Mechanistic and Bioinformatic Studies of Mitochondrial Ribosomes and Auxiliary Translational Factors Domenick Gabriel Grasso

Mechanistic and Bioinformatic Studies of Mitochondrial Ribosomes and Auxiliary Translational Factors

Domenick Gabriel Grasso

A dissertation submitted to the faculty of the University of North Carolina at Chapel Hill in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Department of Chemistry.

Chapel Hill 2007

Approved by:

Dr. Linda Spremulli

Dr. Gary Pielak

Dr. Dorothy Erie

Dr. Kevin Weeks

Dr. Nancy Thompson

ii ABSTRACT

DOMENICK GABRIEL GRASSO: Mechanistic and Bioinformatic Studies of Mitochondrial Ribosomes and Auxiliary Translational Factors (Under the direction of Dr. Linda L. Spremulli)

This body of work is focused on mitochondrial translational systems. This apparatus synthesizes vital components of the electron transport chain facilitating the production of energy in eukaryotic cells. Of particular interest is the role of mammalian mitochondrial initiation factor 2 (IF2mt) in this process. This factor recruits the initiatior fMet-tRNA to begin protein synthesis within this organelle. IF2mt is unique in that it possesses an insertion. The work described here demonstrates that the insertion acts as the factor equivalent of initiation factor 1 (IF1). IF1 is an essential gene in prokaryotes and eukaryotes

(eIF1A) but has yet to be discovered in mitochondria. IF2mt is a GTPase and the guanine ring binding region in this factor was probed to determine the effect of mutations on the X residue of the guanine ring binding motif NKXD. Quite surprisingly despite the connotation of X being random, this residue is very sensitive to perturbation.

Other work included in this dissertation includes a bioinformatics analysis of the proteins found in a Leishmania tarentolae mitochondrial ribosomal small subunit particle. L. tarentolae is a member of the trypanosomitids family, whose members are known to cause

African sleeping sickness and South American Chagas disease. Examination of this organism offers a deeper understanding to the mechanism of action of these disease causing organisms and offers the potential development of therapies for these illnesses. This study explored the

iii limitations of “hits” produced by search algorithms as well as insight into mitochondrial ribosomal protein evolution.

This last chapter of this dissertation attempts to further the understanding of the mechanisms of translation in the mammalian mitochondria by probing for previously unknown translational factors. Multiple approaches are employed to find these factors including alignments to factors known to be involved in the prokaryotic translational apparatus, such as LepA, as well as promoter and import sequence analysis.

ACKNOWLEDGEMENTS

I cannot faithfully say that I have done this on my own. There is one person who, without their guidance, this work would have never been possible: my advisor, mentor and friend Linda Spremulli. She is a beacon of fortitude and kindness that I will always admire and respect. Words cannot express my thanks to her for guiding me these past five years. I hope to live up to the potential she sees in me and make her proud through my accomplishments in science and life.

Of course one cannot escape their genes so my parents deserve a lot of the credit for turning me into the curious creature that I am today. I love them and my family very much.

And last I have been blessed to have been surrounded by amazing and beautiful people. I’d like to thank those who have stood by me, believed in me and loved me. Time and distance can only separate me physically from the people I care for because I will always hold those joyful memories shared between us close to my heart.

TABLE OF CONTENTS

LIST OF TABLES ...... vii

LIST OF FIGURES ...... viii

ABBREVIATIONS ...... x

Chapter

I. MECHANISM OF PROTEIN SYNTHESIS ...... 1

INTRODUCTION ...... 2

PROKARYOTIC PROTEIN SYNTHESIS ...... 4

EUKARYOTIC PROTEIN SYNTHESIS ...... 10

MITOCHONDRIAL PROTEIN SYNTHESIS ...... 14

REFERENCES ...... 20

II. STRUCTURAL AND BIOINFORMATIC STUDIES OF MITOCHONDRIAL RIBOSOMES ...... 23

INTRODUCTION ...... 24

MATERIALS AND METHODS ...... 35

RESULTS ...... 39

DISCUSSION ...... 46

REFERENCES ...... 48

III. ANALYSIS OF THE ROLES OF THE NKXD MOTIF AND THE INSERTION DOMAIN IN MAMMALIAN MITOCHONDRIAL INITIATION FACTOR 2..50

INTRODUCTION ...... 51

vi MATERIALS AND METHODS ...... 60

RESULTS ...... 69

DISCUSSION ...... 79

REFERENCES ...... 83

IV. AUXILIARY TRANSLATIONAL FACTORS OF MAMMALIAN MITOCHONDRIAL RIBOSOMES ...... 87

INTRODUCTION ...... 88

MATERIALS AND METHODS ...... 92

RESULTS ...... 96

DISCUSSION ...... 109

REFERENCES ...... 111

vii LIST OF TABLES

Table

1-1. Comparison of the E. coli and Mammalian Mitochondrial Translational Machinery ...... 16

1-2. Further Comparison of the E. coli and Mammalian Mitochondrial Translational Machinery ...... 19

2-1. The Primers Used to Created MRPL13, MPRL15, MRPS2 and MRPS11 ...... 36

2-2. Preliminary Analysis of the 45S SSU* Particle LC/MS/MS Peptide Hits ...... 43

2-3. Comparison of Leishmania, Yeast and Human Mitochondrial and Cytoplasmic Ribosomal Proteins to that of Eubacteria...... 45

3-1. Primers Used to Created the NKXD Variants, IF2mtΔ37 and EcoIF2::37 ...... 62

4-1. Conservation of Bacterial Translational Factors...... 90

4-2. Mitochondrial Import and Promoter Analysis of Various Factors ...... 103

viii LIST OF FIGURES

Figure

1-1. Placement of tRNAs on the SSU and the Peptide Exit Channel ...... 3

1-2. Overview of the Steps of Initiation in Prokaryotes ...... 6

1-3. Overview of the Elongation Cycle in Prokaryotes ...... 8

1-4. Overview of Termination and Recycling in Prokaryotes ...... 9

1-5. Overview of Eukaryotic Initiation ...... 12

1-6. Overview of Elongation and Termination in Eukaryotes ...... 13

1-7. The Mammalian Mitochondrial Genome...... 15

2-1. The Prokaryotic LSU and SSU ...... 26

2-2. The Secondary and Tertiary Structure of Prokaryotic rRNA ...... 27

2-3. The Mammalian Mitochondrial Ribosome ...... 29

2-4. The Mammalian Mitochondrial rRNA ...... 30

2-5. Structure and Isolation of Leishmania tarentolae Mitochondrial Ribosomes ...... 33

2-6. The Secondary Structure of Leishmania tarentolae SSU 9S rRNA ...... 34

2-7. Modeling of the Mammalian Mitochondrial Proteins for Accessibility ...... 40

2-8. Expression and Western Analysis of MRP Proteins and Antibodies ...... 41

3-1. Domain Organization and Structure of IF2 ...... 52

3-2. The GTPase Center of aIF5B Coordinated with GDPNP ...... 54

3-3. The NKXD Motifs in IF2 ...... 56

3-4. Alignment and Structural Model of the Insertion in IF2mt ...... 57

3-5. Factor Binding Model of the Initiation Complex on the SSU ...... 59

ix 3-6. Creation of IF2mtΔ37 and EcoIF2::37 ...... 63

3-7. Western Analysis of C290A/I/S IF2mt Expression ...... 70

3-8. 70S and 55S Initiation Complex Formation with NKXD Variants ...... 72

3-9. 70S Initiation Complex Formation and IF1 Dosing with IF2mtΔ37 ...... 75

3-10. 70S Initiation Complex Formation and IF1 Dosing with EcoIF2::37 ...... 76

3-11. 55S Initiation Complex Formation and IF1 Dosing with IF2mtΔ37 and EcoIF2::37...... 78

3-12. Space-filled Model of the NKXD Motif ...... 80

4-1. Homology of LepA to EF-G and EF-Tu ...... 91

4-2. NRF-2 Domains in Mitochondrial Translational Factors ...... 93

4-3. Alignment of mtLepA and E. coli LepA ...... 97

4-4. Analysis of the mtLepA L/P Mutation...... 98

4-5. Growth Curves of EF-Gmt and mtLepA ...... 100

4-6. Promoter Analysis of IF3mt, mtLepA, mtOxa1, mtLetm1 and mtCgtAE...... 102

4-7. Alignment of Yeast Mdm38 and Ylh47 Against Human Letm1 ...... 104

4-8. Promoter Analysis of mtERA, YciH Homologs, mtObgH1 and mtSpoT ...... 105

x ABBREVIATIONS

A260 or A280 absorbance at the indicated wavelength (nm) aa amino acid aa-tRNA aminoacyl-tRNA

BME -mercaptoethanol

BSA bovine serum albumin cryo-EM cryo-electron microscopy

DTT dithiothreitol

EDTA ethylenediaminetetraacetic acid fMet-tRNA formylated methionyl-tRNA

GDP guanosine 5'-diphosphate

GDPNP guanosine 5'-[ , -imido]triphosphate

GTP guanosine 5'-triphosphate h hour

Hepes 4-(2-hydroxy-ethyl)-1-piperazineethanesulfonic acid

HPLC high performance liquid chromatography

IPTG isopropyl- -D-thiogalactopyranoside nt nucleotide

PCR polymerase chain reaction

PMSF phenylmethylsulfonyl fluoride poly(A,U,G) poly(adenylic, uridylic, guanylic) acid

PVDF polyvinylidene fluoride

Tris Tris-(hydroxymethyl)aminomethane

CHAPTER I

MECHANISM OF PROTEIN SYNTHESIS

INTRODUCTION

Protein synthesis is a vital part of the cellular machinery. This process was initially investigated in the 1960s. Subsequent work has revealed a diverse array of protein synthesis components across organisms and organelles. The ribosome is the central player in protein synthesis and it acts as a scaffold which facilitates the interaction of the messenger ribonucleic acids (mRNA), containing the coded gene sequence, and the transfer ribonucleic acids (tRNA), which carry each amino acid to the correct position in the growing polypeptide. The ribosome catalyzes peptide bond formation using the peptidyltransferase activity of this particle. All ribosomes are made up of a large subunit (LSU) and small subunit (SSU) and the chemistry of protein synthesis (translation) occurs at the interface of these subunits. The small subunit cradles the mRNA and, therefore, has discrete binding sites for the tRNA, which complements the mRNA’s codons. These sites are referred to as the acceptor-site (A-site), peptidyl-tRNA-site (P-site) and the exit-site (E-site) (Fig. 1-

1A). The roles of these sites will be discussed later. The large subunit contains the exit tunnel for the growing polypeptide chain and the site of peptidyltransferase activity (Fig. 1-

1B). The general outline of the process involves four main steps: initiation, elongation, termination and recycling. In each step there are respective factors that allow the ribosome to complete its task. These processes are carried out on the ribosome in very specific sites on the ribosome.

Initiation of protein synthesis involves initiation factors (IF) that work together to bring the ribosome, the mRNA and the first tRNA to be positioned correctly to start translating the mRNA into protein. Once the ribosome has been correctly oriented into its

2 3 initiation state, the elongation factors (EF) are recruited to continue the process of protein synthesis. These EFs work together bringing in new aminoacyl-tRNAs (aa-tRNAs) and moving the mRNA through the ribosome. There comes a point when the ribosome encounters a stop codon and termination needs to occur. A release factor (RF) is involved in the recognition of termination codons and in the release of the growing polypeptide from the ribosome. Finally the ribosomes are recycled, in most cases, by the ribosome recycling factor (RRF) which further dissociates the mRNA and tRNA from the ribosome and leaves it ready for another round of initiation.

Of course, this simplistic view hides the grace and choreography involved in these processes. Ribosomes and the processes by which proteins are made vary quite dramatically between organisms and even within organelles in the same cell, namely mitochondria in eukaryotes.

PROKARYOTIC PROTEIN SYNTHESIS

The protein synthesis machinery in prokaryotes is the most thoroughly understood of the known translational systems. The ribosome is composed of both ribosomal RNA (rRNA) and proteins. The prokaryotic ribosome is comprised of two subunits, the LSU and the small

SSU. The prokaryotic LSU is a 50S particle, based on its sedimentation in a sucrose gradient, with 33 proteins that decorate the 23S and 5S rRNAs. The prokaryotic SSU is a

30S particle with 21 proteins and a 16S rRNA. These two subunits come together to form a

70S particle.

4 Translation initiation in prokaryotes involves three initiation factors: IF1, IF2 and

IF3. IF1 is thought to bind at the A-site of the SSU thus preventing the initiator tRNA from binding [1-3] and has been implicated as an RNA chaperone [4]. IF2 is a GTPase that

Met Met facilitates the binding of the initiator tRNA, formylmethionyl-tRNAi (fMet- tRNAi ), to the P-site of the SSU [5]. IF3 appears to have many functions [6]: it binds to the 30S ribosomal subunit thereby shifting the equilibrium of the 70S particle toward dissociation, it promotes 30S initiation complex formation by increasing the rate of the P-site codon-

Met anticodon interaction between fMet-tRNAi and the mRNA, and it acts as a proofreading factor by increasing the dissociation rate of noncanonical and pseudo-30S initiation complexes. The mRNA is recruited to the SSU via a base-pairing interaction between the 3' end of the 16S rRNA and the Shine-Dalgarno sequence located upstream of the initiation codon in the mRNA. In addition single stranded regions, U-rich or U(C)/A-rich, regions of mRNA, interact with the ribosomal protein S1 promoting initiation on certain mRNAs [7].

The proposed mechanism of initiation is shown in Figure 1-2. This process begins with the dissociation of the 70S particles into 50S and 30S subunits. IF3 is thought to act as an antiassociation factor, binding to the 30S subunit and sequestering it for initiation. IF1 binds to the IF3:30S complex. IF2 is recruited to the small subunit along with the mRNA

Met and fMet- tRNAi . IF2 interacts with the initiator tRNA and promotes the binding to the P- site, where the initiation codon is located. The order of binding may be random with stable interactions observed only when all three components are present together. IF3 is thought to play a proofreading role in this process and positions the message so that the start, AUG, codon is correctly placed in the P-site of the ribosome. The 50S subunit associates with this

6 complex and IF2 hydrolyzes GTP and is released along with all of the other IFs and translational elongation is ready to begin. Details on the release of the initiation factors are not yet resolved.

Elongation of protein synthesis immediately follows initiation. This process (Fig. 1-

3) involves two GTPase factors, EF-Tu, which binds aminoacylated-tRNAs (aa-tRNAs), and

EF-G, which translocates the ribosome along the mRNA exposing the next codon. With the initiator tRNA in the P-site, the EF-Tu:GTP:aa-tRNA ternary complex delivers the aa-tRNA specific by the next codon into the A-site of the ribosome. Once correct positioning of this tRNA is accomplished, EF-Tu hydrolzes GTP and is released. The accommodation of an incorrect tRNA is sensed by the ribosome-bound complex and that aa-tRNA is rejected.

