ABSTRACT 
 
 
 
 
Title of Document: BUILDING A MAP OF THE DYNAMIC 
RIBOSOME.   
  
 
 Suna Pelin Gulay, Doctor of Philosophy, 2015 
  
 
Directed By: Professor Jonathan D. Dinman 
Cell Biology and Molecular Genetics 
 
 
 
Our understanding of the static structure of the 80S eukaryotic ribosome has 
been enhanced by the emergence of high resolution cryo-electron microscopy and 
crystallography data over the past 15 years. However our understanding of the 
dynamic nature of the ribosome has lagged. High-throughput Selective 2’-Hydroxyl 
Acylation analyzed by Primer Extension (hSHAPE) is easily amenable for 
interrogation of rRNA dynamics. Here we report an improved method of hSHAPE 
data analysis and apply it to translation initiation and elongation complexes of the 
yeast ribosome to identify the changes in rRNA flexibility that occur during these 
processes. Most importantly, we have obtained complete analyses of tRNA binding 
and intersubunit bridge dynamics, as well as overall expansion segment dynamics, as 
the ribosome progresses through the translation elongation cycle. The results from 
these analyses suggest that (1) the yeast P site tRNA binding site is a “hybrid” 
  
between the prokaryotic and mammalian P sites, (2) there may be substates of 
intersubunit rotation, (3) expansion segments may have roles in accommodation. We 
are also able to identify a network of information pathways that connect elongation 
factor binding sites to all tRNA binding sites, five intersubunit bridges and two 
expansion segments. Future directions of this project will focus on improving the 
visualization of our data to better reflect the highly dynamic nature of the yeast 
ribosome and to reveal the underlying causes of the observed rRNA flexibility 
changes.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
  
 
 
 
 
 
 
 
 
BUILDING A MAP OF THE DYNAMIC RIBOSOME 
 
 
 
By 
 
 
Suna Pelin Gulay 
 
 
 
 
Dissertation submitted to the Faculty of the Graduate School of the  
University of Maryland, College Park, in partial fulfillment 
of the requirements for the degree of 
Doctor of Philosophy 
2015 
 
 
 
 
 
 
 
Advisory Committee: 
 
Professor Jonathan Dinman, Chair 
Associate Professor Douglas Julin 
Assistant Professor Vincent Lee 
Professor Anne Simon 
Associate Professor Richard Stewart 
 
 
 
 
 
 
 
  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
© Copyright by 
 
Suna Pelin Gulay 
 
2015 
 
 
 
 
 
 
 
 
 
 
 
 
 
  ii 
 
Dedication 
 
I would like to dedicate this dissertation to my family, especially my sister 
Ece, for lending me her ear at every petty annoyance, my fiancé Darryl, for his 
immense support throughout my studies, peaking in the last three months, and Şeker. 
This would not have become a reality if it were not for all of you. 
 
  iii 
 
Acknowledgements 
 
 
First and foremost, I would like to thank my advisor, Dr. Jonathan Dinman, 
for giving me years of guidance and support, and fulfilling every last-minute request 
of edits or recommendation letters without fail. I learned how to be a scientist from 
working with you. 
I would also like to thank my committee for their time and assistance in 
improving my research. Special thanks to Dr. Anne Simon and Dr. Vincent Lee for 
their support of my future career. 
A big thank you to everyone in the department I had the pleasure of 
interacting with through teaching assistantships and administrative processes, 
especially Dr. John Buchner, Dr. Michelle Brooks, Ms. Gwen Warman and Ms. 
Molly Burke. You made my life easier in ways I cannot express. 
Thank you to everyone that made contributions to my scientific endeavors, 
including classmates, alumni from Dinman lab, and collaborators at UMD, NIH, 
LANL. I am looking forward to continue working with you all. 
Last but not least, very sincere thanks to my support system: friends and 
family that helped make UMD my home away from home. A special note of 
appreciation goes to my “Dinman lab family”: Dr. Sharmishtha, Alicia, Carol, Cassie, 
Drs. Trey & Lara, Vivek, Ryan, and Joe. You all made for an enjoyable experience 
both in and outside the lab. It was a crazy ride, and I appreciate you being there to 
experience it with me. We will always be close friends and colleagues.  
 
  iv 
 
Table of Contents 
 
Dedication ..................................................................................................................... ii 
 
Acknowledgements ...................................................................................................... iii 
 
Table of Contents ......................................................................................................... iv 
 
List of Tables ............................................................................................................... vi 
 
List of Figures ............................................................................................................. vii 
 
List of Abbreviations ................................................................................................... ix 
 
Chapter 1: Introduction ................................................................................................. 1 
1.1. The Ribosome and its static structure ................................................................ 1 
1.2. Translation ......................................................................................................... 7 
1.2.1. Initiation ...................................................................................................... 9 
1.2.2. Elongation ................................................................................................. 13 
1.2.3. Termination and Ribosome Recycling ...................................................... 20 
1.3. Translation Inhibitors ....................................................................................... 24 
1.4. Translational dynamics .................................................................................... 27 
1.5. Experimental approaches related to the ribosome structure and function ....... 31 
1.5.1. X-ray crystallography ............................................................................... 33 
1.5.2. Cryo-EM ................................................................................................... 34 
1.5.3. SmFRET ................................................................................................... 35 
1.5.4. MD simulations ......................................................................................... 37 
1.5.5. Chemical structure probing ....................................................................... 38 
1.6. Project rationale and aims ................................................................................ 44 
 
Chapter 2: The evolution of hSHAPE analyses .......................................................... 47 
2.1. Background and rationale ................................................................................ 47 
2.2. Results .............................................................................................................. 49 
2.2.1. hSHAPE analyses v1.0 ............................................................................. 49 
2.2.2. hSHAPE analyses v2.0 ............................................................................. 51 
2.2.3. Application of method to ribosome complexes and discussion .................... 55 
 
Chapter 3: hSHAPE analyses of translation initiation complexes .............................. 60 
3.1. Background and rationale ................................................................................ 60 
3.2. Results .............................................................................................................. 64 
3.3. Discussion ........................................................................................................ 71 
 
Chapter 4: Walking the ribosome through the translation elongation cycle ............... 75 
4.1. Background and rationale ................................................................................ 75 
4.1.1. Known functional regions of bacterial and eukaryotic rRNAs ................. 76 
  v 
 
4.1.2. Changes in intersubunit bridges ................................................................ 80 
4.1.3. More structural insight into eukaryotic ribosomes ................................... 86 
4.2. Results .............................................................................................................. 88 
4.2.1. Complex assembly and verification .......................................................... 88 
4.2.2. Identification of rRNA functional sites in hSHAPE data ......................... 93 
4.2.3. Intersubunit bridges ................................................................................ 103 
4.2.4. Expansion segments ................................................................................ 109 
4.2.5. Allosteric communication pathways ....................................................... 111 
4.3. Discussion ...................................................................................................... 119 
 
Chapter 5: Conclusion and future directions ............................................................ 134 
 
Chapter 6: Materials and methods ............................................................................ 143 
 
Appendix: 2D difference maps ................................................................................. 156 
 
Bibliography ............................................................................................................. 166 
 
 
 
 
   
 
 
 
 
 
 
  vi 
 
List of Tables 
 
Table 1. Summary of protected and deprotected rRNA nucleotides upon tRNA 
binding. ....................................................................................................................... 78 
Table 2. mRNA-rRNA interactions from hSHAPE data. ........................................... 93 
Table 3. hSHAPE data on tRNA binding sites from bacterial studies........................ 96 
Table 4. hSHAPE data on tRNA binding sites from the mammalian ribosome. ........ 98 
Table 5. EF binding sites from chemical probing studies. ........................................ 101 
Table 6. EF binding sites from X-ray crystallography structures. ............................ 102 
Table 7. hSHAPE data on bacterial, nonrotated conformation bridges. ................... 105 
Table 8. hSHAPE data on eukaryotic, rotated conformation bridges. ...................... 107 
Table 9. Nucleotides involved in reactivity change patterns related to EF binding and 
dissociation. .............................................................................................................. 114 
Table 10. Nucleotides involved in reactivity change patterns related to 
accommodation. ........................................................................................................ 116 
Table 11. Nucleotides involved in reactivity change patterns related to translocation.
................................................................................................................................... 117 
Table 12. Nucleotides involved in reactivity change patterns related to a switch from 
the accommodation to translocation steps of elongation. ......................................... 118 
Table 13. Comparison of the matches between the observed yeast P site and the 
prokaryotic / mammalian P sites. .............................................................................. 121 
 
  vii 
 
List of Figures 
 
Figure 1. The features of bacterial and yeast ribosomes. .............................................. 2 
Figure 2. Ribosomal subunits and functional sites. ...................................................... 3 
Figure 3. 80S ribosome and tRNA, mRNA, EF binding sites. ..................................... 4 
Figure 4. Yeast rRNA domains on 2D and 3D maps. ................................................... 5 
Figure 5. Intersubunit bridges of the yeast ribosome. ................................................... 6 
Figure 6. Overview of translation in bacteria. .............................................................. 7 
Figure 7. Overview of translation in eukaryotes. .......................................................... 8 
Figure 8. Eukaryotic translation elongation cycle. ..................................................... 14 
Figure 9. The structure of A/T state tRNA. ................................................................ 16 
Figure 10. The peptidyltransfer reaction. .................................................................... 17 
Figure 11. The concerted proton shuttle. .................................................................... 18 
Figure 12. Bacterial and broad-spectrum translation inhibitors. ................................. 25 
Figure 13. Eukaryote specific antibiotics. ................................................................... 26 
Figure 14. Motions of the small subunit color coded by the level of displacement. .. 28 
Figure 15. SHAPE reaction......................................................................................... 41 
Figure 16. Mechanisms of RNA SHAPE chemistry. .................................................. 42 
Figure 17. Scheme of highthroughput selective 2’-hydroxyl acylation analyzed by 
primer extension.......................................................................................................... 43 
Figure 18. Box plot analyses indicate extremely skewed distribution of data............ 50 
Figure 19. Frequency histograms also show extremely skewed distribution of data. 51 
Figure 20. Box plot analyses show the data acquire a unimodal, symmetric 
distribution after natural log (ln) transformation and median normalization. ............ 52 
Figure 21. Frequency histograms also show improvement in data distribution upon ln 
transformation. ............................................................................................................ 53 
Figure 22. Difference calculations and data visualization. ......................................... 54 
Figure 23. Data distribution of the reactivity difference between the P site occupied 
complex and the empty ribosome 18S rRNAs. ........................................................... 55 
Figure 24. Representative results from the two statistical approaches. ...................... 56 
Figure 25. The P stalk in images generated by the two statistical approaches. .......... 58 
Figure 26. Model summarizing the interactions between different initiation factors 
and the structural rearrangements of the initiation complex. ...................................... 60 
Figure 27. Possible location of eIF5 and the structure of the individual eIF5 domains.
..................................................................................................................................... 63 
Figure 28. Building initiation complexes.................................................................... 64 
Figure 29. Building initiation complexes.................................................................... 66 
Figure 30. Difference maps of eIF5 complexes and empty SSU. ............................... 67 
Figure 31. Difference maps of eIF5 complexes and empty SSU. ............................... 68 
Figure 32. Difference maps of the full length eIF5 complex and NTD or CTD portion 
complexes. .................................................................................................................. 69 
Figure 33. Difference maps of the full length eIF5 complex and NTD or CTD portion 
complexes. .................................................................................................................. 69 
Figure 34. Possible eIF5 binding sites. ....................................................................... 74 
Figure 35. Expansion segments in 80S yeast ribosome. ............................................. 87 
  viii 
 
Figure 36. The complex assembly scheme. ................................................................ 88 
Figure 37. Results of tRNA binding assays. ............................................................... 90 
Figure 38. Results of dipeptide formation assays. ...................................................... 90 
Figure 39. Rotational status of CHX and ANI complexes.......................................... 91 
Figure 40. Results of eEF2 binding and puromycin (translocation) assays. ............... 92 
Figure 41. Changes in reactivity at the mRNA binding region. ................................. 94 
Figure 42. The yeast P site. ......................................................................................... 99 
Figure 43. P stalk base (GTPase associated center, GAC) in the difference maps... 103 
Figure 44. Regions of protection on 18S rRNA upon subunit joining. .................... 108 
Figure 45. The reactivity changes in 60S interacting nucleotides of the SSU through 
the elongation cycle. ................................................................................................. 109 
Figure 46. hSHAPE reactivity of different expansion segments in the accommodation 
related complexes...................................................................................................... 110 
Figure 47. Pathway between EF binding site and tRNA binding sites. .................... 122 
Figure 48. Pathways between EF binding site and the intersubunit bridges B6, eB11, 
eB13. ......................................................................................................................... 124 
Figure 49. Pathway between EF binding site and the intersubunit bridge B7b/c. .... 125 
Figure 50. Comparison of hSHAPE data to the B1b/c pathway. .............................. 126 
Figure 51. Pathways between EF binding site and the expansion segments ES7L, 
ES27L. ...................................................................................................................... 127 
Figure 52. Possible network of elongation factor binding related communication 
pathways in yeast ribosomes. .................................................................................... 129 
Figure 53. MD-SHAPE analysis of the empty 80S ribosome. ................................. 139 
Figure 54. The hSHAPE data visualized in voxels. .................................................. 141 
 
  ix 
 
List of Abbreviations 
 
1M7   1-methyl-7-nitroisatoic anhydride 
AC   Accommodation corridor 
Ac-Phe-tRNA  N-acetyl-phenylalanyl tRNA 
ANI Anisomycin; used to denote a complex prepared with this 
antibiotic 
ASL   Anticodon stem loop 
CHX Cycloheximide; used to denote a complex prepared with this 
antibiotic 
CP   Central protuberance 
Cryo-EM  Cryogenic electron microscopy 
CTD   Carboxyl terminal domain 
CTT   Carboxyl terminal tail 
DC   Decoding center 
EF   Elongation factor 
ES   Expansion segment 
GAC   GTPase associated center 
hSHAPE  Highthroughput selective 2’-hydroxyl acylation analyzed by
   primer extension 
IF   Initiation factor 
ln   Natural log 
LSU   Large subunit 
NGD   No-go decay 
  x 
 
NTD   Amino terminal domain 
NTT   Amino terminal tail 
ORF   Open reading frame 
Phe-tRNA  Phenylalanyl tRNA 
PIC   Preinitiation complex 
Postacc  Used to abbreviate “postaccommodation” complex 
Posttrans  Used to abbreviate “posttranslocation” complex 
Preacc   Used to abbreviate “preaccommodation” complex 
Pretrans  Used to abbreviate “pretranslocation” complex 
PTC   Peptidyltransferase center 
RF   Release factor 
smFRET  Single molecule Förster resonance energy transfer 
SRL   Sarcin-ricin loop 
SSU   Small subunit 
TC   Ternary complex 
 
 
  1 
 
Chapter 1: Introduction 
 
1.1. The Ribosome and its static structure 
 
Protein synthesis, also called translation, is a crucial step in the central dogma 
of molecular biology. It involves the decoding of genetic information carried by the 
messenger ribonucleic acid (mRNA) by the the ribosome, employing transfer RNAs 
(tRNA) carrying amino acids and protein factors that regulate different steps. To 
understand translation, it is critical to understand the structure, both static and 
dynamic, of the ribosome. The ribosome is a ribonucleoprotein complex consisting of 
two subunits, each of which are composed of ribosomal RNA (rRNA) and proteins. 
The two subunits and the fully associated ribosome are named after their 
sedimentation values. Bacterial 70S ribosomes are made up of the 50S and 30S 
subunits. The 50S subunit is composed of 33 proteins, 5S rRNA (121 bases) and 23S 
rRNA (2904 bases), while 30S is composed of 21 proteins and 16S rRNA (1542 
bases) (Figure 1) 1. Eukaryotic 80S ribosomes are made up of the 60S and 40S 
subunits. In lower eukaryotes, such as Saccharomyces cerevisiae, 60S consists of 46 
proteins, 5S (121 bases), 5.8S (158 bases) and 25S rRNA (3396 bases), while 40S 
consists of 33 proteins and 18S rRNA (1800 bases) (Figure 1) 1. Higher eukaryotes, 
such as Homo sapiens, have even larger ribosomes, in which 60S is composed of 47 
proteins, 5S (121 bases), 5.8S (156 bases) and 28S rRNA (5034 bases), and 40S is 
composed of 33 proteins and 18S rRNA (1870 bases) 1. Although the primary 
sequences of the rRNAs can vary greatly among different species and are used to 
determine phylogenetic relationships, they all fold into complex secondary structures 
  2 
 
that are highly conserved in all domains of life, and can further fold into conserved 
tertiary structures. The eukaryotes have extra rRNA regions in addition to the core 
structure. These are called “expansion segments” (ES) (Figure 1), which get longer 
and more complex as we go higher in the assumed hierarchy of species. 
 
Figure 1. The features of bacterial and yeast ribosomes. 
Comparison of the features of the 80S ribosomes from bacteria and Saccharomyces cerevisiae, and the 
common core between these are shown. Red and blue mark the rRNA and the proteins, respectively, 
with lighter colors used to indicate the elements of the common core while the brighter colors show the 
additional components in each ribosome type. The relative ratios of the conserved and additional 
elements are shown below the structures. Most of the eukaryote specific rRNA components consists of 
expansion segments (ES). Image modified from Melnikov et al, 2012 1. 
 
The architecture of the whole ribosome has prominent features that were first 
identified by electron microscopy. These features, now known to be dictated by 
rRNA domains, include the central protuberance (CP), the L1 stalk, and the L7/L12 
stalk (the P stalk in eukaryotes) in the large subunit (LSU). In the small subunit 
(SSU), the prominent structures are named mostly after body parts (Figure 2). Of these, 
the “head,” the “body,” and the “neck” are the most functionally critical. 
  3 
 
 
Figure 2. Ribosomal subunits and functional sites. 
Shown left is the 40S “small” subunit, right is the 60S “large” subunit. Structurally important regions 
are labeled in black. Proteins are shown in tones of blue, rRNA is shown in grey. Functionally 
important sites are shown in red: DC stands for decoding center, PTC stands for peptidyltransferase 
center, GAC stands for GTPase associated center, SRL stands for sarcin-ricin loop. The structurally 
important, dynamic element P stalk forms the GAC. GAC and SRL together form most of the 
elongation factor (EF) binding site. Image generated by PyMOL, structure from Ben-Shem et al, 2011 
2. 
 
The ribosome is a ribozyme, as the rRNA component harbors its catalytic site: 
the peptidyltransferase center (PTC) where peptide bond transfer between the 
growing peptide chain and a new aminoacyl residue occurs. In addition, while not 
directly catalytic, the GTPase associated center (GAC) which stimulates GTP 
hydrolysis by elongation factors, and the decoding center (DC) where the new 
aminoacyl tRNA recognizes and binds its codon on the mRNA template, as well as 
large portions of the elongation factor, mRNA, and tRNA binding sites are mainly 
composed of RNA. Similarly, a critical structural element required for elongation 
factor binding called the sarcin-ricin loop is composed of a loop at the end of Helix 
95 (H95) of the large subunit, taking its name from the fact that ribotoxins α-sarcin 
and ricin cleave this region. tRNA binding sites are the A (“aminoacyl”) site where 
  4 
 
the incoming aminoacyl tRNA binds, the P (“peptidyl”) site where the growing 
peptide chain is found attached to a tRNA, and the E (“exit”) site from which 
deacylated tRNAs leave the ribosome (Figure 3). Another important regulatory region 
that consists of rRNA is the accommodation corridor (AC), or the path of rRNA 
nucleotides that act as a gate to the acceptor stem and the CCA end of the incoming 
aminoacyl tRNA on its journey to the PTC. 
 
Figure 3. 80S ribosome and tRNA, mRNA, EF binding sites. 
Fully associated 80S ribosome viewed from 40S head - 60S central protuberance side. Some structural 
elements are noted in black. Proteins are shown in tones of blue, rRNA is shown in grey. Functional 
sites are marked with red. mRNA (thick black line) winds around the neck of 40S, to which the tRNAs 
bind (blocks of varying colors). A, P, E sites are shown in different colors. PTC: peptidyl transferase 
center, DC: decoding center, GAC: GTPase associated center, SRL: sarcin-ricin loop. GAC and SRL 
form most of the elongation binding site. 
  
The rRNA folds into many domains that confer both rigidity and flexibility to 
the ribosome, as well as forming the functional sites and pockets mentioned above. 
These domains are shown in Figure 4. The rRNA harbors many modified nucleotides, 
including ribose 2’-O-methylated bases and pseudouridines; these are important for 
ribosome function, as loss of base modifications cause translational fidelity defects 3–
  5 
 
7. Another extraordinary feature of the rRNA is the presence of noncanonical base 
pairs 8. At the secondary and tertiary levels, numerous structural motifs and long 
distance interactions are important both functionally, as indicated by their presence at 
the catalytic sites, and structurally 8. The most common of these are the A-minor 
motifs and interactions. In these motifs, single-stranded adenosines reach into the 
minor groove of a helix, making hydrogen bonds and van der Waals contacts with 
both the bases and the riboses of the interacting nucleotides. A-minor interactions are 
seen between the PTC adenines and the 3’ CCA ends of tRNAs, and between the DC 
adenines and the codon:anticodon duplex. Other motifs and interactions throughout 
the rRNA include tetraloops,  E-loops, U- and K-turns, purine stacks, coaxial 
stacking, and ribose zippers 8,9. 
 
Figure 4. Yeast rRNA domains on 2D and 3D maps. 
  6 
 
A. Two dimensional view of 25S rRNA domains is shown on the left, and the three dimensional view 
(in a similar angle to Figure 1) is shown on the right, keeping the same color code to mark the places of 
different domains. B. Two dimensional view of 18S rRNA domains is shown on the left, and the three 
dimensional view (similar angle to Figure 1) is shown on the right. The domains fold separately both at 
the secondary and tertiary levels, and dictate the structural elements of the whole ribosome (especially 
true for 18S domains). Functional sites are noted on the 2D maps; structurally important regions are 
noted on the 3D maps. Two dimensional views are from Petrov et al, 2014 10, three dimensional views 
are generated using Ribovision 11. 
 
All prominent structural features of the large subunit are functionally 
important, and their protein counterparts are just as critical as the rRNA. For example, 
the L1 stalk includes the ribosomal protein L1, and the L7/L12 contains heterodimers 
of L7 and L12 proteins bound to L10 in prokaryotes, and the acidic P proteins in 
eukaryotes. Intersubunit bridges forming the fully associated ribosomes are other 
regions in which proteins take part (Figure 5) 2. Bridge B1b/c is solely formed by 
protein-protein interactions. The eukaryotic ribosomes have more intersubunit bridges 
compared to the bacterial ribosomes, and these “eukaryote specific” bridges generally 
have a protein component. Finally, uL4 and uL22 are important components of the 
peptide exit tunnel where the growing peptide chain resides, in addition to the rRNA 
bases that line this region.  
 
Figure 5. Intersubunit bridges of the yeast ribosome. 
  7 
 
Conserved bridges are shown in blue, eukaryote specific bridges are shown in red. Except B1b/c, all 
bridges have an rRNA component, with the eukaryote specific bridges solely consisting of 
rRNA:protein interactions. Image from Ben-Shem et al, 2011 2. 
1.2. Translation 
 
Figure 6. Overview of translation in bacteria. 
A simplified version of the whole translation process in bacteria is shown, and comprises of initiation, 
elongation, termination and recycling. Steps 1-3 above form initiation, step 4 belongs to elongation, 
step 5 and 6 are termination and recycling, respectively. Please see text for more detailed information 
on each part of translation. Image modified from Melnikov et al, 2012 1. 
  8 
 
 
Figure 7. Overview of translation in eukaryotes. 
A simplified version of the whole translation process in eukaryotes is shown, and comprises of 
initiation, elongation, termination and recycling. Steps 0-3 above form initiation, step 4 belongs to 
elongation, step 5 and 6 are termination and recycling. Please see text for more detailed information on 
each part of translation. Image modified from Melnikov et al, 2012 1. 
 
Protein translation occurs in four main steps: Initiation, elongation, 
termination and recycling. During initiation, the mRNA binds to the ribosomal small 
subunit and the initiator tRNA binds to the start codon, thus defining the translational 
reading frame. The elongation cycle begins when the next aminoacyl tRNA binds to 
its codon on the mRNA, and continues until a stop codon is reached. Most of peptide 
synthesis occurs during this step of translation through peptide bond transfer. 
  9 
 
Termination occurs when the stop codon is recognized by a variety of trans-acting 
factors, resulting in release of the amino acid chain. This is followed by ribosome 
recycling, which frees the ribosome from the newly translated polypeptide and the 
mRNA, allowing the cycle to be repeated. Overviews of the bacterial and eukaryotic 
translation are shown in Figure 6 and Figure 7. 
1.2.1. Initiation 
 
The key processes of translation initiation in both prokaryotes and eukaryotes 
are the SSU recruitment to the start codon and subunit association. For the first 
process, the protein coding RNAs of most prokaryotes have a highly conserved 
purine-rich sequence, Shine-Dalgarno (SD) sequence, located approximately 8 
nucleotides upstream of the start codon AUG. This is complementary the 3’ end of 
16S rRNA, and base-pairing between the two positions the ribosomal small subunit in 
close proximity with the translation start site. The start codon is heavily biased 
towards AUG in E. coli, however GUG and UUG can be favored in other organisms 
such as Actinobacteria spp. Metagenomics analyses have shown that not all genes 
have the SD sequence, revealing the presence of SD-independent translation initiation 
pathways in bacteria 12. Ribosomal protein S1 is thought to be the key component of 
such SD-independent pathways, as ribosomes with S1 can translate messages with 
weak SD sequences whereas S1-depleted ribosomes cannot, and S1 variants may be 
employed for efficient translation of messages from different prokaryotic species 12–
14. 
Prokaryotic initiation requires three proteinacious initiation factors: IF1, IF2, 
and IF3. These, in complex with the SSU, form the pre-initiation complex, to which 
  10 
 
the mRNA binds through SD-antiSD interactions 15. IF3 is an anti-association factor 
that keeps the two ribosomal subunits separate during initiation and recycling, in 
accordance with its primary binding site on the interface surface of 30S 15–17. IF1 
binds to and blocks the A site of SSU, as well as stimulating activities of the other 
initiation factors. In this respect, it cooperates with IF2 to ensure the correct 
positioning of the initiator tRNA in the P site. IF1 also induces structural 
rearrangements in the SSU, some at the A site, and some longer distance changes that 
influence association and dissociation of the ribosomal subunits 15,16. IF2 is a GTPase 
and its main role is to recruit the initiator tRNA, a formylated methionine tRNA 
(fMet-tRNAfMet) in prokaryotes, to the P site. Its C terminus is highly conserved in 
different species, as well as kingdoms, and contains the recognition site for fMet-
tRNAfMet. The G domain, containing the GTP binding pocket, is structurally related to 
the G domains of other transacting factors 15,16. Structural studies indicate that the 
positioning of IF2 on SSU changes depending on whether it is bound to GTP (ie. 
before GTP hydrolysis), GDPCP (mimics hydrolyzed GTP, before phosphate 
release), or GDP (after phosphate release) 15,16. GTP hydrolysis of IF2-GTP might be 
enhanced during LSU joining 16. It results in the dissociation of the three initiation 
factors, and the complete association of 70S ribosomes, allowing translation 
elongation to resume. 
SSU recruitment to the start codon occurs in two steps in eukaryotes: 
Recruitment of the SSU to the 5’ end of the message (which requires mRNA 
activation), and scanning for the start codon. Switching from a scanning-permissive 
state to scanning-restrictive state upon start codon recognition requires structural 
  11 
 
rearrangements within SSU, in addition to the changes in the components of the large 
initiation machinery. These make eukaryotic initiation overall much more complex 
than the prokaryotic translation initiation. 
Most eukaryotic mRNAs are also posttranscriptionally modified, the most 
common being addition of a 7-methylguanosine cap to the 5’ end, and a polyadenine 
tail to the 3’ end, except histone encoding mRNAs which are not polyadenylated. 
These modifications are thought to be important for labeling nuclear-encoded 
mRNAs as “self”, and to facilitate translation initiation, as different factors bind to 
these regions (eIF4E and poly(A) binding proteins, respectively), and are bridged by 
the large scaffolding protein eIF4G to effectively bring together the ends of the 
mRNA together, forming the closed loop mRNP. The whole eIF4 complex (eIF4F) 
also includes eIF4A, an ATP-dependent DEAD box RNA helicase important to 
resolve the secondary structure of the mRNA, and eIF4B, the cofactor of eIF4A and 
an RNA binding protein that helps to anchor the mRNA and the 18S rRNA. The 
positioning of the eIF4F complex targets the ribosomal subunit to the 5’ end of the 
mRNA, while the closed loop possibly allows for efficient reinitiation on the same 
message. 
Another anchoring protein for binding of the small subunit onto the 
“activated” mRNA is eIF3, which is a component of the 43S preinitiation complex 
(43S PIC). Formation of 43S PIC will be discussed in more detail in Section 3.1, and 
entails binding of eIF1, eIF1A, eIF3, eIF5 and the ternary complex of eIF2, GTP and 
Met-tRNAi
Met to the 40S subunit. eIF1 and eIF3 are known to stimulate ternary 
complex binding to 40S, with the initiator tRNA positioned over the P site 18. eIF2 
  12 
 
and eIF5 are known to interact, and this may be how eIF5 is recruited to the ribosome 
19. eIF1A and eIF1 are the eukaryotic homologs of bacterial IF1 and IF3, respectively. 
In addition to functions homologous to their bacterial counterparts (blocking of the A 
site, and preventing association of ribosomal subunits, respectively), they are 
important in scanning, start codon recognition and the related structural 
rearrangements.  
After 43S PIC is recruited onto the mRNA, scanning for the start codon AUG 
starts. Scanning on structured messages is powered by ATP hydrolysis, which is 
mostly required by eIF4A and eIF4B 20. During scanning, 43S PIC/mRNA is in an 
open conformation induced by eIF1, eIF1A and stabilized by eIF3. eIF1 is also 
important in monitoring tRNA/mRNA interactions, as it regulates the Pout and Pin 
states of the tRNA, where Pin is the fully accommodated state of the initiator tRNA. 
The conformations of the SSU and the initiator tRNA allow the PIC to move on the 
mRNA, until an AUG in good context is encountered. The optimal start codon 
context entails the Kozak consensus sequence, (gcc)gccRccAUGG, where R is a 
purine and the AUG is underlined. Another event that takes place during scanning is 
the GTP hydrolysis of the G protein eIF2 by the GTPase activating protein (GAP) 
eIF5. The inorganic phosphate (Pi) resulting from GTP hydrolysis is not readily 
released, however, and Pi release is thought to be the “committing” step of the 
initiation, as it occurs upon start codon recognition and eIF1 dissociation, signaling 
the formation of correct tRNA/mRNA interactions.  
Upon start codon recognition, 48S PIC forms. The SSU assumes a “closed 
conformation,” possibly provided by the eIF1A-eIF5 interaction. This conformation 
  13 
 
is not permissive for scanning. The initiator tRNA commits to the Pin state, which is 
thought to stabilize the closed conformation. After Pi release, eIF5 and eIF2-GDP 
dissociate as well and are replaced by eIF5B-GTP. eIF5B is homologous to the C 
terminus of prokaryotic IF2, and recruits the large subunit to the 48S complex. eIF1A 
is thought to contribute to subunit association 21. The LSU is prevented from 
associating earlier due to interactions with eIF6 22. eIF6 dissociates upon 
phosphorylation involving RACK1 (Receptor for Activated C Kinase 1) which is a 
scaffolding protein associated with protein kinase C and an ancillary protein of SSU 
23,24. GTP of eIF5B is hydrolyzed upon LSU joining. eIF5B-GDP and eIF1A 
dissociate and the ribosome is now ready to resume with the elongation cycle. 
Even though ribosomal scanning is the main hypothesis for start codon 
recognition in eukaryotic translation initiation, a recent study indicates another 
mechanism called “RNA looping” by the authors 25, whereby a reporter gene may be 
translated independently of the 5’ end of mRNA through looping of the mRNA 
portion between the ribosome binding site and the start codon. This hypothesis, as 
well as the mathematical model that is associated with it, explains the effect of 5’ 
untranslated region length on efficient translation initiation as well. 
1.2.2. Elongation 
 
The elongation cycles of bacteria and eukaryotes are functionally identical, 
the only difference being the names given to the transacting factors that take part in 
the process. The elongation factors of prokaryotes and eukaryotes are homologous in 
sequence and function. An overview is shown in Figure 8. 
  14 
 
 
Figure 8. Eukaryotic translation elongation cycle. 
The ribosome before and after translocation is traditionally named “PRE” and “POST.” In “POST” 
conformation, it has an open A site and a peptidyl tRNA or initiator tRNA on the P site. A ternary 
complex composed of aminoacyl tRNA, eEF1A and GTP binds during the decoding step, followed by 
GTP hydrolysis and accommodation of the incoming tRNA to the LSU A site. This is coupled to a 
“rolling” motion in mammalian ribosomes. Peptidyl transfer (PTF) takes place upon correct 
positioning of the tRNA CCA ends. At this point, any deacylated tRNA remaining on the E site is 
detached, allowing for the “rotation” motion. Rotation is coupled to tRNA CCA end movement to the 
next site on LSU. The rotated conformation locks upon binding of eEF2-GTP. Upon GTP hydrolysis, 
translocation of tRNA anticodons : mRNA codons occur, opening up the A site for the cycle to be 
repeated. eEF2 function is coupled to the reversal of the earlier rolling and rotation motions. Image 
from Budkevich et al, 2014 26. 
 
1.2.2.1. Decoding, accommodation and peptidyl transfer 
 
The cycle begins as a new aminoacyl tRNA, dictated by the next codon on the 
mRNA in a process known as “decoding,” is brought to the A site of the ribosome in 
a ternary complex with the elongation factor Tu (EF-Tu in bacteria, eukaryotic 
elongation factor 1A or eEF1A, in eukaryotes) and GTP. The interaction of the 
  15 
 
aminoacyl tRNA anticodon with the mRNA codon is accompanied by the movements 
of SSU rRNA nucleotides G530 (G580 in yeast), A1492-93 (A1755-56 in yeast) and 
LSU nucleotide A1913 (A2256 in yeast). 
In the vacant ribosomes, the three adenines normally form stacking 
interactions with each other. Upon binding of the anticodon to the codon, A1913 
rotates freeing the nucleobases of A1492-93 to “flip out”, G530 switches to anti- 
conformation from syn-, forming a part of the stabilization network of the codon-
anticodon duplex 27–32. There are two schools of thought on the roles of these 
nucleotides during decoding. Earlier research suggested that these occur in response 
to cognate aminoacyl tRNA anticodon - mRNA codon interactions and can 
discriminate against near-cognate ones, by resulting in the “closure” of the small 
subunit whereby the shoulder moves towards the neck 27,29,30. However recent 
research suggests that these structural changes take place whether or not there is a 
cognate or near-cognate tRNA in the A site. In fact, some of these start happening 
after the initiator fMet-tRNAfMet has bound to the P site, and these nucleotides may 
actually form a “static” part of the DC 28,31,32. 
Right after decoding, the tRNA is conformationally distorted, or “bent,” with 
its anticodon stem loop bound to the mRNA codon in the decoding center and its 
acceptor stem bound by the protein factor in the GTPase associated center 33; this is 
also known as the A/T state (Figure 9). Structural and mutational studies indicate that 
Domain I of EF-Tu is responsible for GTP binding, rearrangements in this domain 
causing a “hydrophobic gate” of amino acids to open and allow the catalytic amino 
acid (His84) to come in contact with GTP. His84 is thought to position and activate a 
  16 
 
water molecule for GTP hydrolysis. The ribosome may be contributing to this process 
as well, as GAC is known to stimulate hydrolysis 2500 fold in bacteria, and SRL is 
located adjacent to the nucleotide binding pocket. 
 
Figure 9. The structure of A/T state tRNA. 
The “bent” structure of A/T state tRNA is shown. This conformation is assumed when the elongation 
ternary complex binds the ribosome and before GTP hydrolysis by EF-Tu (eEF1A in yeast). The 
bending occurs in the anticodon stem (purple tRNA) and is due to the localization of the anticodon to 
the SSU A site, while the CCA end of the tRNA is bound to the elongation factor. Upon GTP 
hydrolysis and EF-Tu dissociation, accommodation occurs and A/A state is achieved (dark blue 
tRNA). Image from Schmeing et al, 2009 33. 
 
 
Upon GTP hydrolysis and release of inorganic phosphate, the affinity of the 
protein factor for the aminoacyl-tRNA is decreased and it detaches from the 
ribosome. The acceptor stem of the aminoacyl-tRNA moves into the PTC through a 
corridor of rRNA nucleotides from LSU helices 89-92, propelled by the “bent” 
conformation. This process is known as “accommodation” and is complete when the 
3' end of the incoming tRNA is placed in a position favoring peptide bond transfer in 
the PTC. 
Peptide bond transfer involves a nucleophilic attack by the amino group of the 
A site tRNA aminoacyl residue on the carbonyl carbon of the peptide on the P site 
tRNA resulting in formation of an amide bond and transfer of the peptidyl group onto 
  17 
 
the A-site tRNA. This occurs through the formation of a zwitterionic tetrahedral 
intermediate (Figure 10), followed by deprotonation into a second intermediate (not 
shown) and finally the decomposition of the second intermediate into products. The 
nucleophilic attack and formation of the intermediates have been shown to be the rate 
limiting step on the ribosome, whereas the final breakdown occurs in a separate, rapid 
step 34–36. The rate limiting step proceeds through a concerted proton shuttle 
mechanism (Figure 11).  
 
Figure 10. The peptidyltransfer reaction. 
Image from Scott A. Strobel, via 37. 
 
