s -
3 3.
in”
z... .

Em.“ 3......
d.

3:
353......
{03.211.- .3...
.u .
Iii _ ..,
2.
. v

I

..

it!!! {2...
liswtiufi

.. .
. i:
5...: . £131.?
:. v. 2, . .. .
. r 03h.
. I 4 ,
.9... . .
a... . . ,
4.1.!!! ..
r 5 .133... l
b 3.2. I
. . , . o sunken: 5)....
. 55.33.... 31.41::
. . .3 3i: .. . .c ..
31345:: I... :fitnn.
. .1 .2... .
. . ’2‘1
. :31: . 3!..m.
. . afifiv.‘.fl.fi.. .3911...
. . .5 2?.3111:.:.- v1.0.3.
.01 ..)¥:l,.l « l}.‘
. u .49.}. . inn...” ..
#9... (£1) 1“.
. Hhi.a...f}.h 1.1.63.
, . v!.-."€v..<§.1xl..h
. . Tui’ll’h... '1 ,
. 3 . 43...: iv
a. $3.. iffhflail
. .v 3.3:»! 35?. . . a
, v. .3... 2.1.5.
519...: u o 3..
x..v.t11...a.m~rzuflnl: . 3....
. . . .i‘. (a: Irl 3.3.1.:
5.1...) 5.3.1!
. , 131.. (-01.79; .
. 1. a. 3...“:
, Jflfiflaiit .....
zl‘l
4!... Hunt}...
. . .rbmfi‘rhiz. . . flag».
21.6.... .I. 2.-.. 25...... .
. .. 1....e.a...a:?».ur!..:o. 2... 9b”... .
.21.”...592eti... 1...... 53b5,]...
.i ”is! d5“n..w..y..i.,:.§.. can?» urn... ska. .
v.3... [£3%.d:€:1:2z ..
, :. a :3
.53....
a.

3
.
9......
.. .3. 4.
-22.! t
E. .2...

. 1t.

11:.

In:

flit": .
a," 1‘}.
v.3.nm... .5:

. ,

3;" 1...”...

 

0:)?

This is to certify that the
dissertation entitled

ELUCIDATING THE STRUCTURE AND KINETICS OF THE
APOCYTOCHROME B mRNA/gRNA COMPLEX IN
TRYPANOSOMA BRUCE] MITOCHONDRIA

presented by

Laura Elizabeth Yu

has been accepted towards fulfillment
of the requirements for the
th degree in

Cell and M lecular Biology

       

 

     

 

 

MSU is an Affirmative Action/Equal Opportunity Institution

 

LIBRARY
Michigan State
University

—~

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:lClRCiDateDue.indd-p.1

 

ELUCIDATING THE STRUCTURE AND KINETICS OF
THE APOCYTOCHROME B mRNA/gRNA COMPLEX IN
TR YPANOSOMA BR UCEI MITOCHONDRIA

By

Laura Elizabeth Yu

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Cell and Molecular Biology Program

2006

ABSTRACT

ELUCIDATING THE STRUCTURE AND KINETICS OF THE APOCYTOCHROME
B mRNA/gRNA COMPLEX lN TRYPANOSOMA BRUCE] MITOCHONDRIA

By
Laura Elizabeth Yu
Expression of mitochondrial genes in Ttypanosoma brucei requires RNA editing of its
mRNA transcripts. During editing, uridylates are precisely inserted and deleted as
directed by the guide RNA (gRNA) template to create the protein open reading frame.
This process involves the bimolecular interaction of the gRNA with its cognate pro-edited
mRNA and the assembly of a protein complex. The importance of RNA structure in
establishing a functional editing complex is poorly understood. Previous experiments
indicate that different mRNA/gRNA pairs can form similar secondary structures
suggesting that a common core architecture may be important for editosome recognition
and function. Using solution structure probing, we have investigated the structure of the
initiating gRNA, gCYb-558, in the mRNA/gRNA complex. The data indicate that the
stem-loop formed by the guiding region of the gRNA alone is maintained in its
interaction with the pre-edited message. In addition, the data suggest that a gRNA stem-
loop structure is maintained through the first few editing events by the use of alternative
base pairing with the U-tail. This suggests that the gRNA stem-loop is an important
component of the initial editing complex.
In trypansomes, RNA editing of the mitochondrial mRNAs is developmentally
regulated in a transcript specific manner. We hypothesize that regulation involves the
structure of the mRN A and its ability to interact with its gRNA. Surface plasmon

resonance was used to measure the kinetics of gRNA binding for three separate

mRNA/gRNA pairs; two mRN A substrates with predicted single stranded anchor binding
sites (ABS) and one mRNA substrate with the ABS located within a thermodynamically
stable stem-loop. The stability of the mRN A stem-loop appears to affect the gRN A
anchor target binding and results in a slower association rate as well as a faster
dissociation rate. In contrast, the mRN As with an open ABS associate with their cognate
gRN As at a much faster rate. In addition, they have a surprisingly slow dissociation rate.
This slow dissociation rate may be necessary to allow the editosome protein complexes to
recognize and assemble onto the mRN A/gRN A pair.

Editing of the mRN A, results in a progressive lengthening of the anchor helix. Using
apocytochrome b (CYb) mRN A and 2 partially edited CYb mRNAs, the kinetic effects of
editing for the CYb mRN A/gRN A interaction were investigated. The CYb mRN A forms
a stable stem-loop that the gRNA anchor has difficulty invading to base pair at the anchor
binding site (ABS). Each editing event appears to result in a decreased dissociation rate
that may be important for editing progression. In addition, the U-tail targets one side of
the stem-loop for binding and this appears to increase the association rate constant two
fold when the gRN A binds the unedited CYb mRNA providing a new and exciting

fimction for the U-tail.

To my husband, Min, and my family and friends for their love and support

iv

ACKNOWLEDGEMENTS

First I would like to thank my husband, Min Yu. Without his love and support I
would never have made it through graduate school. He has been the best helpmate and
sounding board on earth, and for his finite patience and never-ending love I thank him.
I’d like to thank my family and fi'iends fiom home for their love and support; especially
Frank and Sylvia Bean, my parents, and Bob and Christina Bean, my brother and sister,
and all my aunts, uncles and cousins. I’d also like to thank my friend, Angie Blazek, who
knows me best. I’d also like to thank my fi'iend, Michelle Degnan, who I partied and
studied with as an undergrad, and who was crazy like me and went to graduate school.

I would like to thank my advisor, Donna Koslowsky. She is a wonderful example of a
female scientist and an excellent role model. As a mentor, she spent many, many hours
advising me on experiments, writing, and scientific protocol. For her diligence, I thank
her. I would also like to thank my guidance committee members, Shelagh F erguson-
Miller, Jim Geiger, Charles Hoogstraten, and Ron Patterson for their help and dedication.
I owe special thanks to Dr. Patterson and Dr. Hoogstraten. Dr. Patterson was always
available to provide a listening ear and sound advice. Dr. Hoogstraten was invaluable
with good discussion and advice for my Biacore and kinetics data.

The Koslowsky lab has been a wonderful environment for friendship as well as
learning. I’d like to thank former lab member, Dr. Sandra Clement, for her unwavering
support, excellent ideas (scientific and social), and wonderful friendship. I could never
have made it through graduate school without Larissa Reifur, who has been my right

hand woman in the lab. Larissa has been an awesome friend, excellent partner in crime

in the lab, and she is always willing to read and edit everything. Playing volleyball with
Larissa helped me stay sane. I’d also like to thank Melissa Mingler for her spunky
attitude and sparkling personality. The undergraduates coming through our lab have been
invaluable; l’d especially like to thank Remy Brim, Andrea Hingst, Tanojo Mustika, and
Aimee Sutherland.

The faithful members of the MSU RNA Journal Club have been very supportive.
They have been especially helpful with ideas, listening, and evaluation of presentations.
I’d especially like to thank the members of the Donna Koslowsky lab (Donna, Sandi,
Larissa, and Mel), the Charles Hoogstraten lab (Charlie, James, and Kristy), the Ron
Patterson Lab (Ron, Kevin, and Weizhong) and the John Wang lab (John, Rich, and
Patty) for their patience, good ideas, and helpful advice.

I’d like to thank Dr Robert Hausinger, MSU Microbiology and Molecular Genetics,
and Dr. Zachary Burton, MSU Biochemisty and Molecular Biology, for their help with
rate constant equations for my kinetics data and good discussion and advice. I’d also like
to thank Dr. Joseph Leykam and the members of the Macromolecular Structure Facility
in the Biochemistry and Molecular Biology Department for the use of the Biacore
machine.

I’d like to thank the Cell and Molecular Biology program for the support and training
I’ve received as well as the faculty and staff of the Department of Microbiology and
Molecular Genetics for all their‘help and assistance. I’d especially like to thank former
graduate secretary Angie Zell and current graduate secretary Christine VanDeuren for

fixing my messes within the MSU bureaucracy, and there have been a few.

vi

TABLE OF CONTENTS

List of Tables ........................................................................................ x
List of Figures ........................................................................................ xi
List of Equations .................................................................................... xiii
Key to Abbreviations ............................................................................... xiv
CHAPTER 1 INTRODUCTION .................................................................. 1
History of human Afi'ican trypanosomiasis ..................................................... .2
Trypanosomes ........................................................................................ 3
Metabolism of trypanosomes ....................................................................... 6
Kinetoplast ............................................................................................. 7
Developmental regulation of RNA editing ...................................................... 14
Apocytochrome b .................................................................................... 16
The editosome ....................................................................................... 16
Other RNA-RN A interactions ..................................................................... 24
Antisense RN As ..................................................................................... 25
miRN As .............................................................................................. 27
RN A-RN A conclusions ............................................................................ 28
Overview of thesis .................................................................................. 29
Literature cited ....................................................................................... 35

CHAPTER 2 INTERACTIONS OF MRNAS AND GRNAS INVOLVED IN
TRYPANOSOME MITOCHONDRIAL RNA EDITING: STRUCTURE PROBING OF

A GRN A BOUND TO ITS COGNATE MRNA ................................................ 44
Introduction .......................................................................................... 45
Results and Discussion ............................................................................. 47
Crosslinked RN As used for solution structure probing ........................................ 47
Structure of the NgCYb-558 anchor region ...................................................... 55
Structure of NgCYb-558 guiding region ......................................................... 63
Structure of CYbPES3 .............................................................................. 65
Conclusions .......................................................................................... 69
Materials and Methods .............................................................................. 74
Oligodeoxyribonucleotides ........................................................................ 74
Oligoribonucleotide ................................................................................. 74
DNA templates and RNA synthesis ............................................................... 74
RNA crosslinking and end-labeling ............................................................... 76
Structure specific enzymatic probing .............................................................. 76
Solution structure probing with chemicals ....................................................... 77
Acknowledgements .................................................................................. 78
Literature Cited ...................................................................................... 79

vii

CHAPTER 3 INTERACTIONS OF MRNAS AND GRNAS INVOLVED IN
TRYPANOSOME MITOCHONDRIAL RNA EDITING: CALCULATIONS OF THE
REAL TIME KINETICS OF BINDING FOR THREE MRNAS BOUND TO THEIR

GRNAS .............................................................................................. 84
Introduction ........................................................................................... 85
Results ................................................................................................. 9O
mRN A/ gRN A pairs examined using surface plasmon resonance ............................ 90
CYbU-NgCYb-558 .................................................................................. 9O
A6UENDSh-gA6-l4 ................................................................................ 91
mND7550-gND7-550 .............................................................................. 94
Surface plasmon resonance studies ................................................................ 95
The NgCYb-558/CYbU interaction ............................................................... 97
The gA6-l4/A6ENDSh interaction ................................................................ 98
The gND7-550/mND7550 interaction .......................................................... 101
Discussion .......................................................................................... 102
Materials and Methods ............................................................................ 108
Oligodeoxyribonucleotides ....................................................................... 108
RNA synthesis and labeling ...................................................................... 108
Surface Plasmon Resonance Studies ............................................................ 110
Acknowledgements ................................................................................ 1 12
Literature cited ..................................................................................... 113

CHAPTER 4 INTERACTIONS OF MRNAS AND GRNAS INVOLVED IN
TRYPANOSOME MITOCHONDRIAL RNA EDITING: CALCULATIONS OF RNA
BINDING AFFINITY FOR A GRNA BOUND TO ITS MRNA AS EDITING

PROGRESSES ..................................................................................... 1 17
Introduction ......................................................................................... 1 18
Results ............................................................................................... 125
RNA substrates ..................................................................................... 125
Equilibrium binding studies ...................................................................... 126
Surface plasmon resonance studies .............................................................. 130
The NgCYb—558/5’CYbUT Interaction ......................................................... 131
The NgCYb-558/5’CYbPES 1T Interaction .................................................... 142
The NgCYb-558/5’CYbPES3T Interaction .................................................... 150
Photoaflinity crosslinking and mapping of U5 and U10 of the CYb U-tail ............... 152
Discussion .......................................................................................... 169
Materials and Methods ............................................................................ 175
Oligodeoxyribonucleotides ....................................................................... 175
RNA Synthesis and Labeling ..................................................................... 175
Serial Dilutions of mRNA ........................................................................ 176
Gel Band-shift Analysis ........................................................................... 176
Dissociation rate gels .............................................................................. 177
Association rate gels ............................................................................... 178
Surface plasmon resonance ....................................................................... 180
RNA crosslinking and mapping of crosslinks ................................................. 183

viii

Acknowledgements ................................................................................ l 84

Literature Cited ..................................................................................... 185
CHAPTER 5 CONCLUSIONS AND FUTURE RESEARCH .............................. 188
Summary ............................................................................................ 189
Anchor binding site structure influences editing ............................................... 190
CYb mRNA/gRN A complex structure during editing ......................................... 19]
Kinetics of anchor target binding during editing ............................................... 192
Editosome recruitment ............................................................................. 196
Editing accessory factors .......................................................................... 199
MRP1 and MRP2 ................................................................................... 199
RBP16 ................................................................................................ 201
Literature cited ...................................................................................... 203

LIST OF TABLES

Table 1. Regulation of the RNA editing of the T. brucei mRNA transcripts ............... 15
Table 2. RNA editing complexes and the protein components .............................. 18
Table 3. List of Oligodeoxyribonucleotides used in Chapter 2 ............................... 76
Table 4. List of all Biacore substrates used in experiments in Chapter 3 ................... 92

Table 5. List of rate constants and dissociation constants for A6, CYbU, and ND7 from

Biacore experiments ....................................................................... 101
Table 6. List of Oligodeoxyribonucleotides used in Chapter 3 .............................. 109
Table 7. List of oligoribonucleotides used in Chapter 3 ..................................... 1 10
Table 8. Concentrations of mRNAs used in gel band shift assays ......................... 129
Table 9. List of rate constants and dissociation constants calculated from gel band shift

assays .................................................................................... 141
Table 10. List of all Biacore substrates used in Chapter 4 .................................. 143
Table l 1. List of rate constants and dissociation constants calculated from CYb Biacore

data ....................................................................................... 147
Table 12. Oligoribonucleotide in Chapter 4 ................................................... 154
Table 13. List of Oligodeoxyribonucleotides used in Chapter 4 ............................ 154
Table 14. Comparison between gel shift data and SPR data .......................................... 173

LIST OF FIGURES

Figure l. The life cycle of the trypanosome ...................................................... 4
Figure 2. Diagram of energy metabolism in the mammalian host ............................. 8
Figure 3. Diagram of energy metabolism in the insect host .................................... 9
Figure 4. Diagram of mitochondrial respiration complexes .................................. 10
Figure 5. Maxicircle, minicircle, and gRNA diagrams ........................................ 13
Figure 6. Alignment of the DNA/edited mRNA/ and amino acid sequences and
alignments of the gRNAs .............................................................. 17
Figure 7. Diagram of the mRNA/gRNA complex ............................................. 22
Figure 8. Predicted structures of the apocytochrome b RN As .......................................... 30
Figure 9. Secondary structure of the CYb mRNA ............................................. 32
Figure 10. Predicted models for mRNA/gRN A secondary structure ........................ 48
Figure 11. Sequence alignments of gRNA/mRN A substrate pairs ........................... 50
Figure 12. Structure probing of the gRNA, NgCYb-558 ...................................... 52
Figure 13. Structure probing of the partially edited mRN A, CYbPES3 ..................... 56
Figure 14. Chemical structure probing of the partially edited mRNA, CYbPES3.........59

Figure 15. Summary figures from solution structure probing gels for the mRNA/gRN A
complex .................................................................................. 61

Figure 16. Summary figure of CYbPES3 mRNA alone from solution structure probing
gels ........................................................................................ 68

Figure 17. The structure of the MRPl/MRPZ heterotetramer binding two gRNAs ......... 73
Figure 18. Predicted secondary structures for A6, ND7, and CYbU RN As ............... 87

Figure 19. Alignment of the anchor helices for A6, ND7, and CYbU RNAs .............. 93

xi

LIST OF FIGURES (CON’T)

Figure 20. Separate association and dissociation line fits of Biacore graphs for CYb, A6,

and ND7 experiments .................................................................. 99
Figure 21. Predicted models for the CYb mRN A/gRN A complexes ...................... 120
Figure 22. Sequences of the RNAs used for Chapter 4 experiments ....................... 123
Figure 23. Predicted secondary structures for the CYb mRNAs ............................ 127
Figure 24. Gel band shift assays to calculate the apparent dissociation constant... . . ....133

Figure 25. Graphs showing the analysis of the CYb apparent dissociation constants
calculated form the gel band shift assays ........................................... 135

Figure 26. Gel shifts used to calculate the observed rate constants and the dissociation
rate constants ........................................................................... 139

Figure 27. Graphs used to calculate the observed rate constants and the dissociation rate
constants ................................................................................ 140

Figure 28. Separate association and dissociation line fits of the CYb biacore
sensograms ............................................................................. 144

Figure 29. Gel showing the crosslink bands 1 and 2 for the U5 and U10 crosslinks. ...155
Figure 30. Primer extension analysis ofthe crosslink bands 157
Figure 31. RNase H analysis of the U10 crosslinks ........................................... 162
Figure 32. Diagram of the U5 and U10 crosslinks on the mRNA alone and the
mRNA/gRNA complex ............................................................... 165

Images in this dissertation are presented in color.

xii

Equation 1.
Equation 2.

Equation 3.

Equation 4.
Equation 5.
Equation 6.

Equation 7.

Equation 8.

Equation 9.

LIST OF EQUATIONS
Equation to calculate the association constant for Biacore results ........... 1 11
Equation to calculate the dissociation rate constant for Biacore results. . ..1 11

Modified equation to calculate the dissociation rate constant for Biacore

results ................................................................................ 112
Equation to calculate the dissociation constant fi'om rate constants. . . . . . l 12
Equation to calculate dissociation constant from gel shifts ................... 177
Equation to calculate the dissociation rate constant from gel shifts. . . . . l 78

Equation to calculate the observed rate constant using a single exponential

fit ..................................................................................... 179
Equation to calculate the observed rate constant using a double exponential

fit ..................................................................................... 179
Equation to calculate the association rate constant ............................ 180

xiii

A

ABS

C

CYb
CYbU
CYbPESl
CYbPES3
DEPC

G
gCYb-558
gRNA

Kr)

kobs

korr

kon
MBN

NgCYb—558
NgCYb-558sU
nt

RU

SPR

T1

T2

U

U-tail

UV

V1

KEY TO ABBREVIATIONS

adenosine

anchor binding site

cytosine

apocytochrome b

partial CYb mRN A Unedited

partial CYb mRN A Partially Edited through Site 1
partial CYb mRN A Partially Edited through Site 3
diethyl pyrocarbonate

guanosine

initiating CYb gRNA

guide RNA

dissociation constant

observed rate constant

dissociation rate constant

association rate constant

mung bean nuclease

new gCYb—558 gRN A with a U-tail
new gCYb-558 gRN A without a U-tail
nucleotide

resonance unit

surface plasmon resonance

RNase T1

RNase T2

uridine

uridine tail

ultra-violet light

RNase V1

xiv

CHAPTER 1

INTRODUCTION

History of Human African Trypanosomiasis

Trypanosomes are motile, protozoan parasites with a life cycle that includes two hosts,
insects and mammals. Trypanosoma brucei gambiense and T rypanosoma brucei
rhodesiense are African species that infect man and wild game; they cause the disease
human African trypanosomiasis (HAT), also known as sleeping sickness. HAT threatens
over 60 million people in 36 sub-Saharan African countries. Only 3 to 4 million of those
at risk have access to health care that provides screening and treatment (WHO, 2001).

T typanosoma brucei bmcei (T brucez), the form studied in our lab, does not infect man
but causes the wasting disease, nagana, in ruminants. The effects of nagana are equally
devastating to the population as the disease places a major constraint on agricultural
development in many of the poorer countries. Nagana infects 46 million cattle in Africa
and causes the death of three million cattle a year (Unisci, 2001). The economic loss
associated with African trypanosomiasis is estimated to be $4.5 billion each year
(Mattock & Pink, 2003-2004).

T. brucei is transmitted through the saliva of a bite from a tsetse fly (Glossina ssp.) as
it takes a blood meal from the new vertebrate host. The area of Africa where the tsetse
fly breeds covers over 10 million square kilometers (WHO, 2001; Kuzoe & Schofield,
2004) creating a major challenge to both the prevention of parasite transmission as well
as treatment of infected humans and animals. Other problems associated with treating
humans with trypanosomiasis include inadequate resources, surveillance, disease
knowledge, and diagnostic tools. The drugs available are expensive; most cause adverse
side effects and cannot be administered without medical aid (Mattock & Pink, 2003-

2004).

Historically, the spread of infection was primarily controlled by destroying the
breeding grounds of the tsetse fly, spraying of pesticides, and reduction of wild game
populations. Using these techniques, the number of cases of trypanosomiasis (over
50,000 reported cases of infection in 1940) was drastically reduced by the 1960’s (less
than 10,000 reported cases of infection) (Kuzoe & Schofield, 2004). Unfortunately, with
the decrease in the numbers of afflicted, interest in HAT declined resulting in a reduction
of control activities and a resurgence of human and animal trypanosomiasis (with close to
40,000 reported cases of infection by 2000). In addition, political instability, wars, and
civil unrest have led to further blocks in control activities (Kuzoe & Schofield, 2004).
The World Health Organization (WHO) estimates that 300,000 to 500,000 people are in
danger of HAT infection every year, but the instability of the most affected areas means
there are no accurate numbers to support these estimates (Mattock & Pink, 2003-2004).

A renewed interest in HAT prevention, through WHO and other non-profit groups has
begun measures to try to reduce infection. Past methods of tsetse vector control
described earlier are no longer recommended as they cause other secondary
environmental concerns. Newer methods of control include traps and screens to capture
and kill flies, live bait techniques (treating cattle with insecticides), and sterile insect
technique (releasing mass numbers of sterile male tsetse flies) (Kuzoe & Schofield,
2004).

Trypanosomes

The life cycle of the trypanosome includes two hosts. Part of the life cycle is spent in

an insect vector (procyclic stage) and part in a mammalian host (bloodstream stage) (Fig.

1). When an infected tsetse fly takes a blood meal, the trypanosomes are injected through

Slender trypomastigote
Sparse, short tubular
cristae ”“‘

  
    
 
  
     
    
  
 
 

Intermediate trypomastigote
5" Cristae lengthen

Metacyclic

trypo-
ma stigote St
’ umpy

Closely-packed trypomastigote

tubular cristae
Many tubular

cristae

Epimastigote

Midgut and cardia
trypomastigotes

In tsetse mes

Figure 1. The life cycle of the trypanosome includes mammalian and insect hosts. In the
tsetse fly, the mitochondrion is large and reticulated, and ATP is produced through
electron transport and oxidative phosphorylation. In contrast, during the bloodstream
stage of the life cycle, the trypanosome derives most of its energy from glycolysis. The
mitochondrion is less active and has a compact morphology. From: Vickerman, K. 1971.
In Fallis, A.M., ed. Ecology and Physiology of Parasites. University of Toronto Press,
Toronto.

the salivary glands and enter the bloodstream. In the bloodstream, the parasites
proliferate in waves as morphologically slender forms. They are able to thwart the
human immune system by having a protein coat composed of a variant specific
glycoprotein (VSG) with thousands of genetic variations that change with each wave of
parasitemia (V ickerman, 1985). During each proliferative wave, as parasite numbers
increase, differentiation to the short, stumpy form occurs. Cell division is arrested in this
form, and the trypanosome is pre-adapted for transmission to the insect host. When an
insect becomes infected through a trypanosome enriched blood meal, the stumpy forms
travel to the mid-gut of the insect where they develop into the procyclic form Procyclic
forms lose the VSG coat and develop a coat of procyclins. Trypansomes in the mid-gut
arrest in division and migrate to the salivary glands. They attach themselves to the walls
of the salivary gland as epirnastigote forms to proliferate. As epirnastigote numbers rise,
the trypanosomes differentiate into metacyclic forms that are non-proliferative and
develop the VSG coat. The metacyclic forms are then ready to infect a new host
(Matthews, 2005).

One unique feature in trypanosomes is the change in energy metabolism that is
reflected in the extreme morphological difference in mitochondrion structure between the
two hosts (Vickerman, 1965). Trypanosomes do not store carbohydrates, so they must
obtain their energy supply directly from a host. In the insect vector, the trypanosome has
a fully active and more developed mitochondrion that produces ATP through electron
transport and oxidative phosphorylation. During this stage, the mitochondrion grows
long and reticulated and develops a mitochondrial network where most of the cristae are

plate-like (Brown et al., 1973). In the mammalian host, the trypanosome oxidizes

glucose from the bloodstream using an organelle unique to trypanosomes, the glycosome,
while most mitochondrial functions are suppressed. Morphological changes in the
structure of the mitochondrion reflect its suppressed function as it becomes a linear canal
extending the length of the trypanosome and containing some tubular cristae (Brown et
a1., 1973). Glycosomes are peculiar peroxisomes that compartmentalize the first seven
steps of glycolysis as well as the pentose phosphate pathway. The changes in
morphology and content of this organelle lead to large differences in the metabolism of
trypanosomes in the separate hosts. The trypanosome is able to effect a rapid change in
its development, morphology, and energy metabolism in response to an instantaneous
difference in temperature (insect 27°C, mammal 37°C), cellular environment, and
immune response that is brought on by a change of host (Brown & Neva, 1983).
Metabolism of Trypanosomes

The metabolism of trypanosomes allows a large amount of flexibility in order to
respond to sudden changes in nutrition as well as a change of host. In the mammalian
bloodstream, compartmentalization of glycolysis in the glycosome appears to regulate the
amount of ATP that is accessible to hexokinase and phosphofi'uctokinase (Fig. 2). These
glycolytic enzymes are unusual as they are unregulated in trypanosomes and could cause
a toxic accumulation of glycolytic intermediates if allowed unlimited access to ATP
(Hannaert et a1., 2003). The glycosome allows no net change in ATP to occur through
the first seven steps of glycolysis. Afterward, 3-phosphoglycerate is shuttled to the
cytoplasm for the last three steps of glycolysis to produce two net ATP and pyruvate.
Even though the mitochondrion is largely repressed during the mammalian stage, a

putative glycerol 3-phosphate/DHAP shunt runs between the glycosome and the

mitochondrion. This shunt helps to maintain the NAD+/NADH balance with the help of
a glycerol 3-phosphate dehydrogenase and the terminal alternative oxidase (Michels et
a1., 2000). During this stage of the life cycle, the mitochondrion lacks most Krebs cycle
enzymes as well as a cytochrome containing respiratory chain. However, it does
maintain an electron membrane potential through FoFl-ATPase (Nolan & Voorheis,
1992; Schnaufer et a1., 2005).

When living in a procyclic host, the trypanosome has a more elaborate energy and
carbohydrate metabolic network (Fig. 3) that includes the mitochondrial Krebs cycle and
a respiratory chain (Fig. 4) (Hannaert et a1., 2003; Coustou et a1., 2005). The glycosome
enzyme content changes and resembles succinic fermentation in anaerobic organisms
versus lactic acid fermentation favored while in the mammalian bloodstream. This shift
toward succinic fermentation allows the trypanosome to require 50% less pyruvate to
maintain energy production (Hannaert et a1., 2003). During this stage of the life cycle,
the trypanosome is able to metabolize both glucose and amino acids (mainly proline) as
its energy source (Lamour et a1., 2005). However, it appears that neither energy source is
fully converted to C02 despite the presence of a firlly functional Krebs cycle and
respiratory chain. Instead, glucose is converted to succinate, acetate, lactate, and alanine,
while proline is converted to succinate (Coustou et a1., 2005).