The peptidyltransferase activity of the 50S subunit then catalyzes the transfer the peptide from the P-site peptidyl-tRNA to the amino acid on the A-site aa-tRNA resulting in a growing polypeptide chain. Following this transfer, EF-G:GTP is recruited into the A-site of the ribosome, hydrolyzes GTP and translocates the tRNAs from the P- and A-sites to the E- and P-sites, respectively. The EF-G:GDP protein is then released. The ribosome is then ready for the next round of elongation. The E-site tRNA is then ejected and EF-Tu brings in the aa-tRNA corresponding to the next codon on the mRNA and the process repeats itself as the polypeptide chain grows.

When the translational elongation machinery reaches a stop codon (UAA, UGA,

UGG) termination is triggered. An overview of termination and recycling is shown in Figure

1-4. Termination involves three main release factors: RF1, RF2 and RF3. RF1 and RF2, are class 1 release factors which interact with the termination codons in a sequence specific

9 way. RF1 and RF2 can both interact with UAA codons but only RF1 can decode UAG and only RF2 can decode UGA [8]. Once a stop codon is encountered the appropriate release factor RF1 or RF2 is recruited to the A-site aids in the hydrolysis of the peptidyl-tRNA. The polypeptide is released and RF3 is recruited to the complex. RF3 is the class 2 release factor and a GTPase, which is recruited to and stimulates the activities of RF1 and RF2 and also facilitates the release of the growing polypeptide from the tRNA in the P-site and prepares the ribosome for recycling [9]. RF3 is subsequently released and the ribosomal complex is ready for recycling.

Recycling is very closely associated with termination and is facilitated by ribosome recycling factor (RRF), EF-G and IF3 which dissociate the 70S ribosomal complexes into the substituent subunits and free the mRNA and other factors interacting with the ribosome

[10]. RRF and EF-G bind to the terminated ribosomal complex and promoted the dissociation of the subunits. It is then proposed that IF3 aids in the dissociation of the mRNA and tRNA from the small subunit. However there is some controversy in the role of

IF3 in this process. The consensus is that IF3 helps the ribosome-recycling process primarily by converting transiently dissociated subunits into subunits by binding sequestering the 30S subunits [11]. Following recycling the ribosomes are ready for another round of initiation.

EUKARYOTIC PROTEIN SYNTHESIS

Like prokaryotic protein synthesis, eukaryotic protein synthesis requires a unique set of protein factors and has a very specialized set of auxiliary components. The most studied eukaryotic translational system is that of yeast. The yeast ribosomes is an 80S particle and

10 contain a 60S LSU and a 40S SSU. The LSU is made up of 3 rRNAs (5.8S, 25S, 5S) and 46 ribosomal proteins. The SSU is composed of a single 18S rRNA and 32 ribosomal proteins.

Initiation of protein synthesis in eukaryotes, as in prokaryotes, begins with the binding of the mRNA and initiator tRNA to the SSU (Fig. 1-5). In this system, eIF3 acts as an antiassociation factor sequestering the 40S SSU in concert with a variety of other factors including eIF1, eIF1A and eIF3. These proteins interact with each other along with the eIF2B:GTP:fMet-tRNAfMet and mRNA to form the initiation complex. Many of these additional proteins are involved in the recognition of the 5' cap of the mRNA and the poly(A) tail a mechanism of initiation which is not known to exist in prokaryotes. These proteins include eIF3, eIF4A, eIF4B, eIF4F and PAB (poly-A binding protein). Once the mRNA is recruited a 43S:mRNA complex is created. Scanning by this complex to locate the initiation codon, AUG, is facilitated by eIF1 and eIF1A. Subsequently, eIF5 aids in releasing eIF2, which occurs following the hydrolysis of GTP. The 60S LSU is recruited with the aid of eIF5B to form the 80S initiation complex.

In contrast to initiation, the mechanism of elongation is thought to be significantly conserved across all kingdoms of life [12-14]. An exception in eukaryotic elongation (Fig. 1-

6) is the presence of a third elongation factor, eEF-3, which is present in fungal systems [15].

The cognate aa-tRNA is brought into the A-site by eEF1A:GTP. GTP is hydrolyzed and eEF1A is released. The peptidyltransferase activity of the ribosome transfers the growing polypeptide to the A-site aa-tRNA. Then eEF2:GTP then translocates the peptidyl-tRNAs to provide an open A-site for the next aa-tRNA.

Eukaryotic termination is also related to the process in prokaryotes (Fig. 1-6).

However, it functions with a single class 1 release factor (eRF1) which decodes all of the three stop codons UAA, UAG, and UGA [16;17]. The factor eRF1 facilitates the hydrolysis of the peptidyl-tRNA and thus the release of the polypeptide. Eukaryotes also contain a single class 2 release factor (eRF3) which, unlike prokaryotic termination factors, associates with eRF1 in the absence of the ribosome [18].

The factors involved in ribosome recycling in eukaryotes are unknown. There does not appear to be an RRF ortholog in the eukaryotic genome. Various initiation factors in eukaryotes have antiassociation activity in vitro but it is not known if they are relevant to ribosome recycling and instead used to prevent premature association of the subunits during initiation {5624).

MITOCHONDRIAL PROTEIN SYNTHESIS

Mammalian mitochondria contain their own genome, which encodes 2 rRNAs, 22 tRNAs and 13 proteins (Fig. 1-7) {3298}. The 13 proteins encoded in the mitochondrial genome are vital components of the electron transport chain and ATP synthases, which generate over 90

% of cellular energy. These proteins are synthesized inside the mitochondria and are inserted into the inner membrane. Hence, their synthesis require a unique mitochondrial translational apparatus. The mitochondrial ribosome is a 55S complex with a 39S LSU and as 28S SSU

[19]. The LSU contains a 16S rRNA and 48 proteins. The SSU contains a 12S rRNA and 29 proteins. The mechanism of protein synthesis in mitochondria is thought to be similar to the prokaryotic system. A comparison of these systems is provided in Table 1-1.

16 The initiation of protein synthesis in mitochondria is thought to be comparable to that of prokaryotes except that no mammalian mitochondrial IF1 homolog has been discovered

(Table 1-2). Initiation factor 2 (IF2mt) and initiation factor 3 (IF3mt) were discovered and their functions in in vitro assays were confirmed [20-22]. IF2mt selects the initiator tRNA and promotes its binding to the ribosome. IF3mt promotes the dissociation of the 55S mitochondrial ribosome into subunits and may play additional, less-well-understood, roles in initiation complex formation. Despite the lack of an IF1 homolog in mitochondria there is an insertion in IF2mt that is proposed to play the role of IF1 as discussed in greater detail in

Chapter III. Significantly there are no known 5' or 3' untranslated regions (UTR) or 5' caps on mammalian mitochondrial mRNAs and it is not known how the mitochondrial mRNAs are positioned for initiation [23].

Mitochondria contain all of the factors known to be required for elongation in prokaryotes; therefore, the basic steps in the process are, proposed to proceed via the same mechanisms. Mitochondrial EF-Tu and EF-Ts were discovered and assayed for their relative activities [24-26]. Furthermore mitochondrial EF-G has also been characterized and is proposed to play the same role in translocation as the bacterial factor in protein synthesis

[27]. A unique problem in the mitochondrial system is the diverse structures of the tRNAs, some of which lack entire loops which are strongly conserved across all kingdoms. The mitochondrial EF-Tu has to overcome these deficiencies and still deliver the appropriate aa- tRNA to the proper codon.

Mitochondrial termination and recycling have not been explicitly studied but two factors homologous to the prokaryotic factors RF1 and RRF have been detected in silico and

17 mitochondrial RRF (RRFmt) was cloned and expressed in E. coli [28]. No homologs of the bacterial RF2 and RF3 have yet been discovered in mitochondria.

19 REFERENCES

[1] Moazed D, Samaha RR, Gualerzi C, & Noller HF (1995) Specific protection of 16 S rRNA by translational initiation factors. J. Mol. Bio., 248, 207-210.

[2] Carter AP, Clemons J, Brodersen DE, Morgan-Warren RJ, Hartsch T, Wimberly BT, & Ramakrishnan V (2001) Crystal Structure of an Initiation Factor Bound to the 30S Ribosomal Subunit. Science, 291, 498-501.

[3] Laursen BS, Sorensen HP, Mortensen KK, & Sperling-Petersen HU (2005) Initiation of Protein Synthesis in Bacteria. Microbiol. Mol. Biol. Rev., 69, 101-123.

[4] Croitoru V, Semrad K, Prenninger S, Rajkowitsch L, Vejen M, Laursen BS, Sperling- Petersen HU, & Isaksson LA (2006) RNA chaperone activity of translation initiation factor IF1. Biochim., 88, 1875-1882.

[5] Luchin S, Putzer H, Hershey J, Cenatiempo Y, Grunberg-Manago M, & Laalami S (1999) In vitro study of two dominants inhibitory GTPase mutants of E. coli translation initiation factor IF2: Direct evidence that GTP hydrolysis is necessary for factor recycling. J. Biol. Chem., 274, 6074-6079.

[6] Fabbretti A, Pon CL, Hennelly SP, Hill WE, Lodmell JS, & Gualerzi CO (2007) The real-time path of translation factor IF3 onto and off the ribosome. Mol. Cell, 25, 285- 296.

[7] Boni IV (2006) Diverse molecular mechanisms for translation initiation - In prokaryotes. Mol. Bio., 40, 658-668.

[8] Ito K, Uno M, & Nakamura Y (2000) A tripeptide 'anticodon' deciphers stop codons in messenger RNA. Nature, 403, 680-684.

[9] Poot R, Tsareva NV, Boni I, & vanDuin J (1997) RNA folding kinetics regulates translation of phage MS2 maturation gene. PNAS, 94, 10115.

[10] HIROKAWA GO, Nijama RM, Raj VS, Kaji H, Igarashi K, & Kaji A (2005) The role of ribosome recycling factor in dissociation of 70S ribosomes into subunits. RNA, 11, 1317-1328.

[11] Hirokawa G, Demeshkina N, Iwakura N, Kaji H, & Kaji A (2006) The ribosome- recycling step: consensus or controversy? Trends in Biochem. Sci., 31, 143-149.

[12] Ramakrishnan V (2002) Ribosome structure and the mechanism of translation. Cell, 108, 557-572.

[13] Spahn CM, Beckmann R, Eswar N, Penczek PA, Sali A, Blobel g, & Frank J (2001) Structure of the 80S ribosome from Saccharomyces cerevisiae--tRNA-ribosome and subunit-subunit interactions. Cell, 107, 373-386.

20 [14] Nissen P, Hansen J, Ban N, Moore PB, & Steitz TA (2000) The Structural Basis of Ribosome Activity in Peptide Bond Synthesis. Science, 289, 920-930.

[15] Triana-Alonso FJ, Chakraburtty K, & Nierhaus KH (1995) The elongation factor 3 unique in higher fungi and essential for protein biosynthesis is an E site factor. J. of Biol. Chem., 270, 20473-20478.

[16] Frolova L, Le G, X, Rasmussen HH, Cheperegin S, Drugeon G, Kress M, Arman I, Haenni AL, Celis JE, Philippe M, & . (1994) A highly conserved eukaryotic protein family possessing properties of polypeptide chain release factor. Nature, 372, 701- 703.

[17] Dontsova M, Frolova L, Vassilieva J, Piendl W, Kisselev L, & Garber M (2000) Translation termination factor aRF1 from the archaeon Methanococcus jannaschii is active with eukaryotic ribosomes. FEBS Lett., 472, 213-216.

[18] Kapp LD & Lorsch JR (2004) The Molecular Mechanics of Eukaryotic Translation. Ann. Rev. Biochem., 73, 657-704.

[19] Sharma MR, Koc EC, Datta PP, Booth TM, Spremulli LL, & Agrawal RK (2003) Structure of the Mammalian mitochondrial ribosome reveals an expanded functional role for its component proteins. Cell, 115, 97-108.

[20] Liao H-X & Spremulli LL (1991) Initiation of protein synthesis in animal mitochondria: Purification and characterization of translational initiation factor 2. J. Biol. Chem., 266, 20714-20719.

[21] Liao H-X & Spremulli LL (1990) Identification and initial characterization of translational initiation factor 2 from bovine mitochondria. J. Biol. Chem., 265, 13618- 13622.

[22] Koc EC & Spremulli LL (2002) Identification of mammalian mitochondrial translational initiation factor 3 and examination of its role in initiation complex formation with natural mRNAs. J. Biol. Chem., 277, 35541-35549.

[23] Scheffler I (1999) Mitochondria. Wiley-Liss, Inc., New York.

[24] Woriax V, Burkhart W, & Spremulli LL (1995) Cloning, sequence analysis and expression of mammalian mitochondrial protein synthesis elongation factor Tu. Biochim. Biophys. Acta, 1264, 347-356.

[25] Schwartzbach C & Spremulli LL (1989) Bovine mitochondrial protein synthesis elongation factors: Identification and initial characterization of an elongation factor Tu-elongation factor Ts complex. J. Biol. Chem., 264, 19125-19131.

[26] Xin H, Woriax VL, Burkhart W, & Spremulli LL (1995) Cloning and expression of mitochondrial translational elongation factor Ts from bovine and human liver. J. Biol. Chem., 270, 17243-17249.

21 [27] Bhargava K, Templeton PD, & Spremulli LL (2004) Expression and characterization of isoform 1 of human mitochondrial elongation factor G. Protein Expr. Purif., 37, 368-376.

[28] Zhang Y & Spremulli LL (1998) Identification and cloning of human mitochondrial translational release factor 1 and the ribosome recycling factor. Biochim. Biophys. Acta, 1443, 245-250.

CHAPTER II

STRUCTURAL AND BIOINFORMATIC STUDIES OF MITOCHONDRIAL

RIBOSOMES

INTRODUCTION

Mitochondrial ribosomes differ quite significantly then their cytoplasmic counterparts. The relative size and shape of ribosomes have many similarities, but the protein to ribosomal

RNA (rRNA) content is quite variable. The primary architecture of the ribosomal machinery remains the same with a large subunit (LSU) and a small subunit (SSU). Mitochondrial ribosomes have been imaged using cryo-electronmicroscopy (cryoEM) and scanning electron microscopy (SEM) but, as of yet, none has been crystallized. To do a thorough analysis of mitochondrial ribosomes one needs to compare it to the benchmark structure, the crystal structure of the 70S Thermus thermophilus ribosome. To understand ribosomes from different sources, it is important to identify the rRNA and protein composition. Identification of the rRNAs is generally straight forward due to the significant conservation of their structures. However, the identification of the spectrum of ribosomal proteins present requires a more labor intense approach. A proteomics and bioinformatics approach to analyze the protein composition of mitochondrial ribosomes has been carried out on Bos taurus and

Leishmania tarentolae ribosomes.