Despite decades of study, the precise role of the PTC is not entirely clear, as 
catalysis is achieved through the substrates themselves (e.g. possibly the 2’-hydroxyl 
group of A76 of the peptidyl tRNA 38), implying an indirect role for the ribosome 39. 
A2451 in bacteria (A2820 in yeast) is within hydrogen bonding distance of the 
nucleophilic amino group of the aminoacyl tRNA, however it does not have a direct 
effect on peptide bond formation 40. Even though the adenine is conserved in all 
domains of life, the effects of A2451 on peptidyl transfer have been largely attributed 
to the ribose 2’-hydroxyl group through atomic mutagenesis 41. Overall, the PTC is 
thought to be important in accurate positioning of substrates and in water exclusion 
  18 
 
38,42,43. According to this view, A and P loops (LSU Helices 92 and 80, respectively) 
are more crucial than any individual nucleotide for peptide bond formation 44. 
 
Figure 11. The concerted proton shuttle. 
Protons of the concerted shuttle are circled. Image from Rodnina, 2013 36. 
 
1.2.2.2. Rotation and translocation 
 
After peptide bond formation, the next step is to move the mRNA and the 
tRNAs by one codon on the ribosome, opening up the A site again for a new 
aminoacyl tRNA to bind. This process is called “translocation.” It occurs as a result 
of both movements within the ribosome, specifically rotation, and by the action of 
another elongation factor, EF-G in bacteria or eEF2 in eukaryotes.  
After accommodation and during peptidyl transfer, the tRNAs are in the so-
called “classical” A/A and P/P states. Upon peptidyl transfer, the small subunit rotates 
(5-8°) with respect to the large subunit 26,45; this correlates with movement of the 
CCA ends of the A site and P site tRNAs to P and E sites, respectively. The new 
conformation of the tRNAs are called “hybrid” A/P and P/E states. Due to the 
structure of the E site, the P/E state can only be assumed by a deacylated tRNA, 
hence it has to occur after peptidyl transfer. The change in tRNA conformation seems 
to be coupled to movement of L1 stalk in a manner to stabilize the P/E hybrid state 46. 
  19 
 
The elongation factor (i.e. EF-G and eEF2) functions in the movement of the tRNA 
ASL and the corresponding mRNA codons to the P and E sites 47, and elongation 
factor binding locks the ribosome into the rotated state. Domain IV of the elongation 
factor, and possibly a conserved histidine residue at the tip of this domain (His699 in 
yeast; posttranslationally modified to diphthamide 48), is inserted into the decoding 
center, and is thought to displace the decoding center rRNA nucleotides 47. Domain 
III contacts both SSU and LSU, allowing the factor to “sense” ribosome conformation 
47. Domain G harbors the GTP binding site as well as switch regions that prevent 
premature GTP hydrolysis 47. Domain II contains structural elements that contact both 
the shoulder of SSU and switch I region before GTP hydrolysis, hence these elements 
may be how GTP hydrolysis, ribosome conformation, and translocation are coupled 
47. Upon GTP hydrolysis, the ribosome is unlocked with additional tRNA hybrid 
states shown in bacteria 42,49–51, and translocation occurs possibly as a result of: 1) 
interaction of the tip of domain IV with the P site codon-anticodon duplex pushing it 
“forward”, 2) the back-rotation of small subunit, 3) interaction of L1 with P/E tRNA 
pulling it “forward”, and 4) interplay between A and P sites (shown in a recent study 
51). 
The release of E site tRNA and the presence of a ribosomal E site at all are 
active questions in the field. In fungi, an additional elongation factor, eEF3, was 
found to bind close to the E site on posttranslocation ribosomes. It was proposed that 
this factor may function in E site tRNA release in fungi, possibly through L1 mobility 
52. eEF3 is fungi-specific, does not have a homolog in bacteria or in mammals, and is 
required for protein synthesis 52,53. 
  20 
 
1.2.3. Termination and Ribosome Recycling 
 
Translation elongation cycle is repeated until a stop codon (UAA, UGA or 
UAG) is encountered. These codons are not recognized by tRNAs but by release 
factors (RF’s). In bacteria, type I release factors RF1 (specific for UAG and UAA) 
and RF2 (specific for UGA and UAA) recognize stop codons 36,54. The codon specific 
motif of RF1 is PVT of domain 2, whereas it is SPF for RF2. When the RFs bind to 
the ribosome, these motifs are located in the 30S decoding center. Thr186 of RF1 
selects for an adenine in the second position of the stop codon, whereas the Ser206 
can form hydrogen bonds with either adenine or guanine in this position. Additional 
selection occurs at the third position, where Thr194 and Gln181 of RF1 allow for 
either adenine or guanine, whereas Val203 of RF2 restricts this to an adenine. The 
first position of the codon is universally restricted to a uridine since a common 
element in both RF’s can form hydrogen bonds with the Watson-Crick edge of this 
nucleotide 54.  
Peptide bond hydrolysis from the P site tRNA is achieved through the 
conserved GGQ motif found in domain 3. This motif locates to the PTC and functions 
in correct positioning of the substrates for hydrolysis. The glutamine component 
selects for water as the nucleophile and excludes larger molecules. This component 
further aids in the hydrolysis reaction by providing a hydrogen bonding network for 
stabilization of the transition states 36,54. Unless the stop codon is recognized, the 
GGQ motif does not locate to its correct place, as the structural rearrangements 
occurring upon stop codon recognition are required for the correct positioning of this 
motif 54.  
  21 
 
Hydrolysis of the peptide bond occurs through a tetrahedral intermediate that 
breaks down into the free peptide product and deacylated tRNA. It does not require 
proton shuttling as only one proton needs to be transferred 35,36. The PTC also plays a 
role in peptide bond hydrolysis, specifically nucleotides A2451 (A2820 in yeast), 
U2506 (U2875 in yeast), U2585 (U2954 in yeast), and A2602 (A2971 in yeast) 35,44. 
Additionally, coupling of stop codon recognition to peptide bond hydrolysis requires 
rearrangements in the decoding center nucleotides including A1492-93 (A1755-56 in 
yeast) of 16S rRNA and A1913 (A2256 in yeast) of 23S rRNA 54. 
After peptide bond hydrolysis, RF1 and 2 remain tightly bound to the 
ribosome. A class II RF, RF3, which is an EF-G-like GTPase, is necessary for the 
release of RF1 and 2. This is thought to occur by rotation of the small subunit upon 
GTP hydrolysis, weakening the bonds of RF1 and 2 with the ribosome 54. 
The process of ribosomal subunit dissociation and release the deacylated 
tRNA remaining on the ribosome is called “recycling” and occurs through the 
combined action of ribosome recycling factor (RRF), EF-G, and IF3 in bacteria. RRF 
and EF-G function in dissociation of the ribosomal subunits in a GTP-dependent 
manner, whereas IF3 stabilizes this dissociation by binding to the 30S subunits and 
preventing the subunits to reassociate until initiation 55. IF3 is also required for 
ejection of deacylated tRNA and the mRNA from the ribosome, and this constitutes 
the rate limiting step of recycling 56. It is thought that EF-G induces structural 
rearrangements similar to the unlocking observed during translocation, rather than 
translocation itself, during recycling, even though RRF mimics an A site tRNA and 
binds to the A site 56. In turn, RRF interacts with intersubunit bridges B2a and B3, 
  22 
 
and the combined action of EF-G and RRF alters these bridges, resulting in 
dissociation of the subunits 56. 
Termination and recycling events in eukaryotes remain active areas of 
research. In eukaryotes, termination is catalyzed by the collaboration of eRF1 and 
eRF3. eRF1 is a class I RF and is responsible for stop codon recognition and peptide 
bond hydrolysis, whereas eRF3, a class II RF, is an EF-Tu-like GTPase. eRF1 is 
tRNA-shaped like its bacterial counterparts, although it has no evolutionary 
relationship to these proteins 53. It decodes stop codons through a highly conserved 
NIKS motif of its amino terminal domain via interactions resembling those of 
codon:anticodon 57. Another motif that is implicated in stop codon recognition is 
YxCxxxF. The middle (M) domain of eRF1 resembles the acceptor stem of a tRNA 
and extends into the PTC, with a GGQ motif in this region being responsible for 
peptide bond hydrolysis 58. This reaction appears to occur as it does in bacteria.  
eRF3 interacts with carboxyl terminus and M domain of eRF1. eRF1 acts as a 
GTP dissociation inhibitor. GTP hydrolysis by eRF3 promotes eRF1 function by 
positioning the M domain into the PTC 59, essentially like GTP hydrolysis by EF-Tu 
aiding in the accommodation of the aminoacyl tRNA CCA end into the PTC. eRF3 
does not promote dissociation of eRF1 from the ribosome, in fact, eRF1 has been 
shown to remain associated with the ribosome following termination and to be 
required for recycling 60,61.  
Following termination in eukaryotes, the ribosome can either be recycled or 
translation can be “reinitiated” on the same circular message 62. Reinitiation is 
thought to occur through the interaction of eRF3 with PABP 63. 
  23 
 
Recycling in eukaryotes consists of separate steps of subunit dissociation, 
catalyzed by ABCE1/Rli1 61, a cytosolic, essential ATP binding cassette family 
protein, and deacylated tRNA ejection followed by dissociation of 40S subunit from 
mRNA 64. Mechanistic insights into recycling in eukaryotes come from studies on the 
no-go decay (NGD) pathway. Dom34 and Hbs1, key components of NGD, are related 
to eRF1 and eRF3 respectively, but instead of catalyzing peptide release, they 
catalyze subunit dissociation in stalled ribosomes. This observation and subsequent 
research showed that the eukaryotic release factors are able to promote subunit 
dissociation at slow rates themselves 65. ABCE1/Rli1 is proposed to increase the rate 
of subunit dissociation through ATP hydrolysis and possibly by rearrangements of the 
intersubunit bridges 66–69. It also functions in an ATP-independent manner to promote 
peptide release during termination, hence possibly staging the sequence of events of 
termination and recycling 53,70. Structural studies provide support for an EF-Tu-like 
function for eRF3 in promoting eRF1 M domain localization to the PTC, as well as a 
model in which upon GTP hydrolysis by eRF3, no factor dissociates but ABCE1/Rli1 
binds, facilitating eRF1 M domain positioning, hence promoting peptide release, and 
subsequently, ribosome recycling 66–69. 
After subunit dissociation, eIF3 is thought to prevent the reassociation of 
subunits 53. Release of deacylated tRNA and mRNA possibly occur through an open 
conformation of the 40S subunit brought about by Ligatin (also known as eIF2D) and 
the related protein pair MCT-1/DENR 71. Release of tRNA and mRNA has also been 
shown to be stimulated in vitro by canonical initiation factors eIF1, eIF1A and eIF3 
60.  
  24 
 
1.3. Translation Inhibitors 
 
Translation inhibitors comprise approximately half of the available antibiotics. 
As a general rule, the inhibitors that bind to the small subunit bind around the neck, as 
this is the region of the SSU that interacts with mRNA and tRNA anticodon. The 
inhibitors that target the large subunit tend to bind in the vicinity of the PTC and the 
peptide exit tunnel, where the CCA end of the tRNA and the growing peptide chain 
are located. Hence, most known translation inhibitors act during elongation cycle of 
translation, and modify codon:anticodon interactions, block tRNA localization to the 
PTC, or block the new peptide chain from growing 72. Bacterial translation inhibitors 
are shown in Figure 12. Due to the conserved nature of the bacterial and eukaryotic 
ribosomal catalytic sites, a number of these (sparsomycin, aminoglycosides, etc) are 
also are toxic to eukaryotic cells. For example, ribotoxins such as α-sarcin cleave the 
SRL of the LSU rRNA, abolishing the ribosome’s ability to stimulate GTP hydrolysis 
activity. Puromycin is a structural analog of the 3’ end of tyrosyl-tRNA and it triggers 
premature release of the peptide chain by forming an amide linkage with it. 
Sparsomycin is a nucleotide analogue that binds at the PTC and inhibits tRNA 
binding at the A site, while promoting translocation and stabilization of the tRNA at 
the P site. Aminoglycosides such as paromomycin bind at the SSU rRNA whereby 
promoting near-cognate tRNA binding, and in some cases, inhibiting small subunit 
rotation, thereby inhibiting translocation 73,74. 
  25 
 
 
Figure 12. Bacterial and broad-spectrum translation inhibitors. 
The steps of translation inhibited by different translation inhibitors are shown, and color coded 
according to the phase of translation. Image from Sohmen et al, 2009 72. 
 
There are also eukaryote specific translation inhibitors, a comprehensive 
structural study of which was recently published 75. The binding sites of, and the steps 
blocked by these inhibitors are shown in Figure 13. Interestingly, most eukaryote 
specific inhibitors bind to the LSU, some of which bind to the E site, thereby 
inhibiting translocation. The most well-known examples of these are cycloheximide 
and lactimidomycin. These antibiotics are closely related but lactimidomycin has an 
additional lactone ring, slowing its insertion into the E site, making it a “slow” 
inhibitor that can only block translocation during the first elongation cycle, as it 
cannot effectively compete with tRNA In contrast, cycloheximide is very efficient at 
this and can block translocation at any stage during the elongation cycle. A number of 
  26 
 
the eukaryote specific translation inhibitors bind to the PTC as well, mostly in the 
vicinity of the 3’ end of A site tRNA. Some of these, such as T-2 toxin and verrucarin 
A, form extensive interactions with the rRNA bases of the A site and extend towards 
and/or into the peptide exit tunnel. Most of the A site blockers are inhibitors of 
peptidyl transfer and some are competitors of accommodation (eg. anisomycin) 76,77. 
 
Figure 13. Eukaryote specific antibiotics. 
A. The binding sites of the antibiotics studied in de Loubresse et al, 2014 75. B. The step of translation 
blocked by the antibiotics shown in A. Out of these, edeine, geneticin G418, pactamycin and 
blasticidin S are broad-spectrum antibiotics and inhibit translation in prokaryotes as well. Image 
modified from de Loubresse et al, 2014 75. 
    
  27 
 
1.4. Translational dynamics 
 
As discussed in Section 1.2, translation is a highly dynamic process. 
Translational dynamics may refer to time dependent changes in either composition or 
conformation of the translational machinery 78. Compositional dynamics are 
determined by bimolecular association and dissociation rate constants, and are 
affected by intermolecular collision frequencies, electrostatic interactions, proper 
binding orientations, and energy barriers for dissociation of noncovalent 
intermolecular interactions. These events are related to tRNA or protein factor 
binding. Single molecule, fluorescence techniques have been especially helpful in 
determining the order of binding of bacterial initiation factors to the ribosome, as well 
as the hydrogen bond energies of mRNA or tRNA-ribosome interactions, uncovering 
the mechanism of mRNA secondary structure unwinding on the ribosome, and the 
mechanism of IRES-mediated translation. Structural studies with cap-dependent 
eukaryotic initiation complexes are underway to determine the mechanistic aspects of 
eukaryotic initiation factor binding. 
Conformational dynamics encompass electronic motions, bond vibrations, 
local conformational changes of protein side chains, nucleic acid bases or sugars, and 
larger rearrangements such as domain movements. These occur on an intramolecular 
level, such as tRNA, translation factor, or ribosome dynamics. As most of these are 
functionally important, they were discussed above in subsection 1.2.2. Elongation. 
Examples of the changes in tRNA conformation include the switch from the “bent” 
A/T state during the decoding step to A/A state before peptidyl transfer, as well as the 
formation of hybrid A/P and P/E states simultaneously with ribosomal rotation, 
  28 
 
requiring rearrangements of the acyl moieties with respect to key rRNA nucleotides. 
Studies of translation factor dynamics focus mainly on EF-G. Domain IV, which 
shows molecular mimicry with the tRNA anticodon arm, undergoes a hinge-like 
motion with respect to domain G (GTPase domain) upon GTP hydrolysis, disrupting 
the interactions between the tRNA at the A site and the mRNA, and driving 
translocation. Another large conformational change is seen in domain II of RRF, 
induced by EF-G domain IV movement during recycling, which results in disruption 
of ribosomal intersubunit bridges. 
 
Figure 14. Motions of the small subunit color coded by the level of displacement.  
A. Fully ratcheted state, observed in bacterial 30S, characterized by rotation with respect to the 50S, 
and a swivel movement of the head domain with respect to the body. B. Rotated state, observed in 
bacterial and eukaryotic small subunits, characterized by rotation with respect to the large subunit. C. 
Subunit rolling with respect to 60S, recently described in mammalian 40S. Image from Budkevich et 
al, 2014 26. 
 
Structural studies have found evidence of both intersubunit and intrasubunit 
ribosomal dynamics. Intersubunit conformational changes (Figure 14) are mainly 
brought about by rotation (shown in ribosomes from various organisms) and rolling 
  29 
 
(shown in mammalian ribosomes) of the small subunit with respect to the large 
subunit. Rotation (a motion of 5°-8°) and the recently described rolling (a motion of 
6°) occur in planes that are perpendicular to each other. Even though rolling occurs 
during accommodation of aminoacyl tRNA to the A site, the reversal of both rolling 
and rotation occurs during translocation, hence both motions are thought to prepare 
the ribosome for translocation 26. 
One type of intrasubunit movement in the small subunit occurs at the 
decoding center, as A1755-56 (A1492-93 in bacteria) “flip” in response to tRNA 
binding 79. Another important motion, observed in bacteria, is the head swivel which 
is the 14° movement of the head domain with respect to the body of the small subunit, 
occurring in the “rolling” plane. This is also thought to be important in translocation. 
Low resolution structural studies have provided insight into ribosomal dynamics 
during eukaryotic initiation 80,81. Upon eIF1 and eIF1A binding, a connection between 
the head and the shoulder is observed on the solvent side, involving the 18S rRNA 
and S3. The beak and the platform also exhibit altered conformations. A so-called 
“latch,” formed by interactions between h18 and h34 and thought to clamp around 
mRNA to prevent dissociation, is “closed” in empty 40S, but “open” or not visible in 
the 40S-eIF1-eIF1A complex, possibly to allow mRNA binding. While there is 
evidence indicating eIF1’s involvement in influencing the conformation of the 
platform and the positions of the mRNA and the initiator tRNA (whereas eIF1A acts 
like IF1 and seemingly only blocks the A site), both factors are required for the full 
conformational change. These changes are observed in the 43S preinitiation 
complexes (ie. in the presence of other initiation factors and initiator tRNA) as well, 
  30 
 
resulting in the opening of the head-body-platform junction and hence the mRNA 
entry channel, and permitting mRNA binding and scanning for the start codon.  
The intrasubunit motions of the large subunit are mostly observed in the so-
called “hinge” regions, the L1 and L7/L12 stalks (P stalk in eukaryotes), and show 
changes in conformation throughout the bacterial translation 82–84. Information on 
their conformations and functions during eukaryotic translation remains limited as 
their high intrinsic mobilities have interfered with structural studies, however they are 
thought to act similar to their bacterial counterparts. The L1 stalk, formed by the 
protein L1 and H76-78 in both bacteria and eukaryotes, becomes especially mobile 
after peptidyl transfer and upon rotation, moving towards the CP, and locks in this 
position upon EF-G binding 78. After GTP hydrolysis and simultaneously with 
translocation, it goes back to its original position, “pulling” the P/E tRNA to the E 
site as it does so, or possibly out of the ribosome since the original position is 
permissive for bacterial E site tRNA release 78. L7/L12 stalk is formed by H33-34 
(GAC) at its base, the long α helix of L10 bolstered by L7/L12 NTD dimers at its 
middle, and L7/L12 CTDs at the tip (bound to the NTD dimers by long, disordered 
flexible linkers) 83. A similar structure is formed by the acidic phosphoproteins P0, 
P1, P2 in the eukaryotic ribosomes, according to low resolution structural studies. 
The CTDs in the resulting structure are very flexible and are thought to capture 
elongation factors, as L7/L12 stalk forms part of the factor binding site. This stalk 
also has been observed to move towards CP upon TC binding 78,83. Another important 
large subunit motion occurs during accommodation. H92, or the “A loop,” switches 
from a conformation that impedes tRNA accommodation to the PTC to a permissive 
  31 
 
position, resulting in the opening of a 3D gate, formed by U2492 (U2861 in yeast) of 
H89, and C2556 and C2573 (C2925 and C2942 in yeast) of H92 83,85,86. This gate, and 
the whole AC, a corridor of 20 conserved rRNA bases spanning H89-H93 that 
interact with the tRNA during accommodation, “closes” again after the tRNA CCA 
end has completely moved into the PTC. 
1.5. Experimental approaches related to the ribosome structure and 
function 
 
The last five to six decades have seen major breakthroughs in the research of 
translation and ribosomes, allowing us to have an atomic resolution view of the static 
ribosome and to identify many steps and components of translation 87. The advent and 
commercialization of electron microscopy and ultracentrifugation fueled this research 
area in the 1950’s. Biochemical and molecular biology approaches, made possible by 
the use of radioactive labeling, from 1950’s to 2000’s resulted in the discovery of 
tRNA and mRNA, elucidation of the genetic code and the protein composition of the 
ribosome, reconstitution of bacterial ribosomes in vitro, and the sequencing of the 
large rRNAs. Biophysical approaches yielded important results as well. In the late 
1960’s, small angle X-ray and neutron scattering techniques (SAXS and SANS, 
respectively) allowed the overall shapes of the ribosome and its subunits to be 
observed, enabling generation of a “neutron map” of the positions of the proteins in 
the small subunit. This is also when Förster resonance energy transfer (FRET) started 
to be used in this research area to locate ribosomal proteins. The use of FRET began 
to flourish in the mid-2000’s, and has been enhanced by the recent development of 
single molecule techniques. In the mid-1970’s and 1980’s, advances in electron 
  32 
 
microscopy made it possible to identify the shapes and interactions of the ribosomal 
subunits, the locations of specific proteins and resulted in the discovery of the peptide 
exit tunnel. These advances also laid the foundation of the single particle 
reconstruction techniques of today. Cryogenic electron microscopy (cryo-EM), 
developed in mid-1990’s and still used today with major improvements primarily 
driven by advances in computational technologies, opened the door to structural 
dynamics studies. 1980’s and 1990’s also saw the extensive use of chemical and 
enzymatic structure probing methods, resulting in the identification of A, P, E sites on 
the rRNA, as well as the hybrid states of tRNA binding. The use of nuclear magnetic 
resonance (NMR) has been helpful in analyses of rRNA fragments and ribosomal 
protein / RNA fragments. The application of X-ray crystallography to the field began 
with crystals of isolated ribosomal proteins, and moved to the whole ribosome in the 
1970’s when the first useful, but poorly diffracting crystals of ribosomes were 
obtained. The diffraction resolution continually increased. The incorporation of the 
phases of the electromagnetic waves, and not just the intensities, resulted in a 
breakthrough in the field in 1998, resulting in the publication of the first atomic 
resolution structures of the ribosomal subunits in 2000. Overall this method 
contributed immensely to the determination of the atomic structure of ribosomes from 
all domains of life.  
Today the static structures of the 70S and 80S ribosome are well known 
thanks to continued improvements to X-ray crystallography, while cryo-EM and 
single molecule studies continue to be promising to study ribosomal dynamics. 
Additional visualization of dynamics is made possible by molecular dynamics (MD) 
  33 
 
simulations, especially when coupled to cryo-EM, single-molecule FRET (smFRET), 
and highthroughput versions of chemical probing techniques. The following 
subsections will focus on the key methods to study ribosome structure and dynamics. 
1.5.1. X-ray crystallography 
 
Different forms of electromagnetic radiation are useful to visualize objects, 
requiring the wavelength of the radiation to be comparable to the smallest features to 
be resolved. In biological molecules, the smallest features of interest are atomic bond 
lengths (1-2 Å in length), rendering X-rays suitable. X-rays may be produced by in-
house rotating anodes, or by synchrotrons. Upon interaction of X-ray waves with the 
electrons of matter, the electrons become secondary sources of electromagnetic 
radiation, scattering the incident radiation. Scattering from one molecule is generally 
too weak to measure, hence the successful use of X-rays has first been described on 
crystals, which contain millions of ordered and repeating molecules. X-ray scattering 
off the regularly repeating assembly of molecules in a crystal allows reconstruction of 
electron densities, finally arriving at the distribution of the atoms 82. 
X-ray crystallography is undoubtedly a very powerful technique especially for 
de novo structure determination, as it does not have a size limitation, as long as well-
diffracting crystals may be prepared. The structures of many proteins, including the 
Nobel prize winning hemoglobin, have been discovered by this technique 88. 
However, and even though the X-ray structure of the ribosome were also awarded the 
Nobel prize, obtaining well-diffracting crystals, especially as they pertain to 
ribosomes, is a difficult process, and requires very homogeneous samples. 
Homogeneity may be achieved by employing thermophilic bacteria and archaea, 
  34 
 
whose ribosomes are highly stable and resistant to degradation 82. Obtaining 
homogeneous ribosomes from eukaryotes is still challenging, as the only known  
method in use is glucose starvation of yeast 82. This causes ribosomes to be “locked” 
with the Stm1 protein, a suppressor of ribosome activity 2. While glucose starvation 
stabilizes yeast ribosomes in a single, homogenous population, these ribosomes 
cannot be used to make complexes because their ligand binding sites are occupied by 
Stm1 and they are in a permanent “rotated” conformation 2. Furthermore, X-ray 
crystallography is not suitable for studies of dynamics, in any domain of life, as 
molecules may behave differently in crystals and in solution, and more than one 
diffraction experiment on separate subsets of homogeneous ribosomes would be 
required to really get a sense of the change in the “motion” of molecules. 
1.5.2. Cryo-EM 
 
Cryogenic electron microscopy (cryo-EM) is another structure determination 
method that is suitable for use with dynamic systems. A dilute solution of a biological 
complex is flash frozen in a thin layer of vitreous ice (e.g. using liquid ethane) on an 
electron microscopy grid, and 2-dimensional (2D) images (micrographs) are collected 
under 50,000x or more magnification by transmission electron microscopy (TEM). 
The signal to noise ratio with this method is low, resulting in low contrast images, 
requiring hundreds of thousands of images to be collected. Contrast transfer functions 
of these are calculated to explore the relationship between the object of interest and 
the contrast, and to correct for the low contrast. Particles in the micrographs are then 
classified according to orientation, averaged, and merged. Three-dimensional (3D) 
density maps are obtained upon image processing 84. 
  35 
 
The power of cryo-EM in ribosome studies comes from the natural contrast 
between the phosphate backbone and proteins 84. Cryo-EM also does not have a size 
limitation. Common image processing methods to obtain 3D information from 2D 
images include random conical tilt (RCT) and common line (CL) 84. In RCT, the 
electron micrographs of the same area of the grid are generated in tilted and untilted 
condition, giving two angles. A third angle is computed from the 2D alignment of 
chosen particle images. These three angles or views can then be used to compute an 
initial 3D volume. In CL, common lines of any two 2D averaged classes are found as 
a result of the Fourier transform of the original 3D object to compute their relative 
angles. This is suitable for a small set of atoms, as it generates an ab initio model of 
the structure. Larger datasets, with both 3D construction techniques, require a known 
model of the object of interest (such as an X-ray crystallography structure of the 
ribosome). Although cryo-EM is suitable to gather information on ribosomal 
dynamics, its resolution remains limited, and is currently being improved through 
both computational means and new camera technologies 89. Cryo-EM is also a very 
labor-intensive method, although this is currently being solved by automation of some 
processing steps, such as particle classification. 
1.5.3. SmFRET 
 
FRET stands for Förster resonance energy transfer. This type of energy 
transfer occurs between fluorescent dyes showing spectral overlap, such that the 
acceptor dye is excitable by the wavelength of photons the donor dye emits as it 
fluoresces, and is very sensitive to the inter-dye distance. Hence the acceptor 
fluoresces upon coming to close contact (20-80 Å) with the donor. When 
  36 
 
biomolecules are tagged with such FRET pairs, and the photons emitted by the 
acceptor dye detected, interrogation of interactions and dynamics become possible. 
As such, this method and the derivations of it have been very useful in analyses of 
protein conformation, protein-protein interactions and cellular localization of proteins 
78,90.  
FRET, especially in conjunction with single molecule technologies 
(smFRET), has also been extensively used in protein translation research, mostly 
pertaining to translational dynamics, as inter-dye distance can be monitored over time 
78,91. In setups with the ribosome, molecules of interest, either intrinsic or extrinsic to 
the ribosome, are labeled with FRET fluorophore pairs. This primary setup requires 
synchronization of molecules, which is challenging with multistep processes. This 
challenge has been overcome by incorporation of single molecule spectroscopy and 
microscopy (smFRET), allowing single photons to be detected at a time, which would 
result from a single FRET interaction. The ribosome is also immobilized on an 
optically transparent surface, solving the problem of fluorophore diffusion. As a 
process takes place, binding of molecules of interest (eg. tRNAs) may be detected by 
a series of fluorescence bursts and delays, allowing time and rate constants of 
dissociation and association to be resolved 78,92. 
Although smFRET is a powerful quantitative technique for interrogation of 
translational and ribosomal dynamics, it relies heavily on known structures, and 
cannot be used for discovery of novel dynamic processes. Additional limitations 
come from the maximum possible distance between the fluorophores and fluorophore 
photobleaching. Surface immobilization, although conferring some advantages, may 
  37 
 
also perturb behavior of systems, and may result in some nonspecific surface 
interactions. Currently, the method is also challenging and time consuming 78,84.   
1.5.4. MD simulations 
 
Molecular dynamics (MD) approaches computationally explore the time-
dependent changes in the positions of atoms. Simulations are achieved by first 
modelling the molecular interactions within a macromolecule. This requires 
calculation of the total “potential” force acting on each atom, through bonded or non-
bonded interactions. MD simulations are generally coupled to structural studies to be 
able to do these calculations, as well as obtaining information on the initial positions 
and velocities of all particles in the system. From this total force, the position, 
velocity and acceleration of each atom are calculated by solving for Newton’s 
equations of motion. This process is repeated after computationally moving the atoms 
to their estimated position at a time step. After many iterations of this process, the 
results add up to a trajectory of the time evolution of the system at hand. 
Defined this way, MD simulations provide a visualization option for dynamics 
studies. They have been used in conjunction with cryo-EM and smFRET, and 
facilitate “quasi atomic model” building. As described in Chapter 5: Conclusion and 
future directions, our RNA chemical modification studies are being used to help 
refine MD simulations of the yeast ribosome through the translational elongation 
cycle. 
 
 
  38 
 
1.5.5. Chemical structure probing 
 
All chemical structure probing methods have a common scheme of chemical 
treatment of RNA, which results in adduct formation on the RNA, and detection of 
the locations of such adducts by primer extension by a reverse transcriptase. 
Sequencing reactions are conducted in tandem for each region of interest. 
Conventionally, radioactively labeled primers and sequencing gels are employed for 
fragment analysis, providing single nucleotide resolution in reads of 100-200 nt long 
fragments. 
1.5.5.1. Base specific chemicals 
 
Traditional, base specific chemicals, DMS, CMCT, and kethoxal, have 
affinity for the basepairing edges of nucleotides, mostly the Watson-Crick edge. 
Hence they target non-basepaired regions of RNA molecules. The modification 
patterns obtained from treatment of RNA with these chemicals may indicate 
basepairing, tertiary contacts, or protein-RNA interactions. Due to the base specific 
nature of these chemicals, they have to be used in tandem to obtain complete 
structural information on RNA 93,94. An additional step must also be performed to 
stop the reactions with these chemicals. 
DMS (dimethyl sulfate) methylates the N-1 atom of adenosine and the N-3 of 
cytosine bases. It may modify adenines and cytosines that are single stranded, 
basepaired at the end of a helix, or in a base pair next to a GU pair 93,94. DMS has 
affinity for DNA, as well. CMCT (1-cyclohexyl-(2-morpholinoethyl)carbodiimide 
metho-p-toluene sulfonate) modifies the N-3 atom of uracil, and to a smaller extent, 
  39 
 
the N-1 of guanine. Kethoxal (β-ethoxy-α-ketobutyraldehyde) modifies the N-1 and 
N-2 atoms of guanine 93,94.   
1.5.5.2. Hydroxyl radical footprinting 
 
Another chemical probing method, hydroxyl radical footprinting, relies on the 
cleavage of unprotected regions of RNA by hydroxyl radicals. In addition to the 
general probing scheme mentioned above, this method requires the use of a system to 
continually generate such radicals. This is achieved through the Fenton reaction of 
Fe(II) EDTA with hydrogen peroxide and ascorbic acid. Hydroxyl radical 
footprinting can be used with DNA, RNA, or even proteins, as the radicals can break 
phosphodiester and peptide bonds. This method generates information on the tertiary 
structure of a molecule, specifically solvent accessibility, which makes it especially 
useful in identification of protein binding sites on nucleic acids. For such a purpose, 
tethering of the hydroxyl radical generating system on the protein of interest is 
necessary, and is commonly achieved through the use of BABE 
(bromoacetamidobenzyl-EDTA) in conjunction with reduced iron. BABE is a chelate 
labeling reagent that conjugates with sulfhydryl groups 95,96. Another way to 
constantly generate hydroxyl radicals necessary for cleavage is achieved through 
irradiation with synchrotron X-ray beams. This is especially useful to track dynamic, 
rapid processes, such as ribonucleoprotein folding, as it provides analyses on 
timescales of 50-100 milliseconds 97.   
1.5.5.3. In-line probing 
 
Another method that may be employed to probe the RNA structure, 
specifically the local nucleotide dynamics, is in-line probing. This method does not 
  40 
 
require any chemicals, and is not enzymatic, but is based on the natural tendency of 
RNA to degrade differentially according to its structure. Briefly, single stranded RNA 
is able to sample through a range of conformations, one of which is the “in-line” 
conformation of the 2’-oxygen, the phosphorus, and the adjacent 5’-oxygen. In this 
conformation, the 2’-oxygen acts as a nucleophile in the intramolecular displacement 
of the adjacent 5’-oxygen and cleaves the RNA linkage. In contrast to the flexible 
single stranded regions, the linkages in the highly structured regions of the RNA will 
be much less prone to this sort of cleavage 98.  
1.5.5.4. SHAPE 
 
SHAPE, selective 2’-hydroxyl acylation analyzed by primer extension, is a 
chemical probing method that provides for both structure and dynamics interrogation. 
The scheme is very similar to other chemical probing methods, however more than 
one chemical is suitable for SHAPE. The most important common property of 
SHAPE chemicals is that they react with 2’-hydroxyl groups, hence modifying all 
RNA nucleotides with equal specificity. Another common property is that they are 
self-quenching. Examples of these chemicals are N-methylisatoic anhydride (NMIA, 
half-life: 8.3 min at 37°C), 1-methyl-7-nitroisatoic anhydride (1M7, half-life: 14 sec 
at 37°C), and benzoyl cyanide (BzCN, half-life: 0.2 sec at 37°C) 99,100. Unless the 2’-
hydroxyl group of a nucleotide is constrained through basepairing, stacking or other 
interactions, the SHAPE chemical binds to the nucleotide (Figure 15), forming an 
adduct that cannot be resolved during primer extension by a reverse transcriptase. 
This results in fragments of varying lengths. If radioactively labeled primers are used, 
the fragments can be visualized on gels, and essentially single nucleotide resolution is 
  41 
 
obtained. The use of non-base specific chemicals, and the 2’-hydroxyl reactivity 
essentially depending on local flexibility, gives an edge to SHAPE method compared 
to other chemical probing techniques such that it can be employed to obtain 
information on dynamics of all nucleotides on the RNA of interest. The SHAPE 
chemicals also have high dynamic ranges, hence the data obtained are highly 
quantifiable. There have been further improvements to this method (please see 
below). 
 
Figure 15. SHAPE reaction. 
The ribose 2’-hydroxyl group attacks the carbonyl carbon of the SHAPE chemical, shown here is 1M7. 
This attack depends on the vicinity of the neighboring phosphate group, as its negative charge does not 
allow stable binding of the chemical to the 2’-hydroxyl group. After electron rearrangements, part of 
the anhydride chemical breaks off as carbon dioxide and the RNA nucleotide is covalently bound to an 
adduct. Modified from Merino et al, 2005 and Mortimer and Weeks, 2007 100,101. 
 
1.5.5.4.1. SHAPE chemistry 
During the course of development of SHAPE method, it has been shown that 
the acylation of 2’ position selectively in single stranded RNA nucleotides does not 
rely on solvent accessibility or electrostatic factors, but on local nucleotide flexibility, 
as the single stranded RNA tends to assume more conformations than RNA duplexes 
102. In other studies, the chemical reactivity of 2’-hydroxyl has found to be modulated 
by proximity of adjacent 3’-phosphodiester anion 101. According to this, the 2’-
hydroxyl reactivity of AMP is the highest, followed by, in decreasing order, cyclic 
  42 
 
AMP, ATP, 3’-deoxy-ATP, 3’-O-methyl-ATP, adenosine 3’-ethyl phosphate (pAp-
ethyl; unstructured RNA analogue with 2’-hydroxyl and 3’-phosphodiester 
monoanion) and pAp-ddC (dinucleotide analogue). 
 
Figure 16. Mechanisms of RNA SHAPE chemistry. 
A. Hyper-reactive nucleotides from McGinnis et al are superimposed to point out to the fact that the 
nonbridging oxygen groups of the backbone (and the permanent charge) are directed away from the 2′-
OH group, and the 2’-oxyanion is stabilized this way. B. General base catalysis examples are shown. 
These include catalysis by the pyrimidine O2 (upper left), purine N3 (upper right), pro-S oxygen 
(lower left), and alkaline solvent (lower right). For each panel, the number of occurrences observed in 
SHAPE analysis of 16S rRNA crystals is given in parentheses. Phosphate group oxygens are thought 
to have “pro-chirality” as substitution of one results in S chirality (“pro-S”) and the other in R chirality 
(“pro-R”) 103. Image modified from McGinnis et al, 2012 104. 
 
