structure design • Biochemical assays: UV crosslinking, equilibrium dialysis, RNase digestions, restriction endonuclease assays, binding assays, enzyme assays, crystallization trials, and radiolabeling • Instrumentation: phosphorimager and automated nucleic acid synthesizers • Bioinformatics: multiple sequence alignment of nucleic acid and amino acid sequences, secondary structure alignments, homology searches, and secondary structure predictions

Computers • Office software: Microsoft Office (Word, Excel, PowerPoint), Corel Office, Endnote, Adobe software, Sigma Plot, and SSH Secure Shell • Advanced database searching abilities: primary sequences, homology, secondary structures, similarity searches (BLAST) and Protein Data Bank (PDB) coordinate systems of protein and RNA • Programming languages: Q-Basic, UNIX, and LINUX • Graphics software: Chemdraw, Canvas, Adobe Illustrator, Pymol, RasMol, Chimera, and Ribbons • Structural and refinement programs: HKL2000, Denzo, CNS, XtalView, RNAView, Mfold, VNMR, NMRView, and NMRPipe

PEER-REVIEWED PUBLICATIONS • Spectroscopic and biochemical characterization of box C/D sRNP RNA-protein binding interactions. Karen R. Brown*, Terrie L. Moore*, Jack J. Skalicky, and Hong Li. (manuscript in preparation) • Molecular basis of box C/D RNA-protein interactions: co-crystal structure of archaeal L7Ae and a box C/D RNA. Terrie Moore, Yanming Zhang, Marcia O. Fenley, and Hong Li. Structure 2004, 12, 807-18. (evaluated in Faculty of 1000 May 20, 2004) • Synthesis and characterization of the first homoleptic diimene complex of Pt(IV): [Pt(1,10- phenanthroline)3][PF6]4. Robert Granger II, Jill Nelson Granger, Karl D. Sienerth, J. Micah North, Michael H. Wilson, Terrie H. Luong, Stephanie J. Garcia, and Kelly del Campo. The Journal of Undergraduate Chemistry Research 2002, 3, 95. • Isolation of the first homoleptic diimene Pt(IV) compounds: synthesis, DNA-binding, and cell culture studies. R. M. Granger, J. N. Granger, R. L. Davies, A. K. Addington, S. J. Garcia, A. K. Clark, T. Luong, and M. Wilson. Virginia Journal of Science 1997, 48, 2, 105.

INVITED PRESENTATIONS • Molecular basis of box C/D RNA-protein interactions: co-crystal structure of archaeal L7Ae and a box C/D RNA. Terrie Moore, Yanming Zhang, Marcia O. Fenley, Jack J. Skalicky, and Hong Li, Annual Biophysical Society Conference, February 2004, Baltimore, MD. • Box C/D RNA-protein interactions: crystal structure of ribosomal L7Ae and a box C/D RNA at 2.9Å. Terrie L. Moore, Yanming Zhang, Marcia O. Fenley, and Hong Li, Florida Annual Meeting and Exposition of the American Chemical Society Florida Division, May 2003, Orlando, FL. • Synthesis, characterization and DNA studies of the first homoleptic diimene compound of Pt(IV). Terrie Luong, Mike Wilson, Jill Nelson Granger, Robert M. Granger, Virginia Regional meeting of the American Chemical Society, April 1998, University of Virginia, Charlottesville, VA.

ABSTRACTS • Co-crystal structure of the initial box C/D RNA assembly complex: interactions of a kink turn RNA. Terrie L. Moore, Yanming Zhang, Marcia O. Fenley, Jack J. Skalicky, and Hong Li, RNA Society Meeting, June 2004, University of Wisconsin-Madison, Madison, WI. • Box C/D RNA-protein interactions: co-crystal structure of the archaeal sRNP initiation complex at 2.9Å. Terrie L. Moore, Yanming Zhang, Marcia O. Fenley, and Hong Li, RNA Society Meeting, July 2003, Vienna, Austria. • Investigation of RNA-peptide interactions in the autoregulation of stp expression. Terrie L. Moore and Nancy L. Greenbaum, Florida Annual Meeting and Exposition of the American Chemical Society Florida Division, May 2001, Orlando, FL. • Synthesis, characterization, and DNA Studies of the first homoleptic diimene compound of Pt(IV). Terrie Luong, Mike Wilson, Jill Nelson Granger, Robert M. Granger, Virginia Blue Ridge Regional meeting of the American Chemical Society, April 1998, Radford University, Radford, VA. • Synthesis and characterization of the first homoleptic diimene compound of Pt(IV). Terrie Luong, Jill Nelson Granger, Robert M. Granger, DNA-Platinum Research, February 1998, Virginia Polytechnic and State University, Blacksburg, VA. • Isolation of the first homoleptic diimene Pt(IV) compounds: synthesis, DNA-binding, and cell culture studies. R. M. Granger, J. N. Granger, R. L. Davies, A. K. Addington, S. J. Garcia, A.

92 K. Clark, T. Luong, and M. Wilson, Virginia Academy of Science, May 1997, Virginia Polytechnic and State University, Blacksburg, VA.

PROFESSIONAL MEMBERSHIPS American Association for Advancement of Science (AAAS) American Chemical Society Biophysical Society RNA Society