UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Development of peptide-like library Zhou, Miao 2009

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2009_spring_zhou_miao.pdf [ 2.74MB ]
Metadata
JSON: 24-1.0067132.json
JSON-LD: 24-1.0067132-ld.json
RDF/XML (Pretty): 24-1.0067132-rdf.xml
RDF/JSON: 24-1.0067132-rdf.json
Turtle: 24-1.0067132-turtle.txt
N-Triples: 24-1.0067132-rdf-ntriples.txt
Original Record: 24-1.0067132-source.json
Full Text
24-1.0067132-fulltext.txt
Citation
24-1.0067132.ris

Full Text

DEVELOPMENT OF PEPTIDE-LIKE LIBRARY by Miao Zhou B.S., Simon Fraser University, 2006  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE  in  The Faculty of Graduate Studies (Pharmaceutical Sciences)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) May 2009 © Miao Zhou, 2009  Abstract Peptide libraries are important in peptide based drug discovery and investigations of biological targets. Different chemical and biological methods have been developed to screen and decode peptide libraries. mRNA display technology is a biological method that utilizes in vitro translation systems to build peptide libraries, and the screened peptides are decoded by their covalently attached oligonucleotide templates. Like other biological methods, mRNA display technology can generate peptide libraries with high fidelity and speed but the chemical repertoire is limited mostly to natural amino acids. Our research goal is to expand the chemical repertoire of mRNA display technology by using it in combination with an in vitro reconstituted translation system, and build peptides containing stabilizing chemical features such as N-methyl amino acids that would make peptides more proteolytic resistant and drug-like.  Using a streptavidin-dependent gel shift assay, the unnatural amino acids, N-methyl-Lalanine,  N-methyl-L-glutamate,  N-methyl-L-ornithine,  L-abrine,  t-butyl-L-glycine,  biocytin, and sarcosine were shown to be able to incorporate successively into hexapeptides in the mRNA display format. The peptides containing unnatural amino acids were shown to be more resistant to proteinase K proteolysis than peptides containing only the natural amino acid L-alanine. Several mRNA libraries were made in accordance with the mRNA display format. Repeated E. coli in vitro reconstituted translations of these mRNA libraries with varying translation conditions fail to yield peptides of significant lengths. However, a library of tetra-peptides was synthesized in the absence of mRNA display format. Thus, the puromycin linker required by mRNA display technology  ii  seemed to interfere with proper translations in the E. coli in vitro reconstituted translation system. Experiments showed that E. coli in vitro reconstituted translation system is slow at making full-length peptides in the context of mRNA display technology and a majority of products are prematurely terminated translation products, which could be due to slow peptidyl formation in the E. coli in vitro reconstituted translation system, or the result of undesirable interaction between the E. coli ribosome and puromycin linker that interfere with normal translation. This slow translation kinetic issue needs to be addressed with further experiments in order to successfully pursue the development of this technology.  iii  Table of contents Abstract .......................................................................................................................... ii Table of contents............................................................................................................ iv List of tables ................................................................................................................ viii List of figures ................................................................................................................ ix List of abbreviations ..................................................................................................... xii 1 Introduction ................................................................................................................. 1 1.1 Methods of building combinatorial peptide libraries .............................................. 1 1.2 A closer look at mRNA display technology........................................................... 3 1.2.1 Puromycin: the physical linkage ..................................................................... 3 1.2.2 Making mRNA display based peptide libraries by cell-free translation ........... 7 1.3 Introduction of unnatural amino acids into a biological system .............................. 8 1.3.1 Overview of bacterial translation mechanism.................................................. 8 1.3.2 Reconstituted translation system ................................................................... 10 1.3.3 Codon reassignment and tRNA mutagenesis................................................. 12 1.3.4 Choices of unnatural amino acids and selection target................................... 14 1.4 Library design ..................................................................................................... 15 1.5 Research hypothesis ............................................................................................ 15 1.6 Specific aims....................................................................................................... 15 2  Experimental methods........................................................................................... 17 2.1 Derivatizing tRNAs from tRNAalaW by PCR mutagensis ..................................... 20 2.2 Generation of tRNAs by in vitro transcription ..................................................... 23 2.21 PCR............................................................................................................... 23 iv  2.22 in vitro transcription using T7 RNA polymerase ............................................ 24 2.23 Purification of tRNAs.................................................................................... 24 2.3 Chemical modifications of unnatural amino acids................................................ 26 2.31 Synthesis of N-NVOC-N-Me-Asp-cyanomethyl ester and N-NVOC-N-Me-Asp pdCpA .................................................................................................................. 26 2.32 Synthesis of N-NVOC-Aze cyanomethyl ester and N-NVOC-Aze pdCpA..... 28 2.33 Synthesis of N-NVOC-Bio cyanomethyl ester and N-NVOC-Bio pdCpA ...... 30 2.34 Synthesis of N-NVOC-N-Me-Ser cyanomethyl ester...................................... 32 2.35 Synthesis of N-NVOC-N-Me-Tyr cyanomethyl ester and N-NVOC-N-Me-Tyr pdCpA .................................................................................................................. 34 2.36 Synthesis of N-NVOC-Oic cyanomethyl ester ............................................... 36 2.37 Synthesis of N-NVOC-N-Me-Nle cyanomethyl ester ..................................... 37 2.4 Coupling of tRNAs with amino acids .................................................................. 39 2.5 Detection of aminoacyl-tRNAs by acidic denaturing urea PAGE......................... 40 2.6 EF-Tu binding Assay........................................................................................... 41 2.7 Streptavidin binding assay................................................................................... 42 2.8 E. coli in vitro reconstituted translation reaction .................................................. 42 2.9 Purification and analysis of translation products .................................................. 43 2.91 Purification of peptide-oligonucleotide fusions .............................................. 43 2.92 Purification of peptides.................................................................................. 44 2.93 Denaturing urea PAGE analysis of translation products ................................. 45 2.10 Preparation of MALDI-TOF samples ................................................................ 46 2.101 Preparation from rabbit reticulocyte lysate................................................... 46  v  2.102 Preparation from E. coli in vitro reconstituted translation system ................. 47 2.103 MALDI-TOF analysis ................................................................................. 48 2.11 Proteinase K analysis of translation products generated by E. coli in vitro reconstituted translation system................................................................................. 48 2.12 mRNA library design and generation................................................................. 49 2.121 Making V3A, V3Ap, V4, V4p, V5, and V5p libraries.................................... 49 2.122 Making V3B and V3Bp libraries .................................................................. 51 2.123 Translation of mRNA libraries by E. coli in vitro reconstituted translation system ................................................................................................................... 52 3 Results ....................................................................................................................... 53 3.1 Validation of streptavidin assay........................................................................... 53 3.2 Assessment of the E. coli in vitro reconstituted translation system’s efficiency of making full-length peptides in mRNA display format................................................ 56 3.3 Chemical modifications to the selected unnatural amino acids............................. 64 3.4 Compatibility of unnatural amino acids with the E. coli in vitro reconstituted translation system...................................................................................................... 68 3.41 Chemoenzymatic acylation of tRNAggc(-ca) with chemically modified unnatural amino acids............................................................................................ 69 3.42 EF-Tu binding assay...................................................................................... 71 3.43 E. coli in vitro reconstituted translation with unnatural amino acids............... 73 3.44 Confirmation of peptide-oligonucleotide conjugates made by E. coli in vitro reconstituted translation system using MALDI-TOF ............................................. 76  vi  3.45 Confirmation of peptides made by E. coli in vitro reconstituted translation system using proteinase K ..................................................................................... 80 3.5 Generating a tRNA library................................................................................... 85 3.6 Making mRNA libraries ...................................................................................... 86 3.7 Making peptide libraries by E. coli in vitro reconstituted translation system ........ 87 3.71 Making peptide-oligonucleotide libraries....................................................... 87 3.72 Making peptide libraries in the absence of mRNA display format.................. 96 3.73 A time course experiment set out to address the kinetic issue with E. coli in vitro reconstituted translation system................................................................... 100 4 Discussion and conclusions...................................................................................... 103 4.1 Discussion......................................................................................................... 103 4.2 Conclusions....................................................................................................... 109 4.3 Future work....................................................................................................... 110 Bibliography............................................................................................................... 111 Appendix .................................................................................................................... 118  vii  List of tables Table 1 Components of the E. coli in vitro reconstituted translation system................... 12 Table 2 Template sequences.......................................................................................... 19 Table 3 List of tRNAs and primers needed for making tRNA library............................. 22 Supplementary Table 1 A list of all tRNAs in the tRNA library .................................. 118  viii  List of figures Figure 1 A cartoon representation of puromycin's role in mRNA display technology ...... 6 Figure 2 The chemical structure of puromycin.. .............................................................. 7 Figure 3 The synthetic scheme of converting N-Me-Asp to N-NVOC-N-Me-Asp pdCpA .............................................................................................................................. 28 Figure 4 The synthetic scheme of converting Aze to N-NVOC-Aze pdCpA. ................. 30 Figure 5 The synthetic scheme of converting Bio to N-NVOC-Bio pdCpA. .................. 32 Figure 6 The synthetic scheme of converting N-Me-Ser to N-NVOC-N-Me-Ser cyanomethyl ester. ................................................................................................ 33 Figure 7 The synthetic scheme of converting N-Me-Tyr to N-NVOC-N-Me-Tyr pdCpA. .............................................................................................................................. 35 Figure 9 The synthetic scheme of converting N-Me-Nle to N-NVOC-N-Me-Nle cyanomethyl ester. ................................................................................................ 38 Figure 10 Validation of streptavidin assay by 15% denaturing urea PAGE.................... 54 Figure 11 Validation of streptavidin assay by 12% denaturing urea PAGE.................... 56 Figure 12 Aminoacyl-tRNA dependent formation of dipeptide-oligonucleotides........... 60 Figure 13 Aminoacyl-tRNA dependent formation of hexapeptide-oligonucleotide conjugates. ............................................................................................................ 62 Figure 14 A chemical method of preparing an amino acid for ligation to a tRNA. ......... 65 Figure 15 Structures of chemically modified amino acids as confirmed by ESI-MS. ..... 67 Figure 16 A flow chart showing the steps involved in testing the compatibility of each unnatural amino acid. ............................................................................................ 69  ix  Figure 17 Confirmation of chemo-enzymatically acylated aminoacyl-tRNAggc by 15% acidic denaturing urea PAGE.. .............................................................................. 70 Figure 18 EF-Tu binding assay showed EF-Tu has affinity to all tested tRNAggc precharged with different unnatural amino acids ........................................................ 72 Figure 19 Making 35S-labeled hexapeptide-oligonucleotide conjugates consisting of unnatural amino acids by the E. coli in vitro reconstituted translation system.. ...... 74 Figure 20 Confirmation of peptide-puromycin conjugates generated from rabbit reticulocyte lysate by MALDI-TOF....................................................................... 77 Figure 21 MALDI-TOF confirmation of fMet-(Sar)4-Bio-puromycin made by E. coli in vitro reconstituted translation. ............................................................................... 79 Figure 22 Assessment of proteinase K susceptibility of 35S-labeled peptideoligonucleotide conjugates consisting of different amino acids.............................. 82 Figure 23 Percentages of peptide-oligonucleotide conjugates remaining after 3 hr and 20 hr of proteinase K treatment vs. the composition of peptide-oligonucleotide conjugates ............................................................................................................. 83 Figure 24 Size and purity confirmation of tRNA22(-ca) to tRNA33(-ca) by denaturing urea PAGE analysis. ..................................................................................................... 86 Figure 25 Denaturing urea PAGE analysis of mRNA libraries. ..................................... 87 Figure 26 E. coli in vitro reconstituted translations of V3Ap, V4p, and V5p in the presence of unnatural amino acids.. ..................................................................................... 90 Figure 27 E. coli in vitro reconstituted translations of V3Ap in the presence of increasing amount of Sar-tRNAsnn.......................................................................................... 92  x  Figure 28 E. coli in vitro reconstituted translations of V3Bp in the presence of unnatural amino acids.. ......................................................................................................... 95 Figure 29 Testing the translation inhibitory effect of puromycin in the E. coli in vitro reconstituted translation system using V3B.. ......................................................... 98 Figure 30 E. coli in vitro reconstituted translations of CAC5 template in the presence of either natural or unnatural amino acid over a 6 hour period. ................................ 101 Supplementary Figure 1 Amino acids that are successfully coupled to pdCpA and introduced to the E. coli in vitro reconstituted translation system......................... 119  xi  List of abbreviations 10f-THF  N10-formyltetrahydrofolate  A  adenine; DNA base  a  adenine; RNA base  aaRS  aminoacyl-tRNA synthetase  Abr  L-abrine  Ala  L-alanine  APS  ammonium persulfate  Asp  L-aspartic  Aze  L-azetidine  Bio  biocytin  C  cytosine; DNA base  c  cytosine; RNA base  DMF  N,N-dimethylformamide  E. coli  Escherichia coli  EFG  elongation factor-G  EF-Ts  elongation factor-Ts  EF-Tu  elongation factor-Tu  fMet  N-Formylmethionine  fMet-P  fMet-puromycin-CCA(spacer-9)3(A)21  G  guanine; DNA base  g  guanine; RNA base  , N-methyl-L-tryptophan  acid  xii  Glu  L-glutamate  Gly  L-glycine  Hom  homoproline  IF1  initiation factor one  IF2  initiation factor two  IF3  initiation factor three  Met  methionine  MetRS  methionyl-tRNA synthetase  MTF  methionyl-tRNA formyltransferase  mRNA  messenger RNA  MTF  methionine transformylase  Nle  L-aminohexanoic  N-Me  N-methyl  nns  a randomized codon; n = g, c, a or u; s = g or c  NVOC  4,5-dimethyoxy-2-nitrobenzyl chloroformate  Oic  L-octahydroindole-2-carboxylic  Orn  L-ornithine  Pro  proline  RNase A  ribonuclease A  RRL  rabbit reticulocyte lysate  Sar  L-sarcosine  Ser  L-serine  RF1  release factor 1  acid  acid  xiii  RF2  release factor 2  RF3  release factor 3  T  thymine; DNA base  TEA  triethylamine  TEMED  N,N,N',N'-tetramethylethylenediamine  THF  tetrahydrofuran  tRNA  transfer RNA  Tyr  L-tyrosine  u  uracil; RNA base  Val  L-valine  xiv  1 Introduction 1.1 Methods of building combinatorial peptide libraries Combinatorial peptide libraries have been developed to identify ligands with interesting characteristics or desirable biological activities. Such libraries can be built either synthetically or biologically, each of which has its own advantages and disadvantages. A major advantage with synthetic combinatorial libraries is that they allow the incorporation of building blocks beyond the 20 natural α-amino acids [1]. This ability to access an almost unlimited chemical repertoire is particularly important since peptides containing unnatural or non-proteinogenic amino acids are associated with some promising leads in drug discovery [2-4]. Some disadvantages associated with synthetic libraries are complex decoding process after selection [5, 6], and a lack of serial enrichment of substrates with initial weak binding capacity that could be optimized by sequence evolution [1]. There are several biological methods that have been developed to build combinatorial peptide libraries: phage display [7], yeast display [8], bacterial display [9], ribosome display [10, 11], and mRNA display [12, 13]. All biological methods maintain the great advantage of having easily accessible coding information for each peptide/protein displayed. The first three biological display libraries listed have polypeptides or proteins expressed on the respective microorganism’s surface and the accompanying coding information for each polypeptide or protein are retrievable from the microorganism’s genome. These cell-based libraries allow for the screening and enrichment of peptides/proteins with desirable characteristics, but limited mostly to natural amino acids due to presence of translation factors responsible for editing and  1  proofreading, and have inherent selection biases due to the limitations of the hosts [1416]. The last two biological display methods do not require a living cell; they utilize in vitro translation to produce peptides/proteins according to the given genetic information. Their library size is not limited to the transformation efficiency, and avoids the potential problem of cell toxicity due to expressed protein/peptides faced by living cell-based display methods. In addition, they have the potential to incorporate unnatural amino acids into peptides to make them into more stable and favorable drug candidates. For my research, mRNA display technology is the method of choice as it allows for the incorporation of unnatural amino acids, has a simple decoding scheme, and facilitates the affinity maturation process of selected peptides.  2  1.2 A closer look at mRNA display technology 1.2.1 Puromycin: the physical linkage The unique feature that distinguishes mRNA display technology from the rest is its use of a 3’-puromycin moiety to physically link the genetic information (mRNA) to the encoded information (peptide) (Figure 1). Puromycin (Figure 2) is a universal antibiotic that inhibits cell growth in both prokaryotic and eukaryotic cells; it is also a tool for investigating the catalytic activity of ribosomes [17-20]. Structurally, it mimics an aminoacyl-tRNA molecule; it is able to enter the A site of ribosome in a translation factor-independent fashion, and it forms a covalent bond with the growing peptide chain to cause an early translation termination [21, 22].  