An attempt to show which conformations of the 2’-hydroxyl and the 
neighboring positions are permissive to chemical binding comes from a study in 
which SHAPE has been performed on 16S rRNA crystals 104. The hyper-reactive 
nucleotides resulting from in-crystal SHAPE have mostly fallen into one or more of 
these categories: A) long PO-to-2’-OH distance (nucleotides mostly in C3’-endo 
ribose conformation), B) short 2’-OH to pyrimidine O2 / purine N3 distance 
(nucleotides only in C2’-endo ribose conformation), and C) short 2’-OH to non-
bridging O distance (nucleotides mostly in C2’-endo ribose conformation). The first 
category is in accordance with previous studies 101, while the latter two indicate that 
  43 
 
the 2’-hydroxyl reactivity may be modulated by the pyrimidine O2 or purine N3 
group and the nonbridging phosphate oxygen, respectively. These observations and 
functional group substitution experiments from the above study indicates general base 
catalysis as a mechanism for the increased reactivity of 2’-hydroxyl with SHAPE 
chemicals (Figure 16). 
1.5.5.4.2. Highthroughput SHAPE (hSHAPE) 
 
Figure 17. Scheme of highthroughput selective 2’-hydroxyl acylation analyzed by primer 
extension. 
A. The RNA of interest is treated with a hydroxyl-selective electrophile, 1-methyl-7-nitroisatoic 
anhydride in our experiments. The chemical binds to the 2’-hydroxyl groups of nucleotides based on 
local flexibility. B. Unconstrained regions of the RNA, such as bulges, loops and single stranded 
stretches interact with the SHAPE chemical, resulting in 2’-O-adducts (red knobs). C. These adducts 
cannot be resolved by the reverse transcriptase during primer extension, using fluorescently labeled 
primers (blue with yellow stars). D. Four primer extension reactions are conducted per primer region, 
separated by the fluorescent tag. One tube contains the chemical treated RNA, another contains the 
vehicle control RNA (DMSO treated), and the remaining two contain untreated RNA and are for 
sequencing reactions. For rRNA hSHAPE, we have 20 primer sets and use dideoxynucleotides 
(ddNTPs) corresponding to the two most common nucleotides of a region for sequencing reactions 105. 
E. The fragments resulting from primer extension are resolved by capillary electrophoresis and the 
  44 
 
electropherograms are analyzed by ShapeFinder. This analysis entails correct alignment of the 
corresponding chemical treated and vehicle control peaks with the RNA sequence. F. Integrated peak 
areas (area under chemical treated peak minus area under the corresponding control peak) obtained 
from ShapeFinder are put through statistical analyses and mapped onto 2D and 3D maps of the 
ribosome. Image modified from Vasa et al, 2008 106. 
 
As with all chemical probing techniques, visualization of fragments obtained 
from SHAPE was carried out by gel electrophoresis until recently. Today, this 
powerful technique is coupled to capillary electrophoresis, making it high throughput 
(hence “hSHAPE”) since the read lengths have increased from 100-200 nts to 400-
600 nts. Fluorescently labeled primers are also employed, eliminating the need for 
radioactively labeled materials. These improvements enhance the accuracy of 
quantitative analyses related to hSHAPE, make it easier, faster and more automated. 
The hSHAPE scheme is shown in Figure 17. The methodology we use is described in 
detail in Chapter 6. Other improvements to the SHAPE technique are listed in 
Chapter 2.  
1.6. Project rationale and aims 
 
The extensive use of the biophysical techniques discussed above has resulted 
in an accurate view of the static structure of the ribosome and the steps in translation; 
yet uncovering the dynamic nature of this complex machinery and the allosteric 
communication pathways within it remains a challenge. While providing valuable and 
accurate structure information, X-ray crystallography is limited by the formation of 
high quality crystals. Cryo-electron microscopy, while having provided insight into 
different conformational states of the ribosome, is still limited by resolution. Single 
molecule FRET has been especially useful in determining the tRNA movements 
during translation, but is limited by the maximum possible distance between the 
  45 
 
fluorescent probes. Chemical probing of rRNA structure has the advantage of 
simplicity and can provide a good understanding of how the rRNA structure changes. 
However, the traditional chemicals used in this approach have limitations such as 
high noise, difficult optimization, the need for reaction quenching and  low dynamic 
range. These and other technical details have limited the use of traditional chemical 
probing technology to probing of only short stretches (about 100 nucleotides at a 
time) of rRNA.  
Since relatively small changes in the secondary and tertiary structure of rRNA 
might allow transmission of information between different regions of the ribosome, 
probing of rRNA structure with an enhanced technique can give valuable insight into 
ribosome dynamics. hSHAPE is a high-throughput technique with single nucleotide 
resolution, and can provide data on both overall rRNA dynamics and changes in 
individual functional sites. It is not base specific in contrast to other chemical probing 
methods and targets the 2'-OH of all RNA nucleotides. The chemical used in this 
work, 1-methyl-7-nitroisatoic anhydride (1M7), has a high dynamic range, is self-
quenching and yields high signal to noise ratios, making the method easy to optimize 
100. After application of this technique on different elongation complexes and through 
computational modeling, the allosteric communication pathways throughout the 
ribosome can be uncovered in the context of the translation elongation cycle. Such 
knowledge will ultimately contribute tremendously to our understanding of the 
ribosome and structure/function relationships. 
In this context, this work aims to improve hSHAPE analyses for use with 
eukaryotic rRNA, and to interrogate rRNA dynamics using this method in initiation 
  46 
 
and elongation complexes, with the ultimate aim of uncovering the binding site of 
eukaryotic initiation factor 5 (eIF5) on the SSU, and generating a complete, dynamic 
view of the translation elongation cycle. 
  47 
 
Chapter 2: The evolution of hSHAPE analyses 
 
2.1. Background and rationale 
 
Ribonucleic acid (RNA) is a complex biomolecule that mediates, catalyzes, 
and regulates many processes during gene expression. The catalytic roles of RNA can 
be illuminated by analyses of an RNA molecule’s complex tertiary structures, which 
can be determined by X-ray crystallography and nuclear magnetic resonance 
spectroscopy. For molecules without available atomic resolution structures, secondary 
structure prediction is typically the first step toward elucidating functional higher 
order interactions. Single stranded RNA forms numerous secondary structures 
including duplexes, hairpins, bulges, internal loops and junctions. These can be 
predicted and probed by computational (phylogenetic comparative sequence analysis 
or covariation analysis), and biochemical approaches (enzymatic and chemical 
structure probing) respectively. Biochemical methods are based on the partially 
selective interaction of small molecules or ribonucleases with single stranded, 
unstructured RNA regions and detection of these interactions by primer extension to 
probe local RNA structure, flexibility and solvent accessibility 8,107,108. 
Selective 2’-hydroxyl acylation analyzed by primer extension (SHAPE) is a 
chemical probing method in which a hydroxyl selective, self-quenching electrophile 
is employed to interrogate local RNA flexibility 109. When coupled with primer 
extension protocols employing fluorescently labeled primers and capillary 
electrophoresis 110 (thus rendering it high throughput - hSHAPE), reads in the range 
of 400-600 nt long can be generated with single nucleotide resolution. Data may then 
  48 
 
be analyzed using ShapeFinder 111, and processed through fitted baseline adjustment, 
mobility shift correction, signal decay correction, scaling and peak alignment. The 
output consists of an integrated peak value, or the reactivity value, for each 
nucleotide. Overall, the technique yields high signal to noise ratios and has a high 
dynamic range. Modifications to the method include employing electrophiles 
improved for a shorter reaction time 100,112, coupling of hSHAPE with multiplexed 
deep sequencing 113,114, or the use of a two-capillary system 115, all of which increase 
the strength of the technique. 
There have been major improvements to the process of RNA structure 
determination by SHAPE.  These include improving the algorithms associated with 
RNA secondary structure determination 116, improving the signal decay correction 
associated with primer extension electropherogram analyses 117,118, and merging 
mutational analyses with SHAPE chemistry to arrive at the most reliable secondary 
and tertiary structure predictions in the absence of X-ray crystallography data 119–121. 
What remains common to all these studies is the method of normalization, which 
comprises of box-plot analysis to remove peaks greater than 1.5 times the 
interquartile range above the 75th quartile and division of all SHAPE reactivities by 
the average of the highest 10% of intensities after removal of outliers 116,122. This 
approach loses a considerable amount of information and does not lend itself well to 
visualization of complex structures such as ribosomal RNA (rRNA) in the context of 
the ribosome. One attempt at improving the normalization of SHAPE reactivities 
involves separately normalizing the background and signals by their corresponding 
median values before background subtraction using a series of algorithms named Fast 
  49 
 
Analysis of SHAPE Traces (FAST) 123. However this does not overcome the skewed 
data distributions associated with application of SHAPE on eukaryotic rRNA, which 
interferes with normalization and causes loss of data. 
One of the strengths of SHAPE chemistry is that its high dynamic range has 
the potential to retain high resolution information pertaining to the entropic 
environment of individual ribobases.  Such information has a broad range of 
applications, including providing additional restraints upon molecular dynamics 
simulations, nucleotide resolution mapping of intra- and inter molecular complexes, 
and visualization of changes in entropic landscapes in macromolecules.  As noted 
above, the problem is that current analytical methods lose much of this information.  
Here, we report a new method for normalization and analysis of hSHAPE data that 
maximizes information retention while minimizing background noise. The testbed for 
this method is the yeast ribosome, for which atomic resolution structures are available 
but for which the molecular dynamics remains incomplete. Application of this 
method to a well-defined series of yeast ribosome complexes reveals both previously 
known and unknown changes in rRNA entropic states that can be simply visualized. 
2.2. Results 
2.2.1. hSHAPE analyses v1.0 
 
Our original hSHAPE analysis method, based mainly on previous methods 110, 
employs the raw reactivities or median normalized reactivities of rRNA nucleotides. 
Raw reactivities refer to the integrated peak values obtained from ShapeFinder after 
correction of negative values to “zero”s. Median normalized reactivities are generated 
by dividing the raw reactivities by the median value for a primer region. Both types of 
  50 
 
data are then categorized into ranges based on the mean, the median, the mean plus 
one standard deviation and the mean plus two standard deviations prior to the merge 
of all primer regions together. The ranges are labeled as reactivity levels from 1 to 4, 
which can then be color coded and shown on secondary and tertiary rRNA structure 
maps. These reactivity levels are also used to compare data from different rRNAs, 
obtained from different ribosomal complexes in the context of this work, by simply 
taking the difference of the levels on a per nucleotide basis.  
 
Figure 18. Box plot analyses indicate extremely skewed distribution of data.  
A. Box plot analysis of the raw reactivity values from hSHAPE experiments. B. Box plot analysis of 
the median normalized values from hSHAPE experiments. E1 and E2 show the data from two runs of 
the empty small subunit, and Ave indicate the average of the two. 
 
Evaluation of the results of these processes by box plot analyses indicate 
extremely skewed data, with a very large number of outliers above the maximum 
(Figure 18). Frequency histograms of the raw reactivities and median normalized 
  51 
 
reactivities were also generated to further observe the data distribution, confirm the 
right skew of the data and the presence of extremely high outliers (Figure 19). These 
results illuminate the necessity to improve the data analysis step of hSHAPE. 
 
Figure 19. Frequency histograms also show extremely skewed distribution of data.  
A-C. Histogram of the raw reactivity values from hSHAPE experiments. D-F. Box plot analysis of the 
median normalized values from hSHAPE experiments. E1 and E2 show the data from two runs of the 
empty small subunit, and Ave indicate the average of the two. 
2.2.2. hSHAPE analyses v2.0 
 
Examination of the data shown in Figure 18 and Figure 19 revealed that the 
unimodal and right-skewed distributions resemble lognormal data. Data 
transformation is a common application in genome-wide data analyses 124,125, 
however has not been reported in hSHAPE data analyses before. Thus, to improve the 
analyses scheme, the raw reactivity data were transformed by the natural log (ln) 
  52 
 
before further processing, separately for each primer region. The resulting box-plot 
and histogram represent a tremendous improvement from the previous approach 
(Figure 20A and Figure 21A-C), revealing that the non-zero data adopted a normal 
distribution.  
 
Figure 20. Box plot analyses show the data acquire a unimodal, symmetric distribution after 
natural log (ln) transformation and median normalization.  
A. Box plot analysis of the ln transformed values from hSHAPE experiments. B. Box plot analysis of 
the ln transformed and median normalized values from hSHAPE experiments. E1 and E2 show the 
data from two runs of the empty small subunit, and Ave indicate the average of the two. 
 
 
Next, median values for each primer region were employed to normalize the 
ln transformed data. The ln transformed value for each nucleotide was divided by the 
median of the region, and the data were merged. Box plot and histogram analyses 
show that the data distributions were very similar to that of ln transformed values, 
maintaining the normal distribution of non-zero data (Figure 20B and Figure 21D-F). 
  53 
 
With these new analyses, assignment of reactivity levels for visualization purposes is 
performed following comparison of transformed, normalized values from different 
rRNAs, or different ribosomal complexes in the context of this work. For this 
comparison, the difference of the transformed, normalized values from different 
complexes is divided by the transformed, normalized value of the empty ribosome, on 
a per nucleotide basis. (This division step accounts for the overall reactivity of the 
ribosome.) 
 
Figure 21. Frequency histograms also show improvement in data distribution upon ln 
transformation.  
A-C. Histogram of the ln transformed values from hSHAPE experiments. D-F. Box plot analysis of the 
ln transformed and median normalized values from hSHAPE experiments. E1 and E2 show the data 
from two runs of the empty small subunit, and Ave indicate the average of the two. 
 
 
For visualization purposes, a color code is employed that represents ranges of 
comparative reactivity levels. 15 levels of reactivity are chosen by grouping the data 
generated from comparison calculations, based on the frequency histograms. 
  54 
 
Frequency intervals are used to determine the relative reactivity levels. A 
representative histogram is shown in Figure 22A. Level “0” (indicating no change 
between two complexes) comprises most of the data points (highest bars, roughly in 
the range of -0.25 to 0.25 for all complexes), followed by levels “-1” and “1.” The 
most extreme levels “-7” and “7” (indicating a large decrease in reactivity and a high 
increase in reactivity, respectively, in one complex compared to the other) include the 
smallest number of data points. All other levels include approximately the same 
number of data points. The resulting color scale is shown in Figure 22B in the order 
of relative levels, colors, and comparative reactivity ranges.  
 
 
 
Figure 22. Difference calculations and data visualization. 
A. Data distribution of the reactivity difference between posttranslocation and pretranslocation 
complex 18S rRNAs. For more information on the complexes, please refer to Chapters 4 and 6.  B. The 
above data distribution was used to derive the color code shown here to be used in hSHAPE 2D and 
3D difference maps. 
 
  55 
 
The same color scale is maintained in the difference maps of complexes 
versus the empty ribosome as well, even though the data distribution of the results of 
the comparison calculations differ (Figure 23). As the reactivity of a nucleotide might 
decline to zero in a complex, reactivity level “-1” forms a natural limit to complex 
versus empty ribosome calculations, resulting in a non-normal data distribution after 
this comparison. We maintained the color code from comparisons of two complexes 
(Figure 22B), except level “-1” was assigned the color for level “-7”; to be able to 
visually compare and contrast difference maps of any nature and to indicate that level 
“-1” resembles the largest decrease in reactivity in certain cases (i.e. difference maps 
of complexes vs. empty ribosomes). 
 
Figure 23. Data distribution of the reactivity difference between the P site occupied complex and 
the empty ribosome 18S rRNAs. 
 
2.2.3. Application of method to ribosome complexes and discussion 
 
hSHAPE is a powerful, highly quantifiable method to detect changes in RNA 
flexibility. It is easy to optimize, conduct and does not require specialized machinery. 
The signal to noise ratio is also very high with hSHAPE chemicals compared to other 
RNA structure probes. However, it should be noted that the use of fluorescently 
  56 
 
labeled primers, while supplying an avenue to make the technique highthroughput 
and highly automated, introduces noise during primer extension. The effect of the 
noise increases non-linearly as the data sets become larger, such as with the 5000+ 
nucleotides of eukaryotic ribosomal RNA. To overcome the data distribution 
problems observed with hSHAPE and improve normalization for comparison of data 
from rRNAs of different eukaryotic ribosomal complexes, we suggest adding a 
simple natural log transformation step to the analyses.  
 
Figure 24. Representative results from the two statistical approaches.  
The three dimensional structure is from Ben-Shem et al, 2011.2 (A-B) The difference maps of 
posttranslocation and pretranslocation complexes A. after the first version of statistical analyses, B. 
after the updated statistical analyses. (C-D) The difference maps of P site occupied complex and the 
empty ribosome C. after the first version of statistical analyses, D. after the updated statistical 
analyses.  The color code for the first version of analyses is shown on the left, for the updated analyses 
  57 
 
on the right. For more information on the complexes, please refer to Chapter 4: Walking the ribosome 
through the translation elongation cycle and Chapter 6: Materials and methods. 
 
 
Figure 24 compares the outputs from hSHAPE analyses before and after 
natural log transformation. The sample data are from the comparisons of various 
ribosomal complexes. Results with the first version of analyses are shown in panels A 
and C. These images are “noisy,” and give the impression that the reactivities of 
almost all nucleotides have changed, albeit slightly, between the complexes of 
interest. However when the results from the improved analyses are examined (panels 
B and D), it is evident that large changes were masked by smaller changes in 
reactivity in the results of the previous analyses. The changes in reactivity localize to 
specific regions of the ribosome after the improved analyses are applied to the data, 
allowing more complex interpretations of the results. 
Figure 25 focuses on the P stalk region in the difference map of 
posttranslocation vs pretranslocation complex, and compares the effects of the two 
analyses in this region. The updated method (B) has cleared up many of the “slight 
loss of reactivity” regions (indicated with green), which are present in abundance 
after the original analyses are applied (A). The updated analyses have also rendered 
some intersubunit bridge regions visible: A slight protection is now visible at the tip 
of H38 (B1a bridge), and protected nucleotides are visible in H69 and h44 (B2a 
bridge). Other stretches of reactivity decrease or increase are also now visible 
(unmarked), and should be examined further. While eukaryotic structures related to 
translocation are still under study by various groups 26,49,126,127, deprotection of the P 
stalk region and protection of intersubunit bridges are expected changes in the 
  58 
 
ribosomes after translocation. Overall, the new analyses highlight the “true” reactivity 
changes and remove noise from results. 
 
Figure 25. The P stalk in images generated by the two statistical approaches. 
The P stalk area in the difference map of posttranslocation vs. pretranslocation complex is shown after 
A. the first version of analyses, B. the updated analyses. In this view the whole A site area is visible. P 
  59 
 
stalk (red), H38 (dark blue), H69-h44 (light blue), H89, PTC and AC are labeled. The reactivity ladder 
belongs to the updated visualization, however warm colors indicate increase in reactivity and cool 
colors indicate decrease in reactivity in both approaches. Unchanged reactivity is indicated by the color 
black in the first version of analyses. PTC: peptidyl transferase center, AC: accommodation corridor, 
GAC: GTPase associated center. 
 
 
While the improvements described here may be of great utility for ribosomal 
dynamics studies as shown in Figure 24 and Figure 25, users of hSHAPE are advised 
to test other transformation methods on their data, especially given that the data are 
“zero-heavy”; the original hSHAPE values are not of lognormal but of “delta” 
lognormal distribution 128,129. In the context of this work, the zero values were chosen 
to be included in the data processing, as they may be biologically relevant and 
indicating regions of no reactivity. Ln(x+c) transformation, where c = 1 in this work, 
was useful for our purposes, as 99% of the raw reactivities obtained from hSHAPE 
experiments are approximately 102 to 1.5 x 106 times this constant. Along these lines, 
when data transformation was applied to median normalized reactivities instead of 
raw reactivities, the problems with data distribution persisted, as most of these values 
are only 0.5 to 500 times of the constant (data not shown). 
Overall, hSHAPE quantification of large RNA dynamics can be improved 
through data transformation, and increases the amount of information extracted from 
ribosomal RNA studies. Improved hSHAPE analyses then may be applied to 
interrogation of rRNA flexibility in translation initiation and elongation complexes. 
 
  60 
 
Chapter 3: hSHAPE analyses of translation initiation 
complexes 
3.1. Background and rationale 
 
As discussed in Chapter 1, eukaryotic translation initiation is complex, 
beginning with mRNA activation and 43S preinitiation complex (PIC) formation. 43S 
PIC is then recruited to the mRNA, and after further rearrangements, scanning and 
AUG recognition, translation starts. There are still many unknowns, including the 
structural dynamics of 43S PIC formation. Our current understanding of this process, 
based on structural, biochemical and yeast genetics approaches are summarized in 
Figure 26. 
 
Figure 26. Model summarizing the interactions between different initiation factors and the 
structural rearrangements of the initiation complex. 
(i) eIF1 and eIF1A binding rearranges the 40S to an open conformation which is capable of binding the 
ternary complex (TC) composed of eIF2, initiator tRNA and GTP. eIF1A CTT (green) monitors the P 
site and interacts with eIF1, while NTT (red) interacts with eIF1. The CTT has structural elements that 
have been shown to favor of scanning (SE: scanning enhancer). (ii) After eIF5 and mRNA binding, 
scanning starts. The complex is still in the open conformation, and the initiator tRNA is not fully at the 
P site – this is caused by the SE elements of eIF1A. eIF5 may be recruited through interactions of its 
NTD (5N) with eIF2, and its CTD (5C) with the NTT of eIF2 and the CTT of eIF1A. GTP hydrolysis 
occurs but inorganic phosphate (Pi) is not released. (iii) When the start codon is encountered, in a 
complex coordination of events, Pi is released possibly due to a dramatic change in eIF5 binding, 
allowing tRNA anticodon: start codon interactions and leading to displacement of eIF1A CTT. The 
change in eIF5 binding leads to eIF1 release as well and to the assumption of a closed conformation by 
  61 
 
the 40S subunig. The closed conformation is not permissive to scanning. The displaced eIF1A CTT 
now interacts with 5N instead of 5C. Image modified from Saini et al, 2014 130. 
 
Cryo-EM structures of 40S bound with eIF1 and eIF1A have been reported 
131–133. The latest reported resolution is at 3.7Å. These show that eIF1 and eIF1A bind 
to different sides of h44 close to tRNA binding sites. eIF1A binds to the A site (40S 
shoulder); in fact, one of the roles of this factor is to prevent a tRNA from binding to 
the A site. eIF1A has unstructured amino and carboxyl terminal tails (NTT and CTT), 
and its CTT reaches and occludes the P site as well. The eIF1A-CTT may enable 
communication between this factor and eIF1, as eIF1 binds near the P site (40S 
platform), in accordance with its role in regulating the full localization of the initiator 
tRNA ASL into the P site. In the structures of 40S bound with eIF1 and eIF1A, gross 
conformational rearrangements are observed in comparison to empty 40S and to 48S 
PIC after start codon recognition 132–135. The head/body/platform junction appears to 
be “open” in these structures, as well as the “latch” between h18 and h34 which 
clamps around mRNA upon start codon recognition. These data all indicate that the 
mRNA entry channel assumes an “open” conformation that allows the 43S PIC to 
scan along the mRNA in search of a start codon. 
Other components of 43S PIC are eIF3, the ternary complex (TC = eIF2-GTP- 
Met-tRNAi
Met) and eIF5. eIF3 is a very large, multisubunit complex, and recent X-ray 
crystallography structures of the yeast eIF3, coupled with cryo-EM reconstruction, 
cross-linking, mass spectrometry, and structure modeling, indicate that it clamps 
around the 40S subunit, perhaps coordinating the correct positioning of other 
initiation factors, including eIF1 and eIF1A 136. Given that eIF3 is an essential 
element of the reconstituted yeast translation initiation system 19, it may possibly 
  62 
 
function as more than an anti-association factor. However, assembly of 48S PIC in 
the absence of eIF3 has also been reported seemingly with no defects 137, thus 
providing insight into start codon recognition and the related structural 
rearrangements. Additionally, there are structural differences between yeast and 
mammalian eIF3 and the mammalian eIF3 structures have not been mapped to the 
predicted locations of eIF1, eIF5-CTD or eIF2 134,138. 
Initiator TC recruitment occurs through the interaction of eIF2α subunit with 
40S, possibly through ribosomal protein uS7, and is enhanced by the other initiation 
factors of the 43S PIC. In yeast, eIF3 is thought to be in a position to monitor TC 
binding 134. Overall, there are a plethora of interactions between the initiation factors 
that make up the 43S PIC that stabilize the complex and which may possibly 
enhances TC binding. These include (1) eIF5-CTD (carboxyl terminal domain) 
interactions with the eIF2β subunit-NTT, the eIF3c subunit-NTD (amino terminal 
domain), and eIF1; (2) eIF1 interactions with the eIF2β subunit-NTT and the eIF3c-
NTD; and (3) eIF3a subunit-CTD interactions with the eIF2β subunit 134. 
eIF5 is perhaps the most elusive of the initiation factors that are thought to 
form the 43S PIC. It is the GTPase activating protein (GAP) of eIF2, and is known to 
function in GTP hydrolysis by eIF2, regulation of inorganic phosphate release and 
stimulation of the closed conformation of 30S upon start codon recognition 80,130,139; 
hence its functions are relatively well known on the 43S-mRNA and 48S PIC’s. Its 
role in the 43S PIC seems to be to stabilize the complex through its interactions with 
other factors (mostly eIF2). However its exact binding location on the ribosome is not 
known and has not been reported. Biochemical assays indicate direct binding between 
  63 
 
eIF5 (and / or its NTD/CTD) and SSU 19,140. The latest cryo-EM structure of the 48S 
PIC indicates an additional density on the platform which is suspected to be eIF5 137. 
The GAP catalytic site of eIF5 is located on its NTD, while the CTD includes a 
HEAT domain shown to interact with eIF1, eIF2β and eIF3c-NTD. These two 
domains are held together by a linker region. There is no available structure of the 
complete eIF5, but the structures of its NTD and CTD are known 141,142. Current 
known and estimated structural information on eIF5 is summarized in Figure 27. 
 
Figure 27. Possible location of eIF5 and the structure of the individual eIF5 domains. 
A. 48S PIC structure and the densities identified on it. Most of the densities could be attributed to 
relevant factors (eIF1 - light blue, eIF1A - dark blue, eIF2 - orange, purple, red) or tRNA (green) or 
mRNA (bright pink). An unidentified density is marked in pale pink and this may belong to eIF5. This 
density is better seen on the right. h44 density has been labeled in black on the left. B. eIF5-NTD 
structure. R15 (red) is the catalytic residue for GAP function. Some other functionally important amino 
acids are also shown. C. eIF5-CTD structure. These eight helices form an expanded HEAT domain. 
  64 
 
Aromatic and acidic amino acid boxes, AA1 (grey) and AA2 (red) are important for binding to eIF3 
and eIF2. Images modified from Hussain et al, 2014, Conte et al, 2006, and Bieniossek et al, 2006 
137,141,142. 
 
Our aim in this project was to use hSHAPE to map eIF5 onto different 
preinitiation complexes. We started with complexes of 40S with eIF1, eIF1A, with or 
without eIF5, as eIF1 and 1A are known to stably interact with 40S regardless of the 
presence or absence of other factors, thus forming an appropriate baseline for 
subsequent analyses. After discovering a potential footprint for eIF5 using this 
approach, we worked with complexes of 40S with eIF5 or eIF5-NTD or eIF5-CTD, 
obtaining different sets of protected regions. Our analyses suggest that eIF5 binding 
appears to change in the presence and absence of eIF1 and eIF1A, perhaps indicating 
that the initial eIF5 binding to the 40S requires guidance from other factors. 
3.2. Results 
 
 
Figure 28. Building initiation complexes.  
A. Difference map of the eIF1•eIF1A complex minus the empty SSU, B. Difference map of the 
eIF1•eIF1A•eIF5 complex minus the empty SSU, C. Difference map of the eIF1•eIF1A•eIF5 complex 
minus the eIF1•eIF1A complex, indicating the reactivity changes caused by eIF5. The view is from the 
  65 
 
intersubunit face of the eukaryotic rRNA. The eIF1 (yellow-orange circle) and eIF1A (yellow ellipse) 
related changes are shown. The areas inside the black rectangles lose reactivity upon eIF5 binding. The 
green ellipse marks the “latch” area between h18 (shoulder) and h34 (head). There is a reactivity 
decrease on the h18 side upon eIF5 binding. The tertiary structure is from Ben-Shem et al, 20112. 
 
To obtain a footprint of the eukaryotic initiation factor 5, structural analyses 
were performed on different initiation complexes of the yeast ribosome. These 
complexes were assembled by Fan Zhang at NIH. The first set of complexes to be 
probed were ribosomal small subunits of the yeast bound with eIF1, eIF1A, and/or 
eIF5. Results of these are shown in Figure 28 and Figure 29, from the intersubunit 
face and the solvent side, respectively. As can be seen in the panel A of Figure 28, 
binding of eIF1 and eIF1A causes flexibility changes in different places of the 18S 
rRNA. The known eIF1A binding site (on the neck, towards the platform) appears to 
have lost flexibility. The result of eIF1 binding, although not very clear in 40S-eIF1-
eIF1A complex (Figure 28A), is also visible (on the neck, towards the shoulder). 
Addition of eIF5 causes protection in the head domain, the neck, at the back of the 
shoulder, as well as further protection on the shoulder. The changes in the neck 
domain, and the tips of helices 18 and 34 (the so-called “mRNA latch”) may be 
indicative of assumption of a closed conformation by the 40S subunit (Figure 28B 
and C). 
  66 
 
 
Figure 29. Building initiation complexes.  
A. Difference map of the eIF1•eIF1A complex minus the empty SSU. B. Difference map of the 
eIF1•eIF1A•eIF5 complex minus the empty SSU. C. Difference map of the eIF1•eIF1A•eIF5 complex 
minus the eIF1•eIF1A complex, indicating the reactivity changes caused by eIF5. The view is from the 
solvent exposed side of the eukaryotic rRNA. eIF1 (yellow-orange circle) and eIF1A (yellow circle) 
related changes are shown. The reactivity of the area inside the black rectangle decreases upon eIF5 
binding. The tertiary structure is from Ben-Shem et al, 2011.2 
 
We repeated hSHAPE with complexes of 40S with only full length eIF5, 
eIF5-NTD or eIF5-CTD, as eIF5 can bind 40S on its own in vitro 80. While we 
anticipated that this would provide a more accurate view of their respective binding 
sites, the protected regions observed for these domains were different than those 
observed with the previous set of complexes (the locations of black rectangles vs. 
blue ellipses in Figure 30 and Figure 31). The decreased reactivity in the head domain 
appears to have moved towards the top of this structure, accompanied with some 
additional loss of reactivity in the beak region (not marked in the figures). This loss of 
reactivity is seen more clearly in the full length eIF5 complex, rather than with only 
NTD or only CTD, with CTD reflecting more of this reactivity change than NTD. 
The reactivity change in the beak is more visible in the eIF5-CTD complex. A new 
  67 
 
region of reactivity decrease is visible in the middle of the body domain (specifically 
on h44) in the full length eIF5 complex, compared to vacant 40S. This may be a 
protection pattern caused by the physical binding of the eIF5-NTD (Figure 30A and B 
vs. C). Another new region of reactivity decrease is visible towards the back of the 
shoulder and right below the beak (specifically helices 16 and 17). This pattern is 
strongest with the full length eIF5 complex and to some degree in the eIF5-CTD 
complex. The mRNA latch (at least the h18 component, marked by the green ellipse 
in Figure 30) also has decreased reactivity in the full length eIF5 complex. 
 
Figure 30. Difference maps of eIF5 complexes and empty SSU. 
A. Difference map of full length eIF5 minus empty SSU, B. Difference map of the N terminus of eIF5 
minus empty SSU, C. Difference map of the C terminus of eIF5 minus empty SSU. The view is from 
the intersubunit face of the eukaryotic rRNA. The black rectangles are used to show where protected 
regions were seen with the previous set of complexes. The regions of reactivity decrease from the 
current set of complexes are indicated with blue ellipses. The “latch” is indicated by a green ellipse. 
The tertiary structure is from Ben-Shem et al, 2011.2 
 
  68 
 
 
Figure 31. Difference maps of eIF5 complexes and empty SSU. 
A. Difference map of full length eIF5 minus empty SSU, B. Difference map of the N terminal domain 
of eIF5 minus empty SSU, C. Difference map of the C terminal domain of eIF5 minus empty SSU. 
The view is from the solvent exposed side of eukaryotic rRNA. The black rectangle marks the place of 
the loss of reactivity seen in the previous set of complexes. The thick, blue ellipse indicates the loss of 
reactivity observed with the current set of complexes. The thin, blue ellipse marks the h16-17 loss of 
reactivity. The “latch” is marked with a green ellipse. The tertiary structure is from Ben-Shem et al, 
2011.2 
 
The location of decreased reactivity is observed at the solvent exposed face of 
40S is also changed (the black rectangle vs. the thick, blue ellipse in Figure 31). 
Decreased reactivity is observed in h21 (part of expansion segment 6 in eukaryotes), 
to a greater extent with the complex containing full length eIF5, and to a lesser extent 
with the eIF5-CTD containing complex. Decreased reactivity in the h16-17 region is 
visible in this view as well, especially in the eIF5-CTD containing complex. Both 
components of the mRNA latch (h18-h34) show decreased reactivity in the presence 
of the full length eIF5 complex in this view. 
  69 
 
 
Figure 32. Difference maps of the full length eIF5 complex and NTD or CTD portion complexes. 
A. Difference map of the full length eIF5 complex minus the eIF5-NTD complex, indicating the 
reactivity changes caused by the C terminal domain, B. Difference map of the full length eIF5 complex 
minus the eIF5-CTD complex, indicating the reactivity changes caused by the N terminal domain. The 
view is from the intersubunit face of the tertiary structure of eukaryotic rRNA at 3Å resolution. The 
latch (green ellipse), and the regions of reactivity decrease from h16 (dark red ellipse), h17 (dark red 
circle), and h44 (blue ellipse) are marked and show differences in reactivity between the two maps.  
 
 
Figure 33. Difference maps of the full length eIF5 complex and NTD or CTD portion complexes. 
  70 
 
A. Difference map of the full length eIF5 complex minus the eIF5-NTD complex, indicating the 
reactivity changes caused by the C terminal domain, B. Difference map of the full length eIF5 complex 
minus the eIF5-CTD complex, indicating the reactivity changes caused by the N terminal domain. The 
region of loss of reactivity from h21 is indicated with a blue ellipse. A region implicated in the 
previous set of complexes is shown with a black rectangle. The view is from the solvent exposed side 
of the tertiary structure of eukaryotic rRNA at 3Å resolution. 
 