Prokaryotic Ribosomes

As previously stated the prokaryotic ribosome is the benchmark that other ribosomes are compared. This ribosome consists of a 70S particle with two distinct subunits, the 30S SSU and the 50S LSU. Generally speaking the prokaryotic ribosome is rRNA decorated with some ribosomal proteins. By weight, the prokaryotic ribosome is 2/3 rRNA and 1/3

ribosomal proteins and has a total mass 2.3 MDa. The hallmark structural features of this ribosome include the head, neck, shoulder, spur and platform of the 30S SSU. In the 50S

LSU the L1 stalk, central protuberance and the L7/L12 stalk have been defined (Fig. 2-1)

[1;2].

The 70S ribosome contains 3 rRNAs. The 23S and the 5S rRNAs are found in the

LSU and the 16S rRNA is found in the SSU [3] (Fig. 2-2). These rRNAs form a complex secondary and tertiary structure composed of many distinct helices. The 23S rRNA in the

LSU contains 101 helices organized into 5 domains, of which domain V is the site of peptidyl transferase center (PTC). The 5S rRNA in the LSU contains 5 helices [4]. In the

SSU the 16S rRNA contains 45 helices which are organized into 4 domains: the 5' domain, the central domain, the 3' major domain and the 3' minor domain. The 3' minor domain contains helix 44 in which nucleotides A1492 and A1493 reside. These nucleotides “flip out” to inspect that codon:anticodon interaction of the tRNA:mRNA complex in the A-site and result in structural rearrangements of the rRNA. These nucleotides also rearrange upon the binding of IF1 and the tRNAs [1]. The 16S rRNA in the SSU also contains the complementary Shine-Dalgarno sequence which facilitates positioning of the mRNA for the initiation of translation.

The ribosomal proteins bind to the rRNA scaffold. There are a total of 54 proteins in the 70S complex. The 30S SSU has 21 proteins and the 50S LSU has 33. The locations of nearly all of the ribosomal proteins have been mapped. The intersubunit space is primarily mediated by rRNA:rRNA interactions. One protein occupies the intersubunit space, SSU protein S12. Other notable proteins in the ribosome complex is are L7/L12 stalk, which has been implicated in the ribosome's GTP hydrolysis function, and is spatially close to the

sarcin/ricin loop (SRL) on the 23S LSU rRNA. The 50S large subunit contains the peptidyltransferase center and the peptide exit channel.

Mammalian Mitochondrial Ribosomes

Mammalian mitochondrial ribosomes are 55S particles, consisting of a 28S SSU and a 39S

LSU. The composition of the mitochondrial ribosome is very different than that of the prokaryotic ribosome. These ribosomes are composed of approximately 2/3 protein and 1/3

RNA, which is the opposite of the ratio found in prokaryotes. The mass of the mitochondrial

55S ribosomal particle is 2.7 MDa, which is slightly more than the prokaryotic ribosome.

Despite the greater mass, the mitochondrial ribosome has a lower sedimentation coefficient which suggests that it is "fluffier" (i.e. more porous) than the prokaryotic ribosome (Fig. 2-

3A) [5]. Mitochondrial ribosomes also contain the hallmark structural features found in the prokaryotic ribosomes. The 28S SSU clearly has the characteristic head, neck, shoulder, spur and platform (Fig. 2-3B). The 39S LSU also contains the L1 stalk, central protuberance and the L7/L12 stalk as previously defined (Fig. 2-3C).

The 55S mitochondrial ribosome contains 2 rRNAs. The 28S SSU contains a 12S rRNA and the 39S LSU contains a 16S rRNA. In this organelle the rRNAs for both subunits are dramatically truncated relative to the prokaryotic rRNAs (Figs. 2-4) [6;7]. These rRNAs do not contain random deletions but rather distinct regions of secondary structures are absent.

The mitochondrial equivalent of the peptidyltransferase center in the LSU and helix 44 in the

SSU are conserved.

To complement the decrease in rRNA the mitochondrial ribosome has a greater number of ribosomal proteins. This increase in ribosomal proteins was initially thought to replace regions of missing rRNA but cryoEM analysis has revealed that many of those regions are left devoid of both [5]. There are 77 ribosomal proteins that make up the full complement of ribosomal proteins. The mitochondrial 28S SSU has 29 proteins of which 14 have bacterial homologs: S2, S5, S6, S7, S9, S10, S11, S12, S14, S15, S16, S17, S18 and

S21 [6]. The mitochondrial 39S LSU has 48 proteins of which 28 have bacterial homologs:

L1, L2, L3, L4, L7/L12, L9, L10, L11, L13 - L24, L27, L28, L30 and L32 - L36. The additional mitochondrial ribosomal proteins not listed have unknown functions but some may be involved in the interaction at the interface of the inner membrane of the mitochondrion and the ribosome during translation. A striking difference between the intersubunit space in the mitochondrial ribosome and the prokaryotic ribosome is that the mitochondrial ribosome is dominated by protein:protein, as opposed to rRNA:rRNA interactions [5].

Leishmania tarentolae Mitochondrial Ribosomes

Leishmania tarentolae is an organism in the family of trypanosomitids. Members of this family include Trypanosoma brucei and Trypanosoma cruzi, which are known to cause

African sleeping sickness and South American Chagas disease, respectively. To combat these diseases and develop therapies, the mitochondrial (kinetoplast) ribosomes are under study. The L. tarentolae mitochondrial ribosome is studied in lieu of T. brucei and T. cruzi because it is non-pathogenic and relatively easy to cultivate. Interestingly, there have been varying reports of antibiotic sensitivities against the ribosomal particle. Some reports

indicated susceptibility similar to that of the eukaryotic cytoplasmic ribosome [8;9] and others report similarities to the prokaryotic ribosome.

A recent study revealed that the L. tarentolae mitochondrial ribosome was a 50S particle with a 30S SSU and a 40S LSU [10]. Based on EM studies of the 50S particle it has been suggested to have similar structural features to both the prokaryotic and mammalian mitochondrial ribosome such as the head and body of the SSU and the L1 stalk, central protuberance and the L7/L12 stalk on the LSU (Fig. 2-5A/B).

The full protein complement of the 50S particle has not been determined but the rRNAs have been characterized. The 30S SSU contains a 9S rRNA [11]. This rRNA lacks many of the domains relative to the prokarytic 16S SSU rRNA but retains the regions complementary to the A-site in the truncated helix 44 (Fig. 2-6). The 40S LSU contains a

12S rRNA which is also significantly shortened but nevertheless the peptidyl transferase center is conserved. A complete secondary structure for this rRNA is not available.

In addition to the 30S SSU and the 40S LSU, other sedimentation particles have been observed and characterized. The unique sedimentation particles identified are 70S, 65S and

45S. The 65S particle is believed to be a dimer of 40S LSU particles. The 45S particle was characterized as having only SSU 9S rRNA and was therefore proposed to be a functional intermediate of the 30S SSU (Fig. 2-5C). This particle was subsequently renamed the 45S

SSU* [12]. The 70S particle was determined to be a dimer of the 45S SSU* particles. The biological relevance of these particles are unknown and are quite unique to the trypanosomes.

Developing a better understanding of the trypanosome kinetoplast ribosome will help to potentially develop antibiotics and therapies to combat this organism. Therefore a bioinformatics analysis of the mass spectrometry data of the protein composition of the 45S

SSU* particle was thoroughly analyzed and implications for organellar evolution are discussed.

MATERIALS AND METHODS

Materials: Chemicals were purchased from Sigma and Fisher Scientific. Oligonucleotide primers were manufactured at the Nucleic Acids Core Facility (University of North

Carolina). All plasmids were purified using QIAGEN Qiaprep® Miniprep Handbook as described by the manufacturer. Restriction enzymes were purchased from New England

Biolabs. Digested plasmids and PCR products were purified using the Qiagen Gel

Extraction/PCR Purification Kit.

Cloning of Mammalian Mitochondrial Ribosomal Proteins L13, L15, S2 and S11: The cDNAs encoding human mitochondrial ribosomal proteins L13, L15, S2 and S11 (NCBI

GeneID numbers 28998, 29088, 51116 and 64963, respectively) were ordered from

American Type Culture Collection (ATCC). The cDNA were amplified out of the stock pCMV-SPORT6 vector using the primers listed in Table 2-1. The PCR products were digested with NdeI and XhoI along with purified pET(+)21b plasmid. The PCR product was ligated into the pET(+)21b plasmid at 16 °C for 16 hours under standard ligation conditions.

The ligated plasmid was then transformed into Stratagene’s E. coli DH5α, sequenced and subcloned into Stratagene E. coli BL21 (DE3) RIL cells.

Growth and Expression of Mammalian Mitochondrial Ribosomal Proteins L13, L15, S2 and

S11: The E. coli BL21 RIL cells containing the genes for these proteins were used to inoculate 20 mL of LB containing 100 μg/mL ampicillin and 50 μg/mL kanamycin. This sample (20 mL) is used to inoculate 2 L LB media containing the appropriate antibiotic.

These cells are grown with vigorous shaking (about 200 rpm) at 37 ° C to an A595 of 0.6 to

0.8. Expression was induced by the addition of 0.1 mM IPTG and is carried out at 27 °C overnight. Cells are harvested by centrifugation in 1 L bottles in a Sorvall RC3B centrifuge at

5,000 rpm (7,300 x gmax) for 25 min at 4 ° C. The cell pellets are resuspended in buffer containing 20 mM Tris-HCl (pH 7.6) and 10 mM MgCl2 and collected by centrifugation in

50 mL falcon tubes in a RC5B Sorvall centrifuge using the SS34 rotor at 7,500 rpm (6,700 x gmax) for 25 min at 4 ° C. The supernatant was discarded and residual liquid carefully drained from the cell pellets.

Ni-NTA Purification of Mammalian Mitochondrial Ribosomal Proteins L13, L15, S2 and

S11: The purification of these proteins followed the QIAexpressionist denaturing protocol from Qiagen. If cell pellets were frozen they were first thawed for 15 min on ice and then resuspended in buffer B (100 mM NaH2PO4, 10 mM Tris-HCl, 8 M urea, adjusted pH to 8.0) at 5 mL per gram wet weight. Cells were stirred for 60 min at room temperature to lyse the cells. The lysate was centrifuged at 10,000 x g for 30 min at room temperature to pellet the cellular debris. The lysate was placed in a clean 50 mL orange cap tube and 1 mL of a 50%

Ni-NTA slurry in Buffer B was added and the solution was rocked for 60 min at room temperature. The lysate-resin was carefully added to a Qiagen 5 mL polypropolyene column and washed twice with 4 mL Buffer C (100 mM NaH2PO4, 10 mM Tris-HCl, 8 M urea,

adjusted pH to 6.3). Protein was eluted from the Ni-NTA resin in 0.5 mL fractions of Buffer

E (100 mM NaH2PO4, 10 mM Tris-HCl, 8 M urea, adjusted pH to 4.5), following a 30 min incubation for each fraction. The proteins were then concentrated by placing all fractions into a dialysis tube and layering it with PEG 8000 to a final volume ~ 200 μL.

Antibody Production against Mammalian Mitochondrial Ribosomal Proteins L13, L15, S2 and S11: Using the immunoadjuvant TiterMax Gold (R) a 250 μL emulsion of 125 μL

TiterMax to 125 μL of concentrated protein (~250 μg). Each emulsion was divided into two

100 μL fractions per protein in sterile dispoable syringes and delivered to the UNC

Veterinary Services where 4 New Zealand White Rabbits were used to produce the antibodies. The rabbits were injected at 0, 4 and 8 weeks and blood was collected to test the serum at each injection time. The final bleed was performed at 16 weeks and the rabbits were exsanguinated. The antibodies were verified by Western blotting as described [13;14]

Bioinformatics Analysis of the Protein of the L. tarentolae 45S SSU* particle: The peptide fragments obtained from the trypsin digested L. tarentolae mitochondrial 45S SSU* particles

LC/MS/MS proteomic analysis were used in a BLASTP query against bacteria, yeast and eukaryota databases. E-values smaller than 1e-03 were considered significant.

RESULTS

Mammalian Mitochondrial Ribosomes

If one could capture the mitochondrial ribosome with factors intact, we could learn a tremendous amount about protein synthesis in this organelle. To this end, and based on the assumption that mitochondrial ribosomal proteins with prokaryotic homologs would be located on the mitochondrial ribosome in the same relative orientation to that of their prokaryotic counterparts on the prokaryotic ribosome, the crystal structures of the prokaryotic subunits were examined for ribosomal proteins that were accessible and would not interfere in factor binding. After a thorough analysis of both the Haloarcula marismortui

50S LSU and the Thermus thermophilus 30S SSU, four ribosomal proteins with mitochondrial homologs fit the criteria: L13, L15, S2 and S11 (Fig. 2-7).

As described in Materials and Methods, the mammalian mitochondrial ribosomal proteins were all cloned and expressed successfully (Fig. 2-8A). Antibody production proceeded without incident and the antibodies were very specific for a single antigen in the mitochondrial ribosomes (Fig. 2-8B). Interestingly in the case of MRPS11, the antibody revealed isoforms of this protein not previously observed. This isoform was confirmed by using the MRPS11 gene as a query in a BLASTP search against the Homo sapiens EST database at NCBI. Further analysis and experimentation with these antibodies was not performed due to a loss of funding and a refocusing of effort on other projects.

Leishmania tarentole Ribosomal Analysis

The LC/MS/MS peptide fragment data of the trypsin digested 45S SSU* Leishmania tarentole ribosomal particle was first analyzed by MASCOT against the Leishmania major

Gene Database, http://www.genedb.org/genedb/leish/. A representative table of proteomic hits is shown in Table 2-2. A more thorough analysis of the initial hits was performed to determine which hits were significant Leishmania SSU ribosomal proteins.

Seven homologs of small subunit ribosomal proteins identified earlier in the analysis of the mixed 45S/50S fraction [10] (S5, S9, S11, S15, S16, S17, S18) and three additional components (S6, S8, MRPS29) have been found in the 45S SSU* complexes. The degree of sequence conservation varies greatly among these polypeptides. Thus, the identification of the S5, S8, S17 and MRPS29 homologs was based on BLAST searching that yielded highly significant E-values smaller than the conventional 1e-03 cut-off value.

Although the BLAST E-scores for S11 similarity hits were not significant (E ≥ 0.51), this protein was found to contain the RNase H-like SCOP domain (E = 3e-07) present in the ribosomal protein family including S11 and L18, as well as the NCBI conserved domain

RpsK (representing ribosomal protein S11) (E = 0.0001). Similarly, the best BLAST E-score for the S9 homolog was only 0.12. Nonetheless, the identification of this protein is reliable: it was based on finding ribosomal protein S5 domain 2-like SCOP domain (E = 4e-16) present in the S5 and S9 ribosomal protein families; and finding the Ribosomal_S9 Pfam domain (E

=0.0008).