Roberts et al. and Nemoto et al. devised a method to attach puromycin molecule to an mRNA template via a puromycin-containing DNA linker; upon translation of this modified mRNA molecule, the newly made peptide becomes attached to the puromycin and thus remain linked to the coding mRNA [12, 13]. The method was optimized by Liu et al. [23]. From a series of experiments, Liu et al. had made several important discoveries. First, by ligating myc RNA to DNA linkers of varying lengths bearing a 3’ puromycin, they evaluated the percentage of mRNA-myc peptide fusion generated by in vitro translations, and they optimized the DNA linker length and linker sequence. Second, by adding increasing amounts of Mg2+ or K+ ions in post-translational reaction and evaluating the percentage of fusion formation, they found that post-translational addition of these ions increased the fusion yield. Third, through performing co-translation reaction of myc and λPPase mRNAs and observing no cross fusion product, they deduced a cis  3  model for fusion formation: fusion formation only occurs between the peptide and the puromycin coming from the same mRNA molecule. Fourth, based on the gel images, their method of attaching puromycin to the end of mRNA molecule didn’t seem to affect the translation process; i.e. only mRNA fusions with full-length myc and λPPase proteins were observed [23]. The last observation was also reported by Nemoto et al. when they assessed the formation of mRNA-tau protein via puromycin [12].  Because puromycin was originally used as a translation inhibitor, attachment of puromycin to the end of an mRNA molecule should in theory interfere with proper translation and cause premature termination of protein synthesis. Both Starck et al. and Miyamoto-Sato et al. offered possible explanations of why translation inhibition was not observed in mRNA display [24, 25]. By evaluating the amount of globin protein being produced by a eukaryotic in vitro translation system in the presence of increasing amounts of free puromycin or puromycin derivatives, Starck et al. demonstrated that puromycin-oligonucleotide conjugates do not inhibit translation as effectively as puromycin alone. The inhibitory effect decreases as the oligonucleotides chain length increases, and the formation of truncated translation product is only observed when the concentration of puromycin derivative is raised. Miyamoto-Sato et al. demonstrated that at low concentration (0.04 µM), puromycin and its derivatives showed no detectable inhibition of protein synthesis in an E. coli S30 extract system and bound specifically to the C-terminus of the full-length protein. They proposed that at low concentrations, puromycin and its derivatives cannot compete with aminoacyl-tRNAs and only have a chance at bonding to proteins at the stop codon, where there is a translational pause in  4  both E. coli and eukaryotes. Based on all the experimental evidence, the specific attachment of puromycin to the fully translated products in mRNA display technology is likely due to a combination of two factors: reduced puromycin inhibitory effect as a result of its conjugation to DNA linker and its low effective concentration as a result of its attachment to mRNA templates.  Overall, the method of using a puromycin derivative to link nascent peptides/proteins to their mRNAs offers a way to track each peptide by its genetic code and makes the selection  and  enrichment  of  peptides  with  desirable  properties  possible.  5  Figure 1 A cartoon representation of puromycin's role in mRNA display technology. (A) Puromycin derivative (purple) is covalently linked to the 3’-end of an mRNA molecule (red). During in vitro translation, aminoacyl-tRNA on the ribosomal A site engages in peptide formation with peptidyl-tRNA on the ribosomal P site. (B) When ribosome reaches the end of the mRNA molecule, it stalls and puromycin, which resembles the 3’-end of an aminoacyl-tRNA molecule, enters the ribosomal A site and engages in peptide formation with peptidyl-tRNA on the ribosomal P site. The resulting peptidyl-puromycin molecule contains a stable amide linkage. (C) The peptide oligonucleotide conjugate can be purified from the translation reaction by affinity chromatography.  6  Figure 2 The chemical structure of puromycin. Puromycin is an aminonucleoside antibiotic with part of the molecule mimicking tyrosyl-tRNA It can enter the A site of the ribosome and cause premature release of peptides.  1.2.2 Making mRNA display based peptide libraries by cell-free translation Due to the requirement for mRNA modification with puromycin derivatives, mRNA display technology is carried out in cell-free translation systems. The cell-free translation systems can be divided into two major categories; crude cell extract from Escherichia coli (E. coli), rabbit reticulocyte, or wheat germ [26, 27] and purified translation system containing purified components from the translation machinery [28]. Crude cell extract is easy to obtain but faces the challenges of energy depletion and protease and nuclease contamination [29, 30]. Translation machinery supplied by this crude cell extract also has difficulty in incorporating unnatural amino acids, since it contains all natural aminoacyltRNA synthetases (aaRS) whose function is to charge a natural amino acid to its cognate tRNA and hydrolyze any mis-acylated tRNA. Efforts have been made to incorporate unnatural amino acids, using methods such as introducing nonsense suppressor tRNA charged with an unnatural amino acid [31], sense suppressor tRNA charged with an 7  unnatural amino acid in endogenous-tRNA-depleted lysates [32] or engineered aaRS/tRNA pair recognizing a particular unnatural amino acid [33]. Each method has its own limitations or biases [34], and is not suitable for making peptide libraries consisting of a large number of unnatural amino acids. In comparison, reconstituted translation system allows the incorporation of unnatural amino acids in a controlled manner and the omission of aaRSs [34]. For my research, I chose to employ an E. coli-based reconstituted translation system to translate an mRNA display library as it is more amenable to the incorporation of multiple unnatural amino acids.  1.3 Introduction of unnatural amino acids into a biological system 1.3.1 Overview of bacterial translation mechanism Messenger RNAs (mRNAs), transfer RNAs (tRNAs) and ribosomes are the three major participants in translation. mRNAs are the transcription products of DNA, and contain the essential coding information for peptides/proteins. An mRNA molecule usually has three basic elements: a 5’ untranslated region for ribosome binding, an intervening coding sequence that usually starts with the codon AUG and ends with one of the three stop codons (UAA, UAG, UGA). The role of a tRNA molecule is to deliver a specific amino acid to the ribosome for translation. The anticodon region of each tRNA is complementary to the codon region of the mRNA, and this pairing between the anticodon and codon dictates the sequence of the peptide to be translated. The ribosome is compositionally and structurally complex, and consists of ribosomal RNA (rRNA) and  8  proteins. In E. coli, the ribosome has a sedimentation coefficient of 70S, and is composed of 30S and 50S subunits. It is on the ribosome where the translation event unfolds [35].  There are a number of proteins that are important in regulating and carrying out translation. AaRS are responsible for pairing the correct amino acid to the correct tRNA. In E. coli, there is one aaRS for each amino acid, except lysine, which has two aaRSs. Methionine transformylase (MTF) and the small molecule (10f-THF) work together to transfer a formyl group to methionine-tRNAfMet to give to N-formylmethionine-tRNAfMet (fMet-tRNAfMet); fMet-tRNAfMet will serve as E. coli’s initiating residue. Initiation factors 1, 2, and 3 (IF1, IF2, and IF3, respectively) are proteins that are important in translation initiation. Initiation of translation requires the binding of IFs, mRNA and fMet-tRNAfMet to the 30S subunit of the ribosome, which is collectively called the 30S initiation complex. The formation of 30S initiation complex promotes the full assembly of the ribosome by binding to the 50S subunit, which is accompanied by the release of IFs. Once the subunits of ribosome are assembled, translation can start. The elongation process of translation is assisted by elongation factors (EF-Tu, EF-Ts, and EF-G). There are three tRNA binding sites on the ribosome: A (aminoacyl), P (peptidyl) and E (exit) site. Amino acid/tRNA complex is escorted by EF-Tu to the ribosome A site. If the anticodon sequence on the tRNA is complementary to the codon sequence of the template, EF-Tu releases aminoacyl-tRNA through hydrolysis of GTP. The EF-Tu is regenerated by EF-Ts by exchanging GDP for GTP. The A site aminoacyl-tRNA engages in peptide formation with the peptidyl-tRNA on the P site. Once a peptide bond is formed, the ribosome moves one codon downstream of the template and frees up the A site for a new  9  aminoacyl-tRNA. The free tRNA leaves the ribosome from E site. This cycle repeats until the termination signal is reached. There are three termination signals or stop codons: uaa, uag or uga. When a termination signal is reached by the A site, release factors (RF1, RF2 and RF3) bind and cause the peptide to be released from the ribosome, followed by disassembly of the ribosomal complex [35].  Proper translation requires not only the essential translation components mentioned above, it also requires optimal salt concentrations and an energy source [35]. Ribosomal assembly and disassembly can be favored with a high concentration of Mg2+ in the former case, or high concentrations of Na+ or K+ in the latter case. IF-2, EF-Tu and EF-G require GTP to function properly, and aminoacylation of tRNAs requires ATP. Therefore, it is important to ensure that the translation mixture is supplemented with enough ATP and GTP molecule s [35]. With a good understanding of the essential factors involved in the translation machinery, the system maybe manipulated to be more suitable for in vitro translation and incorporation of unnatural amino acids. 1.3.2 Reconstituted translation system The reconstituted translation system is also known as PURE (Protein synthesis Using Recombinant Elements) system. The idea of purifying individual translation components and combining them to make an artificial translation system was first introduced by Weissbach’s group; they were also the first group to define the ‘essential factors’ involved in protein synthesis [36]. Shimizu et al. then introduced the hexa-histidine based purification scheme for the essential protein factors, and the concept of omitting release factors for efficient incorporation of an unnatural amino acid [28]. Around the same time,  10  Forster et al. developed a simplified reconstituted translation system that used a similar purification strategy, and further separated the aminoacylation of tRNAs from the translation reaction, and thus allowed the omission of aaRS from the translation reaction [37]. This idea of omitting aaRSs from the translation reaction is significant in that without these key proofreading factors, the fidelity of the translation reaction can be challenged and the genetic code can be rewritten. Overall, the translation apparatus can be reconstituted for maximal flexibility and manipulation by containing only the key components: ribosomes, initiation and elongation factors, mRNA and aminoacyl-tRNAs (Table 1).  11  Table 1 Components of the E. coli in vitro reconstituted translation system. The system contains buffer that allows energy regeneration, ions for ribosome assembly, factors for charging initiator tRNAfMet (MetRS, tRNAfmet, [35S]Met, 10f-THF, MTF), initiation factors (IF1, IF2, IF3), elongation factors (EF-Tu, EF-Ts, EF-G), ribosomes and templates. The 10X translation buffer contains 80 mM putrescine, 10 mM spermidine, 50 mM K3PO4, 950 mM KCl, 50 mM NH4Cl. 10 mM DTT, 30 mM ATP, 25 mM GTP, 60 mM phosphoenolpyruvate, and 0.1 mg/ml pyruvate kinase. Translation Components  Target Concentration (µM)  10X Translation Buffer  1X  10X CaCl2 (5 mM)  500  50X Mg(OAc)2 (1M)  20,000  IF1  0.5  IF2  0.5  IF3  0.5  EF-Tu  6.6  EF-Ts  0.5  EF-G  0.5  MetRS  0.5  10f-THF  30  MTF  0.5  Ribosomes  0.8  tRNAfmet  0.8  Unnatural amino acids-tRNA  variable  [35S]Met  0.8  Template  0.8  1.3.3 Codon reassignment and tRNA mutagenesis A number of methods have long been used to introduce unnatural amino acids into peptides/proteins. However, most methods are limited by one or more factors such as the  12  type of unnatural amino acids to be incorporated, incorporation efficiency, and the position of incorporation on a peptide/protein. Schultz and coworkers were the first group to introduce a biosynthetic method that allows the site-specific incorporation of a large variety of unnatural amino acids into proteins [38, 39]. This biosynthetic method encompasses several main concepts that were important in governing the development of all subsequent methods for unnatural amino acids incorporation. First, anticodon-codon recognition is independent of the amino acid charged on the tRNA. In other words, any amino acid can be delivered to a codon, as long as the carrier tRNA contains the complementary anti-codon region. Secondly, the translation machinery is tolerant of a wide range of unnatural amino acids. Taking these concepts together, an amino acid of interest can theoretically be incorporated at any desired position of a peptide if it has a unique codon and an existing tRNA with a reassigned complementary anti-codon region, which can be done by standard PCR mutagenesis of tRNA anti-codon region. This mutagenized tRNA can then be charged with the amino acid of interest by a generalized two-step reaction that was first developed by Hecht et al. and improved by Robertson et al. [38, 40, 41]. To increase the versatility of the translation system towards the type and number of unnatural amino acids to be incorporated, Forster et al. introduced the concept of completely reassigning the sense codons to unnatural amino acids in combination with the elimination of competing factors: aaRSs, natural occurring aminoacyl-tRNAs and release factors; they had achieved making peptides containing multiple unnatural amino acids [42]. This work sheds light on the possibility of making a peptide library containing a variety of unnatural amino acids for the discovery of small peptide-based ligands.  13  1.3.4 Choices of unnatural amino acids and selection target Cyclic amino acids can reduce the conformation flexibility of peptides. This conformational constraint may sometimes result in more potent ligands for biological receptors [6]. Likewise, N-methyl amino acid containing substances have been associated with good proteolytic stability and improved pharmacokinetic properties found in several natural products, e.g. vancomycin, cyclosporin, actinomycine D, and other lead compounds [43-48]. The chosen unnatural amino acids are N-methyl and cyclic amino acids to add unique features to the peptides that will be created. Some of the unnatural amino acids that I use in my research have demonstrated to be successfully incorporated into proteins by nonsense tRNA suppression, while the rest are yet to be determined. As a proof of concept, a well-studied pharmaceutical target, thrombin, will eventually be used to test the feasibility of my method. Thrombin is a multi-functional serine protease involved in blood coagulation and thrombosis, inflammation and wound repair in tissue injury. It is a highly specific and tightly regulated enzyme. The discovery of the optimal peptide sequence for binding to thrombin has contributed to the development of direct thrombin inhibitors that may be used as anticoagulants. The optimal cleavage sites for thrombin have been determined to be A-B-Pro-Arg-||-X-Y and Gly-Arg-||-Gly; where A and B are hydrophobic amino acids and X and Y are nonacidic amino acids [49]. Some of the chosen natural/unnatural amino acids will resemble natural amino acids that make up the optimal cleavage site to the target, e.g. proline, N-Me-Orn, Sar, hydrophobic and nonacidic N-Me and cyclic amino acids. This selection bias is intended to increase the likelihood of selecting ligands of thrombin with good affinity and specificity.  14  1.4 Library design When designing an mRNA library for the purpose of generating peptide libraries, several factors need to be taken into consideration: the non-coding region of the sequence, the length of the coding region, and the size of the randomized region in the coding region. The 5’ non-coding region of the mRNA library needs to contain at least a ribosomal binding sequence (e.g., Shine-Dalgarno sequence) [50] and sometimes an epsilon enhancer sequence is also included to enhance ribosomal binding [37]. These sequences allow the binding of ribosome to the mRNA to ensure proper translation initiation in E. coli. The coding region of the library is designed to have a region of random codons to allow for the random incorporation of a pool of amino acids followed by a fixed set of codons that code for a specific amino acid. Since the generated peptide library is intended to target the active site of thrombin, and the basic binding motif for recognition by the active site consists of 3 amino acids, the random coding region needs to contain at least 3 random codons.  1.5 Research hypothesis It is possible to build a library of peptide-oligonucleotide conjugates consisting of unnatural amino acids using an E. coli in vitro reconstituted translation system.  1.6 Specific aims 1. Demonstrate that an E. coli in vitro reconstituted translation system has the ability to make small peptide-oligonucleotide conjugates.  15  2. Chemically conjugate the selected unnatural amino acids to tRNAs to make them accessible to the E. coli in vitro reconstituted translation system. 3. Demonstrate that the selected unnatural amino acids are recognizable by the E. coli in vitro reconstituted translation system and can be incorporated successively. 4. Synthesize tRNAs required for generation of a library of peptide-oligonucleotide fusions. 5. Build mRNA library in preparation for making a library of peptideoligonucleotide fusions. 6. Demonstrate that peptide-oligonucleotide fusions of defined length can be made when mRNA library is translated by an E. coli in vitro reconstituted translation system in the presence of tRNAs charged with unnatural amino acids  16  2 Experimental methods Reagents were obtained from the following sources: Ambion Inc. (Austin, TX, USA) DEPC treated water; RNAsecure; RNase cocktailTM; linear acrylamide Amersham Biosciences (Piscataway, NJ, USA) Oligo(dT)-cellulose type 7; storage phosphor screen; PD-10 desalting column Bachem Americas, Inc. (Torrance, California, USA) N-Me-Asp; Aze; N-Me-Ser, N-Me-Tyr, Oic, N-Me-Nle Bio-Rad Laboratories Ltd. (Mississauga, Ontario, Canada) Mini-PROTEAN Tetra Electrophoresis System; Bio-rad Silver Stain Plus Kit Dr. Jack Szostak at Massachusetts General Hospital (Boston, Massachusetts, USA) T4 RNA ligase EMD Biosciences Inc. (Gibbstown, NJ, USA) Red nova® lysate Fisher Scientific Canada (Ottawa, Ontario, Canada) Acetonitrile; butanol; dichloromethane; diethyl ether; DMF, dioxane; ethanol; 1.5 ml microcentrifuge tubes; ethyl acetate, glacial acetic acid; glycerol; hydrochloric acid; methanol; 5 ml syringe; TEA; THF; tetrabutylammonium acetate; vertical electrophoresis system; slab gel dryer SDG5040 Fluka and Riedel-de Haën, Sigma-Aldrich Laborchemikalien GmbH (Seelze, Germany) Biocytin Integrated DNA Technologies (Toronto, Ontario, Canada) All the primers used in transcription reactions; V3B mRNA library; primer 1 used in DNA library generation Invitrogen (Carlsbad, CA, USA) Taq DNA polymerase 500U; 10bp DNA ladder National Diagnostics (Atlanta, Georgia, USA) SequaGel® and ProtoGel® reagents New Brunswick Scientific (Edison, NJ, USA) Incubator Shakers Innova 44  17  New England Biolabs (Ipswich, Massachusetts, USA) All restriction digest enzymes and their respective buffers; T4 RNA ligase buffer; T4 DNA ligase (100,000 units/ml) Newport corporation (Mountain View, CA, USA) Xenon lamp PerkinElmer Life And Analytical Sciences, Inc. (Waltham, Massachusetts, USA) 35 S-methionine (1175Ci/mmol); [14C]Ala (165.0 mCi/mmol) QIAGEN Science (Maryland, MD, USA) QIAquick Gel Extraction Kit Roche Applied Science (Laval, Quebec, Canada) PWO SuperYield DNA polymerase Sigma Chemical Co. (St. Louis, Missouri, USA) Initiator tRNA and total tRNA from E. coli; phenol:chloroform:isoamyl alcohol 25:24:1 pH 8.0; ribonuclease A from bovine pancreas (for molecular biology, ≥70 Kunitz units/mg protein, lyophilized power); streptavidin from Streptomyces avidinii, lyophilized powder, ≥13 units/mg protein; proteinase K from Engyodontium album, ≥30 units/mg protein Schricks Laboratories (Jona, Switzerland) (6S)-5-formyl-5,6,7,8-tetrahydrofolic acid, calcium salt Stratagene (La Jolla, California, USA) QuickChange® site-directed mutagenesis kit TriLink Biotechnologies (San Diego, California, USA) GCC1 template; CAC5 template; V3 DNA template; V4 DNA template; V5 DNA template; splint; primer 2; pF30P VWR International (Mississauga, Ontario, Canada) Nutating mixer; high performance centrifuge Nanosep®Centrifugation Device 10K MWCO  tubes,  polypropylene,  sterile;  The templates used in the experiments are listed in Table 2. Some of these templates are custom synthesized (GCC1, CAC5, CAC1, V3B) and others are transcribed from DNA sequence and modified accordingly. They all share the following features: a 5’ noncoding sequence that allows for ribosome binding followed by a short coding sequence.  