We generated difference maps of the complex containing the full length eIF5 
vs. complexes with either eIF5-NTD or eIF5-CTD in an effort to observe the 
reactivity changes caused by specific domains (Figure 32 and Figure 33). In these 
maps, the differences between eIF5 vs eIF5-NTD should indicate changes caused by 
the CTD, and vice versa. We observed reactivity differences between eIF5 minus 
eIF5-NTD and eIF5 minus eIF5-CTD in the same regions discussed above. For 
example, components of the latch (green ellipse in Figure 32A vs B) show greater 
degrees of chemical reactivity loss in the eIF5 minus eIF5-NTD map than the eIF5 
minus eIF5-CTD map, implicating the eIF5-CTD in this rearrangement. In contrast, 
the protected area from h44 (blue ellipse in Figure 32B vs A) is more extensive in the 
eIF5 minus eIF5-CTD difference map, implicating the eIF5-NTD as the cause. 
Additionally, the two eIF5 domains seem to affect helices 16 and 17 in an opposite 
manner (dark red ellipse and circle in Figure 32A vs B), with the CTD domain 
possibly affecting h16 more (Figure 32A), and the NTD affecting h17 more (Figure 
32B). Another interesting observation is that some loss of reactivity on the platform is 
visible in Figure 32B (full length eIF5 vs CTD map - unmarked). This is the region 
we initially identified a possible footprint for eIF5, and may be of importance. 
The h21 region shows slight differences between NTD vs CTD as well. eIF5-
CTD may cause a greater extend to decrease in chemical reactivity in this region 
(Figure 33A vs B). The back of the beak loses reactivity in the presence of the eIF5-
CTD as well (Figure 33A - unmarked). In the eIF5 vs eIF5-CTD map, decreased 
  71 
 
reactivity is also observed in a region we identified with the previous set of 
complexes (Figure 33B - black rectangle). The previous regions of reactivity loss are 
only visible in difference maps of full length eIF5 vs eIF5-CTD complexes, 
indicating the effect of the eIF5-NTD on these regions (Figure 32B and Figure 33B). 
3.3. Discussion 
 
Eukaryotic translation initiation consists of 43S PIC formation, activated 
mRNA recruitment to the PIC, and scanning for a start codon in optimal context. One 
of the components of the 43S PIC is the eukaryotic initiation factor 5. The functions 
of eIF5 after mRNA recruitment, during scanning and start codon recognition have 
been well documented and most importantly relate to its interaction with eIF2 / 
initiation ternary complex (TC). Binding assays indicate a direct association between 
eIF5 and the ribosomal small subunit, however a binding site has not been described, 
and a common hypothesis is that it binds through initiation TC. The conformational 
changes that it is thought to induce would require interactions with the SSU directly 
as well, especially upon start codon recognition. In this work, we report a possible 
change in the binding site of eIF5 depending on the presence of other IF’s (i.e. eIF1 
and 1A), and find support for eIF5-induced conformational changes. 
The hSHAPE results obtained with the 40S-1-1A and 40S-1-1A-5 complexes 
indicate two regions of decreased chemical reactivity located on the intersubunit face 
of 40S upon eIF5 binding: one in the head domain and the other on the platform. 
These are in accordance with theories on the location of eIF5, and the interacting 
eIF1A tail domain 80,130,139. Additionally, the reactivity change observed in the 
“mRNA latch” in the same set of complexes may imply that the initiator TC binds to 
  72 
 
open up the mRNA entry tunnel, thus enabling mRNA binding and consecutive 
scanning to take place. 
hSHAPE results on 40S-eIF5 complexes without additional IF’s show 
different regions of decreased chemical reactivity. The reactivity decrease on the head 
domain changes slightly and decreased reactivity is also seen in the middle of the 
body domain, specifically on h44. Another “new” region of reactivity decrease is 
visible towards the back of the shoulder and right below the beak (specifically helices 
16 and 17), mostly in the full length eIF5 complex and to some degree in the eIF5-
CTD complex. This is interesting as, induced by eIF1 and eIF1A, h16 is able to adopt 
“open” (extending away from 40S body) and “closed” (bending towards 40S beak) 
conformations that are thought to regulate mRNA entry and scanning 1. We do not 
obtain directional information from hSHAPE, however, stabilization of h16 is still 
thought provoking. The results further indicate that the “mRNA latch” may be closed 
in the full length eIF5 complex. Together with stabilization of h16, the results may 
indicate a “closed” conformation of the 40S. This would be compatible with the 
known eIF5 functions during eukaryotic initiation, i.e. that is it is a regulator of 
scanning, and may imply that the entire repertoire of eIF’s are needed on the 43S PIC 
to form the “open” conformation, balancing out the effect of eIF5. In fact, based on 
the results from the two sets of complexes, eIF1 and 1A may also be functioning to 
correctly position eIF5, eventually (with eIF2-TC in place for example) resulting in 
the 40S open conformation. Viewed as a whole, this may ensure the correct order of 
binding events. 
  73 
 
Another, less exciting, explanation of the differences observed between the 
two sets of complexes might simply be that eIF5 does not bind as efficiently on 40S 
without eIF1 or 1A present. In fact, the dissociation constants for eIF5, eIF5-NTD 
and eIF5-CTD are quite high: 300 ± 50 nM, 250 ± 30 nM, and 1500 ± 500 nM, 
respectively 80. These, in turn, imply either a high koff value or a low kon value, which 
would influence the length of time available for 1M7 reagent to react with the 
associated complexes. This may also explain the unclear effects of the individual 
domains alone (especially of CTD, as it has the highest dissociation constant). 
While both explanations above are viable, another region of decreased 
reactivity located on the solvent side of the 40S subunit seems to support the former. 
h21 (part of expansion segment 6 in eukaryotes) shows decreased hSHAPE reactivity, 
mostly in the complex with full length eIF5 and to a lesser extent with eIF5-CTD. 
This is interesting as structural rearrangements have been observed in this area during 
transition from 40S to 43S, and are thought be associated with a rotation in the 
platform-body region 143, compatible with a closed conformation and regulating 
mRNA recruitment and scanning. 
A preliminary model is shown in Figure 34, implicating eIF5 as a regulator of 
40S rearrangements, and being led to its correct binding site by the eIF1A tail. 
Experiments with mutant eIF5 and the interacting eIF1A tail, as well as other 
structure probing approaches, will be necessary to test this model. Nonetheless, 
through hSHAPE studies, we are able to show reactivity changes induced by eIF5 in 
agreement with major rearrangements of 40S, possibly towards the closed 
conformation. 
  74 
 
 
Figure 34. Possible eIF5 binding sites. 
A. When eIF1 (green) and 1A (red) are bound to the ribosome, eIF5 (yellow) is “guided” to a site on 
the platform, possibly through interactions with eIF1A CTT. In the actual 43S PIC, interactions with 
eIF2-TC are also crucial. eIF1A CTT interacts with eIF5 NTD, hence the platform may be primarily 
the binding site of this domain, while eIF5 CTD may be binding to the head (possibly through 
interactions with eIF1A NTT 80). eIF5 binding is accompanied by the “closing” of the mRNA latch 
(small rectangle). B. eIF5, having low affinity for the empty 40S, does not bind to its site on the 
platform without presence of eIF1 and 1A. Instead it possibly binds to a site in the middle of h44 
(blue). This is accompanied by rearrangements resembling closed complex formation: mRNA latch 
closure, h16 stabilization, body-platform rotation. 
 
 
  75 
 
Chapter 4: Walking the ribosome through the translation 
elongation cycle 
4.1. Background and rationale 
 
Translation elongation is a very dynamic process that involves both large and 
small scale structural rearrangements (please see Sections 1.2.2 and 1.4 for details). 
Most of these rearrangements have been well described in the catalytic sites of the 
ribosome, especially in bacteria. Eukaryotic studies are underway but are hindered by 
the resolution of the current methodologies (please see Section 1.5 for details). While 
large scale, low resolution studies indicate that translation elongation occurs is very 
similar in prokaryotes and eukaryotes, recent studies indicate some differences, 
including potential differences in tRNA binding sites and intersubunit motions 
26,131,127. Added to these are the facts that eukaryotic ribosomes are larger and have 
additional elements (e.g. expansion segments on rRNA). The observation of “subunit 
rolling” in mammalian ribosomes is especially intriguing 26, as it may indicate 
presence of communication pathways within eukaryotic ribosomes beyond the usual 
functional sites. In sum, these differences highlight the need for further studies on the 
eukaryotic elongation cycle and how ribosome structural dynamics map to each step 
of this process.  
Given that the rRNA harbors important elements including the 
peptidyltransferase center, the decoding center and a large fraction of the intersubunit 
bridges, interrogation of rRNA dynamics could be beneficial to the study of 
eukaryotic translation elongation and for identifying potential functions of the “non-
catalytic” regions of the ribosome. The budding yeast, Saccharomyces cerevisiae, 
  76 
 
provides a powerful model system to assemble and study different eukaryotic 
ribosomal elongation complexes. As discussed in detail in Chapters 1 and 2, 
improved hSHAPE analyses have proven useful for detecting and visualizing changes 
in rRNA flexibility among different ribosomal complexes. Here we report the 
assembly and hSHAPE probing of seven ribosomal complexes that occur during the 
elongation cycle and discuss the important implications of these results. Previously 
described functional regions and intersubunit bridges are mentioned in detail for the 
rest of Section 4.1, as well as what is already known on eukaryote-specific elements. 
Section 4.2 starts with results from complex verification, and goes on to compare the 
hSHAPE data to known functional regions and intersubunit bridges, as well as to 
analyze eukaryote-specific elements. We conclude our analyses with a discussion on 
a network of communication pathways identified in our data. 
4.1.1. Known functional regions of bacterial and eukaryotic rRNAs 
 
There are a number of well-established paradigms upon which our work can 
be built. The first is based on classical studies from the Noller group in which 
bacterial ribosomes were subjected to chemical probing and hydroxyl radical 
cleavage studies 144–149. These provide a strong foundation for identifying pre-
“landmark” rRNA nucleotides whose protection patterns change upon binding of 
mRNA, tRNA or other transacting factors to the ribosome. The second are the 
numerous structural studies, including X-ray crystallographic and cryo-EM, of both 
bacterial and eukaryotic ribosomes in complex with the substrates and accessory 
proteins 84,135,143,150,151. 
  77 
 
4.1.1.1. mRNA - rRNA interactions 
 
The landmark mRNA/16S rRNA interactions in prokaryotic ribosomes are: 
U1381 (C1618 in yeast), C1395 (C1632 in yeast), G413 (A485 in yeast), U421 (U494 
in yeast), G424 (G497 in yeast), A532 (A579 in yeast), G693 (G904 in yeast), U723 
(C934 in yeast), A845 (U1060 in yeast), G1131 (U1362 in yeast), C1132 (U1363 in 
yeast), G1300 (C1537 in yeast), G1338 (G1575 in yeast). Such studies also pre-pre-
dicted mRNA binding sites on the 23S RNA: U1065 (A1240 in yeast; close to GAC) 
and U887 (A1025 in yeast; close to doman II) 152. Structural studies of mammalian 
ribosomes and initiation complexes identified an mRNA cleft and latch on 18S 
rRNA, formed by the nucleotides C565, C575-U578, G904, U999, G1150-A1152, 
C1190-U1191, C1274, A1427, C1620-U1621, C1634-C1641, A1756-U1758, G1760-
A1763, A1765-U1770, A1794 (all yeast numbers) 84,135,143,150. 
4.1.1.2. tRNA - rRNA interactions 
 
The rRNA nucleotides protected upon tRNA binding have been studied 
extensively in bacterial ribosomes 147,148,153–155. The chemical probing results on these 
are summarized in Table 1. These studies have also indicated the presence of hybrid 
states of tRNAs 147,148,154. 
In the early chemical probing and hydroxyl radical cleavage studies, binding 
of tRNAs to different sites on bacterial ribosomes were controlled by magnesium 
concentrations and the presence or absence of an mRNA template. In these 
conditions, the nucleotides taking part in E site binding were acylated or cleaved 
regardless of presence of mRNA. Hydroxyl radical cleavage experiments of the LSU 
identified nucleotides U1859-G1862 (A2219-A2222 in yeast), C1881-G1884 
  78 
 
(A2224-C2227 in yeast), U2074-76 (U2416-G2418 in yeast), C2395-U2401 (C2765-
C2773 in yeast), G2421-C2424 (G2793-G2796 in yeast), U2431-G2436 (G2800-
G2805 in yeast) as residing in the E site. mRNA-dependent cleavages occurring 
under low Mg++ and mRNA-independent cleavages under high Mg++ indicated P site 
related interactions. These P site nucleotides in the LSU were: C1941-C1947 (C2284-
C2290 in yeast), G1954-U1956 (U2297-A2299 in yeast), G1959-A1966 (G2302-
A2309 in yeast), G2251-U2257 (G2619-A2626 in yeast), U2492-G2494 (U2861-
G2863 in yeast), C2594-U2596 (C2963-U2965 in yeast), C2601-G2603 (C2970-
G2972 in yeast). Strictly mRNA dependent cleavages were thought to be related to 
the A site. These A-site nucleotides in the LSU were: U1940-A1970 (G2283-A2313 
in yeast), U2555-U2561 (U2924-A2930 in yeast), A2566-C2573 (U2935-C2942 in 
yeast), A2600 region (A2969 region in yeast), G2659-G2664 (G3026-G3031 in yeast) 
149. 
 
16S 18S 23S 25S 
A 
site 
G529, G530, A1492, 
A1493, (A1408, 
G1494), A892(↑), 
G1405(↑), A532(↑), 
G926(↑), G1338(↑) 
G577-78, C1646, 
A1755-57, A580(↑), 
U1117(↑), A1151(↑), 
U1643(↑), A1576(↑) 
C2254, A2439, 
A2451, G2553, 
U2555, A2602, 
G1041, C1941, 
G1068, G1071, 
U2609, (A1916, 
A1918) 
A1217, G1243, 
G1246, C2284, C2622, 
A2808, A2820, 
G2922, U2924, 
A2971, U2978, 
A2259, G2261 
P 
site 
G693, A794, C795, 
G926, G1401, 
(A532, G1338, 
A1339, G966, G529, 
G530) 
A905, A1151, 
A1005, C1006-07, 
G1638, C1639, 
G577-78, A580, 
A1576-77, C1192 
G2252, G2253, 
A2439, A2451, 
U2506, U2584, 
U2585, (A1916, 
A1918), A1926, 
G2505, A2602(↑) 
G2620-21, A2808, 
A2820, U2269, 
G2874, U2875, 
U2953-54, A2259, 
G2261, A2971(↑) 
E 
site 
  C2394, G2112, 
G2116, (A2169) 
C2764, G2454, A2458, 
A2484 
Table 1. Summary of protected and deprotected rRNA nucleotides upon tRNA binding. 
These have been gathered from chemical probing studies on bacterial rRNA. Please see text for further 
detail, as well as tRNA binding sites emerging from hydroxyl radical cleavage studies and/or other 
structural studies. CCA-end dependent protections are shown in italics. EF-Tu dissociation dependent 
  79 
 
protections (ie. A site tRNA CCA-end / acceptor stem dependent protections) are shown in grey. 
Deprotections are indicated with an “up arrow” in parentheses. Weakly protected nucleotides are in 
parentheses. Yeast nucleotides corresponding to the bacterial counterparts are also listed. 
 
Structural studies on mammalian translation initiation complexes describe the 
SSU P site as nucleotides C1000-A1001, G1169-A1171, C1190-C1192, A1194-
C1195, C1461-G1466, G1574-U1579, G1622, A1635-C1637 of 18S rRNA (all yeast 
numbers) 135. A cryo-EM study of the mammalian pre-translocation complex 
identified nucleotides G577, G1179, C1274, A1756, G1757 as residing in the SSU A 
site; A1001, G1002, C1192, G1462, C1463, G1575, A1576, G1638, C1639, C1759, 
G1760 as constituting the SSU P site (or P/E); and G904-A906, C1162, A1577, 
U1578 as SSU E site bases. The same study identified nucleotides G1178, A1179, 
A1193, G1194, G2414-U2417, A2443, C2444, U2835, C2836, U2996, G2997, 
U3010 in the LSU A site (or A/P); U2408-U2410, U2423, A2424, U2771, C2772, 
G2778, A2779, A2813, G2814, U2861-G2863, U2978, U2979, A3129, A3130 in the 
LSU P site; and A2368-G2370, G2393, G2394, G2614, G2615, A2643 for the LSU E 
site (or P/E) 127. 
4.1.1.3. Elongation factor - rRNA interactions 
 
According to hydroxyl radical probing studies with bacterial ribosomes, the 
main nucleotides of the small subunit rRNA that are in close proximity with EF-G 
(eEF2 in eukaryotes) are:  A33-G39 (U29-U35 in yeast), A160 (A156 in yeast), 
A440-G445 (G510-A515 in yeast), A495-C500 (A542-G548 in yeast), G537-G539 
(A585-C587 in yeast), G790 (G1002 in yeast), U954 (C1180 in yeast), U1211-A1212 
(U1443-A1444 in yeast), C1227-G1230 (C1461-G1464 in yeast), C1398-G1400 
(C1636-G1638 in yeast) 156. 
  80 
 
The same study shows that the sarcin-ricin stem-loop in domain IV (C2646-
G2674 in bacteria; U3013-U3041 in yeast) and the region around A1070 (A1245 in 
yeast) interact with EF-G in the large subunit rRNA. Other EF-G interacting bases 
identified in this study include G2470-A2471 (G2839-C2840 in yeast), U2479-A2482 
(G2848-A2851 in yeast), C1920-G1922 (C2263-C2265 in yeast), C1965 (C2308 in 
yeast), U1065-A1070 (A1240-A1245 in yeast), A1090-C1100 (U1265-C1275 in 
yeast), C2649-G2659 (A3016-G3026 in yeast), C2667-U2672 (C3034-C3039 in 
yeast) 156.  
LSU rRNA protection sites of EF-G reported in a chemical probing study of 
bacterial ribosomes are as follows: A1067 (G1242 in yeast), A1069 (A1244 in yeast), 
G2655 (G3022 in yeast), A2660 (A3027 in yeast), G2661 (G3028 in yeast) 146. The 
same study identified the following nucleotides as involved in EF-Tu (eEF1A in 
eukaryotes) dependent protections of LSU rRNA: G2655 (G3022 in yeast), G2661 
(G3028 in yeast), A2660 (A3027 in yeast) and A2665 (A3032 in yeast) 146. No SSU 
nucleotides are implicated in the chemical probing studies 146, hence the SSU 
nucleotides listed above may only be found in close proximity to, but not in direct 
contact with EF-G. 
4.1.2. Changes in intersubunit bridges 
 
Another topic of interest in the field are the locations of points of contact 
between the SSU and the LSU, i.e. the intersubunit bridges. These are emerging as 
functional centers as well, as translocation involves movement of the subunits relative 
to one another, resulting in structural changes the intersubunit bridges as the ribosome 
transits the elongation cycle. The intersubunit bridges have been identified and 
  81 
 
characterized using many techniques, including crosslinking, chemical probing, 
mutagenesis, cryo-EM and X-ray crystallography 2,157–162. Here all previously known 
intersubunit bridges are summarized. 
In the nonrotated prokaryotic ribosome 161, the B1a bridge forms by the 
interaction of uS13 and nucleotides A886-C888 (G1024-A1026 in yeast) located at 
the tip of H38. In the rotated yeast ribosome 2, it forms by the interaction of 18S 
nucleotides U1239 and U1240, and 25S nucleotide A1025. The B1a bridge is broken 
in the hyper-rotated conformation 2. 
B1b/c bridge forms by the interaction of uS13 and uL5 in both prokaryotes 161 
and yeast 2, although in yeast uS19 also contributes to the formation of this bridge. In 
the hyper-rotated conformation of the yeast ribosome 2, different residues from the 
same proteins take part in the bridge formation. 
The B2a intersubunit bridge is composed entirely of RNA:RNA interactions. 
In bacteria 161, it is formed by  SSU nucleotides C1408-C1410, U1494-C1495 
(C1646-A1648, U1758-C1759 in yeast, respectively) located in h44, and by the LSU 
H69 nucleotides A1913-C1914, A1918 (A2256-C2257, G2261 in yeast, respectively). 
In yeast 2, the interactions are more extensive but still mostly span h44 and H69 (18S: 
U1004, U1643-C1646, U1758-C1759, G1780; and 25S: A2255, C2257-2259, A2262-
U2264). These interactions do not appear to be altered in the hyper-rotated ribosomes 
2. 
B2b is also an RNA:RNA bridge. In bacteria 161, it forms by (1) the 
interactions between h24 nucleotides G784-G785, C794 (U996-G997, C1006 in 
yeast) and H67, H69 nucleotides C1836-C1837, G1922 (C2195-C2196, C2265 in 
  82 
 
yeast), and (2) the interactions between h45 nucleotides G1516-C1519 (G1780-
C1783 in yeast) and H69, H71 nucleotides A1919-C1920, A1932 (A2262-C2263, 
A2275 in yeast). In yeast 2, the same helices take part in formation of this bridge, but 
the nucleotides change slightly: G994-U996, U1779-A1781 of 18S interact with 
C2195-C2196, G2272, A2275 of 25S. These interactions change further in the hyper-
rotated form 2: U996, U1779-A1781 of 18S interact with C2196-C2197, U2274-
A2275 of 25S. 
B2c is another RNA:RNA bridge. In bacteria 161, h24 and h27 nucleotides 
G770-U771, A900-G901 (U982-A983, A1125-G1126 in yeast) interact with H67 
nucleotides C1832-C1833 (U2191-C2192 in yeast). These slightly change in the 
rotated yeast ribosomes 2: U981-G984 of 18S interact with C2151, U2191-C2192 of 
25S; further changes were observed in the hyper-rotated conformation 2: U982-A983 
of 18S interact with U2191 of 25S. 
In bacteria 161, B3, also an RNA:RNA bridge, forms from the interactions of 
h44 nucleotides U1484-G1486 (G1748-A1750 in yeast) and H71 nucleotides C1947-
G1948, A1960-C1961 (C2290-A2291, A2303-C2304 in yeast). In yeast 2, the same 
helices take part in the arrangement of this bridge, except the interactions are much 
more extensive: A1655-U1657, A1659, A1746-A1749 of 18S interact with A1922, 
G2124-A2126, C2290-U2292, U2294, U2301-A2303, G2305 of 25S. This is 
unchanged in hyper-rotated ribosomes 2. This and observations from rotated 
prokaryotic ribosomes, have led to the hypothesis that B3 acts as a pivot during 
rotation 2,163. However, this is challenged by the latest cryo-EM studies of mammalian 
ribosomes 26. 
  83 
 
In prokaryotes 161, the B4 intersubunit bridge is formed by (1) the interactions 
of h20 nucleotides C763-G764 (C975-G976 in yeast) with H34 nucleotides C717-
A718 (A848-C849 in yeast), and (2) the interactions of uS15 with the same H34 
nucleotides. This changes considerably in the rotated yeast ribosomes 2: G628, A630, 
A971-A973 of 18S interact with A846-A847 of 25S, and the proteins eL30 and uS15 
interact. Furthermore, residues from uS15 also interact with nucleotide A847 of 25S. 
In the hyper-rotated ribosomes 2, uS15 interacts with G844-A847 of 25S rRNA, and 
G628, G972-A973 of 18S interact with A846-48 of 25S. 
B5 is an extensive bridge in prokaryotes 161. It forms by the interactions of (1) 
h44 nucleotides G1418-U1419 (U1656-U1657 in yeast) with H64 nucleotides C1768-
U1769 (A2126-U2127 in yeast), (2) h44 nucleotides G1420-22 (G1658-A1660 in 
yeast) with uL14, and (3) h44 nucleotides G1474-U1476 (U1738-A1740 in yeast) 
with H62 nucleotides A1689-90 (A1921-22 in yeast) and H64 nucleotide G1989 
(A2332 in yeast). In yeast 2, B5 forms by the interactions of C411, U1734 of 18S with 
uL14, and the interaction of A1667 of 18S with G1935 of 25S rRNA. In hyper-
rotated ribosomes 2, A1667 of 18S also interacts with A1936 of 25S. 
In nonrotated prokaryotic ribosomes 161, B6 forms through interactions 
between h44 nucleotides A1429-30, G1474-U1476 (A1667-G1668, U1738-A1740 in 
yeast) and H62 nucleotides A1689-90, G1702-A1705 (A1921-22, G1934-U1937 in 
yeast), and the interaction of G1431 of 16S (U1669 in yeast) with bL19. In rotated 
yeast ribosomes 2, bL19, which does not have a eukaryotic homologue, is replaced by 
eL24, and the 18S nucleotides that take place in this bridge are G1670 and U1724 
  84 
 
only. This bridge remains unchanged in hyper-rotated ribosomes 2. This bridge may 
rearrange upon “subunit rolling” in mammalian ribosomes 26. 
B7a is formed by h23 nucleotides C698, G702 (C910, G913 in yeast) and H68 
nucleotides A1848-G1849, G1896 (A2208-U2209, G2239 in yeast), in the nonrotated 
prokaryotic ribosome 161. In the rotated yeast ribosome 2, U909-C910, G913 of 18S 
interact with A2207, G2239-40 of 25S. In the hyper-rotated state 2, U911, G913 of 
18S interact with U2205 of 25S to hold this bridge together. 
B7b (B7b/c in newer studies) is an RNA:protein bridge, and is formed mainly 
through uL2 interactions with h23 and h24 in both nonrotated prokaryotic and rotated 
eukaryotic ribosomes 2,161. In prokaryotes 161, the specific nucleotides that take part 
are SSU G712-13, G773-A776 (A924-G925, G985-A988 in yeast). In the rotated 
yeast ribosome 2, SSU G986-87, U1012-A1013 interact with uL2, while SSU A983 
interacts with eL43. This is changed considerably in hyper-rotated ribosomes 2: SSU 
A892, G984, G986, A988 interact with uL2 and SSU A923, U982-A983, G985, 
G1122 interact with eL43. 
B8 is the last universal bridge, and is formed by the interactions of h14 
nucleotides G345-47 (G418-A420 in yeast) with uL14, in the prokaryotic nonrotated 
ribosome 161. In the rotated yeast ribosome 2, C411-A412, and G419 interact with the 
homologous protein. In the hyper-rotated yeast ribosome 2, the interacting rRNA 
nucleotides remain unchanged, while the interacting residues from uL14 change 
slightly. This bridge may be rearranging upon “subunit rolling” in mammalian 
ribosomes 26. 
  85 
 
The eukaryote specific bridges eB8, eB11-14 have only been described in the 
rotated structures of the eukaryotic ribosome 2,162. eB8 is formed by the interactions 
of eS1 with nucleotides A2536 and U2537 of 25S rRNA. These change in hyper-
rotated ribosomes to interactions between S1e and nucleotides G2533, A2535 of 25S, 
and eL43 2. This bridge may  also rearrange upon “subunit rolling” in mammalian 
ribosomes 26. eB11 forms through interactions between eS8 and the 25S nucleotides 
A2107, G3345, G3353-U3355, with changes to interactions with G3345-U3346, 
G3353-U3354 in hyper-rotated conformation 2. eB12 is defined by interactions 
between uS17and 18S nucleotides U813-G815, A850-A855 with eL19, while in the 
hyper-rotated state 2, eS7 takes the place of uS17. eB13 forms through interactions 
between eS6 and uL3 and eL24 in the rotated eukaryotic ribosomes, while the 18S 
nucleotides G273-C275, U278, U280, A1712-A1714 also interact with eL24 in the 
hyper-rotated state 2. This bridge may also rearrange upon “subunit rolling” in 
mammalian ribosomes 26. Finally, eB14 forms by the interactions between 18S 
nucleotides G1112, G1114-G1118, A1125-G1127, C1641-G1642, G1654-A1655, 
C1773-U1775, G1777-78, A1782-U1785 and eL41. Another nucleotide, A1750 is 
added to the bridge in the hyper-rotated form 2, and the interacting residues of eL41 
change slightly. 
The metazoan ribosomes may have three more bridges, eB15-eB17, indicated 
in the cryo-EM structure of the 80S Drosophila ribosome, formed by the interactions 
between extended ES31L and protein eS27 and the interactions of extended  ES27L 
and eS8 (rotated) or eS27 (nonrotated), respectively 162. 
  86 
 
4.1.3. More structural insight into eukaryotic ribosomes 
 
As previously noted in Chapter 1, there are eukaryote-specific ribosomal 
structures. Not only are there eukaryote-specific ribosomal proteins and eukaryote-
specific extensions to universal proteins, but eukaryote-specific rRNA extensions 
exist.  These are called “expansion segments”. Expansion segments and their 
structures have been studied in the context of ribosome and rRNA evolution. They are 
generally located on the solvent exposed regions of the ribosome (Figure 35). In fact, 
as the eukaryote species complexity increases, an rRNA layer is observed to form on 
the ribosome. Possible functions of some expansion segments have been defined, 
such as interacting with extraribosomal factors and structures 82,164,165,166.  
An example of a relatively well studied expansion segment is ES27L 164. This 
is a long and very flexible region of LSU, comparable in flexibility to the L1 stalk, 
and is found opposite to the P stalk / GAC. At least three conformations of this region 
have been observed through cryo-EM in yeast, wheat germ, fruit fly and human 
ribosomes. ES27L extending towards the L1 stalk is named “ES27Lin”, whereas 
extending towards the peptide exit tunnel is known as “ES27Lout”. Another 
conformation observed in wheat germ and fruit fly ribosomes is “ES27Lint” in which 
ES27L is not as extended to the L1 stalk as human ES27L is. ES27L is thought to 
regulate access of nonribosomal factors, such as chaperones, modifying enzymes, 
SRP, or the translocon, to the exit tunnel. 
  87 
 
 
Figure 35. Expansion segments in 80S yeast ribosome. 
Expansion segments are shown in red. Functional regions are shown in yellow for reference. ES: 
expansion segment, the “S” following the number denotes small subunit ES, whereas “L” denotes large 
subunit ES. PTC: peptidyl transferase center, DC: decoding center. Ribosome structure is from 
Armache et al, 2010 167. 
 
ES27L also interacts with other components of the ribosome, including 
another expansion segment, ES31L, and ribosomal proteins eL34 and eL38 164. The 
interplay between ES31L and ES27L has been observed to change depending on 
presence or absence of eEF2, i.e. the rotational status of the ribosome. 
Another expansion segment, ES7S, located on SSU, is an extension of h26 
and is part of the mRNA exit tunnel 82. Yet another, ES7L, located on LSU, has been 
observed to contact Sbp2, selenocyteine insertion sequence binding protein-2, a 
crucial component of selenocysteine protein synthesis in humans 165,166. Other than 
these the studies on expansion segments have stayed at the structural level 168–171. 
ES3S and ES6S comprise most of the extended rRNA in SSU 82. In LSU, majority of 
expansion segments are found grouped together, one such group forming behind the P 
  88 
 
stalk (ES7L, ES39L), and another behind the L1 stalk (ES19L, ES20L, ES26L, 
ES31L) 164. 
4.2. Results 
4.2.1. Complex assembly and verification 
 
 
Figure 36. The complex assembly scheme. 
The P site occupied complex (P site) is obtained by interacting empty 80S ribosomes with a peptidyl 
tRNA mimic (N-acetyl-phenylalanyl-tRNA or Ac-Phe-tRNA) and a template [polyuridylic acid or 
poly(U)]. From this complex, three others are obtained by addition of the ternary complex (TC) of 
eEF1A and phenylalanyl-tRNA (Phe-tRNA) with GDPNP (Pre-accommodation complex), or TC with 
GTP followed by an antibiotic (cycloheximide) (CHX / classical states complex), or TC with GTP only 
(Post-accommodation complex). From the Post-accommodation complex, the three final complexes are 
obtained by addition of another antibiotic (anisomycin) (ANI / hybrid states complex), or eEF2 with 
GDPNP (Pre-translocation complex), or eEF2 with GTP (Post-translocation complex). 
 
Ribosomal complexes of the elongation cycle were assembled and verified 
prior to hSHAPE probing. Seven such complexes were assembled as described in 
Figure 36 and Chapter 6: Materials and methods. These are: (1) P site occupied 
  89 
 
complex with a P site tRNA; (2) pre-accommodation complex with a P site tRNA and 
TC-GDPNP; (3) post-accommodation complex with A and P site tRNAs; (4) pre-
translocation complex with A and P site tRNAs and eEF2-GDPNP(5) post-
translocation complex with P and E site tRNAs; as well as two complexes with 
translation inhibitors (6) CHX complex with cycloheximide, A and P site tRNAs; and 
(7) ANI complex with anisomycin, A and P site tRNAs. In each of these complexes, 
the tRNA(s) should be in P/P state (P site occupied complex), P/P and A/T states 
(pre-accommodation), A/A-A/P and P/P-P/E states (post-accommodation), A/P and 
P/E states (pre-translocation), P/P and E/E states (post-translocation), A/A and P/P 
states (CHX), and A/P and P/E states (ANI). The tRNA states are also indicative of 
rotational status with “classical” A/A, P/P, E/E states occurring in nonrotated 
complexes, and “hybrid” A/P and P/E states occurring in rotated complexes. 
A series of biochemical assays were performed to verify the assembled 
complexes. The results of tRNA binding assays are shown in Figure 37. Results of 
tRNA binding assays.. P site tRNA binding results are applicable to all complexes and 
indicate P site occupancy to be 90%. Separate A site binding assays were conducted 
for pre-accommodation (employing TC-GDPNP as the A site substrate), post-
accommodation (employing TC-GTP as the A site substrate), and CHX complexes 
(incubating P site occupied complex with cycloheximide before using TC-GTP as the 
A site substrate); these indicate A site occupancy rates in the range of 55-65%. 
Results from post-accommodation complex are applicable to ANI, pre-translocation, 
and post-translocation complexes. 
  90 
 
 
Figure 37. Results of tRNA binding assays. 
P and A site binding assays were performed as described in Chapter 6: Materials and methods. The 
tRNA binding activity from different batches of ribosomes were comparable, at about (A) 85-90% for 
the P site, and (B) 55-65% for the A site. Separate A site tRNA binding reactions were performed for 
pre-accommodation, CHX and post-accommodation complexes. 
 
 
Figure 38. Results of dipeptide formation assays. 
HPLC-based dipeptide formation assays were performed as described in Chapter 6: Materials and 
methods, for further verification of pre-accommodation (left) and post-accommodation (right) 
complexes. Dipeptide formation was observed in post-accommodation but not pre-accommodation 
complex, as expected. Quantification suggests a dipeptide yield of about 50%, consistent with the 
results from binding assays. 
 
Pre-accommodation and post-accommodation complexes were further verified 
through an HPLC-based dipeptide formation assay 172, indicating whether or not 
peptidyl transfer has taken place in these complexes (Figure 38). According to this, 
pre-accommodation complex shows no significant dipeptide formation, whereas post-
accommodation complex does show dipeptide in amounts comparable to the A and P 
site tRNA occupancy rates (about 50%).  
  91 
 
The CHX and ANI complexes were further verified by base-specific structure 
probing of intersubunit bridge (B7a) to indicate their rotational statuses (Figure 39). 
Empty ribosomes are thought to freely rotate, whereas P site occupied complex is 
nonrotated, hence these were used as controls. Results from the CHX and ANI 
complexes indicate that these are nonrotated and rotated, respectively. These results 
also correlated with hSHAPE probing of this bridge. 
 
Figure 39. Rotational status of CHX and ANI complexes. 
Kethoxal probing was performed as described in Chapter 6: Materials and methods, for further 
verification of CHX and ANI complexes. G913 (18S rRNA) component of B7a bridge (top) was 
probed with guanine-specific chemical kethoxal. Lower reactivity (smaller peaks in the 
electropherograms at the bottom) indicated a nonrotated bridge, where G913 and A2207 basepair. 
Higher reactivity (larger peaks) indicated a rotated bridge, where G913 and A2208 form a stacking 
interaction. Based on this, CHX is not rotated, and ANI is, as expected. 
 
Pre-translocation and post-translocation complexes were further verified by 
(1) eEF2 binding assay, and (2) puromycin assay to detect the presence of an 
unoccupied A site (Figure 40). eEF2 was found to bind ribosomes with 70% 
efficiency, and the A site was unoccupied in 70% of post-translocation complexes, 
indicating that translocation occurred in all complexes that could bind eEF2. Since 
  92 
 
puromycin binds to an open LSU A site, the reaction was conducted at 4°C to prevent 
binding to the pre-translocation complex, in accordance with similar studies in 
bacteria 173,174. As a result, there was a clear difference between A-site occupancy of 
pre- and post-translocation complexes. 
 
Figure 40. Results of eEF2 binding and puromycin (translocation) assays. 
These assays were performed as described in Chapter 6: Materials and methods, to verify pre- and 
post-translocation complexes. A. eEF2 binding to ribosomes are about 70%. B. A site vacancy in the 
post-translocation state is about 70%, in accordance with eEF2 binding results, and higher than the pre-
translocation state. 
 
 
 
 
 
 
  93 
 
4.2.2. Identification of rRNA functional sites in hSHAPE data 
 
After assembly and biochemical verification of complexes, hSHAPE was 
performed as described in Chapter 6. Results from known functional sites, obtained 
mostly from bacterial and mammalian ribosome studies were examined. 
4.2.2.1. mRNA - rRNA interactions 
 
SSU Empty P site Average 
 
SSU Empty P site Average 
A579 1.028081 0 0.476597 
 
G577 1.399097 0 0.544874 
G904 1.508937 1.043881 1.067043 
 
G904 1.508937 1.043881 1.067043 
C1537 1.066834 0.555762 0.935147 
 
C1274 1.54062 0.929795 1.37215 
G1575 1.19643 0.836136 0.656089 
 
C1620 0.902895 0 0.953627 
C1634 1.223311 0.852858 1.385665 
 
C1634 1.223311 0.852858 1.385665 
LSU 
    
A1635 1.030116 0.540819 0.936215 
A1025 1.674609 1.007647 1.506515 
 
C1636 1.077452 0.740066 1.131201 
A1240 1.074042 0.734931 0.802454 
 
C1637 1.062668 0.549173 0.694238 
     
G1638 0.865004 0.18962 0.608664 
     
C1640 1.111845 0.657189 1.158289 
Table 2. mRNA-rRNA interactions from hSHAPE data. 
Left columns correspond to the comparison of bacterial studies to the hSHAPE data, right columns 
correspond to the comparison of mammalian studies to the hSHAPE data. Please see text for details. 
Empty: empty ribosome values, P site: P site occupied complex values, Average: average of the values 
from all complexes with an mRNA template. SSU: small subunit, LSU: large subunit. The values are 
obtained directly from complex comparison calculations (please see Chapter 6: Materials and 
methods). 
 
The reactivity values we examined in the hSHAPE data in comparison with 
the mRNA-rRNA contact sites from previous studies were either the P site occupied 
complex values or the average values of all complexes compared to empty ribosomes 
(Figure 41 and Table 2). Of the nucleotides uncovered in prokaryotic ribosome 
chemical probing studies, A579, G904, C1537, G1575 of SSU, and A1025, A1240 of 
LSU were found to be protected in both P site occupied ribosomes, and overall in all 
complexes. Of the nucleotides uncovered in mammalian ribosome structural studies, 
G577, G904, C1274, A1635, C1637, G1638 of SSU were found to be protected in 
  94 
 
both P site occupied ribosomes, and overall in all complexes. These nucleotides also 
show up in the difference map of P site occupied complex minus empty ribosome 
(Figure 41). There are also nucleotides that show protection overall, indicated by the 
average value from all complexes, compared to the empty ribosomes, but not in P site 
occupied ribosomes: C565, C575, A1427 of SSU. The mammalian ribosome 
structural studies have not uncovered any nucleotides from the LSU that might be 
affected by mRNA binding to ribosomes. 
 
Figure 41. Changes in reactivity at the mRNA binding region. 
Part of the 3D difference map of P site occupied complex vs. empty ribosome is shown above. Even 
though there are highly reactive nucleotides throughout, the reactivity of the neck region (and of the 
other mRNA interacting components) is low, possibly due to interactions with the mRNA template. 
 