The S15 homolog was identified based on the similarity with putative mitochondrial

S15 proteins from Theileria annulata (E=0.009) and Schizosaccharomyces pombe (E=0.037)

and the presence of the S15/NS1 RNA binding SCOP domain (E=0.014). The identification of the putative S6, S16 and S18 homologs was not strongly supported. The putative S6 homolog was identified only by a weak similarity to several mitochondrial S6 proteins. The

S16 protein was moderately similar to fungal MRPS24 (S16 family) (E ≥ 0.038). The S18 protein was found to contain the SCOP ribosomal protein S18 domain (1.80e-02) but demonstrated only a weak similarity to the corresponding eubacterial proteins.

To summarize from above, the proteomics analysis indicated the presence of 10 candidates for SSU proteins in the trypanosome SSU* complexes. This number is somewhat smaller than the 14 prokaryotic homologs found in the bovine mitochondrial 28S SSU. With the exception of S8, all of the homologs observed in Leishmania are also observed in mammalian mitochondrial ribosomes. In general the similarity between the Leishmania SSU homologs and the bacterial proteins is somewhat higher than the similarity observed between the Leishmania SSU homologs and homologs observed in other mitochondrial translational systems. This observation suggests that different constraints have been placed on the mitochondria of different organisms during evolution leading to considerable divergence in the ribosomal proteins.

To further this analysis, the cytoplasmic ribosomal proteins of Leishmania major,

Saccharomyces cerevisiae, and Homo sapians and the mitochondrial ribosomal proteins identified in the 45 SSU* analysis were used as a query in a BLASTP analysis against the all eubacteria in the NCBI database. The expected result is homology between the bacterial ribosomal proteins with the mitochondrial proteins from these organisms. This expectation is based on the assumption that the mitochondrial ribosomal proteins are under less selective pressure to evolve from their bacterial protein progenitors. Quite unexpected was the result

that the cytoplasmic ribosomal proteins of Leishmania major are more like bacterial ribosomal proteins than the Leishmania major mitochondrial ribosomal proteins. A summary of the findings are listed in Table 2-3.

DISCUSSION

Mammalian Mitochondrial Ribosomes

Although further analysis of the human mitochondrial ribosomal protein antibodies was not performed these factors will potentially be used in future experiments. They were initially conceived to develop a ribosome capture assay. This assay would have used these antibodies by immobilizing them on beads and mitochondrial lysates would have been washed over them, potentially capturing the ribosomal complexes. In theory, these complexes could contain novel proteins involved in the mitochondrial translation machinery. These complexes would have been enriched and analyzed using a proteomics approach, similar to that of the Leishmania 45S SSU* analysis.

Leishmania tarentole Ribosomal Analysis

The BLASTP analysis performed on the 45 SSU* particle reflects the relationship between the mitochondrial, cytoplasmic and bacterial ribosomal proteins. The classical view of the endosymbiotic origin of mitochondria leads to the idea that components of the translational system in these organelles should be more closely related to those of bacteria than are

components of the cytoplasmic translational system. To examine this issue in Leishmania, we have BLASTed the Leishmania mitochondrial and cytoplasmic ribosomal proteins against all eubacteria in the NCBI database. This analysis was carried out with 5 of the ribosomal proteins, which are represented in all three categories of ribosomes (prokaryotic, cytoplasmic and mitochondrial). As indicated in Table 2-3, for 4 out of the 5 proteins, the Leishmania cytoplasmic ribosomal protein was more closely related to its bacterial homolog than was the corresponding mitochondrial homolog. This observation is somewhat surprising and indicates that the evolutionary divergence of mitochondrial ribosomal proteins in the trypanosomes has been surprisingly rapid. To place this analysis in context, a similar analysis was carried out with yeast and mammalian cytoplasmic and mitochondrial ribosomal proteins. Of the 5 proteins for which an analysis could be carried out, 3 yeast mitochondrial ribosomal proteins are more closely related to bacteria than the corresponding cytoplasmic homolog. In mammals, only 4 proteins could be examined in this way due to the loss of S8.

Of these, two mitochondrial homologs are closer to bacteria, one protein has the same

BLAST score and one cytoplasmic ribosomal protein is closer to bacteria than the mitochondrial homolog. This analysis suggests that the Leishmania mitochondrial ribosomal proteins have diverged more from bacteria than have most other mitochondrial ribosomal proteins. However, surprisingly, mitochondria from a variety of very diverse organism have some mitochondrial ribosomal proteins that have become more distant from bacteria than the corresponding cytoplasmic ribosomal proteins. This observation reflects a more rapid evolution of mitochondrial components to that of the cytoplasmic components.

REFERENCES

[1] Lithgow T (2000) Targeting of proteins to mitochondria. FEBS Lett., 476, 22-26.

[2] Wilson DN, Blaha G, Connell SR, Ivanov PV, Jenke H, Stelzl U, Teraoka Y, & Nierhaus KH (2002) Protein synthesis at atomic resolution: mechanistics of translation in the light of highly resolved structures for the ribosome. Curr. Protein Pept. Sci., 3, 1-53.

[3] Maguire BA & Zimmermann RA (2001) The ribosome in focus. Cell, 104, 813-816.

[4] Lu M & Steitz TA (2000) Structure of Escherichia coli ribosomal protein L25 complexed with a 5S rRNA fragment at 1.8-A resolution. PNAS, 97, 2023-2028.

[5] Sharma MR, Koc EC, Datta PP, Booth TM, Spremulli LL, & Agrawal RK (2003) Structure of the Mammalian mitochondrial ribosome reveals an expanded functional role for its component proteins. Cell, 115, 97-108.

[6] Koc EC, Burkhart W, Blackburn K, Moseley A, & Spremulli LL (2001) The small subunit of the mammalian mitochondrial ribosome: Identification of the full complement of ribosomal proteins present. J. Biol. Chem., 276, 19363-19374.

[7] Koc EC, Burkhart W, Blackburn K, Schlatzer D.M., Moseley A, & Spremulli LL (2001) The large subunit of the mammalian mitochondrial ribosome: Analysis of the complement of ribosomal protein present. J. Biol. Chem., 276, 43958-43969.

[8] Nebohacova M, Maslov DA, Falick AM, & Simpson L (2004) The Effect of RNA Interference Down-regulation of RNA Editing 3'-Terminal Uridylyl Transferase (TUTase) 1 on Mitochondrial de Novo Protein Synthesis and Stability of Respiratory Complexes in Trypanosoma brucei. J. Biol. Chem., 279, 7819-7825.

[9] Horvath A, Nebohacova M, Lukes J, & Maslov DA (2002) Unusual polypeptide synthesis in the kinetoplast-mitochondria from Leishmania tarentolae. Identification of individual de novo translation products. J. Biol. Chem., 277, 7222-7230.

[10] Maslov DA, Sharma MR, Butler E, Falick AM, Gingery M, Agrawal RK, Spremulli LL, & Simpson L (2006) Isolation and characterization of mitochondrial ribosomes and ribosomal subunits from Leishmania tarentolae. Mol. Biochem. Parasitol., 148, 69-78.

[11] de lC, V, Lake JA, Simpson AM, & Simpson L (1985) A minimal ribosomal RNA: sequence and secondary structure of the 9S kinetoplast ribosomal RNA from Leishmania tarentolae. PNAS, 82, 1401-1405.

[12] Maslov DA, Spremulli LL, Sharma MR, Bhargava K, Grasso D, Falick AM, Agrawal RK, Parker CE, & Simpson L (2007) Proteomics and electron microscopic characterization of the unusual mitochondrial ribosome-related 45S complex in Leishmania tarentolae. Mol. Biochem. Parasitol., 152, 203-212.

[13] Ma J & Spremulli LL (1996) Expression, purification and mechanistic studies of bovine mitochondrial translational initiation factor 2. J. Biol. Chem., 271, 5805-5811.

[14] Spencer AC & Spremulli LL (2005) The interaction of mitochondrial translational initiation factor 2 with the small ribosomal subunit. Biochim. Biophys. Acta - Proteins & Proteomics, 1750, 69-81.

CHAPTER III

ANALYSIS OF THE ROLES OF THE NKXD MOTIF AND THE INSERTION

DOMAIN IN MAMMALIAN MITOCHONDRIAL INITIATION FACTOR 2 INTRODUCTION

Structure of IF2: Mammalian mitochondria have a translational system responsible for the synthesis of 13 proteins required for oxidative phosphorylation in the inner membrane. In this system, mitochondrial IF2 (IF2mt) stimulates the binding of fMet-tRNAfMet to the 28S small subunit of the mitochondrial ribosome in the presence of GTP and mRNA [1;2]. As described in Chapter I, IF2 is a guanine nucleotide binding protein (G-protein or GTPase) and recruits fMet-tRNAfMet into the P-site of small subunit of the ribosome forming the initiation complex to begin protein synthesis [3].

The secondary structure of IF2mt is based on the six domain model of E. coli IF2 [4]

(Fig. 3-1A). In E. coli IF2 the role of domain I is not known. The two isoforms of IF2,

IF2α, consisting of domains I - VI, and IF2β, which lacks domain I, are both functionally active in vivo and in vitro [5;6]. Domain II was shown to be directly involved in the binding of this factor to the 30S small subunit (SSU) of the E. coli ribosome [7]. Interestingly, an IF2 variant lacking both domains I and II can bind 30S subunits only in the presence of initiation factor 1 (IF1) [4], suggesting that the N-terminus of IF2 provides an important part of the

SSU binding activity of this factor. It has also been proposed that fMet-tRNAfMet stabilizes the interaction of IF2 with the ribosome for factors that lack the N-terminal domains [8]. No specific role of domain III has been demonstrated beyond its role in connecting domains II and IV [9]. Domain IV is the G-domain, where the GTPase activity of this factor is located, and in all IF2s this domain is very highly conserved. This domain has also been implicated as a potential metabolic sensor [10]. Domain V has no known function but has been proposed that, together with IF1, it could mimic the structure of elongation factor G at the A-

52 site of the ribosome [11]. Domain VI is divided into two subdomains C1 and C2, the latter of which is directly involved in interacting with the fMet-tRNAfMet [12-16]. IF2mt is characterized as having four domains equivalent to domains III – VI of the E. coli factor. It has an insertion between domains V and VI that has a unique function in this factor [17].

The 3-dimensional structure of IF2 is not known but the structure of an archaeal homolog, Methanobacterium thermoautotrophicum aIF5B, which consists of domains IV through VI, reveals a chalice like structure with domains IV - VI C1 forming the cup and VI

C2 forming the base [18] (Fig. 3-1B). The N-terminal region of IF2 was studied using NMR and CD analysis and was found to have a well ordered structure even though this factor can function with these domains truncated [19].

The G-Domain of IF2: Of particular interest is domain IV, the G-domain. In IF2mt the G- domain is 66 % identical to the G-domain of E. coli IF2 [20] and 43 % identical to the G- domain M. thermoautotrophicum aIF5B. It shares a common structural core with the G- domains of other G proteins, including the G-domain of EF-Tu and EF-G [11;21]. A number of G-proteins are involved in protein synthesis and are required in all 3 major steps of translation: initiation, elongation and termination.

G-proteins are characterized by 5 conserved motifs [22] (Fig. 3-2A). G-1

(GXXXXGKS/T) coordinates the α- and β-phosphates of the guanine nucleotide. G-2 (XTX) is thought to coordinate a Mg2+ ion. G-3 (DXXG) aids in the coordination of the Mg2+ and the γ-phosphate of GTP. G-4 (NKXD) and G-5 (SAL/K) in IF2s are important in selecting and positioning the guanine ring. Mutations in the motifs of the G-domain often lead to cold

54 sensitive phenotypes in E. coli IF2 [23;24] or can be lethal [21;25] illustrating of the importance of the G-domain for the function of IF2.

The NKXD motif interacts with the guanine ring through a hydrogen bond between the amine of asparagine (N) and the O6 position of the guanine ring and through a hydrogen bond between the carboxyl group of the aspartatic acid (D) and the 2’ amine of the guanine ring (Fig. 3-2B). In aIF5B, the X residue, isoleucine (I), is pointing away from the guanine ring and is largely buried in the G-domain. The X residue in the NKXD motif is not truly random but partially conserved in homologous proteins (Fig. 3-3). In IF2mt this residue is present as Cys290. In IF2 the X residue is generally methionine (M), valine (V), cysteine (C) or isoleucine (I). In elongation factor EF-Tu, this residue is generally valine (V), methionine

(M), isoleucine (I) or leucine (L). While for release factor 3 it is leucine (L), methionine

(M), tyrosine (Y) or tryptophan (W). In the work described below the Cys290 in bovine

IF2mt was mutated to serine (S), alanine (A) and isoleucine (I) and the effects of these changes on the activity and stability of this protein were assessed.

The Insertion Region of IF2mt: An interesting characteristic of IF2mt is the presence of an insertion sequence between domains V and VI (Fig. 3-1), which is strongly conserved amongst IF2mt across various species [17]. In bovine IF2mt this sequence consists of 37 amino acids which are very hydrophilic with a high percentage of acidic and basic residues

(Fig. 3-4B). A partial structural model of IF2mt based on the pdb coordinates of M. thermoautotrophicum aIF5B was generated by Swiss-Model [26]. The model consists of the conserved regions of domain IV and domains V-VI C2 and indicates that there is a significant degree of structural conservation between IF2mt and aIF5B. The most obvious

57 difference between the two structures lies in the insertion region of IF2mt. The model predicts that this region loops out and protrudes from the surface of the molecule (Fig. 3-4B).

Several models for the binding of IF2 to the small ribosomal subunit [8;27;28] would place the insertion near the A-site where IF1 binds [27;29] (Fig. 3-5).

No homolog of IF1 has been detected in mitochondria from any organism ranging from yeast to man, despite database searching and biochemical probing [30]. IF1 is an essential factor in both prokaryotes and in the eukaryotic cell cytoplasm (eIF1A) and it would be surprising if no equivalent factor were present in mitochondria. The insertion region in

IF2mt is the most reasonable region to serve the role of IF1 for the mitochondrion. Initial experiments on the role of the insertion in IF2mt in translational initiation were investigated by site-directed mutagenesis. Several highly conserved lysines and arginines residues in the insertion were changed to Ala [31] and the resulting mutants showed decreased ability to bind 28S subunits but retained their ability to bind fMet-tRNAfMet. These results indicate that the residues in the insertion may make important ribosomal contacts between this factor and the small ribosomal subunit [17].

In the current work, the role of the insertion domain in IF2mt was examined by removing the entire insertion in IF2mt and replacing it with the equivalent, smaller, region from E. coli IF2 (IF2mtΔ37). In a complementary experiment, the insertion observed in

IF2mt was placed into E. coli IF2 (EcoIF2::37). These hybrid IF2s were assayed for their response to IF1 in initiation complex formation on both E. coli and mitochondrial ribosomes.