Kinetoplast

The mitochondrion of trypanosomes is called a kinetoplast because of the unique and
complex structure formed by the kinetoplast DNA (kDNA). The mitochondrial DNA in
these organisms is incredibly bizarre, consisting of thousands of minicircles (1.0 kb each)

and 40-50 maxicircles (23 kb) that are topologically interlocked forming a compact disk

Energy Metabolism in the Bloodstream-Long,
Slender form

   

Glycosome

   
 

Glucose
Mitochondrion

Terminal

N08 N04
N09 ND7
ND5

Figure 2. Diagram of energy metabolism in the mammalian host. The glycosome is a
peculiar organelle that houses the first seven steps of glycolysis as well as much of the
pentose phosphate pathway. During the bloodstream stage, trypanosomes use ghrcose as
their rmin energy source. The glycosome is thought to act as an energy regulator to keep
glycolytic intermediates from accumulating. The last three steps of glycolysis are
cytoplasmic afier 3-phosphog1ycerate is shuttled to the cytoplasm to result in two net
ATP. The mitochondrion is largely repressed; however, a putative glycerol 3-
phosphate/DHAP shunt to the mitochondrion is thought to help maintain NAD+INADH
balance via glycerol 3-phosphate dehydrogenase and the terminal alternative oxidase.
The transcription and editing of some of the mRNAs from Complex I of the respiratory
chain are upreguhted in the bloodstream stage. A FoFl-ATPase also maintains an
electron membrane potential as well. A6 = A'I'Pase subunit 6, BPGA = 1,3-
bisphosphoglycerate, DHAP = dihydroxyacetone phosphate, G-3-P = glyceraldehydes 3-
phosphate, ND = NADH dehydrogenase, 3PGA = 3-phosphog1ycerate, QHZ/Q =
ubiquinone. Hannaert, V. et a1. (2003) Evolution of energy metabolism and its
compartmentation in Kinetoplastida. Kinetoplastid Biology and Disease. 2:11.

Energy Metabolism in the Procyclic Form
Glycosome

Mitochondrlon

Glycerol-3—phosphate
oxidase ,. ; ‘7 ¢

Substrate-level
Phosphorylation

 

Figure 3. Diagram of energy metabolism during the procyclic stage of the life cycle.
During the insect stage, the trypanosome requires a more elaborate metabolic network.
The glycosome enzyme content switches to succinic fermentation similar to anaerobic
organisms in order to maintain energy production in the absence of a steady glucose
supply. Using the TCA cycle and respiration, the trypanosome uses amino acids (mainly
proline) and any glucose available as its main energy sources. DHAP =
dfliydroxyacetone phosphate, Gly-S-P = glycerol 3-phosphate, PEP =
phosphoenolpyruvate. Hannaert, V. e! at (2003) Evolution of energy metabolism and its
compartmentation in Kinetoplastida. Kinetoplastid Biology and Disease. 2:11.

Figure 4

ease—being... cage—yea
i. \ll/ 8223
.F< 3+ as of

a 0200532
\ into \. on: no
_>_ erEooL gang—Eco _

              

   
 
 

eon—=5» n._.<
> 53:30

  

  

2. xofiEeo

maxeano congamem 3:20:85).

    

. g
. A f 33.582:
_ , «5858.5
A .35..

. .

   

ocean
2.955253...

10

Figure 4. Diagram of the mitochondrial respiration complexes. Complexes 1, HI, IV, and
V are present as well as the alternative oxidase. Complex II is predicted to be in
trypanosome mitochondria, because a succinate dehydrogenase activity and succinate-
dependent respiration are present. However, the actual proteins for the complex have not
been verified. Cyt = cytochrome, Cu = copper center, H+ = proton, FeS = iron/sulfur
cluster, FMN = flavin mononucleotide, NADH = Nicotinamide adenine dinucleotide in
its reduced form, Q = ubiquinone. Hannaert, V. et al. (2003) Evolution of energy
metabolism and its compartmentation in Kinetoplastida. Kinetoplastid Biology and
Disease. 2:11.

11

structure (Lukes et a1., 2002). Kinetoplast maxicircle DNA (Fig. 5A) is similar to other
mitochondrial DNA in that it is circular, and encodes rRNA and some essential protein
components for mitochondrial respiration. However, the majority of the mRN As that are
encoded on the maxicircle are not translatable. These mitochondrial transcripts may
contain conserved flame shifts, internal stop codons, and in some cases lack start codons
(Koslowsky, 2004). RNA editing in kinetoplasts is a post-transcriptional process that
involves the precise insertion and deletion of uridine residues in mitochondrial messenger
RNAs (mRN As). These changes are made by several proteins that collectively form the
editosome complex and are directed by a small RNA molecule called a guide RNA
(gRNA). This editing process is required to make many of the mitochondrial transcripts
translatable.

The first report of RNA editing showed that a flame shift in the cytochrome oxidase II
gene encoded by the mitochondrial DNA was corrected in the mRN A transcript by the
addition of four uridine residues (Benne et a1., 1986). Uridine deletion was also found to
occur (Shaw et a1., 1988) during editing. Editing can repair conserved flame shifts,
internal stop codons, and in some cases create the proper start codon through the precise
insertion or deletion of uridines (Feagin et al., 1988b). Some mitochondrial transcripts
have more than 50 percent of their reading flame added through editing (Feagin et al.,
1988a)

The template for editing is provided by gRNAs that are usually encoded on
minicircles (Fig. 5B) (usually 3 gRNAs per minicircle) (Sturm & Simpson, 1990). Two
gRNAs are encoded on the maxicircle (Clement et al., 2004; Golden & Hajduk, 2005);

one of these appears to use a novel method of in cis gRNA binding for RNA editing of

12

A T. brueel mltoehondrlal Maxlclrcle DNA

 

_:" ",.,_-,.__... .._....,, ....=. . gs; ,t. ..-...... , .. ,- . .........-. ’- uuuuuuuuuuuu
5’ anchor guldlng raglan uridine tail

Figure 5. Maxicircle, minicircle, and gRNA diagrams. A. Maxicircle DNA resembles
other eukaryotic mitochondrial DNA; it encodes rRNAs (9S and 12S) and some essential
components of the mitochondrial protein respiration complexes. Many of these
mitochondrial transcripts require editing before they can be translated. B. Minicircles
encode ~3 gRNAs usually in between 18 bp inverted repeats. The gRNAs provide the
templates for editing the mRNAs. C. A gRNA (50-70 nts) consists of 3 firnctional
elements: an anchor sequence that is complementary to the mRNA, the guiding region
provides a template for RNA editing, and a U—tail added post transcriptionally with
probable multiple functions. 128 = 128 rRNA, 98 = 98 rRNA, A6 = ATPase subunit 6,
CO = cytochrome oxidase, CR3 = C-rich region, CYb = apocytochrome b, Murf=
Maxicircle Unidentified Reading Frame, ND = NADH dehydrogenase, RSP12 =
rrbosomal protein 12. Courtesy of Donna Koslowsky.

13

the C011 transcript (Golden & Hajduk, 2005). Guide RNAs have an average length of
50-70 nts and consist of three functional elements (Fig. 5C). Contained within the 5’ end
of gRNAs is a short sequence known as the gRNA anchor, a 5-21 nucleotide region that
base pairs with a particular mRN A just 3’ of the editing domain (Blum et al., 1990). The
second element, the guiding region, serves as a template for the editing process and is
complementary (allowing G-U base pairs) to the mature mRNA (Blum et a1., 1990).
Finally, at the 3’ end of the gRN A is a poly-uridylate tail (U-tail) that is added post-
transcriptionally (Blum & Simpson, 1990).
Developmental Regulation of RNA Editing

Many of the proteins for energy metabolism that are made in the mitochondria are
thought to be developmentally regulated by RNA editing (Table 1). Cytochrome b (CYb)
and Cytochrome oxidase 11 (C011) are preferentially edited in the procyclic host of the T.
brucei life cycle, while NADH dehydrogenase subunit genes 8 and 9 (ND8, ND9) are
preferentially edited in the bloodstream host (Priest & Hajduk, 1994). NADH
dehydrogenase subunit 7 (N D7) is fiilly edited in bloodstream form but only the 5’
domain is edited in the insect host (Koslowsky et a1., 1992). Other substrates, ATPase 6
subunit (A6) and cytochrome oxidase III (COIH), appear to have consistent levels of
editing in either host (Bhat et a1., 1990).

Developmental regulation of RNA editing may control mitochondrial biogenesis by
only allowing production of respiratory proteins during the correct life cycle. However,
very little is known concerning how editing is regulated. The gRNA transcripts, the

mRNA transcripts, and the editosome protein complexes are always present

14

 

 

mRNA Mitochondrial Fully Higher Steady
Complex/ Edited/Higher State mRNA
Function Levels Level
NDl I Never Edited Constitutive
ND3 I Bloodstream Bloodstream
ND4 I Never Edited Bloodstream
NDS I Never Edited Bloodstream
ND7 I 5’both/3’BS Bloodstream
ND8 I Bloodstream Bloodstream
ND9 I Bloodstream Bloodstream
CYb III Procyclic Procyclic
COI IV Never Edited Procyclic
COII IV Procyclic Procyclic
COIII IV Constitutive Procyclic
A6 V Constitutive Constitutive
MURFI Unknown Never Edited Bloodstream
MU RFII Unknown Constitutive Constitutive
CR3 Unknown Bloodstream Unknown
CR4 Unknown Bloodstream Unknown
RSP12 Ribosomal Bloodstream Bloodstream

 

Table 1. Regulation of the RNA editing of the T. brucei mRNA transcripts. Column 1:
mitochondrial mRNA. Column 2: mitochondrial complex is which the protein is thought
to function. Column 3: life cycle stage that the transcript is preferentially edited in.
Column 4: life cycle stage where steady state level of mRN A is upregulated. The ND7
transcript is differentially edited in the two hosts; the 5’ end is edited in both hosts, while
the 3’ end is edited in the mammalian bloodstream. 128 = 128 rRNA, 9S = 98 rRNA, A6
= ATPase subunit 6, CO = cytochrome oxidase, CR3 = C-rich region, CYb =
apocytochrome b, Murf = Maxicircle Unidentified Reading Frame, ND = NADH
dehydrogenase, RSP12 = ribosomal protein 12. Hajduk SL, Sabatini RS. 1998.
Mitochondrial mRNA Editing in Kinetoplastid Protozoa. In: Grosjean H, Benne R, eds.
Modification and Editing of RNA. Washington, DC: ASM Press. pp 377-411.

15

(Koslowsky et a1., 1992; Hajduk & Sabatini, 1998). However, there may be
developmentally regulated proteins that are involved in regulating RNA editing through
gRNA recruitment, RNA-RN A annealing activity or chaperone activity.
Apocytochrome b

Most of my work will focus on the study of the apocytochrome b (CYb)
mRN A/gRN A pair (Fig. 6). Expression of the CYb mRN A is regulated through RNA
editing (Feagin et al., 1988b) and is needed for mitochondrial respiration in the insect
host (Vickerman, 1965). Editing of the apocytochrome b mRN A requires the insertion of
34 uridylates at 13 sites near its 5’ end (Fig. 6A). The editing process that creates the
CYb initiation codon occurs preferentially during the procyclic (insect) and stumpy
bloodstream stages of the trypanosome life cycle (Feagin et a1., 1987; Feagin & Stuart,
1988). There are three redundant gRNAs that provide the template for initiating CYb
editing (Fig. 6B) and at least one more is thought to exist for downstream editing (Riley
et a1., 1994). The initiating gRNA, gCYb558, directs the first seven editing events where
21 uridylates are inserted.
The Editosome

The latest model for RNA editing in trypanosomes involves >20 proteins (Table 2)
that form a complex structure of proteins called the editosome around an mRNA/gRNA
pair (Simpson et a1., 2004; Stuart et a1., 2005). The naming scheme for these 20 proteins
includes kinetoplast RNA editing (KRE) at the beginning of every editosome protein
name. The insertion and deletion of uridines occurs through successive rounds of

enzymatic reactions. These enzymatic activities are coordinated by the editosome and

16

A CYb DNA aligned to edited mRNA and AA seguence:

1 --------- + --------- + --------- +-- - - -- - --+ - - 42
DNA GTTAAGAATAATGGTTATAAATTTTATATAAA A G CG G AGA A A
RNA-ed GUUAAGAAUAAUGGUUAUAAAUUUUAUAUAAAuAuGuuuCGuuGuAGAuuuuuAuuAuuu
1 --------- + --------- + --------- + --------- + --------- + --------- + 60
AA M F R C R F L L F
1 - - - - - - - - - 9
43 - - ----- + - -------- + --------- + --------- + ------ 86

A A AGAAA G G GTCTTTTAATGTCAGGTTGTTTATATAGAATATAT
uuuuuAuuAuuuAGAAAuuuGuGuuGUCUUUUAAUGUCAGGUUGUUUAUAUAGAAUAUAU
61 --------- + --------- + --------- + --------- + --------- + --------- + 120
F L L F R N L C C L L M S G C L Y R I Y
10 + - - - - - - — - - + - — - - - - - - - 29

B
CYb mRNA aligned with gCYb558:

5’AuAuGuuuCGuuGuAGAuuuuuAuuAuuuuuuuuAuuAuuuAGAAAuuuGuGuuGUCUUUUAAUGUCAG

||===|||||=|==l|=|||||====||||||||||l|#=l

gCYb-SSB-B’AAGGGAAAUAGUGGAUUUUUAAGUGUAACAGAAAAUUAGGG5'

gCYb RNA seguences aligned:

gCYbS 6 GA GGAGAUAGUAAAAGACAAUG UAGAUUUCUGAGUAAUGGGGAGGAUAACUACUCUCUAGGGAAGAAAAU
gCYbS 6 03 GGAGAUAGUAAAAGACAAUGUAGAUUUCUGAGUAAUAGGGAGGAUAACUACUCUCUAGGGAAGAAAAU
gCYbS 6 0 C GGAGAUAGUAAAAGACAAUG UAGAUUUCUGAGUAAUGGGGGGGAUAACUACUCUCUAGGGAAGAAAAU
gCYbS 5 8 GGGAGAU - - UAAAAGACAAUGUGAAUUUUUAGGUGAUAAAGGGAAUAAUUA

.tiiiii. . .*i***********'****.*. .ii.**. . .*.*.****.** ..................

Figure 6. A. The DNA sequence of the 5’ end of the apocytochrome b gene is aligned
with the 5’ end of the edited RNA transcript (RNA-ed) and the amino acid sequence
(AA). (A. Estevez and L. Simpson, Unpublished,
http://dna.kdna.ucla.edu/trypanosomelseqs/tbcybedmap.html). The 34 uridine residues
that are inserted into the 13 editing sites near the 5’ end of the gene are shown in
lowercase. The DNA codes for 1118 nucleotides. After editing, the mature RNA
transcript is 1152 nucleotides long and codes for a protein that is 370 amino acids long.
B. Part of the edited mRN A is aligned with the wild-type initiating gRNA, gCYb558.
The wildtype gCYb gRN As are aligned below the mRNA/gRNA alignment (gRNA
sequences flom http://biosun.bio.tu-dannstadt.de/goringer/gRNA/gRNA.html). # =
mismatched sequence, “z” = GU base pair, “l” for Watson-Crick base pairing, * =
matching sequence, - = gap.

17

 

 

 

 

 

 

Protein U insertion! Function Functional Domains
Name deletion
KREPA1 Both Interaction 1 zinc finger,1 zinc-like finger, OB fold
KREPAZ U deletion Interaction 2 zinc finger, OB fold
KREPA3 Both Interaction 2 zinc finger, 08 fold
KREPA4 Both Interaction OB fold-like
KREPAS Interaction OB fold-like?
KREPA6 Both Interaction OB fold-like
KREN1 U deletion endonuclease U1-Iike zinc finger, RNaselll, dsRBD
KREPBZ Both endonuclease U1-like zinc finger, RNaselll, dsRBD
KRENZ U insertion endonuclease U1-Iike zinc finger, RNaselll, dsRBD
KREPB4 Both Interaction U1-Iike zinc finger, RNaselIl-like, Pumilio
domain
KREPBS Both Interaction U1-Iike zinc finger, RNasellI-Iike, Pumilio
domain
KREPB6 Interaction U1-like zinc finger
KREPB'I U insertion Interaction U1-like zinc finger
KREPBB U deletion Interaction U1-Iike zinc finger
KREX1 U deletion Exanse 5‘3'exonuclease, endo, exo, phosphatase
KREX2 Both? Exanse 5'3'exonuclease, endo, exo, phosphatase
KREL1 U deletion RNA ligase Ligase, tau (microtubule assoc), kinesin light
chain
KREL2 U insertion RNA ligase Ligase, tau (microtubule assoc), kinesin light
chain
KRET2 U insertion TUTase Nt. transferase domain, core,
. Poly(A)polymerase associated domain
KREH1 transient helicase Helicase
KRET1
complex
KRET1 ? gRNA TUTase zinc finger, poly-A polymerase catalytic and
associated domains
7
MRP
Complex
MRP1 ? RNA R-rich domain
matchmaking
MRP2 ? RNA R-rich domain
matchmaking
Others
RBP16 ? Interaction cold shock domain, RGG RNA binding domain
REAP-1 ? Interaction 21 AA repeat
TbRGG1 ? Interaction RGG RNA bindinLdomain

 

Table 2. RNA editing complexes and the protein components. Column 1: name of the
proteins. Column 2: involved in U insertional editing, U deletional editing, or both.

Column 3: putative function. Column 4: lists the motifs found in each protein. Stuart KD,
Schnaufer A, Ernst NL, Panigrahi AK. 2005. Complex management: RNA editing in trypanosomes. Trends
Biochem Sci 30:97-105. Carnes J, Trotter JR, Ernst NL, Steinberg A, Stuart K 2005. An essential RNase
lll insertion editing endonuclease in Trypanosoma brucei. Proc Natl Acad Sci U S A 102: 16614-16619.
Panigrahi AK, Ernst NL, Domingo GJ, Fleck M, Salavati R, Stuart KD. 2006. Compositionally and
functionally distinct editosomes in Trypanosoma brucei. RNA 12: 1038-1049.

18

templated by the gRN A (Blum et a1., 1990; Blum & Simpson, 1990; Seiwert & Stuart,
1994; Adler & Hajduk, 1997; Simpson et a1., 2004; Stuart et a1., 2005). After gRNA
binding, editing begins with the endonucleolytic cleavage of the pre-mRN A at the first
editing site via either an insertional (KREN2) or deletional (KREN1) endonuclease
containing an RNase 111 domain (Panigrahi et al., 2003b; Carnes et a1., 2005; Trotter et
a1., 2005; Kang et a1., 2006). Uridine residues not base paired with the gRNA are thought
to be removed via an editosome U-specific 3’ to 5’ exonuclease (Exanse); there are two
candidate Exanses with 5’ to 3’ exonuclease motifs (KREX1 and KREX2) (Aphasizhev
et a1., 2003a; Panigrahi et al., 2003b). KREX1 appears to be involved in U deletion;
however KREX2 appears to be involved in both U insertion and U deletion and may be a
“U-trimmer” (Panigrahi et a1., 2006). Uridine residues are added via a terminal uridylyl
transferase (TUTase) (Aphasizhev et a1., 2002; Ernst et a1., 2003; Panigrahi et al., 2003b).
RNAi studies indicate that the most likely candidate for the editing TUTase is KRET2
(Aphasizhev et al., 2003c). A second TUTase (KRET1) is found in an independent
complex and is thought to be involved in addition of the gRNA U-tail (Nebohacova et a1.,
2004). After uridine addition or deletion has occurred, an RNA ligase is required to
ligate the two halves of the mRNA. Again, there are two RNA ligases called kinetoplast
RNA editing ligase (KREL) 1 and 2 that appear to be the editing ligases (McManus et a1.,
2001; Panigrahi et al., 2001a; Schnaufer et a1., 2001; Cruz-Reyes et a1., 2002). Similar to
the endonucleases and exonucleases, one ligase, KRELl, appears to be for U deletion,
while the other, KRELZ, appears to be for U insertion (Huang et a1., 2001). However,
there is some question as to how strictly these roles for the ligases in the separate U

insertion or deletion complexes are upheld (Gao & Simpson, 2003). Additional protein

19

studies find many editosome proteins have RNA binding activities, are essential for
editosome complex assembly, and appear to be involved in RNA-protein interactions and
these include KREPA1-6 and KREPB4-8 (Panigrahi et al., 2001b; O'Hearn et a1., 2003;
Panigrahi et al., 2003a; Panigrahi et al., 2003b; Wang et a1., 2003; Kang et al., 2004;
Brecht et a1., 2005; Salavati et aL, 2006).

Significant progress has also been made in identifying proteins that appear to be
accessory factors that transiently act during the editing process (Table 2) (Koller et a1.,
1997; Missel et a1., 1997; Madison-Antenucci et al., 1998; Vanhamme et a1, 1998;
Hayman & Read, 1999; Aphasizhev et al., 2003b; Pelletier & Read, 2003). A few of
these appear to enhance the mRNA/gRNA annealing process (Muller et a1., 2001; Miller
et a1., 2006). The mitochondrial RN A-binding proteins (MRP), MRP1 and MRP2, are
accessory factors to the editing process that are arginine rich and bind to gRNAs (Koller
et a1., 1997; Blom et al., 2001). Both MRP proteins co-purify in a protein complex, and
when both proteins are knocked down in RN Ai experiments, there are very low amounts
of CYb mRN A editing (Vondruskova et a1, 2005). RBP16 is a 16 kDa protein that
binds U-rich sequences such as the gRNA U-tail. It contains an N-terminal cold shock
domain and a C-terminal region rich in arginine and lysine residues (Hayman & Read,
1999). The in vitro association between RBP16 and gRNA is increased in the presence of
p22, a human p32 homologue (Hayman et a1., 2001). RNAi knock-down of RBPl 6
results in a ~98% reduction in CYb mRN A editing (Pelletier & Read, 2003).

Under electron microscopy, the editosome appears to be a structure composed of 4
protein bulges (Stuart et a1., 2002). During the purification process for the editosome

complex, protein complexes of various protein compositions and sizes are found to co-

20

exist. In 1992, Pollard et al. discovered two different complexes could be obtained
through separation on glycerol gradients, a 408 complex and a 198 complex (Pollard et
al., 1992). Since then, discovery of these complexes as well as others has led to new
protein discoveries. These additional complexes that have been isolated by separate labs
appear to have different editing capabilities, and this suggests that insertion and deletion
editing could be physically and fiinctionally separate (Goringer et al., 1994; Peris et al.,
1994; Corell et al., 1996; Peris et a1., 1997; Rusche etal., 1997; Cruz-Reyes et al., 1998;
Stuart et al., 2002; Aphasizhev et al., 2003a; Panigrahi et al., 2003b; Schnaufer et a1.,
2003; Panigrahi et al., 2006). Many of the editosome proteins mentioned above appear to
have evolved in pairs. It has been suggested that these pairs then insert specifically
either into a U-deletion or a U-insertion complex or subcomplex. These subcomplexes
may form one large complex that coordinates the entire editing process, or the editosome
complexes may load and unload flom the mRNA/gRNA complex during editing
(Panigrahi et al., 2006).

Despite this progress, very little is known about how the editing complex is assembled
onto specific RNAs. There are hundreds of different mRNA/gRNA pairs, but no
conserved sequence domains have been found in mRNAs. The only conserved sequence
domain shared between the gRNAs is the U-tail. Previous work in this lab suggests that
different mRNA/gRNA pairs can form similar structures. Initially, computer predicted
models of three separate mRNA/gRNA pairs gave three very different structures. Upon
incorporating 3’ crosslinking results into a secondary structure modeling program, all
three mRNA/gRNA pairs formed similar structures of three helices that form around and

appear to expose the first few editing sites (Leung & Koslowsky, 1999). The three

21

3!

G “600“...
H. CGG E51 . \ ,
. / / new i co°°$\\\\p (’65
30¢,” MM 644 6 see \\ 096
UU 09,0
u
”AA AA Anchor Helix
U-tail Helix U \ A U
:I \6‘ G u6
\ u
3: u
c \ 9
AA U
A A gRNA stem-loop
U G
AGUG

Figure 7. Diagram of the mRN A/gRN A complex. The predicted model for
mRNA/gRNA complex secondary structure has three predicted helices: an mRNA/gRNA
anchor helix, a gRNA stem-loop of the guiding region, and a U-tail/mRNA duplex. ESl
shows the first editing site is just 5’ (mRN A) of the anchor helix.

22

helices consisted of a gRNA/mRN A anchor duplex, a U-tail/mRN A duplex and a gRNA
stemloop (Fig. 7). Initial solution structure probing of CYb mRN A crosslinked to gCYb-
558 supports this model (Leung & Koslowsky, 2001a). We believe that structure
recognition of the mRNA/gRNA complex may be important for efficient editosome
assembly with the mRNA/gRNA pair.

One possibility for editosome recruitment is that proteins with RNA binding sites are
needed to recognize different structural elements of the mRN A/gRN A complex and that
the local structure at the editing sites may determine which protein editing complex is
recruited. The first step in editing is the gRN A anchor binding the mRN A anchor
binding site and this anchor helix must form in order to correctly position the gRNA for
efficient editing. The endonucleases of the editosome have double-stranded RNA
binding domains that may target the anchor helix for binding to cleave the mRN A strand
(Panigrahi et al., 2006). In addition, one of the proteins that binds RNA, KREPA4,
appears to bind gRNA U-tails or U-rich sequences with a putative Sl motif similar to a
cold-shock domain (Salavati et al., 2006). Proteins that prefer U-rich sequence could
function as sensors to detect U’s needing deletion by targeting deletion complexes to
deletion sites. Additionally, there may be proteins of the editosome that bind various
elements of RNA structure in order to correctly position the editosome on the complex.
Discovering the structure of an mRN A/gRN A pair such as CYb may be the first step in
unraveling what RNA structure attracts the editosome, and how the RNA structure
changes when in contact with the editosome.

Our lab is interested in how the editosome group of proteins recognizes the

mRNA/gRN A complex. There appear to be many different specialized complexes for

23

editing that target hundreds of different pairs of mRN A/gRN A complexes. Our lab is
interested in gRNA targeting of the mRN A for binding and how this affects editing
efficiency. Specifically, we are interested in the mRNA/gRNA complex interaction and
understanding how the anchor sequence, U-tail, and guiding region interact with the
mRN A during RNA editing. Since there appears to be no sequence homology between
the many mRN As and their gRNAs that proteins can recognize, a common structure may
be what the editosome proteins recognize. We are interested in knowing what the
structure of the CYb gRNA/mRN A complex is as well as what role the structure of the
CYb mRN A plays in the binding affinity of the gRNA to the mRN A. We are also
interested in what role the U-tail has in the mRNA/gRNA complex interaction.
Additionally we would like to know what effects partial editing of the mRN A has on the
gRN A/mRN A complex structure as well as the kinetics of its formation.
Other RN A-RNA Interactions

Studying other well characterized RNA-RNA interactions may aid in our
understanding of this complex interaction. RNA-RNA interactions are important
regulators and tools in every organism. RNA structure is critical for many RNA-RNA
and RNA-protein interactions to display nucleotides in the correct orientation for efficient
interaction. The versatility of RNA interactions allow organisms to control multiple
processes and relies on diverse structures, mechanisms, and biological roles (Altuvia &
Wagner, 2000). There appears to be some repetition in structure and method of target
identification that resembles and contrasts with the gRNA/mRN A binding interaction.
Here are a few well-studied examples of RNA target recognition of mRNAs that may aid

in the understanding of our gRNA/mRN A interaction.

24

Antisense RNAs

Antisense RNA regulation in prokaryotes is one example of regulatory RNAs that use
target binding to achieve post-transcriptional regulation of gene expression (Altuvia &
Wagner, 2000). The studies of antisense RN As flom bacteria also elucidate ways that the
structure of RNA affects target recognition and binding progression. Many antisense
RNAs function as regulators of plasmid copy number in prokaryotes. Antisense RNAI
binds its target RNAII to control the copy number of plasmid, ColEl. It does this by not
allowing RNAII to bind the plasmid near the origin of replication where RNase H would
digest RNAII to provide primers for DNA replication of ColEl (Tomizawa et al., 1981).
The RNAI-RNAII complex uses the preferred antisense target binding strategy of kissing
loops. This describes the first helix nucleation event that begins when two
complementary stem-loops, one or more flom each RNA molecule, begin forming a helix
with the two loop regions. This loop-loop interaction initiates helix formation
(Tomizawa, 1984). The RNAI-RNAII kissing complex is recognized by the Rom protein
that then stabilizes the complex and stops RNAII flom binding the plasmid (Eguchi &
Tomizawa, 1990, 1991). Most antisense RN As present their complementary sequence in
a loop structure that appears to be optimized to achieve rapid binding (Franch et al.,
1999). Another antisense RNA, CopA, regulates the copy number of plasmid R1 through
the translational repression of repA; CopA binds its target site CopT in the leader region
of the repA mRN A. Studies of the CopA antisense RNA show that modification of loop
regions and removal of bulge regions alter this RNA-RNA interaction. The bulges
facilitate binding between the RNAs by acting as helix destabilizing elements in the

binding intermediate structures (Hjalt & Wagner, 1995). When modifying loop regions,

25

disruption of a putative U-turn disrupts the structure of the loop in CopA (Slagter-Jager &
Wagner, 2003) as well as sok RNA (Franch et al., 1999). The lack of U-tum inhibits
binding at the nucleation site of the kissing loops, delaying formation of the stable
inhibitory structure to stop translation. Rapid binding of CopA and formation of the
inhibitory four helix junction is necessary to stop the ribosome flom binding (Slagter-
Jager & Wagner, 2003). The rate limiting step of binding of CopA to CopT is formation
of the loop-loop interaction that rapidly forms an early intermediate structure. Once the
kissing loops form, the probability of dissociation is small compared to conversion to the
stable inhibition complex (Nordgren et al., 2001). Antisense RNAs such as ColE 1 ,
CopA, etc are efficient as inhibitors, because they have high affmity target binding (Kolb
et al., 2001). These antisense RN As do not require protein co-factors for efficient target
binding. It has been postulated that the structure of these RNAs has evolved for optimal
target recognition leading to plasmid copy inhibition (Kolb et al., 2001). For antisense
RNAs, a fast association rate is more important than the thermodynamic stability (AG)
between the interacting RNAs (N ordstrom & Wagner, 1994; Altuvia & Wagner, 2000).
Antisense regulation requires co-transcriptional binding i.e. the antisense RNA must bind
before the ribosome binds to repress translation (Altuvia & Wagner, 2000). Therefore,
these antisense RNAs appear to have evolved structures that maximize the association
rate.