18  Table 2 Template sequences. The name of each template is listed on the left column, and the sequences are listed on the right column. The small letters represent RNA, the capital letters represent DNA, the bold letters represent the coding region of each template, the non-bold letters represent the non-coding region of each template, and the spacer-9 is triethylene glycol. Templates  Sequences (5’-3’)  GCC1  ggaggacgaa aug gcc (A)21(Spacer-9)3ACC(puromycin)  CAC5  ggaggacgaa aug gcc gcc gcc gcc cac (A)21(Spacer-9)3ACC(puromycin)  CAC1  ggaggacgaa aug cac gcc gcc gcc gcc (A)21(Spacer-9)3ACC(puromycin)  V3A  gggttaactttagtaaggaggtagatatacagg aug (nns)3 ccc ggg aaa ggg ccc aaa  V3Ap  gggttaactttagtaaggaggtagatatacagg aug (nns)3 ccc ggg aaa ggg ccc aaa (A)21(Spacer- 9)3ACC(puromycin)  V3B  ggaggacgaa aug (nns)3 aaa ggg ccc aaa ccc ggg  V3Bp  ggaggacgaa aug (nns)3 aaa ggg ccc aaa ccc ggg (A)21(Spacer9)3ACC(puromycin)  V4  gggttaactttagtaaggaggtagatatacagg aug (nns)4 ccc ggg aaa ggg ccc aaa  V4p  gggttaactttagtaaggaggtagatatacagg aug (nns)4 ccc ggg aaa ggg ccc aaa (A)21(Spacer-9)3ACC(puromycin)  V5  gggttaactttagtaaggaggtagatatacagg aug (nns)5 ccc ggg aaa ggg ccc aaa  V5p  gggttaactttagtaaggaggtagatatacagg aug (nns)5 ccc ggg aaa ggg ccc aaa (A)21(Spacer-9)3ACC(puromycin)  19  2.1 Derivatizing tRNAs from tRNAalaW by PCR mutagenesis tRNAs with different anticodon sequences were generated by mutating the DNA sequence coding the anti-codon region of the tRNAalaW from E. coli. The gene for E. coli alaW tRNA was cloned into the pUC19 vector by the summer student Adarsh Patel (2006). Two mutagenic primers were designed for each tRNA. Forward primer has the sequence of 5'- GCT GGG AGA GCG CTT GCA TNN NAT GCA AGA GGT CAG CGG -3', and the reverse primer has the sequence 5'- CCG CTG ACC TCT TGC ATN NNA TGC AAG CGC TCT CCC AGC -3', where NNN stands for the targeted anticodon/codon sequence (refer to Table 3 for details). There were 12 pairs of primers designed; each pair of primers would give to a tRNA with a unique anti-codon. Each 50 µl mutagenic PCR reaction was set up to contain 10 ng of pUC19-alaW, 0.25 µM of each forward and reverse primers, 0.2 mM dNTP and 0.5 µl of Pwo enzyme in 1X Pwo buffer. PCR was performed as follows: initial denaturation at 95oC for 30 s, denaturation at 95oC for 30 s, primer annealing at 55oC for 1 min, extension at 68oC for 6 min, 18 cycles, final extension at 68oC for 2 min followed by 16oC at the end of the PCR. At the end of each PCR reaction, 1 µl of Dpn I was added to each reaction tube to remove pUC19-alaW, and the reaction was incubated at 37oC for 1.5 hr. 5 µl of each crude PCR mixture was then added to 50 µl of XL-1 Blue competent cells and mixed gently. The cells were incubated on ice for 30 minutes, heat-shocked at 42°C for 40 s, and immediately placed on ice for 2 min. 950 µl of LB broth was then added to the cells and cells were incubated at 37°C for 1 hr in the incubator (Innova 44) with shaking at 250 rpm. The cells were then spun down at 1,699 x g for 1 min, and 900 µl of the supernatant was removed. The cells were suspended in the remaining 100 µl supernatant and plated on pre-warmed LB/Amp plates.  20  The cells were allowed to grow at 37°C overnight. 2 colonies were picked from each plate the next day and each colony was inoculated in 5 ml LB containing ampicillin and grew overnight. The plasmids were then isolated using QIAprep Spin Miniprep Kit. The isolated DNAs were then sent to UBC Nucleic Protein Sequencing (NAPS) Unit for sequencing confirmation.  21  Table 3 List of tRNAs and primers needed for making tRNA library. Each tRNA has its own assigned number, followed by the codon it would recognize by Watson-Crick paring and the mutagenic primers needed to make the genes coding for these tRNAs. tRNA  Codon  22  aug  23  agg  24  agc  25  acc  26  uug  27  ucc  28  guc  29  ggc  30  gcg  31  cug  32  cgc  33  ccg  Mutagenic Forward Primer (5’-3’)  Mutagenic Reverse Primer (5’-3’)  GCT GGG AGA GCG CTT GCA TCA TAT GCA AGA GGT CAG CGG GCT GGG AGA GCG CTT GCA TCC TAT GCA AGA GGT CAG CGG GCT GGG AGA GCG CTT GCA TGC TAT GCA AGA GGT CAG CGG GCT GGG AGA GCG CTT GCA TGG TAT GCA AGA GGT CAG CGG GCT GGG AGA GCG CTT GCA TCA AAT GCA AGA GGT CAG CGG GCT GGG AGA GCG CTT GCA TGG AAT GCA AGA GGT CAG CGG GCT GGG AGA GCG CTT GCA TGA CAT GCA AGA GGT CAG CGG GCT GGG AGA GCG CTT GCA TGC CAT GCA AGA GGT CAG CGG GCT GGG AGA GCG CTT GCA TCG CAT GCA AGA GGT CAG CGG GCT GGG AGA GCG CTT GCA TCA GAT GCA AGA GGT CAG CGG GCT GGG AGA GCG CTT GCA TGC GAT GCA AGA GGT CAG CGG  CCG CTG ACC TCT TGC ATA TGA TGC AAG CGC TCT CCC AGC CCG CTG ACC TCT TGC ATA GGA TGC AAG CGC TCT CCC AGC CCG CTG ACC TCT TGC ATA GCA TGC AAG CGC TCT CCC AGC CCG CTG ACC TCT TGC ATA CCA TGC AAG CGC TCT CCC AGC CCG CTG ACC TCT TGC ATT TGA TGC AAG CGC TCT CCC AGC CCG CTG ACC TCT TGC ATT CCA TGC AAG CGC TCT CCC AGC CCG CTG ACC TCT TGC ATG TCA TGC AAG CGC TCT CCC AGC CCG CTG ACC TCT TGC ATG GCA TGC AAG CGC TCT CCC AGC CCG CTG ACC TCT TGC ATG CGA TGC AAG CGC TCT CCC AGC CCG CTG ACC TCT TGC ATC TGA TGC AAG CGC TCT CCC AGC CCG CTG ACC TCT TGC ATC GCA TGC AAG CGC TCT CCC AGC  GCT GGG AGA GCG CTT GCA TCG GAT GCA AGA GGT CAG CGG  CCG CTG ACC TCT TGC ATC CGA TGC AAG CGC TCT CCC AGC  22  2.2 Generation of tRNAs by in vitro transcription 2.21 PCR To generate tRNAs that lack the last two dinucleotides (c, a) at the 3’ end, 136-bp fragments were first amplified from the coding plasmids by PCR using Pwo (Roche) with the universal M13 forward primer 5’-GTAAAACGAC GGCCAGT-3’ and reverse primer -CAalaW 5’- GTGGAGCTAA GCGGG -3’. To generate full-length tRNAs, the M13 forward primer and reverse primer FLalaW 5’- TGGTGGAGCT AAGCGG -3’ were used. Each 200 µl PCR reaction contained approximately 200 ng of plasmids, 1 µM forward primer, 1 µM reverse primer, 0.5 mM dNTPs, and 4 µl of Pwo in 1X Pwo buffer. PCR was performed as the following: initial denaturation at 95oC for 1 min, denaturation at 95oC for 30 s, primer annealing at 54oC for 1 min, extension at 72oC for 1 min, 29 cycles, final extension at 72oC for 10 min followed by 16oC at the end of the PCR. After the PCR reaction, an equal volume of phenol:chloroform:isoamyl alcohol 25:24:1 pH 8.0 was added to the PCR reaction tube; the mixture was vortexed for 10 s followed by centrifugation at 20,817 x g for 1 min. The top aqueous layer was then carefully transferred to a new Eppendorf tube, into which 1/10th volume of 3.0 M sodium acetate (pH 5.2) and 2.5X volume of 100% ethanol were added and well mixed. The mixture was then kept at -80oC for 30 to 40 min to allow PCR products precipitate; the precipitated PCR product was recovered by centrifugation at 20,817 x g, 4oC for 10 min. The supernatant was decanted and 500 µl 70% ethanol was added to the pellet to wash away extra salts and the tube was centrifuged again at 20,817 x g, 4oC for 5 min. The supernatant was again decanted. The pellet was dried in a speed vacuum centrifuge and suspended in 200 µl DEPC treated H2O. A 1 µl aliquot of the sample was diluted in 99 µl  23  of deionized H2O to determine its concentration using the equation C (µg/ml) = A260/0.020. Typically, a PCR reaction yielded 40 µg DNA, and this provided enough templates to perform 1 ml in vitro transcription reaction using T7 RNA polymerase.  2.22 In vitro transcription using T7 RNA polymerase Approximately 40 µg of PCR products were used as templates in 1 ml transcription reaction to generate tRNA. The transcription reaction contained 200 µl of PCR products (40 µg), 200 µl of 5X transcription buffer (0.2 M Tris-HCl pH 8.0, 40 mM MgCl2, 10 mM spermidine-(HCl)3, 125 mM NaCl), 50 µl of 100 mM DTT, 250 µl of 20 mM NTP, and 200 µl of RNA Secure; the 900 µl mixture was incubated at 60oC for 10 min, immediately on ice for 2 min, and to which 100 µl of 0.35 mM homemade T7 RNA polymerase was added. The mixture was well mixed by gentle pippetting before it was placed in a 37oC incubator to allow the reaction to proceed overnight (approximately 16 hr). The reaction was quenched by adding 1/20th volume of 0.5 mM EDTA (pH 8.0) and incubated at room temperature for 10 min. The reaction mixture was then extracted with phenol:chloroform:isoamyl alcohol 25:24:1 pH 8.0 and ethanol-precipitated to remove proteins and salts (refer to section 2.11 for experimental procedure). The RNA pellets were air dried on ice and suspended in approximately 200 µl DEPC treated H2O.  2.23 Purification of tRNAs The crude transcription products were further purified by 12% denaturing urea PAGE. A (20 x 20 cm) 12% urea gel was set up as follows: 48 ml of SequaGel concentrate, 42 ml of SequaGel diluent, and 10 ml of SequaGel buffer were individually measured and 24  pooled in an Erlenmeyer flask. The mixture was degassed while stirring under a vacuum for 1 min. The gel apparatus (Thermo Fisher Scientific) was set up per manufacture’s instruction. When ready to pour, 800 µl of 10% ammonium persulfate (APS) was added to the mixture, and the mixture was swirled to mix before 40 µl of TEMED was added and mixed again. Once poured, the gel was allowed to solidify at room temperature for 30 min to 1 hr. The vertical electrophoresis system is used to carry out gel electrophoresis. The gel was pre-run in 1X TBE buffer (0.1M Tris/Borate/EDTA) for 30 min at 20 W. When ready, the wells of the gel were flushed using a syringe to remove urea. An equal volume of 2X RNA gel loading dye (95% formamide, 5 µM EDTA, 0.0025% w/v SDS, 0.0025% w/v bromophenol blue) was added to the crude transcription products and the mixture was heated at 65oC for 3 min before loaded onto the gel. Around 100 µl of sample was loaded into each well and transcription products from 1 ml transcription usually took 4 or 5 wells. In a separate well, 5 µg of control tRNA with a size identical to the sample was loaded as a size control. Electrophoresis was carried out at 50 W until the dye front reached the bottom. The gel was then removed from the glass plates and placed on a saran wrap on top of a TLC plate. By shining a UV lamp on the gel, RNA products were visualized by UV shadowing and were cut out from the gel using a razor. The gel pieces were crushed by passing through a sterile 5 ml syringe into a 15 ml centrifuge tube containing 10 ml of gel elution buffer (0.5 M ammonium acetate, 0.1 mM EDTA, 0.1% SDS). The tRNAs were allowed to elute from the gel pieces by incubation at room temperature overnight while mixing on a nutating mixer. The next day, the tube was centrifuged at 1000 x g for 5 min to spin down the gel pieces and the supernatant was carefully transferred to a 50 ml centrifuge tube. The gel pieces were washed with 5 ml of  25  deionized H2O, spun down, and the supernatant was combined with the isolated supernatant from the previous step. To the collected supernatant, 30 ml of butanol was then added, mixed by shaking, and centrifuged at 500 x g for 5 min. The butanol layer was removed, and 5X volume of butanol was added to the remaining aqueous layer and the procedure repeated until the aqueous layer was reduced to around 1 ml. The aqueous layer was then transferred to Eppendorf tubes and subjected to ethanol precipitation (refer to 2.21 for details). The tRNA pellets were air dried on ice and suspended in a total volume of around 100 µl of deionized H2O. The concentration of the tRNAs was quantified by UV absorbance at 260 nm using the equation C (µg/ml) = A260/0.025. For purified tRNAs that lacked the last dinucleotides (tRNA(-ca)), their concentrations were adjusted to 0.67 µg/µl in 6.7 mM HEPES pH 7.5 and refolded by heating at 94ºC for 3 min and slowly cooling to ambient temperature. The purified tRNA were stored at -80 ºC.  2.3 Chemical modifications of unnatural amino acids 2.31 Synthesis of N-NVOC-N-Me-Asp-cyanomethyl ester and N-NVOC-N-Me-Asp pdCpA The overall reaction scheme is shown in Figure 3. In a 25 ml round bottom flask, 50 mg of N-Me-Asp-OH was dissolved in 1 ml of 10% NaCO3 and 0.5 ml of dioxane and cooled to 4oC in an ice bath. To N-Me-Asp-OH, 93.7 mg of 4,5-dimethyoxy-2-nitrobenzyl chloroformate (NVOC-Cl, Sigma) dissolved in 1.6 ml dioxane:THF (1:1) was slowly added. The reaction was protected by aluminum foil and stirred at 0oC for 1 hr followed by room temperature overnight. The reaction was quenched by addition of 30 ml of  26  distilled water, and extracted twice with 10 ml diethyl ether. The aqueous layer was kept and acidified to pH 2 using concentrated HCl, and extracted twice with 10 ml ethyl acetate. The organic layer was rotary evaporated to give to yellow colored solid. To the N-NVOC-N-Me-Asp-OH, 1.5 ml dry DMF, 1.5 ml chloroacetonitrile and 1 ml dry TEA were added while the reaction was stirred under nitrogen at room temperature. The reaction was allowed to proceed for 24 hr while protected by aluminum foil. The solvents of the crude reaction mixture were removed by rotary evaporation, and the product NNVOC-N-Me-Asp cyanomethyl ester was purified by silica gel chromatography in 99:1 CH2Cl2:MeOH. The collected fractions were rotary evaporated and reduced to yellow oil. The yield was 40.8% (20.8 mg) and the mass was confirmed by ESI-MS, expected [M+H]+, 465.4; observed [M+H]+, 465.2. The dinucleotide pdCpA (5 mg, synthesized by Dr. Galpin) was dissolved in 120 µl of dry DMF and transferred to the flask containing 18.9  mg  N-NVOC-N-Me-Asp  cyanomethyl  ester.  A  catalytic  amount  of  tetrabutylammonium acetate was added to catalyze the reaction. The reaction was stirred at room temperature for 2 hr. The crude sample was purified by C18 semiprep HPLC with the following condition: Solvent A: 25 mM NH4OAc:CH3CN (pH 4.5), 95:5. Solvent B: 25 mM NH4OAc:CH3CN (pH 4.5), 10:90. Gradient: 0%-100%B in 70 min. Flow: 10 ml/min. The HPLC was monitored at both 260 nm (pdCpA ε260 = 23000 cm1  M-1) and 350 nm (NVOC ε350 = 6336 cm-1M-1). The peaks that showed absorbance at  260 nm that were at least 1X greater than the absorbance in 350 nm were collected. Two fractions were collected, one had a retention time of 29 min and the other had a retention time of 38.5 min. The collected fractions were lyophilized and the solids were redissolved in 5 ml of 10 mM acetic acid/CH3CN and lyophilized again. This procedure  27  was repeated one more time before a sample was taken from the pale yellow solids to confirm the mass by ESI-MS. For N-NVOC-N-Me-Asp pdCpA, the expected [M+H]+ is 1043.3. No signals matching [M+H]+ or [M+Na]+ were detected.  Figure 3 The synthetic scheme of converting N-Me-Asp to N-NVOC-N-Me-Asp pdCpA.  2.32 Synthesis of N-NVOC-Aze cyanomethyl ester and N-NVOC-Aze pdCpA The overall synthetic scheme is shown in Figure 4. In a 25 ml round bottom flask, 28.3 mg of Aze was dissolved in 1 ml of 10% NaCO3 and 0.5 ml of dioxane and cooled to 4oC in an ice bath. NVOC-Cl (77.2 mg, Sigma) was dissolved in 1.6 ml dioxane:THF (1:1) 28  and slowly added to Aze. The reaction was protected by aluminum foil and stirred at 0oC for 1 hr followed by room temperature over night. The reaction was quenched by addition of 15 ml of distilled water. The reaction was extracted twice with 5 ml diethyl ether. The aqueous layer was kept and acidified to pH 2 using concentrated HCl, and extracted twice with 5 ml ethyl acetate. The organic layer was dried with MgSO4, filtered, and rotary evaporated to give to a yellow colored film. To the N-NVOC-Aze, 1.5 ml dry DMF, 1.5 ml chloroacetonitrile and 1 ml dry TEA were added while the reaction was stirred under nitrogen at room temperature. The reaction was allowed to proceed for 24 hr while protected by aluminum foil. The solvents of the crude reaction mixture were removed by rotary evaporation, and the product N-NVOC-Aze cyanomethyl ester was purified by silica gel chromatography in 97:3 CH2Cl2: MeOH. The collected fraction was rotary evaporated and reduced to yellow oil. The yield was 71% (0.1 g) and the mass was confirmed by ESI-MS, expected [M+H]+, 380.3; observed [M+H]+, 380.2, [M+Na]+, 402.2. The dinucleotide pdCpA (5 mg) was dissolved in 120 µl of dry DMF and transferred to the flask containing 15.2 mg N-NVOC-Aze cyanomelthyl ester. A catalytic amount of tetrabutylammonium acetate was added to catalyze the reaction. The reaction was stirred at room temperature for 5 hr. The crude sample was purified by C18 semiprep HPLC with the following condition; Solvent A: 25 mM NH4OAc:CH3CN (pH 4.5), 95:5. Solvent B: 25 mM NH4OAc:CH3CN (pH 4.5), 10:90. Gradient: 0%-100%B in 70 min. Flow: 10 ml/min. Two fractions with retention time of 25 min and 28 min respectively, were collected and lyophilized. The solids were redissolved in 5 ml of 10 mM acetic acid/CH3CN and lyophilized again. This procedure was repeated one more time before the pale yellow solids was subjected to ESI-MS to confirm the mass. The calculated mass  29  of N-NVOC-Aze pdCpA was [M+H]+, 959.8. No signals corresponding to [M+H]+ or [M+Na]+ were detected in either fractions.  Figure 4 The synthetic scheme of converting Aze to N-NVOC-Aze pdCpA.  2.33 Synthesis of N-NVOC-Bio cyanomethyl ester and N-NVOC-Bio pdCpA The overall synthetic scheme is shown in Figure 5. In a 50 ml round bottom flask, 100 mg of Bio and 57.24 mg of sodium carbonate were dissolved in a mixture of 15 ml water and 10 ml THF. The mixture was cooled to 4oC in an ice bath. Meanwhile, 74.4 mg of  30  NVOC-Cl, was dissolved in 10 ml THF and slowly added to the Bio mixture. The reaction was protected by aluminum foil and stirred at 0oC for 1 hr followed by room temperature over night. The reaction mixture was rotary evaporated to dryness. The NNVOC-Bio was in yellow solid form. To the N-NVOC-Bio, 3 ml dry DMF, 3 ml chloroacetonitrile and 800 µl dry TEA were added while the reaction was stirred under nitrogen at room temperature. The reaction was allowed to proceed for 24 hr while protected by aluminum foil. The solvents of the crude reaction mixture were removed by rotary evaporation, and the product N-NVOC-Bio cyanomethyl ester was purified by silica gel chromatography in 90:10 CH2Cl2:MeOH. The collected fraction was rotary evaporated and reduced to yellow oil. The yield was 94% (166 mg) and the mass was confirmed by ESI-MS, expected [M+H]+, 651.7; observed [M+H]+ 651.3. The dinucleotide pdCpA (15 mg) was dissolved in 120 µl of dry DMF and transferred to the flask containing 24.3 mg N-NVOC-Bio cyanomethyl ester. A catalytic amount of tetrabutylammonium acetate was added to catalyze the reaction. The reaction was stirred at room temperature for 2 hr. The crude sample was purified by C18 semiprep HPLC. Solvent A: 25 mM NH4OAc:CH3CN (pH 4.5), 95:5. Solvent B: 25 mM NH4OAc:CH3CN (pH 4.5), 10:90. Gradient: 0%-100%B in 70 min. Flow: 10 ml/min. Two fractions with retention time of 15.7 min and 16.6 min respectively, were collected and lyophilized. The solids were redissolved in 5 ml of 10 mM acetic acid/CH3CN and lyophilized again. This procedure was repeated one more time before the pale yellow solids was subjected to ESI-MS to confirm the mass. For both isolation fractions, the calculated mass for NNVOC-Bio pdCpA was [M+H]+, 1230; the observed masses were [M+H]+, 1230.2; [M+Na]+, 1252.3.  31  Figure 5 The synthetic scheme of converting Bio to N-NVOC-Bio pdCpA.  2.34 Synthesis of N-NVOC-N-Me-Ser cyanomethyl ester The overall synthetic scheme is shown in Figure 6. In a 10 ml round bottom flask, 33.4 mg of N-Me-Ser-OH was dissolved in 1 ml of 10% NaCO3 and 0.5 ml of dioxane and cooled to 4oC in an ice bath. Meanwhile, 77.2 mg of NVOC-Cl was dissolved in 1.6 ml dioxane:THF (1:1) and slowly added to N-Me-Ser-OH. The reaction was protected by aluminum foil and stirred at 0oC for 1 hr followed by room temperature over night. The reaction was quenched by addition of 15 ml of distilled water. The reaction was extracted 32  twice with 5 ml of diethyl ether, the aqueous layer was kept and acidified to pH 2 using concentrated HCl, and extracted twice with 5 ml of ethyl acetate. The organic layer was dried with MgSO4, filtered and rotary evaporated to give to a yellow colored film. To the N-NVOC-N-Me-Ser-OH, 1.5 ml dry DMF, 1.5 ml chloroacetonitrile and 1 ml dry TEA were added while the reaction was stirred under nitrogen at room temperature. The reaction was allowed to proceed for 24 hr while protected by aluminum foil. The solvents of the crude reaction mixture were removed by rotary evaporation, and the product NNVOC-N-Me-Ser cyanomethyl ester was purified by silica gel chromatography in 97:3 CH2Cl2: MeOH. The collected fraction was rotary evaporated and reduced to yellow oil. The yield was 37% (32.4 mg) and the mass was confirmed by ESI-MS, expected [M+H]+, 398.3; observed [M+H]+, 398.0.  Figure 6 The synthetic scheme of converting N-Me-Ser to N-NVOC-N-Me-Ser cyanomethyl ester.  