Furthermore, in vivo assays by our collaborators in India (Dr. Umesh Varshney) with the wild-type IF2mt and these hybrid factors were evaluated for their ability to rescue an E. coli

IF2 knock-out and a double IF1/IF2 knock-out using a P1 transduction strategy. In E. coli it

59 should be noted that both IF1 and IF2 genes are essential and a knockout of either one is lethal [6;32].

MATERIALS AND METHODS

Materials: Chemicals were purchased from Sigma and Fisher Scientific. Oligonucleotide primers were manufactured at the Nucleic Acids Core Facility (University of North

Carolina). [35S]Met was purchased from Perkin Elmer. [35S]fMet-tRNAfMet was prepared as described [33;34]. E. coli and mitochondrial ribosomes and ribosomal subunits were prepared as described [34-36]. E. coli IF3 and mammalian mitochondrial IF3 (IF3mt) were prepared as described [30]. The previously cloned plasmids carrying the mature form of the

IF2mt coding sequence [33] in the pET21b plasmid, E. coli factors in pET21b plasmids and all other plasmids were purified using QIAGEN Qiaprep® Miniprep Handbook as described by the manufacturer. Restriction enzymes were purchased from New England Biolabs.

Digested plasmids and PCR products were purified using the Qiagen Gel Extraction/PCR

Purification Kit.

Site-Directed Mutagenesis of the X residue of NKXD motif in IF2mt: Purified plasmid DNA

(pET21b IF2mt) was used for site directed mutagenesis using Stratagene’s QuikChange®

Site-Directed Mutagenesis Kit as described by the manufacturer. The primers used are listed in Table 3-1. The PCR reactions (100 µL) contained 10 µL 10x Pfu Buffer (Stratagene),

~100 ng pET21b IF2mt DNA, 0.5 mM dNTPs, ~250 ng forward primer, ~250 ng reverse primer and 2.5 units of PfuTurbo DNA polymerase. These reactions were run in an Ericomp

60 Thermocycler under the following conditions: Cycle 1 (95 °C, 1 min) x 1, Cycle 2 (95 °C, 1 min; 55 °C, 1 min; 65 °C, 8 min) x 18, Cycle 3 (4 °C, 99 h). The reaction mixtures were transformed into E. coli ER2267 competent cells using Stratagene’s protocols. The presence of the mutation was confirmed by DNA sequencing at the UNC-CH Genome Analysis

Factility. Variant IF2mt plasmids were then transformed into E. coli BL21(DE3) cells, containing a pArgU218 plasmid (provided by Dr. Yamada, Mitsubishi Chemical Corp.,

Yokohama, Japan), for expression.

Subcloning of the IF2mt with Insertion Replaced by the E. coli Domain V/VI Junction: Our collaborators in India, Dr. Umesh Varshney and Rahul Gaur (Bangalore, Indian Institute of

Technology), created a variant of IF2mt with the insertion deleted and replaced with the corresponding sequence in E. coli IF2 (EcoIF2), creating the variant IF2mtΔ37, in a pTrc vector (Fig. 3-6). The IF2mtΔ37 gene was amplified from the pTrc vector by PCR under the same conditions as above, with primers listed in Table 3-1. The IF2mtΔ37 PCR product was digested with NdeI and NcoI and ligated into the expression vector pET24b, that provides kanomycin resistance and provides a C-terminal 6x His tag, under standard ligation conditions (16 h at 16 °C). The ligated pET24b IF2mtΔ37 plasmid was transformed into

Stratagene E. coli DH5α competent cells for plasmid production and Stratagene E. coli BL21

RIL competent cells for expression.

Creation of EcoIF2 with IF2mt Insertion (EcoIF2::37): Using the pET21b EcoIF2 stock previously cloned in our lab, two site directed mutagenesis reactions were performed to insert

PmlI, and BamHI sites into the E. coli IF2 sequence creating silent mutations. These two

63 restriction sites were chosen because they were not present in the gene sequence and they required minimal mutations. The silent mutation primers used are listed in Table 3-1 and the

PCR conditions are the same as the NKXD site-directed mutations. For PmlI a G to A silent mutation in the EcoIF2 gene was created as underlined (1950CACGTG1955). For BamHI two silent mutations were created in the EcoIF2 gene, a T to A and T to C as underlined

(2095GGATCC2100). The sequence to be inserted into the pET21b EcoIF2 was cloned out of

IF2mt with primers listed in Table 3-1, PCR was performed under the same conditions as above, and the PCR product and mutated pET21 EcoIF2 plasmid were digested with PmlI and BamHI and subsequently ligated under standard conditions for 16 h at room temperature. The ligated pET21b EcoIF2::37 plasmid was transformed into Stratagene E. coli DH5α competent cells for plasmid production and then into E. coli BL21(DE3) cells, containing a pArgU218 plasmid.

Growth of Cells, Induction and Preparation of Cell Extracts: A sample of a frozen stock of E. coli BL21 cells carrying the coding region for the variant IF2 in the corresponding pET vector was used to inoculate 20 mL of LB containing the appropriate antibiotics as indicated.

The BL21(DE3) pArgU218 pET21 IF2mt, and all IF2mt NKXD variants, required 100

μg/mL ampicillin and 50 μg/mL kanamycin. The BL21(DE3) pET21 EcoIF2 required 100

μg/mL ampicillin. The BL21(DE3) RIL pET24d IF2mtΔ37 required 34 μg/mL chloramphenicol and 50 μg/mL kanamycin. The BL21(DE3) pArgU218 pET21 EcoIF2::37 required 50 μg/mL kanamycin and 100 μg/mL ampicillin. This bacterial sample (20 mL) was used to inoculate 2 L LB media containing the appropriate antibiotics. These cells were grown with vigorous shaking (about 200 rpm) at 37 ° C to an A595 of 0.6 to 0.8. Expression

64 was induced by the addition of 0.1 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) and was carried out at 27 °C overnight. Cells were harvested by centrifugation in 1 L bottles in a

Sorvall RC3B centrifuge at 5,000 rpm (7,300 x gmax) for 25 min at 4 °C. The cell pellets were resuspended in buffer I (20 mM Tris-HCl (pH 7.6) and 10 mM MgCl2) and collected by centrifugation in 50 mL falcon tubes in a RC5B Sorvall centrifuge using the SS34 rotor at

7,500 rpm (6,700 x gmax) for 25 min at 4 °C. The supernatant was discarded and residual liquid carefully drained from the cell pellets. The pellets were either immediately lysed by sonication or fast frozen in a dry ice-isopropanol bath and stored at -80 °C.

The cell pellet (~ 8 g) was resuspended in five volumes (5 mL/g cells) of ice cold

Buffer II (20 mM Tris-HCl (pH 7.6), 10 mM MgCl2, 1 M KCl, 10 mM imidazole, 5 mM β- mercaptoethanol and 0.1 mM phenylmethylsulfonyl fluoride) by gentle mixing. Lysozyme

(~ 1 mg) was added to the mixture followed by DNase I (to 5 µg/mL). The mixture was placed in an ice bath and sonicated in 1 s bursts with 9 s off for 7 cycles (70 sec total) using a

Branson SONIFER Cell Disruptor 185 with a power setting of 9 (130 W). The solution was then adjusted to a final concentration 1 M NH4Cl by the addition of solid NH4Cl while stirring. The sample was stirred for an additional 15 min at 0 °C after the salt concentration had been adjusted. The solution was subjected to centrifugation in a Beckman L8-70 ultracentrifuge using a Sorvall TFT 50.38 rotor for 1 h at 28,000 rpm (71,000 x gave) at 4 °C.

The supernatant was removed and further processed with Ni-NTA resin as described below.

Purification of IF2 variants by Ni-NTA resin: The sample from sonication (~50 mL) was mixed for 40 min at 4 °C with a 50 % slurry of Ni-NTA agarose resin (0.2 mL resin per g original cell weight) equilibrated in Buffer II. Incubation was carried out on a shaking

65 platform so that the sample was mixed gently during incubation with the resin. This mixture was transferred to a 5 mL polypropylene column (Qiagen). The resin was washed with an excess of Buffer II (at least 50 mL) and the variant IF2 was eluted from the resin using approximately 10-15 mL Buffer III (20 mM Tris-HCl (pH 7.6), 10 mM MgCl2, 50 mM KCl,

150 mM imidazole, 5 mM β-mercaptoethanol and 0.1 mM phenylmethylsulfonyl fluoride) using 1-2 mL/g original cell weight. This step was done by three sequential elutions of about

5 mL each. Each aliquot of Buffer III was incubated with the resin for 15 min at 4 °C before it was collected. The aliquots were combined and dialyzed against a 100-fold excess of

Buffer IV (20 mM Hepes-KOH (pH 7.6), 10 mM MgCl2, 50 mM KCl, 10 % glycerol and 6 mM β-mercaptoethanol) at 4 °C for 1.5 h with one change of buffer after 45 min.

Purification of NKXD IF2mt variants by HPLC: The dialyzed sample from the Ni-NTA resin was passed through a 0.45 μm filter and then injected into a 10 mL loop on the Rainin

Rabbit HPLC injection port. The sample was applied to a TSKgel DEAE 5PW HPLC column (7.5 cm x 7.5 mm, TosoHaas Inc., Japan) equilibrated in Buffer IV at a flow rate of

0.5 mL/min. This step was carried out at 4 °C controlled by the Macintosh computer sitting outside a cold box. The column was washed for at least 20 min with Buffer IV and then developed with a linear gradient (50 mL) from 0.05 to 0.25 M KCl in Buffer IV at a flow rate of 0.5 mL/min. The absorbance at 280 nm was monitored with an ISCO UA6 UV Visible detector and a chart recorder set to the 0.5 scale with a chart speed of 30 cm/h. The flow cell has a path length of 0.5 cm. Fractions (0.5 mL) were collected in Eppendorf tubes using an

ISCO RETRIEVER 500 fraction collector. The Eppendorf tubes had the lids removed and were balanced in 13 x 100 mm glass test tubes. The fractions were fast frozen in a dry ice

66 isopropanol bath and stored at -70 °C. Fractions containing the IF2mt variant were identified by 10 % SDS polyacrylamide gel electrophoresis and staining with Coomassie Blue using 20

μL aliquots of the appropriate fractions. The presence of IF2mt was verified by Western blotting as described [17;33]. The appropriate fractions are pooled accordingly. The IF2mt variants eluted at a KCl concentration of approximately 0.12 M.

Purification of IF2 Variants on a Gravity DEAE-Sepharose Column: The dialyzed sample from the Ni-NTA column was applied to a ~10 ml (5.7 x 1.5 cm) DEAE-Sepharose anion exchange column. The column was washed with 100 mL Buffer IV containing 100 mM KCl until the absorbance at 280 nm was near the baseline. A linear gradient (40 mL) from 100 mM - 300 mM KCl in Buffer IV was applied to the column at a flow rate of 0.75 mL/min.

The absorbance at 280 nm was monitored with an ISCO UA6 UV Visible detector at a setting of 0.5 and a chart speed of 30 cm/h. Fractions (0.75 mL) were collected in Eppendorf tubes at a flow rate of 0.75 mL/min. Fractions containing mature IF2mt were identified by

10 % SDS-polyacrylamide gel electrophoresis (Fig. 3B) and pooled. The sample was fast frozen in a dry-ice isopropanol bath and stored at -70 °C. IF2mt eluted at a KCl concentration of approximately 0.22 M.

Assays of IF2 variants on 70S E. coli Ribosomes: Reaction mixtures (0.1 mL) contained 50 mM Tris-HCl (pH 7.8), 5 mM MgCl2, 80 mM KCl, 0.25 mM GTP, 1.25 mM phospho(enol)pyruvate, 0.37 units of pyruvate kinase, 1 mM dithiothreitol (DTT), 12.5 µg poly(A,U,G), 0.06 µM (6 pmol) [35S]fMet-tRNAfMet [approximately 70,000 counts per min

(cpm)/pmol], 0.24 µM (24 pmol) E. coli IF3, 0.24 µM (58 µg) E. coli 70S ribosomes, and

67 the various amounts of the IF2 variants. These assays were also performed in the presence of

0.50 μM (50 pmol) IF1 where indicated. When dosing with IF1 instead of with IF2 the concentration of IF2 was held constant at the indicated concentration. Samples were incubated at 37 °C for 15 min. After incubation, each sample was rapidly diluted with 3-4 mL of cold Mg5 Wash Buffer (50 mM Tris-HCl (pH 7.8), 5 mM MgCl2, 80 mM KCl) and filtered through a nitrocellulose membrane (Millipore HAWG) that had been wetted with cold Mg5 Wash Buffer. Filtration was facilitated by use of a gentle vacuum created by a pump. The filters were washed with 3 aliquots (3-4 mL each) of cold Mg5 Wash Buffer with the vacuum on. Filters were dried at 100 °C for 5-7 min and counted in 5 mL toluene containing 5 g/L 2,4 diphenyloxazole (PPO) scintillation cocktail.

Assays of IF2 variants on Bos taurus (Bovine) 55S Mitochondrial Ribosomes: Reaction mixtures (0.1 mL) contained 50 mM Tris-HCl (pH 7.8), 7.5 mM MgCl2, 35 mM KCl, 0.25 mM GTP, 1.25 mM phospho(enol)pyruvate, 0.37 units of pyruvate kinase, 1 mM DTT, 12.5

µg poly(A,U,G), 0.06 µM (6 pmol) [35S]fMet-tRNAfMet (approximately 70,000 cpm/pmol),

0.2 µM IF3mt, 0.06 µM (6 pmol) 55S mitochondrial ribosomes (1 A260 is 32 pmol of 55S ribosomes) and the desired amount of the IF2 variants. These assays were also performed in the presence of 0.50 μM (50 pmol) E. coli IF1 where indicated. When dosing with IF1 instead of IF2, the concentration of IF2 was held constant at the indicated concentration.

Samples were incubated at 37 °C for 15 min and then treated as described above except that

Mg7.5 Wash Buffer (50 mM Tris-HCl (pH 7.8), 7.5 mM MgCl2, 35 mM KCl) was used.

68 RESULTS

NKXD Variants of IF2mt

Preparation and Expression of the NKXD IF2mt Variants: Analysis of sequence alignments of IF2s using the National Center for Biotechnology Information’s (NCBI) BLAST and

Conserved Domain Database (CDD) programs indicates that the X residue of the NKXD motif is not random but rather is generally present as isoleucine, valine, methionine or cysteine (Fig. 3-3). This observation suggests that a limited repertoire of residues can be accommodated as the X residue in IF2. There is no clear pattern between organisms (gram negative, gram positive or organelle) and the X residue in IF2s. To investigate the range of residues that can be accommodated at the X position, this residue in bovine IF2mt (cysteine at position 290) was mutated to alanine , isoleucine and serine.

The wild-type and mutated proteins were expressed in E. coli as His-tagged proteins and their expression was examined by Western blotting following purification on Ni-NTA resins. As indicated in Fig. 3-7 the wild-type protein was present as a major band at 73 kDa.