4.2.2.2. tRNA - rRNA interactions 
 
First, the tRNA binding sites that emerged from bacterial chemical probing 
studies were examined in the hSHAPE data (Table 3). Protections at 18S P site 
residues are observed upon N-Ac-Phe-tRNA binding to yeast 80S. These occur at 
positions G577, A579, G904, A905, A1005, C1006, G1575, A1577, G1638, and can 
be seen in both numerical data and the difference maps. Some additional nucleotides 
show decreased reactivity values compared to empty ribosomes, but these do not 
appear in difference maps. Protection at 25S P site residues are also observed in the P 
  95 
 
site occupied complex compared to empty ribosomes. These occur at positions 
U2269, A2820, and U2954. While no changes were observed at nucleotides G2620 or 
G2621, C2622 shows strong protection in this analysis. These nucleotides are all in 
the distal tip of Helix 80 which is known as the “P site loop,” hence this may be an 
important protection. As with SSU nucleotides, some additional LSU nucleotides 
show decreased reactivity values but these are not present in difference maps. 
Trapping the ribosome before accommodation (i.e. pre-accommodation 
complex) appears to render the 18S flexible in general. Unfortunately, most of the 
18S A site (positions A1755-G1757) could not be probed due to their proximity to the 
3’ end of the SSU rRNA it is difficult to discern a clear A site signature on the small 
subunit with hSHAPE. However, it is clear that U578 is less reactive in the pre-
accommodation complex relative to the P site occupied complex suggesting the start 
of A tRNA binding. In the color maps, most 25S A site residues become slightly 
more reactive or do not show changes in reactivity, with the exception of A2259 
which becomes strongly protected compared to the P site occupied complex. A2259 
is among the nucleotides the protection of which occurs independently of eEF1A (EF-
Tu) dissociation. 
SSU Site Empty P site Pre-acc Post-acc Pre-trans Post-trans 
G577 A, P 1.399097 0 1.176051 0.440576 0 1.107741 
U578 A, P 0 0 0 1.154491 1.492553 1.409848 
A579 P 1.028081 0 1.058643 0.359615 0 0.964728 
G904 P 1.508937 1.043881 1.237881 1.070823 1.065315 0.917316 
A905 P 1.399623 1.003750 0.955063 1.320124 0.670896 1.423297 
A1005 P 0.976133 0.126593 1.218276 0.241482 0.360543 1.295051 
C1006 P 0.615169 0.193700 0 0.837512 0 0.972633 
G1575 P 1.196430 0.836136 0 0 1.434292 1.010017 
A1577 P 0.987951 0.543608 1.107272 0 0.911948 0.922447 
G1638 P 0.865004 0.189620 0.912904 0.968274 0.633609 0.338911 
U1643 A (↑) 0 0 1.503339 1.285889 1.743357 0.593165 
LSU 
 
Empty P site Pre-acc Post-acc Pre-trans Post-trans 
  96 
 
A2259 A 0 0.822130 0 0 0 0.844950 
G2261 A 1.326057 1.115180 1.479520 2.125404 1.170622 0 
U2268 P 1.656955 0.649949 1.410646 1.177815 1.363268 1.335987 
U2269 P 1.184619 0.544774 1.091499 1.439345 0 0.960000 
G2454 E 1.073323 0.902117 1.164097 1.616188 1.200907 1.114971 
A2458 E 1.564251 1.358675 1.705848 2.239336 1.440975 1.445277 
A2484 E 0.999342 0.710854 1.068599 1.511473 1.130644 1.091234 
C2622 A 1.079750 0 1.137126 1.064209 1.091083 1.074191 
C2764 E 1.878895 0.986822 1.062229 1.071432 1.027051 0.505065 
A2808 A, P 0 0 0.813131 1.177240 0 0.699794 
A2820 A, P 1.506094 0.914349 1.201215 0 1.389515 1.227199 
U2924 A 0.996548 1.900586 0.947680 1.006964 1.494715 0.998771 
U2954 P 1.591385 0 1.370176 1.901525 2.047267 1.780146 
A2971 A 1.595470 1.679157 1.528251 0 1.724749 1.658034 
U2978 A 1.427609 1.174853 1.361624 0.863808 1.345808 1.502811 
Table 3. hSHAPE data on tRNA binding sites from bacterial studies. 
Please see text for details. Empty: empty ribosome values, P site: P site occupied complex values, Pre-
acc: pre-accommodation complex values, Post-acc: post-accommodation complex values, Pre-trans: 
pre-translocation complex values, Post-trans: post-translocation complex values. SSU: small subunit, 
LSU: large subunit. The “Site” column indicates if the nucleotide belongs to the bacterial A, P, and/or 
E sites. The values are obtained directly from complex comparison calculations (please see Chapter 6: 
Materials and methods). 
 
After GTP hydrolysis, accommodation and peptidyltransfer (i.e. in the post-
accommodation complex), tRNA protection patterns become clearer, especially in the 
18S rRNA. The 18S nucleotides that show increased protection are: A1005, A1576, 
A1577, U1643, and C1646. Nucleotides G577-A579, G904, G1575, which were 
protected in the previous complexes, also remain protected in post-accommodation 
complex. The 25S nucleotides that show continued and / or enhanced protection are: 
A2259, U2875, and G2794. These were protected in the pre-accommodation complex 
as well. The nucleotides showing increased protection or novel protection compared 
to the pre-accommodation complex are A2820, A2971, and U2978. Also, some E site 
tRNA interactions can be observed at nucleotides C2212, and A2802 151. 
The pre-translocation complex, i.e. ribosomes bound with eEF2 and prior to 
GTP hydrolysis, show continued or increased protection of tRNA binding related 
nucleotides, especially in the 25S rRNA. The 25S rRNA protections occur at 
  97 
 
positions G2261, U2269, C2284, G2620, G2621, A2808, U2875, and U2953. 
Nucleotides A1217 and G1243 are also protected, and even though chemical probing 
studies list these in the A site (Table 1), other structural studies implicate them in 
transacting factor binding 33,47. Some of the listed protections are unexpected as they 
should interact with the A site tRNA more than the P site or E site tRNAs according 
to chemical probing studies (Table 1). The key A site residues A2971 and U2978 
become reactive again, with values comparable to empty ribosomes. This observation 
clashes with studies from the Noller laboratory 144,147,148, however it might also 
indicate an empty LSU A site, as would be expected. The loss of reactivity at 
nucleotides G2454, A2458, A2484, and an overall protected L1 stalk base (H76-78), 
indicates E site tRNA protections. We also observed novel or enhanced protections at 
A905, C1006, and G1638 of 18S.  
SSU 
Empty P site Pre-acc Post-acc Pre-trans Post-trans 
A site 
G577 1.399097 0 1.176051 0.440576 0 1.107741 
P or P/E 
      C1000 0.958233 0 0.995246 1.179447 1.203922 1.155918 
G1002 1.164849 0 1.379372 1.3597 1.416736 1.400864 
G1462 0 0.978274 0.970021 1.220028 1.136653 1.09194 
G1575 1.19643 0.836136 0 0 1.434292 1.010017 
G1638 0.865004 0.18962 0.912904 0.968274 0.633609 0.338911 
C1639 0 0 0 0.986505 1.051786 1.488376 
E site 
      G904 1.508937 1.043881 1.237881 1.070823 1.065315 0.917316 
A905 1.399623 1.00375 0.955063 1.320124 0.670896 0.423297 
A906 0 1.063726 0.83807 0 1.580771 0.956454 
LSU 
Empty P site Pre-acc Post-acc Pre-trans Post-trans 
A or A/P 
A1179 0.964898 0.685139 1.261124 0.426619 1.181147 0 
G2414 0.849775 0.706048 1.06291 0 1.086422 0.892657 
U2416 0.864446 0.52094 0.611094 0 0 0 
U2996 1.483941 1.664498 1.568466 0 1.628608 1.64789 
G2997 0 0 1.04262 0.840714 0.977197 0 
P site 
      U2408 0.935348 0 0.945575 1.012484 0 0.946127 
A2424 1.024822 0 0.94651 1.11736 1.019969 0.916552 
  98 
 
G2778 1.028703 0 0.32769 0.929118 0.956259 0.849439 
A2779 1.068401 0 0 1.344604 1.202072 1.20289 
A3130 1.152856 0 0 1.077206 0 1.014833 
E or P/E 
      A2368 0.851326 0.720796 0.781305 0 0.80203 0 
G2369 0.977028 0.652577 1.005705 1.412935 0 0.914007 
G2393 0.914806 0 0.868926 1.319257 0.91636 0.992166 
G2614 0.885197 1.294668 0.97423 1.295094 0.962226 1.067243 
A2643 1.007919 1.077436 0.906561 1.299254 1.010257 0.936591 
Table 4. hSHAPE data on tRNA binding sites from the mammalian ribosome. 
Empty: empty ribosome values, P site: P site occupied complex values, Pre-acc: pre-accommodation 
complex values, Post-acc: post-accommodation complex values, Pre-trans: pre-translocation complex 
values, Post-trans: post-translocation complex values. SSU: small subunit, LSU: large subunit. rRNA 
nucleotides that are thought to interact with A, P, E tRNAs, as well as tRNAs in hybrid A/P and P/E 
states are labeled. The values are obtained directly from complex comparison calculations (please see 
Chapter 6: Materials and methods). 
 
In the post-translocation complex, we observed a more open A site compared 
to the pre-translocation complex, especially in 18S. SSU nucleotides G577 and A580 
become reactive, indicative of an open A site. Also in the SSU, loss of reactivity is 
observed at G904, G1575, G1638, U1643 and C1646. These might be due to re-
established P/P state binding, or in the case of U1643 and C1646 may indicate 
rearrangements of specific intersubunit bridges. In the 25S rRNA, positions A2259, 
C2284, and A2808, as well as nucleotides from H43-44 become reactive. A2971 and 
U2978 remain reactive, indicating an open A site. The E site remains protected, and a 
novel protection is seen at C2764.  
Table 4 shows a comparison of the hSHAPE data to published mammalian 
tRNA binding sites 127. Compared to the results from bacterial sites, higher numbers 
of complete loss of reactivity are seen with mammalian sites, especially in the P site 
occupied and post-accommodation complexes. 
Examination of Table 3 and Table 4 reveal differences between bacterial and 
fungal tRNA binding sites on rRNA, and indicate that mammalian sites may be a 
better match to yeast sites, especially considering the distances of the listed 
  99 
 
nucleotides from a model tRNA. Based on this, we employed the latest cryo-EM 
structures of yeast ribosomes to visually analyze the P tRNA binding site further 175. 
The P site occupied vs empty ribosome differences were mapped onto the nonrotated 
80S ribosome, and the protected nucleotides located close to the model P site tRNA 
were examined (Figure 42). By this analysis, G2249-A2252, C2265-U2268, A2309, 
C2622, C2646, C2654-A2656, C2693-A2695, G2968-A2969 emerge as the new 
nucleotides of the 25S (LSU) P site. Some of these (A2309, C2622, G2968-A2969) 
were found to be in proximity to the P site tRNA by the hydroxyl radical cleavage 
experiments of the bacterial large subunit 149. G2620-21, A2808, A2820, U2953-54 
remain unchanged between bacteria and yeast. C1000-G1002, C1463-C1465, A1576-
U1579, C1636-C1637 emerge as the new nucleotides of the 18S (SSU) P site, and 
this is in accordance with mammalian ribosome structures 135,127. Only interactions 
with A1576-77 are conserved from bacteria. 
 
Figure 42. The yeast P site. 
The nucleotides that have decreased reactivity in the P site occupied complex compared to the empty 
ribosome were mapped to the nonrotated yeast ribosome structure 175. The model P site tRNA (grey) 
was used to locate the rRNA nucleotides that possibly interact with the tRNA. The labeled nucleotides 
in the figure are located 3-12 Å away from different parts of the tRNA (except G2677-A2679, shown 
with a transparent circle, which are 18 Å away on the average). The model mRNA is shown in tan. 
Please see text for nucleotide identities. 
 
  100 
 
Analysis of post-translocation complex vs. empty ribosome reveals 
differences with the described P site. G2250-51, C2265, G2619, A2694-96, A2808, 
A2820, A2969 in the LSU, and G1466, A1576, G1638 in the SSU remain unchanged 
between P site occupied complex and the post-translocation complex. The post-
translocation complex reveals additional decreases in hSHAPE reactivity at SSU 
C1644, U1647 and LSU A1006, C2285, C2405, G2623, A2647, U2668, A2680-
U2681, G2863. Of these C2285 and G2863 are in proximity to the P site related 
cleavages from bacterial studies, making the bacterial 50S P site still a better match 
for the yeast 60S P site than the mammalian one. All the listed nucleotides from both 
complexes are in close proximity (averaging at 10 Å) with the model P/P tRNA. 
 
4.2.2.3. Elongation factor - rRNA interactions 
 
The complexes that retain an elongation factor are the pre-accommodation 
(eEF1A) and pre-translocation (eEF2) complexes. Therefore, we examined the 
nucleotides that were found to be protected in these complexes upon elongation factor 
binding in chemical probing experiments in bacteria. In the GAC, A1244, and G3028, 
A3032 of the SRL show protection in the difference map of the pre-accommodation 
and P site occupied complexes. A3032 shows protection in the hSHAPE difference 
map of the pre-translocation and post-accommodation complexes (Table 5).  
Chemical probing 
    LSU Empty P site Pre-acc Post-acc Pre-trans Post-trans 
A1244 1.202057 0.574830 0 0 1.116745 1.106730 
G3028 0 1.145025 0.961712 0.908052 1.093411 1.195303 
A3032 0.710629 1.244453 1.042951 0.806668 0.608918 0.938253 
       Hydroxyl radical cleavage 
   SSU Empty P site Pre-acc Post-acc Pre-trans Post-trans 
U29 0.975940 1.171044 0.960857 0.561885 0 0.821889 
U33 0.898435 1.167619 1.073215 0.732314 0.388697 1.078317 
  101 
 
G34 0 0.723018 0.944016 0.490194 0 1.043170 
G510 0.821488 0 0.965594 1.004035 0 0 
A511 1.114740 1.074892 0.969845 1.033155 0 0 
A512 0.982187 1.086948 0 0.968909 0 0 
U513 1.166411 1.150214 0.702178 0.959734 1.139413 0.998140 
G514 1.189051 1.120822 0.928289 1.013778 0 0.739962 
A544 1.170754 1.049323 0.733191 0.878903 1.198156 0 
U547 0.965420 1.092893 0.776802 1.003396 0.460939 0 
G548 0.993852 0.985371 0.905813 0.867271 0 0 
G586 0.829594 1.075882 0.905225 1.007972 0.732587 0.919999 
C587 0.833604 0 0 0.991647 0 0.853817 
A1444 0 1.245539 1.103465 1.045523 0.945114 1.591058 
C1461 0 0 1.122910 0 1.230382 1.141285 
G1462 0 0.978274 0.970021 1.220028 1.136653 1.091940 
G1464 1.237013 0 1.133840 0.828773 0 0.841558 
G1638 0.865004 0.189620 0.912904 0.968274 0.633609 0.338911 
LSU Empty P site Pre-acc Post-acc Pre-trans Post-trans 
G1243 1.760012 0 0.951139 1.679863 0.964481 1.150582 
A1244 1.202057 0.574830 0 0 1.116745 1.106730 
A1270 1.293568 0.713459 0 1.074548 1.318914 1.315441 
C2308 1.120530 0.952449 1.447967 1.842303 1.363942 1.324106 
C2849 1.076941 1.156672 0.896894 0.891885 0.588945 0.838247 
G2850 0 0.949812 1.015095 1.488754 0 0.900230 
A2851 1.073177 1.365186 1.060798 1.038430 0.923093 1.085299 
A3017 0 1.025394 0.997964 0 0.712744 1.327650 
C3018 0 1.086792 0.780153 0 0 0.935807 
U3019 0.859554 1.188375 0 1.198593 0.722542 0 
U3023 0.926946 1.089361 0 0.858891 0.713559 0.989456 
A3024 0.903170 1.054250 0.695538 0.856185 0.986207 0.901661 
C3025 0.000000 1.086760 0.980006 1.112488 0.778146 1.032909 
G3026 0.928263 0.773983 0.869914 1.244164 0.975608 1.100397 
A3035 0 1.063433 0 0.947684 1.005874 1.114901 
G3036 1.003952 0.928459 0 1.006930 0.841481 0.852623 
U3038 0 1.255474 1.135425 0.780687 1.050342 0 
Table 5. EF binding sites from chemical probing studies.  
Empty: empty ribosome values, P site: P site occupied complex values, Pre-acc: pre-accommodation 
complex values, Post-acc: post-accommodation complex values, Pre-trans: pre-translocation complex 
values, Post-trans: post-translocation complex values. SSU: small subunit, LSU: large subunit. The 
values are obtained directly from complex comparison calculations (please see Chapter 6: Materials 
and methods). 
 
X-ray crystallographic structures of the bacterial ribosome with EF-Tu 
(stabilized by kirromycin) or EF-G (stabilized by fusidic acid) were also examined for 
nucleotides that interact with the elongation factors 33,47. In the hSHAPE data, SSU 
nucleotides G430, C431, and LSU nucleotides A1270, C2844, G3028, A3032 emerge 
  102 
 
as eEF1A (EF-Tu) interacting nucleotides. SSU nucleotides U52, C433, G434, A436, 
A441, C442, and LSU nucleotides G1243, G3030-A3032 emerge as eEF2 (EF-G) 
interacting nucleotides. Results are shown in Table 6. 
X-ray crystallography 
     SSU Factor Empty P site Pre-acc Post-acc Pre-trans Post-trans 
U52 EF-G 0.994683 0.814413 1.000532 0.721838 0 0.729098 
G430 EF-Tu 1.132694 1.039492 0.405961 1.070965 0.998279 1.220954 
C431 EF-Tu 0 0.944068 0.534706 0 1.310621 1.247968 
C433 EF-G 1.078705 1.168738 0.970594 0.529084 0 1.336763 
G434 EF-G 0 1.086352 1.115286 0.966473 0.107163 0 
A436 EF-G 1.206836 1.099728 1.454033 1.463126 0 1.447335 
A441 EF-G 1.035357 1.102187 0.978408 1.068419 0 0 
C442 EF-G 0.896014 1.080354 0.908404 0.980458 0.742448 1.062566 
LSU Factor Empty P site Pre-acc Post-acc Pre-trans Post-trans 
G1243 EF-G 1.760012 0 0.951139 1.679863 0.964481 1.150582 
A1270 EF-Tu 1.293568 0.713459 0 1.074548 1.318914 1.315441 
C2844 EF-Tu 1.240611 1.553191 1.160410 1.470473 1.971848 1.143636 
G3028 EF-Tu 0 1.145025 0.961712 0.908052 1.093411 1.195303 
G3030 EF-G 0.829820 1.157142 0.968730 0.499957 0 1.023816 
G3031 EF-G 0.609064 1.144072 1.008770 0.739089 0 0 
A3032 EF-Tu, EF-G 0.710629 1.244453 1.042951 0.806668 0.608918 0.938253 
Table 6. EF binding sites from X-ray crystallography structures. 
Empty: empty ribosome values, P site: P site occupied complex values, Pre-acc: pre-accommodation 
complex values, Post-acc: post-accommodation complex values, Pre-trans: pre-translocation complex 
values, Post-trans: post-translocation complex values. SSU: small subunit, LSU: large subunit. 
Nucleotides implicated in binding of specific elongation factors are labeled in the “Factor” column. 
EF-Tu is the bacterial counterpart of eEF1A, EF-G is the bacterial counterpart of eEF2. The values are 
obtained directly from complex comparison calculations (please see Chapter 6: Materials and 
methods). 
 
Comparison of the whole GAC (nucleotides 1230-1280) in all complexes 
yielded interesting results. There is a decrease in reactivity of this region in the P site 
occupied complex compared to empty ribosomes, followed by increase in reactivity 
in the pre-accommodation and, to a lesser extent, post-accommodation complexes 
(Figure 43). The reactivity decreases again in the pre-translocation complex, in 
accordance with eEF2 binding, and increases in specific bases in the post-
translocation complex. The results from sarcin-ricin loop (SRL; nucleotides around 
3020-30 in the above tables) indicate an overall loss of reactivity in pre-
  103 
 
accommodation and pre-translocation complexes, in accordance with EF binding, 
hence the deviance in the reactivity of the corresponding GAC’s is thought 
provoking. 
 
Figure 43. P stalk base (GTPase associated center, GAC) in the difference maps. 
Upper row left to right: P site occupied - empty, pre-accommodation - P site occupied, post-
accommodation - pre-accommodation difference maps. Lower row left to right: pre-translocation - 
post-accommodation, post-translocation-pre-translocation difference maps. P stalk base is circled. 
4.2.3. Intersubunit bridges 
 
The hSHAPE data were also compared with structural studies from bacteria 
and yeast that identified LSU:SSU contacts, i.e. intersubunit bridges, in nonrotated 
and rotated ribosomes, respectively 2,161. Nonrotated complexes (P site occupied, 
CHX, post-translocation complexes) were mainly compared to the empty ribosomes 
and the complexes that are immediately before or after them, to evaluate the bridges 
  104 
 
from nonrotated structures. Rotated complexes (post-accommodation, ANI, pre-
translocation) were used in the evaluation of rotated ribosome bridges. An nonrotated 
/ rotated ribosome bridge component was assumed to be visible in hSHAPE data if 
any of the nonrotated / rotated complexes showed protection compared to the empty 
ribosomes and the other complexes surrounding them. We only considered the rRNA 
components, as hSHAPE does not give data on proteins. These results are 
summarized in Table 7 and Table 8. 
SSU Bridge Empty P site Post-acc Pre-trans Post-trans CHX ANI 
C1646 B2a 0.86574 0.812653 1.1208662 1.997161 0.895131889 0.8498122 0.98274652 
U1647 B2a 1.372284 1.057123 1.6139609 0.8485179 0 1.6351251 1.38117491 
A1648 B2a 0 0.742094 1.1133393 1.6802686 1.588106108 1.0646258 0.71909374 
G997 B2b 0.860241 0 1.0414039 1.1617747 1.007135251 1.1099782 0.93796598 
C1006 B2b 0.615169 0.1937 0.8375123 0 0.972632686 0 1.02024292 
U982 B2c 0.840464 0 1.1998926 0.9471848 1.238487449 1.2533042 1.18279604 
A983 B2c 0.887246 1.055425 1.0466288 0.9057969 1.08382522 0.9540696 1.02080427 
A1125 B2c 0.987227 0 1.2357653 1.5300019 0.740117114 0 0.8400954 
G1126 B2c 0.938292 1.21013 0.8838819 0.9293682 1.015187456 0 0.92730647 
G976 B4 0.575106 0 1.086813 1.0538135 1.246054331 0.8116136 0.60522325 
G1658 B5 0.923964 0 1.1386782 1.0804183 0.932762866 1.1155992 1.30347527 
A1660 B5 0 0 0.8819223 0 0.960882592 0.7230418 1.4543778 
U1738 B5, B6 0 0 1.0290032 
  
0 0.88136261 
C1739 B5, B6 0 
 
0.9020902 
  
1.0537561 1.12512227 
A1667 B6 0.767705 0.69241 0.8728328 0 0.833540216 0.6982577 1.34177178 
G1668 B6 0.714619 0 0.8697098 0 0.730278628 0.8622518 0.91567792 
U1669 B6 0 0.357769 0.882194 0.9626738 0.727681563 1.0102266 1.48336323 
C910 B7a 0.778511 0.973588 0 1.1725574 0 0 0.89769034 
G913 B7a 1.210575 0.970306 1.3410615 1.4194958 0 0.9552165 1.28318313 
G914 B7a 0.918626 0.862648 0 1.0614172 0 0.8893325 0.79972507 
G925 B7b 0.711435 0 0 0.8698684 1.057758313 0.8889053 0.88565041 
G985 B7b 0.864954 1.056574 1.0360825 1.0577368 1.162144775 0.9615991 0.99307325 
G986 B7b 0.582086 1.109588 0.9386496 0.8580121 0.832449337 0 1.0021138 
G987 B7b 0.881483 0 1.1338049 1.2882377 1.247737575 1.1666039 1.11202167 
G419 B8 1.05883 0 1.4406986 0.8919211 0.968572086 1.0668742 0.95748056 
A420 B8 1.228887 0 1.7342675 0.4392027 0.95474109 1.3536546 0.9589811 
         LSU Bridge Empty P site Post-acc Pre-trans Post-trans CHX ANI 
A1025 B1a 1.674609 1.007647 2.117648 1.6129892 1.048101287 1.0372475 1.99149157 
A1026 B1a 1.682716 1.528939 2.2137263 1.7148504 1.375389568 1.0343607 2.09227145 
A2256 B2a 1.289989 0.385102 0.6972292 0.8441767 1.407282296 0 0 
C2257 B2a 1.429314 0 2.0814382 1.4846587 1.028994542 1.0996502 1.58934739 
G2261 B2a 1.326057 1.11518 2.1254041 1.1706219 0 1.4751843 1.78384405 
  105 
 
C2196 B2b 0.918888 0 0 0.552581 0.783829875 0.8032884 0 
A2262 B2b 1.269554 0 1.8641947 1.356294 1.08741044 1.2967673 1.38128648 
C2265 B2b 1.166487 0.82812 1.0906709 1.1640379 1.288430234 0.8209109 0 
A2275 B2b 1.030825 0 1.4941747 0 1.093299901 1.3906824 0 
C2192 B2c 0.971161 0 0 0.769856 0 0.9621819 0.96012941 
C2290 B3 1.021718 0.765861 1.6213262 1.2229734 1.147171389 1.0716945 1.10985335 
A2291 B3 1.069814 0.913487 1.484852 1.3096802 1.133128308 0.9413655 0.86338599 
A2303 B3 0.792156 0.824726 1.2298419 0.877768 0 0 0.8179269 
C2304 B3 0 0 0 0 1.022986958 0 0.75423998 
C849 B4 0 0.990276 0.7764342 0 0.739915883 0.6506956 0.8063262 
A2332 B5 0.79065 0 1.4466433 0.9000802 0.863621087 1.0152965 0 
G1934 B6 0.883854 1.024268 0.8707546 1.1903695 1.023360696 1.1747014 1.40425196 
A2208 B7a 1.396026 0.758025 1.2931782 1.1020973 0.927880223 0.8090739 1.59736329 
U2209 B7a 1.635583 1.356684 2.2219065 1.4285737 1.531487289 1.7098194 1.77030469 
G2239 B7a 1.120313 0.904275 1.5196005 1.1541604 0.771308043 1.075435 1.07088127 
Table 7. hSHAPE data on bacterial, nonrotated conformation bridges. 
The bridge components were obtained from Yusupov et al, 2001 161. The bridge column indicates 
which intersubunit bridge the nucleotide belongs to. Expected decreases in reactivity in the nonrotated 
complexes are highlighted green. The values are obtained directly from complex comparison 
calculations (please see Chapter 6: Materials and methods). 
 
The bridge components at which all complexes of a specific rotational status 
showed loss of reactivity compared to the complexes of the “opposing” status were 
especially valuable; strikingly, these are very few in number. We also observed that 
the nonrotated complexes, especially the P site occupied complex, showed the highest 
number of expected reactivity levels in the bridges. These findings are consistent with 
the notion that there may be “sub-rotational conformations” occurring throughout 
elongation and that the intersubunit interface may be more “fluid” in structure than 
previously thought. These sub-conformations may manifest themselves as (A) the 
formation of some bridges but not others, or, (B) the formation of some interactions 
that make up a bridge but not others, or, (C) combinations of both. As indicated by 
the better fit of nonrotated complex hSHAPE data to the list of nonrotated bridges, 
compared to the rotated complex data and rotated bridges, there may be fewer “sub-
nonrotated” conformations than rotated ones. 
  106 
 
The eukaryotic bridges implicated in “subunit rolling”, i.e. B6, B8, eB8, and 
eB13, were also examined 26. Some effects were also to be expected in eB11 and 
eB12. While this analysis captured some changes in these bridges (highlighted grey in 
Table 8), we note that they also overlap with SSU rotation. 
SSU Bridge Empty P site Post-acc Pre-trans Post-trans CHX ANI 
C1644 B2a 1.197942 0.80129 1.2385371 0 0.57737618 0.2824204 0.99134972 
G1645 B2a 1.191097 0.68625 0.7017392 1.5028355 0.33794883 0.5609593 0 
U982 B2c, B7b/c 0.840464 0 1.1998926 0.9471848 1.23848745 1.2533042 1.18279604 
A1655 B3, eB14 1.172668 0.91913 1.1980247 1.1473936 1.13295112 1.103525 0 
A1746 B3 1.107333 
 
0.4917296 
  
1.18558 
 
G1747 B3 1.155208 
 
0.7777143 
  
1.0384968 
 
G1748 B3 0.849402 
 
0 
  
0.8211643 
 
A1749 B3 0.805794 
 
0.6787745 
  
0.9816615 
 
G628 B4 0.983325 0.95369 1.1084418 0.9176349 1.28509855 0.9806069 0.98876165 
C411 B5, B8 0.979727 0 0 0 0.92383781 0 1.15636695 
A1667 B5 0.767705 0.69241 0.8728328 0 0.83354022 0.6982577 1.34177178 
U1734 B5 0.777364 0 1.0561588 0 
 
1.163628 0.6512412 
G1670 B6 0 0.87795 0.9599856 0 0 1.0997787 0 
U909 B7a 0.804306 0.9771 0.6761169 0.9856134 1.02802159 0.764185 0.77536369 
C910 B7a 0.778511 0.97359 0 1.1725574 0 0 0.89769034 
U911 B7a 1.442609 1.14016 1.6040537 1.435276 1.30660978 1.2548329 0.81765021 
A923 B7b/c 0.541091 0 0.9651395 0.288727 0.7841185 0.6040875 0.89618312 
G986 B7b/c 0.582086 1.10959 0.9386496 0.8580121 0.83244934 0 1.0021138 
U1012 B7b/c 0.740227 0 1.092217 0 1.0207736 1.1491892 1.20778133 
G1122 B7b/c 0.895671 1.20101 1.2928013 1.4566045 1.15371639 1.2369045 0 
A412 B8 1.406518 1.16878 0.7639067 1.1158622 1.16966659 1.1471297 1.23205285 
U851 eB12 1.109696 1.09448 0.9380267 0 1.02969013 0.7401936 0.72600599 
G273 eB13 1.066791 1.14048 1.3061394 0.9359146 1.22184896 1.053235 1.12253314 
G274 eB13 0 0 1.5934351 1.2150526 1.46781917 0 0.82992983 
C275 eB13 1.300059 1.52884 2.1362754 2.1539131 2.38308732 1.8034771 1.47732532 
A1714 eB13 0.946517 0.71593 0 0.9875311 0 0.9851416 0 
G1114 eB14 0.299353 1.13198 0.9823288 0.9257716 0.95046597 0.9258818 1.03726958 
U1115 eB14 0 0 1.0249099 0.8024211 1.03311253 1.0592058 1.16080222 
G1126 eB14 0.938292 1.21013 0.8838819 0.9293682 1.01518746 0 0.92730647 
G1127 eB14 0.847594 1.25346 0.8140357 0.749936 0 0.4022362 0.68906548 
C1641 eB14 0.947436 0.74338 0.3160873 1.9428717 1.42852197 0.9327305 1.08301684 
G1642 eB14 0.862858 0.81917 0.7312791 0 1.55778182 0.9029685 1.11540635 
         
LSU Bridge Empty P site Post-acc Pre-trans Post-trans CHX ANI 
A2259 B2a 0 0.82213 0 0 0.84494963 0 1.60620842 
C2263 B2a 0.920025 0.72067 0 0 0.59286677 0.801735 0.84492549 
C2196 B2b 0.918888 0 0 0.552581 0.78382988 0.8032884 0 
C2197 B2b 0.956444 0 1.3292934 0.914977 1.02848798 1.0841475 0 
A2275 B2b 1.030825 0 1.4941747 0 1.0932999 1.3906824 0 
  107 
 
C2192 B2c 0.971161 0 0 0.769856 0 0.9621819 0.96012941 
A1922 B3 0.858229 1.1692 0.7506318 1.062651 1.01319101 0.9216749 0.8793968 
A2291 B3 1.069814 0.91349 1.484852 1.3096802 1.13312831 0.9413655 0.86338599 
U2301 B3 0.876703 0.89921 1.0899571 0 0.98564701 0.743509 0.97125693 
G2302 B3 0.995543 0.89406 0 1.008746 0.97549001 1.1056209 1.02460239 
G2305 B3 1.342081 1.17167 2.0275786 1.2708593 1.37998697 1.4428183 0 
G844 B4 1.033777 1.29458 1.1662149 1.3124769 1.28052769 1.2576829 0 
A848 B4 1.371455 1.20915 0.8673768 0 1.12450725 1.0884484 1.02659477 
G1935 B5 1.027573 1.16948 0.8775504 1.2388827 1.01987166 1.1395794 1.49197697 
A1936 B5 1.139345 1.23549 0.6964552 1.0620109 1.02945407 0.9721565 1.14264889 
G2240 B7a 1.181148 0.755 1.6445104 1.2295566 1.21894703 1.2051912 0 
G2533 eB8 0.821709 0 0 0 0 0.8387015 0 
A2535 eB8 1.071347 0.68267 0 1.0482945 1.07665899 1.0188318 0.92474591 
G3345 eB11 0.967349 
 
0 
  
0.9833053 
 Table 8. hSHAPE data on eukaryotic, rotated conformation bridges. 
The bridge components were obtained from Ben-Shem et al, 2011 2. The bridge column indicates 
which intersubunit bridge the nucleotide belongs to. Bridges implicated in subunit rolling are 
highlighted in grey. Expected decreases in reactivity in the rotated complexes are highlighted green. 
The values are obtained directly from complex comparison calculations (please see Chapter 6: 
Materials and methods). 
 
Since there were more unexpected data on the intersubunit bridges than the 
regions examined in the earlier sections, and since intersubunit rearrangements 
remain an active field of research in eukaryotic ribosomes, we examined our data 
from one more angle. A difference map of empty 80S and 40S was generated (Figure 
44), to highlight the reactivity decreases caused by subunit association. As expected, 
most of these are found on the known subunit interface of the ribosome, clustering in 
the SSU h44, and in the head and platform domains. Interestingly, the SSU nucleotide 
numbers so identified were not a perfect match to the known list of yeast intersubunit 
bridges 2. 
We then focused on the same nucleotides in all difference maps (Figure 45), 
with surprising results. Except for the platform and a small portion of h44, the 
possible “interface nucleotides” are very reactive in the nonrotated, P site occupied 
complex compared to empty ribosomes. In the pre-accommodation complex, most of 
these become even more reactive compared to the P site complex, whereas in the 
  108 
 
post-accommodation complex a portion of the head domain and most of h44 have 
decreased reactivity compared to pre-accommodation ribosomes. The nucleotides 
become more reactive again in the pre-translocation complex compared to post-
accommodation ribosomes. Decreased reactivity in the post-translocation complex is 
visible but not dramatic. Finally, the rotated complex with hybrid state tRNAs shows 
decreased reactivity in the head domain and in parts of h44 compared to the 
nonrotated complex with classical state tRNAs. 
 
Figure 44. Regions of protection on 18S rRNA upon subunit joining. 
The nucleotides that have lost reactivity in 18S rRNA after 60S joining (i.e. difference map of 80S vs 
40S) are highlighted in green and mostly map to the subunit interface. 
 
These rather unexpected results reveal some novel insights into ribosome 
structural dynamics.  First, a region on the platform located close to one nucleotide of 
the B7b/c bridge, becomes protected upon subunit association and in nonrotated, P 
site occupied ribosomes. Second, the rearrangements observed in the subunit interface 
suggest that the pre-accommodation complex may be in a rotated state. Finally, 
different bridges or regions of the intersubunit interface are affected in different 
complexes, regardless of rotational status, supporting the notion that the ribosome 
may acquire many different rotated and nonrotated sub-states. 
  109 
 
 
Figure 45. The reactivity changes in 60S interacting nucleotides of the SSU through the 
elongation cycle. 
The regions obtained from the difference map of empty 80S and 40S (Figure 44) were examined in the 
difference maps of elongation cycle complexes. A. P site occupied complex vs. empty ribosome, B. 
Pre-accommodation vs. P site occupied complex, C. Post-accommodation vs. Pre-accommodation 
complex, D. Pre-translocation vs. Post-accommodation complex, E. Post-translocation vs. Pre-
translocation complex, F. Complex with hybrid state tRNAs (ANI) vs. Complex with classical state 
tRNAs (CHX). 
4.2.4. Expansion segments 
 
The eukaryote specific components of the rRNA are known as expansion 
segments (ES). Graphing the hSHAPE reactivities of ES’s from different elongation 
complexes yielded interesting results. Overall, most ES’s show accommodation-
specific changes in reactivity, such that the reactivity decreases or increases as 
ribosomes progress from the P site occupied complex to the post-accommodation 
  110 
 
complex. Graphs of ES’s that display the most dramatic changes are shown in Figure 
46. These are ES3S from SSU, ES7L from the P stalk region, ES27L from the “back” 
of LSU, across from the P stalk, and ES31L from the L1 stalk region. Further 
analyses indicate that the reactivity returns to pre-accommodation levels in pre-
translocation and post-translocation complexes for ES7L, ES27L and ES31L. For 
ES3S, it stays roughly at the post-accommodation levels.  
 