While CopA and other antisense RN As have been shown to bind their targets unaided,
other antisense RNA interactions with their targets are mediated by a protein co-factor
called qu. In E. coli, this sm-like protein was able to improve complex formation

between an antisense RNA (spot42) and its mRNA target (gal) by increasing the affinity

26

of the anti-sense RNA for its target ~150 fold (Moller et al., 2002). It does not appear to
alter the RNA structure of spot42 when bound, while it increases the affinity of the
antisense RNA for its target mRN A (gal). qu may act as a general co-factor that
facilitates RNA-RN A interactions (Moller et al., 2002). qu is also able to melt the
structure of a target mRNA (sodB) making it accessible to RyhB, a small RNA regulator
(Geissmann & Touati, 2004). Another study describes qu as a chaperone, because qu
is required to facilitate the OxyS RNA-RNA interaction but is dispensable after binding
has taken place (Zhang et al., 2002). There is speculation that an ancestral sm-like
protein played decisive roles in the folding, binding, and functions of RN As.
Establishment of short stretches of base pairing in bimolecular RNA interactions would
allow any short segment of exposed RNA to be used for target recognition (Moller et al.,
2002).
miRNAs

MicroRNAs (miRN As) also employ RNA-RNA target binding to control post-
transcriptional regulation of mRN As. Until recently, miRN As along with other non-
protein coding RN As were considered evolutionary debris (Mattick & Makunin, 2005).
However, it is becoming clear that an extensive RNA regulatory network exists in all
organisms that works independently and in parallel with the protein coding system
(Mattick, 2004).

MicroRN As are ~21-25 nt RNAs made flom longer hairpin precursors (primary
microRNA transcripts) (Lee et a1., 2003). These small RNAs target mRNAs for cleavage
or translational repression (Lai, 2003) through complementarity with a ~7 bp “seed”

sequence near the 5’ end (Lewis et al., 2003). This seed sequence is sufficient to confer

27

strong regulation by the miRN A (Brennecke et al., 2005). The miRN A is separated flom
its complement after integration into the RNA Induced Silencing Complex (RISC)
(Tomari et al., 2004; Matranga et al., 2005). The varying members of the RISC complex
determine the destiny of the transcript. However, the amount of complementarity
between the miRN A and the mRN A is thought to determine the fate of the mRN A
through recruitment of the different RISC complexes (Hutvagner & Zamore, 2002; Tang
et al., 2003).

MicroRNAs are excellent examples of mRN A regulation through target binding. The
seed sequence of miRNAs is very similar to the gRNA anchor sequence. It is
complementary to its target RNA and allows G-U bps. The complementarity between the
gRN A and the mRNA determines the amount and type of editing of the mRN A. This
correlates well with miRN As and how the amount of miRN A complementarity to its
target directs transcript cleavage or repression. The gRNA directs the editosome proteins
for U addition or deletion. Studies of the editosome show that the protein content is
dynamic with various editosome complexes composed of different complexes of proteins.
It would appear that U addition and U deletion complexes have differing complexes of
proteins (similar to RISC) suggesting alternate complexes must bind the mRNA/gRNA
complex in order to complete editing (Panigrahi et al., 2006). This alternate loading of
editosome complexes is directed by the gRNA; similar to the miRN A directing the
loading of the proper RISC complex.

RNA-RNA conclusions
Historically, antisense RNA studies focused on RNA-RN A interactions. The studies

that combine binding rate with structure changes are important first steps that laid the

28

groundwork for qu and antisense studies that were discovered later. The miRNA
involvement with the mRN A mirrors the gRNA binding to its target mRN A, while
loading of the RISC complexes onto the miRN A appear to be similar to the editosome
complexes loading onto the mRNA/gRNA complex.

Overview

The focus of this work is on the structure and thermodynamic interaction between the
CYb gRNA, gCYb-558, and its mRN A during the editing process. The many RNA-RNA
examples provide ideas and templates for discovering how this interaction proceeds. The
CYb interaction does not appear to copy any of these interactions, and yet elements of the
interaction are similar to other RNA-RNA interactions. It will be interesting to see how
this interaction compares to these examples of RNA target binding.

The structure of the CYb mRNA/gRNA pair was predicted to form three helices
(Leung & Koslowsky, 1999). Solution structure probing of the unedited CYb mRN A
bound to the gRNA verified two of the predicted helices (Fig. 8) (Leung & Koslowsky,
2001a). Curiously, the secondary structure predictions of two partially edited substrates,
with crosslinking data incorporated, produced similar structures with the same three
predicted helices. The anchor region becomes progressively longer after editing begins,
and the U-tail interaction with the purine rich region becomes increasingly shorter (Fig.
8). The gRNA stem-loop was predicted to be maintained by feeding portions of the U-
tail into this helix (Fig. 8) (Leung & Koslowsky, 2001b). Crosslinking data verified the
5’ end (gRN A orientation) of the anchor duplex as well as the 3’ position (gRNA

orientation) of the U-tail. The CYb U-tail was found to interact with the same 5

29

Accra-see

3’
UUUUUUUUUUUUUB — t G U G U— AGGGS'

A u A—U
- —u
u“ U if A
{:3 GMA
66" ”A ,
'.3
° ‘23 B CYbU+NgCYb-558 3,33
A cu fie. e 0 “Xe“
UA 3002/ / 90° \\ °¢ o
vol/I/Uouxcofigopr’
”(Q/owl], o}
“u A
u A
“AM: ‘90“:
C CYbPES3+NgCYb-558 z :3
i" 3:
" c a.
u
AA A
AU 63
,o ‘60
¢’/ valemuuucucuueucuuuuw ucaeouuee'
2 llllllll'IKIillllAléllllldll5.
cccc‘fi’u cu
\\ \ °
7 3 °
’7‘ 04 all

Figm'e 8. Predicted structures of the apocytochrome b RNAs. A. Predicted structure of
gCYb-558 flom Schmid, B. etal. (1995) NAR 23:3093-3102. The gRNA appears to form
two loop regions. One is a small anchor loop, while the second, larger stem-loop forms
in the guiding region. B. Predicted structure of 5’CYbUT+NgCYb—558 flom Leung, 8.8.
and Koslowsky, DJ. (2001) RNA 7:1803-1816. The solution structure probing
experiments prove that the mRNA is involved in two helices with the gRNA with the first
few editing sites being sensitive to single-stranded specific nucleases. C. Predicted
structure of 5’CYbPBS3T+NgCYb—558 flom Leung, 8.8. and Koslowsky, DJ. (2001)
NAR 29:703-709. The red U’s represent uridines inserted through RNA editing and
result in the anchor helix doubling in length. The 3’ end of the U-tail is predicted to
interact with the same region of the mRNA even after the third editing event even as the
U-tail is predicted to become incorporated into the gRNA stem-loop.

30

nucleotides upstream of the anchor duplex when base paired with the unedited and two
partially edited CYb mRN As (Fig. 8) (Leung & Koslowsky, 1999). In chapter 2 of this
thesis, the solution structure probing of the gRN A interacting with both unedited and
partially edited CYb mRNAs is presented which experimentally confirms the computer
predicted structures. This suggests that the formation of three helices surrounding the
editing site may be an important structural feature and that the U-tail may play an
important structural role. By employing an U-tail, the gRNA may increase the
mRNA/gRNA complex stability, while still allowing U-tail migration within the complex
during editing.

Additionally, comparisons of the thermodynamic binding aff'mities for two different
mRNA/gRNA pairs have been investigated. The affinity of the CYb mRN A/gRN A pair
was compared to the A6 mRNA/gRNA pair. The ATPase 6 subunit (A6) mRN A and its
initiating gRNA, gA6-14, is used in all in vitro studies of editing and is constitutively
edited. In contrast, the CYb mRN A is only edited in the insect and short stumpy stages
of the life cycle (Feagin & Stuart, 1988) and will not undergo a full round of editing in
the in vitro assay. The A6 mRNA/gRNA pair is predicted to have an open structure with
an accessible anchor binding site, while the CYb mRN A (Fig. 9) has been shown through
solution structure probing to form a stable stemloop structure that incorporates the
mRNA anchor binding site within the stem (Leung & Koslowsky, 2001a). The
difference in mRN A structure and accessibility of the anchor binding site is clearly
reflected in the measured apparent equilibrium constants. The A6 gRNA, gA6-l4, has a
very high affinity for its cognate mRN A, with measured Kp’s in the single digit nM

range. In contrast, the CYb gRNA, gCYb-558, has a very low afl'mity for its cognate

31

AG 6 /Es1
A G
G-@

I I
O>©<§

C)

c>c>>>>oo°o>°
I

I
>C>CCCCCG

Figure 9. Predicted secondary structures of the unedited CYb mRN A. 5’CYbUT, forms
a stable stem-loop with the first few editing sites positioned within the terminal loop of 5
base pairs. The hollow nucleotides represent the anchor binding site (ABS) where the
gRN A anchor binds to form the mRNA/gRNA complex. ES] = Editing site 1.

32

mRNA with measured Kp’s in the uM range. This suggests that for efficient gRNA
interaction a RNA chaperone is probably required and that the stable stem-loop formed
by the CYb mRNA may be a way to regulate its editing. In chapter 3 of this dissertation,
an in depth analysis using real time kinetics was used to study the thermodynamics of
three different mRNA/gRNA pairs. The CYb mRN A/gRN A binding interaction is
compared to two unedited mRNA/gRNA substrates that have predicted single stranded
anchor binding sites (ABS) using surface plasmon resonance (SPR). The A6 and NADH
dehydrogenase 7 (N D7) substrates are constitutively edited in contrast to the CYb
substrate. Interestingly, the A6 and ND7 mRNAs had 2000 fold and 8400 fold higher
affinity binding to their gRNAs respectively than the CYb mRNA/gRNA pair. The SPR
data provides additional rate constant data. The A6 pair has a ~20 fold faster association
rate and a ~20 fold slower dissociation rate than CYb. The ND7 pair has a ~90 fold
faster association rate and a ~20 fold slower dissociation rate than CYb. This provides
further evidence and information concerning how the mRN A structure around the
immediate editing domain can strongly affect the gRN A target binding.

In chapter 4, the interaction between two partially edited CYb mRN As and the
initiating gRNA, gCYb-558, is investigated through the use of gel shift assays and
surface plasmon resonance to find the dissociation constants as well as the association
and dissociation rate constants. Both methods confirm that the addition of two uridines
in the first editing site results in a significant decrease in the dissociation constant (Kn),
while the next two editing events (+4 U’s) results in an additional decrease. The binding
affinity for the CYb gRNA/mRN A improves with each editing event; this suggests that

with each editing event, the mRN A becomes more likely to finish editing. Surprisingly,

33

the presence of the U-tail appears to increase the association rate of the gRNA binding
the unedited CYb mRN A. Crosslinking data shows that the U-tail binds one side of the
helix that hides the anchor binding site. This suggests that the U-tail is helping disrupt
the CYb mRN A stem-loop through base pairing with the upstream purine rich region. In
chapter 5, I discuss the results of my research and the future directions for research on

target binding between the CYb gRNA and mRN A and editosome recruitment.

34

Literature Cited

Adler BK, Hajduk SL. 1997. Guide RNA requirement for editing-site-specific
endonucleolytic cleavage of preedited mRN A by mitochondrial ribonucleoprotein
particles in Trypanosoma brucei. Mol Cell Biol 17:5377-5385.

Altuvia S, Wagner EG. 2000. Switching on and off with RNA. Proc Natl Acad Sci U S A
97:9824-9826.

Aphasizhev R, Aphasizheva 1, Nelson RE, Gao G, Simpson AM, Kang X, Falick AM,
Sbicego S, Simpson L. 2003a. Isolation of a U-insertion/deletion editing complex
flom Leishmania tarentolae mitochondria. Embo J 22:913-924.

Aphasizhev R, Aphasizheva 1, Nelson RE, Simpson L. 2003b. A 100-kD complex of two
RN A-binding proteins flom mitochondria of Leishmania tarentolae catalyzes

RNA annealing and interacts with several RNA editing components. RNA 9:62-
76.

Aphasizhev R, Aphasizheva I, Simpson L. 2003c. A tale of two TUTases. Proc Natl
Acad Sci U S A.

Aphasizhev R, Sbicego S, Peris M, Jang SH, Aphasizheva 1, Simpson AM, Rivlin A,
Simpson L. 2002. Trypanosome mitochondrial 3' terminal uridylyl transferase
(TUTase): the key enzyme in U-insertion/deletion RNA editing. Cell 108:637-
648.

Benne R, Van den Burg J, Brakenhoff JP, Sloof P, Van Boom JH, Tromp MC. 1986.
Major transcript of the flameshifted coxII gene flom trypanosome mitochondria
contains four nucleotides that are not encoded in the DNA. Cell 46:819-826.

Bhat GJ, Koslowsky DJ, Feagin JE, Smiley BL, Stuart K 1990. An extensively edited
mitochondrial transcript in kinetoplastids encodes a protein homologous to
ATPase subunit 6. Cell 61:885-894.

Blom D, Burg J, Breek CK, Speijer D, Muijsers A0, Benne R. 2001. Cloning and
characterization of two guide RNA-binding proteins flom mitochondria of
Crithidia fasciculata: gBP27, a novel protein, and gBP29, the orthologue of
Trypanosoma brucei gBP21. Nucleic Acids Res 29:2950-2962.

Blum B, Bakalara N, Simpson L. 1990. A model for RNA editing in kinetoplastid
mitochondria: "guide" RNA molecules transcribed flom maxicircle DNA provide
the edited information. Cell 60:189-198.

Blum B, Simpson L. 1990. Guide RNAs in kinetoplastid mitochondria have a

nonencoded 3' oligo(U) tail involved in recognition of the preedited region. Cell
62:391-397.

35

Brecht M, Niemann M, Schluter E, Muller UF, Stuart K, Goringer HU. 2005. TbMP42, a
protein component of the RNA editing complex in Aflican trypanosomes, has
endo-exoribonuclease activity. Mol Cell I 7:621-630.

Brennecke J, Stark A, Russell RB, Cohen SM. 2005. Principles of microRNA-target
recognition. PLoS Biol 3ze85.

Brown HW, Neva FA. 1983. Ch. 4: Blood and Tissue Protozoa of Man. Basic Clinical
Parasitology. Norwalk, CT: Appleton Century Croft. pp 45-53.

Brown RC, Evans DA, Vickerman K. 1973. Changes in oxidative metabolism and
ultrastructurc accompanying differentiation of the mitochondrion in Trypanosoma
brucei. Int J Parasitol 32691-704.

Carnes J, Trotter JR, Ernst NL, Steinberg A, Stuart K. 2005. An essential RNase III
insertion editing endonuclease in Trypanosoma brucei. Proc Natl Acad Sci U S A
102:16614-16619.

Clement SL, Mingler MK, Koslowsky DJ. 2004. An intragenic guide RNA location
suggests a complex mechanism for mitochondrial gene expression in
Trypanosoma brucei. Eukaryot Cell 32862-869.

Corell RA, Read LK, Riley GR, Nellissery JK, Allen TE, Kable ML, Wachal MD,
Seiwert SD, Myler PJ, Stuart KB. 1996. Complexes flom Trypanosoma brucei
that exhibit deletion editing and other editing-associated properties. Mol Cell Biol
16:1410-1418.

Coustou V, Besteiro S, Riviere L, Biran M, Biteau N, Franconi JM, Boshart M, Baltz T,
Bringaud F. 2005. A mitochondrial NADH-dependent fiimarate reductase

involved in the production of succinate excreted by procyclic Trypanosoma
brucei. J Biol Chem 280: 16559-16570.

Cruz-Reyes J, Rusche LN, Sollner-Webb B. 1998. Trypanosoma brucei U insertion and
U deletion activities co-purify with an enzymatic editing complex but are
differentially optimized. Nucleic Acids Res 26:3634-3639.

Cruz-Reyes J, Zhelonkina AG, Huang CE, Sollner-Webb B. 2002. Distinct functions of
two RNA ligases in active Trypanosoma brucei RNA editing complexes. Mol Cell
Biol 22:4652-4660.

Eguchi Y, Tomizawa J. 1990. Complex formed by complementary RNA stem-loops and
its stabilization by a protein: function of CoIEl Rom protein. Cell 60:199-209.

Eguchi Y, Tomizawa J. 1991. Complexes formed by complementary RNA stem-loops.

Their formations, structures and interaction with ColEl Rom protein. J Mol Biol
220:83 1-842.

36

Ernst NL, Panicucci B, lgo RP, Jr., Panigrahi AK, Salavati R, Stuart K 2003. TbMP57 is
a 3' terminal uridylyl transferase (TUTase) of the Trypanosoma brucei editosome.
Mol Cell 11:1525-1536.

Feagin JE, Abraham JM, Stuart K. 1988a. Extensive editing of the cytochrome c oxidase
III transcript in Trypanosoma brucei. Cell 53:413-422.

Feagin JE, Jasmer DP, Stuart K. 1987. Developmentally regulated addition of nucleotides
within apocytochrome b transcripts in Trypanosoma brucei. Cell 49:337-345.

Feagin JE, Shaw JM, Simpson L, Stuart K. 1988b. Creation of AUG initiation codons by
addition of uridines within cytochrome b transcripts of kinetoplastids. Proc Natl
Acad Sci U S A 85:539-543.

Feagin JE, Stuart K. 1988. Developmental aspects of uridine addition within
mitochondrial transcripts of Trypanosoma brucei. Mol Cell Biol 8: 1259- 1265.

Franch T, Petersen M, Wagner EG, Jacobsen JP, Gerdes K. 1999. Antisense RNA
regulation in prokaryotes: rapid RN A/RN A interaction facilitated by a general U-
m loop structure. J Mol Biol 294:1115-1125.

Gao G, Simpson L. 2003. Is the Trypanosoma brucei REL] RNA ligase specific for U-
deletion RNA editing, and is the REL2 RNA ligase specific for U-insertion
editing? J Biol Chem 278:27570-27574.

Geissmann TA, Touati D. 2004. qu, a new chaperoning role: binding to messenger RNA
determines access for small RNA regulator. Embo J 23 2396-405.

Golden DE, Hajduk SL. 2005. The 3'-untranslated region of cytochrome oxidase II
mRNA functions in RNA editing of Aflican trypanosomes exclusively as a cis
guide RNA. RNA 11:29-37.

Goringer HU, Koslowsky DJ, Morales TH, Stuart K. 1994. The formation of
mitochondrial ribonucleoprotein complexes involving guide RNA molecules in
Trypanosoma brucei. Proc Natl Acad Sci U S A 91:1776-1780.

Hajduk SL, Sabatini RS. 1998. Mitochondrial mRNA Editing in Kinetoplastid Protozoa.
In: Grosjean H, Benne R, eds. Modification and Editing of RNA. Washington,
DC: ASM Press. pp 377-411.

Hannaert V, Bringaud F, Opperdoes FR, Michels PA. 2003. Evolution of energy
metabolism and its compartmentation in Kinetoplastida. Kinetoplastid Biol Dis
2:11.

Hayman ML, Miller MM, Chandler DM, Goulah CC, Read LK 2001. The trypanosome
homolog of human p32 interacts with RBP16 and stimulates its gRNA binding
activity. Nucleic Acids Res 29:5216-5225.

37

Hayman ML, Read LK. 1999. Trypanosoma brucei RBP16 is a mitochondrial Y-box
family protein with guide RNA binding activity. J Biol Chem 274: 12067-12074.

H jalt TA, Wagner EG. 1995. Bulged-out nucleotides in an antisense RNA are required
for rapid target RNA binding in vitro and inhibition in vivo. Nucleic Acids Res
23:580-587.

Huang CE, Cruz-Reyes J, Zhelonkina AG, O'Heam S, Wirtz E, Sollner-Webb B. 2001.
Roles for ligases in the RNA editing complex of Trypanosoma brucei: band IV is
needed for U-deletion and RNA repair. Embo J 20:4694-4703.

Hutvagner G, Zamore PD. 2002. A microRNA in a multiple-nunover RNAi enzyme
complex. Science 297 :2056-2060.

Kang X, Falick AM, Nelson RE, Gao G, Rogers K, Aphasizhev R, Simpson L. 2004.
Disruption of the zinc finger motifs in the Leishmania tarentolae LC-4

(=TbMP63) L-complex editing protein affects the stability of the L-complex. J
Biol Chem 279:3893-3899.

Kang X, Gao G, Rogers K, Falick AM, Zhou S, Simpson L. 2006. Reconstitution of full-
round uridine-deletion RNA editing with three recombinant proteins. Proc Natl
Acad Sci U S A.

Kolb FA, Westhof E, Ehresmann C, Ehresmann B, Wagner EG, Romby P. 2001. Bulged
residues promote the progression of a loop-loop interaction to a stable and
inhibitory antisense-target RNA complex. Nucleic Acids Res 29:3145-3153.

Koller J, Muller UF, Schmid B, Missel A, Kruft V, Stuart K, Goringer HU. 1997.
Trypanosoma brucei gBP2 1. An arginine-rich mitochondrial protein that binds to
guide RNA with high affmity. J Biol Chem 272:3749-3757.

Koslowsky DJ. 2004. A historical perspective on RNA editing: how the peculiar and
bizarre became mainstream. Methods Mol Biol 265: 161-197.

Koslowsky DJ, Riley GR, F eagin JE, Stuart K 1992. Guide RNAs for transcripts with
developmentally regulated RNA editing are present in both life cycle stages of
Trypanosoma brucei. Mol Cell Biol 12:2043-2049.

Kuzoe FAS, Schofield CJ. 2004. Strategic review of traps and targets for tsetse and
African trypanosomiasis control. Special programme for research and training in
tropical diseases (T DR). Geneva: UNICEF/UNDP/World Bank/WHO,
http://wwwwho.int/tdr/publications/publicationsitsetse_traps.htm.

Lai EC. 2003. microRNAs: runts of the genome assert themselves. Curr Biol 13:R925-
936.

38

Lamour N, Riviere L, Coustou V, Coombs GH, Barrett MP, Bringaud F. 2005. Proline
metabolism in procyclic Trypanosoma brucei is down-regulated in the presence of
glucose. J Biol Chem 280211902-11910.

Lee Y, Ahn C, Han J, Choi H, Kim J, Yim J, Lee J, Provost P, Radmark 0, Kim S, Kim
VN. 2003. The nuclear RNase III Drosha initiates microRNA processing. Nature
425:415-419.

Leung SS, Koslowsky DJ. 1999. Mapping contacts between gRNA and mRN A in
trypanosome RNA editing. Nucleic Acids Res 27:778-787.

Leung SS, Koslowsky DJ. 2001a. Interactions of mRN As and gRNAs involved in
trypanosome mitochondrial RNA editing: structure probing of an mRN A bound to
its cognate gRNA. RNA 7:1803-1816.

Leung SS, Koslowsky DJ. 2001b. RNA editing in Trypanosoma brucei: characterization
of gRN A U-tail interactions with partially edited mRN A substrates. Nucleic Acids
Res 29:703-709.

Lewis BP, Shih II-I, Jones-Rhoades MW, Bartel DP, Burge CB. 2003. Prediction of
mammalian microRN A targets. Cell [15:787-798.

Lukes J, Guilbride DL, Votypka J, Zikova A, Benne R, Englund PT. 2002. Kinetoplast
DNA network: evolution of an improbable structure. Eukaryot Cell 1:495-502.

Madison-Antenucci S, Sabatini RS, Pollard VW, Hajduk SL. 1998. Kinetoplastid RNA-
editing-associated protein 1 (REAP-1): a novel editing complex protein with
repetitive domains. Embo J I 7:6368-6376.

Matranga C, Tomari Y, Shin C, Bartel DP, Zamore PD. 2005. Passenger-strand cleavage

facilitates assembly of siRN A into AgoZ-containing RNAi enzyme complexes.
Cell [23:607—620.

Matthews KR. 2005. The developmental cell biology of Trypanosoma brucei. J Cell Sci
[18:283-290.

Mattick J S. 2004. RNA regulation: a new genetics? Nat Rev Genet 5 :3 16-323.

Mattick JS, Makunin IV. 2005. Small regulatory RNAs in mammals. Hum Mol Genet 14
Spec No I:R12l-l32.

Mattock N, Pink R. 2003-2004. Making health research work for poor people. WHO
Tropical Disease Research 1 7th Annual Progress Report.

Geneva: UNICEF/UNDP/World Bank/WHO,
_h_ttp://www.who. int/tdr/publications/publications/pr1 7.htm.

McManus MT, Shirnamura M, Grams J, Hajduk SL. 2001. Identification of candidate
mitochondrial RNA editing ligases from Trypanosoma brucei. RNA 7 :167-1 75.

39

Michels PA, Hannaert V, Bringaud F. 2000. Metabolic aspects of glycosomes in
trypanosomatidae - new data and views. Parasitol Today 16:482-489.

Miller MM, Halbig K, Cruz-Reyes J, Read LK 2006. RBP16 stimulates trypanosome
RNA editing in vitro at an early step in the editing reaction. RNA 12: 1292-1303.

Missel A, Souza AE, Norskau G, Goringer HU. 1997. Disruption of a gene encoding a
novel mitochondrial DEAD-box protein in Trypanosoma brucei affects edited
mRNAs. Mol Cell Biol 17:4895-4903.

Moller T, Franch T, Hojrup P, Keene DR, Bachinger HP, Brennan RG, Valentin-Hansen
P. 2002. qu: a bacterial Sm-like protein that mediates RN A-RNA interaction.
Mol Cell 9:23-30.

Muller UF, Lambert L, Goringer HU. 2001. Annealing of RNA editing substrates
facilitated by guide RNA-binding protein gBP21. Embo J 20:1394—1404.

Nebohacova M, Maslov DA, Falick AM, Simpson L. 2004. The effect of RNA
interference Down-regulation of RNA editing 3'-terminal uridylyl transferase
(TUTase) 1 on mitochondrial de novo protein synthesis and stability of
respiratory complexes in Trypanosoma brucei. J Biol Chem 279:7819-7825.

Nolan DP, Voorheis HP. 1992. The mitochondrion in bloodstream forms of Trypanosoma
brucei is energized by the electrogenic pumping of protons catalysed by the F1 F0-
ATPase. Eur J Biochem 209:207-216.

Nordgren S, Slagter-Jager JG, Wagner GH. 2001. Real time kinetic studies of the

interaction between folded antisense and target RN As using surface plasmon
resonance. J Mol Biol 310:1125-1 134.

Nordstrom K, Wagner EG. 1994. Kinetic aspects of control of plasmid replication by
antisense RNA. Trends Biochem Sci 1 91294-300.

O'Hearn SF, Huang CE, Hemann M, Zhelonkina A, Sollner-Webb B. 2003. Trypanosoma
brucei RNA editing complex: band 11 is structurally critical and maintains band V
ligase, which is nonessential. Mol Cell Biol 23:7909-7919.

Panigrahi AK, Allen TE, Stuart K, Haynes PA, Gygi SP. 2003a. Mass spectrometric
analysis of the editosome and other multiprotein complexes in Trypanosoma
brucei. J Am Soc Mass Spectrom 14:728-735.

Panigrahi AK, Ernst NL, Domingo GJ, Fleck M, Salavati R, Stuart KD. 2006.
Compositionally and functionally distinct editosomes in Trypanosoma brucei.
RNA 12:1038-1049.

Panigrahi AK, Gygi SP, Ernst NL, lgo RP, Jr., Palazzo SS, Schnaufer A, Weston DS,
Carmean N, Salavati R, Aebersold R, Stuart KD. 2001a. Association of two novel

40

proteins, TbMP52 and TbMP48, with the Trypanosoma brucei RNA editing
complex. Mol Cell Biol 21:380-389.

Panigrahi AK, Schnaufer A, Carmean N, Igo RP, Jr., Gygi SP, Ernst NL, Palazzo SS,
Weston DS, Aebersold R, Salavati R, Stuart KD. 2001b. Four related proteins of
the Trypanosoma brucei RNA editing complex. Mol Cell Biol 21:6833-6840.

Panigrahi AK, Schnaufer A, Ernst NL, Wang B, Carmean N, Salavati R, Stuart K. 2003b.
Identification of novel components of Trypanosoma brucei editosomes. RNA
9:484—492.

Pelletier M, Read LK. 2003. RBP16 is a multifunctional gene regulatory protein involved
in editing and stabilization of specific mitochondrial mRN As in Trypanosoma
brucei. RNA 9:457-468.

Peris M, Frech GC, Simpson AM, Bringaud F, Byrne E, Bakker A, Simpson L. 1994.
Characterization of two classes of ribonucleoprotein complexes possibly involved
in RNA editing from Leishmania tarentolae mitochondria. Embo J 13:1664-1672.

Peris M, Simpson AM, Grunstein J, Liliental JE, Frech GC, Simpson L. 1997. Native gel
analysis of ribonucleoprotein complexes from a Leishmania tarentolae
mitochondrial extract. Mol Biochem Parasitol 85 :9-24.

Pollard VW, Harris ME, Hajduk SL. 1992. Native mRNA editing complexes fi‘om
Trypanosoma brucei mitochondria. Embo J 11:4429-4438.

Priest JW, Hajduk SL. 1994. Developmental regulation of Trypanosoma brucei
cytochrome c reductase during bloodstream to procyclic differentiation. Mol
Biochem Parasitol 65:29 1-304.