33  2.35 Synthesis of N-NVOC-N-Me-Tyr cyanomethyl ester and N-NVOC-N-Me-Tyr pdCpA The overall synthetic scheme is shown in Figure 7. In a 10 ml round bottom flask, 54.7 mg of N-Me-Tyr-OH was dissolved in 1 ml of 10% NaCO3 and 0.5 ml of dioxane and cooled to 4oC in an ice bath. NVOC-Cl (77.2 mg, Sigma) was dissolved in 1.6 ml dioxane:THF (1:1) and slowly added to N-Me-Tyr-OH. The reaction was protected by aluminum foil and stirred at 0oC for 1 hr followed by room temperature over night. The reaction was quenched by addition of 15 ml of distilled water. The reaction was extracted twice with 5 ml of diethyl ether, the aqueous layer was kept and acidified to pH 2 using concentrated HCl, and extracted twice with 5 ml of ethyl acetate. The organic layer was dried with MgSO4, filtered and rotary evaporated to give to a yellow colored film. To the N-NVOC-N-Me-Tyr-OH, 1.5 ml dry DMF, 1.5 ml chloroacetonitrile and 1 ml dry TEA were added while the reaction was stirred under nitrogen at room temperature. The reaction was allowed to proceed for 24 hr while protected by aluminum foil. The solvents of the crude reaction mixture were removed by rotary evaporation, and the product NNVOC-N-Me-Tyr cyanomethyl ester was purified by silica gel chromatography in 97:3 CH2Cl2: MeOH. The collected fraction was rotary evaporated and reduced to yellow oil. The yield was not determined and the mass was confirmed by ESI-MS, expected [M+H]+, 474.4; observed [M+H]+, 474.2, [M+Na]+, 496.2. The dinucleotide pdCpA (5 mg) was dissolved in 120 µl of dry DMF and transferred to the flask containing 21 mg N-NVOCN-Me-Tyr cyanomethyl ester. A catalytic amount of tetrabutylammonium acetate was added to catalyze the reaction. The reaction was stirred at room temperature for 7 hr. The crude sample was purified by C18 semiprep HPLC. Solvent A: 25 mM NH4OAc:CH3CN  34  (pH 4.5), 95:5. Solvent B: 25 mM NH4OAc:CH3CN (pH 4.5), 10:90. Gradient: 0%100%B in 70 min. Flow: 10 ml/min. Two fractions with retention time of 25 min and 26 min respectively, were collected and lyophilized. The solids were redissolved in 5 ml of 10 mM acetic acid/CH3CN and lyophilized again. This procedure was repeated one more time before the pale yellow solids were subjected to ESI-MS to confirm the mass. The calculated mass of N-NVOC-N-Me-Tyr pdCpA was [M+H]+, 1053.9. No signals corresponding to [M+H]+ or [M+Na]+ were observed in either fractions.  Figure 7 The synthetic scheme of converting N-Me-Tyr to N-NVOC-N-Me-Tyr pdCpA.  35  2.36 Synthesis of N-NVOC-Oic cyanomethyl ester The overall synthetic scheme is shown in Figure 8. In a 50 ml round bottom flask, 90.4 mg of Oic was dissolved in 1 ml of 10% NaCO3 and 0.5 ml of dioxane and cooled to 4oC in an ice bath. NVOC-Cl,(148.8 mg) was dissolved in 1.6 ml dioxane:THF(1:1) and slowly added to Oic. The reaction was protected by aluminum foil and stirred at 0oC for 1 hr followed by room temperature over night. The reaction was quenched by addition of 15 ml of distilled water. The reaction was extracted twice with 5 ml diethyl ether, the aqueous layer was kept and acidified to pH 2 using concentrated HCl, and extracted twice with 5 ml ethyl acetate. The organic layer was dried with MgSO4, filtered and rotary evaporated to give to a yellow colored film. To the N-NVOC-Oic, 1.5 ml dry DMF, 1.5 ml chloroacetonitrile and 1 ml dry TEA were added while the reaction was stirred under nitrogen at room temperature. The reaction was allowed to proceed for 24 hr while protected by aluminum foil. The crude reaction was diluted with ethyl acetate (15 ml) and extracted with 0.5 N HCl (15 ml). The organic layer was then extracted with saturated aqueous sodium bicarbonate (15 ml) and saturated sodium chloride (15 ml), dried over MgSO4, and concentrated by rotary evaporation. The product N-NVOC-Oic cyanomethyl ester was purified by silica gel chromatography in 99:1 CH2Cl2: MeOH. The collected fraction was rotary evaporated and reduced to yellow oil. The mass was confirmed by ESI-MS, expected [M+H]+, 448.6; observed [M+H]+, 448.1, [M+Na]+, 470.0.  36  Figure 8 The synthetic scheme of converting Oic to N-NVOC-Oic cyanomethyl ester.  2.37 Synthesis of N-NVOC-N-Me-Nle cyanomethyl ester The overall synthetic scheme is shown in Figure 9. In a 50 ml round bottom flask, 78.4 mg of N-Me-Nle was dissolved in 1 ml of 10% NaCO3 and 0.5 ml of dioxane and cooled to 4oC in an ice bath. NVOC-Cl (148.8 mg, Sigma) was dissolved in 1.6 ml dioxane:THF (1:1) and slowly added to N-Me-Nle. The reaction was protected by aluminum foil and stirred at 0oC for 1 hr followed by room temperature over night. The reaction was quenched by addition of 15 ml of distilled water. The reaction was extracted twice with 5 ml of diethyl ether, the aqueous layer was kept and acidified to pH 2 using concentrated HCl, and extracted twice with 5 ml ethyl acetate. The organic layer was dried with MgSO4, filtered and rotary evaporated to give to a yellow colored film. To the N-NVOCN-Me-Nle, 1.5 ml dry DMF, 1.5 ml chloroacetonitrile and 1 ml dry TEA were added while the reaction was stirred under nitrogen at room temperature. The reaction was allowed to proceed for 24 hr while protected by aluminum foil. The crude reaction was 37  diluted with ethyl acetate (15 ml) and extracted with 0.5 N HCl (15 ml). The organic layer was then extracted with saturated aqueous sodium bicarbonate (15 ml) and saturated sodium chloride (15 ml), dried over MgSO4, and concentrated by rotary evaporation. The product  N-NVOC-N-Me-Nle  cyanomethyl  ester  was  purified  by  silica  gel  chromatography in 99:1 CH2Cl2: MeOH. The collected fraction was rotary evaporated and reduced to yellow oil. The mass was confirmed by ESI-MS, expected [M+H]+, 423.5; observed [M+H]+, 424.0, [M+Na]+, 446.0.  Figure 9 The synthetic scheme of converting N-Me-Nle to N-NVOC-N-Me-Nle cyanomethyl ester.  38  2.4 Coupling of tRNAs with amino acids Two couplings methods are summarized in this section. The first one is chemoenzymatic coupling between chemically modified amino acids and tRNA(-ca). The second one is enzymatic coupling between natural amino acids and tRNA.  Per coupling reaction using chemically modified amino acids, 31 µl of ddH2O, 30 µl of 0.67 µM refolded tRNA(-ca), 8 µl of 10X T4 RNA ligase buffer, 8 µl of 3 mM NNVOC-amino acid pdCpA and 5 µl of homemade T4 RNA ligase I (10 µM) were added in sequential order in an Eppendorf tube kept on ice. The mixture was mixed by gentle pippetting before placed in a 37ºC incubator for 1 hr. The reaction was quenched by adding deionized H2O to a final volume of 300 µl and 1/10th volume of 3 M sodium acetate  (pH  5.2).  The  protein  was  removed  from  the  reaction  by  phenol:chloroform:isoamyl alcohol 25:24:1 pH 5.0 (adjusted by 1 M HCl) extraction, and the aqueous layer was subjected to ethanol precipitation (refer to section 2.21 for details). The pellet was air dried while kept on ice, and resuspended in less than 10 µl of deionized H2O. 0.5 µl of the sample was then added in 99.5 µl of ddH2O to determine its concentration based on the equation, C (µg / ml) = A260 / 0.025. The aminoacyl-tRNAs were kept at -80oC until use.  Coupling of full-length tRNAggc with amino acid L-alanine was done using alanyl-tRNA synthetase [51]. tRNAggc (full-length) was first refolded by adjusting its concentration to 0.5 µg/µl in 6.7 mM HEPES pH 7.5 and heating at 94ºC for 3 min and slowly cooled to ambient temperature. Per coupling reaction contained 50 µg of tRNAggc (full-length), 12  39  µCi [14C]Ala, 2.4 µM alanyl tRNA synthetase in a buffer containing 40 mM Tris, pH 7.5, 15 mM MgCl2, and 10 mM ATP. The reaction was allowed to proceed for 1.5 hr at 37oC. The reaction was quenched by adding 1/10th volume of 3 M sodium acetate (pH 5.2). The enzyme  and  Ala  were  removed  from  the  reaction  by  extraction  with  phenol:chloroform:isoamyl alcohol 25:24:1 pH 5.0 (adjusted using 1 M HCl). The aqueous layer was subjected to ethanol precipitation (refer to section 2.21 for details). The RNA pellet was air dried while kept on ice. Ala-tRNAggc was then resuspended in deionized H2O to achieve a final concentration of 2 µg/µl.  2.5 Detection of aminoacyl-tRNAs by acidic denaturing urea PAGE Chemical acylation of tRNAs was confirmed using a modified protocol [52]. A 15% acidic denaturing urea gel (8.6 x 6.8 cm) was set up as follows; 6 ml of SequaGel concentrate, 3 ml of SequaGel diluent and 1 ml of 1 M sodium acetate pH 5.0 were added in sequential order into a clean beaker, swirled to mix, and 80 µl 10% of APS and 4 µl of TEMED were added. The gel mixture was then poured into gel casts (Mini-Protean Tetra Electrophoresis System), before mini-PROTEAN combs (10-well, 0.75 mm) were placed. The gels were allowed to polymerize for around 20 min. The gels were pre-run in 100 mM sodium acetate for 30 min at 70 V. When ready, the wells of the gel were flushed using a syringe to remove urea. Around 0.2 µg of aminoacyl-tRNA was mixed with 10 µl of 1X acidic denaturing urea PAGE loading buffer (100 mM sodium acetate, 8 M urea, 0.05% Bromophenol blue). The samples were separated by gel electrophoresis at 100-150 V until the dye front reached the bottom of the gel. The gels were then removed from the  40  glass plates and stained with 0.03% w/v ethidium bromide for 1 min, rinsed with distilled H2O for another 1 min. The stained gels were viewed using Alphaimager®HP.  2.6 EF-Tu binding Assay This assay was used to detect EF-Tu’s affinity to unnatural amino acid-tRNA. It was performed using the protocol developed by Doi et al. [53]. A 5 µl aliquot of 10 µM EFTu was first pre-incubated at 37°C for 15 minutes in a buffer containing 1 mM GTP, 70 mM HEPES-KOH (pH 7.6), 52 mM NH4OAc, 8 mM Mg(OAc)2, 30 mM KCl, 0.8 mM DTT, 6% glycerol, 10 mM phosphoenolpyruvate, and 0.08 U/µl pyruvate kinase. 3 µl of 8.3 µM NVOC-unnatural amino acid-tRNA (in 1 mM NaOAc) was deprotected using a xenon lamp (Newport Corporation) with a 315 nm cut-off filter at 500 W for 7 min. The deprotected unnatural amino acid-tRNA was then added to the preincubated EF-Tu solution, followed by 2 µl of ternary complex buffer (150 mM HEPES-KOH, pH 7.6, 195 mM NH4OAc, 30 mM Mg(OAc)2). The mixture was further incubated at 37°C for 10 minutes. To set up a 8% native polyacrylamide gel, a 4.5 ml mixture containing 1.3 ml of Protogel (30%) in 50 mM Tris-HCl pH 6.8, 65 mM NH4OAc, 10 mM Mg(OAc)2 was first made; polymerization was initiated with the addition of 50 µl APS and 3 µl TEMED. The gels were cast using Mini-Protean Tetra Electrophoresis System; and the solidified gels were kept at 4oC until use. The electrophoresis was performed at 4°C in 50 mM TrisHCl pH 6.8, 65 mM NH4OAc and 10 mM Mg(OAc)2, and was stopped when the dye fronts were 3-4 cm from the wells. Gels were visualized using Bio-Rad Silver Stain Plus Kit.  41  2.7 Streptavidin binding assay Streptavidin binding assay was used to detect Bio-containing molecules and was based on a modified protocol [52]. Streptavidin sample loading buffer contained 0.2 mg/ml streptavidin from Streptomyces Avidinii (≥13 units/mg), 37 mM piperazine, 37 mM EDTA, 6 M urea and 0.05% Bromophenol blue. 10 µl of streptavidin sample loading buffer was added to each sample and the samples were then subjected to either normal or acidic denaturing-PAGE analysis. To stain gels with ethidium bromide, the gels were agitated in 0.03% w/v ethidium bromide for 1 min followed by rinsing with distilled H2O for another 1 min. The ethidium-bromide stained gels were viewed using alphaimager®HP. To stain gels with Coomassie Blue, the gels was fixed in fixing solution (10% v/v acetic acid and 50% v/v methanol) for 10 min, stained with staining solution (0.1% w/v Coomassie Brilliant Blue G-250, 10% v/v acetic acid and 45% v/v methanol) by microwaving the submerged gel for 10 s, and destained with destaining solution (10% v/v acetic acid and 10% v/v methanol) by gentle agitation in the solution until the gels were fully destained. The stained gels were viewed using an alphaimager®HP. Preparing gels for phosphorimaging is described in section 2.93.  2.8 E. coli in vitro reconstituted translation reaction The translation reactions were performed in a modified polymix buffer pH 7.6 (20 mM Mg(OAc)2, 0.5 mM CaCl2, 8 mM putrescine, 1 mM spermidine, 5 mM K3PO4, 95 mM KCl, 5 mM NH4Cl, 1 mM DTT, 3 mM ATP, 2.5 mM GTP, 6 mM phosphoenolpyruvate, and 0.01 mg/ml pyruvate kinase) supplemented with 10 mM creatine phosphate and 30 µM 10-formyl-5,6,7,8-tetrahydrofolic acid. The reconstituted translation reactions 42  contained final concentrations of approximately 0.5 µM of each MTF, Met-RS, IF1, IF2, IF3, EF-Ts, EF-G, 6.6 µM EF-Tu, 0.8 µM purified ribosomes, 0.8 µM [35S]Met, 0.4 µM tRNAfMet, and precharged aminoacyl-tRNAs. The concentration of Bio-tRNA was always 2.4 µM. For other pre-acylated aminoacyl-tRNAs, unless specified otherwise, the amount of each was calculated using the equation: Amount (pmol) = Total amount of templates (pmol) x Number of codons recognized by this type of tRNA x 4. For aminoacyl-tRNA made by the chemo-enzymatic method, the N-NVOC group on the amino acid needs to be first photo-cleaved before the aminoacyl-tRNAs can be introduced into the translation mixture; and this is done by placing the aminoacyl-tRNA at the bottom of a PCR tube in front of a xenon lamp with a 315 nm cut-off filter at 500 W for 7 minutes. Translation was initiated by addition of 0.8 µM of the template, and the translation mixture was incubated at 37°C for 1 hr before stop salt was added to adjust the final salt concentration of KCl and Mg(OAc)2 to 550 and 50 mM, respectively [23]. The translation mixtures were incubated further at room temperature for 1 hr followed by -20oC overnight.  2.9 Purification and analysis of translation products 2.91 Purification of peptide-oligonucleotide fusions If translation was performed in a mRNA display format, i.e. generated peptides would be in fusion with their coding templates, unless specified otherwise, purification of the translation products was done as follows; To each crude translation mixture, 200 µl of 2.5% Oligo (dT)-cellulose Type 7 and 1 ml of incubation buffer (0.5 M KCl, 0.01 M Tris, pH 7.5) were added. The mixture was mixed on a nutating mixer at room temperature for  43  1 hr. The resins were spun down at 2000 x g for 30 s. The supernatant was disposed. The resins were then washed by 500 µl of incubation buffer, and the supernatant was disposed after the resins were spun down. The washing step was repeated 4 more times. The translation products were eluted by adding 400 µl of deionized H2O to the resin and spinning the resin down at 10621 x g for 1 min. The supernatant was pipetted into a clean Eppendorf tube, and this elution procedure was repeated one more time. To each 400 µl elution, 1/10th volume of sodium acetate (pH 5.2), 25 µg of linear acrylamide and 1 ml of 100% ethanol were added. The translation products were allowed to precipitate at -80oC for 40 min, and the precipitates were collected by centrifuging the mixture at 20,817 x g for 10 min. The pellets from the same translation mixture were resuspended in a total volume of 8 µl deionized H2O. The mRNA portion of the translation products were removed by incubating with 4 µg of RNase A (≥70 Kunitz units/mg protein) at 37oC for 30 min; 0.5 µl of the sample was then used for scintillation counting; the rest sample would be subjected either to further treatment or analyzed using denaturing urea PAGE.  2.92 Purification of peptides If translation was performed using templates without puromycin attachment, i.e. nascent peptides not fused with their coding templates, crude translation mixture was purified as follows; deionized H2O was added to the reaction mixture to give a final volume of 100 µl followed by 1/10th volume of 3 M sodium acetate and 3X volume of 100% ethanol. The mixture was stored at -80oC for 20 min before the oligonucleotide components of the reaction mixture were collected by centrifugation at 20,817 x g for 10 min. The pellet was washed by adding 500 µl of 70% ethanol and centrifuging at 20,817 x g for 10 min.  44  The ethanol wash was repeated once. The pellet was resuspended in 40 µl of deionized H2O, and to which 30 µg of RNase A was added and incubated at 37oC for 1 hr. The mixture was then passed through a nanosep centrifugation device (MWCO 10K) by centrifugation at 14,000 x g for 3 min. The flow through was collected and 1/6th fraction was then used for denaturing urea PAGE analysis.  2.93 Denaturing urea PAGE analysis of translation products To set up two 12% denaturing urea gels, a mixture containing 4.8 ml SequaGel concentrate, 4.2 ml SequaGel diluent and 1 ml SequaGel diluent was made. The gels were cast using Mini-Protean Tetra Electrophoresis System. The gels were pre-run at 70 V for 30 min. To each sample, either 10 µl of 6 M urea buffer (6 M urea and 0.05% Bromophenol blue) or streptavidin sample loading buffer was added. Gel wells were washed with a 5 ml syringe before samples were loaded. Gel electrophoresis was performed at 150 V and stopped when the dye reached the bottom of the gels.  After gel electrophoresis, gels were removed from the gel running apparatus, and rinsed with water containing 5% glycerol (50 ml/gel) for 5 min. The gels were placed on 2 layers of filter paper and covered with saran wrap before transferred to a slab gel dryer (SGD5040) for drying at 80oC for 20 min/gel. The dried gels were exposed to a storage phosphor screen at room temperature for at least 12 hr. The phosphor screen was scanned by a Typhoon imager (Typhoon 8600 Variable Mode Imager). The gel image was quantified by using ImageQuant 5.2 software. Percentage of gel shift (%) was calculated using the equation: Percentage of gel shift (%) = Signal of gel shift/Total signal x 100%  45  2.10 Preparation of MALDI-TOF samples 2.101 Preparation from rabbit reticulocyte lysate A 100 µl non-radiolabeled reaction and a 25 µl radiolabeled reaction were prepared according to the manufacture’s instructions (Red Nova® Lysate). The non-radiolabeled reaction contained 0.1 M potassium acetate, 0.5 mM magnesium acetate, 40 µM Met, 2 µM CAC1 templates, and 40% lysate in 1X translation mix. The radiolabeled reaction contained 0.8 µM of [35S]Met in place of non-radiolabeled Met and 0.8 µM of CAC1 template. The reaction was incubated at 30oC for 1 hr, and the stop salt was added to adjust the final salt concentration of KCl and Mg(OAc)2 to 550 and 50 mM, respectively. The translation mixtures were incubated at room temperature for 1 hr followed by -20oC overnight.  To each crude reaction mixture, 0.25 µl/(pmol of templates) of RNase cocktail was added and incubated at room temperature for 3 hr. The crude translation products were purified by oligo (dT)-cellulose using the method described in section 2.91. The purified translation products were resuspended in a total volume of 7 µl of ddH2O, 1 µl of 10X DNase I buffer and 2 µl of DNase I (2,000 units/ml) and incubated at 37oC for 2 hours. Radioactive samples were then analyzed as described in section 2.93. The nonradioactive sample was subjected to MALDI-TOF analysis.  46  2.102 Preparation from the E. coli in vitro reconstituted translation system The translation reactions were performed in a modified polymix buffer pH 7.6 (20 mM Mg(OAc)2, 0.5 mM CaCl2, 8 mM putrescine, 1 mM spermidine, 5 mM K3PO4, 95 mM KCl, 5 mM NH4Cl, 1 mM DTT, 3 mM ATP, 2.5 mM GTP, 6 mM phosphoenolpyruvate, and 0.01 mg/ml pyruvate kinase) supplemented with 10 mM creatine phosphate and 30 µM 10-formyl-5,6,7,8-tetrahydrofolic acid. 250 µl of translation reaction was set up for each tested amino acid; 225 µl was set up using non-radioactive Met, and the other 25 µl was set up using radioactive [35S]Met as a monitoring reaction. The reconstituted translation reactions contained final concentrations of approximately 0.5 µM of each MTF, Met-RS, IF1, IF2, IF3, EF-Ts, EF-G, 6.6 µM EF-Tu, 0.8 µM heat activated purified ribosomes (42oC for 30 min), 0.4 µM tRNAfMet, and 17 µM deprotected amino acid-tRNAs. Deprotection of aminoacyl-tRNAs was done by placing the aminoacyltRNA at the side of a PCR tube in front of a xenon lamp (Newport Corporation) with a 315 nm cut-off filter at 350 W for 2 minutes to remove the photo-cleavable protecting group. The non radiolabeled reactions contained 40 µM Met and the radiolabeled reactions contained 0.8 µM [35S]Met. Translation was initiated by addition of 0.8 µM of CAC5 template, and the translation mixture was incubated at 37°C for 2 hr before adding stop salt to adjust the final salt concentration of KCl and Mg(OAc)2 to 550 and 50 mM, respectively. The translation mixtures were further incubated at room temperature for one hour followed by -20oC overnight.  The radiolabeled reactions were purified and analyzed as described in section 2.9. To each non-radiolabeled translation mixture (225 µl), 45 µl of RNase cocktail was added  47  and incubated at room temperature for 2 hr. The translation products were then purified using oligo (dT)-cellulose as described in section 2.91. The final pellets from each translation reaction were resuspended in a total volume of 7 µl of ddH2O, 1 µl of 10X DNase I buffer and 2 µl of DNase I (2,000 Units / ml). The mixture was then incubated at 37oC for 2 hr and stored at -80oC until ready to be sent off for MALDI-TOF analysis.  2.103 MALDI-TOF analysis Upon treatment with RNase A and DNase I, the nucleotide portion was removed from peptide-oligonucleotides to give to peptide-puromycin conjugates, which were then subjected to MALDI-TOF-analysis (UBC laboratory of molecular biophysics). All MALDI data were obtained using Applied Biosystems Voyager System 4311 operating in linear positive and negative ion modes with delayed extraction. The accelerating voltage was set at 20,000 V, grid voltage at 74%, guide wire at 0.002%, and extraction delay time at 350 ns. The calibration matrix used was α-cyano-4-hydroxycinnamic acid.  2.11 Proteinase K analysis of translation products generated by E. coli in vitro reconstituted translation system To prepare for proteinase K analysis, 125 µl of E. coli in vitro reconstituted translation was set up for each amino acid tested using CAC5 as the template (section 2.8). After quenching each translation reaction, 25 µl RNase cocktail was added to the crude translation reaction mixture and incubated at room temperature for 2 hr. The translation products were then purified using oligo (dT)-cellulose using procedure described in  48  section 2.91, except the precipitation was carried out without linear acrylamide. The final pellets from each translation reaction were resuspended in a total volume of 30 µl of 20 mM Tris-HCl, pH 7.5. The purified translation products from each reaction were then aliquoted into 3 equal portions of 7 µl; 5 ug of proteinase K from Engyodontium album (≥30 units / mg) was then added to two of the aliquots, one of which was incubated at 37oC for 3 hr and the other for 20 hr. After proteinase K treatment, the enzyme was inactivated by heating the samples at 95oC for 10 minutes. To each sample, 10 µl of 6 M urea buffer was added before the samples were analyzed using 12% denaturing urea PAGE as described in section 2.93.  