 
 
 
 
Figure 46. hSHAPE reactivity of different expansion segments in the accommodation related 
complexes. 
The reactivities of ES3S, ES7L, ES27L and ES31L in P site occupied, preaccommodation and 
postaccommodation complexes are shown, and indicate overall increase or decrease in reactivity as 
accommodation proceeds. 
  111 
 
These results indicate that expansion segments go through structural 
rearrangements during elongation, most notably during accommodation, suggesting 
potential roles for these eukaryote-specific elements at this step of translation. 
4.2.5. Allosteric communication pathways 
 
The ribosome is a very large protein-RNA complex and the functions of only 
few regions are well understood. It is generally assumed that the whole molecule is 
required for structural stability and that the main communication occurs between DC, 
PTC, and GAC, through tRNAs and transacting factors. However, studies have found 
that mutations within these regions may cause changes in rRNA structure extending 
into other regions 176–178. Furthermore, ribosomal protein mutations that map to 
regions that are away from catalytic sites can affect both ribosome function and 
structure 179–182. These indicate that information may be transferred to and from 
functional sites to distal regions of the ribosome, possibly through allostery. To find 
such “allosteric communication pathways” in the hSHAPE data, patterns of reactivity 
changes were examined. Nucleotides that did not show reactivity changes in one or 
more difference maps were omitted. 
These analyses reveal that there is no single nucleotide whose reactivity 
continually increases or decreases as the ribosome progresses through the elongation 
cycle. There are also no nucleotides that only lose reactivity in pre-accommodation 
complex, or only in post-accommodation complex, or only in pre-translocation 
complex, or only in post-translocation complex. A single nucleotide, A3008, of LSU 
H94 is only deprotected in P site occupied ribosomes. There are no nucleotides that 
only lose reactivity in P site occupied ribosomes; this was expected since most or all 
  112 
 
features of this complex are present in all of the other complexes as well (e.g. all of 
the complexes contain both a tRNA in the P-site and an mRNA). 
Specific patterns were observed that highlight the relatively direct effects of 
(1) trans-acting factor binding, (2) rRNA structural rearrangements that occur during 
a specific step (e.g. accommodation or translocation), (3) structural rearrangements 
that occur between steps (i.e. accommodation and translocation). A five-symbol 
annotation is used to label the patterns, where (-) denotes a reactivity decrease while 
(+) denotes an increase, and the order of (-)s and (+)s match the difference maps: P 
site occupied vs empty → pre-accommodation vs P site occupied → post-
accommodation vs pre-accommodation → pre-translocation vs post-accommodation 
→ post-translocation vs pre-translocation. 
The most common two patterns in the hSHAPE data are (-+-+-) and (+-+-+), 
i.e. the nucleotides sharing these patterns go through reactivity increases or decreases, 
respectively, only in pre-accommodation and pre-translocation complexes. Of the 
5354 total nucleotides examined, 60 and 76 nucleotides, respectively, share these 
patterns. These patterns are mainly related to EF binding. The cyclical nature of these 
patterns is especially thought provoking as it may reinforce idea that of the sequential 
binding of these transacting factors ensures the unidirectionality of eukaryotic 
elongation.  
4.2.5.1. Trans-acting factor binding patterns 
 
As discussed in the previous section, the (-+-+-) and (+-+-+) patterns were of great 
interest (Table 9). Mapping of these enabled us to identify connections between 
nucleotides located in different functional regions, including H69 (decoding, the B2a 
  113 
 
bridge), H82 (the P and E sites), H88 (the E site), H90 (the A site), and the SRL. 
Nucleotides from these regions showed decreased reactivity upon factor binding. 
Other known important regions were also highlighted as a result of different 
nucleotides either decresing or gaining reactivity upon factor binding: these include 
h23 (the B7b/c bridge), and h44 (decoding, and many bridges).  
Helices (-+-+-)  Helices (+-+-+)  Helices (--+-+)  Helices (++-+-) 
 5.8S   5.8S   18S   18S 
SS A41  H6 C57  B h24 C1006  h6a C99 
SS A61   18S  h34 C1279  B h8 U144 
SS U74  h6 A86  B h42 U1552  B h13 C402 
SS G75  ES3 210  h43 A1592  SS C435 
 18S  ES3 236   25S  h17 C530 
B h5 U44  B ES3 246  H42 C1216  B h23 U896 
B h6 A65  h11 U318  H43 C1254  B h34 A1275 
h6 C87  h11 G325  SS C1257  h34 C1280 
SS U101  h11 G329  H41 U1336  h35 G1292 
SS U137  h11 C338  SS G1493  h35 C1327 
h8 G154  h11 G356  H54 U1626  h39 C1339 
h9 A182  h12 G363     SS U1413 
h12 G365  h12 U374     h38 U1415 
SS A425  h12 C382     h41 U1532 
B h18 G552  h16 G487     B h42 A1543 
h20 U632  h22 G879     h42 G1548 
ES6 C696  h23 U921     B h42 G1553 
h23 G901  h23a C937     B h43 C1596 
B h23a A939  B h22 A952     B h43 U1598 
B h27 A1131  SS A963     B h43 C1599 
SS A1287  h20 U968     h43 C1606 
B h39 U1347  h27 C1128     SS U1613 
h39 U1350  B h33 A1244     h28 U1617 
h39 G1364  h33 C1248     B h28 C1618 
SS C1634  h34 G1271     h28 U1621 
h44 G1713  h40 G1408     h28 U1626 
h44 G1718  h38 G1418     h28 U1627 
h44 U1724  h41 G1498     h44 A1648 
 25S  B h41 G1523     h44 G1649 
B H11 G33  h43 U1604     h44 U1669 
H17 G171  h44 G1680      25S 
H17 G244  h44 U1682     B H11 A35 
H16 G258  h44 U1725     B H11 G41 
B H21 G282   25S     H41 A1182 
B H31 A761  H11 U44     SS A1190 
H31 G774  H13 A89     H50 A1498 
H33 U905  B H13 A99     H50 G1517 
H37 G964  H15 G139     H50 U1518 
  114 
 
ES12 G1089  H15 G143     SS U1555 
B H41 A1179  H14 A319     SS C1556 
H46 U1384  H2 U414     H53 C1598 
H26a A1446  ES7 G495     H54 G1624 
H26a U1448  ES7 G505     B H59 C1761 
H47 U1470  ES7 U506     H76 U2505 
H52 C1550  ES7 G510     H97 G3109 
ES19 U1564  ES7 C515     H98 G3288 
ES19 G1565  B ES7 A521     H98 G3289 
H53 C1597  ES7 C539     H101 C3324 
H54 U1616  ES7 U587       
ES26 C1633  ES7 C599       
H55 G1652  B ES7 G600       
H57 A1676  B H30 G726       
H58 G1727  H27 G787       
ES27 U1992  B H35 U871       
H67 C2192  H33 A904       
H61 C2350  H41 G1172       
H93 A2969  H57 U1682       
H96 G3052  H54 A1823       
H96 G3083  ES27 G1968       
ES39 C3249  ES27 G1987       
SS A3299  H69 U2254       
H101 G3327  SS U2334       
   H82 U2650       
   SS G2658       
   H86 A2721       
   H88 A2769       
   H74 A2813       
   H90 G2939       
   H90 G2947       
   B 95 C3025       
   H96 G3059       
   B H96 G3080       
   H97 U3111       
   ES39 C3190       
   ES39 G3200       
   H98 U3287       
   B H100 A3307       
   B H101 U3329       
Table 9. Nucleotides involved in reactivity change patterns related to EF binding and 
dissociation. 
The first two patterns focus on changes upon factor binding, the second two focus on factor 
dissociation. Eukaryote-specific components are shown in light blue. Known functional components 
(e.g. involved in tRNA binding / intersubunit bridges / decoding center / P and L1 stalks / peptide exit 
tunnel) are shown in dark blue. A nucleotide close to the PTC but not part of the known functional 
regions is shown in red. Nucleotides belonging to a domain at the back of the LSU (roughly opposite 
from the central protuberance) are italicized (please see the discussion section). SS denotes nucleotides 
from single stranded regions. A “B” in front of a helix name indicates the nucleotide is located in a 
loop or other single stranded portion of the helix.  
 
  115 
 
We also examined (--+-+) and (++-+-) as other patterns related to trans-acting 
factors (Table 9). These appear be more related to factor dissociation than binding. A 
striking result from this analysis is the increase in reactivity of nucleotides located in 
the SSU head domain. Many nucleotides of the head domain become unreactive upon 
factor dissociation (i.e. in post-accommodation and post-translocation complexes). 
This may be a consequence of factor dissociation, of the simultaneous events taking 
place at these steps (e.g. rotation), or of both. As expected, the P stalk base shows 
decreased reactivity upon elongation factor binding (H42-44). 
As the results discussed in Section 4.2.2.3. Elongation factor - rRNA 
interactions suggest, the reactivity of the EF binding site (especially the P stalk base) 
changes dramatically in the P site occupied complex compared to the empty 
ribosome. The (--+-+) and (++-+-) patterns may also be indicative of nucleotides that 
are affected in both P site occupied complex and the trans-acting factor bound 
complexes. In this case, the above interpretation might be modified from “EF 
dissociation” to “loss/gain of reactivity in P stalk base / GAC” as the cause of the 
changes listed in Table 9 (patterns (--+-+) and (++-+-)). 
4.2.5.2. Patterns indicating rearrangements at a specific step 
 
Next, we focused on nucleotides belonging to reactivity change patterns in 
accordance with rearrangements during the accommodation or translocation steps of 
elongation. For this analysis, patterns in which decreased reactivity was observed in 
only one complex were compared to their “partners” at specific steps in the 
elongation cycle (Table 10 and Table 11). 
 
  116 
 
 
Helices (+-+++)  Helices (++-++) 
 18S   18S 
h10 C282  h3 C31 
SS A1139  h4 C36 
h44 G1695  h26 G1046 
 25S  SS A1137 
H41 C1183  SS A1321 
ES27 G1998  h42 U1567 
H95 A3035  h28 G1629 
H96 U3058  h44 A1678 
ES39 G3199   25S 
   H83 C2708 
Table 10. Nucleotides involved in reactivity change patterns related to accommodation. 
Eukaryote specific components are shown in light blue. Known functional components are shown in 
dark blue. Nucleotides close to the PTC but not part of the known functional regions are shown in red. 
SS denotes nucleotides from single stranded regions. A “B” in front of a helix name indicates the 
nucleotide is located in a loop or other single stranded portion of the helix. 
 
Comparison of accommodation complexes using this rubric suggest that more 
communication pathways may be active in the pre-accommodation complex than 
post-accommodation, indicated by the number of nucleotides belonging to 
functionally important helices. The pattern (+-+++) may reflect eEF1A specific 
rearrangements as well. In this case, the identification of nucleotides in ES27 and 
ES39 suggest some specific involvement of these expansion segments in 
accommodation. Additionally, the eB13 intersubunit bridge that is involved in subunit 
rolling in mammalian ribosomes, may slightly rearrange in the pre-accommodation 
complex. The only LSU nucleotide that follows the (++-++) pattern (i.e. decreased 
reactivity in post-accommodation complex) is 2708, which is located in H83, part of 
the central protuberance (known to go through rearrangements upon tRNA binding 
183,184). 
Similar to the results from accommodation complexes, additional 
communication pathways may be active in the eEF2-bound pre-translocation complex 
than post-translocation. The pattern (+++-+) may reflect eEF2 specific 
  117 
 
rearrangements. In this case, the presence of a nucleotide from ES7 suggests some 
specific involvement of this expansion segment in translocation. Contrary to the pre-
accommodation complex, there may also be some involvement of the PTC in the pre-
translocation complex (red items in Table 11). No LSU nucleotide was observed to 
undergo a decrease in reactivity only in the post-translocation complex. This is an 
interesting, yet understandable observation, as most rearrangements between pre- and 
post-translocation complexes would be expected to take place in the SSU due to 
intersubunit rotation and translocation of mRNA:tRNA interaction by one codon on 
the SSU neck. Accordingly, most of the nucleotides following the (++++-) pattern are 
from the head and neck region of the SSU. 
Helices (+++-+)  Helices (++++-) 
 5.8S   18S 
H4 U27  h16 C484 
 18S  h39 G1352 
B h23 U886  B h42 A1556 
SS A1256  h28 G1622 
h34 G1267  h44 C1706 
h42 U1564    
h44 G1690    
h44 G1692    
 25S    
SS G400    
ES7 U540    
H46 C1411    
H68 A2220    
SS A2280    
SS C2362    
B H73 A2397    
H83 G2662    
H97 U3107    
SS G3173    
SS C3181    
Table 11. Nucleotides involved in reactivity change patterns related to translocation. 
Eukaryote specific components are shown in light blue. Known functional components are shown in 
dark blue. Nucleotides close to the PTC but not part of the known functional regions are shown in red. 
SS denotes nucleotides from single stranded regions. A “B” in front of a helix name indicates the 
nucleotide is located in a loop or other single stranded portion of the helix.  
 
  118 
 
4.2.5.3. Patterns indicating rearrangements in-between steps of the elongation 
cycle 
 
Patterns of changes in base reactivity that suggest rearrangements leading to a 
switch between accommodation and translocation were also examined. For this 
analysis, patterns that showed opposing reactivity changes in accommodation and 
translocation complexes were identified (Table 12). There are no nucleotides belonging 
to the (---++) pattern, i.e. no nucleotides that are protected in P site occupied, pre-
accommodation, post-accommodation complexes, yet deprotected in translocation 
complexes. 
Helices (+++--)  Helices (-++--)  Helices (+--++) 
 18S   5.8S   18S 
ES6 G845  SS G63  B h1 G10 
 25S   18S  h2 U12 
H25 U430  h11 G330  h3 U27 
SS G1646   25S  B h7 U304 
H68 U2226  SS G420  h18 G561 
B H72 G2377  SS G421  h23 A884 
SS U3319  ES7 G474  B h23 U916 
   ES7 G538  SS U1232 
   ES7 G575   25S 
   H27 C670  H11 U32 
   H30 G728  H39 U1138 
   H31 A751  H60 U1871 
   H33 G826  H63 G1949 
   H45 A1363  ES27 G2042 
   H46 G1379  ES27 C2064 
   H46 C1396  SS C2797 
   B H68 A2198  H94 C3004 
   H68 G2236  H95 A3017 
   B H69 C2257  B H95 G3028 
   B H69 U2260  H96 G3085 
   H69 A2262  B H97 A3106 
   H26a C2359  H99 U3297 
   H73 G2395    
   H76 G2442    
   H75 A2601    
   H84 C2682    
   B H85 A2694    
   H101 C3321    
Table 12. Nucleotides involved in reactivity change patterns related to a switch from the 
accommodation to translocation steps of elongation. 
  119 
 
In these patterns, accommodation and translocation complexes have opposing changes in reactivity 
(e.g. nucleotides losing reactivity through accommodation, but gaining through translocation are 
indicated by the last pattern). No nucleotide belonged to the pattern (---++). Eukaryote specific 
components are shown in light blue. Known functional components are shown in dark blue. 
Nucleotides close to the PTC but not part of the known functional regions are shown in red. SS denotes 
nucleotides from single stranded regions. A “B” in front of a helix name indicates the nucleotide is 
located in a loop or other single stranded portion of the helix.  
 
One possible structural switch may be reflected by decreased reactivity in 
ES27 and ES7 in the accommodation and translocation complexes, respectively. 
Additionally, PTC rearrangements appear to be more highly implicated in 
translocation rather than accommodation (red nucleotides in Table 12). Decreased 
reactivity in a nucleotide of H76 in pattern (-++--) is in accordance with L1 stalk 
rearrangements related to E site tRNA interactions and translocation 185,186. 
Interestingly, this pattern of decreased reactivity is also observed in the P site 
occupied complex, and thus it may be more appropriate to interpret this as a change 
that is not present in accommodation, rather than one present in translocation. 
Different bridges are impacted at different steps, too, such as B2a (H69) and eB11 
(H101) in translocation, and B7b/c (h23) in accommodation.  
4.3. Discussion 
 
The eukaryotic translation elongation cycle is similar to its bacterial 
counterpart, with the same main steps and homologous transacting factors. However, 
given the increased size of the eukaryotic ribosome and the recently described novel 
intersubunit motions in mammalian ribosomes, the differences may be greater than 
previously thought. In this study, we walked the ribosome through an entire round of 
elongation, and using hSHAPE to determine the changes in rRNA flexibility, we 
obtained a complete dynamic view of the rRNA bases in the elongating ribosome. 
  120 
 
Our analyses included validation by comparison to previously reported changes 
localized in conserved functional regions, to novel changes in eukaryote-specific 
regions of the ribosome. We also identified the nucleotides that share reactivity 
change patterns throughout the elongation cycle, supporting the model of linked long 
range interactions that constitute complex-wide information exchange pathways.  
The hSHAPE analyses of ribosomes harboring tRNAs reveal that the yeast P 
site shares “hybrid” features between bacterial and mammalian sites (Figure 42, Table 
13). In the 40S subunit, the yeast P site protection data closely matches the 
tRNA/rRNA interactions observed in mammalian cryo-EM structures. In contrast, the 
yeast 60S P site hSHAPE protection patterns are a better match to the 50S P 
site/tRNA cleavage patterns obtained from bacterial hydroxyl radical cleavage 
studies. These findings may provide some novel insight into the evolution of the 
ribosome, suggesting that the 40S subunit was first reengineered in the eukaryotes to 
accommodate a more intricate initiation process, followed by further evolution of the 
60S subunit in higher eukaryotes. Results obtained from the posttranslocation 
complexes reinforce the similarities between the yeast and bacterial P sites (through 
hydroxyl radical cleavage), while highlighting the differences between the respective 
P sites in the small subunits (most dramatically observed by loss of reactivity at SSU 
nucleotides A1001-G1002). One caveat to this comparison however is the differences 
between hSHAPE and hydroxyl radical cleavage: the former modifies specific 
nucleotides while the latter targets both exposed bases and their near-neighbors. 
Overall, further experimentation and higher resolution structures will be required in 
order to define the yeast P site, as well as A and E sites at atomic resolution, 
  121 
 
improving on the low resolution structures available currently 187–190. Archaeal 
ribosomes may be another intermediate to examine 191, however the available high 
resolution structures do not contain the fully associated 70S archaeal ribosome and 
rely heavily on the eubacterial ribosome as a model 192–195. 
Prokaryotic P site Mammalian P site 
SSU LSU SSU LSU 
G577-A580 C2284-C2290 C1000-G1002 U2408-U2410 
A905 U2297-A2299 G1169-A1171 U2423-A2424 
A1005-C1007 G2302-A2309 C1190-C1192 U2771-C2772 
A1151 G2619-A2626 A1194-C1195 G2778-A2779 
C1192 A2808 C1461-G1466 A2813-G2814 
A1576-77 A2820 G1574-U1579 U2861-G2863 
G1638-C39 U2861-G2863 G1622 U2978-79 
  U2953-54 A1635-C1639 A3129-30 
  C2963-G2972     
Table 13. Comparison of the matches between the observed yeast P site and the prokaryotic / 
mammalian P sites. 
Nucleotides from previous bacterial and mammalian studies are listed, and the nucleotide ranges that 
match the observed yeast P site (based on the results from both P site occupied complex and 
posttranslocation complex) are marked in green. We cannot obtain data between 18S nucleotides 
G1140-G1200 (grey) by hSHAPE due to presence of highly modified bases in the region, inducing 
strong stops in the primer extension reaction. 
 
Analyses of the hSHAPE data pertaining to the intersubunit bridges indicate 
that one size does not fit all, as there seem to be major rearrangements or reactivity 
differences in the bridge regions among ribosomes of similar rotational status (i.e. 
rotated or nonrotated). This implies that the notions of “rotated” and “nonrotated” 
states are overly simplistic. Indeed, time-resolved cryo-EM has revealed over 50 
conformations of the bacterial ribosome 196, and other studies have identified tRNA 
binding “substates” wherein various portions of tRNAs make interactions with the 
ribosome that are “in-between” previously defined tRNA binding sites 51,197,126. 
Hence our hSHAPE data reinforce the notion that the ribosome can assume numerous 
“sub-rotational statuses”. 
  122 
 
Analyses of the data pertaining to the expansion segments indicate consistent 
rearrangements in most of these elements as translation elongation progresses from 
the P site occupied complex to postaccommodation complex. This suggests that they 
undergo dynamic rearrangements during the process of aa-tRNA accommodation. An 
interesting pattern revealed by hSHAPE is that some expansion segments either lose 
(eg. ES27L) or gain (eg. ES3S, ES7L, ES31L) reactivity in their entirety through the 
accommodation process, but that during translocation, only shorter stretches show 
reactivity changes. 
 
Figure 47. Pathway between EF binding site and tRNA binding sites. 
Please see text for details. In this view, tRNA binding sites are visible between the 40S (SSU) head and 
60S (LSU) central protuberance. Nucleotides (spheres) which are only protected in factor bound 
complexes and the helices in which they are located are shown in bright green. Nucleotides protected 
in both factor bound complexes and P site occupied complex (and the corresponding helices) are 
shown in pale green. Nucleotides that are only reactive in factor bound complexes (and the 
corresponding helices) are shown in red. Nucleotides reactive in both factor bound complexes and P 
site occupied complex (and the corresponding helices) are shown in light red. The peptidyltransferase 
center is shown in yellow. The green helices, except H69, lie in a “branched line” covering parts of all 
tRNA binding sites, bridging these and the EF binding site.  
 
Analyses of the patterns of reactivity changes in the hSHAPE data also 
appears to indicate, that elongation factor binding (either eEF1A or eEF2) triggers 
similar reactivity changes in a set of nucleotides located in different helices. These 
  123 
 
include but are not limited to h23 and h44 in the SSU, and H69, H82, H88, H90, 
SRL, H101 in the LSU. The reactivity of these nucleotides all change in the same 
direction as the ribosome progresses from one complex to the other. In addition, they 
are all protected in both the pre-accommodation and pre-translocation complexes. 
These observations suggest the existence of a physical linkage or “information 
exchange pathway” encompassing the EF binding site, the afforementioned 
functionally important helices, and others that we do not know the function of. Some 
of the helices that are revealed in this analysis, including h23, h38, h39, h41, h43, 
H74, H82, H86, and H88, are a remarkable match to a previously defined 
communication pathway spanning the two subunits through B1b/c intersubunit bridge 
that were identified using mutants of ribosomal protein L11 179. 
As Figure 47 shows, the hSHAPE data provides a pathway between the 
elongation binding site (P stalk and H95 - sarcin ricin loop) and A, P and E sites. 
Nucleotides from H42, H43, H95, H90, H69, H74, H82, H86 and H88 lose reactivity 
in the factor bound complexes (Section 4.2.5.1). H42 and 43 also have low reactivity 
in the P site occupied complex compared to empty ribosomes. H42-43 and H95/SRL 
form part of the elongation factor binding site. H90 forms part of the accommodation 
corridor leading to the A site of peptidyltransferase center / PTC. H69 forms part of 
the decoding center, as well as part of the B2a bridge. H74 is in close proximity to the 
PTC and seems to form a bridge towards H86 and H88 which harbors part of the E 
site. H82 is also in close proximity to the PTC and forms part of the P site. 
Nucleotides from H76 and H93 gain reactivity in the factor bound complexes. H76 
also has high reactivity in the P site occupied complex compared to empty ribosomes. 
  124 
 
H76 forms part of the L1 stalk base and is associated with E site tRNA movements. 
H93 forms part of the A site. Taken as a whole, we suggest an allosteric  
communication pathway that starts at the EF binding site, reaches into the PTC, and 
from there separates into branches, one of which spans the A, P and E sites, and a  
second that may employ the A site tRNA to reach to H69 and across the B2a bridge 
to the decoding center.  
 
Figure 48. Pathways between EF binding site and the intersubunit bridges B6, eB11, eB13. 
Please see text for details. Coloring is the same as Figure 47, with additions. Helices that harbor both 
protected and reactive nucleotides are shown in black. Proteins are shown in blue. Peptidyltransferase 
and decoding centers are shown in yellow. Bridges are shown as orange stars. Information may flow 
through H96 to bridges B6, eB13 (through eL24 and h44), or eB11 (through H101 and eS8 - not 
present in this structure) towards the SSU from H95/SRL. 
 
The flexibility of select nucleotides in H76 (following different patterns of 
reactivity changes) may be due to the bending of the L1 stalk towards or away from 
the E site, depending upon the step of elongation. Different conformations may be 
exposing different nucleotides of this helix to chemical attack. On the other hand, the 
flexibility of select nucleotides in H93 may be due to changes in the occupancy of the 
LSU A site, possibly acting as a sensor for the A site tRNA on the suggested 
pathway. The opposing reactivity of H69 may be on par with the role of H93, or lack 
  125 
 
thereof, since, in contrast to the LSU, the SSU A site is not vacant in any of the factor 
bound complexes. 
 
Figure 49. Pathway between EF binding site and the intersubunit bridge B7b/c. 
Please see text for details. Coloring is the same as Figure 47 and Figure 48. Another bridge implicated 
in the reactivity change pattern results is B7b/c. Information may flow to SSU through uL2 - h23/24 
interaction from EF binding site. Components that bridge uL2 and EF binding site are H97, H90 and 
H74. 
 
Examination of the same sets of nucleotides suggests the presence of 
additional pathways connecting EF binding to intersubunit bridges. Factor binding, 
especially eEF2, causes major rearrangements in the bridge regions hence these 
pathways are of great interest. Figure 48 shows a potential pathway governing part of 
the bridges B6, eB11 and eB13. H95/SRL may connect to H101 through H96, which 
have low reactivity nucleotides in close proximity, possibly allowing interaction. 
H101 in turn interacts with eS8 (not present in the ribosome structures used), forming 
most of eB11. H96 also contacts eL24, which interacts with h44 in return to form 
parts of B6 and eB13 bridges. All these bridges have been implicated in subunit 
rolling of the mammalian ribosomes. The increase in reactivity on H96 and h44 
clusters around eL24 contact points, hence in factor bound complexes B6 and eB13 
  126 
 
may be dramatically rearranged. This is partially in accordance with the results of 
Section 4.2.3. 
 
Figure 50. Comparison of hSHAPE data to the B1b/c pathway. 
Please see text for details. Coloring is the same as Figure 47 and Figure 48. Information may be 
flowing through B1b/c from LSU towards the SSU. The opposing reactivities indicate “switch”es 
throughout elongation. h23 reactivity changes may be due to rearrangements of B7b/c rather than 
B1b/c. 
 
The hSHAPE data also reveal a cluster of h44 nucleotides close to the 
decoding center that go through reactivity changes in correlation with EF binding and 
dissociation. This, together with the reactivity change seen at the tip of H69, and the 
reactivity changes in h11 and h12 of SSU, may indicate how the decoding center is 
networked with the whole ribosome. SSU helices 11 and 12 lie in close proximity to 
h44 decoding center 198, even though their exact roles in translation, if any, are not 
known.  
  127 
 
Figure 49 shows a possible allosteric communication pathway between the EF 
binding site and B7b/c bridge. Upon EF binding, rearrangements occur in H90 both 
through H95/SRL and H97. This passes onto H74 which is in close proximity to uL2. 
uL2 in turn contacts h23-24 of small subunit to form B7b/c. The presence of 
nucleotides of opposing reactivities at the h23-uL2 interface may be indicative of a 
switch from nonrotated to rotated conformations (and vice versa) upon factor binding 
(and dissociation). 
 
Figure 51. Pathways between EF binding site and the expansion segments ES7L, ES27L. 
Please see text for details. Coloring is the same as Figure 47 and Figure 48. Information may flow 
through H41 from P stalk to ES7L, while Domain III may be important in a pathway towards ES27L. 
 
As mentioned previously, the helices and loop regions that were described to 
take place in a communication pathway governing the B1b/c bridge 179 aligned well 
with the hSHAPE data and patterns of reactivity changes in the current study (Figure 
50). Nucleotides from h23, h34, h38, h39, h41, h42, H82, H84, and H88 were 
  128 
 
identified as participating in this pathway. Additional nucleotides mapped to regions 
in-between helices, implicating H74 and H86 as well. In our data, even though 
nucleotides from H84 display decreased reactivity in one or both complexes with an 
elongation factor, there is a point at which no noticeable reactivity change occurs. We 
do not recover nucleotides that show no reactivity change at any given point with the 
pattern search yet, hence H84 nucleotides are not classified into patterns currently. 
Improving the pattern search to include lack of changes in reactivity may solve this 
problem, supporting the B7b/c bridge branch of our network of pathways, and may 
enable other previously described communication pathways 181,199 to surface in our 
data. 
The B1b/c bridge pathway in the hSHAPE data may be extending from EF 
binding sites, passing through the PTC, and then moving through H82 to the central 
protuberance and uL5. In fact, a nucleotide in a single stranded region between H82 
and H83, A2658, loses reactivity in the factor bound complexes, and is in close 
proximity to uL5. uL5 in turn contacts uS13. Both uS13 and another interacting 
protein, uS19, then contact h41 and 42, and the “information” is transferred further 
into the SSU to h34, 38 and 39. According to our data, h23 is part of a pathway 
governing the B7b/c bridge. Hence an explanation for the presence of a h23 
nucleotide in the previously described B1b/c pathway may be that the uL5 mutations 
introduced cause a shift in rotational equilibrium, impacting the B7b/c bridge as well. 
We also identified pathways that connect the EF binding site and expansion 
segments ES7L and ES27L (Figure 51). The results from Sections 4.2.4 and 4.2.5, 
and the location of ES7L impart a potential role to this expansion segment in EF 
  129 
 
binding. In addition to this, H41 may contact both the P stalk helices and ES7L, 
allowing information flow. For a potential ES27L pathway, the solvent exposed 
“back” of the SSU, specifically Domain III, may be important. Most nucleotides from 
this domain (which contacts 5.8S rRNA, and contains Helices 47-59 including H50 - 
part of the peptide exit tunnel) are reactive in the factor bound complexes, and lose 
reactivity upon factor dissociation. Thus, a communication pathway between 
PTC/H74 and ES27L may form in the absence of EFs. 
 
Figure 52. Possible network of elongation factor binding related communication pathways in 
yeast ribosomes. 
According to this model, “information” is generally received by the PTC from EF binding sites and is 
then distributed to different parts of the ribosome, including bridges and expansion segments. A. View 
  130 
 
through the tRNA binding sites. PTC may distribute signals to tRNA binding sites, and/or to SSU 
through B7b/c bridge, and/or to the “back” of the ribosome, to Domain III and ES27L. B. View 
through the P stalk. PTC may distribute signals to SSU through B1b/c and/or B6, eB11, eB13 bridges 
as well. There may also be a path from the P stalk / SRL to ES7L that does not employ the PTC. These 
pathways may also be acting in reverse (please see text). 
 
Figure 52 summarizes the possible allosteric information exchange pathways 
that allow different functional regions of the ribosome to communicate with one 
another as the ribosome progresses through the elongation cycle. This model suggests 
that EF binding and/or dissociation may influence individual intersubunit bridge 
conformations, supporting the hypothesis that the ribosome goes through many 
different rotational statuses throughout the elongation cycle. Further work, including 
mutational analyses, probing of ribosomal mutants by hSHAPE, and improvements to 
computational algorithms employed, are required to further map the dynamic features 
of these intra-ribosomal communication networks. 
Pattern searching is a very recent and promising addition to analyses of 
hSHAPE results. Hence there is room for improvement. As noted above, the search 
should not remain limited to reactivity increases or decreases in difference maps, but 
should also include instances of no change. This way, in-depth pathways may be 
obtained, e.g. it may be possible to “connect the dots” from H90 to H74, as this 
information transmission pathway lies at the center of our current model. In fact, 
preliminary manual pattern searching in PTC reveals bases C2821 and U2865-66 as 
nucleotides that lose reactivity in factor bound complexes, lowering the distance 
between H74 and H90 to about 8Å from 30Å. Nucleotide C2821 continues to lose 
reactivity in the posttranslocation complex, while U2865-66 do not show a change in 
reactivity between pre- and postaccommodation complexes, hence not appearing in 
the results of our current pattern search. These and other nucleotides of the PTC 
  131 
 
should be subjected to more complex analyses for patterns of reactivity changes, 
since they may be important not only in bridging the EF binding site to other 
locations but in many more pathways possibly not related to the effects of EF 
binding. 
Even though in the scope of this work we focused on patterns that are quite 
specific to certain events, some of the patterns obtained may be reflective of more 
than one type of rearrangement in the complexes, such as rotational status and / or 
elongation factor binding and / or conformational changes occurring at specific steps 
of elongation (i.e. accommodation or translocation), hence opening our results to 
different interpretations. Still, nucleotides sharing the same pattern through the 
elongation cycle may implicate allosteric changes in reactivity between diverse 
regions of the ribosome, including non-catalytic sites, and should be explored further 
through mutational and structural analyses. 
Finally, our findings may be summarized in a list: 
 mRNA binding results (Table 2, Figure 41) are in accordance with the reported 
and expected changes in the literature 135,200, validating our experimental approach. 
 Elongation factor binding results (Tables 5 and 6) are generally in accordance 
with the reported and expected changes as well 33,201,188,202. There are nucleotides that 
deviate from the expected changes which should be examined further. A major 
unexpected finding is the protection of the P stalk base in the P site occupied 
ribosomes Figure 43. Given the possible communication pathways that connect the P 
stalk base and / or the sarcin-ricin loop to the small subunit (Figures 48-50), this may 
be due to the effects of the rotational status of this complex on the elongation factor 
  132 
 
binding site. This, in turn, would indicate the possibility that the network of 
communication pathways shown in Figure 52 may be active in both directions; 
whereby different regions of the ribosome influence elongation factor binding site 
rearrangements. 
 tRNA binding results (Tables 3 and 4) both revealed nucleotides that act in 
accordance with the reported and expected changes in the literature 127,149, and ones 
that deviated from these. Visual analyses of the yeast P site employing a model tRNA 
175 (Table 13 and Figure 42) revealed the possibility that the P site of the yeast may be 
an intermediate between the bacterial and mammalian P sites (please see above for 
details). 
 Intersubunit bridge results (Tables 7 and 8, Figure 44) were quite different than 
what was expected in the literature 2,203, with the P site occupied complex giving the 
best fit to the list of reported intersubunit contact points in the nonrotated 
conformation. We conclude that the interface between the two ribosomal subunits is 
more “fluid” than what has been reported until now, in X-ray crystallography studies 
and in low resolution cryo-EM studies, revealing the presence of multiple rotated and 
nonrotated conformations (please see above for details). This may be possible by the 
different elongation factor binding site rearrangements we see with various 
complexes, and, in turn, the individual effects of these rearrangements on intersubunit 
bridges, through the revealed network of communication pathways. Molecular 
dynamics simulations of our data (please see Chapter 5: Conclusion and future 
directions) and examination of our elongation complexes with high resolution cryo-
EM may shed more light on the data obtained from the intersubunit bridges. 
  133 
 
 Our data provide novel insight into the rearrangements of the expansion segments 
of the rRNA throughout the elongation cycle (Figure 46), indicating possible roles of 
these in accommodation specifically (please see above for details). The revealed 
network of communication pathways further imply that ES7L rearrangements may be 
related to elongation factor binding (Figure 51). 
 The data are also a rich source for examination of nucleotides that lose or gain 
reactivity together at the same steps of the elongation cycle (Tables 9-12). 
Highlighting these in the three dimensional structure of the yeast ribosome reveals a 
network of possible communication pathways with the elongation factor binding sites 
at the center (Figures 47-52). Our main hypothesis states that, upon elongation factor 
binding, information flows from the factor binding site to the peptidyl transferase 
center, where it is distributed further to different regions of the ribosome. There are 
pathways of information flow that do not pass through the peptidyl transferase center 
as well. A more comprehensive interpretation of all of our major findings indicate 
that these pathways may also be functioning in reverse, such that the rearrangements 
at intersubunit bridges (i.e. the rotational status) of the ribosome may be influencing 
the rearrangements of the elongation factor binding site. 
 