Riley GR, Corell RA, Stuart K. 1994. Multiple guide RNAs for identical editing of
Trypanosoma brucei apocytochrome b mRN A have an unusual minicircle
location and are developmentally regulated. J Biol Chem 269:6101-6108.

Rusche LN, Cruz-Reyes J, Piller KJ, Sollner—Webb B. 1997. Purification of a functional
enzymatic editing complex from Trypanosoma brucei mitochondria. Embo J
[624069-4081.

Salavati R, Ernst NL, O'Rear J, Gilliam T, Tarun S, Jr., Stuart K. 2006. KREPA4, an
RNA binding protein essential for editosome integrity and survival of
Trypanosoma brucei. RNA 12:819-831.

Schnaufer A, Clark-Walker GD, Steinberg AG, Stuart K. 2005. The Fl-ATP synthase
complex in bloodstream stage trypanosomes has an unusual and essential
function. Embo J 24 :4029-4040.

41

Schnaufer A, Ernst NL, Palazzo SS, O'Rear J, Salavati R, Stuart K. 2003. Separate
insertion and deletion subcomplexes of the Trypanosoma brucei RNA editing
complex. Mol Cell 12:307-319.

Schnaufer A, Panigrahi AK, Panicucci B, Igo RP, Jr., Wirtz E, Salavati R, Stuart K.
2001. An RNA ligase essential for RNA editing and survival of the bloodstream
form of Trypanosoma brucei. Science 291 :2159-2162.

Seiwert SD, Stuart K. 1994. RNA editing: transfer of genetic information from gRN A to
precursor mRN A in vitro. Science 266:114-117.

Shaw JM, Feagin JE, Stuart K, Simpson L. 1988. Editing of kinetoplastid mitochondrial
mRNAs by uridine addition and deletion generates conserved amino acid
sequences and AUG initiation codons. Cell 53:401-411.

Simpson L, Aphasizhev R, Gao G, Kang X. 2004. Mitochondrial proteins and complexes
in Leishmania and Trypanosoma involved in U-insertion/deletion RNA editing.
RNA 10:159-170.

Slagter-Jager JG, Wagner EG. 2003. Loop swapping in an antisense RN A/target RNA
pair changes directionality of helix progression. J Biol Chem 278:35558-35563.

Stuart K, Panigrahi AK, Schnaufer A, Drozdz M, Clayton C, Salavati R. 2002.
Composition of the editing complex of Trypanosoma brucei. Philos Trans R Soc
Lond B Biol Sci 35 7:71-79.

Stuart KD, Schnaufer A, Ernst NL, Panigrahi AK. 2005. Complex management: RNA
editing in trypanosomes. Trends Biochem Sci 30:97-105.

Sturrn NR, Simpson L. 1990. Kinetoplast DNA minicircles encode guide RNAs for
editing of cytochrome oxidase subunit III mRNA. Cell 61:879-884.

Tang G, Reinhart BJ, Bartel DP, Zamore PD. 2003. A biochemical framework for RNA
silencing in plants. Genes Dev 1 7:49-63.

Tomari Y, Matranga C, Haley B, Martinez N, Zamore PD. 2004. A protein sensor for
siRN A asymmetry. Science 306: 1377-1380.

Tomizawa J. 1984. Control of ColEl plasmid replication: the process of binding of RNA
I to the primer transcript. Cell 38:861-870.

Tomizawa J, Itoh T, Selzer G, Som T. 1981. Inhibition of ColEl RNA primer formation
by a plasmid-specified small RNA. Proc Natl Acad Sci U S A 78: 1421-1425.

Trotter JR, Ernst NL, Cames J, Panicucci B, Stuart K. 2005. A deletion site editing
endonuclease in Trypanosoma brucei. Mol Cell 20:403-412.

42

Unisci. 2001. Artificial cow helps to eradicate sleeping sickness. In: Radler D, ed. Daily
University Science News: UniScience News Net, Inc.

Vanhamme L, Perez-Morga D, Marchal C, Speijer D, Lambert L, Geuskens M,
Alexandre S, Ismaili N, Goringer U, Benne R, Pays E. 1998. Trypanosoma brucei
TBRGGl, a mitochondrial oligo(U)-binding protein that co-localizes with an in
vitro RNA editing activity. J Biol Chem 273221825-21833.

Vickerman K. 1965. Polymorphism and mitochondrial activity in sleeping sickness
trypanosomes. Nature 208:762-766.

Vickerman K. 1985. Developmental cycles and biology of pathogenic trypanosomes. Br
Med Bull 41: 105-1 14.

Vondruskova E, van den Burg J, Zikova A, Ernst NL, Stuart K, Benne R, Lukes J. 2005.
RNA interference analyses suggest a transcript-specific regulatory role for
mitochondrial RNA-binding proteins MRP1 and MRP2 in RNA editing and other
RNA processing in Trypanosoma brucei. J Biol Chem 280:2429-2438.

Wang B, Ernst NL, Palazzo SS, Panigrahi AK, Salavati R, Stuart K. 2003. TbMP44 Is
Essential for RNA Editing and Structural Integrity of the Editosome in
Trypanosoma brucei. Eukaryot Cell 2:578-587.

WHO. 2001 . Afiican trypanosomiasis or sleeping sickness. Fact Sheet 259: World Health
Organization.

Zhang A, Wassarman KM, Ortega J, Steven AC, Storz G. 2002. The Sm-like qu protein
increases OxyS RNA interaction with target mRN As. Mol Cell 9:11-22.

43

CHAPTER 2
INTERACTIONS OF MRNAS AND GRNAS INVOLVED IN TRYPANOSOME
MITOCHONDRIAL RNA EDITING: STRUCTURE PROBING OF A GRNA
BOUND TO ITS COGNATE MRNA

(Portions of this chapter were previously published in RNA (2006) 1221-1 1.)

Introduction

In kinetoplastid protozoa, guide RNAs (gRNAs) serve as templates for the precise
insertion and deletion of uridylates during RNA editing of the mitochondrial mRN A
transcripts. This post-transcriptional process produces translatable mRNAs by creating
the open reading fi'ames as well as initiation and termination signals. In addition, there is
a transcript-specific mechanism that regulates editing during the life cycle of the parasite.
The gRN As are key components of the reaction as they provide the information for the
nucleotide (nt) alterations and direct the cleavage and ligation events (Simpson et al.,
2004; Stuart et al., 2005). Guide RNAs have an average length of 50-70 nts and consist
of three functional elements. Contained within the 5’ end of gRNAs is a short sequence
known as the gRNA anchor, a 5-21 nucleotide region that base pairs with a particular
mRNA just 3’ of the editing domain. The second element, the guiding region, serves as a
template for the editing process and is complementary (allowing G-U base pairs) to the
mature mRN A. Finally, at the 3’ end of the gRNA is a poly-uridylate tail (U -tail) that is
added post-transcriptionally (Blum et al., 1990; Blum & Simpson, 1990).

The latest model for RNA editing in trypanosomes involves >20 proteins that form a
complex structure called the editosome around an mRNA/gRNA pair (Simpson et al.,
2004; Stuart et a1, 2005). The insertion and deletion of uridines occur through
successive rounds of enzymatic reactions coordinated by the editosome and templated by
the gRNA (Blum et al., 1990; Blum & Simpson, 1990; Seiwert & Stuart, 1994; Adler &
Hajduk, 1997). Significant progress has been made in identifying proteins involved in
the editing reaction, including both proteins in the complex as well as proteins that appear

to be accessory factors that transiently act during the editing process (Koller et al., 1997;

45

Missel et al., 1997; Madison-Antenucci et al., 1998; Vanhamme et al., 1998; Hayman &
Read, 1999; Aphasizhev & Simpson, 2001; McManus et al., 2001; Panigrahi et al.,
2001a; Panigrahi ct al., 2001b; Rusche et al., 2001; Schnaufer et a1, 2001; Aphasizhev et
al., 2003a; Aphasizhev et al., 2003b; Panigrahi et al., 2003a; Panigrahi et al., 2003b;
Pelletier & Read, 2003). However, despite this progress, very little is known about how
the editing complex is assembled onto specific RNAs. No conserved sequence domains
have been found in mRN As, and the only conserved sequence domain shared between the
gRNAs is the U-tail. Previous work in this lab suggests that different mRNA/gRNA
pairs can form similar structures (Leung & Koslowsky, 1999), and we believe that
structure recognition may be important for efficient editosome assembly with the
mRNA/gRNA pair.

In previous studies, we used a RNA folding program to look for conserved structures
between three different mRNA/gRNA pairs. The predicted secondary structures for these
mRNA/gRNA‘pairings incorporated a distance constraint based on photoaffinity
crosslinks (Leung & Koslowsky, 1999). Further structure analyses for the
apocytochrome b mRNA/gRNA complex were undertaken to look at mRNA/gRNA
structure as editing progresses up through the third editing site. The predicted secondary
structures for these mRNA/gRNA pairings are based on photoaffinity crosslinking studies
of the 5’ and 3’ ends of NgCYb-558 that were mapped along 5’CYbUT (first 88 nts. of
CYb mRNA), 5’CYbPESlT (edited through site 1), and 5’ CYbPES3T (edited through
site 3) (Leung & Koslowsky, 2001b). In addition, a solution structure probing study of
5’CYbUT while paired with NgCYb-558 or NgCYb-558sU (no U-tail) was done (Leung

& Koslowsky, 2001a). With the data from these studies, we proposed a secondary

46

structure model that includes the following features: a gRNA/mRN A anchor helix, a U-
tail/mRNA helix, and a gRNA stem-loop (Fig. 11A) (Leung & Koslowsky, 1999, 2001a,
b). The predicted structure of the gRNA interaction with the partially edited mRN A was
particularly interesting because it suggests that the U-tail could fimction to maintain the
gRNA stem-loop during the first few editing events (Fig. 11B). The focus of this paper is
on the solution structure probing of the secondary structure of the gRN A, NgCYb-558,
paired with unedited and partially edited mRN As, as we would like to understand how
the gRNA structure may be changing during the early steps of the editing process. In
addition, we have also probed the structure of the partially edited mRN A, CYbPES3, to
corroborate our gRNA structure data. Our data indicate that the stem-loop formed by the
guiding region of the gRN A alone is maintained in its interaction with the pre-edited
message as predicted in the computer models. In addition, our data support the predicted
structure for the gRNA interaction with its partially edited mRN A in that the gRNA stem-
loop structure appears to be maintained through the first few editing events by the use of
alternative base pairing with the U-tail.
Results and Discussion
Crosslinked RNAs used for Solution Structure Probing

Editing of the apocytochrome b mRN A is developmentally regulated with the
insertion of 34 uridylates at 13 sites near its 5’ end. The editing process that creates the
CYb initiation codon occurs preferentially during the procyclic (insect) and stumpy
bloodstream stages of the trypanosome life cycle (Feagin et aL, 1987; Feagin & Stuart,
1988). In these experiments, we used modified versions of 5’CYbUT (Koslowsky et al.,

1992; Koslowsky et a1., 1996). CYbU (U = unedited) is identical in sequence to

47

. U A U
UtallI-lellx 4:: Glue
6‘ \u
G ‘U
c ‘9
44 ‘U
A g gRNAstem-loop
”4606
B e .3‘

4 U 4 gRNA stem-loop

Figure 10. The predicted models for mRN A/gRN A complex secondary structure have three
helices: a mRN A/gRN A anchor helix, a gRNA stem-loop within the guiding region, and an U-
tail/mRNA duplex. A. CYbU-NgCYb-558 pair with the first editing site just 5’ (mRNA) of the
anchor indicated by ES]. B. CYbPES3—NgCYb-558 pair with the fourth editing site just 5’
(mRN A) of the anchor indicated by E84.

48

5’CYbUT except for the deletion of 12 nts at the 5’most end (Fig. 11A) and is composed
of 88 nts from the 5’ end of the apocytochrome b mRNA (beginning at +13) including all
13 editing sites as well as a 24 nt vector tag at the 3’ end. The partially edited substrate,
CYbPES3 (PES = Partially Edited through Site 3), is identical to CYbU except for the 6
uridines added so that the first 3 editing sites are fiilly edited (Fig. 11B). Our gRNA
construct NgCYb-558 (Fig. 11C) is very similar to wildtype gCYb-558 and two other
redundant gRNAs, gCYb560A, B. All three gRNAs can act as the editing initiating
gRNA that direct the editing of the first seven editing sites (Riley et a1., 1994). NgCYb-
558 is 59 nucleotides long including a 15 nt uridylate-tail (U -tail) that is added via
ligation with T4 DNA ligase and a bridging deoxyoligonucleotide (Moore & Sharp,
1992; Leung & Koslowsky, 2001a).

The secondary structure of gCYb-558 was reported by Schmid et a1. (1995) and
described as two thermodynamically unstable hairpin loops separated by a single-
stranded region. The smaller hairpin at the 5’ end contains the anchor binding sequence
that selects the mRNA for editing. The guiding region forms a longer stem-loop and the
U-tail appears to be single-stranded (Schmid et al., 1995). The sequence of the gRNA
used in this study differs slightly fi'om the gCYb-558 used in the Schmid et al. study.
One significant difference is that a cytosine residue replaces a uridine at position 25, a
nucleotide change also seen in wild type gCYb-560A, B (Fig. 11C). This U to C change
creates a strong G-C base pair within the guiding region that appears to stabilize a single
gRNA stem-loop that differs slightly from the Schmid et al. (1995) report.

In order to determine the structure of the gRN A in its interaction with its cognate

mRN A, we used solution structure probing techniques and 3’ end labeled gRN As that

49

. .2793 on. =a E EVE—boa: 98 Ben E .wmm-n>UmZ 2 36.808 .20 dam u - .3526an

$2038 u _. .wmm-n>0w 09$ 23, 2 339:8 megabyte E V was 98 98-3 =86§>2§ E m. a memes—05 wac— moflaoo—o—E an

e ammész sagas Seaman“ Ea amoerom .Ezmw £6 messes 25 2:, 55 BE? 8538 $250? 2: .9 533.20
3268:: 05 .«o sermon 05 8.865 382 .358 83 xetQEmaB no.“ L: 53 83 DO u .8533 3:855? n a 53, 5527.

m_ 52: 5.96 .855 05 55,28 98 <ZMw 2: 56389 btficofioifioo 2:. 6860 biz 0.3 3% magma m “mac 05 35 8 Gotta
AF: 3.8830: 852.5 o 05 Sm 3083 250 3 30:52 $ .mflEfiU £738 @330 33:3 2E. .585 & £98532 9 3:33
mmmaero Ea 558.2 3 3% £985 33 33%“ u may $2.50 2: .m e8 .m 2: a we “98> E E a a =03 a 8% mazes

2 =m wig—2: Am: “a 9:5:on <ZME a oEeuteSAoer 05 mo EB .m on“ Bob 3: mm mo 389:8 m_ :86 .565 m_ wnmAcrUmZ
2 vanes :66 ea accuse. €268: u 2 2:6 =3 2: .< £8 nagana <zm§<zam Eco usage; 2a 853% .: as»;

.................... eeee.e.e...ae.ee..aheeec..eeeaeeeeecaae...ea.e.
DDDDDDDDDDDDDDDDfidddUfldGO0<30DUDU<DUd¢S¢UO<GUOUD<¢DUflGDWDDD<D¢DGD<¢U¢U<¢¢¢DG<WMU<UUemunflommunfluu
DDDDDDDDDDDDDDDunu4&54d000ddd5éUDUOfiDDDDD<¢ODUD¢¢U¢0¢¢¢dDD<04000.m uuuuu ommflhum

DDDDDDDDDDDDDDD¢Dbdd5<¢000€<<Dd0D0045MDDDddUDUD<¢U<O¢d¢¢DD 1:000.mnommnkbmnlbz

H3:2. mammfiwu 392
U

OOQ4DD‘d‘dO‘UddbUDU‘dfiDDUDfiUODO¢D<<¢OUGddDddDDDDDDDDDDDDDDDnammunwumz

_...____________:.._______ ...........................................

UGOdDdDfidUdDdDflDDDUDDOOdUDUDd‘DDDDUDOflfluflaflflaflddfiddddddwfiwoUOddddDdDfiDDDD‘ddDfiDDUU iiiiiiiiiii

£3-..wa Z 3 to: z: ammmmmn>v.m

oommommmmsozzm
usemoo:mUwa¢s¢s¢¢w¢s¢D<DDDGDD004UDwD¢¢DDDDUDw530505354¢¢0¢<¢4¢¢0<00Uw¢¢<¢D<D¢DDDD¢<<D<DDOU

”a tom—3.5 <72:— ooaoa om :mmmmma>0.m
m

OOO£D9££4464U¢¢DUD044DDDUD¢00DUdDdddUOOddbddbDDDDDDDDDDDDDDmmmunwumz

_£__________ .......................................................

wobdodbcdwdbdbdbbbODDOGdUbObddDDDbUDoww<¢¢Uddddddodwoowddddbdsdspbbdmdsdbbwo ................

.mmm-a>U Z 3 to: =« =mDn>U.m

oommommm
mzuzzmmsomoo:mOUD4D¢D<4U<D<DdbbbUDDOOflUDOD¢¢DDDDUD000<<46¢¢¢¢d40¢60UGddddDflDdDDDDflddDdDDOO

; 28:225.: 8:2. am .5:ng
<

 

 

 

 

 

50

had been crosslinked to the cognate mRNA at the 3’ end of the anchor duplex (mRN A
orientation). It has been previously shown that these structures can assemble with
enzymatically-active editing proteins and are biologically relevant (Leung & Koslowsky,
2001a).

Crosslinked substrates were generated by transcribing gRN As in the presence of
guanosine 5’-monophosphorothioate (GMPS) (Burgin & Pace, 1990; Harris & Christian,
1999). The gRNAs were 3’end labeled using T4 RNA ligase and [5’-32P] cytidine 3’, 5’-
bisphosphate (pCp) when probing the gRN A; the mRN As were 5’ end labeled using T4
Kinase and [y 32P] rATP, when probing the mRNA structure. After labeling, the thiol
group of GMPS at the 5’ end of the gRNA was coupled to azidophenacyl bromide (APA)
and UV crosslinked as previously described (Burgin & Pace, 1990; Leung & Koslowsky,
1999). The resultant crosslink is covalently linked to the mRN A at the 3’ end (mRN A
orientation) of the gRNA/mRNA anchor duplex (Fig. 11A, B). Therefore, these
crosslinked RNAs maintain the anchor duplex, an element required for the process of
editing. The higher order structure of NgCYb-558 crosslinked to its cognate mRN A was
probed using ribonuclease accessibility experiments (Fig. 12). Single-stranded, non-
stacked conformations were monitored using ribonucleases Tl (G specific), T2 (slight
preference for A), and Mung Bean Nuclease (MBN) (slight preference for A). RNase V1
was used to determine nucleotides involved in base paired helical interactions or in
stacked, structured elements (Lowman & Draper, 1986). For comparative purposes, an

analysis of the structure of the free NgCYb—558 was compared to the structure of the

51

Figure 12 A Mung Bean Nuclease

 

  

 

 

NgCYb—558
Alone +CYbU +CYbPES3

02 0 2 5 "In

' ' E

0

s

s
a:
g a

ma. 5

, _. 5.

2

> 5

3

0

, - -

U U ” U_

U U U"

u u U:

U U U

U U U

B RNase T2
NgCYb-558
Alone +CYbU +CYbPES3

01015T101015U2010

 

52

A
an
extended anchor g

anchor

Guldlng nylon

 

c c C
[Hall

Figure 12 continued C RNase T1

NgCYb-558

Alone +CYbU +CYbPES3
0 1o 15 T1 0 15 02 0 1o 15 "In

 

10

    

 

 

    

3

8

a

I
E:
gs

c

2

mum.' m 3

E

g

3

0

D RNase v1
_N_QCYb-558
Alone +CYbU +CYbPEs3

5T10V2‘5u2025MIn
é

0X80“

Guldlng roglon

li‘tt

 

 

cc:
Utall

53

Figure 12. Structure Probing of the gRNA, NgCYb-558. Each solution structure probing gel has
three sections showing the digestion pattern of NgCYb-558 alone, NgCYb-558 crosslinked to
CYbU (+CYbU), and NgCYb-558 crosslinked to CYbPES3 (+CYbPES3). Digestion times in
minutes are shown at the top. RNase T1 and U2 ladders with nt numbers are indicated. A line
diagram indicating the anchor (bold line), extended anchor (double line), guiding region (thin
line), and U-tail of the gRN A is shown next to each digestion pattern. A. Mung Bean Nuclease
gel, single-strand specific. B. RNase T2, single-strand. C. RNase T1, single-strand guanosine
specific. D. RNase V1, double-strand specific.

54

gRNA crosslinked to either unedited CYb (CYbU) or to a partially edited CYb
(CYbPES3). In addition, to complement the gRNA data, the structure of CYbPES3
crosslinked to both NgCYb-558 and NgCYb-558sU (N gCYb-558 lacking a U—tail) was I
also examined. The mRN A secondary structure was probed by both ribonuclease
accessibility and chemical reactivity (diethyl pyrocarbonate (A, N7)) (Fig. 13, 14). The
data were used to construct secondary structure models with the data summarized in
Figure 15. In the summary figures the intensity of the cut is represented by large or small
symbols for each enzyme or chemical. The high intensity cleavages (large symbols)
represent hot spots that are unprotected by secondary and tertiary interactions. Cleavages
of medium and low intensity were harder to classify and may include both sites where
tertiary structure incompletely blocks access and some secondary cuts that are not
reflective of the native structure (small symbols). In addition, the lower intensity
cleavages could be due to alternative structures formed by a small percentage of the RNA
population. However these cleavages appeared consistently in all gels.
Structure of the NgCYb-558 anchor region

Structure probing of unpaired NgCYb-558 by single-strand specific nucleases
indicates that the anchor region (Gl-C13) is mostly single-stranded (Fig. 12). Under our
conditions (100 mM KCl and lmM MgClz) distinct cleavage products were clearly
visible in the anchor region using Mung Bean Nuclease (MBN), RNase T2 (T2), and
RNase T1 (T1). As expected, when the gRNA was crosslinked to its cognate mRN A
(+CYbU and +CYbPES3), the anchor-duplex formed and was no longer sensitive to the
single-strand specific RNases. The RNase V1 (V l) sensitivity pattern generally

complemented these data. We did observe two V1 cleavages in the unpaired gRNA at

55

Figure 13 A Mung Bean Nuclease

 

 

 

 

 

 

 

 

 

  
   

 

anchor

CYbPES3
+NgCYb-558 +NgCYb-558
Alone no U-tall glue U-tall
0 2 s 2 5 U2 0 5
131’
«341%
H efltm
f: Qn1
5.... A10
4.... A10
A10
3
A14
B RNaserz
CYbPES3
+NgCYb-558 +NgCYb-558
Alone no U-tall plus U-tall
0101520 11 0101520 uz 01015 20am
' ’5’?
g m
:2; 625 """ "” "" R m ".
.... 622 “' "‘ “" ..I.
: A20
- A19
- A18
3.“
4. A10
is
. A14

 

 

M
ON

 

Figure 13 continued C RNase T1

 

 

 

   

 

 

CYbPES3
+NgCYb-558 +NgCYb-558
Alone no U-tall glue U-tall
0101520 0101520020101520
1. ass
- s: 643
..: $5.1 641
ml 634
at: m .m an»
.- 625 e -" ‘f’
an 624 ' ' w
4.. ~
D RNase V1
CYbPES3
+N9CYb-558 +NgCYb-558
Alone no U-tall plus U-tall

 

 

 

025101102510020251ogf'ln

anchor

 

 

 

 

1
at:

57

Figure 13. Structure Probing of the partially edited CYb mRN A, CYbPES3. Each solution
structure probing gel has three sections showing the digestion pattern of CYbPES3 alone,
CYbPES3 crosslinked to NgCYb-5583U (no U-tail), and CYbPES3 crosslinked to NgCYb-558.
Digestion times in minutes are shown at the top. RNase T1 and U2 ladders with nt numbers are
indicated A diagram of the mRN A is next to each digestion pattern showing the orientation 3’
to 5’ with the extended anchor (bold line). A. Mung Bean Nuclease gel, single-strand specific B.
RNase T2, single-strand C. RNase T1, single-strand guanosine specific. D. RNase V1, double-
strand specific.

58

DEPC
CYbPES3
+NgCYb-558 +NgCYb-558
Alone no U-tail plus U-tail

No No No
DEPC 1F 29 T1 DEPC 15 2E DEPC 1F 25 -OH:V

     

anchor

» A20
gm A19

 

A16

A14

 

 

 

Figure 14. Chemical structure probing of the partially edited CYb mRNA, CYbPES3. The
diethyl pyrocarbonate (DEPC) chemical structure probing gel has three sections showing the
difference in digestion pattern of CYbPES3 alone, CYbPES3 crosslinked to NgCYb—558sU (no
U-tail), and CYbPES3 crosslinked to NgCYb-558. Digestion times in minutes are shown at the
top. RNase T1 and U2 ladders with at numbers are indicated; an alkaline hydrolysis ladder (OH-)
was included as well. A diagram of the mRNA is next to each digestion pattern showing the
orientation 3’ to 5’ with the extended anchor (bold line).

59

Figure 15

A NgCYb-558
39
uUUUU
DU 0
U ooo 00.. ”O. Q
U III:.. ...‘IIIAAOIaaaafii
UgUUUU 0A} :gUGUMCAGW 6
"AU _ AAIO 5'66
oAA — UA
utt— U‘
AG -— "A
A6 -— CA
A
.“As UAIO
"A 6| «0
.- u G‘
o. I EU? ...
. I
3’
(9°09

3UUu Ulla/J I I ll I 31 0‘0“
50 -"IAA '20 AAalo
. I‘QU ‘A‘AGUG.
as“: u‘ e I.
‘6 _~ “A
. f3 - 3: Single Strand Specific
"AA ‘ ”If-I. ..Q MungBoanNuclaeao
0 IA” G I . II. RNaeoTZ
o' :G£I. +..RN880T1
.- 2 '. 1- Double Strand .9ch
. ,‘A RNaaeV1

came . .-

fi 8.2”. to...

up .85. II-

; ooazz 44< 300.032 zoom 95! .0.
2:88 226 .333 058..» .825 case

'
tun
funny“? mom/st 0:
Q
...zfigifiweaéfia . ”new no.
. \ u .
.8 5.5 30303034.? 5.. \u
* “V ..1.“
‘OVVG\“
«mm.»
.‘
n e
OWO
'.
mom...

Figure 15 confirmed

3359223230 0

61

Figure 15. Summary figures from solution structure probing gels that show predicted secondary
structures with cleavages from all 4 nucleases and DEPC chemical. Three sizes of symbols
indicate the intensity of cleavage. Circles indicate mung bean nuclease (single-strand specific)
cleavages, squares indicate RNase T2 (single-strand specific) cleavages, crosses indicate RNase
T1 (single-strand guanosine specific) cleavages, cylinders indicate DEPC (single-strand
adenosine specific), and triangles indicate RNase V1 (double-strand specific) cleavages. A.
NgCYb-558 alone B. CYbU/NgCYb-558 complex C. CYbPES3/NgCYb-558 complex. The red
U’s represent the uridylates added during editing.

62

positions A12 and C13 (Fig. 12D), located near the end of the anchor region that were
also sensitive to the single-strand specific nucleases.