2.12 mRNA library design and generation 2.121 Making V3A, V3Ap, V4, V4p, V5, and V5p libraries The following DNA oligos were ordered from TriLink. V3 template: 5’- TTT GGG CCC TTT CCC GGG SNN SNN SNN CAT GTA TAT CTA CCT CCT TAC T -3’. V4 template: 5’- TTT GGG CCC TTT CCC GGG SNN SNN SNN SNN CAT GTA TAT CTA CCT CCT TAC T -3’. V5 template: 5’- TTT GGG CCC TTT CCC GGG SNN SNN SNN SNN SNN CAT GTA TAT CTA CCT CCT TAC T -3’. Splint: 5’- TTT TTT TTT TTT TTT TTT GGG CCC TTT CCC GGG -3’. Primer 2 (reverse primer): 5’2’OMe(UU) TGG GCC CTT TCC CGG G -3’. pF30P: 5’- (phosphate) AAA AAA AAA AAA AAA AAA AAA XXX ACC (puromycin) -3’. The oligo primer 1 (forward primer): 5’- AGT GAA TTC TAA TAC GAC TCA CTA TAG GGT TAA CTT TAG TAA GGA GGT AGA TAT ACA TG -3' was ordered from IDT. The letter N = G, C, A or T; S = G  49  or C. Oligos V3A, V4, V5, primer 2 and splint were purified by 15% denaturing PAGE, and the splint was further desalted using a PD-10 column prior to use. V3A DNA library was made as follows; 10 PCR reactions (50 µl/reaction) were set up according Taq DNA polymerase protocol, each reaction contained 1 µM primer 1, 1 µM primer 2, and 0.2 µM V3A template. The PCR cycle was set as the following; initial denaturation at 94oC for 1 min, denaturation at 94oC for 30 s, primer annealing at 64oC for 20 s, extension at 72oC for 30 s, 14 cycles, final extension at 72oC for 2 min followed by 16oC at the end of the PCR. The PCR mixture was loaded onto a 4% agarose gel and 2 µg of 10 bp DNA ladder was used as a size control. The expected 86 bp PCR product was cut out and extracted using QIAquick Gel Extraction Kit. The purified PCR product was quantified using the equation Concentration (µg/ml) = A260/0.020. V4 (89 bp) and V5 (92 bp) DNA libraries were generated using the same procedure. Each DNA library was converted to mRNA library by T7 in vitro transcription (section 2.22). The mRNA libraries were then purified by 17% denaturing PAGE. The bands corresponding to 60 nucleotides (V3A RNA library), 63 nucleotides (V4 RNA library) and 66 nucleotides (V5 RNA library) were extracted from the gel using the procedure described in section 2.23. The purified RNAs were desalted using PD-10 columns (Amersham). The RNA libraries were quantified by 407.6 OD/µmol for V3A, 431.3 OD/µmol for V4, and 457.9 OD/µmol for V5. The RNA libraries were ligated to 5’-phosphorylated puromycin oligonucleotides by setting up a mixture of 4,000 pmol RNA library and 6,000 pmol splint and heated in H2O to 94oC for 1 min and immediately cooled on ice for 15 min, to which, 4,000 pmol pF30P, T4 DNA ligase buffer, and 3 µl of 100,000 units/ml T4 DNA ligase were then added. The ligation reaction was performed at room temperature for 1 hr. The ligation mixture was extracted  50  with an equal volume of phenol:chloroform:isoamyl alcohol 25:24:1 pH 8.0 and ethanolprecipitated using the procedure described in section 2.21. The full-length mRNApuromycin templates were purified on 17% denaturing PAGE. The purified templates were quantified by, 872.7 OD/µmol for V3Ap, 901.4 OD/µmol for V4p, and 927.6 OD/µmol for V5p.  2.122 Making V3B and V3Bp libraries The following oligos were ordered from IDT. V3B (RNA) 5’- ggaggacgaa aug (nns)3 aaa ggg ccc aaa ccc ggg -3’; New splint (DNA) 5’- TTT TTT TTT TTT TTT TTT CCC GGG TTT GGG -3'. V3B RNA library was first purified by 12% denaturing urea PAGE and desalted by a PD-10 desalting column. The purified V3B RNA library was quantified by 2.5 nmols / OD260. The purified V3B RNA library was ligated to 5’-phosphorylated puromycin oligonucleotides by setting up a mixture of 8,488 pmol V3B library and 12,750 pmol splint, the mixture was heated in H2O to 94oC for 1 min and immediately cooled on ice for 15 min, to which 17,000 pmol pF30P, T4 DNA ligase buffer and 5 µl of 100,000 units/ml T4 DNA ligase were added. The ligation reaction proceeded at room temperature for 1 hr. The reaction was then extracted once with an equal volume of phenol:chloroform:isoamyl alcohol 25:24:1 pH 8.0 and ethanol-precipitated using the procedure described in section 2.21. The full-length mRNA-puromycin templates were purified on 10% denaturing urea PAGE. The V3Bpuromycin was quantified by 701.70 OD260 / µmol.  51  2.123 Translation of mRNA libraries by the E. coli in vitro reconstituted translation system First, a tRNA(-ca) master mix was first made prior to translation reactions (please see Supplementary Table 1 in Appendix for the full list of tRNAs). For translations performed using templates V3Ap, V4p, and V5p, the tRNA(-ca) master mix was made by taking 10 µl aliquot from each of refolded tRNAs(-ca) # 1-7, 9-12, 14-19, 21-33; and 68 µl aliquot from each of the refolded tRNAs(-ca) # 8, 13. For translations performed using templates V3B and V3Bp, the tRNA(-ca) master mix was made by taking 10 µl aliquot from each of the refolded tRNAs(-ca) #1-19, 21-33 (see Supplementary Table 1 in Appendix for the full list of tRNA). Second, the tRNAs(-ca) master mix was used as the tRNAsnn source for ligation to N-NVOC-Sar pdCpA, N-NVOC-N-Me-Val pdCpA or NNVOC-Bio pdCpA. tRNAuuu(-ca) (tRNA#20(-ca)) was ligated to N-NVOC-Bio pdCpA. The ligation procedure was described in section 2.4. Third, when setting up the E. coli in vitro reconstituted translation reactions, the procedure was the same as described in section 2.8 except the following: the amount of Sar-tRNAsnn or N-Me-Val-tRNAsnn was calculated using the formula: Amount of aminoacyl-tRNA (pmol) = Template Amount (pmol) x Number of Random Codons of the Template x N, where N = intended number of library coverage. The amount of Bio-tRNAsnn and Bio-tRNAuuu were used at 2.4 µM. The final concentration of template was 0.4 µM.  52  3 Results 3.1 Validation of streptavidin assay When the streptavidin binding assay was first adapted from Murakami et al. [52], an acidic 15% denaturing urea gel was set up to assess how streptavidin binding would affect the migration of Bio-containing molecules (Figure 10). Two groups of samples were prepared. In the first group, 0.3 µg tRNAggc(-ca) and 0.2 µg N-NVOC-Bio-tRNAggc were prepared in 1X acidic denaturing PAGE loading buffer, and another 0.2 µg NNVOC-Bio-tRNAggc was prepared in streptavidin sample loading buffer. In the second group, 0.3 µg of tRNAgug(-ca) and 0.2 µg N-NVOC-Bio-tRNAgug were prepared in 1X acidic denaturing PAGE loading buffer, and another 0.2 µg N-NVOC-Bio-tRNAgug were prepared in streptavidin sample loading buffer. For this experiment, tRNAggc(-ca) and tRNAgug(-ca) were made using the procedure described in section 2.2. Bio was converted N-NVOC-Bio pdCpA as described in section 2.33. N-NVOC-Bio pdCpA was then enzymatically ligated to each type of tRNA(-ca) as described in the section 2.4.  53  Figure 10 Validation of streptavidin assay by 15% denaturing urea PAGE. The top image shows the gel detected by ethidium bromide staining, and the bottom image shows the same gel detected by coomassie blue staining. Lane 1 is tRNAggc(-ca) analyzed in acidic denaturing urea PAGE loading buffer, lane 2 is N-NVOC-Bio-tRNAggc analyzed in acidic denaturing urea PAGE loading buffer, lane 3 is N-NVOC-Bio-tRNAggc analyzed in streptavidin sample loading buffer, lane 4 is tRNAgug(-ca) analyzed in acidic denaturing urea PAGE loading buffer, lane 5 is N-NVOC-Bio-tRNAgug analyzed in acidic denaturing urea PAGE loading buffer, and lane 6 is N-NVOC-Bio-tRNAgug analyzed in streptavidin sample loading buffer. The bands corresponding to streptavidin-dependent gel shifts are indicated by the black arrows.  The ethidium bromide stained gel shows 1 major band in the lanes containing tRNA(-ca) (Figure 10, lanes 1 and 4, upper gel), and two bands in the lanes containing N-NVOCBio-tRNAs (Figure 10, lanes 2 and 5, upper gel), where the band with slower mobility corresponds to N-NVOC-Bio-tRNA, and the band with faster mobility corresponds to non-ligated tRNA(-ca). The band for N-NVOC-Bio-tRNA observed in lanes 2 and 5 are  54  replaced with two much slower mobility bands when analyzed in streptavidin sample loading buffer (Figure 10, lanes 3 and 6, upper gel). When the same gel was stained with Coomassie Blue, the location of streptavidin protein overlaps with the location of gel shifted N-NVOC-Bio-tRNA (Figure 10, lanes 3 and 6, lower gel), and confirmed the gel shift was due to the binding of the Bio-containing molecule to streptavidin. The appearance of more than one band observed for the streptavidin-dependent gel shift was likely due to the presence of different oligomeric forms of streptavidin [54].  A 12% denaturing urea gel was set up to further test the specificity of streptavidin binding assay (Figure 11). Two samples were prepared in streptavidin sample loading buffer: 0.5 µg tRNAggc(-ca) and 0.5 µg N-NVOC-Bio-tRNAggc. Only the lane containing N-NVOC-Bio-tRNAggc shows a gel shift (Figure 11, lane 2). Based on this experiment, the streptavidin based assay is shown to be specific to Bio-containing molecules.  55  Figure 11 Validation of streptavidin assay by 12% denaturing urea PAGE. The gel is detected by ethidium bromide staining. Lane 1 shows tRNAggc(-ca) analyzed in streptavidin sample loading buffer, and lane 2 shows N-NVOC-Bio-tRNAggc analyzed in streptavidin sample loading buffer. The bottom band corresponds to tRNAggc(-ca), and the streptavidin-dependent gel shift is indicated by the black arrow.  3.2 Assessment of the E. coli in vitro reconstituted translation system’s efficiency of making full-length peptides in mRNA display format As an initial experiment to test the E. coli in vitro reconstituted translation system’s ability to make small peptide-oligonucleotide conjugates, a template encoding two amino acids was used. This template, named GCC1, was custom synthesized. The sequence of the GCC1 is listed Table 2. It is a RNA/DNA hybrid molecule and consisted of 3 regions: the 5’ non-coding region, the coding region and the 3’ non-coding region. The 5’ constant non-coding region has the RNA sequence ggagg needed to ensure proper translation initiation. The coding region has two codons with the RNA sequence aug gcc, where aug 56  codes for fMet-tRNAfmet, and gcc normally codes for Ala, but was re-assigned to code for Bio-tRNAggc. The 3’ non-coding region is a DNA linker with a covalently attached puromycin molecule; its sequence is A21(spacer-9)3ACC(puromycin), where spacer-9 is triethylene glycol phosphate and is believed to provide the flexibility the linker needed to form peptide-mRNA fusion.  The codon gcc was assigned to an unnatural amino acid Bio for two reasons. First, Bio is a modified lysine molecule with a biotin group attached to its side chain, and can be recognized and bound by streptavidin with high and specific affinity. Second, this unnatural amino acid has demonstrated successful incorporation into peptides [32]. By assigning Bio to be the second amino acid incorporated in a dipeptide, the streptavidin assay validated earlier could be used to detect the incorporation of Bio, and deduce the efficiency (%) of making full-length peptides based on the proportion of the translated peptides bound to streptavidin.  The procedure of setting up the E. coli in vitro reconstituted translation is described in section 2.8. Seven translation reactions were set up with Bio-tRNAggc. N-NVOC-Bio was pre-charged to tRNAggc as described in the section 2.4 and deprotected to Bio-tRNAggc just before setting up translation reactions. At the end of translation reactions, the addition of the stop salt solution and the subsequent freezing of crude translation mixture at -20oC overnight promoted ribosomal complex disassembly and the release of mRNApeptide fusions. The mRNA-peptide fusions were a heterogeneous mixture that consisted of incomplete translation product [35S]fMet-GCC1 and full-length translation product  57  [35S]fMet-Bio-GCC1. Taking advantage of the poly (dA) sequence present in the GCC1 and peptide-GCC1 fusions, the translation products were purified from the crude translation mixture by oligo (dT) cellulose. The mRNA portion attached to the translation products were then removed by the RNase A treatment to give to [35S]fMet-puromycinCCA(spacer-9)3(A)21, ([35S]fMet-P) and [35S]fMet-Bio-puromycin-CCA(spacer-9)3(A)21. These samples were then analyzed using denaturing urea PAGE in the presence and absence of streptavidin sample loading buffer.  Except for the negative control (translation performed without Bio-tRNAggc), the purified translation products show two bands in both the presence and absence of streptavidin sample loading buffer (Figure 12, lanes 2-7). The intensity of the upper band increases and that of lower band decreases as more Bio-tRNAggc was added to the translation reaction. The upper band migrates much slower when samples were analyzed in the presence of streptavidin sample loading buffer (Figure 12, lanes 2-7, upper gel). Based on the above evidence, the upper band corresponds to the Bio-containing full-length translation product: [35S]fMet-Bio-puromycin-CCA(spacer-9)3(A)21. The single band in the negative control lane corresponds to the incomplete translation product: [35S]fMet-P (Figure 12, lane 1). The proportion of the translation products that was [35S]fMet-Biopuromycin-CCA(spacer-9)3(A)21 was calculated using the equation: Percentage of gel shift (%) = Intensity of upper band / Total intensity x 100%. The result shows that translation performed with the highest amount of Bio-tRNAggc gave the highest percentage of gel shift. Based on the result, the following conclusion can be drawn: the formation of full-length translation product [35S]fMet-Bio-puromycin-CCA(spacer-  58  9)3(A)21 is dependent on the concentration of Bio-tRNAggc, and this dependence is an indication that the E. coli in vitro reconstituted translation machinery is functional and capable of making at least a dipeptide-oligonucleotide conjugates, in which one of the amino acid can be unnatural.  59  Figure 12 Aminoacyl-tRNA dependent formation of dipeptide-oligonucleotides. 35Slabeled peptide-oligonucleotide conjugates were made from the E. coli in vitro reconstituted translations of GCC1 template with increasing concentrations of BiotRNAggc and analyzed in 12% denaturing urea PAGE. Seven translation reactions were performed using GCC1 template; each reaction contained a different concentration of Bio-tRNAggc, and is reflected by Bio-tRNAggc/template ratio. The upper gel shows the translation products analyzed in the presence of streptavidin sample loading buffer, and the bottom gel shows the translation products analyzed in 6 M urea buffer. The streptavidin-dependent gel shift is indicated by the black arrow (upper gel), and the percentage of gel shift is calculated using the equation: gel shift (%) = intensity of gel shift / total sample intensity x 100%; where the intensity of each band is determined using ImageQuant 5.2.  To test whether the E. coli in vitro reconstituted translation system is capable of making longer peptide-oligonucleotide conjugates, a different template was used. This template is named CAC5 and encodes a hexapeptide. Its sequence is shown in Table 2. It has the 60  same non-coding regions as GCC1. The coding region consists of six codons and has the sequence of aug gcc gcc gcc gcc cac, where aug codes for fMet-tRNAfMet, gcc for AlatRNAggc, and cac for Bio-tRNAgug. The reconstituted translation was set up as described in section 2.8. As the source of aminoacyl-tRNA, 7.5 µg Ala-tRNAggc was enzymatic acylated (section 2.4) and 3 µg N-NVOC-Bio-tRNAgug was chemoenzymatically acylated and deprotected before introduced to the translated mixture. Three translation reactions were set up: one containing only Ala-tRNAggc, one containing only Bio-tRNAgug, and one containing both Ala-tRNAggc and Bio-tRNAgug. The translation products were purified by oligo (dT)-cellulose. Purified translation products from each translation were divided into two equal portions: one was analyzed in 6 M urea buffer and the other in streptavidin sample loading buffer (Figure 13).  61  Figure 13 Aminoacyl-tRNA dependent formation of hexapeptide-oligonucleotide conjugates. 35S-labeled peptide-oligonucleotide conjugates were made from the E. coli in vitro reconstituted translations of CAC5 template with different combinations of aminoacyl-tRNA and analyzed using 12% denaturing urea PAGE. Three translations were set up containing different combinations of aminoacyl-tRNAs required to make fulllength peptide-oligonucleotide conjugates. One reaction was set up containing only BiotRNAgug (lanes 1 and 4); one was set up containing only Ala-tRNAggc (lanes 2 and 5), and one was set up containing both Ala-tRNAggc and Bio-tRNAgug (lanes 3 and 6). Purified translation products were analyzed in the absence (lanes 1-3) and presence (lanes 4-6) of streptavidin sample loading buffer. The streptavidin-dependent gel shift is indicated by the black arrow.  The translation reaction performed with only Bio-tRNAgug resulted in the formation of prematurely terminated product [35S]fMet-P, and this means that Bio-tRNAgug was not used in peptide elongation and additional aminoacyl-tRNAs were needed in order to  62  make full-length translation products (Figure 13, lanes 1 and 4). Translation performed with only Ala-tRNAggc resulted in the formation of [35S]fMet-P and [35S]fMet-(Ala)npuromycin-CCA(spacer-9)3(A)21, where n is less or equal to 4, and the latter appears as a faint band on top of [35S]fMet-P band, and this shows that peptide elongation required the presence of Ala-tRNAggc (Figure 13, lanes 2 and 5). Translation performed with both Ala-tRNAggc and Bio-tRNAgug showed formation of [35S]fMet-P and [35S]fMet-(Ala)4Bio-puromycin-CCA(spacer-9)3(A)21, and the latter migrates much slower compared to incomplete translation product fMet-(Ala)n-puromycin-CCA(spacer-9)3(A)21 (Figure 13, lane 2 vs. lane 3) and gel-shifts when streptavidin was added (Figure 13, lane 6). Due to the poor resolution of the gel, it is unclear how much fMet-(Ala)n-puromycinCCA(spacer-9)3(A)21 was formed when translation was performed with both aminoacyltRNAs (Figure 13, lanes 3 and 6).  Based on this result, several conclusions can be drawn. First, the formation of full-length hexa-peptide is dependent on the presence of both Ala-tRNAggc and Bio-tRNAgug, and this implies that incorporation of amino acids is not random and is dictated by the mRNA sequence. Second, despite tRNAgug being an artificial tRNA derived from mutating the anticodon region of tRNAggc (procedure is described in section 2.1), it still retains the ability to be recognized by the reconstituted translation system and deliver the amino acid of interest. Third, even when all required aminoacyl-tRNAs are present, a majority of the translation products is: [35S]fMet-P.  63  3.3 Chemical modifications to the selected unnatural amino acids In order to conjugate unnatural amino acids to their assigned tRNAs, the unnatural amino acids need to be first chemically modified by protecting their N-termini with the photoprotecting group 4,5-dimethyoxy-2-nitrobenzyl chloroformate (NVOC-Cl), the NNVOC-amino acids are then C-terminal activated to N-NVOC-amino acid-cyanomethyl ester, which then coupled with the dinucleotides pdCpA to give to N-NVOC-amino acid pdCpA (Figure 14). The N-NVOC-amino acid pdCpA are purified by reverse phase HPLC. The HPLC purification is monitored at wavelengths 260 nm and 350 nm where ε260 = 23,000 M-1cm-1 for pdCpA and ε350 = 6336 M-1cm-1 for NVOC. The peaks that show a ratio of A260 /A350 greater than 2 are collected and lyophilized.  64  Figure 14 A chemical method of preparing an amino acid for ligation to a tRNA. An amino acid is first N-terminally protected with 4,5-dimethoxy-2-nitrobenzyl. The crude mixture was reacted with chloroacetonitrile and chromatographed on silica gel. The esters were coupled to pdCpA at the 2’/3’-OH in dry DMF with catalytical amount of tetrabutylammonium acetate and the mono-acylated products were isolated by HPLC after several hours of reaction. The fractions were lyophilized; the solid was redissolved in 5 ml of 10 mM acetic acid and lyophilized again. Products were obtained as yellow solid and confirmed by ESI-MS.  Seven unnatural amino acids: N-Me-Asp-OH, Aze, Bio, N-Me-Ser-OH, N-Me-Tyr-OH, Oic and N-Me-Nle were chosen to undergo the chemical synthesis as described in section 2.3. Figure 15 shows the structures of final modified unnatural amino acids as confirmed by ESI-MS. Of all the unnatural amino acids, Bio was successfully coupled to pdCpA to give to N-NVOC-Bio pdCpA; all the rest amino acids were activated to N-NVOC-amino  65  acid cyanomethyl esters, some of which (N-NVOC-N-Me-Asp cyanomethyl ester, NNVOC-Aze cyanomethyl ester and N-NVOC-N-Me-Tyr cyanomethyl ester) were coupled to pdCpA but did not yield any products with the expected masses.  66  Figure 15 Structures of chemically modified amino acids as confirmed by ESI-MS.  67  One of the possible reasons that the couplings of N-NVOC-N-Me-Asp cyanomethyl ester and N-NVOC-N-Me-Tyr cyanomethyl ester to pdCpA didn’t yield desired products was that they all had reactive side chains which would interfere with the coupling of cyanomethyl ester with pdCpA by forming unwanted by-products and thus decrease the yield of expected products. The other possible reason is that the residual tetrabutylammonium acetate remained in the collected and salt-exchanged HPLC fractions suppressed ionization of the products during ESI-MS and made them undetectable.  3.4 Compatibility of unnatural amino acids with the E. coli in vitro reconstituted translation system The compatibility of each unnatural amino acid was checked using EF-Tu binding assay, streptavidin assay, MALDI-TOF and proteinase K assay. The overall experiment is described in Figure 16.  