  134 
 
Chapter 5: Conclusion and future directions 
 
Recent structural advances in the disciplines of ribosome structural biology 
and translation have yielded an accurate, high resolution static view of ribosomes 
from different species, including the eukaryotic model organism S. cerevisiae. The 
race is now on to uncover as much information on ribosomal dynamics as possible, to 
generate a more complete view of this highly dynamic machinery and to connect 
structural changes to functional consequences. 
In this project, our aim was to use rRNA flexibility as a tool to obtain 
information on ribosomal dynamics. To this end, our main system was the eukaryotic 
elongation cycle. Comparison of our results with bacterial studies revealed significant 
differences, adding to a line of work that is revealing eukaryotic elongation to be 
more complex in some ways than the bacterial counterpart. Important implications of 
this work include (1) differences in tRNA binding sites, (2) an improved view of 
rotation and intersubunit motions, (3) possible communication networks linking the 
elongation factor binding sites with various regions of the ribosome, including 
intersubunit bridges and eukaryote-specific elements. 
Paving the way to these important observations were modifications to the 
method used to detect changes in rRNA flexibility, hSHAPE. Chapter 2 describes 
these improvements. To the present advantages of this method, including the ease of 
optimization and the high dynamic range of the non-base specific chemicals 
employed, new ones were added, including accurate data normalization and 
decreasees in data loss, making hSHAPE suitable for use with eukaryotic rRNA. 
  135 
 
Chapter 3 shows that the improved hSHAPE analyses can be used to detect 
fine changes in rRNA flexibility, which indicated that the different initiation 
complexes examined resembled the closed, scanning-repressive state of the small 
subunit. We also attempted to locate the binding site of eukaryotic initiation factor 5, 
and observed different candidate sites depending on presence or absence of other 
initiation factors on the small subunit. 
We performed detailed analyses on rRNA derived from complexes resembling 
different steps of translation elongation cycle in Chapter 4. We used functional 
regions, such as tRNA binding sites, to validate our method, then expanded to 
examine an area of active research in the field with analyses of intersubunit bridges, 
and finally delved into the unknown with analyses of the dynamics of the eukaryotic 
expansion segments, and generated preliminary maps of potential allosteric 
communication pathways. 
Our approach, while giving us promising results for future, was not without 
pitfalls. One very important problem is that we worked with heterogeneous 
populations of complexes, as indicated by the results from biochemical assays. The P 
site occupancy of our ribosomes was about 90%; leaving 10% of empty ribosomes 
with each complex we built on top of the P site occupied ones. The A site occupancy 
was even lower at about 60%; leaving about 30% of complexes with only a P site 
tRNA in our experiments. Translocation may have introduced more variables, as the 
efficiency of this was slightly larger than 60%, at 70%, meaning that on top of the 
post-accommodation complexes, eEF2 could have reacted with 10% more of the 
ribosome population in our preparations. These only account for experimental 
  136 
 
heterogeneity; resulting from technical problems with ribosome purification and / or 
binding of various substrates and elongation factors to the ribosome. Cryo-EM and 
sm-FRET studies indicate that there are naturally occurring heterogeneous 
populations of ribosomes, as well: for example there are pre-accommodation 
ribosomes with slightly but noticeably different conformations (e.g. before or after 
codon recognition) 26. Overall, both different natural populations of ribosomes and 
the substoichiometric occupancies of these populations by the substrates of interest 
(eg. A site tRNA binding) may have effects on our rRNA probing results, decreasing 
the signal magnitude. 
The most direct way to address the problem of naturally occurring 
heterogeneous ribosome populations would be to employ single molecule techniques. 
However, there may be relatively simpler solutions to experimental heterogeneity. 
Use of differentially tagged ribosomes and substrates may be an option. Crosslinking 
would be necessary for affinity purification of complexes since complex assembly in 
vitro relies on a state of equilibrium with bound and unbound substrates. Another 
option may be coupling massively parallel sequencing with hSHAPE 204, allowing for 
detection of chemical adducts on single RNA strands, similar to mutational profiling. 
The technique is currently applicable to short RNA molecules, as identifying two 
different primer reads as coming from the same RNA molecule is challenging. For 
hSHAPE of rRNA, optimization could be possible by performing this only on empty 
ribosomes at first, obtaining all possible locations the chemical could bind to in any 
of the reads, especially focusing on identifying regions such as the bridges, and then 
  137 
 
subtracting these from other complexes as we advance. Still, solving this problem will 
be challenging.  
Another problem with the hSHAPE analyses of rRNA is that currently we 
employ 3’ end primers that are complementary to the 3’ end of rRNA, causing us to 
lose some data from these ends. This could be solved by ligation of a short stretch of 
RNA after the hSHAPE reaction and RNA purification, before primer extension, and 
employing 3’ end primers that are complementary to this “linker.” 
A general problem with assembly of translation complexes in vitro is use of 
high magnesium concentrations (10-15 mM), which are actually inhibitory to 
translation compared to in vivo conditions (1-2 mM) 205. The translation inhibition 
may be related to changes in conformation of ribosomes under high magnesium 
levels. This is a current problem with all ribosomal dynamics studies, as it would 
require purification of complexes from cell lysates, which only yields a handful of 
complexes as some states are possibly very short-lived. Another potential problem 
that impacts the value of our results for translation in vivo is our starting complex. 
Due to the relative simplicity of translation initiation in bacteria, studies have been 
conducted where elongation complexes have been built after initiation takes place. 
Replicating this with eukaryotic ribosomes is challenging. In vitro reconstituted 
eukaryotic translation initiation systems are now available, and have been shown to 
allow subunit joining to yield 80S ribosomes 19,206. In fact, they have also been used 
to reconstitute elongation and termination events 207. While it would potentially be 
lucrative to adapt this system to structural probing, it may introduce too much 
heterogeneity to the ribosome populations, and overwhelm the potential solutions to 
  138 
 
the hetereogeneity problem. The use of an mRNA template with Kozak sequence 175, 
while less exciting than using in vitro reconstituted yeast translation systems, may 
provide a slight relief to the problem of “artificial” start complex. This would at least 
mimic the interactions of the rRNA with the full start codon context. 
There are also challenges to the use of hSHAPE data directly for insight into 
ribosomal dynamics. One such challenge is the lack of information on ribosomal 
protein dynamics. Compared to bacterial ribosomes, both the rRNA and the 
ribosomal protein repertoire of the eukaryotic ribosomes increase, with additional 
sequences added to both types of molecules. In fact, the ratio of the total amounts of 
rRNA and proteins in the ribosome increases from 2:1 to 1:1 in yeast. Ribosomal 
proteins are thought to have active roles in eukaryotic translation 180,208–210, and some 
eukaryote-specific proteins engage in eukaryotic intersubunit bridges, indicating that 
proteins may be dictating the intersubunit motions to a larger degree in eukaryotes. 
Added to the fact that the protein components of the dynamic P (L7/L12 in bacteria) 
and L1 stalks are known to be functionally important in all domains of life, data based 
on rRNA may be losing valuable information on ribosomal dynamics. It may be 
possible to recover some of this information through coupling hSHAPE with 
predictive tools. 
Another hSHAPE-specific challenge is the importance of the same regions in 
rRNA for more than one process, e.g. different intersubunit bridges may partially 
overlap with each other or with tRNA binding sites. A similar challenge results from 
the fact that hSHAPE is sensitive to all nucleotide interactions, including stacking and 
basepairing, as well as any other source of stabilization of nucleotides. Hence upon 
  139 
 
encountering reactivity changes, it is difficult to estimate if the change is due mostly 
to variations in the same plane with the nucleotide (i.e. basepairing) or in different 
planes (i.e. stacking and others). This may be solved through careful, computational, 
delineation of all “forces” acting on rRNA nucleotides, and interpretation of data to 
include all possible explanations. Three dimensional visualization and analyses of 
hSHAPE data would be very valuable to overcome the listed protocol-specific 
challenges. 
 
Figure 53. MD-SHAPE analysis of the empty 80S ribosome. 
Please see text for details. The flexibility color code is shown on the right. Proteins are shown in grey. 
Exposed rRNA regions are highly flexible, whereas regions covered by proteins are not. These 
analyses, with improved coloring, will be very useful in determination of the underlying causes of 
flexibility changes, including protein-RNA interactions as potential causes of such changes. Image 
generated by our collaborators, Dr. Serdal Kirmizialtin and Dr. Karissa Sanbonmatsu (LANL). 
 
Along these lines, one of the current directions of this project is to use our 
data to impose additional constraints in molecular dynamics simulations. As 
explained in Chapter 1, the simulations require calculation of the total force acting on 
  140 
 
each atom and molecule of a system and tracking of the changes in positions of said 
atoms and molecules in a time-dependent manner. MD simulations are initially based 
on published structures resembling the complexes of interest. We suggest that they 
can be fine-tuned by the high resolution hSHAPE data. hSHAPE reactivities may be 
added to the total force calculations by conversion into “pseudo”-free energy values 
116,211, in the same way they are used for structure prediction. Since the simulations 
are based on other structural studies, they are useful in predicting ribosomal protein 
dynamics as well, and nicely indicate whole domain movements vs. base flipping in 
rRNA. As such, MD-SHAPE will be valuable in generating a “movie” of ribosomal 
dynamics changes in a set of elongation complexes. Automated mining of the three 
dimensional information contained in simulations would also be another direction to 
consider to obtain lists of position changes of all nucleotides allowing classifications 
of different motions and simplifying comparisons of simulations of different 
complexes. Purely visual comparisons are also possible directly with MD-SHAPE. 
While MD simulations are most valuable as trajectories of movements, Figure 53 
shows a static view of the empty ribosome after MD-SHAPE, implicating protein-
RNA interactions as major causes of the flexibility differences throughout the 
ribosome as exposed surfaces are highly flexible. Hence MD simulations will 
possibly be useful in identifying the reasons underlying flexibility changes in 
elongation complexes as well. 
Another computational direction we would like to take with this project is to 
define the hSHAPE reactivities in voxels - volumetric pixels of a certain size. This is 
both an attempt to simplify and improve visualization of our data, and definition of 
  141 
 
the voxels in hydrogen-bonding distances (i.e. about 3.5 x 3.5 x 3.5 Å3 to contain 
interacting nucleotides in the same voxel) would allow us to have a sense of which 
interactions may be rearranging as the reactivities change from one complex to the 
next. Figure 54 shows visualization of our data in voxels of 6 x 6 x 6 Å3 coupled to 
volumetric rendering highlighting the maximum flexibility value among the group of 
nucleotides contained in a single voxel. According to this, in post-accommodation 
complex the 25S rRNA undergoes rearrangements whereas in the pre-translocation 
complex it is the 18S rRNA that undergoes changes. Even at this preliminary 
resolution, this analysis is already useful in visualization. Obtaining information on 
which nucleotides are “trapped” in each voxel would allow for further analyses and 
discovery of communication pathways throughout the ribosome in the context of 
translation elongation. 
 
Figure 54. The hSHAPE data visualized in voxels. 
  142 
 
25S and 18S rRNAs are visualized from the P stalk side. The entire volume of the ribosome is divided 
into voxels of 6Å dimensions. For each voxel, the maximum flexibility value among the nucleotides 
contained inside is computed. In the image, volumetric rendering is used to show maximum flexibility 
values of each voxel. The more flexible a voxel is, the more opaque it is. The color code for the 
flexibilities is shown on the right. Image generated by our collaborators, Dr. Sujal Bista and Dr. 
Amitabh Varshney (UMIACS, UMD). 
 
Other than the computational directions we are starting to take through 
collaborations with Los Alamos National Laboratories and Computer Sciences 
Department at University of Maryland, a future direction of this project is to use the 
biochemical and structural tools that were employed here, for “walking” mutant 
ribosomes through the elongation cycle. This could potentially be a part of the 
mutational analyses of ribosomal mutants, and would show (1) at precisely which 
step the mutant ribosome performs poorly and (2) due to which conformational 
change, compared to wild type ribosomes. Currently, hSHAPE is being used on only 
empty mutant ribosomes. Ribosomal mutants may also be employed to check for the 
authenticity of the communication pathways obtained computationally. In fact, the 
current repertoire of ribosomal mutants from our laboratory could be employed for 
this initially. Another way to validate these communication pathways may be through 
repeating our project with ribosomes from other organisms. 
The results of this project identify hSHAPE as a valuable technique for studies 
of ribosomal dynamics. Improvement to address the issues listed above, as well as 
following through with the future directions, has the potential to take similar studies 
to a whole new level, finally obtaining a high resolution dynamic map of the yeast 
ribosome. 
 
  143 
 
Chapter 6: Materials and methods 
Strains.  
Yeast strain JD1370 (MATa trp1 ura3 leu2 PEP4::HIS3 NUC1::LEU2) was 
used for purification of ribosomes. Yeast strains TKY865 (renamed to JD1540; 
MATα leu2-3,112 his4-713 ura3-52 trp1Δ tef2Δ2 tef1::LEU2 met2-1 pTKB779 (TRP1 
2u TEF1-His6) and TKY675 (renamed to JD1334; MATα ade2 leu2 ura3 his3 trp1 
eft1::HIS3 eft2::TRP pEFT1-LEU2 -CEN) were kind gifts from Dr. Terry G. Kinzy 
(Rutgers University) and were used to purify eukaryotic elongation factors 1A and 2 
(eEF1A and eEF2), respectively. Bacterial strain M15 pREP4-pJD833, cloned by Dr. 
Sharmishtha Musalgaonkar from our laboratory, was used to purify phenylalanyl 
tRNA synthetase.  
Ribosome preparation.  
Ribosome purification was essentially performed as previously described 212 
followed by puromycin treatment to remove tRNAs 213. In brief, yeast strain JD1370 
was grown to an OD595 of 1.2 in 2L of YPAD. Cells were broken using 0.5 mm 
glass beads at 4°C with a Biospec Mini bead beater. Cellular debris was removed by 
centrifugation at 30,000xg for 30 minutes in a Beckman-Coulter Optima Max E 
ultracentrifuge. The S30 supernatant was purified using Sulfolink resin. After elution 
from the column, ribosomes were treated with 1 mM GTP and 1 mM puromycin (pH 
7.0) at 30°C for 30 minutes. This was followed by overnight centrifugation in a high 
salt glycerol cushion (20 mM HEPES-KOH pH 7.6, 60 mM NH4Cl, 500 mM KCl, 10 
mM Mg(OAc)2, 2 mM DTT, 25% glycerol) at 100,000xg. The ribosome pellets were 
resuspended in the presence of 20 mM Mg++ and incubated at 4°C for 1 hour to 
  144 
 
ensure maximal subunit association. The resuspension, diluted 1:1 in a buffer without 
Mg++ to lower the final concentration of the ion to 10 mM, was passed through a 10-
30% sucrose cushion (20 mM HEPES-KOH pH 7.6, 100 mM KCl, 10 mM 
Mg(OAc)2, 2 mM DTT) at 45,000xg for 15 hours. 80S fractions were collected. After 
buffer exchange (to 20 mM HEPES-KOH pH 7.6, 50 mM NH4Cl, 10 mM 
Mg(OAc)2, 1 mM DTT, 25% glycerol), 80S ribosomes were quantified and stored at -
80°C. 
Elongation factor purification.  
Yeast strains JD1540 and JD1334 were grown to an OD595 of 1.5 in 2L of 
YPAD. Cell lysis was essentially the same as the ribosome purification procedure, 
except the buffer composition was 20 mM sodium phosphate buffer, pH 7.6, 500 mM 
NaCl and 10 mM imidazole. Cellular debris and ribosomes were removed by 
centrifugation at 100,000xg for 3 hours at 4°C. The supernatants were filtered 
through a 0.2 micron syringe filter. Filtered eEF1A supernatant was mixed and 
incubated with Qiagen Ni-NTA agarose and washed under gravity flow. His-tagged 
eEF1A was eluted in presence of 250 mM imidazole. Eluates from different 
preparations were pooled. Imidazole was removed by buffer exchange. eEF1A 
activity was determined by A site tRNA binding assays (described below). Filtered 
eEF2 supernatant was passed through a 5-ml GE Healthcare HisTrap HP column and 
washed using a peristaltic pump. His-tagged eEF2 was eluted in presence of 250 mM 
imidazole. Imidazole was removed by buffer exchange. eEF2 activity was determined 
through ADP ribosylation by diphtheria toxin and ribosome binding assay 210. Protein 
concentrations were determined by Bradford assay. 
  145 
 
Aminoacyl-tRNA synthetase purification. 
Bacterial strain M15 pREP4-pJD833 was grown to OD600 = 0.5 and induced 
by 1mM IPTG for six hours prior to collection of cell pellets. Cells were washed with 
20 mM sodium phosphate buffer, pH 7.6, 500 mM NaCl, 10 mM imidazole. 
Lysozyme was added to a concentration of 0.1 mg/ml and sonication was employed 
to lyse cells. The rest of the procedure was essentially the same as elongation factor 
purification. Briefly, cellular debris was removed by ultracentrifugation and the 
supernatant was incubated with Ni-NTA agarose and washed under gravity flow. His-
tagged aminoacyl-tRNA synthetase was eluted in the presence of 250 mM imidazole. 
Imidazole was removed by buffer exchange. Protein concentration was determined by 
Bradford assay. The activity was determined by small scale aminoacylation reactions 
and detected as incorporation of [14C] phenylalanine, under increasing protein 
amount. 
tRNA charging. 
 
CCA-adding enzyme was purified by Dr. Alexey Petrov, an alumnus of our 
laboratory, by salting out different fractions of S30 supernatants of Baker’s yeast. The 
activity was determined by Dr. Sharmishtha Musalgaonkar based on quantification of 
[α-32P] ATP incorporation in both time-course assays and under increasing enzyme 
amount. 
Yeast tRNAPhe was incubated with 10x ATP and CTP in 50 mM Glycine-
NaOH, pH 9.0, 20 mM MgCl2 and 1 mM DTT. 0.3 µg of CCA-adding enzyme was 
added to the reaction per pmole tRNA. Pyrophosphatase was also employed to 
remove the inhibitory byproduct pyrophosphate from the reaction (one unit liberates 1 
  146 
 
µmole of inorganic orthophosphate per min at pH 7.2 at 25°C). The reaction took 
place at 30°C for 24 min. 300 mM sodium acetate, pH 5.0, was used to stop the 
reaction, as well as phenol-chloroform extraction to remove proteins. CCA-repaired 
tRNA was collected by overnight ethanol precipitation and resuspension of pellet in 3 
mM sodium acetate. When pyrophosphatase was not used, this solution was passed 
through a G25 column to remove pyrophosphate before moving on to aminoacylation. 
CCA-repaired tRNA was incubated with 10x non-radioactive phenyalanine or 
3x [14C] phenylalanine, in 100 mM HEPES-KOH, pH 7.6, 10 mM ATP, 10 mM KCl, 
20 mM MgCl2 and 1 mM DTT. 0.5 µg of aminoacyl-tRNA synthetase was added per 
pmole of tRNA. Pyrophosphatase was also employed. The reaction took place at 
30°C for 30min. The next steps were similar to the CCA repair procedure, to obtain 
phenylalanyl tRNA (phenol-chloroform extraction, ethanol precipitation). The tRNA 
pellet was resuspended in 3 mM sodium acetate and 10 mM MgCl2. To obtain N-
acetyl-phenylalanyl tRNA, directly after stopping aminoacylation, fresh acetic 
anhydride was added to a concentration of 5% (V/V). This was incubated on ice for 1 
hour. Addition of acetic anhydride and incubation were repeated, and phenol-
chloroform extraction and ethanol precipitation proceeded. The tRNA pellet was 
resuspended in 3 mM sodium acetate and 10 mM MgCl2. 
High performance liquid chromatography (HPLC) was employed to purify 
aminoacylated or acetylated and aminoacylated tRNA from deacyl tRNA. A 
Bakerbond Wide-Pore™ C4 column (250 x 4.6 mm) was equilibrated with buffer A 
(20 mM ammonium acetate pH 5.0, 400 mM NaCl and 10 mM magnesium acetate). 
Sample was injected and eluted over a linear gradient of 0-50% buffer B (buffer A in 
  147 
 
60% methanol) with a flow rate of 0.5 ml/min. The specific steps are: 1. buffer A for 
15 min, 2. linear gradient of buffer B from 0-10% until the 20th min, 3. linear gradient 
of buffer B from 10-50% until the 90th min, 4. run 50% buffer B for 10 min. The 
aminoacylated fractions eluted around 85 min, the acetylated-aminoacylated fractions 
around 95 min, and the deacylated tRNA fractions around 60min. Fractions of 
interests were pooled, ethanol precipitated and resuspended, and stored at -80°C. 
Depending on the phenylalanine species, either concentration or radioactivity was 
measured to quantify the charged tRNAs. 
Ribosome/tRNA interactions.  
To assay binding of aminoacyl-tRNA to the ribosomal A-site, 2 sets of 
reactions were set up in parallel. A mix containing 100 µg of polyuridylic acid 
(poly(U)), 500 pmoles of ribosomes, a 4-fold molar excess of tRNAPhe, all in binding 
buffer (80 mM Tris-HCl, pH 7.4 at 30o C, 160 mM NH4Cl, 15 mM magnesium 
acetate, 2 mM spermidine, 0.5 mM spermine, 6 mM β-mercaptoethanol) in 150 µl 
total volume was prepared and incubated for 30 min at 30°C to block the P-site. To 
prepare the ternary complex ([14C]Phe-tRNAPhe•eEF1A•GTP), 100 µg of eEF1A, 
1mM GTP, 512 pmoles [14C]-Phe-tRNAPhe were mixed in 60 µl total volume binding 
buffer and incubated for 30 min at 30⁰C. After incubation, serial 2-fold dilutions of 
the ternary complex reaction mix were prepared, resulting in 8 fractions containing 
decreasing amounts of ternary complex (128 - 1 pmoles), in 15 µl each. An equal 
amount of the ribosome mix (25 pmoles of ribosomes, 15 µl) was added to each 
dilution, followed by incubation for 30 min at 30oC. The mixtures were applied onto 
pre-wetted nitrocellulose Millipore HA (0.45 micron) filters, washed with binding 
  148 
 
buffer, and radioactivity was measured via scintillation counting. Background control 
reactions without ribosomes were performed at each ligand dilution and subtracted 
from experimental ones. Binding data were fitted into “single binding site with ligand 
depletion” models using GraphPad Prism. To test binding of tRNA to the P-site, N-
acetyl-[14C]-Phe-tRNAPhe was used as the ligand, without prior incubation with 
tRNAPhe, and 11 mM magnesium acetate was used in the binding buffer. Assays were 
conducted for each complex.   
Ribosome/eEF2 interaction.  
5 pmoles ribosomes were incubated with increasing concentrations of eEF2 
(0.5 – 64 pmoles), 10 µg of polyU, and 4x molar excess of [14C] NAD over ribosomes 
in 50 µl total volume of binding buffer (50 mM Tris-HCl pH 7.5, 50 mM ammonium 
acetate, 10 mM magnesium acetate, 2 mM DTT, 100 μM GDPNP) at 30°C for 20 
minutes. Diphtheria toxin (0.2 µg) was added, and reactions were incubated for 30 
min at 30°C. After precipitation with TCA (final concentration 15% V/V) and 15 min 
incubation on ice, reaction mixtures were applied onto GF/C filters, washed with 5% 
TCA, and the amount of [14C]-ADP ribosylated eEF2 was determined by scintillation 
counting. Counts resulted from unbound eEF2 and were subtracted from total eEF2 to 
obtain the amounts bound.  
Ribosomal complex assembly for hSHAPE analyses.  
Elongation: 50 pmoles of 80S yeast ribosomes programmed with 100 µg 
polyuridine were used for all complexes. Complex assembly conditions were based 
on binding assays. To assemble P-site occupied ribosomes, 150 pmoles of N-acetyl-
phenylalanyl (Phe) tRNAPhe were incubated with ribosomes in binding buffer (80 mM 
  149 
 
HEPES pH 7.5, 50 mM NaCl, 11 mM magnesium acetate, 6 mM β-mercaptoethanol) 
at 30°C for 30 minutes. For pre- and post-accommodation ribosome assembly, P-site 
occupied ribosomes were incubated with a ternary complex of eEF1A-GDPNP-Phe-
tRNAPhe or eEF1A-GTP-Phe-tRNAPhe, respectively, in binding buffer with 15 mM 
magnesium acetate at 30°C for 30 minutes. Ternary complexes were prepared by 
incubating 250 pmoles of Phe-tRNAPhe, 1 nmole of eEF1A and 1 mM GDPNP or 
GTP. The hybrid states complex was prepared by incubating post-accommodation 
ribosomes with 1 mM anisomycin. The accommodation intermediate complex and the 
classical states complex were prepared by incubating P-site occupied ribosomes with 
1 mM anisomycin or 1 mM cycloheximide, respectively, before addition of the 
ternary complex and then proceeding with the remainder of post-accommodation 
ribosome preparation. To assemble pre- and post-translocation ribosomes, post-
accommodation ribosomes were prepared in large scale (starting with 500 pmoles 
ribosomes) and centrifuged for 2 hours at 4°C at 100,000xg through 25% glycerol 
cushion in binding buffer with 15 mM magnesium acetate. The ribosome complex 
pellets were resuspended in binding buffer with 15 mM magnesium acetate and 
quantified. 50 pmoles were incubated with 300 pmoles eEF2 and 1 mM GDPNP or 
GTP. 
Initiation: These complexes were assembled by Fan Zhang (NIH). 50 pmoles 
of empty 40S was used as baseline. 100 pmoles eIF1, 100 pmoles eIF1A, and 250 
pmoles eIF5 were added to 50 pmoles 40S for assembly of the complexes 40S-eIF1-
eIF1A and 40S-eIF1-eIF1A-eIF5. The reactions proceeded in reconstitution buffer 
(30 mM HEPES-KOH pH7.4, 100 mM potassium acetate, 3 mM magnesium acetate 
  150 
 
and 2 mM DTT) for 15 min at 26°C. 150 pmoles of full length eIF5 or 150 pmoles of 
eIF5-NTD or 450 pmoles of eIF5-CTD were added to 50 pmoles 40S for assembly of 
40S-eIF5, 40S-eIF5-NTD, or 40S-eIF5-CTD, respectively. The reactions proceeded 
in reconstitution buffer for 6 min at 26°C. 
Dipeptide formation analysis by HPLC. 
The analysis of dipeptide formation by HPLC was adapted from Marquez et 
al, 2004 172. Complexes to be tested were assembled using 50 pmoles P-site occupied 
ribosomes and ternary complexes made of 250 pmoles [14C]-Phe-tRNAPhe, 1 nmole 
eEF1A and 1mM GTP or GDPNP. Phenol-chloroform extraction was performed. The 
aqueous phase was collected and was used in ethanol precipitation overnight. Pellets 
obtained from ethanol precipitation were treated with 0.5M NaOH to hydrolyze RNA. 
The mixtures were neutralized with 1M HCl, loaded onto a Thermo Scientific 
Hypersil™ C18 column (150 x 4.6 mm) equilibrated with 0.1% trifluoroacetic acid 
(TFA), and were subjected to reverse phase HPLC. A binary linear gradient was 
applied at 0.5 ml/min using 0.1% TFA as Buffer A and 60% acetonitrile in 0.1% TFA 
as Buffer B. Fractions of 900 µl were collected and the radioactivity measured by 
liquid scintillation. The results were plotted. 
Translocation assay. 
The puromycin assay to test for translocation was adapted from Rheinberg 
and Nierhaus, 1986, and Nierhaus, 1990 173,174. 250 pmoles post-accommodation 
complex prepared with [14C]-Phe-tRNAPhe tRNA were passed through A-site tRNA 
binding buffer (80 mM Tris-HCl, pH 7.4 at 30o C, 160 mM NH4Cl, 15 mM 
Mg(OAc)2, 2 mM spermidine, 0.5 mM spermine, 6 mM β-mercaptoethanol) with 25% 
  151 
 
glycerol at 100,000xg for 2 hours, to remove unbound tRNAs and the eEF1A. The 
amount of complex collected was measured through detecting radioactivity by 
scintillation counting. For the translocation assay, 5 pmoles of radioactively labeled 
post-accommodation complex was incubated with 35 pmoles eEF2 with 1 mM GTP, 
or 35 pmoles eEF2 with 1 mM GDPNP, or no eEF2 (used as background control). 
Translocation proceeded for 5 minutes at 30°C. The puromycin reaction was 
conducted at 4°C 173,174. 10 mM puromycin was added to each tube and the reactions 
were stopped at 0, 5, 10, 20 and 40 min by ethyl acetate extraction. The organic layer 
containing puromycin was collected and the radioactivity was measured by 
scintillation counting. 
1M7 treatment.  
1M7 treatment was performed as previously described 210. The complexes 
were divided into two aliquots. For “reaction” samples, 1M7 was added to a final 
concentration of 6mM. For “control” samples, the same volume of anhydrous DMSO 
was added. Reactions proceeded for 10 minutes at 30°C and subsequently were 
precipitated overnight in 70% ethanol and 300 mM sodium acetate. rRNAs were 
purified using RNAqueous Micro kit (Ambion). rRNA concentrations were 
determined using a Nanodrop 1000. 
Kethoxal treatment.  
The complexes to be tested were prepared from 100 pmoles of ribosomes and 
divided into four aliquots. For “reaction” samples, 0.5 µl, 1 µl, and 2 µl of kethoxal 
(diluted 1:100 in ethanol) were used. For “control” samples, 1 µl ethanol was used. 
Reactions proceeded for 10 minutes at 30°C and were quenched in 250 mM 
  152 
 
potassium borate, pH 7.0, and 150 mM sodium acetate. rRNAs were precipitated, 
purified and quantitated as described above. 
hSHAPE extension.  
Primer extensions were performed as previously described 210. Four primers 
were constructed by Dr. Jonathan Leshin, an alumnus of our laboratory, with the 
following fluorescent molecules for each region: 6-FAM labeled primers were used 
for the 1M7 treated samples, VIC labeled primers were used for the DMSO treated 
samples, and NED and PET labeled primers were used for sequencing reactions. To 1 
µg of rRNA, 2.5 pmoles of primer was added. The samples were incubated at 65°C 
for 5 minutes then 50°C for 5 minutes. A master mix was made of 5 units Superscript 
III Reverse Transcriptase (Invitrogen), 0.5 mM dNTP mixture, 1X Superscript Buffer 
(Invitrogen) and 5 mM DTT and added to the samples to a final volume of 10 µl. In 
addition, 1 mM of appropriate ddNTP was added to the sequencing samples.105 
Samples were incubated at 50°C for one minute, 52°C for 45 minutes and 65°C for 5 
minutes. The four extension reactions for each region were combined and precipitated 
overnight in 70% ethanol at -20°C. Samples were washed twice with 70% ethanol 
and resuspended in Hi-Di Formamide (Applied Biosystems) and sent for fragment 
analyses using capillary electrophoresis at Genewiz.  
ShapeFinder analysis.  
hSHAPE data were aligned to rRNA sequences using ShapeFinder 210,105,106. 
Fitted baseline adjustment, mobility shift correction, signal decay correction, scaling, 
alignment features of the software were employed and the aligned peaks were fitted 
into a Gaussian function. Integrated peak values were obtained as the areas under 
  153 
 
1M7 treated peaks minus the areas under corresponding negative control peaks, and 
used for further analyses (please see below).  
Box-plot analysis.  
Integrated peak values (1M7 reactivity values for every nucleotide of the 
ribosome) obtained from ShapeFinder were corrected for negative values and 
analyzed by box-plots in R 3.1.0 using robust statistics. The following features of the 
integrated peak data (x1-xn) were determined: the minimum allowed value (Qmin), the 
25th percentile (Q25), the sample median (x̃), the 75
th percentile (Q75), the maximum 
allowed value (Qmax), and the interquartile range (IQR), where: 
 
    x((n+1)/4),   if n is odd      1 
Q25 =     xn/4 + xn/4+1,   if n is even 
           2 
 
 x((n+1)/2),   if n is odd      2 
x̃ =  xn/2 + xn/2+1,   if n is even 
        2 
 
 
    x(3(n+1)/4),   if n is odd      3 
Q75 =     x3n/4 + x3n/4+1,  if n is even 
            2 
 
 
IQR = Q75 - Q25         4 
 
Qmin = Q25 - 1.5 * IQR         5 
 
Qmax = Q75 + 1.5 * IQR        6 
 
As nonreactive nucleotides (i.e. integrated peak values of “zero”) were 
expected, outliers were identified as values above Qmax and were removed in the first 
  154 
 
version of hSHAPE data analyses. A need for data transformation was evident from 
the box plot analyses, and led to improved data analyses as explained below. 
Data processing and normalization. 
Natural log transformation was employed for data processing for improved 
analyses. For each data point in the dataset x1-xn (zero-corrected integrated peak 
values from a single primer region), the transformed values y1-yn were defined as: 
y1,…,yn = ln((x1,…,xn)+c), where c=1       7 
The transformed datasets y1-yn from each primer region were normalized by 
the median for that region and merged to obtain transformed, normalized data for the 
whole yeast ribosomal RNA except 5S rRNA. The merged data were analyzed by 
box-plots and resulted in unimodal, symmetric distributions. Outliers were identified 
and removed as described above.  
Complex comparison.  
To compare reactivity values from different 80S ribosomal complexes, the 
following calculations were performed: 
Complex vs. Empty 80S =    (Complex - Empty 80S), per nucleotide  8 
             Empty 80S 
 
Complex A vs. Complex B = (Complex A - Complex B), per nucleotide  9 
              Empty 80S 
Data grouping and frequency histograms.  
For visualization purposes, a color code was employed that represented ranges 
of comparative reactivity values, or reactivity levels. 15 levels were chosen by 
grouping the data based on frequency in MS Excel and plotting the corresponding 
histograms in R 3.1.0 (Figures 22 and 23). According to this, the values obtained 
  155 
 
from the above complex comparison calculations lay in the interval -15 to 15 for all 
complexes. Frequency intervals were used to determine the relative reactivity levels. 
Level “0” (indicating no change between two complexes) comprised most of the data 
points (highest bars, values between -0.25 to 0.25). The most extreme levels “-7” and 
“7” (indicating a large decrease in reactivity and a high increase in reactivity, 
respectively, in one complex compared to the other) included the smallest number of 
data points (values below -7 and above 7, respectively). All other levels included 
approximately the same number of data points. The resulting color scale is shown in 
Figure 22B in the order of relative levels, colors, and comparative reactivity ranges. 
Generation of difference maps of eukaryotic rRNA.  
The color-coded comparative reactivity values were mapped onto the 3.0 Å X-
ray crystallography structure of Saccharomyces cerevisiae ribosome 2 using a 
PyMOL script 105, and the secondary structure mapping was performed by the online 
tool, Ribovision, following the instructions of the developers 11. 
Pattern discovery on the hSHAPE data. 
MS Excel was employed to examine patterns of reactivity changes in the 
hSHAPE data. 32 patterns were scanned for in five difference maps (P site occupied 
vs empty, preaccommodation vs P site occupied, postaccommodation vs 
preaccommodation, pretranslocation vs postaccommodation, posttranslocation vs 
pretranslocation). Only reactivity decreases and increases in sequence were scanned 
for, and made up the 25 (=32) possible patterns. 
 
  156 
 
Appendix: 2D difference maps 
 
 
  157 
 
 
  158 
 
 
 
  159 
 
 
  160 
 
 
 
  161 
 
 
  162 
 
 
  163 
 
 
 
  164 
 
 
  165 
 
 
 
  166 
 
Bibliography 
1. Melnikov, S. et al. One core, two shells: bacterial and eukaryotic ribosomes. 
Nat. Struct. Mol. Biol. 19, 560–7 (2012). 
2. Ben-Shem, A. et al. The structure of the eukaryotic ribosome at 3.0 Å 
resolution. Science 334, 1524–9 (2011). 
3. Jack, K. et al. rRNA pseudouridylation defects affect ribosomal ligand binding 
and translational fidelity from yeast to human cells. Mol. Cell 44, 660–666 
(2011). 
4. Baxter-Roshek, J. L., Petrov, A. N. & Dinman, J. D. Optimization of ribosome 
structure and function by rRNA base modification. PLoS One 2, e174 (2007). 
5. Decatur, W. A. & Fournier, M. J. rRNA modifications and ribosome function. 
Trends Biochem. Sci. 27, 344–351 (2002). 
6. Persaud, C. et al. Mutagenesis of the modified bases, m(5)U1939 and psi2504, 
in Escherichia coli 23S rRNA. Biochem. Biophys. Res. Commun. 392, 223–7 
(2010). 
7. Kimura, S. & Suzuki, T. Fine-tuning of the ribosomal decoding center by 
conserved methyl-modifications in the Escherichia coli 16S rRNA. Nucleic 
Acids Res. 38, 1341–52 (2010). 
8. Noller, H. F. RNA structure: reading the ribosome. Science (80-. ). 309, 1508 
(2005). 
9. Moore, P. B. Structural motifs in RNA. Annu. Rev. Biochem. 68, 287–300 
(1999). 
10. Petrov, A. S. et al. Secondary structures of rRNAs from all three domains of 
life. PLoS One 9, e88222 (2014). 
11. Bernier, C. R. et al. RiboVision suite for visualization and analysis of 
ribosomes. Faraday Discuss. 169, 195–207 (2014). 
12. Metagenomics: Current Innovations and Future Trends. 295 (Horizon 
Scientific Press, 2011). at 
<http://books.google.com/books?id=YGc1P0RiAOgC&pgis=1> 
13. Farwell, M. A., Roberts, M. W. & Rabinowitz, J. C. The effect of ribosomal 
protein S1 from Escherichia coli and Micrococcus luteus on protein synthesis 
in vitro by E. coli and Bacillus subtilis. Mol. Microbiol. 6, 3375–83 (1992). 
  167 
 
14. Bernstein, J. R., Bulter, T., Shen, C. R. & Liao, J. C. Directed evolution of 
ribosomal protein S1 for enhanced translational efficiency of high GC 
Rhodopseudomonas palustris DNA in Escherichia coli. J. Biol. Chem. 282, 
18929–36 (2007). 
15. Simonetti, A. et al. A structural view of translation initiation in bacteria. Cell. 
Mol. Life Sci. 66, 423–36 (2009). 
16. Simonetti, A. et al. Structure of the 30S translation initiation complex. Nature 
455, 416–420 (2008). 
17. Marintchev, A. in Biophysical approaches to translational control of gene 
expression 83–101 (2013). 
18. Jackson, R. J., Hellen, C. U. T. & Pestova, T. V. The mechanism of eukaryotic 
translation initiation and principles of its regulation. Nat. Rev. Mol. Cell Biol. 
11, 113–27 (2010). 
19. Acker, M. G., Kolitz, S. E., Mitchell, S. F., Nanda, J. S. & Lorsch, J. R. 
Reconstitution of yeast translation initiation. Methods Enzymol. 430, 111–45 
(2007). 
20. Pestova, T. V & Kolupaeva, V. G. The roles of individual eukaryotic 
translation initiation factors in ribosomal scanning and initiation codon 
selection. Genes Dev. 16, 2906–22 (2002). 
21. Acker, M. G., Shin, B.-S., Dever, T. E. & Lorsch, J. R. Interaction between 
eukaryotic initiation factors 1A and 5B is required for efficient ribosomal 
subunit joining. J. Biol. Chem. 281, 8469–75 (2006). 
22. Ceci, M. et al. Release of eIF6 (p27BBP) from the 60S subunit allows 80S 
ribosome assembly. Nature 426, 579–84 (2003). 
23. Sengupta, J. et al. Identification of the versatile scaffold protein RACK1 on the 
eukaryotic ribosome by cryo-EM. Nat. Struct. Mol. Biol. 11, 957–962 (2004). 
24. Nilsson, J., Sengupta, J., Frank, J. & Nissen, P. Regulation of eukaryotic 
translation by the RACK1 protein: a platform for signalling molecules on the 
ribosome. EMBO Rep. 5, 1137–41 (2004). 
25. Paek, K. Y. et al. Translation initiation mediated by RNA looping. Proc. Natl. 
Acad. Sci. 112, 201416883 (2015). 
26. Budkevich, T. V et al. Regulation of the mammalian elongation cycle by 
subunit rolling: a eukaryotic-specific ribosome rearrangement. Cell 158, 121–
31 (2014). 
  168 
 