While V1 sensitive cleavages were clearly visible within the anchor duplex of the
paired gRN A (+CYbU, Fig. 12) and within the extended anchor duplex (+CYbPES3, Fig.
12), it was interesting that the cleavages were concentrated in the 5’ half (mRN A
orientation) of the anchor duplex. Lack of cleavage in the Gl-A7 region may be due to
the presence of the crosslink. For NgCYb-558 paired with CYbU, V1 cleavages within
the anchor were surprisingly weak and limited to A8-C13 (Fig. 12D). This region was
also very poorly cleaved in the NgCYb-558/CYbPES3 pairing, where the anchor is
extended to U26. In this extended anchor, strong V1 cleavages were observed in the 3’
most end of the duplex (A21-U26, Fig. 12D). The limitation of strong RNase V1
cleavage to the A21-U26 region of the extended anchor duplex was unexpected and
suggests that most of the anchor duplex may be protected by tertiary interactions.
Structure of NgCYb-558 guiding region

Secondary structure analyses of several unpaired gRN As indicate that the guiding
regions (nucleotides between the 5’ anchor and 3’ U-tail) form a gRNA stem-loop
element (Schmid et al., 1995). In our digests of unpaired NgCYb-558, we also observe
the presence of a gRNA stem-loop (Fig. 15). Weak but consistent single-strand specific
cleavages at nts A14-A20 indicate that the area adjacent to the anchor (containing the
template for the first few editing sites) is single-stranded until nt A21 (Fig. 12). Distinct
cleavages using the single-strand specific RNases were also observed for nts A27-A36.
The lack of single-stranded specific cleavages in the U22-U26 and G37-G39 flanking

regions and weak cleavages of A41 to U42 suggested that the U22-U42 region base pairs

63

to the G37-U42 region and forms a stem with a large 10 nt terminal loop (A27-A36) (Fig.
12). The presence of this stem-loop is confirmed by the pattern of V1 sensitive cleavages
~ofnts U22-26 and nts G37-U42 (Fig. 12D). When the gRNA was paired with the
unedited mRNA (+CYbU) or the partially edited CYbPES3, the patterns of nuclease
cleavages for the guiding region were almost identical to those observed in the gRN A
alone (Fig. 12). This was particularly surprising for the NgCYb-558/CYbPES3 pair, as
the anchor duplex has doubled in length (26 bp vs 13 bp) and has incorporated the U22-
U26 5’-stem region from our gRNA stem-loop. All three gRNA structures (gRNA alone,
paired with CYbU and paired with CYbPES3) showed a pattern of single-strand RNase
sensitivity in the A27-A36 region flanked by RNase V1 sensitivity in the A21-U26 and
G38-A42 regions (Figs. 12, 15). While the RNase cleavage patterns of the guiding
region were very similar, we did see some consistent differences. For the NgCYb-
558/CYbU pair, weak but consistent MBN and RN ase T2 cleavage sensitivity could be
seen in the A15 to U18. In addition, both G17 and G19 are accessible to RNase T1,
suggesting the presence of a 5-6 nt bulge between the anchor duplex and the gRN A
guiding region stem (Fig. 12A, B, C). The corresponding region is completely protected
when the gRNA is paired with CYbPES3 as would be expected with the extension of the
anchor duplex to U26. The NgCYb-558 paired to CYbPES3 also showed a subtle
difference in the pattern of RNase V1 cleavage in the A21-U26 region (Fig. 12). Both
the gRNA alone and gRN A paired with unedited CYb had stronger RNase V1 cleavages
at U22 and C25 when these nts are forming the initial gRNA stem-loop. When the
gRNA is paired with CYbPES3 and nts U22 and C25 have become part of the anchor

duplex, the U22 V1 cleavage appears to be intense, while the other sites in this region had

relatively equal cleavage intensities (Fig. 12D). We also observed a slight increase in the
RNase V1 sensitivity in the U33-G37 region for the gRNA when paired to CYbPES3
(Fig. 12D, panel 3). In addition, the single-strand specific cleavages in the A27-A36
region were consistently less intense for the gRN A when paired with CY bPES3 (Fig. 12).
The cleavage patterns of the gRN A were incorporated into summary figures based on
computer modeling of the predicted secondary structure of the gRN A alone (Fig. 15A) as
well as the mRNA/gRN A complex with both unedited and partially edited mRNA (Fig
153, C). Our cleavage patterns support the computer predicted models and illustrate the
change in gRNA structure produced by editing of the mRNA. The unpaired gRNA forms
a gRNA stem-loop that is maintained when the gRNA pairs with the pre-edited mRN A
(Fig. 15A, B). Editing of the first three sites extends the anchor duplex by 13 bp,
incorporating one side of the gRNA stem into the anchor helix. The nucleotides that
were located within the terminal loop remain single-stranded, but are now located within
a bulge region between the anchor duplex and the new gRN A stem-loop. However, a
new gRN A stem-loop is formed by alternative base pairing of A36-A41 with part of the
U-tail (Fig. 15C).
Structure of CYbPES3

In order to corroborate the structure of the gRNA interaction with the partially edited
CYbPES3, we also probed the secondary structure of CYbPES3. For comparative
purposes, we investigated the structure of CYbPES3 alone as well as its structure when
paired with NgCYb-558 with and without the U-tail. In previous work, the structure of
the unedited CYb mRNA alone was found to be a strongly base paired stem-loop with the

first three editing sites contained in a terminal 5 nt loop (A35-U41) (Leung &

65

Koslowsky, 2001a). Editing of the first three sites, appears to add an additional 6
uridines to the terminal loop, but otherwise does not significantly change the structure
found within the stem region (Figs. 13, 14, 16). This is shown (Figs. 13, 14) by MBN
(U42-A36), T2 (U42-U38), and T1 (G41, G43) cleavages within the terminal loop (Fig.
l3, l6). RNase Vl data complements the data from the single-stranded nucleases
(Fig.13D, 16).

The structure of CYbPES3 crosslinked to its cognate gRNA (N gCYb-558) was much
more difficult to interpret. Pairing of the partially edited mRN A with its gRNA clearly
changed the structure of the mRN A, making it much more susceptible to the single-strand
specific nucleases (Fig. 13). Interestingly, while the anchor duplex region was mostly
protected, we did observe distinct cleavages in the U39-G43 regions for all three single-
strand specific nucleases (Fig. 13). This corresponds to a run of four G:U base pairs
within the anchor duplex created by the second and third RNA editing events. Within
this region, the RNase V1 analyses did complement the single-strand specific nucleases
with distinct V1 cleavages localized in two regions flanking the U39-G43 region; A33-
A37 and U47-A54. No corresponding single-strand nuclease sensitivity was observed
when probing the gRN A; however, this region was also protected in our RNase V1
analyses suggesting that tertiary interactions may be restricting nuclease access. It has
been proposed that G-U base pairs may locally destabilize helical regions widening the
major groove (Chow et al., 1992).

Upstream of the anchor duplex, nts A14-G27, were also much more sensitive to the
single-stranded nucleases, indicating that interaction with the gRN A opens up the stable

stem-loop structure of the mRNA (Fig. 13, 15C). Surprisingly, the run of six adenosine

66

Ac ‘ 6 Single Strand Specific
00% Mung Bean Nuclease
A _ u II. RNase T2

A A .. u +IIII RNase T1

A A .. U . g ' DEPC

A U -— A Double Strand Specific
AA-U nAA RNaseV1

67

Figure 16. Summary figure of CYbPES3 alone from solution structure probing gels that show
predicted secondary structure with cleavages from all 4 nucleases and DEPC using three sizes of
symbols to indicate intensity of cleavage. Circles indicate mung bean nuclease (single-strand
specific) cleavages, squares indicate RNase T2 (single-strand specific) cleavages, crosses indicate
RNase T1 (single-strand guanosine specific) cleavages, cylinders indicate DEPC (single-strand
adenosine specific), and triangles indicate RNase V1 (double-strand specific) cleavages.

68

residues (A28-A32), located just 5’ of the anchor duplex were clearly protected from
cleavage by both MBN and RNase T2 (Fig. 13). RNase V1 cleavages were observed
within this region. However, RNase V1 cleavages were also observed for nts A14-G27,
which also show distinct sensitivity to the single stranded nucleases (Fig. 13D). The
ability of RNase V1 to cleave stacked, highly structured regions and junctions found
between two helices may explain some of the overlapping nuclease sensitivity (Lowman
& Draper, 1986). In an effort to resolve our conflicting nuclease results, additional
structure probing experiments, using diethyl pyrocarbonate (DEPC), were performed
(Fig. 14). This chemical was chosen because it is reactive with purines and because their
reactivity can be monitored via chain scission, a necessity with our crosslinked substrates.
DEPC is very sensitive to the stacking of base rings; therefore, adenines within a helix
are not reactive. In these experiments, while the anchor is clearly protected, nts A26-A32
are reactive with DEPC suggesting that this region is in fact single-stranded and is not
highly structured (Figs. 14, 15C). Adenosine residues located further upstream (A14-
A21) are not reactive suggesting that this region is helical in nature (Figs. 14, 15C).

In comparing the cleavage patterns when CYbPES3 was paired to its gRNA with and
without (N gCYb-558sU) a U-tail, we saw very little difference. Previous crosslinking
experiments indicate that the U-tail does interact with the purine rich region around G25;
however, in these experiments the crosslinking efficiency was low indicating that this
interaction is not stable.

Conclusions
In this study, we investigated the structure of NgCYb—558 alone and its structure when

paired with its cognate unedited mRN A or partially edited mRN A. From our previous

69

structure studies (Leung & Koslowsky, 1999), the computer predicted models suggested
that the gRNA stem-loop (A29-U42) is maintained when the initial mRNA/gRNA
complex is formed (Fig. 15B). In addition, 3’ crosslinking studies with partially edited
substrates suggested that a gRNA stem-loop structure is preserved as editing proceeds
through the third editing site via interaction with the U-tail (A36-U53)(Fig. 15C). The
results of this study support the predicted models and suggest that the gRN A stem-loop
may be an important structural component of the initial editing complex. We hypothesize
that the formation of multiple helices surrounding the first few editing sites may allow for
tertiary interactions that help stabilize the initial gRNA/mRN A interaction. Increasing
the amount of editing increases the number of base pairs within the anchor duplex, and
this may negate the need for tertiary interactions during the later stages of the gRNA
interaction. Alternatively, the multiple helices may limit the number of potential

gRN A/mRN A interaction sites, and this may help increase the accuracy of the editing
process. This suggests that the U-tail may enhance the editing process in a number of
ways. The U-tail is added post-transcriptionally and has been shown to be unnecessary
for in vitro cleavage of editing substrates (Kable et al., 1996; Seiwert et al., 1996;
Burgess et al., 1999). However, sequence modifications that disrupt upstream base
pairing interactions severely diminish formation of edited product. Likewise, sequence
modifications that increase the stability of the upstream duplex are known to increase the
efficiency of in vitro editing, suggesting that the upstream gRNA/mRNA interaction is
important (Burgess et al., 1999; Kapushoc & Simpson, 1999; lgo et a1., 2000; Cruz-Reyes
et al., 2001). Previous work in our lab using gel shift analyses indicates that the U-tail is

very important for stabilization of the interaction of some mRNA/gRNA pairs

70

(Koslowsky et al., 2004). RNAi studies have also shown that when the RNA editing
terminal uridylyl transferase (RETl), responsible for adding the U-tail to gRNAs is down
regulated, there is a decrease in edited mRN As and inhibited grth of the trypanosome
(Aphasizhev et al., 2002). This suggests that the U-tail is necessary for in vivo editing
(Gott, 2003; Nebohacova et al., 2004). We know that the RNA editing process must
involve the formation and disruption of intramolecular helices to form a number of
intermediate complexes (Nordgren et al., 2001). Guide RNAs appear to be able to take
advantage of U-tail flexibility through the ability of uridines to base pair with both purine
bases. By employing an uridylate tail, the gRNA may increase mRNA/gRNA complex
stability without hampering the U-tail migration needed within the complex during the
editing process. In addition, the editing process may take advantage of the special
properties of G-U base pairs. The G-U base pair has a distinctly different geometry and
may locally destabilize helical regions creating protein recognition elements (Chow et al.,
1992; Batey & Williamson, 1996). Editing also creates G-U base pairs raising interesting
possibilities for protein recognition of the edited mRNA/gRNA complex, and the
recruitment of RNA helicases necessary for multiple gRNA utilization.

Recent work on the editing accessory factors MRP1 and MRP2 (Koller et al., 1997;
Blom et al., 2001; Aphasizhev et al., 2003b) provide a crystal structure for the
matchmaking proteins MRP1 and 2 binding a gRNA (Fig. 17). Both MRP1 and 2 are
required for a stable MRP complex to form and knockdown of this complex through
RNAi results in a reduction of CYb editing (Vondruskova et al., 2005). MRP1 was found

to have a RNA annealing activity that increases gRNA/mRNA complex formation

71

Schumacher, M.A. etal. (2006)
Cell 126:701-711.

 

Figure 17. The structure of the MRP1/MRP2 heterotetramer binding two gND7-506
gRNAs. This is a ribbon diagram of the MRP1/MRP2-gND7-506 complex. MRP1 is
colored magenta, while MRPZ is yellow. The gRNA nucleotides for the anchor region
inchrde green nucleotides representing a small stern-loop as well as white nucleotides
representing the region between the anchor helix and the guiding region stem-loop. Red
nucleotides represent the gRNA stem-loop of the guiding region. MRP = mitochondrial
RNA binding proteins. ND7 = NADH dehydrogenase subunit 7. This figure is fi'om
Schumacher, M.A. et al. (2006) Cell 126:701-711.

72

(Muller et al., 2001) through a putative charge reduction fimction between the
mRNA/gRNA anchor helix (Muller & Goringer, 2002). MRP1 and 2 form a
heterotetramer potentially binding two gRNAs. MRP1 binds the gRNA anchor and holds
it in a single stranded conformation similar to our CYb gRNA anchor structure, while
MRP2 binds the predicted gRNA stem-loop (Fig. 17) (Schumacher et al., 2006). The
structure data for the gRNA in this complex correlates well with our gRN A alone
structure for NgCYb-558 providing evidence that the RNA structures are relevant, and
that the gRN A anchor region is most likely in a single stranded conformation during
anchor helix formation. MRP1 presents the anchor sequence with the bases exposed and
ready for binding, while providing charge neutralization for the anchor phosphate
backbone. In addition, these accessory factors appear to recognize structure not
sequence as the protein/RNA binding interaction occurs primarily through electrostatic-
phosphate contacts. This appears to verify our hypothesis that the structure and not the
sequence of the mRNA/gRNA complex is recognized by the editosome proteins as it is
known that MRP1 and MRP2 are transient members of the editosome complex (Allen et
al., 1998). Interestingly, the structure of the gRNA is not altered by the MRP complex
(Hermann et al., 1997; Schumacher et a1., 2006), and the gRNA guiding region stem-loop
is maintained providing additional evidence that the gRNA stem-loop appears to be an
important structure for the editing process.

We have reported here the structure of the naked RNA for the dynamic CYb
mRNA/gRNA complex. It will be interesting to discover in future studies how much

these RNA structures and dynamics change or are maintained in the presence not only of

73

the accessory factors but also with the editosome proteins assembling and disassembling
from the mRNA/gRNA complex.

MATERIALS AND METHODS

Oligodeoxyribonucleotides:

All oligodeoxynucleotides (Table 3) were ordered fiom Integrated DNA
Technologies, Inc. (Coralville, IA).

Oligoribonucleotide:

Oligoribonucleotide was ordered fi'om Dharrnacon Research, Inc. (Lafayette, CO).
Uls-tail: 5’UUUUUUUUUUUUUUU3’, 15 nt.

DNA templates and RNA synthesis:

DNA templates of the partial mRN As: CYbU and CYbPES3 templates were created
by ligating two Oligodeoxyribonucleotides, the 5’ (5’ShortCYb and 5’CYbPES3 Short,
Table 3) and 3’ (3’ShortCYb and 3’CYbPES3Short, Table 3) halves of the molecules
using T4 DNA ligase (Roche) and a complementary DNA bridge (ShortCYb bridge and
CYbPES3 ShortDNAbridge, Table 3) (Moore & Sharp, 1992). The ligated templates
were then PCR amplified with T7 and Big SK oligonucleotides (Table 3) and Taq
polymerase (Promega) as per the manufacturer’s directions. The gRNA constructs
(N gCYb558 and NgCYb558(sU), Table 3) have been previously described (Leung &
Koslowsky, 2001b). The mRN A transcripts were synthesized either by T7 RNA
polymerase (Ribomax, Promega) according to the manufacturer’s directions or in the
presence of 5 mM guanosine as described previously (Leung & Koslowsky, 2001a).
NgCYb-558 and NgCYb-558sU were synthesized using a T7 RNA polymerase using a

Ribomax kit (Promega) in the presence (Burgin & Pace, 1990; Harris & Christian, 1999)

74

 

 

Name Sequence Length
(nts.)

Big SK 5’-GGCCGCTCTAGAACTAGTGG-3’ 20

T7 5’- AATTAATACGACTCACTATAG-3’ 22

NgCYb558(sU) 5 ’-TTATTCCCTTTATCACCTAGAAAT 65
TCACATTGTCTTTTAATCCCTATAGT
GAGTCGTA'I‘TAAATT-3’

NgCYb558 5’-AAAAAAAAAAAAAAATTATTCCCTT 80
TATCACCTAGAAATTCACATI‘GTCTTTT
AATCCCTATAGTGAGTCGTATTAAATT-

3,

gCYb558(sU) bridge 5’-AAAAAAAAAAAAAAATTATTCCC 32
'I‘I'I‘ATCACC-3 ’

5’ShortCYb 5’-CTTI‘CTTI"I'ITCTCCGCTI‘TTATA 59
TAAAATTTATAACCTATAGTGAGTC
GTATTAAA'I'T-3 ’

3 ’ShortCYb 5 ’-CCGCTCTAGAACTAGTGGATCCA 58
TATATTCTATATAAACAACCTGACAT
TAAAAGACC-3’

ShortCYb bridge 5’-GGAGAAAAAAGAAAGGGTCTTTT 31

5’CYbPES3Short

3’CYbPES3 Short

CYbPES3 ShortDNAbridge

AATGTCAG-3’
5’-CAAATTTCTTTTTTCTCCGCTTITA
TATAAAATTTATAACCCTATAGTGAG
TCGTATTAAATT-3 ’
5’-CCGCTCTAGAACTAGTGGATCCAT
ATATTCTATATAAACAACCTGACATTA
AAAGACAACA-3 ’
5’-GGAGAAAAAAGAAATTTGTGTTGT
CTT'ITAATGTCAG-3 ’

Table 3. List of Oligodeoxyribonucleotides

75

63

61

37

or absence of Guanosine 5’-monophosphorothioate (GMPS)(Biolog Life Science
Institutes, Germany or Harris Lab). If GMPS was incorporated at the 5’ end of the
NgCYb-558 transcripts (with and without U-tail), then 20 uCi of [320.P] rATP (Perkin-
Elrner) was added for visualization, the rGTP was reduced to 80 nmols, and 8 pl of 100
mM MgC12 was added per 80 pl reaction. The mRNAs and gRNAs were gel purified on
6% and 15% 7 M urea acrylamide gels respectively (Leung & Koslowsky, 2001a). The
Oligoribonucleotide Uls-tail was phosphorylated using T4 kinase(1nvitrogen) and ligated
to NgCYb-558sU using 25 U T4 DNA ligase (Roche) and the NgCYb558(sU) bridge
(Table 3) (Moore & Sharp, 1992; Leung & Koslowsky, 2001a).

RNA crosslinking and end-labeling:

Attachment of p-azidophenacyl bromide (Sigma) to GMPS NgCYb-558 and GMPS
NgCYb-558sU transcripts as well as crosslinking of gRNAs and mRNAs was done as
described previously (Leung & Koslowsky, 1999, 2001a, b). The mRNA (50 pmols of
free and 5 pmols of crosslinked mRNA) was 5’ end labeled using 50-100 uCi of [y 32P]
rATP (Perkin-Eirner) and T4 Kinase (Invitro gen) as per the manufacturers’ directions and
gel purified as above. The gRNA was 3’ end labeled by ligating 0.5 nmols of Uls-tail to
130 uCi of[5’-32P] cytidine 3’,5’-bisphosphate (pCp) (Perkin-Elmer) using T4 RNA
ligase (New England Biolabs) as per the manufacturer’s directions. The labeled Urs-tail
was ethanol precipitated, 5’ phosphorylated with T4 Kinase (Invitrogen) according to the
manufacturer’s directions, and ligated to NgCYb-558sU (no U-tail) using 25 U high
concentration T4 DNA Ligase (Roche) and the NgCYb558(sU) bridge (Table 3) as
previously described.

Structure specific enzymatic probing:

76

In these experiments, 150-200 kcpm of labeled RNA sample and 10 pg of tRNA were
heated to 50°C, cooled slowly, and incubated at 27°C for 20 minutes. Enzyme was then
added to the sample (0.1 U RNase T1 (Industrial Research, LTD), 0.18 U RNase V1
Cobra Venom (PierceMB), 0.4 U RNase T2 (Invitrogen), 1 U Mung Bean
Nuclease(GibcoBRL)). For RNases T1 and T2, aliquots were taken at 5, 10, and 15
minutes of digestion, and for RNase V1 and Mung Bean Nuclease aliquots were taken at
2, 5, and 10 minutes. The aliquots were immediately phenol/chloroform extracted,
ethanol precipitated, and then loaded on 12 or 15% denaturing acrylamide gels.
Reactions were performed in 10 mM Tris pH 7.5, 100 mM KCl, 1 mM MgC12 for all
enzymes except MBN. MBN had buffer conditions of 10 mM Tris pH 7.5, 10 mM
NaOAc pH 5.0, 0.1 mM ZnOAc, 50 mM NaCl, 1 mM L-cysteine, 5% glyceroL 1 mM
MgC12. RNA sequence ladders were made by denaturing 150 kcpm of labeled RNA and
2 ug of tRNA and digesting with 0.07 U RNase T1 and 0.2 U RNase U2 (Industrial
Research, LTD) for 10 minutes at 55°C in the following buffers: [l X Buffer I(T1): 19.8
mM NaCitrate pH 5.0, 1 mM EDTA, 4.2 M Urea, 0.02% Xylene Cyanol, 0.05%
Bromophenol blue], [1 X Buffer 11(U2): same as above except NaCitrate pH 3.5](buffer
recipes courtesy of Dr. Brenda Peculis, NIH, personal communication). The gels were
fixed, dried, and exposed on a phosphorirnaging screen overnight.

Solution Structure Probing with Chemicals:

For chemical modification and cleavage, the 5’ end labeled mRNA was annealed to
the gRNA by heating to 50°C and cooling slowly to 27°C where they were held for 30
minutes in 50ul of 1 X HE buffer (25 mM Hepes pH 7.5, 2 mM MgOAc, 50 mM KCl,

0.5 mM DTT, and 0.1 mM EDTA) and then crosslinked along with controls as above.

77

The mRN A alone samples as well as untreated samples were treated as above without
crosslinking. Ten pg of tRNA plus 40U of rRNAsin (Promega) were added to all
samples and a no chemical control aliquot (10pl) was taken. Four pl of 97% diethyl
pyrocarbonate (DEPC, Aldrich) was added and the samples incubated at room
temperature (Brunel & Romby, 2000). A 20pl aliquot was taken from each reaction at 15
and 25 minutes. The aliquots were immediately ethanol precipitated twice in the
presence of 0.2 M NaOAc pH 7.0, followed by a 70% ethanol rinse. Crosslinks and
controls were then gel purified on 6% acrylamide, 7 M urea gels as described above
(Leung & Koslowsky, 1999, 2001a, b). Strand scission of all chemical samples was
accomplished by incubating samples in 20 p1 of 1 M Aniline (Sigma-Aldrich) pH 4.5 for
10 minutes at 55°C, ethanol precipitating as above (Brunel & Romby, 2000), and
analyzed on 15% denaturing acrylamide gels. T1 RNA sequencing ladders were made as
described above. An alkaline hydrolysis ladder of mRN A was generated by incubating
100 kcpm of RNA and 10 pg tRNA in 50 mM NaHC03/Na2CO3 pH 9.0, lmM EDTA pH
8.0 at 90°C for 5 minutes.
Acknowledgements

This work was supported by National Institutes of Health Grant AI34155 to D.K
We’d like to thank Dr. Ron Patterson, Dr. Charles Hoogstraten, Dr. John Wang, Dr.
Sandra Clement, Larissa Reifur, and Melissa Mingler and all other members of the MSU
RNA Journal Club for critical reading of the manuscript and helpful discussions as well
as to Dr. Brenda Peculis at NIH for sharing her buffer recipes and digestion conditions.
We would also like to thank Dr. Michael Harris fi‘om Case Western University for

samples of GMPS and technical advice.

78

Literature Cited

Adler BK, Hajduk SL. 1997. Guide RNA requirement for editing-site-specific
endonucleolytic cleavage of preedited mRN A by mitochondrial ribonucleoprotein
particles in Trypanosoma brucei. Mol Cell Biol 1 7:5377-5385.

Allen TE, Heidmann S, Reed R, Myler PJ, Goringer HU, Stuart KD. 1998. Association of
guide RNA binding protein gBP2] with active RNA editing complexes in
Trypanosoma brucei. Mol Cell Biol 18:6014-6022.

Aphasizhev R, Aphasizheva 1, Nelson RE, Gao G, Simpson AM, Kang X, Falick AM,
Sbicego S, Simpson L. 2003a. Isolation of a U-insertion/deletion editing complex
fi'om Leishmania tarentolae mitochondria. Embo J 22:913-924.

Aphasizhev R, Aphasizheva 1, Nelson RE, Simpson L. 2003b. A 100-kD complex of two
RNA-binding proteins from mitochondria of Leishmania tarentolae catalyzes
RNA annealing and interacts with several RNA editing components. Rna 9:62-76.

Aphasizhev R, Sbicego S, Peris M, Jang SH, Aphasizheva 1, Simpson AM, Rivlin A,
Simpson L. 2002. Trypanosome mitochondrial 3' terminal uridylyl transferase
(TUTase): the key enzyme in U-insertion/deletion RNA editing. Cell 108:637-
648.

Aphasizhev R, Simpson L. 2001. Isolation and characterization of a U-specific 3'-5'-
exonuclease from mitochondria of Leishmania tarentolae. J Biol Chem
276:21280-21284.

Batey RT, Williamson JR. 1996. Interaction of the Bacillus stearothermophilus ribosomal
protein 815 with 16 S rRNA: II. Specificity determinants of RNA-protein
recognition. J Mol Biol 261 :550-567.

Blom D, Burg J, Breek CK, Speijer D, Muijsers AO, Benne R. 2001. Cloning and
characterization of two guide RNA-binding proteins from mitochondria of
Crithidia fasciculata: gBP27, a novel protein, and gBP29, the orthologue of
Trypanosoma brucei gBP2]. Nucleic Acids Res 29:2950-2962.

Blum B, Bakalara N, Simpson L. 1990. A model for RNA editing in kinetoplastid
mitochondria: "guide" RNA molecules transcribed from maxicircle DNA provide
the edited information. Cell 60: 189-198.

Blum B, Simpson L. 1990. Guide RNAs in kinetoplastid mitochondria have a
nonencoded 3' oligo(U) tail involved in recognition of the preedited region. Cell
62:391-397.

Bnmel C, Romby R. 2000. Probing RNA Structure and RNA-Ligand Complexes with
Chemical Probes. In: Abelson JN, Simon MI, eds. Methods in Enzymologz: RNA-
Ligand Interactions. San Diego: Academic Press. pp 3-21.

79

Burgess ML, Heidmann S, Stuart K. 1999. Kinetoplastid RNA editing does not require
the terminal 3' hydroxyl of guide RNA, but modifications to the guide RNA
terminus can inhibit in vitro U insertion. Rna 5 :883-892.

Burgin AB, Pace NR 1990. Mapping the active site of ribonuclease P RNA using a
substrate containing a photoaffinity agent. Embo J 9:4111-4118.

Chow CS, Behlen LS, Uhlenbeck OC, Barton JK. 1992. Recognition of tertiary structure
in tRNAs by Rh(phen)2phi3+, a new reagent for RNA structure—function
mapping. Biochemistry 31:972-982.

Cruz-Reyes J, Zhelonkina A, Rusche L, Sollner-Webb B. 2001. Trypanosome RNA
editing: simple guide RNA features enhance U deletion 100-fold. M01 Cell Biol
21:884-892.

Feagin JE, Jasmer DP, Stuart K 1987. Developmentally regulated addition of nucleotides
within apocytochrome b transcripts in Trypanosoma brucei. Cell 49:337-345.

Feagin JE, Stuart K. 1988. Developmental aspects of uridine addition within
mitochondrial transcripts of Trypanosoma brucei. Mol Cell Biol 8: 1259-1265.

Gott JM. 2003. Two distinct roles for terminal uridylyl transferases in RNA editing. Proc
Natl Acad Sci U S A 100: 10583-10584.

Harris ME, Christian EL. 1999. Use of circular permutation and end modification to
position photoaffinity probes for analysis of RNA structure. Methods 18:51-59.

Hayman ML, Read LK 1999. Trypanosoma brucei RBP16 is a mitochondrial Y-box
family protein with guide RNA binding activity. J Biol Chem 274 : 12067-12074.

Hermann T, Schmid B, Heumann H, Goringer HU. 1997. A three-dimensional working
model for a guide RNA from Trypanosoma brucei. Nucleic Acids Res 25 :231 1-
2318.

lgo RP, Jr., Palazzo SS, Burgess ML, Panigrahi AK, Stuart K. 2000. Uridylate addition
and RNA ligation contribute to the specificity of kinetoplastid insertion RNA
editing. Mol Cell Biol 20:84-47-8457.

Kable ML, Seiwert SD, Heidmann S, Stuart K 1996. RNA editing: a rmchanism for
gRNA-specified uridylate insertion into precursor mRNA. Science 273:1189-
1195.

Kapushoc ST, Simpson L. 1999. In vitro uridine insertion RNA editing mediated by cis-
acting guide RNAs. RNA 5 :656-669.

Koller J, Muller UF, Schmid B, Missel A, Kruft V, Stuart K, Goringer HU. 1997.
Trypanosoma brucei gBP21. An arginine-rich mitochondrial protein that binds to
guide RNA with high aff’mity. J Biol Chem 272:3749-3757.

80

Koslowsky DJ, Goringer HU, Morales TH, Stuart K 1992. In vitro guide RNA/mRN A
chimaera formation in Trypanosoma brucei RNA editing. Nature 356:807-809.

Koslowsky DJ, Kutas SM, Stuart K 1996. Distinct differences in the requirements for
ribonucleoprotein complex formation on differentially regulated pre-edited
mRN As in Trypanosoma brucei. Mol Biochem Parasitol 80: 1-14.

Koslowsky DJ, Reifur L, Yu LE, Chen W. 2004. Evidence for U-Tail Stabilization of
gRNA/mRNA Interactions in Kinetoplastid RNA Editing. RNA Biology 1:28-34.

Leung SS, Koslowsky DJ. 1999. Mapping contacts between gRN A and mRN A in
trypanosome RNA editing. Nucleic Acids Res 27 2778-787.

Leung SS, Koslowsky DJ. 2001a. Interactions of mRNAs and gRNAs involved in
trypanosome mitochondrial RNA editing: structure probing of an mRN A bound to
its cognate gRNA. RNA 7: 1803-1816.

Leung SS, Koslowsky DJ. 2001b. RNA editing in Trypanosoma brucei: characterization
of gRN A U-tail interactions with partially edited mRN A substrates. Nucleic Acids
Res 29:703-709.