68  Figure 16 A flow chart showing the steps involved in testing the compatibility of each unnatural amino acid with the E. coli in vitro reconstituted translation system.  3.41 Chemoenzymatic acylation of tRNAggc(-ca) with chemically modified unnatural amino acids Chemically modified unnatural amino acids, N-NVOC-N-Me-Ala pdCpA, N-NVOC-NMe-Glu pdCpA, N-NVOC-N-Me-Orn pdCpA, N-NVOC-Abr pdCpA, N-NVOC-tBu pdCpA, N-NVOC-Bio pdCpA and N-NVOC-Sar pdCpA were ligated to tRNAggc(-ca) according to the procedure described in section 2.4 (these pdCpA coupled amino acids were synthesized by Vesna Stojanovic and Lin Gao; see Supplementary Figure 1 in appendix for the structures of these unnatural amino acids) .The aminoacyl-tRNAs were then checked using 15% acidic denaturing urea PAGE as described in the section 2.5.  69  The gel is detected by ethidium bromide staining (Figure 17). There are eight sample lanes on the gel, the first lane is tRNAggc(-ca), and it serves as a size control. Lanes 2-8 show the ligation products between tRNAggc(-ca) and chemically modified unnatural amino acids. Lanes 2, 3, 4, 7 and 8 show two bands, where one band corresponds to tRNAggc(-ca) as it has similar migration compared to the control; the other band corresponds to N-NVOC-unnatural amino acid-tRNAggc and it migrates slower due to the additional amino acid except for N-NVOC-N-Me-Glu-tRNAggc, which migrates faster due to the presence of extra negative charge on the glutamate side chain. Lanes 5 and 6 show one major band that migrate slower than the control and this indicates that majority of the ligation products were N-NVOC-Abr-tRNAggc and N-NVOC-tBu-tRNAggc, respectively. Overall, this gel image confirms that tRNAggc was successfully coupled to these unnatural amino acids.  Figure 17 Confirmation of chemo-enzymatically acylated aminoacyl-tRNAggc by 15% acidic denaturing urea PAGE. The gel is detected by ethidium bromide staining. Lane 1 shows tRNAggc(-ca); lanes 2-7 show tRNAggc(-ca) ligated to different unnatural amino acids. Lane 2: N-NVOC-N-Me-Ala-tRNAggc; lane 3: N-NVOC-N-Me-GlutRNAggc; lane 4: N-NVOC-N-Me-Orn-tRNAggc; lane 5: N-NVOC-Abr-tRNAggc; lane 6: N-NVOC-tBu-tRNAggc; lane 7: N-NVOC-Bio-tRNAggc; lane 8: N-NVOC-Sar-tRNAggc.  70  3.42 EF-Tu binding assay Because not all chosen unnatural amino acid have demonstrated successful incorporation into peptides by translation machinery, the unnatural amino acids were tested individually for their compatibility with EF-Tu before introduced into the E. coli in vitro reconstituted translation reactions, and this was done by EF-Tu binding assay [53]. EF-Tu is an elongation factor involved in prokaryotic translation; and its role is to bind aminoacyltRNAs and deliver them to the ribosome during peptide elongation. The EF-Tu binding assay detects conformational changes in EF-Tu upon binding to its substrates, and this conformational change can be detected by a change in migration pattern on native gel. A detailed procedure is described in section 2.6. The native gel is visualized by silver staining (Figure 18). The unnatural amino acid-tRNAggc tested were N-Me-Ala-tRNAggc, N-Me-Glu-tRNAggc, N-Me-Orn-tRNAggc, Abr-tRNAggc, tBu-tRNAggc, Bio-tRNAggc and Sar-tRNAggc. The negative control used was tRNAggc(-ca).  71  Figure 18 EF-Tu binding assay showed EF-Tu has affinity to all tested tRNAggc precharged with different unnatural amino acids. The assay was performed on 8% native polyacrylamide gel and detected by silver staining. Lane 1 shows EF-Tu binding to tRNAggc(-ca) and lanes 2-8 show the EF-Tu binding to the following amino acid-tRNAggc: N-Me-Ala-tRNAggc (lane 2), N-Me-Glu-tRNAggc (lane 3), N-Me-Orn-tRNAggc (lane 4), Abr-tRNAggc (lane 5), tBu-tRNAggc (lane 6), Bio-tRNAggc (lane 7) and Sar-tRNAggc (lane 8). EF-Tu incubated with tRNAggc(-ca) shows a major band just below the well followed by a faint broad band (Figure 18, lane 1). EF-Tu incubated with tRNAggc precharged with unnatural amino acids also show a major band at the same position as the negative control, but the broad band shown below the major band is much more intense (Figure 18, lanes 2-8). The appearance of more than one band on the native gel reflects the different conformations of EF-Tu: one in its inactive form and one in its aminoacyl-tRNA bound active form. Based on EF-Tu migration pattern in the absence of its substrate (Figure 18, lane 1), the major band observed just below the well is most likely the inactive form of the EF-Tu. As unnatural aminoacyl-tRNAs were added, the faster migrating band appears  72  more intense as a result of EF-Tu binding to the aminoacyl-tRNA. Because the broad, faster migrating band appear more intense in all lanes 2-8, this EF-Tu assay seems to suggest that EF-Tu has affinity for all the tested unnatural amino acid-tRNAs, and these unnatural amino acid-tRNAs could potentially be incorporated into peptides.  3.43 E. coli in vitro reconstituted translation with unnatural amino acids The template used to perform E. coli in vitro reconstituted translation experiment using unnatural amino acids was CAC5 (Table 2). The coding sequence of CAC5 was aug gcc gcc gcc gcc cac, where aug coded for fMet-tRNAfMet, gcc coded for the amino acid to be tested (amino acid-tRNAggc), and cac coded for Bio-tRNAgug. A 100-µl translation reaction was set up for each type of amino acid. Except Ala-tRNAggc, which was enzymatically  acylated,  all  the  other  unnatural  amino  acid-tRNAggc  were  chemoenzymatically acylated using modified tRNAggc(-ca) as described in section 2.4. The translations were set up as described in section 2.8. After translation, the translation products were purified (2.91) and analyzed in the presence of the streptavidin sample loading buffer (section 2.93).  The translation reactions yielded two major populations of translation products that can be clearly distinguished on the gel (Figure 19). The lower band on the gel corresponds to the prematurely terminated translation product that can be represented as [35S]fMet(amino acid)n-puromycin-CCA(spacer-9)3(A)21, where n = 0, 1, 2, 3 or 4. The upper band is the streptavidin-dependent gel shift and corresponds to the fully translated product: [35S]fMet-(amino acid)4-Bio-puromycin-CCA(spacer-9)3(A)21. The calculated percentage  73  of gel shift is: 6% with Ala, 19% with N-Me-Ala, 12% with N-Me-Glu, 17% with N-MeOrn, 19% with Abr, 20% with tBu, 31% with Bio and 18% with Sar.  Figure 19 Making  35  S-labeled hexapeptide-oligonucleotide conjugates consisting of  unnatural amino acids by the E. coli in vitro reconstituted translation system. Translation reactions were performed with enzymatically acylated Ala-tRNAggc (lane 1), chemoenzymatically acylated N-Me-Ala-tRNAggc (lane 2), chemoenzymatically acylated N-Me-Glu-tRNAggc (lane 3), chemoenzymatically acylated N-Me-Orn-tRNAggc (lane 4), chemoenzymatically acylated Abr-tRNAggc (lane 5), chemoenzymatically acylated tButRNAggc  (lane  6),  chemoenzymatically  acylated  Bio-tRNAggc  (lane  7)  and  chemoenzymatically acylated Sar-tRNAggc (lane 8). The bottom band corresponds to 35Slabeled prematurely terminated translation products. The gel shift indicated with the black arrow corresponds to  35  S-labeled fully translated products. The percentage of gel  shift was calculated using the equation: gel shift (%) = intensity of gel shift / total sample intensity per lane x 100%. The signal intensities were quantified using ImageQuant 5.2.  The percentage of gel shift is a reflection of the proportion of molecules containing Bio, and since Bio is the last amino acid to be incorporated into the hexapeptide-  74  oligonucleotide fusions encoded by the CAC5 template, the percentage of gel shift can be used to deduce the proportion of full-length translated products from total translation products. An exception to this is the E. coli in vitro reconstituted translation reaction performed  with Bio-tRNAggc,  which  generated  peptide-oligonucleotide fusions  containing one or more Bio molecules. Based on the percentage of gel shift values, translation performed with natural amino acid Ala appears to give the least amount of fully translated products. The amino acid Als was the only amino acid acylated to the tRNAggc using an enzymatic method that required alanyl-tRNA synthetase, and the salt concentration as well as the yield of aminoacyl-tRNA could differ from the other unnatural aminoacyl-tRNAs. In addition, the amount of aminoacyl-tRNA was difficult to measure because aminoacyl-tRNA cannot be distinguished from unacylated tRNA on a 15% acidic denaturing urea gel due to the small size difference. The percentage of gel shift obtained after in vitro translation with Bio-tRNAggc correlates to the formation of peptide-oligonucleotides [35S]fMet-(Bio)n-puromycin-CCA(spacer-9)3(A)21, where n = 1, 2, 3, 4 or 5. For translations performed with other unnatural amino acids, the calculated percentage of gel shift values correlate to the formation of [35S]fMet-(unnatural amino acid)4-Bio-puromycin-CCA(spacer-9)3(A)21, and are in the range of 10% - 20%, and there doesn’t seem to be a correlation between the structure of an unnatural amino acid and its efficiency in incorporating into full-length peptide-oligonucleotide conjugates.  75  3.44 Confirmation of peptide-oligonucleotide conjugates made by the E. coli in vitro reconstituted translation system using MALDI-TOF MALDI-TOF mass spectrometry was used to confirm that E. coli in vitro reconstituted translation system could make hexapeptide-oligonucleotide conjugates consisting of unnatural amino acids. The first step in performing MALDI-TOF analysis was to set up a positive control translation that could be used to optimize the experimental conditions for performing MALDI-TOF.  The template used was CAC1 (Table 2). It was translated using rabbit reticulocyte lysate (RRL), an in vitro eukaryotic translation system. RRL was used because it is a highly efficient in vitro translation system, and there is no prematurely terminated translation products observed when used in an mRNA display format: i.e. only a single band corresponding to fully translated peptide-oligonucleotide conjugate is observed by denaturing urea PAGE analysis. RRL translation was set up as described in section 2.101. Translation of CAC1 template with RRL would give rise to the peptide-oligonucleotide fusions: Met-His-(Ala)4-puromycin-CCA(spacer-9)3(A)21-ccg ccg ccg ccg ccg ccg cac gua aagcaggagg. The fusions were first treated with RNase A to remove the mRNA portion of the molecule, purified from the translation mixture by oligo (dT) cellulose, then DNase I treated to remove the DNA portion of the molecule. At the end of the purification process, peptides with the sequence Met-His-(Ala)4-puromycin were submitted for MALDI-TOF (see section 2.103 for MALDI-TOF conditions).  76  MALDI-TOF spectrum shows a signal at m/z 1099.87 (Figure 20). The mass 1099.87 corresponds to the sodium/potassium adduct of peptide-puromycin conjugates Met-His(Ala)4-puromycin, [M-4H+3Na+K]+, with a calculated mass of 1099.23, where methionine is oxidized and puromycin depurinated.  Figure 20 Confirmation of peptide-puromycin conjugates generated from rabbit reticulocyte lysate by MALDI-TOF.  77  The same MALDI-TOF condition used to analyze peptide-oligonucleotide conjugates from RRL was applied to the analysis of peptide-oligonucleotide conjugates from the E. coli in vitro reconstituted translation reactions. The E. coli in vitro reconstituted translations were set up using CAC5 as described in 2.102. The aminoacyl-tRNAs used were Sar-tRNAggc. The purified and treated translation products should consist of fMet(Sar)n-puromycin, n = 0, 1, 2, 3, or 4 and fMet-(Sar)4-Bio-puromycin.  The resulting MALDI-TOF spectrum shows a signal at m/z 1437.8 (Figure 21); and this corresponds to a sodium adduct of fMet-(Sar)4-Bio-puromycin, [M-3H+4Na]+ with a calculated mass of 1437.48. No signals are detected that would match any prematurely terminated translated product: fMet-(Sar)n-puromycin.  78  Figure 21 MALDI-TOF confirmation of fMet-(Sar)4-Bio-puromycin made by E. coli in vitro reconstituted translation.  E. coli in vitro reconstituted translation reactions were set up using each of the following amino acids N-Me-Ala, N-Me-Glu, N-Me-Nle, tBu, Hom, Oic and Pro (see Supplementary Figure 1 in Appendix for all the structures). The purified and nucleasetreated translation products should consist of fMet-(amino acid of interest)n-puromycin, n = 0, 1, 2, 3, or 4 and fMet-(amino acid of interest)4-Bio-puromycin; and the presence of these two major population of molecules was confirmed by streptavidin-dependent gel shift on 12% denaturing urea gels (see section 2.102 for details). However, the resulting  79  MALDI-TOF spectra give poor signal-to-noise ratios, and no signals matching any of the expected peptide-puromycin conjugates were detected. Overall, the MALDI-TOF results seem to be inconsistent.  3.45 Confirmation of peptides made by the E. coli in vitro reconstituted translation system using proteinase K Since MALDI-TOF mass spectrometry didn’t confirm the masses of peptideoligonucleotide conjugates consisting of multiple unnatural amino acids made by the E. coli in vitro reconstituted translation system; an alternative proteinase K treatment experiment was proposed. Proteinase K is a non-specific endonuclease that can cleave peptide bonds adjacent to the carboxyl group of aliphatic and aromatic amino acids. Since most of the unnatural amino acids used in the E. coli in vitro reconstituted translation system were N-Me amino acids, peptide bonds found in the peptides consisting of these unnatural amino acids should be tertiary amide bonds and resist to proteinase K digestion.  E. coli in vitro reconstituted translations were set up using Sar-tRNAgug and tRNAggc precharged with each of the following amino acids: N-Me-Orn, Abr, tBu, N-Me-Nle, Oic, Bio, Pro and Ala. These translations would lead to the formation of [35S]fMet-(amino acid)n-puromycin-CCA(spacer-9)3(A)21 where n = 0, 1, 2, 3, or 4; and [35S]fMet-(amino acid)4-Sar-puromycin-CCA(spacer-9)3(A)21. An additional reaction was set up without adding any tRNAs pre-charged with amino acids, and this would lead to the formation of only [35S]fMet-P. The translated products were purified and subjected to proteinase K treatment before analyzed using 12% denaturing urea PAGE as described in section 2.11.  80  Purified products from each translation reaction were shown after 0 hr, 3 hr and 20 hr proteinase K treatments (Figure 22). The intensity of each band was quantified using ImageQuant 5.2 software; and the percentage remaining (%) after proteinase K treatment for each type of peptide-oligonucleotide conjugates was calculated using the equation: Percentage Remaining (%) = Intensity after Proteinase K treatment / Intensity with no proteinase K treatment x 100%. The percentage remaining (%) value for each type of peptide-oligonucleotide conjugates is plotted on a bar graph as shown in Figure 23.  81  Figure 22 Assessment of proteinase K susceptibility of  35  S-labeled peptide-  oligonucleotide conjugates consisting of different amino acids. The peptideoligonucleotides were translated with the E. coli in vitro reconstituted translation system. The peptide-oligonucleotides containing N-Me-Orn (lane 1), Abr (lane 2), tBu (lane 3), N-Me-Nle (lane 4), Oic (lane 5), Bio (lane 6), Pro (lane 7), only fMet-P (lane 8), and Ala (lane 9) were analyzed. Each type of peptides were divided into three equal proportions; one served as no proteinase K treatment group, one was treated with 5 µg of proteinase K for 3 hr and one was treated with 5 µg of proteinase K for 20 hr.  82  Figure 23 Percentages of peptide-oligonucleotide conjugates remaining after 3 hr and 20 hr of proteinase K treatment vs. the composition of peptide-oligonucleotide conjugates. The peptide composition is shown on the x-axis, the percentage remaining after 3 hr (blue) and 20 hr proteinase K treatment (purple) are shown on the y-axis. The black dotted line marks the percentage remaining (%) value for peptide-oligonucleotides conjugates containing Ala. Based on the percentage remaining (%) values, translation products consisting only of [35S]fMet-P showed some degree of proteinase K resistance despite the fact that its only peptide bond is a secondary amide bond, and this resistance might be due to its small size and/or the non- peptide-like composition. Since this fMet-P also exists in translated products from translations reactions performed with different unnatural amino acids as shown in previous results (Figure 12, Figure 13 and Figure 19), the percentage remaining values (PR) for peptide-oligonucleotide conjugates containing different unnatural amino  83  acids would be partially contributed by the proportion of fMet-P present in each sample, and the contribution from the actual peptide composition (PRP or percentage remaining due to actual peptide composition) should be calculated using the equation: PRP = PR – (Proportion of fMet-P) x (percentage remaining of fMet-P after proteinase K treatment); where the proportion of fMet-P may vary in each type of peptideoligonucleotide conjugates, and the percentage remaining of fMet-P is 24% after 20 hr of proteinase K treatment.  Based on the equation, proteinase K resistant peptide-oligonucleotide conjugates should show a PR value greater that 24%, where 24% is the calculated percentage remaining value for fMet-P at 20 hr of proteinase K digestion. As shown in Figure 23, all peptideoligonucleotide conjugates show PR values higher than 24%, except the peptideoligonucleotide conjugates translated using the natural amino acid Ala, which suggests that peptide-oligonucleotide conjugates consisting of Ala are not resistant to proteinase K digestion. A range of proteinase K resistance is observed in other peptide-oligonucleotide conjugates after 20 hr of proteinase K digestion; however, the differences in proteinase K resistance can be contributed by a number of factors: the differences in proportion of fMet-P in each type of peptide-oligonucleotide conjugates due to different incorporation efficiency of each unnatural amino acid, potential structural bias in proteinase K recognition, and the composition differences in peptide-oligonucleotide conjugates.  Overall, the proteinase K experiment demonstrates that peptide-oligonucleotide conjugates translated using different amino acids show differences in resistance to the  84  proteinase K treatment; and the peptide-oligonucleotide conjugates containing Ala show the least resistance. These differences in the proteolytic resistance observed amongst peptides translated using different amino acids indirectly reflect the differences in peptide compositions, and imply that these amino acids were incorporated into peptides by the E. coli in vitro reconstituted translation system.  3.5 Generating a tRNA library In order to generate a peptide-like library consisting of unnatural amino acids by the E. coli in vitro reconstituted translation system, all codons present in the library need to be re-assigned to allow unnatural amino acid incorporation. Since unnatural amino acids are delivered to their assigned codons by tRNAs containing complementary anti-codons, making a peptide-like library first requires making a library of tRNAs that are capable of delivering unnatural amino acids. In the tRNA library, each tRNA has two features: a unique anti-codon region and a modified 3’ end that lacks the ca di-nucleotides to allow its subsequent ligation to an unnatural amino acid. Genes coding for tRNAs with different anti-codon regions were derived from a single gene coding for tRNAggc by PCR mutagenesis as described in section 2.1. After the genes were amplified and confirmed by sequencing, the tRNA(-ca) were then in vitro transcribed from these genes as described in section 2.2. Each type of tRNA generated was assigned a number (see Supplementary Table 1 in Appendix for a full list of tRNAs in the library). The purified and refolded tRNAs were then analyzed using 12% denaturing urea PAGE to confirm their size and purity (Figure 24).  85  Figure 24 Size and purity confirmation of tRNA22(-ca) to tRNA33(-ca) by denaturing urea PAGE analysis. The gel is detected by ethidium bromide staining, and the identity of each tRNA is indicated by its assigned number. Denaturing urea PAGE analysis of purified tRNAs(-ca) from tRNA22(-ca) to tRNA33(-ca) shows all tRNAs migrated similarly and each tRNA appeared as one single band (Figure 24). This confirms that these purified tRNAs were uniform in size and there was no nuclease contamination.  3.6 Making mRNA Libraries The mRNA libraries were designed to contain the following features: a) 5’ non-coding sequence to ensure proper translation initiation; b) a coding sequence that starts with the codon aug, followed by a few random codons nns, where n = g, c, a or u; s = g or c; by combination, each nns codon would allow the random incorporation of 4 x 4 x 2 = 32 different unnatural amino acids; following the random codons, there were six more codons that would code for fixed amino acid; c) most of the mRNA libraries also had a 3’ non-coding region with the sequence (A)21(spacer-9)3ACC(puromycin), and this region would be important in linking the nascent peptides to the mRNA templates and thus achieving the mRNA display format. There are five mRNA libraries: V3Ap, V3B, V3Bp, V4p, and V5p. V3Ap, V4p and V5p were made using the procedure described in section  86  2.121; V3B and V3Bp were made using the procedure described in section 2.122. The sequences of these libraries are shown in Table 2. The purity of the mRNA libraries was checked using denaturing urea PAGE and detected by ethidium bromide staining as shown in Figure 25. All mRNA libraries analyzed using the denaturing urea gels show single band, which indicates that the libraries are pure and ready for in vitro translation.  