27. Youngman, E. M., Cochella, L., Brunelle, J. L., He, S. & Green, R. Two 
distinct conformations of the conserved RNA-rich decoding center of the small 
ribosomal subunit are recognized by tRNAs and release factors. Cold Spring 
Harb. Symp. Quant. Biol. 71, 545–9 (2006). 
28. Demeshkina, N., Jenner, L., Westhof, E., Yusupov, M. & Yusupova, G. A new 
understanding of the decoding principle on the ribosome. Nature 1–5 (2012). 
doi:10.1038/nature10913 
29. Ogle, J. M., Murphy, I. V, Frank, V., Tarry, M. J. & Ramakrishnan, V. 
Selection of tRNA by the ribosome requires a transition from an open to a 
closed form. Cell 111, 721–732 (2002). 
30. Ogle, J. M., Carter, A. P. & Ramakrishnan, V. Insights into the decoding 
mechanism from recent ribosome structures. Trends Biochem. Sci. 28, 259–66 
(2003). 
31. Jenner, L., Demeshkina, N., Yusupova, G. & Yusupov, M. Structural 
rearrangements of the ribosome at the tRNA proofreading step. Nat. Struct. 
Mol. Biol. 17, 1072–8 (2010). 
32. Demeshkina, N., Jenner, L., Westhof, E., Yusupov, M. & Yusupova, G. New 
structural insights into the decoding mechanism: translation infidelity via a 
G·U pair with Watson-Crick geometry. FEBS Lett. 587, 1848–57 (2013). 
33. Schmeing, T. M. et al. The crystal structure of the ribosome bound to EF-Tu 
and aminoacyl-tRNA. Science 326, 688–94 (2009). 
34. Hiller, D. a, Singh, V., Zhong, M. & Strobel, S. a. A two-step chemical 
mechanism for ribosome-catalysed peptide bond formation. Nature 476, 2–6 
(2011). 
35. Kuhlenkoetter, S., Wintermeyer, W. & Rodnina, M. V. Different substrate-
dependent transition states in the active site of the ribosome. Nature 476, 351–
4 (2011). 
36. Rodnina, M. V. The ribosome as a versatile catalyst: reactions at the peptidyl 
transferase center. Curr. Opin. Struct. Biol. 23, 595–602 (2013). 
37. Borman, S. Protein Factory Reveals its Secrets. Chem. Eng. News 85, 13–16 
(2007). 
38. Zaher, H. S., Shaw, J. J., Strobel, S. a & Green, R. The 2’-OH group of the 
peptidyl-tRNA stabilizes an active conformation of the ribosomal PTC. EMBO 
J. 1–9 (2011). doi:10.1038/emboj.2011.142 
  169 
 
39. Bieling, P., Beringer, M., Adio, S. & Rodnina, M. V. Peptide bond formation 
does not involve acid-base catalysis by ribosomal residues. Nat. Struct. Mol. 
Biol. 13, 423–428 (2006). 
40. Polacek, N., Gaynor, M., Yassin, A. & Mankin, A. S. Ribosomal peptidyl 
transferase can withstand mutations at the putative catalytic nucleotide. Nature 
411, 498–501 (2001). 
41. Lang, K., Erlacher, M., Wilson, D. N., Micura, R. & Polacek, N. The role of 
23S ribosomal RNA residue A2451 in peptide bond synthesis revealed by 
atomic mutagenesis. Chem. Biol. 15, 485–92 (2008). 
42. Voorhees, R. M. & Ramakrishnan, V. Structural basis of the translational 
elongation cycle. Annu. Rev. Biochem. 82, 203–36 (2013). 
43. Weinger, J. S., Parnell, K. M., Dorner, S., Green, R. & Strobel, S. A. Substrate-
assisted catalysis of peptide bond formation by the ribosome. Nat. Struct. Mol. 
Biol. 11, 1101–1106 (2004). 
44. Youngman, E. M., Brunelle, J. L., Kochaniak, A. B. & Green, R. The active 
site of the ribosome is composed of two layers of conserved nucleotides with 
distinct roles in peptide bond formation and peptide release. Cell 117, 589–599 
(2004). 
45. Frank, J. & Agrawal, R. K. A ratchet-like inter-subunit reorganization of the 
ribosome during translocation. Nature 406, 318–22 (2000). 
46. Julián, P. et al. Structure of ratcheted ribosomes with tRNAs in hybrid states. 
Proc. Natl. Acad. Sci. U. S. A. 105, 16924–7 (2008). 
47. Gao, Y.-G. et al. The structure of the ribosome with elongation factor G 
trapped in the posttranslocational state. Science 326, 694–9 (2009). 
48. Ortiz, P. A., Ulloque, R., Kihara, G. K., Zheng, H. & Kinzy, T. G. Translation 
elongation factor 2 anticodon mimicry domain mutants affect fidelity and 
diphtheria toxin resistance. J. Biol. Chem. 281, 32639 (2006). 
49. Frank, J. Intermediate states during mRNA-tRNA translocation. Curr. Opin. 
Struct. Biol. 22, 778–85 (2012). 
50. Fischer, N., Konevega, A. L., Wintermeyer, W., Rodnina, M. V & Stark, H. 
Ribosome dynamics and tRNA movement by time-resolved electron 
cryomicroscopy. Nature 466, 329–333 (2010). 
  170 
 
51. Zhou, J., Lancaster, L., Donohue, J. P. & Noller, H. F. How the ribosome 
hands the A-site tRNA to the P site during EF-G-catalyzed translocation. 
Science 345, 1188–91 (2014). 
52. Andersen, C. B. F. et al. Structure of eEF3 and the mechanism of transfer RNA 
release from the E-site. Nature 443, 663–668 (2006). 
53. Dever, T. E. & Green, R. The elongation, termination, and recycling phases of 
translation in eukaryotes. Cold Spring Harb. Perspect. Biol. 4, a013706 (2012). 
54. Zhou, J., Korostelev, A., Lancaster, L. & Noller, H. F. Crystal structures of 
70S ribosomes bound to release factors RF1, RF2 and RF3. Curr. Opin. Struct. 
Biol. 22, 733–42 (2012). 
55. Hirokawa, G. et al. The role of ribosome recycling factor in dissociation of 70S 
ribosomes into subunits. RNA 11, 1317–28 (2005). 
56. Peske, F., Rodnina, M. V & Wintermeyer, W. Sequence of steps in ribosome 
recycling as defined by kinetic analysis. Mol. Cell 18, 403–412 (2005). 
57. Fan-Minogue, H. et al. Distinct eRF3 requirements suggest alternate eRF1 
conformations mediate peptide release during eukaryotic translation 
termination. Mol. Cell 30, 599–609 (2008). 
58. Frolova, L. Y. et al. Mutations in the highly conserved GGQ motif of class 1 
polypeptide release factors abolish ability of human eRF1 to trigger peptidyl-
tRNA hydrolysis. RNA 5, 1014–20 (1999). 
59. Frolova, L. et al. Eukaryotic polypeptide chain release factor eRF3 is an eRF1- 
and ribosome-dependent guanosine triphosphatase. RNA 2, 334–41 (1996). 
60. Pisarev, A. V, Hellen, C. U. T. & Pestova, T. V. Recycling of eukaryotic 
posttermination ribosomal complexes. Cell 131, 286–99 (2007). 
61. Pisarev, A. V et al. The role of ABCE1 in eukaryotic posttermination 
ribosomal recycling. Mol. Cell 37, 196–210 (2010). 
62. Skabkin, M. A., Skabkina, O. V, Hellen, C. U. T. & Pestova, T. V. Reinitiation 
and other unconventional posttermination events during eukaryotic translation. 
Mol. Cell 51, 249–64 (2013). 
63. Uchida, N., Hoshino, S.-I., Imataka, H., Sonenberg, N. & Katada, T. A novel 
role of the mammalian GSPT/eRF3 associating with poly(A)-binding protein 
in Cap/Poly(A)-dependent translation. J. Biol. Chem. 277, 50286–92 (2002). 
  171 
 
64. Jackson, R. J., Hellen, C. U. T. & Pestova, T. V. Termination and post-
termination events in eukaryotic translation. Advances in protein chemistry 
and structural biology 86, 45–93 (Elsevier Inc., 2012). 
65. Shoemaker, C. J., Eyler, D. E. & Green, R. Dom34:Hbs1 promotes subunit 
dissociation and peptidyl-tRNA drop-off to initiate no-go decay. Science (80-. 
). 330, 369–372 (2010). 
66. Becker, T. et al. Structure of the no-go mRNA decay complex Dom34-Hbs1 
bound to a stalled 80S ribosome. Nat. Struct. Mol. Biol. 18, 715–720 (2011). 
67. Becker, T. et al. Structural basis of highly conserved ribosome recycling in 
eukaryotes and archaea. Nature 482, 501–506 (2012). 
68. Gao, N. et al. Mechanism for the disassembly of the posttermination complex 
inferred from cryo-EM studies. Mol. Cell 18, 663–74 (2005). 
69. Preis, A. et al. Cryoelectron microscopic structures of eukaryotic translation 
termination complexes containing eRF1-eRF3 or eRF1-ABCE1. Cell Rep. 8, 
59–65 (2014). 
70. Shoemaker, C. J., Eyler, D. E. & Green, R. Dom34:Hbs1 promotes subunit 
dissociation and peptidyl-tRNA drop-off to initiate no-go decay. Science 330, 
369–72 (2010). 
71. Skabkin, M. A. et al. Activities of Ligatin and MCT-1/DENR in eukaryotic 
translation initiation and ribosomal recycling. Genes Dev. 24, 1787–801 
(2010). 
72. Sohmen, D., Harms, J. M., Schlünzen, F. & Wilson, D. N. Enhanced SnapShot: 
Antibiotic inhibition of protein synthesis II. Cell 139, 212–212.e1 (2009). 
73. Sohmen, D., Harms, J. M., Schlünzen, F. & Wilson, D. N. SnapShot: 
Antibiotic inhibition of protein synthesis I. Cell 138, 1248.e1 (2009). 
74. Tsai, A. et al. The impact of aminoglycosides on the dynamics of translation 
elongation. Cell Rep. 3, 497–508 (2013). 
75. De Loubresse, N. G. et al. Structural basis for the inhibition of the eukaryotic 
ribosome. Nature 513, 517–522 (2014). 
76. Hansen, J. L., Moore, P. B. & Steitz, T. A. Structures of five antibiotics bound 
at the peptidyl transferase center of the large ribosomal subunit. J. Mol. Biol. 
330, 1061–1075 (2003). 
  172 
 
77. Rakauskaite, R. & Dinman, J. D. Mutations of highly conserved bases in the 
peptidyltransferase center induce compensatory rearrangements in yeast 
ribosomes. RNA 17, 855–64 (2011). 
78. Petrov, A., Chen, J., O’Leary, S., Tsai, A. & Puglisi, J. D. Single-molecule 
analysis of translational dynamics. Cold Spring Harb. Perspect. Biol. 4, 
a011551 (2012). 
79. Zeng, X., Chugh, J., Casiano-Negroni, A., Al-Hashimi, H. M. & Brooks, C. L. 
Flipping of the ribosomal A-site adenines provides a basis for tRNA selection. 
J. Mol. Biol. 426, 3201–13 (2014). 
80. Nanda, J. S., Saini, A. K., Muñoz, A. M., Hinnebusch, A. G. & Lorsch, J. R. 
Coordinated movements of eukaryotic translation initiation factors eIF1, 
eIF1A, and eIF5 trigger phosphate release from eIF2 in response to start codon 
recognition by the ribosomal preinitiation complex. J. Biol. Chem. 288, 5316–
29 (2013). 
81. Hinnebusch, A. G. & Lorsch, J. R. The mechanism of eukaryotic translation 
initiation: new insights and challenges. Cold Spring Harb. Perspect. Biol. 4, 
a011544– (2012). 
82. Jenner, L., Ben-Shem, A., Demeshkina, N., Yusupov, M. & Yusupova, G. in 
Biophysical approaches to translational control of gene expression (ed. 
Dinman, J. D.) 1–25 (Springer New York, 2013). 
83. Sanbonmatsu, K. Y., Blanchard, S. C. & Whitford, P. C. in Biophysical 
approaches to translational control of gene expression 51–68 (Springer New 
York, 2013). 
84. Datta, P. & Chatterjee, A. in Biophysical approaches to translational control 
of gene expression 27–50 (Springer New York, 2013). 
85. Whitford, P. C. et al. Accommodation of aminoacyl-tRNA into the ribosome 
involves reversible excursions along multiple pathways. RNA 16, 1196–204 
(2010). 
86. Sanbonmatsu, K. Y., Joseph, S. & Tung, C. S. Simulating movement of tRNA 
into the ribosome during decoding. Proc. Natl. Acad. Sci. 102, 15854 (2005). 
87. Biophysical approaches to translational control of gene expression. (Springer 
New York; 1 edition, 2013). doi:10.1007/978-1-4614-3991-2 
88. Perutz, M. F. in Nobel Lectures Chemistry 653–673 (Elsevier Publishing 
Company, 1964). 
  173 
 
89. Bai, X.-C., Fernandez, I. S., McMullan, G. & Scheres, S. H. W. Ribosome 
structures to near-atomic resolution from thirty thousand cryo-EM particles. 
Elife 2, e00461 (2013). 
90. Roy, R., Hohng, S. & Ha, T. A practical guide to single-molecule FRET. Nat. 
Methods 5, 507–16 (2008). 
91. Blanchard, S. C. Single-molecule observations of ribosome function. Curr. 
Opin. Struct. Biol. 19, 103–9 (2009). 
92. Wang, Y., Xiao, M. & Li, Y. Heterogeneity of single molecule FRET signals 
reveals multiple active ribosome subpopulations. Proteins 82, 1–9 (2014). 
93. RNA-Protein Interactions : A Practical Approach: A Practical Approach. 366 
(Oxford University Press, 1998). at 
<http://books.google.com/books?id=ObiBuB_s24cC&pgis=1> 
94. Weeks, K. M. Advances in RNA structure analysis by chemical probing. Curr. 
Opin. Struct. Biol. 20, 295–304 (2010). 
95. Hermanson, G. T. Bioconjugate Techniques. 1200 (Academic Press, 2013). at 
<http://books.google.com/books?id=6aO-207lhdgC&pgis=1> 
96. Heilek, G. M., Marusak, R., Meares, C. F. & Noller, H. F. Directed hydroxyl 
radical probing of 16S rRNA using Fe(II) tethered to ribosomal protein S4. 
Proc. Natl. Acad. Sci. U. S. A. 92, 1113–6 (1995). 
97. Woodson, S. A., Deras, M. L. & Brenowitz, M. Time-resolved hydroxyl 
radical footprinting of RNA with X-rays. Curr. Protoc. Nucleic Acid Chem. 
Chapter 11, Unit 11.6 (2001). 
98. Regulski, E. E. & Breaker, R. R. In-line probing analysis of riboswitches. 
Methods Mol. Biol. 419, 53–67 (2008). 
99. Leshin, J. A., Meskauskas, A. M. & Dinman, J. D. in Biophysical approaches 
to translational control of gene expression 69–81 (Springer New York, 2013). 
100. Mortimer, S. a & Weeks, K. M. A fast-acting reagent for accurate analysis of 
RNA secondary and tertiary structure by SHAPE chemistry. J. Am. Chem. Soc. 
129, 4144–5 (2007). 
101. Merino, E. J., Wilkinson, K. A., Coughlan, J. L. & Weeks, K. M. RNA 
structure analysis at single nucleotide resolution by selective 2’-hydroxyl 
acylation and primer extension (SHAPE). J. Am. Chem. Soc. 127, 4223–4231 
(2005). 
  174 
 
102. Chamberlin, S. I. & Weeks, K. M. Mapping Local Nucleotide Flexibility by 
Selective Acylation of 2‘-Amine Substituted RNA. J. Am. Chem. Soc. 122, 
216–224 (2000). 
103. Roat-Malone, R. M. Bioinorganic Chemistry: A Short Course. 243 (John 
Wiley & Sons, 2007). at <https://books.google.com/books?id=g--28L1G-
MkC&pgis=1> 
104. Mcginnis, J. L., Dunkle, J. A., Cate, J. H. D. & Weeks, K. M. The Mechanisms 
of RNA SHAPE Chemistry. (2012). 
105. Leshin, J. A., Heselpoth, R., Belew, A. T. & Dinman, J. D. High throughput 
structural analysis of yeast ribosomes using hSHAPE. RNA Biol. 8, 478–487 
(2011). 
106. Vasa, S. M., Guex, N., Wilkinson, K. a, Weeks, K. M. & Giddings, M. C. 
ShapeFinder: a software system for high-throughput quantitative analysis of 
nucleic acid reactivity information resolved by capillary electrophoresis. RNA 
14, 1979–90 (2008). 
107. McGinnis, J. L., Duncan, C. D. S. & Weeks, K. M. High-throughput SHAPE 
and hydroxyl radical analysis of RNA structure and ribonucleoprotein 
assembly. Methods Enzymol. 468, 67–89 (2009). 
108. Batey, R., Rambo, R. & Doudna, J. Tertiary Motifs in RNA Structure and 
Folding. Angew. Chem. Int. Ed. Engl. 38, 2326–2343 (1999). 
109. Wilkinson, K. A., Merino, E. J. & Weeks, K. M. Selective 2’-hydroxyl 
acylation analyzed by primer extension (SHAPE): quantitative RNA structure 
analysis at single nucleotide resolution. Nat. Protoc. 1, 1610–6 (2006). 
110. Wilkinson, K. a et al. High-throughput SHAPE analysis reveals structures in 
HIV-1 genomic RNA strongly conserved across distinct biological states. PLoS 
Biol. 6, e96 (2008). 
111. Vasa, S. M., Guex, N., Wilkinson, K. A., Weeks, K. M. & Giddings, M. C. 
ShapeFinder : A software system for high-throughput quantitative analysis of 
nucleic acid reactivity information resolved by capillary electrophoresis 
(2008). doi:10.1261/rna.1166808 
112. Mortimer, S. a & Weeks, K. M. Time-resolved RNA SHAPE chemistry: 
quantitative RNA structure analysis in one-second snapshots and at single-
nucleotide resolution. Nat. Protoc. 4, 1413–21 (2009). 
  175 
 
113. Lucks, J. B. et al. Multiplexed RNA structure characterization with selective 
2’-hydroxyl acylation analyzed by primer extension sequencing (SHAPE-Seq). 
Proc. Natl. Acad. Sci. U. S. A. 108, 11063–8 (2011). 
114. Loughrey, D., Watters, K. E., Settle, A. H. & Lucks, J. B. SHAPE-Seq 2.0: 
systematic optimization and extension of high-throughput chemical probing of 
RNA secondary structure with next generation sequencing. Nucleic Acids Res. 
gku909– (2014). doi:10.1093/nar/gku909 
115. Karabiber, F., McGinnis, J. L., Favorov, O. V & Weeks, K. M. QuShape: 
rapid, accurate, and best-practices quantification of nucleic acid probing 
information, resolved by capillary electrophoresis. RNA 19, 63–73 (2013). 
116. Low, J. T. & Weeks, K. M. SHAPE-directed RNA secondary structure 
prediction. Methods 52, 150–8 (2010). 
117. Kladwang, W., VanLang, C. C., Cordero, P. & Das, R. Understanding the 
errors of SHAPE-directed RNA structure modeling. Biochemistry 50, 8049–56 
(2011). 
118. Yoon, S. et al. HiTRACE: high-throughput robust analysis for capillary 
electrophoresis. Bioinformatics 27, 1798–805 (2011). 
119. Siegfried, N. A., Busan, S., Rice, G. M., Nelson, J. A. E. & Weeks, K. M. 
RNA motif discovery by SHAPE and mutational profiling (SHAPE-MaP). Nat. 
Methods 11, 959–65 (2014). 
120. Kladwang, W., VanLang, C. C., Cordero, P. & Das, R. A two-dimensional 
mutate-and-map strategy for non-coding RNA structure. Nat. Chem. 3, 954–62 
(2011). 
121. Tian, S., Cordero, P., Kladwang, W. & Das, R. High-throughput mutate-map-
rescue evaluates SHAPE-directed RNA structure and uncovers excited states. 
RNA 20, 1815–26 (2014). 
122. Kladwang, W. et al. Standardization of RNA chemical mapping experiments. 
Biochemistry 53, 3063–5 (2014). 
123. Pang, P. S., Elazar, M., Pham, E. a & Glenn, J. S. Simplified RNA secondary 
structure mapping by automation of SHAPE data analysis. Nucleic Acids Res. 
39, e151 (2011). 
124. Fang, G. et al. Genome-wide mapping of methylated adenine residues in 
pathogenic Escherichia coli using single-molecule real-time sequencing. Nat. 
Biotechnol. 30, 1232–9 (2012). 
  176 
 
125. Xing, H., Mo, Y., Liao, W. & Zhang, M. Q. Genome-wide localization of 
protein-DNA binding and histone modification by a Bayesian change-point 
method with ChIP-seq data. PLoS Comput. Biol. 8, e1002613 (2012). 
126. Ramrath, D. J. F. et al. Visualization of two transfer RNAs trapped in transit 
during elongation factor G-mediated translocation. Proc. Natl. Acad. Sci. U. S. 
A. 110, 20964–9 (2013). 
127. Budkevich, T. et al. Structure and Dynamics of the Mammalian Ribosomal 
Pretranslocation Complex. Mol. Cell 44, 214–224 (2011). 
128. Crow. Lognormal Distributions: Theory and Applications. 387 (CRC Press, 
1987). at <https://books.google.com/books?id=B8kNa1khS4QC&pgis=1> 
129. Wang, S. C. Analysis of Zero-Heavy Data Using a Mixture Model Approach. 
(Virginia Tech, 1998). at <http://vtechworks.lib.vt.edu/handle/10919/30357> 
130. Saini, A. K. et al. Eukaryotic translation initiation factor eIF5 promotes the 
accuracy of start codon recognition by regulating Pi release and conformational 
transitions of the preinitiation complex. Nucleic Acids Res. 42, 9623–40 
(2014). 
131. Lomakin, I. B. & Steitz, T. A. The initiation of mammalian protein synthesis 
and mRNA scanning mechanism. Nature 500, 307–11 (2013). 
132. Passmore, L. A. et al. The eukaryotic translation initiation factors eIF1 and 
eIF1A induce an open conformation of the 40S ribosome. Mol. Cell 26, 41–50 
(2007). 
133. Weisser, M., Voigts-Hoffmann, F., Rabl, J., Leibundgut, M. & Ban, N. The 
crystal structure of the eukaryotic 40S ribosomal subunit in complex with eIF1 
and eIF1A. Nat. Struct. Mol. Biol. 20, 1015–7 (2013). 
134. Hinnebusch, A. G. The scanning mechanism of eukaryotic translation 
initiation. Annu. Rev. Biochem. 83, 779–812 (2014). 
135. Lomakin, I. B. & Steitz, T. A. The initiation of mammalian protein synthesis 
and mRNA scanning mechanism. Nature 500, 307–11 (2013). 
136. Erzberger, J. P. et al. Molecular Architecture of the 40S⋅eIF1⋅eIF3 Translation 
Initiation Complex. Cell 158, 1123–35 (2014). 
137. Hussain, T. et al. Structural Changes Enable Start Codon Recognition by the 
Eukaryotic Translation Initiation Complex. Cell 159, 597–607 (2014). 
  177 
 
138. Hashem, Y. et al. Structure of the mammalian ribosomal 43S preinitiation 
complex bound to the scanning factor DHX29. Cell 153, 1108–19 (2013). 
139. Luna, R. E. et al. The interaction between eukaryotic initiation factor 1A and 
eIF5 retains eIF1 within scanning preinitiation complexes. Biochemistry 52, 
9510–8 (2013). 
140. Algire, M. A. et al. Development and characterization of a reconstituted yeast 
translation initiation system. RNA 8, 382–397 (2002). 
141. Conte, M. R. et al. Structure of the eukaryotic initiation factor (eIF) 5 reveals a 
fold common to several translation factors. Biochemistry 45, 4550–8 (2006). 
142. Bieniossek, C. et al. The crystal structure of the carboxy-terminal domain of 
human translation initiation factor eIF5. J. Mol. Biol. 360, 457–65 (2006). 
143. Gilbert, R. J. C. et al. Reconfiguration of yeast 40S ribosomal subunit domains 
by the translation initiation multifactor complex. Proc. Natl. Acad. Sci. 104, 
5788 (2007). 
144. Moazed, D. & Noller, H. F. Interaction of tRNA with 23S rRNA in the 
ribosomal A, P, and E sites. Cell 57, 585–597 (1989). 
145. Moazed, D. & Noller, H. F. Binding of tRNA to the ribosomal A and P sites 
protects two distinct sets of nucleotides in 16 S rRNA. J. Mol. Biol. 211, 135–
45 (1990). 
146. Moazed, D. & Robertson, J. M. Interaction of elongation factors EF-G and EF-
Tu with a conserved loop in 23S RNA. Nature 334, 362–364 (1988). 
147. Moazed, D. & Noller, H. F. Sites of interaction of the CCA end of peptidyl-
tRNA with 23S rRNA. Proc. Natl. Acad. Sci. 88, 3725–3728 (1991). 
148. Moazed, D. & Noller, H. F. Intermediate states in the movement of transfer 
RNA in the ribosome. Nature 342, 142–148 (1989). 
149. Joseph, S. & Noller, H. F. Mapping the rRNA neighborhood of. 15, 910–916 
(1996). 
150. Fraser, C. S., Berry, K. E., Hershey, J. W. B. & Doudna, J. A. eIF3j is located 
in the decoding center of the human 40S ribosomal subunit. Mol. Cell 26, 811–
9 (2007). 
151. Khade, P. & Joseph, S. Functional interactions by transfer RNAs in the 
ribosome. FEBS Lett. 584, 420–6 (2010). 
  178 
 
152. Wollenzien, P., Expert-Bezancon, A. & Favree, A. Sites of Contact of mRNA 
with 16S rRNA and 25S rRNA. Biochemistry 30, 1788–1795 (1991). 
153. Moazed, D. & Noller, H. F. Binding of tRNA to the ribosomal A and P sites 
protects two distinct sets of nucleotides in 16 S rRNA. J. Mol. Biol. 211, 135–
45 (1990). 
154. Moazed, D. & Noller, H. F. Interaction of tRNA with 23S rRNA in the 
ribosomal A, P, and E sites. Cell 57, 585–597 (1989). 
155. Moazed, D. & Noller, H. F. Transfer RNA shields specific nucleotides in 16S 
ribosomal RNA from attack by chemical probes. Cell 47, 985–994 (1986). 
156. Wilson, K. S. & Noller, H. F. Molecular movement inside the translational 
engine. Cell 92, 337–349 (1998). 
157. Mitchell, P., Osswald, M. & Brimacombe, R. Identification of intermolecular 
RNA cross-links at the subunit interface of the Escherichia coli ribosome. 
Biochemistry 31, 3004–3011 (1992). 
158. Merryman, C., Moazed, D., Daubresse, G. & Noller, H. F. Nucleotides in 23S 
rRNA protected by the association of 30S and 50S ribosomal subunits1. J. 
Mol. Biol. 285, 107–113 (1999). 
159. Merryman, C., Moazed, D., McWhirter, J. & Noller, H. F. Nucleotides in 16S 
rRNA protected by the association of 30S and 50S ribosomal subunits. J. Mol. 
Biol. 285, 97–105 (1999). 
160. Maivali, U. Definition of bases in 23S rRNA essential for ribosomal subunit 
association. RNA 10, 600–604 (2004). 
161. Yusupov, M. M. et al. Crystal structure of the ribosome at 5.5 A resolution. 
Science 292, 883–96 (2001). 
162. Anger, A. M. et al. Structures of the human and Drosophila 80S ribosome. 
Nature 497, 80–5 (2013). 
163. Zhang, W., Dunkle, J. a & Cate, J. H. D. Structures of the ribosome in 
intermediate states of ratcheting. Science 325, 1014–7 (2009). 
164. Wilson, D. N. & Doudna Cate, J. H. The structure and function of the 
eukaryotic ribosome. Cold Spring Harb. Perspect. Biol. 4, a011536– (2012). 
165. Kossinova, O., Malygin, A., Krol, A. & Karpova, G. The SBP2 protein central 
to selenoprotein synthesis contacts the human ribosome at expansion segment 
7L of the 28S rRNA. RNA 20, 1046–56 (2014). 
  179 
 
166. Caban, K. & Copeland, P. R. Selenocysteine insertion sequence (SECIS)-
binding protein 2 alters conformational dynamics of residues involved in tRNA 
accommodation in 80 S ribosomes. J. Biol. Chem. 287, 10664–73 (2012). 
167. Armache, J.-P. et al. Cryo-EM structure and rRNA model of a translating 
eukaryotic 80S ribosome at 5.5-A resolution. Proc. Natl. Acad. Sci. U. S. A. 
107, 19748–53 (2010). 
168. Alkemar, G. & NYGÅRD, O. Secondary structure of two regions in expansion 
segments ES3 and ES6 with the potential of forming a tertiary interaction in 
eukaryotic 40S ribosomal subunits. RNA 10, 403 (2004). 
169. Alkemar, G. & Nygård, O. Probing the secondary structure of expansion 
segment ES6 in 18S ribosomal RNA. Biochemistry 45, 8067–78 (2006). 
170. Larsson, S. L. & Nygård, O. Proposed secondary structure of eukaryote 
specific expansion segment 15 in 28S rRNA from mice, rats, and rabbits. 
Biochemistry 40, 3222–31 (2001). 
171. Nygård, O., Alkemar, G. & Larsson, S. L. Analysis of the secondary structure 
of expansion segment 39 in ribosomes from fungi, plants and mammals. J. 
Mol. Biol. 357, 904–16 (2006). 
172. Marquez, V. et al. Maintaining the Ribosomal Reading Frame:: The Influence 
of the E Site during Translational Regulation of Release Factor 2. Cell 118, 
45–55 (2004). 
173. Rheinberger, H. J. & Nierhaus, K. H. Allosteric interactions between the 
ribosomal transfer RNA-binding sites A and E. J. Biol. Chem. 261, 9133–9 
(1986). 
174. Nierhaus, K. H. The allosteric three-site model for the ribosomal elongation 
cycle: features and future. Biochemistry 29, 4997–5008 (1990). 
175. Svidritskiy, E., Brilot, A. F., Koh, C. S., Grigorieff, N. & Korostelev, A. A. 
Structures of Yeast 80S Ribosome-tRNA Complexes in the Rotated and 
Nonrotated Conformations. Structure 22, 1210–8 (2014). 
176. Burakovsky, D. E. et al. Mutations at the accommodation gate of the ribosome 
impair RF2-dependent translation termination. RNA 16, 1848–53 (2010). 
177. Sergiev, P. V. V et al. The conserved A-site finger of the 23 S rRNA: Just one 
of the intersubunit bridges or a part of the allosteric communication pathway? 
J. Mol. Biol. 353, 116–123 (2005). 
  180 
 
178. Rakauskaite, R. & Dinman, J. D. An arc of unpaired “hinge bases” facilitates 
information exchange among functional centers of the ribosome. Mol. Cell. 
Biol. 26, 8992–9002 (2006). 
179. Rhodin, M. H. J. & Dinman, J. D. An Extensive Network of Information Flow 
through the B1b/c Intersubunit Bridge of the Yeast Ribosome. PLoS One 6, 
e20048 (2011). 
180. Rhodin, M. H. J. & Dinman, J. D. A flexible loop in yeast ribosomal protein 
L11 coordinates P-site tRNA binding. Nucleic Acids Res. 38, 8377–8389 
(2010). 
181. Meskauskas, A. & Dinman, J. D. A molecular clamp ensures allosteric 
coordination of peptidyltransfer and ligand binding to the ribosomal A-site. 
Nucleic Acids Res. 38, 7800–13 (2010). 
182. Koosha, H., Cameron, D., Andrews, K., Dahlberg, A. E. & March, P. E. 
Alterations in the peptidyltransferase and decoding domains of ribosomal RNA 
suppress mutations in the elongation factor G gene. RNA 6, 1166–73 (2000). 
183. Agrawal, R. K., Lata, R. K. & Frank, J. Conformational variability in 
Escherichia coli 70S ribosome as revealed by 3D cryo-electron microscopy. 
Int. J. Biochem. Cell Biol. 31, 243–54 (1999). 
184. Malhotra, A. et al. Escherichia coli 70 S ribosome at 15 A resolution by cryo-
electron microscopy: localization of fMet-tRNAfMet and fitting of L1 protein. 
J. Mol. Biol. 280, 103–16 (1998). 
185. Fei, J., Kosuri, P., MacDougall, D. D. & Gonzalez, R. L. Coupling of 
ribosomal L1 stalk and tRNA dynamics during translation elongation. Mol. 
Cell 30, 348–359 (2008). 
186. Trabuco, L. G. et al. The role of L1 stalk-tRNA interaction in the ribosome 
elongation cycle. J. Mol. Biol. 402, 741–60 (2010). 
187. Spahn, C. M. T. et al. Structure of the 80S ribosome from Saccharomyces 
cerevisiae–tRNA-ribosome and subunit-subunit interactions. Cell 107, 373–
386 (2001). 
188. Taylor, D. J. et al. Structures of modified eEF2 80S ribosome complexes 
reveal the role of GTP hydrolysis in translocation. EMBO J. 26, 2421–2431 
(2007). 
189. Taylor, D. J. D. J. et al. Comprehensive molecular structure of the eukaryotic 
ribosome. Structure 17, 1591–1604 (2009). 
  181 
 
190. Spahn, C. M. T. et al. Domain movements of elongation factor eEF2 and the 
eukaryotic 80S ribosome facilitate tRNA translocation. EMBO J. 23, 1008–
1019 (2004). 
191. Petrov, A. S. et al. Evolution of the ribosome at atomic resolution. Proc. Natl. 
Acad. Sci. U. S. A. 111, 10251–10256 (2014). 
192. Schroeder, S. J., Blaha, G., Tirado-Rives, J., Steitz, T. A. & Moore, P. B. The 
structures of antibiotics bound to the E site region of the 50 S ribosomal 
subunit of Haloarcula marismortui: 13-deoxytedanolide and girodazole. J. Mol. 
Biol. 367, 1471–1479 (2007). 
193. Greber, B. J. et al. Cryo-EM structure of the archaeal 50S ribosomal subunit in 
complex with initiation factor 6 and implications for ribosome evolution. J. 
Mol. Biol. 418, 145–60 (2012). 
194. Schmeing, T. M. et al. A pre-translocational intermediate in protein synthesis 
observed in crystals of enzymatically active 50S subunits. Nat. Struct. Mol. 
Biol. 9, 225–230 (2002). 
195. Yonath, A. The search and its outcome: high-resolution structures of ribosomal 
particles from mesophilic, thermophilic, and halophilic bacteria at various 
functional states. Annu. Rev. Biophys. Biomol. Struct. 31, 257–73 (2002). 
196. Fischer, N., Konevega, A. L., Wintermeyer, W., Rodnina, M. V & Stark, H. 
Ribosome dynamics and tRNA movement by time-resolved electron 
cryomicroscopy. Nature 466, 329–33 (2010). 
197. Yamamoto, H. et al. EF-G and EF4: translocation and back-translocation on 
the bacterial ribosome. Nat. Rev. Microbiol. 12, 89–100 (2014). 
198. Ramakrishnan, V., Brodersen, D. E., Carter, A. P., Wimberly, B. T. & 
Clemons Jr., W. M. Crystal structure of the 30s ribosome. (2001).  
199. Rakauskaite, R., Dinman, J. D. & Rakauskait\.e, R. rRNA mutants in the yeast 
peptidyltransferase center reveal allosteric information networks and 
mechanisms of drug resistance. Nucleic Acids Res. 36, 1497 (2008). 
200. Berk, V., Zhang, W., Pai, R. D. & Cate, J. H. D. Structural basis for mRNA 
and tRNA positioning on the ribosome. Proc. Natl. Acad. Sci. 103, 15830 
(2006). 
201. Stark, H. et al. Ribosome interactions of aminoacyl-tRNA and elongation 
factor Tu in the codon-recognition complex. Nat. Struct. Mol. Biol. 9, 849–854 
(2002). 
  182 
 
202. Gao, Y.-G. et al. The structure of the ribosome with elongation factor G 
trapped in the posttranslocational state. Science 326, 694–9 (2009). 
203. Yusupov, M. M. et al. Crystal structure of the ribosome at 5.5 Å resolution. 
Science. 292, 883 (2001). 
204. Homan, P. J. et al. Single-molecule correlated chemical probing of RNA. Proc. 
Natl. Acad. Sci. U. S. A. 111, 13858–13863 (2014). 
205. Shenvi, C. L., Dong, K., Friedman, E. M., Hanson, J. A. & Cate, J. H. D. 
Accessibility of 18S rRNA in human 40S subunits and 80S ribosomes at 
physiological magnesium ion concentrations - implications for the study of 
ribosome dynamics. RNA 11, 1898-1908 (2005).  
206. Algire, M. A. et al. Development and characterization of a reconstituted yeast 
translation initiation system. RNA 8, 382–97 (2002). 
207. Eyler, D. E. & Green, R. Distinct response of yeast ribosomes to a miscoding 
event during translation. RNA 17, 925–32 (2011). 
208. Graifer, D. & Karpova, G. Roles of ribosomal proteins in the functioning of 
translational machinery of eukaryotes. Biochimie 109C, 1–17 (2014). 
209. Musalgaonkar, S., Moomau, C. A. & Dinman, J. D. Ribosomes in the balance: 
structural equilibrium ensures translational fidelity and proper gene expression. 
Nucleic Acids Res. 42, 13384–92 (2014). 
210. Sulima, S. O. et al. Eukaryotic rpL10 drives ribosomal rotation. Nucleic Acids 
Res. 42, 2049–63 (2014). 
211. Deigan, K. E., Li, T. W., Mathews, D. H. & Weeks, K. M. Accurate SHAPE-
directed RNA structure determination a SHAPE experiment can be interpreted 
as a pseudo-free energy high accuracy . Free energy minimization , by using 
SHAPE pseudo-. 2008, (2008). 
212. Leshin, J. A, Rakauskaitė, R., Dinman, J. D. & Meskauskas, A. Enhanced 
purity, activity and structural integrity of yeast ribosomes purified using a 
general chromatographic method. RNA Biol. 7, 354–60 (2010). 
213. Triana, F., Nierhaus, K. H. & Chakraburtty, K. Transfer RNA binding to 80S 
ribosomes from yeast: evidence for three sites. Biochem. Mol. Biol. Int. 33, 909 
(1994).