Lowman HB, Draper DE. 1986. On the recognition of helical RNA by cobra venom V1
nuclease. J Biol Chem 261:5396-5403.

Madison-Antenucci S, Sabatini RS, Pollard VW, Hajduk SL. 1998. Kinetoplastid RNA-
editing-associated protein 1 (REAP-1): a novel editing complex protein with .
repetitive domains. Embo J I 726368-6376.

McManus MT, Shimamura M, Grams J, Hajduk SL. 2001. Identification of candidate
mitochondrial RNA editing ligases from Trypanosoma brucei. Rna 7: 167-175.

Missel A, Souza AE, Norskau G, Goringer HU. 1997. Disruption of a gene encoding a
novel mitochondrial DEAD-box protein in Trypanosoma brucei affects edited
mRNAs. Mol Cell Biol 17:4895-4903.

Moore MJ, Sharp PA. 1992. Site-specific modification of pre-mRNA: the 2'-hydroxyl
groups at the splice sites. Science 256:992-997.

Muller UF, Goringer HU. 2002. Mechanism of the gBP2 l-mediated RNA/RNA
annealing reaction: matchmaking and charge reduction. Nucleic Acids Res
30:447-455.

Muller UF, Lambert L, Goringer HU. 2001. Annealing of RNA editing substrates
facilitated by guide RNA-binding protein gBP21. Embo J 20: 1394-1404.

Nebohacova M, Maslov DA, Falick AM, Simpson L. 2004. The effect of RNA
interference Down-regulation of RNA editing 3'-terminal uridylyl transferase

81

(TUTase) l on mitochondrial de novo protein synthesis and stability of
respiratory complexes in Trypanosoma brucei. J Biol Chem 279:7819-7825.

Nordgren S, Slagter-Jager JG, Wagner GH. 2001. Real time kinetic studies of the
interaction between folded antisense and target RNAs using surface plasmon
resonance. J Mol Biol 310:1 125-1134.

Panigrahi AK, Allen TE, Stuart K, Haynes PA, Gygi SP. 2003a. Mass spectrometric
analysis of the editosome and other multiprotein complexes in Trypanosoma
brucei. J Am Soc Mass Spectrom 14:728-735.

Panigrahi AK, Gygi SP, Ernst NL, lgo RP, Jr., Palazzo SS, Schnaufer A, Weston DS,
Carmean N, Salavati R, Aebersold R, Stuart KD. 2001a. Association of two novel
proteins, TbMP52 and TbMP48, with the Trypanosoma brucei RNA editing
complex. Mol Cell Biol 21:380-389.

Panigrahi AK, Schnaufer A, Carmean N, Igo RP, Jr., Gygi SP, Ernst NL, Palazzo SS,
Weston DS, Aebersold R, Salavati R, Stuart KD. 2001b. Four related proteins of
the Trypanosoma brucei RNA editing complex. Mol Cell Biol 21 :6833-6840.

Panigrahi AK, Schnaufer A, Ernst NL, Wang B, Carmean N, Salavati R, Stuart K 2003b.
Identification of novel components of Trypanosoma brucei editosomes. Rna
9:484-492.

Pelletier M, Read LK 2003. RBP16 is a multifunctional gene regulatory protein involved
in editing and stabilization of specific mitochondrial mRNAs in Trypanosoma
brucei. RNA 92457-468.

Riley GR, Corell RA, Stuart K 1994. Multiple guide RNAs for identical editing of
Trypanosoma brucei apocytochrome b mRNA have an unusual minicircle
location and are developmentally regulated. J Biol Chem 269:6101-6108.

Rusche LN, Huang CE, Piller KJ, Hemann M, Wirtz E, Sollner-Webb B. 2001. The two
RNA ligases of the Trypanosoma brucei RNA editing complex: cloning the
essential band IV gene and identifying the band V gene. Mol Cell Biol 21:979-
989.

Schmid B, Riley GR, Stuart K, Goringer HU. 1995. The secondary structure of guide
RNA molecules from Trypanosoma brucei. Nucleic Acids Res 23 :3093-3 102.

Schnaufer A, Panigrahi AK, Panicucci B, lgo RP, Jr., Wirtz E, Salavati R, Stuart K
2001. An RNA ligase essential for RNA editing and survival of the bloodstream
form of Trypanosoma brucei. Science 291 :2159-2162.

Schumacher MA, Karamooz E, Zikova A, Trantirek L, Lukes J. 2006. Crystal Structures
of T. brucei MRP1/MRP2 Guide-RNA Binding Complex Reveal RNA
Matchmaking Mechanism. Cell 126:701-71 1.

82

Seiwert SD, Heidmann S, Stuart K 1996. Direct visualization of uridylate deletion in
vitro suggests a mechanism for kinetoplastid RNA editing. Cell 84:831-841.

Seiwert SD, Stuart K 1994. RNA editing: transfer of genetic information from gRNA to
precursor mRNA in vitro. Science 266:114-117.

Simpson L, Aphasizhev R, Gao G, Kang X. 2004. Mitochondrial proteins and complexes
in Leishmania and Trypanosoma involved in U-insertion/deletion RNA editing.
RNA 10:159-170.

Stuart KD, Schnaufer A, Ernst NL, Panigrahi AK 2005. Complex management: RNA
editing in trypanosomes. Trends Biochem Sci 30:97-105.

Vanhamme L, Perez-Morga D, Marchal C, Speijer D, Lambert L, Geuskens M,
Alexandre S, Ismaili N, Goringer U, Benne R, Pays E. 1998. Trypanosoma brucei
TBRGGl, a mitochondrial oligo(U)-binding protein that co-localizes with an in
vitro RNA editing activity. J Biol Chem 2 73:21825-21833.

Vondruskova E, van den Burg J, Zikova A, Ernst NL, Stuart K, Benne R, Lukes J. 2005.
RNA interference analyses suggest a transcript-specific regulatory role for
mitochondrial RNA-binding proteins MRP1 and MRP2 in RNA editing and other
RNA processing in Trypanosoma brucei. J Biol Chem 280:2429-2438.

83

CHAPTER 3
INTERACTIONS OF MRNAS AND GRNAS INVOLVED IN TRYPANOSOME
MITOCHONDRIAL RNA EDITING: MEASUREMENTS OF THE REAL TINIE
KINETICS OF BINDING FOR THREE MRNAS BOUND TO THEIR GRNAS

(The ND7 experiments were run by Aimee B. Ogden, a former undergraduate, under my
supervision.)

84

Introduction

RNA editing in kinetoplasts is a post-transcriptional process that involves the precise
insertion and deletion of uridine residues in mitochondrial mRN As. These changes are
made by more than twenty proteins collectively known as the editosome and are directed
by a small RNA called a guide RNA (gRNA). The gRNA plays key roles in the template
directed cleavage, uridylate insertion or deletion and finally religation of the transcript
(Simpson et al., 2004; Stuart et al., 2005). Therefore, the ability of the gRNAs to target
their cognate substrates efficiently is an important part of the editing process. The
gRNAs that direct the editing process are short RNA molecules (~50-70 nucleotides)
made of three functional elements. At the 5’ end of the gRNA is a 5-21 nucleotide region
known as the anchor that is complementary to the mRNA just 3’ of the editing domain.
The second element is known as the guiding region and serves as the template for proper
editing of the mRN A. The guiding region is complementary to the mature mRN A
(allowing G-U base pairs). The poly-uridylate tail (~15 nts) is the third element, and it is
added post-transcriptionally (Blum et al., 1990; Blum & Simpson, 1990).

Little is known about how fast editing occurs, so it is unknown if the gRNA/mRNA
complex requires additional protein factors for a faster association rate. However, the
stability of the mRNA/gRNA complex may be important for the editing process. In
previous work, different mRNA/gRNA pairs were investigated to find the difference in
the binding interaction each gRNA had for its cognate mRNA: the unedited ATPase 6
subunit (A6) mRNA and its initiating gRN A, gA6-14, and the unedited apocytochrome b
(CYb) mRNA with its initiating gRNA, gCYb-558 (Koslowsky et al., 2004). The A6

mRN A is constitutively edited and is used for the in vitro studies of the editing process.

85

The apparent KD (Knapp) for the gA6-14/A6UENDSh interaction is in the nM range
indicating that this guide RNA has high affmity for its cognate message. Editing of the
CYb mRN A is developmentally regulated occurring preferentially in the insect host
(procyclic) (Feagin et al., 1985). In contrast to gA6-14, the interaction between gCYb-
558 and its cognate unedited mRNA is very weak with apparent Kp’s in the pM range.
Solution structure probing indicates that the CYb mRN A forms a stable stem-loop
structure, with the anchor binding site (ABS) found within the double-stranded stem (Fig.
20A) (Leung & Koslowsky, 2001a). Our hypothesis is that the stable secondary structure
of the CYb mRNA plays an important role in the regulation of the editing process for this
transcript. The stability of the CYb mRN A stem-loop suggests that a RNA helicase is
probably required for effective gRN A interaction. RNA structure is critical for many
RNA-RNA and RNA-protein interactions to display nucleotides in the correct orientation
for efficient interaction. It is clear that the structure of the immediate editing domain can
profoundly affect the efficiency of the interaction between the gRN A and the target
mRNA. To gain a better understanding of how structure surrounding the anchor binding
site may affect gRNA targeting, we analyzed the kinetic parameters of binding for three
different mRN A/gRN A pairs using surface plasmon resonance technology. The
dissociation rate constants (kofi‘), the association rate constants (ken), and the dissociation
equilibrium constants (K0) are compared at 2 mM magnesium for the unedited
apocytochrome b (CYb) mRN A/gRN A interaction as well as two other initiating

mRNA/gRNA pairs, ATPase 6 (A6) and NADH

86

m
G
e e.
n n
s A
W AGAA A
E U UAAGAUCA
U A ..__.. .
6 We @UUUUAGU
m are a w
1 u
a 6
AA
AM
A G G G
UUUU UCA GUUUUUAUA
.__. _— .___.. _.
AAAAAAGAGGCGAAAA Au

3’

3"

C mND7550

0666

m.
e

O
07
V@@\0A c A G r

87

Figure 18 continued D NgCYb-553

3", 59¢
Uu U G40
uGUAACAG
UAAU AACi
A — u
A _ U E gA6-14 A u
e -— u A
21°. 3. °
— Uu _
U A G ”6 uA A? -
u
SCAA”
uA Au _ AA
A- u
F gND7-550 e - c
5'Gc GU - AA
3, ‘G A- u
””00 A G - U
”U” UGAAU“ U U
U u - AUG 6 U
Ac 6
u -A A
u — e
u-AU
”u—e
u — A
A- u
u -A
AA- 0AA
G A
Us U

88

Figure 18 continued
G CYbU+NgCYb-558

000:3
a G6“I; 09‘") \\ ”is“ (’65.
"IUIUU u/JUI/l/ ,Gf‘lAG 61°8\\;;,#° 6
c)
AA AA Anchor Helix
U-tall Hollx U .... 4A6 GU
1:0 " .
c \ 6! 3 .
Gd, . p
4° \ 3 f 3036
A A 0 F
‘u c 31
‘6 0° 3‘ H A6UENDSh+gA6-14

  

 

o
ofio
o o
”y. 881
I ,Fe Gull/
, mND7550+gND7-550 0 won uuoerurIAque'

3' '° 8“ a: lééélc'lilllé.

c-
; g2§ 531 “2'8“ A“

c- ..
' c-G / . e-c
I c—O A c3 a Anchorl-lallx
" “—9 AG Ac \H“ .

c—>c Ac Hull couGG°5 U-A

1,1u 1 “$5 “UAGGUGA A A-U r;

u. .. / u°'“:: ,

u-c“ . Anchorl-Ialix GACAGJ ,

U-A =0

A-U E

u-A g.

aA-UAA

chuA g

Figure 18. Predicted secondary structures of the A6, CYb, and ND7 RNAs. A. CYbU,
the unedited mRNA, forms a stable stem-loop (5 bases) with the first few editing sites
positioned within the terminal loop and most of the anchor binding site is base paired. B.
A6UENDSh, an unedited mRN A with a loose open structure and part of the anchor
region in a loop. C. mND7550, another unedited mRNA, has a loose open structure with
a mostly single-stranded anchor binding site. The hollow letters represent nucleotides
that are part of the anchor binding site. E8] = editing site 1. D. NgCYb-558 gRNA
secondary structure. E. gA6-14 gRNA secondary structure. F. gND7-550 gRNA
secondary structure. All gRNAs are predicted to have a mostly single stranded anchor
sequence, a gRNA stem-loop, and a single stranded U-tail. G. CYbUSh+NgCYb-558
mRNA/gRNA complex secondary structure. H. A6UENDSh+gA6~l4 mRNA/gRNA
complex predicted secondary structure. I. mND7-550+gND7-550 mRNA/gRNA
complex predicted secondary structure. All three mRNA/gRNA complexes contain
mRNA/gRNA anchor helices, a gRNA stem-loop, and a mRNA/U-tail helix.

89

dehydrogenase 7 (ND7). In contrast to the CYb substrate, the A6 and ND7 mRN As are
constitutively edited and predicted to have relatively little structure surrounding their
single stranded anchor binding sites.

In this study, the CYb mRN A/gRN A interaction appears to be severely affected by the
structure of the mRN A. The association rate for CYb is unusually slow compared to the
other mRN A/ gRN A substrates. In addition, the dissociation rate is considerably faster,
resulting in a calculated equilibrium dissociation constant (K0) in the micromolar range.
The A6 and ND7 substrates show that in the absence of significant mRN A structure (such
as the CYb mRN A) and when compared to other RNA-RNA binding interactions, the
mRN A/gRN A interaction appears to have a relatively normal association rate coupled
with a much slower dissociation rate. This study allows us to begin to dissect what
elements of the mRN A and gRN A, sequence and structure, are important to achieve
stable duplex formation. The high afl'mity observed in the A6 and ND7 mRNA/gRN A
interactions is the result of a very slow dissociation rate. This complex stability may be
required for editing to occur.

Results
mRNA/gRNA pairs examined using surface plasmon resonance
CYbU-NgCYb558

The CYbU and NgCYb-558 RNAs used in this study have been previously described
(Yu & Koslowsky, 2006). CYbU is composed of the first 88 nts fi'om the CYb mRNA 5’
end. Editing of this mRN A is developmentally regulated with the insertion of 34
uridylates at 13 sites occurring preferentially during the procyclic stage of the life cycle

(Feagin et al., 1987; Feagin & Stuart, 1988). The gRNA construct, NgCYb-558, is 59

90

nucleotides long including a 15 nt. uridylate tail (U-tail) (Table 4) (Leung & Koslowsky,
2001b). NgCYb-558 is almost identical to wild type and directs 21 uridylate insertions at
the first 7 editing sites. The secondary structures for NgCYb-558, the CYbU mRN A
substrate, and the gRNA/mRN A complex have been determined using both solution
structure probing and UV-crosslinking techniques (Schmid et al., 1995; Leung &
Koslowsky, 1999, 2001a, b; Yu & Koslowsky, 2006). The CYbU mRNA substrate alone
folds into a stable stem-loop (AG = -24.47 kCal/mol) (Reifur et al., 2006), with the
terminal loop containing the first three editing sites (Fig. 18A). The anchor binding site
(ABS) of the unedited CYb mRNA is composed of 13 nts and is found sequestered
within the stable mRNAstem (Leung & Koslowsky, 2001a). The CYb gRNA has a
mostly single-stranded anchor. The guiding region is contained within a stem-loop that
forms before editing begins whereas the U-tail appears to have a single-stranded
conformation (Fig. 18D) (Schmid et al., 1995; Yu & Koslowsky, 2006). Duplex
formation between the gRNA and mRNA results in the formation of a three helical
structure surrounding the first few editing sites. These include a gRN A/mRN A anchor
helix, gRNA stem-loop, and U—tail/mRNA helix (Fig. 186) (Leung & Koslowsky, 2001a;
Yu & Koslowsky, 2006). The anchor helix formed between the gRNA and mRN A
involves 12 base pairs composed of eight A-U bps, three G-C bps, and one G-U bp (Fig.
19A)
A6UENDSh-gA6-l4

ATPase 6 subunit (A6) mRN A appears to have consistent levels of editing in either
host (Bhat et a1., 1990) and is extensively edited with multiple gRNAs. The A6UENDSh

substrate used in this study represents 80 nts fi'om the 3’ end of the ATPase 6 subunit

91

 

Analyte No.
RNA or Sequence of
Ligand nts.

CYbU+BigSK Ligand 5’GGUUAUAAAUUUUAUAUAAAAGCGG
AGAAAAAAGAAAGGGUCUUUUAAUGUC

AGGUUGUUUAUAUAGAAUAUAUGGAUC 100
atatggaagtaattagaagga-biotin3’

CYbU+ Ligand S’GGUUAUAAAUUUUAUAUAAAAGCGGA

Biatagl GAAAAAAGAAAGGGUCUUUUAAUGUCAG
GUUGUUUAUAUAGAAUAUAUGGAUCcacua 98

guucuagagcggcc-biotinB’

NgCYb558 Analyte S’GGGAUUAAAAGACAAUGUGAAUUUCUA
GGUGAUAAAGGGAAUAAUUUUUUUUUUU

UUUU3’ 59
mND7550+Bi Ligand 5’AAAAACAUGACUACAUGAUAAGUACAAG
gSK AGGAGACAGACGACAGUGUCCACAGCACC 91

CGUUUCAGCACAGatatggaagtaattagaagga-biotin3 ’

gND7-550 Analyte 5’GGGAUGCAGUGGAUUAAGUGUAAGAUGA
UAUAAAUGUGAAUAUUUUUUUUUUUUUUU3’ 57

A6UENDSh+ Ligand S’GGAAAGGUUAGGGGGAGGAGAGAAGAAA

BigSK GGGAAAGUUGUGAUUUUGGAGUUAUAGAA
UAAGAUCAAAUAAGUUAAUAAUAcactagttcta 99
gagcggcc-biotin3’

gA6- l 4 Analyte 5 ’AUAUACUAUAACUCCGAUAACGAAUC AG
AUUUUGACAGUGAUAUGAUAAUUAUUUUU 68
UUUUUUUUUUU3 ’

Table 4. List of all Biacore substrates used in Biacore experiments. The biotinylated
DNA tags were ligated on using DNA bridges (listed in Table 6) and are indicated by
lowercase letters. Column 1: RNA name. Column 2: Ligands were biotinylated and
attached to streptavidin coated SA chips (Biacore, Uppsala, Sweden) and Analytes were
injected in running buffer and flowed across the surface of the chip to bind the Ligand.
Column 3: Sequence of the Biacore substrates. Column 4: number of nucleotides.

92

.9555 on“ 3 39592» 8 3:3?

unease 89528 2F .0 £223 05 a 3-9% 2 8:93 atoms: smnzma,‘
BF .m fiancee 05 a wmmb>omz 8 conga aroma—E D56 2:. .< .23 2a €235
wedge 05 .8538 $880388 n _ £8 83 DO u H .8538 vegans—£8 u a
.8538 “38> n vacuo— oze: 688033 .392 05 .8 80:2. 8:05 2: .2 oSwE
§4306§¢B§e§o§§4§s05053533238259:o m m - 52m

///*__¢___¢__n .............................................
0¢U¢UO¢UDDDUUUU¢UUdU‘UUDDDU4U4GU<G¢U¢G¢GU¢U<40ébU¢¢D<UD40450¢wD<U¢¢¢¢<

“ammnhA—Z 3 to: 2a 5.5.— cmmR—ZE
U

0000D¢b4¢UDUUO¢DddU0¢¢DU<G<DDDDG¢U¢UDU¢D¢DU¢S¢¢SD<asn«anmdmz

.§__________._ .............................................

«sfisfipsoéag pmofiogggpo‘oopppagoopsogooogufiofigo40000048005- ..
”em-e4 3 to: an Amfizgc<
m

UGO‘bbldddaflUddDUDOddDDDUD¢0050¢D¢d¢000¢¢D<¢DDDDDDDDDDDDDDDmmminflumz

ooz<s§<4w§§¢§ossoo¢upobd<§g§§¢80§5<94§4§5¢8we ................
£3-56 z 3 to: =a :56
<

 

 

 

93

mRNA. There are 26 editing sites within this substrate. Solution structure probing of the
mRN A alone indicates that the A6UENDSh substrate forms a short stem-loop structure
(stern = 8 bps, AG = -8.00 kCal/mol) (Reifiir et aL, 2006). However, in contrast to the
CYb mRNA, the ABS for the A6 mRNA is mostly located within the single stranded
loop region (Fig. 18B) (Reifur & Koslowsky, 2006). The initiating gRNA, gA6-l4,
directs insertion of 57 uridines in 23 sites and deletion of 6 uridines at 3 sites (Table 4).
Solution structure probing of the A6 gRNA suggests that it forms a structure similar to
NgCYb-558 with a mostly single-stranded anchor, a modified gRN A stem-loop, and a
single-stranded U-tail (Fig. 18E) (Schmid et al., 1995; Reifur & Koslowsky, 2006). UV
crosslinking analyses combined with computer modeling indicate that the A6
mRNA/gRNA complex also forms a three helical structure including an anchor helix,
gRNA stem-loop, and U-tail/mRN A helix (Leung & Koslowsky, 1999). However,
between the anchor helix and the gRNA stem-loop, there is a bulge as well as an
additional short helical region (Fig. 18H). This bulge 5’ (mRNA) of the anchor helix
contains the first editing site, and the upstream short helical region is composed of 4
consecutive bps (1 G-C, 2 A-U, and l G-U) that may contribute to complex stability
(Reifur & Koslowsky, 2006).
mND7550-gND7-550

‘ Editing of the NADH dehydrogenase subunit 7 mRNA (ND7) is unusual in that
editing occurs within two distinct domains separated by 59 nts (homology region 3
(HR3)) that are not edited (Koslowsky et al., 1990). The mND7-550 mRNA substrate
(Table 4) used in this study consists of the 5’ most 70 nts of the ND7 transcript. It

contains the anchor binding site for gND7-550; a span of 14 nts that are not edited in the

94

mature transcript. The computer predicted structure for mND75 50 suggests that this
substrate has very little structure surrounding the ABS (AG = -10.83 kCal/mol) (Reifur et
aL, 2006). While the AG suggests that mN D7550 is slightly more stable than
A6UENDSh, it is characterized by multiple short helices. The mND7550 ABS is
predicted to be mostly single stranded, spanning the region between two short helices
(Fig. 18C). The gND7-550 gRNA directs 33 uridine insertions into 12 editing sites and
appears to initiate editing (Koslowsky et al., 1990; Corell et al., 1993). The gND7-550
gRNA is composed of 57 nts and is predicted to have a single-stranded anchor sequence,
a modified gRN A stem-loop that incorporates six uridylates from the U-taiL and
additional single-stranded uridylates in the U-tail (Fig. 18F) similar to the general
predicted structure for most gRNAs (Schmid et al., 1995). The mND7550/gND7-550
anchor helix is composed of 3 A-U bps, 8 G-C bps, 2 mismatches, and l G-U bp (Fig.
19C). The ND7 mRNA/gRNA complex is also predicted to form a three helical structure
composed of an anchor helix, gRN A stem-loop, and U—tail/mRN A helix. Between the
anchor helix and the gRN A stem-loop there are two predicted bulges, a mismatch, and a
short helical region (Fig. 181). The first predicted bulge 5’ (mRNA) of the anchor helix
contains the first editing site, and further editing would disrupt the upstream short helical
regions. However, the additional bps upstream of the anchor contain 3 G-C and 2 A-U
bps, and these would have additional stabilizing effects for complex formation The ND7
anchor helix contains more G-C bps than A6 and CYb; however, there are also 2 internal
mismatches.

Surface Plasmon Resonance Studies

95

The association and dissociation rate constants of the gRNA/mRN A interactions were
measured using surface plasmon resonance (SPR). The BIACORE 2000 (Biacore,
Uppsala, Sweden) monitors intermolecular interactions through SPR that arises when
incident light is reflected from a thin gold film between two layers. A change of the
refiactive index changes the angle of light which is recorded by the detector as a change
in resonance units (RU). The change in refractive index depends on the material close to
the non-illuminated side of the gold surface where a binding reaction takes place and is
closely correlated to a change in mass near the gold surface. The rate at which this
change occurs provides information about the association and dissociation rates of the
molecules (Katsamba et a1, 2002).

In these experiments, a deoxyoligonucleotide tag with a 3’ biotin label was ligated
onto the 3’ ends of the target mRNAs (Moore & Sharp, 1992). The biotin labeled
mRN A was then immobilized to the streptavidin covered surface of the SA chip
(Biacore, Uppsala, Sweden) in two of the four channels of a BIACORE 2000 (Biacore,
Uppsala, Sweden). One channel remained empty and was used as a reference surface
and one contained the biotinylated tag as a control for background binding. When
gRNA was injected over the channels and bound mRNA, there was a change in
refractive index which was displayed as a sensogram. A continuous flow of gRNA
solution at various concentrations (30 to 1500 nM) was injected over the immobilized
mRN As to monitor the gRNA association with its target mRNA. The dissociation phase
was monitored by chasing the gRNAs with buffer alone.

In order to see reliable gRNA binding to the mRNA, 50 to 350 resonance units of

mRN A was attached to the chip surface. For the CYb substrate, binding did not come to

96

equilibrium even with high gRNA concentrations and long injection times. In addition,
the long injection times and the regeneration procedure used between two binding assays
at different RNA concentrations progressively affected the mRN A integrity during the
experiments (von der Haar & McCarthy, 2003). These limitations make it difficult to
generate curves amenable to simple Scatchard-type analyses and thermodynamically
' valid rate constants for the different mRN A/gRN A pairs under similar conditions. The
CYb interaction has a very slow association rate and appears to be at the limit of
detection for the Biacore machine for the on-rate; however, it is within the range of the
off-rate. In addition, line fitting using global analysis requires extremely high quality
experimental data. Since CYbU kinetics is at the detection limit for the machine, this
method of analysis proved impractical (Myszka, 1997). However, using the separate fits
function of Biaevaluation 3.0 (Biacore, Uppsala, Sweden) with a global analysis for each
association and dissociation curve separately allowed an analysis of the individual rate
constants. The individual rate constants were averaged and the equilibrium dissociation
constant was calculated fi'om the rate constants. The errors reported were based on the
variances of all curves obtained (Nordgren et al., 2001).
The NgCYb-558/5’CYbUT Interaction

For the binding of NgCYb-558 to its cognate mRNA target, CYbU, 150-250 pl of 2
mM magnesium buffer containing 200 to 1500 nM gCYb-558 were injected, at a flow
rate of 10 ill/min over the surface of the sensor chip. After each injection of gRNA,
buffer was allowed to flow for 15-20minutes at 10 ul/min for dissociation rate
information. A one minute injection of 8 M Urea or 10 mM EDTA was run between

every cycle to strip the gRNA fi'om the mRN A.

97

Using surface plasmon resonance (SPR), for CYbU+NgCYb-558 (Table 4) we saw an
association rate constant (kw) for CYbU+NgCYb-558 of 570 M'ls'l (5.7 x 102 M‘ls”).
The dissociation rate constant (km) for the CYb complex was 2.7 x 10'3 5'1 (Table 5).
These numbers were calculated fi'om separate line fits of the association and dissociation
curves from the Biacore sensograms in BIAevaluation 3.0 (Biacore, Uppsala, Sweden)
(Fig. 20A). The dissociation equilibrium constant (Kn) of 4.7 pM (4.7 x 1043 M) was
calculated from the rate constants above (Table 5) using the equation K0 = ken/k0". This
KD is ten fold higher than the apparent Kn found for CYb using gel shift assays
(Koslowsky et a1., 2004).

The association rate for CYbU is very slow compared to loop-loop RNA-RNA
interactions such as TAR-TAR“ and CopA-CopT; however, the dissociation rate is
comparable to these RNA-RNA interactions (Persson et al., 1988; Nair et a1., 2000;
Nordgren et al., 2001). The antisense RNAs such as CopA-CopT are diffusion-controlled
relying on a rapid association in conjunction with high antisense concentrations in order
to achieve post—transcriptional control of plasmid propagation (Nordgren et aL, 2001).
The CYb mRNA stem-loop does not allow complex formation to be diffusion-controlled
(i.e. mRNA and gRNA do not bind every time they come into contact). The slowness of
the CYb association rate suggests that the CYbU complex probably rarely forms in vivo
without help from an annealing factor.

The gA6-14 /A6ENDSh Interaction

For the binding of gA6-14 to its cognate mRNA target, A6UENDSh, 100-300 pl of 2

mM magnesium buffer containing 50 to 200 nM gA6-14 were injected, at a flow rate of 5

ill/min over the surface of the sensor chip. After each injection of gRNA, buffer was

98

Association Curve Dissoclatlon Curve

”A CYbUSh+NgCYb-558

 

400 1200 2000

 

 

m
It... I 1.2 (10.2) x 10“M"s" k...- 3.0 (11.5) x 10%"
300 mo 1900 am 7360

C muorsso-gum-sso

t

 

   

 

l

Reeponee (RU)
8

 

 

= 5.1 (1:0.8) x 10*4 It‘s“ kd,‘ 2.9 (10.4) x 10‘ e"
1m 2000 3000 eooo 5000

Time (a)
Figure 20. Separate association and dissociation line fits ofthe unedited CYb, A6, and
ND7 Biacore sensograms. The concentration ongNA is indicated on each graph above
each line. A. Sensogram of CYbU + NgCYb558 with first the association line fit and
then the dissociation line fit. B. Sensogram of A6UENDSh + gA6-l4 with first the
association line fit and then the dissociation line fit. C. Sensogram of mND7550 +
gND7-550 with first the association line fit and then the dissociation line fit Each
associationrateconstant Manddissociationratcconstamam) is listcdwiththeermr
in parentheses RU = resonance units and s = seconds.