In addition to the full-length protein product a number of other bands were observed in these preparations. Western blot analysis indicated that the majority of these were degradation products (Fig. 3-7) which could be removed by HPLC [17]. Expression of the C290S variant was similar to that of the wild-type IF2mt. This observation suggests that replacement of the cysteine residue by serine allows the protein to fold correctly. In contrast to wild-type IF2mt and the C290S variant, replacement of C290 by alanine resulted in a severe reduction in the amount of full-length protein obtained. Essentially no full-length polypeptide product was

70 observed in these preparations. Western blotting indicated the presence of significant amounts of degradation products. These observations suggest that replacement of C290 by alanine results in a protein in which the G-domain fails to fold properly resulting in an unstable factor. When C290 was replaced by isoleucine, a common residue in the X position in IF2s, a similar problem with the stability of the protein was observed. Again, this variant of IF2mt failed to fold and no full-length protein was observed, only degradation products were visualized. The potential that these proteins went into inclusion bodies was evaluated by purifying them under denaturing conditions. However, no additional bands were observed

(data not shown).

Activity in Initiation Complex Formation of the NKXD Variants: The low level of expression of the C290A and C290I variants precluded testing their activities in initiation complex formation. However, the activity of the C290S variant was tested on both E. coli

70S ribosomes and mitochondrial 55S ribosomes. The C290S was as active as wild-type

IF2mt in initiation complex formation on the 70S ribosome (Fig. 3-8A). This variant was also as active as wild-type IF2mt in promoting the binding of fMet-tRNAfMet to mitochondrial 55S ribosomes (Fig. 3-8B).

Role of the Insertion Between Domains V and VI on the Activity of IF2mt and its

Response to IF1

Preparation and Expression of IF2mtΔ37 and EcoIF2::37: IF2mt has a conserved insertion between domains V and VI that has been investigated to determine its role in initiation.

72 Based on structural analysis of E. coli IF2 (EcoIF2) and its placement on the prokaryotic SSU, the mitochondrial insertion has been proposed to interact in the same region on the SSU as IF1. To further examine the role of the insertion domain the entire insertion in IF2mt was removed and replaced with the equivalent region from E. coli IF2

(Fig. 3-6). IF2mtΔ37 was prepared by our collaborators in India, Dr. Umesh Varshney and

Rahul Gaur. This variant of IF2mt has the insertion deleted and replaced with the corresponding sequence in E. coli IF2, forming IF2mtΔ37. The region of E. coli IF2 spanning the insertion was used to replace the insertion and a number of flanking residues using conserved sequences in domains V and VI to anchor the ends of the deletion. For this purpose, amino acid residues 654-698 of EcoIF2 were used to replace residues 444-525 of

IF2mt (Fig. 3-6). Silent mutations were engineered to introduce the necessary restriction sites. The IF2mt insertion region is flanked by residues homologous to E. coli IF2, 440

AREV 403 and 526 GSVE 529 on the N and C terminal borders of the insertion, respectively. This strategy is analogous to that used in deleting the coiled-coil region of E. coli EF-Ts and its replacement with a bridging linker derived from EF-Tsmt [37;38]. In the complementary experiment to that of the IF2mtΔ37 hybrid, the insertion observed in IF2mt was inserted into E. coli IF2 using the cloned His-tagged form of this factor to create the

EcoIF2::37 hybrid. The same flanking regions as above were used to amplify the IF2mt insertion and ligate it into EcoIF2 using the PmlI and BamHI restriction sites, previously described (see scheme Fig. 3-6D).

Cells expressing both IF2mtΔ37 and EcoIF2::37 were grown, expressed and purified according to the IF2mt variant purification protocol. Both proteins expressed well and their activities were analyzed in 70S (Figs. 3-9, 3-10) and 55S (Fig. 3-11) initiation complex

73 formation assays in the presence and absence of IF1. Both IF2mtΔ37 and EcoIF2::37 were assayed separately using the same controls, namely wild-type EcoIF2 and IF2mt. EcoIF2 and IF2mt served as benchmarks to compare each assay because their activities have been previously characterized.

The activity of IF2mtΔ37, IF2mt lacking the insertion region, in the presence and absence of IF1 on 70S prokaryotic ribosomes is shown in Figure 3-9. In the absence of IF1

(Fig. 3-9A) IF2mtΔ37 shows a modest activity. The activities of wild-type EcoIF2 and

IF2mt were modest and quite active, respectively. The modest activity of EcoIF2 is not unexpected; this factor is known to require IF1 to be fully functional. The IF2mtΔ37 has approximately the same activity as EcoIF2. In the presence of IF1 (Fig. 3-9B), the activity of

IF2mtΔ37 appears stimulated about 5-fold. Additionally, the activity of EcoIF2 is stimulated nearly 10-fold, as expected, and is comparable to that of IF2mt. Dosing IF1 (Fig. 3-9C) against constant amounts of IF2s reveals that IF2mtΔ37 shows a clear response to the addition of this factor and now mimics the response of EcoIF2. IF2mt shows no IF1 dose response.

The activity of EcoIF2::37, E. coli IF2 with the IF2mt insertion, in the presence and absence of IF1 on 70S prokaryotic ribosomes is shown in Figure 3-10. In the absence of IF1

(Fig. 3-10A) EcoIF2::37 shows an anemic activity. The activities of wild-type EcoIF2 and

IF2mt were the same as the previous experiment. In the presence of IF1 (Fig. 3-10B), the activity of EcoIF2::37 appears slightly improved, although still very anemic. EcoIF2 and

IF2mt respond as previously noted. Dosing of IF1 (Fig. 3-10C) against constant amounts of

IF2 revealed that EcoIF2::37 shows negligible response to the addition of this factor and now resembles the response of IF2mt.

76 The 55S mitochondrial initiation complex activity assays were performed on IF2mt,

EcoIF2, IF2mtΔ37 and EcoIF2::37 in the presence and absence of IF1 (Fig. 3-11A/B) and also with increasing concentrations of IF1 (Fig. 3-11C) as previously described. These assays revealed little because EcoIF2, IF2mtΔ37 and EcoIF2::37 are not active on 55S mitochondrial ribosomes. Clearly the insertion in IF2mt is very important in initiation complex formation on 55S mitochondrial ribosomes. The IF2mt variant lacking the insertion, IF2mtΔ37, has approximately 5 % of the activity of the wild-type IF2mt. Despite the increased activity of IF2mtΔ37 in the presence of IF1 on 70S prokaryotic ribosomes, on

55S mitochondrial ribosomes IF1 has no effect. The lack of an IF1 effect is not totally unexpected on 55S mitochondrial ribosomes because this factor is not a component of the mitochondrial proteome; whereas, IF1 is an essential gene in prokaryotes. Following the logic of IF1’s lack of activity on 55S mitochondrial ribosomes, EcoIF2 also has no activity.

The experiment to add the IF2mt insertion to the EcoIF2 factor, EcoIF2::37, to gain activity on 55S mitochondrial ribosomes was unsuccessful.

These observations suggest that the insertion in IF2mt permits this factor to function effectively in the absence of IF1 and that its deletion results in an IF2 derivative whose activity is more dependent on the addition of IF1. Such a dependency agrees with the idea that IF2mtΔ37 can no longer support an infA deletion, a property also evident in transduction experiments performed by our collaborators, discussed later. Addition of the insertion region to EcoIF2 (EcoIF2::37) leads to a factor with very low activity (Fig. 3-10). It shows little stimulation by IF1 but the very anemic activity of this construct makes a clear interpretation of this particular result difficult. However, it remains to be seen if further improvements in

78 the design of such a construct e. g., wherein the 37 amino acid insertion may be placed in an appropriate structural context, may allow evolution of a gain of function chimera of EcoIF2.

DISCUSSION

NKXD IF2mt Variants

The results presented here clearly indicate that the X residue in the NKXD motif is quite restricted in the side chain that can be accommodated in a specific protein. The structure of aIF5B indicates that the X residue points into the interior of G-domain and is relatively inaccessible to solvent (Fig. 3-12). Any distortion of conformation in this region could lead to significant structural perturbations and the subsequent degradation of the protein.

Structural distortion can easily arise when residues of significantly different size replace interior residues of a protein. In the case of IF2mt we have replaced cysteine (86 Å3) with alanine (67 Å3), serine (73 Å3) and isoleucine (124 Å3) [39;40]. The C290S mutation results in the least net change in volume compared to the alanine or isoleucine mutations.

This observation suggests that replacement of the sulfur of cysteine with the oxygen of serine permits an acceptable fold in the G-domain despite the somewhat smaller size of this residue. In contrast, replacement of the sulfur of cysteine by the hydrogen of alanine creates too large a hole in the G-domain. This hole may permit the penetration of water into the hydrophobic core of the G-domain destabilizing its structure.

Numerous studies of protein folding indicate that the creation of cavities can destabilize the hydrophobic core of a domain destabilizing its entire structure. Eriksson et al.

80 [41] have carried out an extensive mutagenic analysis with phage T4 lysoszyme. Their data indicate that mutation of leucine to alanine at a variety of positions resulted in a significant loss in the stability of the protein. Cavities can lead to substantial repacking of the interior.

A single interior mutation can decrease the stability of a protein by up to 9 kcal/mol [39] potentially causing the protein to unfold.

In contrast to the potential cavity formed in the C290A variant the bulky side chain of

Ile would create significant steric clashes in the tightly packed interior of the G-domain.

These steric clashes would be expected to distort the structure of this domain preventing the proper folding of the protein and perhaps exposing hydrophobic residues of the core to solvent. Thus, although the designation of a residue as (X) suggests that the amino acid in this position is random, proteins are highly ordered macromolecules and many positions can accommodate a limited repertoire of amino acids.

IF2mt Insertion Hybrids

Our working hypothesis is that the insertion in IF2mt plays the role of IF1 due to the lack of an identifiable IF1 in the mitochondria and the observation that a factor equivalent to IF2 is necessary in both prokaryotic and eukaryotic cytoplasmic systems. Models of IF2mt have suggested that the insertion projects from the body of the protein. Fitting this model into

Cryo-EM density of IF2 on the 30S SSU places the IF2mt insertion near the region that IF1 occupies in crystallographic studies. Therefore, the role of the insertion domain in IF2mt was analyzed both by removing the insertion in IF2mt and by placing the IF2mt insertion into E. coli IF2. Removal of the insertion in IF2mtΔ37 resulted in a factor that shows some

81 dependence on IF1 for maximal activity in initiation complex formation. This observation implicates the insertion of IF2mt in playing the role of IF1 or at least providing the contacts with the prokaryotic ribosome to attain maximal initiation complex activity. However, addition of this region to E. coli IF2 (EcoIF2::37) resulted in a factor that has lost all activity in initiation complex formation. This loss of activity could have arisen from a distortion in the protein structure. In creating these hybrids we were concerned about damaging the fold of the protein and some structural distortions have probably resulted as suggested by the decreased binding activity of these hybrids.

In vivo work carried out by our collaborators in India revealed that IF2mt can not only replace IF2 in E. coli IF2 deletion strains but can also rescue the double knock-out

IF1/IF2 E. coli strains. Experiments have shown that the IF2/IF1 knock-out cannot be rescued by E. coli IF2 since the IF1 gene is essential in bacteria. However, the double deletion can be rescued by IF2mt strongly suggesting that IF2mt provides the function of both IF1 and IF2. Clearly since the IF1 deletion is lethal there is some inherent ability of

IF2mt to replace the functionality of this factor. Both IF2mtΔ37 and EcoIF2::37 can rescue the strain in which the IF2 gene has been knocked out. The in vivo work strongly supports the idea that IF2mt serves the role of both IF2 and IF1. These gain-of-function experiments are a powerful indication that the insertion in IF2mt in translation does in fact replace need for a mitochondrial IF1 factor.

82 REFERENCES

[1] Liao H-X & Spremulli LL (1991) Initiation of protein synthesis in animal mitochondria: Purification and characterization of translational initiation factor 2. J. Biol. Chem., 266, 20714-20719.

[2] Liao H-X & Spremulli LL (1990) Identification and initial characterization of translational initiation factor 2 from bovine mitochondria. J. Biol. Chem., 265, 13618- 13622.

[3] Laursen BS, Sorensen HP, Mortensen KK, & Sperling-Petersen HU (2005) Initiation of Protein Synthesis in Bacteria. Microbiol. Mol. Biol. Rev., 69, 101-123.

[4] Moreno JMP, Dyrskjøtersen L, Kristensen J, Mortensen K, & Sperling-Petersen H (1999) Characterization of the domains of E. coli initiation factor IF2 responsible for recognition of the ribosome. FEBS Lett., 455, 130-134.

[5] Nyengaard NR, Mortensen KK, Lassen SF, Hershey JWB, & Sperlingpetersen HU (1991) Tandem Translation of Escherichia-Coli Initiation-Factor If2-Beta - Purification and Characterization Invitro of 2 Active Forms. BBRC, 181, 1572-1579.

[6] Sacerdot C, Vachon G, Laalami S, Morel-Deville F, Cenatiempo Y, & Grunberg- Manago M (1992) Both forms of translational initiation factor IF-2 ( and ) are required for maximal growth of Escherichia coli. J. Mol. Biol., 225, 67-80.

[7] Moreno JMP, Kildsgaard J, Siwanowicz I, Mortensen KK, & Sperling-Petersen HU (1998) Binding of E. coli initiation factor IF2 to 30S ribosomal subunits: a functional role for the N-terminus of the Factor. BBRC, 252, 465-471.

[8] Moreno JMP, Sorensen M, Mortensen KK, & Sperling-Petersen HU (2000) Molecular mimicry in translation initiation: A model for the initiation factor IF2 on the ribosome. Life, 50, 347-354.

[9] Mortensen K, Kildsgaard J, Moreno JMP, Steffensen S, Egebjerg J, & Sperling- Petersen H (1998) A six-domain structural model for Escherichia coli translation initiation factor 2. Characterization of twelve surface epitopes. BBRC, 46, 1027-1041.

[10] Milon P, Tischenko E, Tomsic J, Caserta E, Folkers G, La Teana A, Rodnina MV, Pon CL, Boelens R, & Gualerzi CO (2006) The nucleotide-binding site of bacterial translation initiation factor 2 (IF2) as a metabolic sensor. PNAS, 103, 13962-13967.

[11] Brock S, Skaradkiewicz K, & Sprinzl M (1998) Initiation factors of protein biosynthesis in bacteria and their structural relationship to elongation and termination factors. Mol. Micro., 29, 409-417.

[12] Guenneugues M, Caserta E, Brandi L, Spurio R, Meunier B, Pon CL, Boelens R, & Gualerzi CO (2000) Mapping the fMet-tRNA binding site of initiation factor IF2. EMBO, 19, 5233-5240.