Figure 25 Denaturing urea PAGE analysis of mRNA libraries. The gels are detected by ethidium bromide staining. The following libraries are analyzed using 17% denaturing urea PAGE: V3Ap (lane 1); V4p (lane 2) and V5p (lane 3). The following libraries are analyzed using 10% denaturing urea PAGE: V3B (lane 4) and V3Bp (lane 5).  3.7 Making peptide libraries by the E. coli in vitro reconstituted translation system 3.71 Making peptide-oligonucleotide libraries To test whether the E. coli in vitro reconstituted translation system can make peptidesoligonucleotide conjugates consisting of unnatural amino acids from translating mRNA libraries, Sar and Bio are introduced into the translation reactions of V3Ap, V4p and V5p. 87  The amino acid Sar would be ligated to the tRNA library to give to Sar-tRNAsnn, and Bio would be ligated to tRNAuuu(-ca) to give to Bio-tRNAuuu. During in vitro translation, SartRNAsnn would deliver sarcosine amino acids to all 32 possible codons in the random coding region of the library; and Bio-tRNAuuu would deliver Bio amino acids only to the aaa codons present in the constant coding region of the library. At the end of the translation reactions, the translation products should consist of [35S]fMet-(Sar)n-template, n = 0, 1, 2, 3, 4, 5 (when used V3Ap), 6 (when used V4p), or 7 (when used V5p); and [35S]fMet-(Sarcosine)m-Bio-template, m= 5 (when used V3Ap), 6 (when used V4p), or 7 (when used V5p). Streptavidin-dependent gel shift assay is then used to detect the proportion of translation products that had successful Bio incorporation.  V3Ap, V4p and V5p were translated by the E. coli in vitro reconstituted translation system as described in section 2.123. For each library, two translation reactions were set up; one served as a negative control that didn’t contain any unnatural amino acid-tRNAs and the other served as the experimental group that contain Sar-tRNAsnn and Bio-tRNAuuu. Half of the purified translation products from the negative control group and all purified translation products from the experimental group were RNase A treated before analyzed using 12% denaturing urea PAGE in the presence of streptavidin sample loading buffer.  For the translations performed with V3Ap, the purified translation products from the negative control translation were analyzed in both the absence and presence of RNase A (Figure 26, lanes 1 and 2). The expected translation product is [35S]fMet-V3Ap before RNase A treatment, and [35S]fMet-P after RNase A treatment. The gel shows that the  88  RNase A treated translated products migrate much faster than the non-treated sample and this change in migration is due to the loss of the mRNA portion in the purified translation products. The experimental translation performed with V3Ap resulted in translation products with the same migration pattern as the negative control (Figure 26, lane 3 vs. lane 2). Similar observations are made in translations performed with V4p and V5p (Figure 26, lanes 4-9).  89  Figure 26 E. coli in vitro reconstituted translations of V3Ap, V4p, and V5p in the presence of unnatural amino acids. All translation products were analyzed in the presence of streptavidin sample loading buffer. The amount of Sar-tRNAsnn is indicated by the Sar-tRNAsnn to template ratio or library coverage. Lanes 1, 4 and 7 show the nonRNase A treated translation products from the negative control reactions with V3Ap V4p, and V5p, respectively; Lanes 2, 5 and 8 show the RNase A treated translation products from the negative control reactions with V3Ap V4p, and V5p, respectively; Lanes 3, 6 and 9 show the RNase A treated translation products from the experimental translations with V3Ap V4p, and V5p, respectively. Based on this experiment, two conclusions can be drawn. First, the change in migration pattern of the translation products upon RNase A treatment implies that mRNA libraries were translated in the mRNA display format. Second, since translations with unnatural amino acid-tRNAs yielded translation products showing similar migration pattern as  90  [35S]fMet-P (translation product of negative control translations) and didn’t show any streptavidin-dependent gel shift, the E. coli in vitro reconstituted translation system seems to have low efficiency of incorporating amino acids into peptides when the templates are V3p, V4p and V5p. . Since initial E. coli in vitro translation experiment with mRNA libraries V3Ap, V4p and V5p didn’t yield peptides of significant lengths; V3Ap, the smallest library was studied in detail. Four E. coli in vitro reconstituted translation reactions were set up using V3Ap as the template. The amount of Sar-tRNAsnn introduced to the translation reactions varied from 0 µg to 20 µg and was calculated using the equation: Amount of Sar-tRNAsnn (pmol) = 20 (pmol of V3Ap) x 3 (number of random codons) x N; where N = intended number of library coverage. Bio-tRNAuuu was added to all translation reactions. The translation products were purified, treated with RNase A and analyzed using 12% denaturing urea PAGE in the presence of streptavidin sample loading buffer. The streptavidin-dependent gel shift was used as a detection method to see if peptide-oligonucleotide conjugates [35S]fMet-(Sar)5-Bio-puromycin-CCA(spacer-9)3(A)21were made.  Translations performed without any Sar-tRNAsnn led to the formation of [35S]fMet-P and does not show any gel shift as expected (Figure 27, lane 1). Translations performed with increasing amounts of Sar-tRNAsnn should in theory result in an increased formation of [35S]fMet-(Sar)n-puromycin-CCA(spacer-9)3(A)21 (n =1, 2, 3, 4, or 5) and possibly [35S]fMet-(Sar)5-Bio-puromycin-CCA(spacer-9)3(A)21. However, the purified translation products from these reactions look the same as the negative control and no gel shift is  91  detected (Figure 27, lanes 2-4). It seems that majority of the translation products from translation reactions containing sarcosine-tRNAsnn and Bio- tRNAsnn are [35S]fMet–P.  Figure 27 E. coli in vitro reconstituted translations of V3Ap in the presence of increasing amount of Sar-tRNAsnn. Streptavidin sample loading buffer was added to all samples prior to PAGE analysis. The amount of Sar-tRNAsnn is indicated by the SartRNAsnn to template ratio or library coverage. Lane 1 shows translation products from translation without any unnatural amino acid-tRNAs. Lanes 2-4 show translation products from translations performed with Bio-tRNAuuu and increasing concentration of Sar-tRNAsnn. This result confirms the previous result shown by Figure 26 that the E. coli in vitro reconstituted translation system has trouble making peptide-oligonucleotides of significant length when the templates are V3p, V4p and V5p. Several hypotheses are 92  raised on why the E. coli in vitro reconstituted translation system doesn’t work well with these templates. One possibility is that the random coding region of the library templates may be biased. In theory, there would be 32 equally represented possible codons in the random region of templates; in reality, the 32 possible codons might not be in equal proportion, and some codons might be found at higher frequency than others. If some codons with particular sequences were significantly over-represented, the amount of SartRNAsnn introduced into the translation mixture might not be enough to cover the randomized region and would thus increase the likelihood of forming prematurely terminated translation products. The second possibility is that Sar amino acids might have successfully incorporated into the random coding region of the V3p templates to form [35S]fMet–(Sar)n-puromycin-CCA(spacer-9)3(A)21, but due to the small molecular size of Sar, [35S]fMet–(Sar)n-puromycin-CCA(spacer-9)3(A)21 was indistinguishable from [35S]fMet–P on 12% denaturing urea PAGE. The third possibility is that the puromycin attachment at the end of each library template interferes with the in vitro translation of the template.  To test the first and second hypotheses, a new mRNA library named V3B was designed and ordered from a different company (Integrated DNA Technologies). This library’s random region consists of 3 random codons and the aaa codon is placed immediately after the random codons. This library was purified and ligated to the puromycincontaining oligonucleotides as described in section 2.122, and the ligated library was named V3Bp (Figure 25). Two unnatural amino acids, N-Me-Val and Bio, were used in the translation of V3Bp. Three E. coli in vitro reconstituted translation reactions were set  93  up: one positive control reaction and two experimental reactions. For the positive control reaction, Bio was ligated to the tRNA library to give to Bio-tRNAsnn, in vitro translation using Bio-tRNAsnn would lead to the formation of [35S]fMet-(Bio)n-V3Bp, n = 0, 1, 2 or 3. For the experimental reactions, N-Me-Val was ligated to the tRNA library to give to NMe-Val-tRNAsnn, and Bio was ligated to tRNAuuu(-ca) to give to Bio-tRNAuuu. In vitro translations using N-Me-Val-tRNAsnn and Bio-tRNAuuu would lead to the formation of [35S]fMet-(N-Me-Val)n-V3Bp, n = 0, 1, 2, or 3; and [35S]fMet-(N-Me-Val)3-Bio-V3Bp. The streptavidin-dependent gel shift assay was used to detect the proportion of translation products that had successful Bio incorporation. The experimental procedure is described in section 2.123.  Translation products from the positive control reaction should show streptavidindependent gel shift, however, the phosphorimage shows only one band corresponding to [35S]fMet-P; there is no distinct streptavidin-dependent gel shift detected that would confirm the presence of [35S]fMet-(Bio)n-puromycin-CCA(spacer-9)3(A)21 (Figure 28, lane 1). Purified translation products from the experimental reactions also show only one band corresponding to [35S]fMet-P and no streptavidin-dependent gel shift is observed (Figure 28, lanes 2 and 3). Overall, translations using V3Bp shows that the E. coli in vitro reconstituted translation system can’t incorporate a detectable amount of unnatural amino acids after [35S]fMet.  94  Figure 28 E. coli in vitro reconstituted translations of V3Bp in the presence of unnatural amino acids. All samples were analyzed in streptavidin sample loading buffer. The amount of N-Me-Val-tRNAsnn is indicated by the N-Me-Val-tRNAsnn to template ratio or library coverage. Lane 1 shows the translation product from translation performed with Bio-tRNAsnn. Lanes 2 and 3 show the translation products from translations performed with Bio-tRNAuuu and increasing concentrations of N-Me-ValtRNAsnn.  Based on this and previous results, the failure in making small peptides consisting of unnatural amino acids from mRNA libraries is not caused by potential problems associated with library sequences since similar results are shown consistently across all libraries tested so far. Based on the result from positive control translation performed with V3Bp, incorporation of unnatural amino acids seems to be hindered completely. Therefore, failure to detect peptides of significant lengths due to the poor resolution 95  associated with denaturing urea PAGE analysis is also unlikely. To test the possibility that puromycin attachment at the end of each library template interferes with the in vitro translation, experiments were carried out to test what would happen if V3B, instead of V3Bp was translated by the E. coli in vitro reconstituted translation system.  3.72 Making peptide libraries in the absence of mRNA display format E. coli in vitro reconstituted translations of V3B template were set up as described in section 2.123. Two reactions were set up: a positive control reaction using Bio-tRNAsnn and an experimental reaction using N-Me-Val-tRNAsnn and Bio-tRNAuuu. Because V3B templates do not have puromycin attachment at the end, their translation products are not covalently attached to puromycin-containing oligonucleotides that would facilitate their purification by oligo (dT) cellulose. Instead, the translation products were partially purified using the method described in the section 2.92. They were then analyzed in the absence and presence of streptavidin buffer by 10% denaturing urea PAGE.  Translation products from the positive control translation should consist of [35S]fMet and [35S]fMet-(Bio)n where n = 1, 2 or 3, where only the latter could be detected by the streptavidin-dependent gel shift. A streptavidin-dependent gel shift is observed for the positive control reaction and this indicates that [35S]fMet-(Bio)n is made (Figure 29, lane 2). Translation products from the experimental translation reaction should consist of [35S]fMet, [35S]fMet-(N-Me-Val)n where n = 1, 2 or 3, and [35S]fMet-(N-Me-Val)3-Bio(N-Me-Val)m where m = 0, 1, 2, and only molecules containing Bio could be detected by streptavidin-dependent gel shift. A streptavidin-dependent gel shift is observed in lane 4  96  (Figure 29). The gel shift indicates that some of the translation products are [35S]fMet-(NMe-Val)3-Bio-(N-Me-Val)m. In addition to the gel shift, an unknown intensely stained band is observed in all lanes, and based on its large size, it is likely to be some kind of translation factor that was purified along with the peptide and would bind to [35S]fMet either specifically or non- specifically, or [35S]fMet-tRNAfMet.  97  Figure 29 Testing the translation inhibitory effect of puromycin in the E. coli in vitro reconstituted translation system using V3B. Partially purified translation mixture from translation of V3B in the presence of Bio-tRNAsnn was analyzed in the absence and presence of streptavidin loading buffer in lanes 1 and 2 respectively. Partially purified translation mixture from translation performed with N-Me-Val-tRNAsnn and Bio-tRNAuuu was analyzed in the absence and presence of streptavidin loading buffer in lanes 3 and 4 respectively. The amount of N-Me-Val-tRNAsnn is indicated by the N-Me-Val-tRNAsnn to template ratio or library coverage. The gel shift is indicated with the black arrow. The result from E. coli in vitro reconstituted translation of V3B shows that small peptides can be made. This is in contrary to the results generated by E. coli in vitro reconstituted translations of V3Bp, V3Ap, V4p and V5p. The major difference between these two sets of  98  experiments is that the former used templates without puromycin attachment and the latter used templates with puromycin attachment. Therefore, the presence of puromycin molecule seems to affect the peptide elongation process in the E. coli in vitro reconstituted system, in other words, peptides cannot be made from the E. coli in vitro reconstituted system in mRNA display format.  When mRNA display technology was first developed in rabbit reticulocyte lysate, puromycin conjugation to prematurely terminated translation products was not observed, i.e. only fully translated products conjugated to puromycin molecules were observed; and puromycin conjugation to the peptides were believed to be a slow process and can only occur when there is translational pause which happens at translation termination [12, 23]. Based on the observation that most translation products generated were [35S]fMet-P from E. coli in vitro reconstituted translation of V3Ap, V3Bp, V4p and V5p, it is reasonable to speculate that the translational pause occurs more often in E. coli in vitro reconstituted translation system, and thus gives puromycin the opportunity to conjugate to the elongating peptides and prematurely terminate translations. If significant translational pauses have occurred during E. coli in vitro reconstituted translation, the speed of making full-length peptide should decrease dramatically, and a time course experiment should show an increase in the amount of full-length peptides being made over time.  99  3.73 A time course experiment set out to address the kinetic issue with the E. coli in vitro reconstituted translation system The time course experiment was performed as follows: two 150-µl translation reactions were set up as described in section 2.8 using template CAC5 except the reconstituted translation reactions contained final concentrations of 0.8 µM purified and heat activated ribosomes (40oC for 30 min) and 17 µM of either deprotected Ala-tRNAggc or SartRNAggc. Deprotection of aminoacyl-tRNAs was performed by placing the aminoacyltRNAs at the side of a PCR tube in front of a xenon lamp (Newport Corporation) with a 315 nm cut-off filter at 350 W for 1 min to remove the photocleavable protecting group. A 25 µl aliquot from the translation mixture was taken out at each of the following time point: 0.5 hr, 1 hr, 1.5 hr, 2 hr, 3 hr, and 6 hr. The stop salt solution was immediately added to the sample to quench the reaction by adjusting the final concentration of KCl and Mg(OAc)2 to 550 and 50 mM, respectively. The quenched translation mixtures were incubated at room temperature for 1 hr followed by -20oC overnight. The translation products were purified and analyzed as described in the section 2.9.  Translations performed using natural amino acid Ala show increased formation of fully translated products (from 10% to 14%) as the translation time increased from 0.5 hr to 3 hr (Figure 30, lanes 1-6). For translations performed using unnatural amino acid Sar, the formation of full-length translation products increased from 3% to 12% as the translation time increased to 3 hr. There was no further increase in gel shift as translation time went over 3 hr (Figure 30, lanes 7-12).  100  Figure 30 E. coli in vitro reconstituted translations of CAC5 template in the presence of either natural or unnatural amino acid over a 6 hour period The translations were performed using either Ala-tRNAggc (left phosphorimage) or Sar-tRNAggc (right phosphorimage) and the purified translation products were analyzed in the presence of streptavidin loading buffer. The top band in the phosphorimages corresponds to the gel shift or fully translated product, and the bottom band corresponds to the prematurely terminated translation products. Translation was quenched at different time point: 0.5 hr, 1 hr, 1.5 hr, 2.0 hr, 3.0 hr, and 6.0 hr. The percentage of gel shift was calculated using the formula: intensity of top band / total signal x 100%.  Based on Figure 30, several conclusions can be drawn. First, the gradual increase in streptavidin-dependent gel shift over time observed in translations performed using Ala and Sar suggests that the E. coli in vitro reconstituted translation system is much slower than E. coli in vivo translation system (typically 12-21 aa/s [55]) at making full-length peptides, regardless the type of amino acids being incorporated. Second, the change in percentage of gel shift seems to be more pronounced from 0.5 to 1 hr. Third, after the 3  101  hr translation period, the proportion of fully translated product reaches a plateau and no more full-length translation products are made.  102  4 Discussion and conclusions 4.1 Discussion The overall goal of this research project is to make a library of peptide-like compounds consisting of unnatural amino acids in an mRNA display format using the E. coli in vitro reconstituted translation system. The project entails the validation of a streptavidindependent gel shift assay for detection of unnatural amino acid incorporation by the E. coli in vitro reconstituted translation system, chemoenzymatic conjugation of unnatural amino acids to tRNAs, the test for unnatural amino acids’ compatibility with the E. coli in vitro reconstituted translation system, identity-confirmation of purified peptides by MALDI-TOF and proteinase K treatment, generations of tRNAs and mRNAs libraries, and the manufacture of peptide-like libraries.  In order to assess the efficiency of the E. coli in vitro reconstituted translation system in generating full-length peptide-oligonucleotide conjugates consisting of one or more unnatural amino acids, the unnatural amino acid Bio is assigned to be the last amino acid to be incorporated in each peptide, and the full-length peptide-oligonucleotide conjugates can be detected by streptavidin-dependent gel shift assay. The streptavidin assay is validated to be specific to Bio-containing molecules and the detected gel shift is the result of retarded migration of Bio-containing molecules upon binding to streptavidin. Using the streptavidin assay, translation reactions performed with GCC1 and CAC5 have shown that the E. coli in vitro reconstituted translations are capable of making full-length dipeptide-oligonucleotide  conjugates  and  hexapeptide-oligonucleotide  conjugates,  103  respectively. In both cases, the formation of full-length translation product requires the presence of aminoacyl-tRNAs to recognize the codons of templates, and formation of prematurely  terminated  translation  product  [35S]fMet-(amino  acid)n-puromycin-  CCA(spacer-9)3(A)21, n = 0, 1, 2, 3, or 4, is observed.  