 

mRNA gRNA Kn (M) k... (M"s") Rods")

 

CYbU NgCYb-558 4.7x 1045 5.7 ($0.9)X 1o+2 2.7 ($0.5) x 10'3
A6UENDSh gA6-l4 2.5x10'9 1.2 ($0.2) x10”4 3.0 (1:15) x10‘5
mND7550 gND7-550 5.6x10'10 5.1 ($0.8) x10“ 2.9 (:04) x10‘5

 

Table 5. Table of rate constants and dissociation constants calculated from CYbU,
A6UENDSh, and mND7550 Biacore data from a global analysis of separate line fits.
Column 1: mRNA, Column 2: gRNA, Column 3: dissociation equilibrium constants,
Column 4: association rate constants, and Column 5: dissociation rate constants from
surface plasmon resonance experiments run at 2 mM magnesium.

100

allowed to flow for 30 minutes at 5 mein for dissociation rate information. A one
minute injection of 8 M Urea was run between every cycle to strip the gRNA from the
mRNA.

Using SPR, the dissociation rate constant (km) was found to be 3.0 x 10'5 s", and the
association rate constant (k...) was 12000 M‘s" (1.2 x 104 M's") (Fig 2013, Table 5).
The equilibrium dissociation constant (KD) at 2 mM magnesium was calculated to be 2.5
nM (2.5 x 10'9 M) (Fig. 18B, Table 5). The A6 mRNA/gRNA complex had ~2000 fold
higher affinity binding than CYbU. The single stranded, unstructured ABS allowed a
~20 fold faster association rate coupled with a dissociation rate being ~20 fold slower
(Fig. 20A, B). When compared to the TAR-TAR" and CopA-CopT RNA-RNA
interactions, the association rate of the A6 mRNA/gRNA complex is eighteen fold
slower. The dissociation rate of A6 is 90 fold slower than these loop-loop interactions
(Persson et aL, 1988; Nair et al., 2000; Nordgren et al., 2001). When studying aptamers
targeting the TAR RNA, the kinetic analysis of mutants showed that the stability of the
complex formation is controlled by the off-rate (km) and not the on-rate (km) (Duconge
et al., 2000). This appears to be true for the A6 mRNA/gRNA complex. This slow
dissociation rate suggests that once the mRNA/gRNA complex forms it may be stable
enough to allow formation of the editosome complex.

The gND7-550/mND7-550 Interaction

For the binding of gND7-550 to its cognate mRNA target, mND7550, 150-250 pl of 2
mM magnesium buffer containing 8.5 to 270 nM gND7-550 were injected, at a flow rate
of 10 mein over the surface of the sensor chip. After each injection of gRNA, buffer

was allowed to flow for 60 minutes at 10 pl/min for dissociation rate information. A one

101

minute injection of 8 M Urea was run between every cycle to strip the gRNA fiom the
mRNA

Using SPR, the dissociation rate constant (kg) was found to be 2.9 x 10'5 s", and the
association rate constant (ken) was 51000 M'ls'l (5.1 x 104 M's") (Fig. 18C, Table 5).
Surprisingly, while the dissociation rate was almost identical to that determined for the
gA6-l4/A6UENDSh interation, the ND7 association rate was ~5 fold faster than A6 and
~90 fold faster than CYbU. The faster association rate resulted in a calculated
equilibrium dissociation constant (KB) of 0.56 nM (5.6 x 1040 M) (Table 5), this was
significantly lower than the K9 calculated for A6 (Table 5).
Discussion

Substantial differences are seen in both association rate and dissociation rates for all
three substrates. The A6 and ND7 mRNA substrates bind with their initiating gRN As
with ~2000 and 8400 fold higher affinity respectively than the CYb mRNA binds its
gRNA. It is interesting that the ND7 substrate binds its gRNA with a five fold faster
association rate than the A6 substrate even though they are both predicted to have single
stranded anchor binding sites. One possibility for this difference could be the G-C
content of the anchor binding sites. The ND7 ABS contains 7 G-C base pairs (Fig 19C),
while the A6 ABS contains only 4 G-C pairs (Fig. 19B). The CYb ABS contains 3 G-C
pairs; however, two of the three residues are already in G-C bps in the mRN A stem-loop
so the gRNA must compete for anchor binding (Fig. 19A, 18A). The G—C rich ABS
appears to be an advantage for ND7, but may be a disadvantage for the CYb stem-loop
resulting in a slower anchor binding association rate. The dissociation rate constants for

A6 and ND7 are ~20 fold slower than the CYb rate. When compared to the kinetically

102

favorable loop-loop interaction, CYb has a similar dissociation rate, while A6 and ND7
have significantly slower dissociation rates. This suggests that in the absence of the
competition of inhibitory RNA structures (such as the CYb stem-loop) the mRNA/gRN A
complex is fairly stable and this stability may allow editing to occur. In addition, the A6
and ND7 mRNA/gRNA complexes have longer anchor helices than CYb. There are
additional bps predicted after the editing bulge that may increase the stability of the
complex that could result in a slower dissociation rate. A baseline study like this gives
tentative insights into how different aspects of the two interacting RNA molecules
contribute to fonnation of the binary complex. The mRN A structure can have a profound
effect on the interaction of the gRNAs and their mRNAs. In addition, the three helix
junction formed by the mRNA/gRN A complex may contribute to the stability of the
complex resulting in the very slow dissociation rate.

RNA-RNA interactions are important regulators and tools in every organism. The
kissing loop-loop interaction is a well characterized RNA-RNA interaction that appears
to occur in all organisms. It is a kinetically favored base pairing intermediate (Chang &
Tinoco, 1997) that appears to have evolved RNA structures optimized for rapid RNA-
RNA interaction (Franch et al., 1999). Binding of the TAT protein to the TAR RNA is
required for HIV replication and results in transcriptional activation. The TAR loop can
be targeted by a complementary hairpin, TAR“, and has a very fast association rate of 105
M's" coupled with a dissociation rate of 10" s". TAR-TAR* binding results in the
inhibition of TAT binding as well as transcription (Nair et aL, 2000). Another well
characterized RN A-RNA interaction is the CopA-CopT antisense RNA that forms a

stable loop-loop intermediate with its target RNA resulting in transcriptional inhibition of

103

the RepA protein; the RepA protein is required for plasmid replication, so CopA-CopT
formation blocks plasmid replication. The CopA-CopT interaction also has a fast
association rate of 105-10‘5 M’s" coupled with a dissociation rate of 10'3 s'1 (Persson et
al., 1988; Nordgren et al., 2001). The TAR-TAR" interaction and the CopA-CopT
binding interaction both result in transcriptional inactivation.

The mRNA/gRN A interaction appears to be different fi'om these loop-loop RNA-
RNA interactions. When compared to the mRN A/ gRN A complexes, the TAR-TAR“ and
CopA-CopT complexes have a 370 fold faster association rate than CYbU, eighteen fold
faster than A6, and four fold faster than ND7. The TAR-TAR and CopA-CopT
complexes dissociate at a rate similar to CYbU and 90 fold faster than A6 and ND7.
Interestingly, the dissociation rate of the complex for A6 and ND7 appears to be
significantly slower than the loop-loop interactions. This suggests that formation of the
anchor helix does not require a fast association rate but higher complex stability to
achieve editing and may have evolved to dissociate slowly to allow editing to begin.

The CYb mRN A is edited preferentially in the insect host. The developmental nature
of CYb editing (Feagin et al., 1988) is one way that trypanosomes regulate the formation
of the respiration complexes when occupying different hosts (Feagin & Stuart, 1988).
One of the ways that editing of CYb is developmentally controlled may be the formation
of a stable mRNA stem-loop that the gRNA must invade in order to bind the mRNA.

The CYb unedited mRN A has a slower kinetic interaction between the mRN A and
gRNA than A6 or ND7. The CYb mRNA/gRN A interaction likely requires a protein co-
factor to initiate the intermolecular base pairing event of the unedited mRNA/gRNA

complex as the binding interaction is very slow and unstable. This appears to be a

104

common requirement in other organisms. In E. coli, a srn-like protein, qu, was able to
improve complex formation between an anti-sense RNA and its mRN A target by
increasing the affinity of the anti-sense RNA for its target ~150 fold (Moller et aL, 2002).
A developmentally regulated protein co-factor such as qu could control CYb editing by
selectively increasing the binding rate of the mRNA to the gRNA whether through
chaperone binding activity (encourage helix formation) or helicase activity (alter RNA
structure). Some mRNA/gRNA pairs already have relatively fast and stable kinetics of
complex formation on their own (such as A6 and ND7) and might be less affected by
regulation of this co-factor during trypanosome mitochondrial development. These
RNAs could continue to be edited; although, the rate of editing might be slightly reduced
in the absence of the initiating co-factor. However, the CYb mRN A editing would be
severely reduced in the absence of this co-factor. There are trypanosome protein co-
factors that are RNA binding proteins and may be involved in editing. One possibility is
RBP16 (Hayman & Read, 1999); it has an RGG RNA-binding motif as well as a cold
shock domain and binds the U-tails of gRNAs with the help of p22 (Pelletier et al., 2000;
Hayman et aL, 2001; Miller & Read, 2003). Another interesting possibility is the MRP1
(Koller et al., 1997) and MRP2 (Blom et al., 2001) protein complex that also binds poly
U tracts and is transiently associated with the proteins of the editing complex by binding
the RNAs for editing (Allen et al., 1998; Lambert et aL, 1999). MRP1 has been shown to
accelerate the anchor formation of annealing mRNAs and gRNAs (Muller et al., 2001)
and is expected to be released once hybridization of the anchors has occurred (Muller &
Goringer, 2002). RNAi knockdown of both RBP16/p22 and MRP1/MRP2 have been

shown to disrupt CYb mRNA editing (Pelletier & Read, 2003; Vondruskova et a1., 2005).

105

It will be interesting to see if these proteins are developmentally regulated in the different
hosts of the life cycle. One or both of these proteins may facilitate the mRNA/gRNA
complex formation and be released once the editosome binds the RNA complex.

Constitutively edited substrates such as A6 and the 5’ end of the ND7 mRN As have an
easily accessible ABS that may encourage editing. A better understanding of gRN A
hybridization to the mRN A ABS and the effects caused by secondary structure may
provide a better understanding of the forrmtion of mRN A/gRN A complexes during the
editing process (Schwille et al., 1996). Currently, there is no data reporting how fast the
rate of editing occurs, so it is unknown if an acceleration of these rate constants is
necessary for in viva editing.

The mND7550+gND7-550 pair has a faster mRNA/gRN A association rate than A6,
but it is not edited in vitro (Reifur et al., 2006). Apparently, the structure of the mRN A
and the ease of the gRNA binding are not the only factors controlling RNA editing
(Reifur et al., 2006). Previous gel shift assays have shown multiple bands of migration of
the mRNA/gRNA complex suggesting multiple conformations (Koslowsky et al., 2004;
Reifur et a1., 2006). In the hepatitis delta virus, the secondary structure of its RNA
controls editing of this RNA by its host. Only RNAs with a branched secondary structure
are edited (Linnstaedt et al., 2006). Perhaps only mRNA/gRNA complexes that adopt the
correct tertiary structure can be recognized by the editing machinery in vitro. Of the
mRNA/gRN A complexes studied to date, it appears that A6 is the only mRNA/gRNA
complex that folds into the correct tertiary complex in the absence of chaperones. So
while editing of CYb appears to be controlled kinetically, editing of other mRNAs such

as ND7 may be controlled structurally.

106

Other RNAs, such as ribozymes, containing three helix junctions similar to the
mRN A/gRN A complex can exist in the inactive extended conformations for long periods
of time and may require regulatory ligands to transition to the correct tertiary structure
(Klostermeier & Millar, 2000). In the same manner, the mRNA/gRNA complex may
require a chaperone to transition to the correct tertiary structure for editing. In addition,
some RNA binding proteins may be required to stabilize the RNA structure; however, in
the absence of proper folding these proteins may exacerbate kinetic traps (Mohr et al.,
2002) instead of stabilizing the correct structure for editing.

Surface plasmon resonance (SPR) is a label free method capable of determining the
kinetics of binding of the mRNA/gRNA complex formation. The equilibrium
dissociation constants (Kn) calculated using BIACORE are similar to those calculated
using gel shift assays for ND7 and A6 mRNA/gRNA complexes (Reifur et al., 2006),
whereas the CYb BIACORE determined K9 is ten fold higher than that previously
reported using gel shift assays (Koslowsky et al., 2004). Previous reports of SPR suggest
that, although SPR works well for some RNA-RNA substrates, the k0,. values measured
for some substrates are lower when compared to other methods (Slagter-Jager, 2003).
This will be discussed in more depth in Chapter four.

In conclusion, editing may be controlled on many different levels. Substrates such as
the CYb mRNA appear to be regulated kinetically with a slow on-rate and a fast off-rate,
while substrates such as ND7 and A6 are not kinetically regulated as shown by the
relatively fast on—rate and very slow off-rate. The slow dissociation rates of the A6 and
ND7 substrates may have evolved to allow the correct editosome complex to assemble on

a stable mRNA/gRN A complex. This study provides information on how the

107

mRNA/gRNA complexes interact; fixture studies using proteins will show how these
interactions differ in the presence of proteins.
MATERIALS AND METHODS
Oligodeoxyribonucleotides
All Oligodeoxyribonucleotides (Table 6) were ordered from Integrated DNA
Technologies, Inc. (Coralville, IA).
RNA Synthesis and Labeling

The templates and sequences for CYbU have been previously described and were
amplified using primers, CYb(-BigSK) and T7CYbShort, listed in Table 6 (Y u &
Koslowsky, 2006). The template for A6UENDSh (5’-
AATTTAATACGACTCACTATAGGAAAGGTTAGGGGGAGGAGAGAAGAA
AGGGAAAGTTGTGATTTTGGAGTTATAGAATAAGATCAAATAAGTTAATAAT
A-3’) was amplified using primers, T7A6short and 3’A6END-4, listed in Table 6.
NgCYb-558 (Leung & Koslowsky, 1999, 2001a, b) as well as gA6-l4 and gND7—550
(Table 4) were synthesized using the Uhlenbeck single-stranded T7 transcription method
(Milligan et al., 1987) with the oligonucleotides described in Table 6. All RNAs were
synthesized by T7 RNA polymerase using a Ribomax Large Scale RNA production kit
(Promega) according to the manufacturer’s directions. RNAs were gel purified as
described previously (Koslowsky et al., 2004). The mND7550 transcript (Table 7) was

purchased from Dharmacon (Boulder, CO).

108

 

 

Name Sequence Length
(nts.)

T7 22mer 5'-AATTTAATACGACTCACTATAG-3 ’ 22

NgCYb558 5 ’-AAAAAAAAAAAAAAATTATTCCCTT 80
TATCACCTAGAAATTCACATTGTCTTTT
AATCCCTATAGTGAGTCGTATTAAATT-3 ’

NgA6- l4 5’-AAAAAAAAAAAAAAATAATTATCATATC 86
ACTGTCAAAATCTGAT'TCGT'TATCGGAGTTA
TAGCCCTATAGTGAGTCGTATTAAATT-3 ’

gND7-550 5 ’-AAAAAAAAAAAAAAATATTCACATTTA 78
TATCATCTTACACTTAATCCACTGCATCCC
TATAGTGAGTCGTATTAAATT-3 ’

T7CYbS hort 5 ’-AAT'1'TAATACGACTCACT ATAGGGT 32
TATAAAT-3 ’

CYb(-BigSK) S’CATCCATATATTCTATATAAACAA 33
CCTGACATT-3’

T7A6short 5’-AATTTAATACGACTCACTATAGGAAA 28
66-3 ’

3’A6END-4 5’-TATTA'1'1‘AACTTATTTGATCITATTCT 37
ATAACTCCAA-3 ’

3’BigSK—biotin 5’-CACTAGTTCTAGAGCGGCC-biotin-3’ 19

CYbbigskBIAbridge 5 ’GCTCTAGAACTAGTGGATCCATATAT 30
TCTA-3 ’

A6bigskBIAbridge 5 ’GCTCIAGAACTAGTGTATTATIAAC 32
TTA'ITI‘G-3 ’

ND7550BigSKBIA 5 ’GCTCTAGAACTAGTGCTGTGCTGAAAC 30

bridge GGG3 ’

Biatagl 5’-ATATGAGTAGGTAA'ITAGTA-biotin-3’ 20

BiataglCYbbridge 5’-ATTACCTACTCATATGATCCATATA 30

TTCTA-3 ’

Table 6. List of Oligodeoxyribonucleotides ordered fiom Integrated DNA Technologies
(Coralville, IA).

109

 

Name Sequence ' Lengt

 

h(nts.L
mND7550 5’-AAAAACAUGACUACAUGAUAAGUACAAGAGG 70
AGACAGACGACAGUGUCCACAGCACCCGUUUCAGCACA
G-3’

 

Table 7. Oligoribonucleotide ordered from Dharmacon (Boulder, CO).
Surface Plasmon Resonance Studies

All Biacore mRNA and gRN A substrates are listed in Table 4. All the bridging
oligonucleotides are listed in Table 6. Measurements of the association and dissociation
rate constants were performed in a BIACORE 2000 instrument (BIACORE, Uppsala,
Sweden). All the solutions used in the binding studies were filtered through a 0.22 pm
polyethersulfone membrane (Coming) or a 0.22 pm Millex-GS membrane (Millipore)
and degassed. The CYbU, A6, and ND7 mRN As were ligated to the BigSK-biotin DNA
tag. CYbU was also ligated to Biatagl , a different biotinylated DNA tag; the
measurements of CYbU were similar regardless of the DNA tag that was attached.
BigSK-biotin or Biatagl was annealed to the 3’ end of the appropriate RNA using DNA
bridging oligonucleotides (Table 6) as described previously (Y 11 & Koslowsky, 2006).
The biotinylated mRN As were gel extracted without phenol and purified using ultra-free
MC membranes (Millipore) and microcon tubes (YM-50, Millipore) according to the
manufacturer’s directions. The gRNAs were transcribed as described above and purified
using Ultra-free MC and the microcon tubes (YM-lO, 30) (Millipore) as above. The
RNA samples were then diluted in running buffer (100 mM Tris pH 7.5, 0.1 mM EDTA,
2 mM MgC12, and 100 mM KCl). The biotinylated mRNA was diluted to 10 nM and
between 50-400 resonance units (RU) of mRNA was attached at 5 pl/min to a

streptavidin coated SA sensor chip (BIACORE, Uppsala, Sweden); the better the

110

mRNA/gRNA interaction the less mRNA was attached. Two cells were immobilized
with mRN A, one was left unmodified to serve as a reference cell, and one cell was
immobilized with the biotinylated DNA tag as a control cell. Binding studies were
carried out running all four cells in series with repetitive cycles of 100-300 pl gRN A
injection at 5-10 mein (association of varying concentrations of gRNA), buffer flow
(dissociation of gRN A) 5—10 mein for 15-60 minutes, and regeneration (50 pl injection
of regeneration buffer, two 50 p1 injections of running buffer at 50p1/min). These
experiments were all run at 27° C. The regeneration buffer for the CYb experiments was
10 mM EDTA or 8 M Urea, the regeneration buffer for the A6 and ND7 experiments was
8 M Urea. The determination of rate constants was performed by fitting theoretical
curves to the experimental curves we obtained using BIAevaluation 3.0 software
(BIACORE, Uppsala, Sweden). The equation used to calculate the association rate
constant was the 1:1 (Langmuir) association formula describing analyte (gRNA) binds to

ligand (mRNA):

k C R — -
a * onc* max ,{1 —e( (ka *Conc +kd)* (t t0)))+ R], ka = association rate

(k a * Cone + kg)—

constant (kw), Conc = molar analyte concentration, R1 = bulk refiactive index effect

(RU), Rmax = Ym = maximum analyte binding capacity (RU), t = time, (0 = time at
start, k d = dissociation rate constant (kofi‘). The equation used to calculate the

dissociation rate constants for A6 and ND7 from the Biacore sensograms was the 1:1

(Langmuir) dissociation formula describing analyte (gRNA) dissociates from surface

complex (mRNA/gRNA complex): R0 It e(‘ kd * (I - (0)) + Offset , R0 = Ymax =

maximum analyte binding capacity (RU), k d = dissociation rate constant, I = time, 10 =

111

time at start, Offset = residual response at infinite time (RU). The equation was slightly

altered for the CYb dissociation rate constant: (R0 — Offset) * e(_ kd * (t _ 10) ) + Offset.

Separate fits for each association and dissociation curve were analyzed globally fiom
each experiment to obtain km. and kofl‘, individually, and the results were averaged. The

dissociation equilibrium constant (K9) was calculated from the averages of the rate

It
constants using the equation: K D 2 2‘1. The errors reported for the rate constants were

0
based on the variances of all curves obtained (Nordgren et a1., 2001).
ACKNOWLEDGEMENTS

This work was supported by National Institutes of Health Grant AI34155 to D.K. We’d
like to thank Larissa Reifur, Dr. Ron Patterson, Dr. John Wang, Dr. Charles Hoogstraten,
and all other members of the MSU RNA Journal Club for critical reading of the
manuscript and helpful discussions. We would like to thank Dr. Joseph Leykam and the
members of the Macromolecular Structure Facility in the Biochemistry and Molecular

Biology Department for the use the Biacore machine.

112

Literature Cited

Allen TE, Heidmann S, Reed R, Myler PJ, Goringer HU, Stuart KD. 1998. Association of
guide RNA binding protein gBP21 with active RNA editing complexes in
Trypanosoma brucei. Mol Cell Biol 18:6014-6022.

Bhat GJ, Koslowsky DJ, Feagin JE, Smiley BL, Stuart K 1990. An extensively edited
mitochondrial transcript in kinetoplastids encodes a protein homologous to
ATPase subunit 6. Cell 61:885-894.

Blom D, Burg J, Breek CK, Speijer D, Muijsers AO, Benne R. 2001. Cloning and
characterization of two guide RN A-binding proteins fi'om mitochondria of
Crithidia fasciculata: gBP27, a novel protein, and gBP29, the orthologue of
Trypanosoma brucei gBP21. Nucleic Acids Res 29:2950-2962.

Blum B, Bakalara N, Simpson L. 1990. A model for RNA editing in kinetoplastid
mitochondria: "guide" RNA molecules transcribed from maxicircle DNA provide
the edited information. Cell 60:189-198.

Blum B, Simpson L. 1990. Guide RNAs in kinetoplastid mitochondria have a
nonencoded 3' oligo(U) tail involved in recognition of the preedited region. Cell
62:391-397.

Chang KY, Tinoco I, Jr. 1997. The structure of an RNA "kissing" hairpin complex of the
HIV TAR hairpin loop and its complement. J Mol Biol 269:52-66.

Corell RA, Feagin JE, Riley GR, Strickland T, Guderian JA, Myler PJ, Stuart K. 1993.
‘ Trypanosoma brucei minicircles encode multiple guide RNAs which can direct
editing of extensively overlapping sequences. Nucleic Acids Res 21 :43 13-4320.

Duconge F, Di Prirno C, Toulrne JJ. 2000. Is a closing "GA pair" a rule for stable loop-
loop RNA complexes? J Biol Chem 275 :21287-2 1294.

Feagin JE, Jasmer DP, Stuart K. 1985. Apocytochrome b and other mitochondrial DNA
sequences are differentially expressed during the life cycle of Trypanosoma
brucei. Nucleic Acids Res I 3 :4577-4596.

F eagin JE, Jasmer DP, Stuart K. 1987. Developmentally regulated addition of nucleotides
within apocytochrome b transcripts in Trypanosoma brucei. Cell 49:337-345.

Feagin JE, Shaw JM, Simpson L, Stuart K 1988. Creation of AUG initiation codons by
addition of uridines within cytochrome b transcripts of kinetoplastids. Proc Natl
Acad Sci U S A 85 :539-543.

Feagin JE, Stuart K 1988. Developmental aspects of uridine addition within
mitochondrial transcripts of Trypanosoma brucei. Mol Cell Biol 8: 1259-1265.

113

Franch T, Petersen M, Wagner EG, Jacobsen JP, Gerdes K. 1999. Antisense RNA
regulation in prokaryotes: rapid RNA/RNA interaction facilitated by a general U-
tum loop structure. J Mol Biol 294:1115-1125.

Hayman ML, Miller MM, Chandler DM, Goulah CC, Read LK. 2001. The trypanosome
homolog of human p32 interacts with RBP16 and stimulates its gRN A binding
activity. Nucleic Acids Res 29:5216-5225.

Hayman ML, Read LK. 1999. Trypanosoma brucei RBP16 is a mitochondrial Y-box
family protein with guide RNA binding activity. J Biol Chem 274:12067-12074.

Katsamba PS, Park S, Laird-Offringa IA. 2002. Kinetic studies of RNA-protein
interactions using surface plasmon resonance. Methods 26:95-104.

Klostermeier D, Millar DP. 2000. Helical junctions as detemrinants for RNA folding:
origin of tertiary structure stability of the hairpin ribozyme. Biochemistry
39: 12970- 12978.

Koller J, Muller UF, Schmid B, Missel A, Kruft V, Stuart K, Goringer HU. 1997.
Trypanosoma brucei gBP21. An arginine-rich mitochondrial protein that binds to
guide RNA with high affinity. J Biol Chem 2 72:3749-3757.

Koslowsky DJ, Bhat GJ, Perrollaz AL, Feagin JE, Stuart K. 1990. The MURF3 gene of
T. brucei contains multiple domains of extensive editing and is homologous to a
subunit of NADH dehydrogenase. Cell 62:901-911.

Koslowsky DJ, Reifur L, Yu LE, Chen W. 2004. Evidence for U-Tail Stabilization of
gRN A/mRN A Interactions in Kinetoplastid RNA Editing. RNA Biology 1:28-34.

Lambert L, Muller UF, Souza AE, Goringer HU. 1999. The involvement of gRNA-
binding protein gBP21 in RNA editing-an in vitro and in vivo analysis. Nucleic
Acids Res 27: 1429-1436.

Leung SS, Koslowsky DJ. 1999. Mapping contacts between gRNA and mRN A in
trypanosome RNA editing. Nucleic Acids Res 27:778-787.

Leung SS, Koslowsky DJ. 2001a. Interactions of mRNAs and gRNAs involved in
trypanosome mitochondrial RNA editing: structure probing of an mRN A bound to
its cognate gRNA. RNA 7:1803-1816.

Leung SS, Koslowsky DJ. 2001b. RNA editing in Trypanosoma brucei: characterization
of gRN A U-tail interactions with partially edited mRN A substrates. Nucleic Acids
Res 29:703-709.

Linnstaedt SD, Kasprzak WK, Shapiro BA, Casey JL. 2006. The role of a metastable
RNA secondary structure in hepatitis delta virus genotype III RNA editing. RNA
12:1521-1533.

114

Miller MM, Read LK. 2003. Trypanosoma brucei: functions of RBP16 cold shock and
RGG domains in macromolecular interactions. Exp Parasitol 105: 140-148.

Milligan JF, Groebe DR, Witherell GW, Uhlenbeck OC. 1987. Oligoribonucleotide
synthesis using T7 RNA polymerase and synthetic DNA templates. Nucleic Acids
Res 15:8783-8798.

Mohr S, Stryker JM, Lambowitz AM. 2002. A DEAD-box protein functions as an ATP—
dependent RNA chaperone in group I intron splicing. Cell 1092769-779.

Moller T, Franch T, Hojrup P, Keene DR, Bachinger HP, Brennan RG, Valentin-Hansen
P. 2002. qu: a bacterial Sm-like protein that mediates RNA-RN A interaction.
Mol Cell 9:23-30.

Moore MJ, Sharp PA. 1992. Site-specific modification of pre-mRNA: the 2'-hydroxyl
groups at the splice sites. Science 256:992-997.

Muller UF, Goringer HU. 2002. Mechanism of the gBP21-mediated RNA/RN A
annealing reaction: matchmaking and charge reduction. Nucleic Acids Res
30:447—455.

Muller UF, Lambert L, Goringer HU. 2001. Annealing of RNA editing substrates
facilitated by guide RNA-binding protein gBP21. Embo J 20: 1394-1404.

Myszka DG. 1997. Kinetic analysis of macromolecular interactions using surface
plasmon resonance biosensors. Curr Opin Biotechnol 8:50-57.

Nair TM, Myszka DG, Davis DR. 2000. Surface plasmon resonance kinetic studies of the
HIV TAR RNA kissing hairpin complex and its stabilization by 2-thiouridine
modification. Nucleic Acids Res 28: 1935-1940.

Nordgren S, Slagter-Jager JG, Wagner GH. 2001. Real time kinetic studies of the
interaction between folded antisense and target RN As using surface plasmon
resonance. J Mol Biol 310:1 125-1 134.

Pelletier M, Miller MM, Read LK. 2000. RNA-binding properties of the mitochondrial
Y-box protein RBP16. Nucleic Acids Res 28:1266-1275.

Pelletier M, Read LK. 2003. RBP16 is a multifunctional gene regulatory protein involved
in editing and stabilization of specific mitochondrial mRN As in Trypanosoma
brucei. RNA 9:457-468.