83 [13] Krafft C, Diehl A, Laettig S, Behlke J, Heinemann U, Pon CL, Gualerzi CO, & Welfle H (2000) Interaction of fMet-tRNA(fMet) with the C-terminal domain of translational initiation factor IF2 from Bacillus stearothermophilus. FEBS Lett., 471, 128-132.

[14] Spurio R, Brandi L, Caserta E, Pon C, Gualerzi C, Misselwitz R, Krafft C, Welfle K, & Welfle H (2000) The C-terminal subdomain (IF2 C-2) contains the entire fMet- tRNA binding site of initiation factor IF2. J. Biol. Chem., 275, 2447-2454.

[15] Szkaradkiewicz K, Zuleeg T, Limmer S, & Sprinzl M (2000) Interaction of fMet- tRNAfMet and fMet-AMP with the C-terminal domain of Thermus thermophilus translation initiation factor 2. Eur. J. Biochem., 267, 4290-4299.

[16] Misselwitz R, Welfle K, Krafft C, Welfle H, Brandi L, Caserta E, & Gualerzi C (1999) The fMet-tRNA binding domain of translational initiation factor IF2: Role and environment of its two cys residues. FEBS Lett., 459, 332-336.

[17] Spencer AC & Spremulli LL (2005) The interaction of mitochondrial translational initiation factor 2 with the small ribosomal subunit. Biochim. Biophys. Acta - Proteins & Proteomics, 1750, 69-81.

[18] Roll-Mecak A, Cao C, Dever TE, & Burley SK (2000) X-Ray structures of the universal translation initiation factor IF2/eIF5B. Conformational changes on GDP and GTP binding. Cell, 103, 781-792.

[19] Laursen BS, Kjaergaard AC, Mortensen KK, Hoffman DW, & Sperling-Petersen HU (2004) The N-terminal domain (IF2N) of bacterial translation initiation factor IF2 is connected to the conserved C-terminal domains by a flexible linker. Protein Sci., 13, 230-239.

[20] Spremulli LL, Coursey A, Navratil T, & Hunter SE (2004) Initiation and elongation factors in mammalian mitochondrial protein synthesis. (Moldave K, ed), pp. 211-261.

[21] Luchin S, Putzer H, Hershey J, Cenatiempo Y, Grunberg-Manago M, & Laalami S (1999) In vitro study of two dominants inhibitory GTPase mutants of E. coli translation initiation factor IF2: Direct evidence that GTP hydrolysis is necessary for factor recycling. J. Biol. Chem., 274, 6074-6079.

[22] Sprang S (1997) G protein mechanisms: Insights from structural analysis. Ann. Rev. Biochem., 66, 639-678.

[23] Larigauderie G, Laalami S, Nyengaard NR, Grunberg-Manago M, Cenatiempo Y, Mortensen KK, & Sperling-Petersen HU (2000) Mutation of Thr445 and Ile500 of initiation factor 2 G-domain affects Escherichia coli growth rate at low temperature. Biochim., 82, 1091-1098.

[24] Laursen BS, Siwanowicz I, Larigauderie G, Hedegaard J, Ito K, Nakamura Y, Kenney JM, Mortensen KK, & Sperling-Petersen HU (2003) Characterization of

84 Mutations in the GTP-binding Domain of IF2 Resulting in Cold-sensitive Growth of Escherichia coli. J. Mol. Biol., 326, 543-551.

[25] Laalami S, Timofeev A, Putzer H, Leautey J, & Grunberg-Manago M (1994) In vivo study of engineered G-domain mutants of Escherichia coli translation initiation factor IF2. Mol. Micro., 11, 293-302.

[26] Schwede T, Kopp J, Guex N, & Peitsch MC (2003) SWISS-MODEL: an automated protein homology-modeling server. Nuc. Acids Res., 31, 3381-3385.

[27] Ramakrishnan V (2002) Ribosome structure and the mechanism of translation. Cell, 108, 557-572.

[28] Marzii ST, Knight WI, Brandi LE, Aserta EN, Soboleva NA, Hill WE, Gualerzi CO, & Lodmell JS (2003) Ribosomal localization of translation initiation factor IF2. RNA, 9, 958-969.

[29] Carter AP, Clemons J, Brodersen DE, Morgan-Warren RJ, Hartsch T, Wimberly BT, & Ramakrishnan V (2001) Crystal Structure of an Initiation Factor Bound to the 30S Ribosomal Subunit. Science, 291, 498-501.

[30] Koc EC & Spremulli LL (2002) Identification of mammalian mitochondrial translational initiation factor 3 and examination of its role in initiation complex formation with natural mRNAs. J. Biol. Chem., 277, 35541-35549.

[31] Lefevre F, Remy M-H, & Masson J-M (1997) Alanine-stretch scanning mutagenesis: a simple and efficient method to probe protein structure and function. Nuc. Acids Res., 447-449.

[32] Cummings H & Hershey J (1994) Translation initiation factor IF-1 is essential for cell viability in Escherichia coli. J. Bact., 176, 198-205.

[33] Ma J & Spremulli LL (1996) Expression, purification and mechanistic studies of bovine mitochondrial translational initiation factor 2. J. Biol. Chem., 271, 5805-5811.

[34] Graves M & Spremulli LL (1983) Activity of Euglena gracilis chloroplast ribosomes with prokaryotic and eukaryotic initiation factors. Arch. Biochem. Biophys., 222, 192- 199.

[35] Steege DA, Graves MC, & Spremulli LL (1982) Euglena gracilis chloroplast small subunit rRNA. Sequence and base pairing potential of the 3' terminus, cleavage by colicin E3. J. Biol. Chem., 257, 10430-10439.

[36] Matthews DE, Hessler RA, Denslow ND, Edwards JS, & O'Brien TW (1982) Protein composition of the bovine mitochondrial ribosome. J. Biol. Chem., 257, 8788-8794.

85 [37] Karring H, Bjornsson A, Thirup S, Clark BFC, & Knudsen CR (2003) Functional effects of deleting the coiled-coil motif in Escherichia coli elongation factor Ts. Eur. J. Biochem., 270, 4294-4305.

[38] Hamilton CM, Aldea M, Washburn BK, Babitzke P, & Kushner SR (1989) New method for generating deletions and gene replacements in Escherichia coli. J. Bact., 171, 4617-4622.

[39] Kubarenko A, Sergiev P, & Rodnina M (2005) GTPases of the Translation Apparatus. Mol. Bio., 39, 646-660.

[40] Mitra K, Schaffitzel C, Shaikh T, Tama F, Jenni S, Brooks CL, Ban N, & Frank J (2005) Structure of the E. coli protein-conducting channel bound to a translating ribosome. Nature, 438, 318-324.

[41] Eriksson A, Baase W, Zhang X-J, Heinz D, Blaber M, Baldwin E, & Matthews B (1992) Response of a protein structure to cavity-creating mutations and its relation to the hydrophobic effect. Science, 255, 178-183.

CHAPTER IV

AUXILIARY TRANSLATIONAL FACTORS OF MAMMALIAN

MITOCHONDRIAL RIBOSOMES

INTRODUCTION

Mammalian mitochondrial DNA (mtDNA) encodes 13 proteins, 2 rRNAs and 22 tRNAs. The heart mitochondrial proteome was characterized using 1D PAGE separation and mass spectrometry analysis and revealed 615 distinct protein products [1]. It is proposed that there are up to 1500 proteins that compose the mitochondrial proteome [2]. Therefore, there is quite a large gap in our understanding of the proteins involved in mitochondrial biosynthesis and the internal functions of this organelle. Of important interest is the mechanism used to synthesize the 13 polypeptides encoded in the mitochondrial genome.

This biosynthetic process is divided into 4 stages of translation: initiation, elongation, termination and recycling. All of these stages require specific auxiliary factors (initiation factors, elongation factors and recycling factors) as well as factors involved in ribosome biogenesis and function. There are a number known translational factors in both prokaryotes and yeast mitochondria that have yet to be identified in the mammalian mitochondrial system. In fact, there are still translational factors being discovered in the prokaryotic system, for example the recently described protein LepA [3].

LepA is the first gene in a bicistronic operon in E. coli. The second gene is the leader peptidase or lep gene which encodes the signal peptidease Lep [4]. A knockout of the LepA gene in E. coli has no discernible phenotype [5] but this protein was found associated with

Staphylococcus aureus ribosomes in antibiotic crosslinking studies [6]. LepA, a ubiquitous protein found in all bacteria and mitochondria, was recently characterized as being an additional elongation factor (and thus a proposed renaming to EF4) after EF-Tu, EF-Ts and

EF-G [3]. The sequence homology to other translational GTPases such as IF2 and EF-G

strongly suggests its interaction with the ribosome. LepA is a highly conserved protein having an amino acid identity across all bacteria from 55% - 68% (Table 4-1). This conservation suggests an important biological role. The structure of LepA is thought to be very similar to that of EF-G, which is a mimic of the EF-Tu:GTP:aa-tRNA ternary complex structure (Fig. 4-1). As detailed in Chapter I the steps of elongation in bacteria are clearly defined by EF-Tu delivering an aa-tRNA to the A-site of the ribosome while EF-G translocates the complex following peptidyl transfer of the growing polypeptide from the P- site tRNA to the A-site tRNA. Where does this leave LepA? Experimental evidence suggests that LepA recognizes the ribosome that has undergone a defective translocation action. LepA corrects this defect by back translocating the tRNAs and mRNAs giving EF-G another chance to carry out translocation correctly [3]. In this Chapter the cloning of the mammalian mitochondrial LepA homolog is reported.

The discovery of new mitochondrial translational factors such as LepA using a bioinformatics approach has prompted us to search the human expressed sequence tag (EST) sequence for other factors involved in prokaryotic and yeast mitochondrial ribosomal biogenesis, ribosome receptors and other accessory ribosomal factors. For the identification of human mitochondrial homologs of other factors required in protein synthesis or ribosome biogenesis, a number of criteria must be applied. First, homology search using tools such as the BLAST program at the National Center for Biotechnology Information (NCBI) must indicate a reasonable degree of amino acid identity at least in portions of the identified protein. As a minimum a 20 % identity is desired to proceed with further analysis. Once a candidate protein has been identified, further examination of the sequence must be carried out to assess whether it could potentially be localized to mitochondria. Several criteria can

be applied here. First, many proteins localized to the mitochondria have N-terminal import sequences. These sequences can be identified with about 75 % accuracy using prediction programs such as MitoProtII and PSORT [7;8]. Second, specific promoter sequences in nuclear genes encoding mitochondrial proteins can be used to identify these factors.

Previous work has analyzed the roles of various promoters in respiratory gene expression in mammalian cells and identified two nuclear respiratory factors (NRF) sequences, NRF-1 and

NRF-2, which play a key role [9]. A recent study of the promoter of IF2mt revealed that nuclear regulatory factor 2 (NRF-2) binding sites are present as tandem pairs. This promoter architecture is also found in many ribosome associated proteins such as EF-Tumt, EF-G1mt,

RF1mt, MetRSmt, rpL3mt and rpS12mt [10;11] (Fig. 4-2). Other factors localized to mitochondria that are known to be controlled by NRF-1 and NRF-2 include TOM20, succinate dehydrogenase subunits B, C and D of complex II, and cytochrome oxidase subunit

Vb of complex IV [9;11;12]. Therefore a traditional sequence homology alignment and promoter analysis of known mitochondrial translational factors are a good start to identify potential mitochondrial proteins which may have a role in translation.

MATERIALS AND METHODS

Materials: Chemicals were purchased from Sigma and Fisher Scientific. Oligonucleotide primers were perpared at the Nucleic Acids Core Facility (University of North Carolina). All plasmids were purified using QIAGEN Qiaprep® Miniprep Handbook as described by the manufacturer. Restriction enzymes were purcfhased from New England Biolabs. Digested

plasmids and PCR products were purified using the Qiagen Gel Extraction/PCR Purification

Kit.

Cloning of mtLepA: Human mtLepA was identified in the ESTs by BLAST analysis using the E. coli LepA sequence as a query. The cDNA was ordered from American Type Culture

Collection (ATCC). The coding sequence was amplified out of the starting pCMV-SPORT6 vector using the following primers: Forward primer with NcoI restriction site 5'-CCA ACC

ATG GCG CCG ACC CTT GGG G-3' and Reverse primer with XhoI restriction site 5'-GGT

TCT CGA GTT TAG AAG ATT GTG TTT TCA GAA CTT TTA TGA AAG-3'. This PCR product was digested with NcoI and XhoI along with purified pET24d(+) plasmid. The PCR product was ligated into the pET24d(+) plasmid at 16 °C for 16 h under standard conditions.

The ligated plasmid was then transformed into Stratagene DH5α, sequenced and subcloned into Stratagene BL21 (DE3) RIL cells according to the Stratagene protocols.

Growth Studies of mtLepA: Overnight cultures of E. coli BL21 (DE3) RIL cells containing either pET24d mtLepA, pET21b mtEF-G or no plasmid were grown in 20 mL of LB with 34

μg/mL chloramphenicol (for the RIL plasmid present in all cells) and the appropriate secondary plasmid antibiotic, 50 μg/mL kanamycin for pET24d and 100 μg/mL ampicillin for pET21b. These overnight cultures were used to inoculate, using a 1 % dilution, 200 mL of LB with the appropriate antibiotic in 500 mL side armed (Nephelo) flasks. The growth of the cells was monitored every 30 minutes by measuring the A600 on a Bausch and Lomb

Spectronic 20 spectrophotometer. At approximately 0.8 A600 the cells were induced with 0.1

M IPTG. Following induction the cells were diluted 10-fold in LB containing the

appropriate antibiotic and 0.1 M IPTG and the effect of induction on the growth of the cells was monitored to a total time of 600 minutes.

Elucidation of Auxiliary Mammalian Mitochondrial Translational Factors: Sequences of known prokaryotic and yeast mitochondrial factors involved in translation were submitted for a TBLASTN search against Homo sapian RefSeq proteins and ESTs at the National Center for Biotechnology Information (NCBI) website, http://www.ncbi.nlm.nih.gov/BLAST/. The top scores were analyzed and top candidates were submitted subjected to MITOPROT analysis at http://ihg.gsf.de/ihg/mitoprot.html to determine if a mitochondrial import sequence was present. If an import sequence was verified then the mature truncated form was used to realign the sequence via CLUSTLW to determine the percent identity to the query sequence.

Promoter Analysis: Once the putative mammalian mitochondrial translation factor gene has been identified, the cDNA or gene name is queried using BLAT alignment tool or Genome

Browser at the UCSC Genome Bioinformatics website, http://genome.ucsc.edu, which contains the latest build of the human genome. Once the gene has been located on the particular chromosome the upstream promoter sequences can be obtained and analyzed using the Transcription Element Search System (TESS) at http://www.cbil.upenn.edu/cgi- bin/tess/tess.

RESULTS

mtLepA

Human mitochondrial LepA (mtLepA) was identified by a BLAST search against the Homo sapians database using the E. coli LepA factor (NP_417064) as a query sequence.