The unnatural amino acids are introduced into the E. coli in vitro reconstituted translation system by a chemo-enzymatic method [38, 40, 41]. This method involves the chemical coupling of unnatural amino acids with dinucleotide pdCpA to give to N-NVOC-amino acid pdCpA, PCR amplification followed by T7 polymerase directed in vitro transcription to generate tRNAs that lack the last ca dinucleotides at 3’ end, and the enzymatic ligation of pdCpA coupled unnatural amino acids to tRNAs(-ca). The rate limiting step in this procedure is the chemical coupling of unnatural amino acids with pdCpA, unnatural amino acids with hydrophobic side chains tend to give higher yields; and unnatural amino acids with reactive side chains should be protected first before going through the procedure to minimize side reactions. The ligation of N-NVOC-amino acids pdCpA to tRNA(-ca) yields N-NVOC-amino acid-tRNAs and the latter can be distinguished from tRNA(-ca) by 15% acidic denaturing urea PAGE. The yield of ligated products does not seem to depend on the structures of N-NVOC-amino acids-pdCpA used for ligation and this observation is in agreement with some of other published findings [38, 56]. After ligation is confirmed to be successful, the ligation products, N-NVOC-unnatural amino acids-tRNAs, are photo-deprotected to give to unnatural amino acids-tRNAs before they are tested for recognition by the E. coli in vitro reconstituted translation system. As the enzyme responsible for delivering aminoacyl-tRNAs to the ribosomes during translation,  104  EF-Tu needs to recognize unnatural amino acids-tRNAs first. EF-Tu binding assay detects EF-Tu’s change in migration pattern on a native polyacrylamide gel as a result of its change in conformation upon binding to its substrates [53]. The assay shows that EFTu has affinity for all tested amino acids-tRNAggc, even though the pairings between amino acids to tRNAggc are orthogonal and the structures of amino acids differ. This observation is in agreement with other published findings, which have shown that translation machinery can accommodate a variety of tRNAs precharged with different unnatural amino acids [39]. When these unnatural amino acids-tRNAs are tested for their incorporation into peptides by the E. coli in vitro reconstituted translation system, streptavidin assay detects full-length translation products in all cases, and the formation of these full-length translation products relies on four successive incorporations of unnatural amino acids. The proportion of full-length translation products, as detected by streptavidin assay, varies among translations performed with different amino acids, however there doesn’t seem to be a clear correlation between the efficiency of making full-length products and the structures of unnatural amino acids. Other published findings have found differences in incorporation efficiency amongst N-Me-amino acids, however the observed differences are variable depending on the methods used to acylated tRNAs with N-Me-amino acids and assays applied to detect the unnatural amino acid incorporation [57-60]. There is also evidence showing that the incorporation efficiency for an unnatural amino acid also depends on the identity of its carrier tRNA, whether it be natural occurring or unnaturally made, mRNA codons used for incorporation, the type of amino acid preceding incorporation and the location of incorporation [58, 59, 61].  105  To further confirm that small peptides-mRNA conjugates consisting of unnatural amino acids are made by the E. coli in vitro reconstituted translation system, the translation products are checked by MALDI-TOF for mass confirmation. The positive control is Met-His-(Ala)4-puromycin translated by rabbit reticulocyte lysate (RRL) and MALDITOF detects the peptides with oxidized methionine and depurinated puromycin. MALDITOF data for peptide-puromycin conjugates from the E. coli in vitro reconstituted translation system shows a clear m/z signal that matches the formation of full-length peptide-puromycin conjugates fMet-(Sar)4-Bio-puromycin with non-oxidized methionine and intact puromycin, but shows no signals that match any of the other expected peptidepuromycin fusions translated using other amino acids. Overall, the MALDI-TOF data seems to be inconsistent: while DNA fragmentation and adduct formation can be a potential problem in MALDI-TOF [62], and methionine is susceptible to oxidation, these events are not observed consistently across all samples though all samples were purified and treated the same. It is also unclear why MALDI-TOF fails to yield signals for most peptide-oligonucleotide conjugates translated from the E. coli in vitro reconstituted translation system. There may be a number of factors that have contributed to these conflicting results: low product yields from the E. coli in vitro reconstituted translation system, some peptide-puromycin conjugates are more stable than others, the matrix may not be suitable for detecting peptide-puromycin conjugates, and some of the experimental conditions used in MALDI-TOF may not be compatible with the detection of peptide/puromycin conjugates from the E. coli in vitro reconstituted translation system. The last two potential problems may have a significant impact on the interpretation of MALDI-TOF spectra. Indeed, the only published method for performing MALDI-TOF  106  mass confirmation of peptide-nucleotides conjugates has shown an observed m/z signal that differs from the calculated m/z by at least 8 Da as a result of DNA attachment [63].  An alternative method to indirectly assess whether the peptide mRNA conjugates consisting unnatural amino acids is made is to perform a proteinase K experiment. Proteinase K is an endonuclease that hydrolyzes peptide bonds after aliphatic and hydrophobic amino acids. Peptides containing N-Me-amino acids should be more resistant to proteolysis compared to those containing only α-amino acids. The proteinase K experiment shows that translation products containing unnatural amino acids have more proteinase K resistance after a 20 hr treatment compared to those containing only the natural amino acids Ala; however, there is no clear trend to show that N-Me-unnatural amino acids-containing peptides are more resistant than α-unnatural amino acidscontaining peptides. The latter observation may be partially explained by the variations in incorporation efficiency of different unnatural amino acids into peptides by the E. coli in vitro reconstituted translation system and the potential structural bias in proteinase K recognition.  To build a library of peptide-oligonucleotides conjugates by the E. coli in vitro reconstituted translation system, libraries of mRNA templates and tRNAs need be designed and made. Four mRNA libraries V3Ap, V3Bp, V4p and V5p are made in mRNA display format; all these libraries have a randomized coding region to allow random incorporation of unnatural amino acids, a constant coding region to code for a specific type of amino acid, and puromycin-containing oligonucleotides after the coding region.  107  To facilitate the incorporation of unnatural amino acids; tRNA(-ca) with different anticodon regions are designed to cover every possible sequence found in the random coding region of mRNA templates and they are derived from tRNAalaW by PCR mutagenesis followed by in vitro transcriptions. All four mRNA libraries, when translated by the E. coli in vitro reconstituted translation system, fail to yield significant amount of peptidemRNA conjugates longer than one amino acid. However, V3B, an mRNA library without the mRNA display format, can be successfully translated by the E. coli in vitro reconstituted translation system to yield tetra-peptides as reflected by streptavidindependent gel shift. Therefore, it appears that the puromycin attachment at the ends of V3Ap, V3Bp, V4p and V5p contribute to the failed translations. Since premature conjugation of puromycin to incomplete peptides is not observed in eukaryotic translation systems and it is believed that puromycin conjugation is a slow process that only occurs when there is translational pause that happens at the end of translation process [12, 23], it is speculated that excessive translational pauses during translational elongation may have occurred in the E. coli in vitro reconstituted translation system that gives puromycin the opportunity to conjugate to the incompletely translated peptides. A time course experiment performed using CAC5 templates showed that the formation of full-length peptides increases over a 3 hr period and this is in contrast to the translation speed in a natural in vivo E. coli translation system (12-21 aa/s [55]). This confirms that the E. coli in vitro reconstituted translation system is slow at making peptides-oligonucleotide conjugates. This may be the result of a variety of potential issues present in the E. coli in vitro reconstituted translation system. First, additional translation factors, other than the ones found in the E. coli in vitro reconstituted translation system, had been reported that  108  may have an effect on peptide synthesis [64, 65]; second, there could be contaminating aminoacyl-tRNA synthetases present in the system that deacylate unnatural amino acids from the unnatural amino acid-tRNAs and thus dramatically decrease their working concentration; third, the tRNAs we used in our system do not contain the nucleoside modifications that are normally present in an in vivo system and the unnatural amino acid-tRNA pairs are orthogonal, both of which will have effects on the recognition by EF-Tu / ribosomes, and may slow down the translation process [59, 66].  4.2 Conclusions Introducing unnatural amino acids into the E. coli in vitro reconstituted translation system has the great potential of making peptide-like compounds consisting of a variety of structures. Combining this method with mRNA display technology opens up the possibility of screening for novel peptidomimetics using a biological system. However, based on experiments performed with a number of different assays, this approach shows a number of limitations: first, despite that our E. coli in vitro reconstituted translation system has the ability to incorporate a number of different unnatural amino acids into peptide-mRNA conjugates as detected by streptavidin-dependent gel shift, a majority of the peptide-mRNA conjugates consist of truncated peptide-mRNA fusions. This low efficiency of making full-length peptides become more pronounced when translating a library of peptide-mRNA conjugates is attempted. Second, except by streptavidindependent gel shift assay, full-length peptide-mRNAs conjugates that consist of unnatural amino acids cannot be confirmed by MALDI-TOF, and proteinase K experiment can only  109  show some differences in proteolytic resistances between peptide-mRNA conjugates consisting of Ala and those consisting of unnatural amino acids.  4.3 Future work There are several issues that need to be addressed in response to the limitations listed in section 4.2: First, it is necessary to assess the translation efficiency of the E. coli in vitro reconstituted translation system for incorporating unnatural amino acids without the mRNA display format; this requires the development of new methods to purify and detect the incomplete and complete translation products; second, MALDI-TOF experiments should be attempted on peptides rather than peptide-puromycin conjugates to minimize the potential problems associated with DNA instability and the low yield issue when mRNA display technology is used with E. coli reconstituted translation machinery, and provide a more direct assessment on the incorporation of unnatural amino acids by the E. coli in vitro reconstituted translation machinery. So far, we have demonstrated that the E. coli in vitro reconstituted translation system has the ability to make small peptide libraries consisting mainly of unnatural amino acids. More effort is needed to link the peptide library to their coding templates to achieve mRNA display.  110  Bibliography 1.  Brenner, S., et al., Encoded combinatorial chemistry. Proc Natl Acad Sci U S A,  1992. 89(12): p. 5381-3. 2.  Feng, S., et al., Molecular basis for the binding of SH3 ligands with non-peptide  elements identified by combinatorial synthesis. Chem Biol, 1996. 3(8): p. 661-70. 3.  Jiracek, J., et al., Development of highly potent and selective phosphinic peptide  inhibitors of zinc endopeptidase 24-15 using combinatorial chemistry. J Biol Chem, 1995. 270(37): p. 21701-6. 4.  Terrett, N.K.B., et al., J.Bioorg.Med.Chem Lett, 1995. 5(917).  5.  Nazif, T., et al., Global analysis of proteasomal substrate specificity using  positional-scanning libraries of covalent inhibitors. Proc Natl Acad Sci U S A, 2001. 98(6): p. 2967-72. 6.  Lam, K.S., et al., The "One-Bead-One-Compound" Combinatorial Library  Method. Chem Rev, 1997. 97(2): p. 411-448. 7.  Smith, G.P., et al., Phage Display. Chem Rev, 1997. 97(2): p. 391-410.  8.  Schreuder, M.P., Brekelmans, S., van der Ende, H., and Klis, F.M., Targeting of a  heterologous protein to the cell wall of Saccharomyces cerevisiae. Yeast, 1993. 9: p. 399409. 9.  Little, M., et al., Bacterial surface presentation of proteins and peptides: an  alternative to phage technology? Trends Biotechnol, 1993. 11(1): p. 3-5. 10.  Hanes, J., et al., In vitro selection and evolution of functional proteins by using  ribosome display. Proc Natl Acad Sci U S A, 1997. 94(10): p. 4937-42.  111  11.  Mattheakis, L.C., et al., An in vitro polysome display system for identifying  ligands from very large peptide libraries. Proc Natl Acad Sci U S A, 1994. 91(19): p. 9022-6. 12.  Nemoto, N., et al., In vitro virus: bonding of mRNA bearing puromycin at the 3'-  terminal end to the C-terminal end of its encoded protein on the ribosome in vitro. FEBS Lett, 1997. 414(2): p. 405-8. 13.  Roberts, R.W., et al., RNA-peptide fusions for the in vitro selection of peptides  and proteins. Proc Natl Acad Sci U S A, 1997. 94(23): p. 12297-302. 14.  Francisco, J.A., et al., Production and fluorescence-activated cell sorting of  Escherichia coli expressing a functional antibody fragment on the external surface. Proc Natl Acad Sci U S A, 1993. 90(22): p. 10444-8. 15.  Georgiou, G., et al., Display of heterologous proteins on the surface of  microorganisms: from the screening of combinatorial libraries to live recombinant vaccines. Nat Biotechnol, 1997. 15(1): p. 29-34. 16.  Traut, R.R., et al., The Puromycin Reaction and Its Relation to Protein Synthesis.  J Mol Biol, 1964. 10: p. 63-72. 17.  Green, R., et al., In vitro complementation analysis localizes 23S rRNA  posttranscriptional modifications that are required for Escherichia coli 50S ribosomal subunit assembly and function. Rna, 1996. 2(10): p. 1011-21. 18.  Noller, H.F., et al., Unusual resistance of peptidyl transferase to protein  extraction procedures. Science, 1992. 256(5062): p. 1416-9. 19.  Samaha, R.R., et al., A base pair between tRNA and 23S rRNA in the peptidyl  transferase centre of the ribosome. Nature, 1995. 377(6547): p. 309-14.  112  20.  Sharma, D., et al., EF-G-independent reactivity of a pre-translocation-state  ribosome complex with the aminoacyl tRNA substrate puromycin supports an intermediate (hybrid) state of tRNA binding. Rna, 2004. 10(1): p. 102-13. 21.  Monro, R.E., et al., Ribosome-catalysed reaction of puromycin with a  formylmethionine-containing oligonucleotide. J Mol Biol, 1967. 25(2): p. 347-50. 22.  Monro, R.E., et al., Ribosome-catalysed peptidyl transfer: effects of some  inhibitors of protein synthesis. J Mol Biol, 1967. 28(1): p. 161-5. 23.  Liu, R., et al., Optimized synthesis of RNA-protein fusions for in vitro protein  selection. Methods Enzymol, 2000. 318: p. 268-93. 24.  Starck, S.R., et al., Puromycin oligonucleotides reveal steric restrictions for  ribosome entry and multiple modes of translation inhibition. Rna, 2002. 8(7): p. 890-903. 25.  Miyamoto-Sato, E., et al., Specific bonding of puromycin to full-length protein at  the C-terminus. Nucleic Acids Res, 2000. 28(5): p. 1176-82. 26.  Jermutus, L., et al., Recent advances in producing and selecting functional  proteins by using cell-free translation. Curr Opin Biotechnol, 1998. 9(5): p. 534-48. 27.  Stiege, W., et al., The potentials of the in vitro protein biosynthesis system. J  Biotechnol, 1995. 41(2-3): p. 81-90. 28.  Shimizu, Y., et al., Cell-free translation reconstituted with purified components.  Nat Biotechnol, 2001. 19(8): p. 751-5. 29.  Matveev, S.V., et al., Effect of the ATP level on the overall protein biosynthesis  rate in a wheat germ cell-free system. Biochim Biophys Acta, 1996. 1293(2): p. 207-12. 30.  Kitaoka, Y., et al., Cooperativity of stabilized mRNA and enhanced translation  activity in the cell-free system. J Biotechnol, 1996. 48(1-2): p. 1-8.  113  31.  Bruce, A.G., et al., Replacement of anticodon loop nucleotides to produce  functional tRNAs: amber suppressors derived from yeast tRNAPhe. Proc Natl Acad Sci U S A, 1982. 79(23): p. 7127-31. 32.  Frankel, A., et al., In vitro selection for sense codon suppression. Rna, 2003. 9(7):  p. 780-6. 33.  Wang, L., et al., Expanding the genetic code of Escherichia coli. Science, 2001.  292(5516): p. 498-500. 34.  Forster, A.C., et al., Pure translation display. Anal Biochem, 2004. 333(2): p.  358-64. 35.  Mathews, C.K., Van Holde, K.E. and Ahern, K.G., Biochemistry. 1999.  36.  Kung, H.F., et al., DNA-directed in vitro synthesis of beta-galactosidase. Studies  with purified factors. J Biol Chem, 1977. 252(19): p. 6889-94. 37.  Forster, A.C., et al., A simplified reconstitution of mRNA-directed peptide  synthesis: activity of the epsilon enhancer and an unnatural amino acid. Anal Biochem, 2001. 297(1): p. 60-70. 38.  Robertson, S.A., et al., A General and Efficient Route for Chemical  Aminoacylation of Transfer RNAs. J Am Chem Soc, 1990. 113: p. 2722-2729. 39.  Cornish, V.W., et al., Probing Protein Structure and Function with an Expanded  Genetic Code. Angew Chem Int Ed Engl, 1995. 34: p. 621-633. 40.  Heckler, T.G., et al., T4 RNA ligase mediated preparation of novel "chemically  misacylated" tRNAPheS. Biochemistry, 1984. 23(7): p. 1468-73. 41.  Hecht, S.M., et al., "Chemical aminoacylation" of tRNA's. J Biol Chem, 1978.  253(13): p. 4517-20.  114  42.  Forster, A.C., et al., Programming peptidomimetic syntheses by translating  genetic codes designed de novo. Proc Natl Acad Sci U S A, 2003. 100(11): p. 6353-7. 43.  Susan M. Miller, R.J.S.S.N.R.N.Z.J.M.K.W.H.M., Comparison of the proteolytic  susceptibilities of homologous L-amino acid, D-amino acid, and N-substituted glycine peptide and peptoid oligomers. Drug Development Research, 1995. 35(1): p. 20-32. 44.  Fahr, A., Cyclosporin clinical pharmacokinetics. Clin Pharmacokinet, 1993. 24(6):  p. 472-95. 45.  Cody, W.L., et al., Design of a potent combined pseudopeptide endothelin-  A/endothelin-B receptor antagonist, Ac-DBhg16-Leu-Asp-Ile-[NMe]Ile-Trp21 (PD 156252): examination of its pharmacokinetic and spectral properties. J Med Chem, 1997. 40(14): p. 2228-40. 46.  Haviv, F., et al., Effect of N-methyl substitution of the peptide bonds in luteinizing  hormone-releasing hormone agonists. J Med Chem, 1993. 36(3): p. 363-9. 47.  Cruz, M., et al., Inhibition of beta-amyloid toxicity by short peptides containing  N-methyl amino acids. J Pept Res, 2004. 63(3): p. 324-8. 48.  Zhang, S., et al., Poly-N-methylated alpha-peptides: synthesis and X-ray structure  determination of beta-strand forming foldamers. Chem Commun (Camb), 2006(5): p. 497-9. 49.  Chang, J.Y., et al., Thrombin specificity. Selective cleavage of antibody light  chains at the joints of variable with joining regions and joining with constant regions. Eur J Biochem, 1985. 151(2): p. 225-30. 50.  Shine, J., et al., Determinant of cistron specificity in bacterial ribosomes. Nature,  1975. 254(5495): p. 34-8.  115  51.  Pestova, T.V., et al., Reconstitution of eukaryotic translation elongation in vitro  following initiation by internal ribosomal entry. Methods, 2005. 36(3): p. 261-9. 52.  Murakami, H., et al., A highly flexible tRNA acylation method for non-natural  polypeptide synthesis. Nat Methods, 2006. 3(5): p. 357-9. 53.  Doi, Y., et al., Elongation Factor Tu Mutants Expand Amino Acid Tolerance of  Protein Biosynthesis System. J Am Chem Soc, 2007. 54.  Bayer, E.A., et al., Postsecretory modifications of streptavidin. Biochem J, 1989.  259(2): p. 369-76. 55.  Bremer, H., et al., Modulation of chemical composition and other parameters of  the cell by growth rate. 2nd ed. Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology. Vol. 2. 1996, Washington, D.C.: ASM Press. 56.  Bain, J.D., et al., Site-specific incorporation of nonnatural residues during in  vitro protein biosynthesis with semisynthetic aminoacyl-tRNAs. Biochemistry, 1991. 30(22): p. 5411-21. 57.  Kawakami, T., et al., Messenger RNA-programmed incorporation of multiple N-  methyl-amino acids into linear and cyclic peptides. Chem Biol, 2008. 15(1): p. 32-42. 58.  Zhang, B., et al., Specificity of translation for N-alkyl amino acids. J Am Chem  Soc, 2007. 129(37): p. 11316-7. 59.  Pavlov, M.Y., et al., Slow peptide bond formation by proline and other N-  alkylamino acids in translation. Proc Natl Acad Sci U S A, 2009. 106(1): p. 50-4. 60.  Merryman, C., et al., Transformation of aminoacyl tRNAs for the in vitro  selection of "drug-like" molecules. Chem Biol, 2004. 11(4): p. 575-82.  116  61.  Dale, T., et al., Amino acid specificity in translation. Trends Biochem Sci, 2005.  30(12): p. 659-65. 62.  Nordhoff, E., et al., Mass spectrometry of nucleic acids. Mass Spectrometry  Reviews, 1997. 15(2): p. 67-138. 63.  Millward, S.W., et al., A general route for post-translational cyclization of mRNA  display libraries. J Am Chem Soc, 2005. 127(41): p. 14142-3. 64.  Green, R.H., et al., Requirements for in vitro reconstruction of protein synthesis.  Biochem Biophys Res Commun, 1985. 126(2): p. 792-8. 65.  Ganoza, M.C., et al., Isolation and point of action of a factor from Escherichia  coli required to reconstruct translation. Proc Natl Acad Sci U S A, 1985. 82(6): p. 164852. 66.  Ogle, J.M., et al., Structural insights into translational fidelity. Annu Rev  Biochem, 2005. 74: p. 129-77.  117  Appendix Supplementary Table 1 List of all the tRNAs in the tRNA library. Assigned number of each tRNA 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33  Anti-codon gaa cac gua cua cga gca cca ggg ccg gag guc cuc ccc guu cuu cgu gau gug cug uuu ggc cau ccu gcu ggu caa gga gac gcc cgc cag gcg cgg  Corresponding codon uuc gug uac uag ucg ugc ugg ccc cgg cuc gac gag ggg aac aag acg auc cac cag aaa gcc aug agg agc acc uug ucc guc ggc gcg cug cgc ccg  118  Supplementary Figure 1 Amino acids that are successfully coupled to pdCpA and introduced to the E. coli in vitro reconstituted translation system.  119  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0067132/manifest

Comment

Related Items