Alignment revealed the human homolog with 47% identity, Unigene sequence

Hs.546419. The protein was analyzed using MitoProtII to determine the presence of a mitochondrial import sequence and with a 94% probability the import sequence was defined as the first 32 residues. Therefore the proposed mature sequence begins at residue

33. An alignment of these two factors is shown in Figure 4-3, which clearly shows the significant identity between these two factors.

The cDNA of this protein was ordered from ATCC and subcloned into the expression vector. Once this gene was sequenced several differences from the original cDNA clone, as reported at NCBI, were indicated as possible point mutations which resulted in a number of amino acid changes. These mutations were analyzed using a TBLASTN analysis of the

Homo sapiens mtLepA sequence against the human expressed sequence tag (EST) database. All mutations except one were revealed to be errors in the original NCBI sequence. As shown in Figure 4-4, the single P/L mutation at residue 58 turns out to be an isoform of mtLepA, where both P and L are present in equal proportions in various ESTs.

The variant that was cloned has a L residue at position 58. This isoform lies within a region that has no homology to E. coli LepA. Therefore the ESTs confirm the mtLepA sequence that was cloned was indeed correct and could be used for further experiments.

The primary experiment to test the activity of mtLepA was an in vivo expression monitoring the growth of the E. coli BL21 (DE3) RIL cells prior to and following induction of the protein using IPTG. The effects of in vivo induction with IPTG on E. coli LepA in

BL21 cells was previously performed and revealed a quite dramatic and unrecoverable effect on cell growth [3]. The experiments performed on pET24d mtLepA in BL21 (DE3) RIL cells revealed that the cells can recover after induction (Fig. 4-5B). mtLepA shows a significant lag in growth relative to the E. coli BL21 (DE3) RIL cells that did not contain a plasmid and EF-Gmt. Following the 1/10 dilution and induction with IPTG both EF-Gmt and mtLepA have a significant lag time but nevertheless they both grow. These results differ quite dramatically from the E. coli factors (Fig. 4-5A).

Auxiliary Translational Factors

To begin the search for auxiliary mitochondrial translational factors, known bacterial and yeast mitochondrial factors were subjected to a BLASTP analysis against the RefSeq Homo sapiens database. Previous analysis using cyber searching had revealed the presence of mitochondrial initiation factor 3 (IF3mt) [13], the prokaryotic LepA [3], and prokaryotic

YidC. YidC is known to be homologous to yeast mitochondrial Oxa1p [14-16] which is involved in the insertion of membrane proteins into the mitochondrial membrane. MitoProtII revealed that all of these proteins have very strong mitochondrial import sequences, with >

93% probability of import. Following MitoProtII analysis the first 30 amino acids of each protein sequence was entered into a BLAT search at the UCSC Genome Database to find the gene on its respective chromosome. The genes were then selected and the genomic

100

sequences were gathered -200 and +100 nucleotides from the transcription start sites. The promoter regions of IF3mt, mtLepA and mtOxa1 all contained strong NRF-2 sequences and the mtLepA sequence also contained a putative NRF-1 site (Fig. 4-6). The mitochondrial import, NRF-1 and NRF-2 sequence data for each factor was compiled into Table 4-2.

To extend this analysis other factors known to be involved in the mitochondrial translational apparatus in yeast were analyzed. These factors were Mba1, Mdm38 and Ylh47

[17;18]. Mba1 has been suggested to interact with Oxa1p and the large subunit of the yeast mitochondrial ribosome recruiting it to the mitochondrial membrane [17]. This factor is a good target for analysis because there is a Oxa1 homolog in mammalian mitochondria.

However, a simple BLASTP analysis using Mba1as a query revealed that there were no known human or prokaryotic homologs of this protein. If such a homolog exists, its sequence must have diverged too much for BLAST analysis to find it. Analysis was continued on the yeast factors Mdm38 and Ylh47 which have known homology to human factor Letm1 [18] (Fig. 4-7). Letm1 is thought to be an inner mitochondrial membrane protein which plays a role in potassium and hydrogen ion exchange [19]. A BLASTP analysis using Mdm38 and Ylh47, as queries against the human RefSeq database confirmed that they were most closely related to Letm1. This Letm1 has no mitochondrial import sequence but analysis of the promoter region of this factor revealed two NRF-2 motifs,

Figure 4-6 and Table 4-2.

Additional prokaryotic factors known to interact with the translational machinery were also evaluated including CgtAE, SpoT, Era, YciH and ObgE (Fig. 4-6, Fig. 4-8 and

Table 4-2). CgtAE is involved in the assembly of the 50S large subunit [20]. The human homolog of CgtAE has a predicted mitochondrial import signal with 99% probability and the

101

102

103

104

105

promoter sequence contains five NRF-2 binding sites. The factor SpoT, which is a ppGpp synthase/hydrolase, copurified with CgtAE on the prokaryotic ribosome [21] and BLASTP analysis of its protein reveals that only the N-terminus of this factor is homologous to a human factor. The N-terminal region of this SpoT homolog was analyzed and no mitochondrial import sequence was revealed. However two NRF-2 motifs are present in its promoter region. Era is a prokaryotic factor implicated in small subunit assembly [22]. The

BLAST analysis of this factor was meager but analysis of the top human homolog revealed a predicted mitochondrial import of 95% as well as four NRF-2 motifs. The prokaryotic factor

YciH is homologous to eukaryotic factor eIF1 but it is not an essential gene and its function is not known [23]. BLASTP analysis of YciH reveals that this factor is homologous to human eIF1 and SUI1/eIF1b, which is an isoform of eIF1. The human homologs of YciH are not alternatively spliced proteins rather they are encoded in separate genes. Neither has a mitochondrial import sequence. The eIF1 factor has two NRF-2 factor binding sites and one strong NRF-1 motif. The SUI1/eIF1b factor has three NRF-2 binding sites and one core

NRF-1 motif. The last prokaryotic factor analyzed was ObgE. This factor has been implicated in ribosome biogenesis and the human factors homologous to this protein, ObgH1 and ObgH2, have previously been studied [24;25]. ObgH1 and ObgH2 are encoded by two different genes and only ObgH1 has a mitochondrial import signal, with 99% probability of import. The import sequence along with biochemical studies [24;25] suggests the both the cytoplasmic and mitochondrial ribosomes require their own factors for ribosome biogenesis. Promoter analysis of the mitochondrial factor, ObgH1, revealed no NRF-2 factor binding sites.

106

In addition to auxiliary translational factors, the nuclear encoded mitochondrial ribosomal protein gene’s promoter regions were subjected to the same analysis as above. It was observed that of the approximately 80 proteins that constitute the mitochondrial ribosome only one lacks a core NRF-2 factor binding site (CGGAAG) and only seven contain NRF-1 factor binding sites ((C/T)GCGCA(C/T)GCGC(A/G)).

DISCUSSION

The cloning and growth studies of mtLepA reveal the presence of a nearly universal translational factor for the mitochondrial system which, that sequence homology and growth study comparisons reveal, plays the role of back translocating mistranslocated mitochondrial ribosomes. This function could provide a vital mechanism for correction the in mitochondrial system. Any error in the translation of the mitochondrial genome can result is significant problems with energy generation in the cell. Surprisingly the growth of mtLepA and EF-Gmt differ quite significantly from their E. coli counterparts. EF-Gmt grows normally before induction but following induction it exhibits a similar but less severe lag in growth as mtLepA. mtLepA exhibits an extended lag phase in growth both before and after induction. The delayed growth could be due to leaky expression of this protein from the pET vector resulting in a disrupted protein synthesis, as seen in the E. coli growth studies.

Furthermore, the extended lag following induction and subsequent exponential growth after

300 min could be the result of a mutant pickup.

The search for new mitochondrial translational proteins has revealed a very diverse and complex array of proteins. Using known bacterial and yeast mitochondrial proteins

107

involved in translation provided as excellent starting point for future analysis and understanding of the mitochondrial translational machinery. Furthermore the presence of

NRF-1 and NRF-2 promoters which have been shown to be important upstream elements involved in the translation of mitochondrial targeted genes in the nuclear genome. Further biochemical experimentation needs to be done to determine the validity of the cyber-probing but the potential elucidation of novel mitochondrial factors such as mtCgtAE and mtERA which may potentially play the role of their prokaryotic homologs, assembly of the large and small subunit respectively, provide new avenues of research. Additionally the presence of

Letm1, a homolog of yeast mitochondrial proteins Mdm38 and Vlh47, suggests another pathway in machinery involved in the insertion of proteins into the inner membrane.

108

REFERENCES

[1] Taylor SW, Fahy E, Zhang B, Glenn GM, Warnock DE, Wiley S, Murphy AN, Gaucher SP, Capaldi RA, Gibson BW, & Ghosh SS (2003) Characterization of the human heart mitochondrial proteome. Nat. Biotech., 21, 281-286.

[2] Taylor SW, Fahy E, & Ghosh SS (2003) Global organellar proteomics. Trends Biotechnol., 21, 82-88.

[3] Qin Y, Polacek N, Vesper O, Staub E, Einfeldt E, Wilson DN, & Nierhaus KH (2006) The Highly Conserved LepA Is a Ribosomal Elongation Factor that Back- Translocates the Ribosome. Cell, 127, 721-733.

[4] March PE & Inouye M (1985) Characterization of the Lep Operon of Escherichia- Coli - Identification of the Promoter and the Gene Upstream of the Signal Peptidase-I Gene. J. Biol. Chem., 260, 7206-7213.

[5] Dibb NJ & Wolfe PB (1986) Lep Operon Proximal Gene Is Not Required for Growth Or Secretion by Escherichia-Coli. J. Bact., 166, 83-87.

[6] Colca JR, McDonald WG, Waldon DJ, Thomasco LM, Gadwood RC, Lund ET, Cavey GS, Mathews WR, Adams LD, Cecil ET, Pearson JD, Bock JH, Mott JE, Shinabarger DL, Xiong LQ, & Mankin AS (2003) Cross-linking in the living cell locates the site of action of oxazolidinone antibiotics. J. Biol. Chem., 278, 21972- 21979.

[7] Nakai K & Kanehisa M (1992) A knowledge base for predicting protein localization sites in eukaryotic cells. Genomics, 14, 897-911.

[8] Claros MG & Vincens P (1996) Computational method to predict mitochondrially imported proteins and their targeting sequences. Eur. J. Biochem., 241, 779-786.

[9] Scarpulla RC (2006) Nuclear control of respiratory gene expression in mammalian cells. J. Cell. Biochem., 97, 673-683.

[10] Hayashi R, Ueda T, Farwell MA, & Takeuchi N (2007) Nuclear respiratory factor 2 activates transcription of human mitochondrial translation initiation factor 2 gene. Mitochondrion, 7, 195-203.

[11] Overman J, Enderle PJ, Farrow JM, Wiley JE, & Farwell MA (2003) The human mitochondrial translation initiation factor 2 gene (MTIF2): transcriptional analysis and identification of a pseudogene. Biochim. Biophys. Acta - Gene Structure and Expression, 1628, 195-205.

[12] Scarpulla RC (2002) Nuclear activators and coactivators in mammalian mitochondrial biogenesis. Biochim. Biophys. Acta - Gene Structure and Expression, 1576, 1-14.

109

[13] Koc EC & Spremulli LL (2002) Identification of mammalian mitochondrial translational initiation factor 3 and examination of its role in initiation complex formation with natural mRNAs. J. Biol. Chem., 277, 35541-35549.

[14] Bonnefoy N, Kermorgant M, Groudinsky O, Minet M, Slonimski PP, & Dujardin G (1994) Cloning of a Human Gene Involved in Cytochrome Oxidase Assembly by Functional Complementation of an Oxa1- Mutation in Saccharomcyes cerevisiae. PNAS, 91, 11978-11982.

[15] Reif S, Randelj O, ska G, Dian A, Krimmer T, Motz C, & Rassow J (2005) Conserved Mechanism of Oxa1 Insertion into the Mitochondrial Inner Membrane. J. Mol. Biol., 354, 520-528.

[16] Preuss M, Ott M, Funes S, Luirink J, & Herrmann JM (2005) Evolution of Mitochondrial Oxa Proteins from Bacterial YidC: Inherited and Acquired Functions of a Conserved Protein Insertion Machinery. J. Biol. Chem., 280, 13004-13011.

[17] Ott M, Prestele M, Bauerschmitt H, Funes S, Bonnefoy N, & Herrmann JM (2006) Mba1, a membrane-associated ribosome receptor in mitochondria. EMBO J., 25, 1603-1610.

[18] Frazier AE, Taylor RD, Mick DU, Warscheid B, Stoepel N, Meyer HE, Ryan MT, Guiard B, & Rehling P (2006) Mdm38 interacts with ribosomes and is a component of the mitochondrial protein export machinery. J. Cell Biol., 172, 553-564.

[19] Nowikovsky K, Froschauer EM, Zsurka G, Samaj J, Reipert S, Kolisek M, Wiesenberger G, & Schweyen RJ (2004) The LETM1/YOL027 gene family encodes a factor of the mitochondrial K+ homeostasis with a potential role in the Wolf- Hirschhorn syndrome. J. Biol. Chem., 279, 30307-30315.

[20] Jiang M, Datta K, Walker A, Strahler J, Bagamasbad P, Andrews PC, & Maddock JR (2006) The Escherichia coli GTPase CgtAE is involved in late steps of large ribosome assembly. J. Bacteriol., 188, 6757-6770.

[21] Wout P, Pu K, Sullivan SM, Reese V, Zhou S, Lin B, & Maddock JR (2004) The Escherichia coli GTPase CgtAE cofractionates with the 50S ribosomal subunit and interacts with SpoT, a ppGpp synthetase/hydrolase. J. Bacteriol., 186, 5249-5257.

[22] Sharma MR, Barat C, Wilson DN, Booth TM, Kawazoe M, Hori-Takemoto C, Shirouzu M, Yokoyama S, Fucini P, & Agrawal RK (2005) Interaction of Era with the 30S Ribosomal Subunit Implications for 30S Subunit Assembly. Mol. Cell, 18, 319-329.

[23] Lomakin IB, Shirokikh NE, Yusupov MM, Hellen CU, & Pestova TV (2006) The fidelity of translation initiation: reciprocal activities of eIF1, IF3 and YciH. EMBO J., 25, 196-210.

110

[24] Sato A, Kobayashi G, Hayashi H, Yoshida H, Wada A, Maeda M, Hiraga S, Takeyasu K, & Wada C (2005) The GTP binding protein Obg homolog ObgE is involved in ribosome maturation. Genes Cells, 10, 393-408.

[25] Hirano Y, Ohniwa RL, Wada C, Yoshimura SH, & Takeyasu K (2006) Human small G proteins, ObgH1, and ObgH2, participate in the maintenance of mitochondria and nucleolar architectures. Genes Cells, 11, 1295-1304.

111