Persson C, Wagner EG, Nordstrom K. 1988. Control of replication of plasmid R1:
kinetics of in vitro interaction between the antisense RNA, CopA, and its target,
CopT. Embo J 7:3279-3288.

Reifur L, Cruz-Reyes J, vanHartesvelt M, Koslowsky DJ. 2006. Anchor binding site
accessibility and RNA editing in Trypanosoma brucei. manuscript in prep.

115

Reifur L, Koslowsky DJ. 2006. Secondary structure of the ATPase 6 mRN A/gRN A
complex. East Lansing, MI: Personal communication.

Schmid B, Riley GR, Stuart K, Goringer HU. 1995. The secondary structure of guide
RNA molecules from Trypanosoma brucei. Nucleic Acids Res 23:3093-3102.

Schwille P, Oehlenschlager F, Walter NG. 1996. Quantitative hybridization kinetics of
DNA probes to RNA in solution followed by diffusional fluorescence correlation
analysis. Biochemistry 35:10182-10193.

Simpson L, Aphasizhev R, Gao G, Kang X. 2004. Mitochondrial proteins and complexes
in Leishmania and Trypanosoma involved in U-insertion/deletion RNA editing.
RNA 10:159-170.

Slagter-Jager JG. 2003. CopA and CopT: The Perfect RNA Couple. Department of Cell
and Molecular Biology. Uppsala, Sweden: Uppsala University. pp 1-47.

Stuart KD, Schnaufer A, Ernst NL, Panigrahi AK. 2005. Complex management: RNA
editing in trypanosomes. Trends Biochem Sci 30:97-105.

von der Haar T, McCarthy JE. 2003. Studying the assembly of multicomponent protein
and ribonucleoprotein complexes using surface plasmon resonance. Methods
29: 167-174.

Vondruskova E, van den Burg J, Zikova A, Ernst NL, Stuart K, Benne R, Lukes J. 2005.
RNA interference analyses suggest a transcript-specific regulatory role for
mitochondrial RNA-binding proteins MRP1 and MRP2 in RNA editing and other
RNA processing in Trypanosoma brucei. J Biol Chem 280:2429-2438.

Yu LE, Koslowsky DJ. 2006. Interactions of mRNAs and gRNAs involved in
trypanosome mitochondrial RNA editing: structure probing of an gRN A bound to
its cognate mRNA. RNA 12:1050-1060.

116

CHAPTER 4
INTERACTIONS OF MRNAS AND GRNAS INVOLVED IN TRYPANOSOME

MIT OCHONDRIAL RNA EDITING: MEASUREMENTS OF RNA BINDING
RATES FOR A GRNA BOUND TO ITS MRNA AS EDITING PROGRESSES

117

INTRODUCTION

In trypanosomes, many of the protein coding regions of mitochondrial transcripts are
untranslatable until post-transcriptional editing occurs. This process involves >20
proteins, collectively known as the editosome, precisely inserting and deleting uridylates
as directed by a guide RNA (gRNA) template (Blum et al., 1990; Blum & Simpson,
1990; Seiwert & Stuart, 1994; Adler & Hajduk, 1997; Simpson et a1., 2004; Stuart et a1.,
2005). The editing process can extensively alter mRNA transcripts by adding and
deleting hundreds of uridylates and may require several gRN As. Editing creates the
proper open reading frame for mitochondrial transcripts including the correct initiation
and termination codons. The gRNAs that direct the editing process are short RNA
molecules (~50—70 nucleotides) made of three functional elements. At the 5’ end of the
gRNA is a 5-21 nucleotide region known as the anchor that is complementary to the
mRNA just 3’ of the editing domain. The second element, known as the guiding region,
region is complementary to the mature mRN A (allowing G-U base pairs) and serves as
the template for proper editing of the mRNA. The post-transcriptionally added poly-
uridylate tail (~15 nts) is the third element (Blum et a1., 1990; Blum & Simpson, 1990).
It appears to play a versatile role in the editing process.

The gRNA plays key roles in the template directed cleavage, uridylate insertion or
deletion and finally religation of the transcript (Simpson et al., 2004; Stuart et al., 2005).
Therefore, the ability of the gRN As to target their cognate substrates efficiently is an
important part of the editing process. In previous work, different mRNA/gRNA pairs
were used to measure the affinity each initiating gRN A had for its cognate mRNA. One

of these substrates was the unedited apocytochrome b (CYb) mRN A with its initiating

118

gRNA, NgCYb-558 (Koslowsky et al., 2004). Editing of the CYb mRNA is
developmentally regulated occurring preferentially in the insect host (procyclic) (F eagin
et al., 1985). The interaction between NgCYb—558 and its cognate unedited mRNA is
very weak with apparent Kp’s in the pM range. Solution structure probing indicates that
the CYb mRNA forms a stable stem-loop structure, with the anchor binding site (ABS)
sequestered within the double-stranded stem (Leung & Koslowsky, 2001a). We
hypothesize that the regulation of the editing of CYb mRN A is thermodynamically
controlled by the structure of the mRNA. The initiating gRNA, NgCYb558, has to
invade the mRN A secondary structure to bind to the anchor binding site of the mRN A
and almost certainly requires a protein co-factor for efficient interaction. In contrast, for
two other gRNAs which show very high affinity for their cognate targets (gA6- 14 and
gND7-550) no significant secondary structure need be disrupted (Reifirr et a1., 2006).
The structure of the immediate editing domain can profoundly affect the efficiency of the
interaction between the mRN A and gRN A. This interaction between the mRNA/gRNA
complex leads to editing that results in changes of sequence and structure of the
mRNA/gRNA complex.

The structure of the CYb mRN A/gRN A complex has been previously studied.
Solution structure probing of the CYb mRNA/gRNA complex reveals a 3 helical junction
surrounding the first few editing sites (Fig. 21A) that while modified by editing, can be
maintained through the third editing site (Fig. 21B,C) (Y u & Koslowsky, 2006). As
editing increases the length of the anchor duplex, the gRNA stem-loop is maintained by
alternate base pairs that include the U-tail. This results in a decrease in the number of

base-pair interactions between the U-tail and the purine-rich mRN A located upstream of

119

 

A ’
...GCGG E51 9
. I / [Item l \)\“")\\\\t-"’(’5
30¢, II/ c e G e°°\\\ o e
”Uu / / / / ‘4 A \
\ “UAA AA Anchor Helix
U-tall Helix U \ AA u
A a u ‘5 u6
A \
U
3 ‘* u
c ‘ C
AA \ U
A g gRNA atom-loop
"4 c uG
B
“Go
0
000//// £52
”0002/ / / 04‘ 133°
0002/ A G \3
- 4
U tall Helix «I ..... A
‘ U
G ~ c
‘3 ~ 0
AA A gRNA stem-loop
AU G
4606
c Mue“c.u3,
534 um) \ \
Gucu \ \ \ \ 55'
., ...t

/ / /
\ “c U
U AGG
U43" HOIIX U \‘ M0110! HOIIX
U\ u A G"
U
0:4
U G
:G
a s.
‘4 o 4 gRNA stem-loop

4
I u
3WJ/I4‘44AAG”AH \\ ”(,5ch

120

Figure 21. The predicted model for mRNA/gRNA complex secondary structure has
three predicted helices: an mRNA/gRNA anchor helix, a gRNA stem-loop of the guiding
region, and a U-tail/mRNA duplex. A. 5’CYbUT-NgCYb-558 pair with ESl showing
the first editing site is just 5’ (mRNA) of the anchor. B. 5’CYbPES1T-NgCYb-558 pair
with E82 showing the second editing site just 5’ (mRNA) of the anchor. C.
5’CYbPES3T-NgCYb—558 pair with ES4 showing the fourth editing site just 5’ of the
anchor. Notice the anchor has doubled in length.

121

the editing site. In this study, the kinetic effects of these sequence and structural changes
during the initial editing events were investigated. Using both native electrophoretic
mobility shift assays (EMSA) and surface plasmon resonance, the ability of NgCYb-558'
to pair with the unedited CYb substrate (CYbU) with a substrate edited through site one
(ESl, CYbPESl) and a substrate edited through editing site three (ES3, CYbPES3) (Fig.
22) was investigated. Surprisingly, the editing of one site (ESl, the addition of two
uridines) decreases the equilibrium dissociation constant (Kn) significantly. Editing
through site three (CYbPES3, the addition of 4 uridines and a doubling of the size of the
anchor helix) also decreased the dissociation constant. The effects of editing on the
stability of the mRNA/gRNA complex were interesting. Analysis of the binding kinetics
using the surface plasmon resonance data indicate that stability is controlled by the off-
rate (dissociation rate) rather than by the on-rate (association rate). In contrast, data fi'om
the EMSA gels indicate that the binding kinetics is controlled both by the on-rate and by
the off-rate. Surprisingly, kinetic measurements indicate that the absence of the U-tail
results in a two fold slower association rate constant for the gRNA binding the unedited
CYb mRNA using both methods. This results in a two fold decrease in the affinity of
binding (Kn). In addition, while the U-tail played a significant role in stabilizing the
interaction of NgCYb-558 with the unedited substrate, it did not significantly increase the
stability of the partially edited substrates. The U-tail was found to interact with the loop
and bulge region that surround the anchor binding site. This suggests that the U-tail is
helping the CYb gRN A/mRN A complex form by helping disrupt the mRNA stem-loop

through binding one side of the helix that hides the anchor binding site.

122

Figure 22

A

5’CYhUT transcript:
GGGCGAAUUGGGUACCGUUAAGAAUAAUGGUUAUAAAUUUUAUAUAAAAGCGGAGAAAAA

AGAAAGGGUCUUUUAAUGUCAQGUUGUUUAUAUAGAAUAUAUGGauccacuaguucuaga
gcggcc

S’CYbUT fled to NgQXESSS:

GGUUAUAAAUUUUAUAUAAAAGCGGAGAAAAAAGAAAGGGUCUUUUAAUGUCAGGUUGUUUAUAUAGA..

NgCYbSS8AAUAAGGGAAAUAGUGGAUCUUUAAGUGUAACAGAAAAW

B

S’CYbPESIT transcript:
GGGCGAAUUGGGUACCGUUAAGAAUAAUGGUUAUAAAUUUUAUAUAAAAGCGGAGAAAAA

AQAAAGGUUGUCUUUUAAUGUCAGGUUGUUUAUAUAQAAUAUAUGGauccacuaguucua
gagcggcc

S’CYbPESlT aligped to NgQYb-SSS:

GGUUAUAAAUUUUAUAUAAAAGCGGAGAAAAAAGAAAGGUUGUCUUUUIADGUCAGGUUGUUUAUAUAG..

NgCYb- 558- AAUAAGGGAAAUAGUGGAUCUUUAAGUGUAACAGAAAAW

C

5’CYbPES3T transcript:
GGGCGAAUUGGGUACCGUUAAGAAUAAUGGUUAUAAAQgQQAUAUAAAAGCGGAGAAAAA

AGAAAuuuGuGuuGUCUUUUAAUGUCAGGUUGUUUAUAUAQAAUAUAUGGauccacuagu
ucuagagcggcc

5’CYbPES3T aligped to NgQXESS8:

GGUUAUAAAUUUUAUAUAAAAGCGGAGAAAAAAGAAAuuuGuGuuGUCUUUDIADGUCAGGUUGUUUA..

-------------------------------- IIIIIII: ===||||||||||||*'|
mama

NgCYb- 558- UUUUAAUAAGGGAAAUAGUGGAUCUUUAAGUW

D
NgCYb-558 transcript):

5’GGGAUUAAAAGACAAUGUGAAUUUCUAGGUGAUAAAGGGAAUAAUUAUUUUUUUUUUUUUUU3'

NgCYb558sU transcript):

5’GKKHHHHMAAAGACmUHKHKEAAUUUCIUKKHKEMIAAAGGGQJHHUMIUA3’

29 _Y_ 1_)558AnchorComplement:

5’TTGTCTTTTAATCCC3'

& X p558longAnchorComplement:

5’AGAAATTCACATTGTCTTTTAATCCC 3’

gCYb55810ngAnchorC0mplement
5 ' AGAAATTCACATTGTCT‘I‘TTAATCCCB '

N gCYb-558sU :AAUAAGGGAAAUAGUGGAUCUUUAAGUGUAACAGAAAAUUAGGGS '

||||||l||||||||
5'TTGTCTTTTAATCCC3'

gCYb558AnchorComplement

123

Figure 22. The sequences of the RN As used for Chapter 4 experiments. Lowercase, italic letters
= vector sequence, # = mismatched sequence, : = GU base-pair, | = complementary sequence, the
aligned anchors are bold. A. The 5’CYbUT transcript and 5’CYbUT aligned to NgCYb-558 at the
anchors. The CYbU transcript is represented by the underlined portion of S’CYbUT. B. The
5’CYbPESlT transcript and 5’CYbPES 1T aligned to NgCYb-558 at the anchors. The CYbPESl
sequence is represented by the underlined portion of 5’CYbPES 1T. C. The 5’CYbPES3T
transcript and 5’CYbPE83T aligned to NgCYb-558 at the anchors. The CYbPES3 sequence is
represented by the underlined portion of 5’CYbPES3T. Notice that the anchor for 5’CYbUT has
13 base-pairs, the anchor for 5’CYbPES IT has become 16 base-pairs, and the anchor for
5’CYbPES3T has doubled to become 26 base-pairs. D. The NgCYb-558 transcript and the
NgCYb-5585U transcript with no U-tail. The gCYb558AnchorComplement sequence is the
oligonucleotide added to the CYbU and CYbPESl dissociation rate constant gels, while the
gCYb55810ngAnchorC0mplement is the oligonucleotide added to the CYbPES3 dissociation rate
constant gels. The oligonucleotides are aligned with the NgCYb-558 sequence.

124

RESULTS

The binding of the gRNA to mRNA is a filndamental step in RNA editing. From
miRNA studies, it is known that a “seed” sequence composed of ~5 nts at the 5’ end of
the miRNA is responsible for targeting its cognate mRNA for destruction or repression '
(Lewis et al., 2005). Similarly, substrate recognition between the gRNA and mRNA is
mediated by base pairing interactions between the 5’ anchor sequence (~5-21 nts) of the
gRNA and its mRN A (Blum et al., 1990). Understanding the elements that confer both
affinity and specificity on this interaction is critical to understanding the editing process.

In this study, electrophoretic mobility shift assays (EMSA), or gel shift assays, were
used to find the equilibrium dissociation constant (KD) for the binding interaction of the
gRNA with the mRNA during editing. Additionally, EMSA and surface plasmon
resonance were used to discover how the dissociation rate constants (keg) as well as the
association rate constants (ken) change as editing proceeds. Taking a direct measurement
of association and dissociation rates allowed for a more accurate assessment of the
kinetics of binding (Young & Wagner, 1991).

RNA substrates

Three progressively edited CYb partial mRN As composed of the first 88 nts from the
CYb 5’ end were used (Fig. 22). The 5’CYbUT sequence is an unedited substrate and
has been previously described. The 5’CYbPES IT and 5’CYbPES3T substrates were
edited through the first editing site (2 uridylate residues inserted) and the first three
editing sites (a total of 6 uridylates inserted), respectively, and have also been previously
described. The CYbU, CYbPESl, and CYbPES3 substrates were identical to the “T”
substrates except the tag vector sequence had been removed (Fig. 22) (Leung &

Koslowsky, 2001a).

125

In previous work, the structure of the unedited CYb mRNA alone was found to be a
strongly base paired stem-loop with the first three editing sites contained in a terminal 5
nt loop (Fig. 23A) (Leung & Koslowsky, 20013). Editing of the first site was predicted
to add 2 uridylates to the terminal loop (Fig. 23B). Editing of the first three sites
appeared to add an additional 6 uridines to the terminal loop, but otherwise did not
significantly change the structure found within the stem region (Fig. 23C) (Yu &
Koslowsky, 2006). Editing changed the structure of the mRNA. The anchor binding site
(ABS) of the unedited CYb mRNA was composed of 13 base pairs. With the first editing
event, 2 uridines were inserted at the 5’ end of the ABS and were predicted to extend it 3
bps (15 bps total) exposing the three residues in a terminal loop of 9 bps (Fig. 23B).
After the third editing event, the mRNA (5’CYbPES3T) structure changed again and now
had 4 additional uridylates inserted into two more editing sites. The ABS was composed
of 26 bps (doubled in length) and the loop region of 13 bps was single stranded and
available to bind the CYb gRNA (Fig. 23C).

The gRNA construct, NgCYb-558, is 59 nucleotides long including a 15 nt. uridylate
tail (U-tail) and has been described previously (Fig. 22) (Leung & Koslowsky, 2001b).
NgCYb-558 is almost identical to wildtype and directs 21 uridylate insertions at the first
7 editing sites. The effects caused by the absence of the U-tail during the binding
interaction was also investigated using a 44 nt. NgCYb-558sU (no U-tail) gRN A (Leung
& Koslowsky, 2001b).

Equilibrium binding studies
The CYb mRN A and gRN A substrates for the EMSA (electrophoretic mobility shift

assays) experiments were transcribed in vitro. The gRNAs were 5’ end labeled by

126

A 5’CYbUT B 5’CYbPES1T C 5’CYbPES3T
e MUGUG
A G G / 581 G U $3 by
A G @
A-Ua/ANCIHI“ A-UKMCW“ A- 'U
G _@ _ @_ _@ KANGIHJQR
A - U A-U A- -U
A-U A—U A- U
A-U A-u A—U
- - A- 0
AA UAA AA UAA A AA
A 11 A u ‘3 u
GA-uG GA—UG A-UG
GG-CA GG-CA GG-CA
c c G
c-o C-G c—c
c—u c-u G-U
A—U A-U A-U
A-UG A-uG A-UG
A—U A-U A—U
A-U A-U A-U
U-A U-A U-A
A-U A-U A-U
U-A U-A U-A

Figure 23. Predicted secondary structures of the CYb mRN As. A. 5’CYbUT, the
unedited mRN A, forms a stable stem-loop with the first few editing sites positioned
within the terminal loop of 5 base pairs. B. 5’CYbPESlT, a partially edited mRNA, also
forms a stem-loop with the two uridines added at the first editing site enlarging the
terminal loop by three bases. C. 5’CYbPE83T, another partially edited mRNA, forms a
stem-loop with a large 11 base terminal loop that includes the six uridines inserted into
the first three editing sites. The hollow nucleotides represent the anchor binding site
(ABS) where the gRNA anchor binds to form the mRNA/gRNA complex.

127

treatment with calf intestinal phosphatase (Invitrogen) using T4 kinase and [y 32P] rATP,
combined in the indicated concentrations of magnesium (0 mM, 1 mM, 2 mM, and 10
mM Mg”). To determine equilibrium dissociation constants (K0), the gRNAs were
combined with increasing concentrations of unlabeled mRN A substrate (Table 8),
denatured at 70° C for two minutes and slowly cooled to room temperature. The
mRNA/gRNA pairs were allowed to anneal for 3 hours to make sure the mRNA/gRNA
complex formation reached equilibrium. Samples were loaded under current onto 6%
polyacrylamide native gels with magnesium concentrations of the gel and running buffer
matching that of the annealing buffer. The apparent equilibrium dissociation constants
were derived by quantifying the finols/ pl mRN A/gRNA complex formation for a range
of mRN A concentrations from 4 gels. There was a large amount of diffuse signal
between the band shifts of the complex and the free gRNA indicating that a lot of
dissociation was occurring during electrophoresis.

The EMSA experiments for the apparent dissociation rate constants were similar to
the experiments above except that one concentration of mRN A (two times the apparent
KD) and gRN A were incubated for 20 hours at 27’ C. Then an oligonucleotide, ’
complementary to the gRNA anchor, was added and time points were taken at the
indicated times and snap-frozen on dry ice. Each sample was individually thawed and
run as above. The EMSA gels for the apparent observed rate constants also used one
concentration of mRNA and gRN A and were incubated at 27° C for the indicated times,
snap-fiozen, and run as above. The apparent observed rate constant for CYbU was

calculated using a single exponential fit based on the reaction mechanism

1:
mRNA + gRNA if gRN% RNA where the mRNA and gRNA bind to form the
[£017

128

 

 

Lanes mRNA Concentration
1, 2 5’CYbUT 0 nM
3, 4 5 ’CYbUT 75 nM
5, 6 5’CYbUT 125 nM
7, 8 5’CYbUT 250 nM

9, 10 5’CYbUT 500 nM

11, 12 5’CYbUT 750 nM

l3, l4 5’CYbUT 1000 nM

15, 16 5’CYbUT 1500 nM

17, 18 5’CYbUT 2000 nM

19, 20 5’CYbUT 3000 nM
l, 2 5’CYbPES1T 0 nM
3, 4 5’CYbPESlT 15.125 nM
5, 6 5’CYbPESlT 31.25 nM
7, 8 5’CYbPESlT 62.5 nM

9, 10 5’CYbPESlT 125 nM

11, 12 5’CYbPESlT 250 nM

l3, l4 5’CYbPESlT 500 nM

15, 16 5’CYbPESlT 750 nM

17, 18 5’CYbPESlT 1000 nM

19, 20 5’CYbPES1T 2000 nM
1, 2 5’CYbPES3T 0 nM
3, 4 5’CYbPES3T 7.563 nM
5, 6 5’CYbPES3T 15.125 nM
7, 8 5’CYbPES3T 31.25 nM
9, 10 5’CYbPES3T 62.5 nM

11, 12 5’CYbPES3T 125 nM

l3, l4 5’CYbPES3T 250 nM

15, 16 5’CYbPES3T 500 nM

17, 18 5’CYbPES3T 750 nM

19, 20 5’CYbPES3T 1000 nM

 

Table 8. Concentrations of 5’CYbUT, S’CYbPESIT, and 5’CYbPES3T mRNAs used in
the gel shifts to determine the dissociation constant (KO).

129

mRNA/gRNA complex. For the partially edited mRNAs, CYbPESl and CYbPES3, the

best line fit for the data points was a double exponential fit using the reaction mechanism:

It kon2
mRNA“ +gRNA —°—"1—ng~% RN A <—> mRNA + gRNA where the mRNA* represents an
kofl

alternate conformation of the mRN A that allowed a significantly faster association rate
with no noticeable dissociation along with a slower mechanism of binding similar to the
unedited CYb reaction. The first observed rate constant for CYbPESl and CYbPES3
could only be measured by the first data point at one minute, and therefore the first
observed rate constant could not be reliably measured using EMSA. Since the first
binding event was too rapid, only the second, slower observed rate constant was reported
and used to calculate the apparent association rate constant (ken). This first binding event
may have been a helix nucleation event or initial interaction that is not stable enough to
measure using EMSA. The on-rates calculated from the apparent observed rate constant
and apparent dissociation rate constant (kampp) were larger than those calculated from the
K0 app and kompp. However, the apparent observed rate constant gel band shift assays
visually demonstrated the thermodynamic trends of association for the CYb
mRNA/gRNA complex during editing.
Surface Plasmon Resonance Studies

The association and dissociation rate constants were then measured using surface
plasmon resonance (SPR). This real time kinetic technique also allowed measurements
of the binding rate of the mRNA/gRNA complex at different stages of the editing
process. The hybridization kinetic rate constants generated from the Biacore
experiments had much less variation than the EMSA experiments and were probably

more accurate. (The SPR experiments were described more in depth in chapter 3.)

130

In these experiments, a deoxyoligonucleotide tag with a 3’ biotin label was ligated
onto the 3’ ends of the target mRNAs using T4 DNA ligase (Moore & Sharp, 1992).
The biotin labeled mRNA was then immobilized to the streptavidin covered surface of
the SA chip (Biacore, Uppsala, Sweden) in two of the four channels of a BIACORE
2000 (Biacore, Uppsala, Sweden). In order to see reliable gRN A binding to the mRN A,
100 to 600 resonance units of mRN A was attached to the chip surface. One channel
remained empty and was used as a reference surface and one contained the biotinylated
tag as a control for background binding. A continuous flow of gRN A solution at various
concentrations (50 to 7400 nM) was injected over the immobilized mRNAs to monitor
the gRNA association with its target mRNA. The dissociation phase was monitored by
chasing the gRNAs with buffer alone.

The analysis method of the CYbU interactions with NgCYb—558 was described in
chapter 3. The CYbU and CYbPESl dissociation rate constant equation was slightly
altered (see materials and methods) to create a better line fit for the data. The CYbPES3
dissociation rate constant data had a better line fit using the 1:1 (Langmuir) dissociation
formula (see Materials and Methods). In addition, the experiments done without the U-
tail had very high background and should probably be repeated using another method that
can detect slower association rates. The individual rate constants were averaged and the
equilibrium dissociation constant was calculated fiom the rate constants. The errors
reported were based on the variances of all curves obtained (Nordgren et a1., 2001).

The NgCYb-558/5’CYbUT Interaction
Previous experiments involving the unedited CYb mRNA/gRNA complex suggest that

the mRNA structure hinders the mRNA/gRNA interaction (Fig. 23A) (Leung &

131

Koslowsky, 2001a; Koslowsky et a1., 2004). Using EMSA experiments, the binding
kinetics of the unedited CYb mRNA/gRNA interaction was studied at different
concentrations of magnesium. No mRNA/gRN A interaction was detected when no
magnesium was added. At 1 mM magnesium, the unedited mRNA/gRNA interaction did
not come to equilibrium and had a Kn = 0.8 pM with the U-tail. The KD of the
interaction could not be calculated in the absence of the U-tail at 1 mM magnesium.
Using higher concentrations of magnesium, the dissociation constant for the unedited
CYb mRNA/gRNA interaction was ~05 pM when the U-tail was present and ~09 pM
when the U-tail is absent. The presence of the U-tail greatly increased the binding
affinity of the unedited CYb mRNA/gRNA interaction (Koslowsky et al., 2004). These
previous experiments had an incubation time of 45 minutes; however, only the EMSA
experiments using the higher concentrations of magnesium were able to come to
equilibrium. In an effort to understand this reaction at physiological magnesium levels,
these experiments were repeated and allowed to incubate for three hours. When the
unedited mRNA/gRNA complex was allowed to incubate for three hours, there was a
slight amount of complex formation at 0 mM magnesium not observed using shorter
incubation times (Fig. 24A); however, a believable dissociation constant could not be
calculated. The unedited CYb mRNA/gRNA interactions were able to come to
equilibrium with three hours incubation time at 1 mM magnesium and the dissociation
constants were similar to the rates calculated previously for the 2 mM magnesium
interactions (Figs. 24A, 25A). At 10 mM magnesium, the dissociation constants

appeared to be slightly reduced with the longer incubation time (Figs. 24A, 25A).

132

Figure 24

A 5’CYbUT B 5’CYbPES1T

————_ ————-_
OmMMg” GmMMgit ' '

       

”Q’fi0¢..-a¢~-¢..qa.

’ 9. 9.! 10mMMg”

10 mM Mg”

  
 

as it”
17131920 2 4 e 891011121314151617181920

133

Figure 24 continued C 59CYbPEs3-r
———_

0 mM Mg"

*ufléfii
“ ‘u'v7s’sfls’s

. ”Chis:
13111» fl.'

1 mM 1192+

 

«hafififiigfitfi

rrfiefigfiggyioegi,

t" ..w_

«aimiiisa as
Cduhoea-QQ-e... not... .-1“.
10mM Mg2t

aznsaeghfii

  

+U1fi3-i‘ _
1 2 3 4 5 a 7 a 7 1011121314151617181720

Figure 24. Gel band shift assay looking at the CYb gRNA binding increasing
concentrations of the cognate CYb mRN As. Representative autoradiographs of 6%
polyacrylamide native gels are shown with the concentration of magnesium indicated in
the top left-hand comer of each gel. The odd number lanes have 5’ end labeled NgCYb-
558 with a U-tail(+U-tai1), while even numbers have 5’ end labeled NgCYb-558 without
a U-tail (-tail). The positions of the free gRN As are indicated with bold arrows.

A. 5’CYbUT + NgCYb-558, B. 5’CYbPESlT + NgCYb-558, and C. 5CYbPES3T
+NgCYb-558.

134

 

Figure 25 ..._ man an Complex

nM Complex nM Complex on Complex

nM Complex

'- ' Mull III Complex

 

 

A 5’CYbUT B 5’CYbPES1T

0 M

O

..8

2

.3
O
3

O‘NUhflO‘NU‘MO‘NUhMO‘Nth

 

0 2 4 0 0.4 0.8
[CYbUT] (nM) [CYbPES‘IT] (9")

135

 

_... +u¢all III Complex
-' ' Mall nM Complex

Figure 25 continued

 

C 5’CYbPES3T

o
5

1
4321054

 

2
210543210543210

one—3:30 .2: Eva—:00 E: lei—coo 5... 80.32.00 2:

[CYbPES3T] (FM)

136

Figure 25. Graphs showing the analysis of CYb gRNA/mRNA complexes at various
magnesium concentrations. The x axis represents the concentration of mRN A used in
pmols/pl (uM), while the y axis represents the amount of mRNA/gRNA complex
formation in frnols/ul (nM). The solid line indicates complexes containing NgCYb-558
(with U-tail); the dashed line indicates complexes containing NgCYb-SSSSU (no U-tail).
The error bars indicate the standard deviation in complex formation observed between
gels. The dissociation constants (uM) are written below the curves with the calculated
error shown in parentheses A. 5’CYbUT (unedited mRNA) binding NgCYb-558. B.
5’CYbPESlT (partially edited mRNA, first editing site contains 2 uridines) binding
NgCYb-558. C. 5’CYbPES3T (partially edited mRNA, first three editing sites contain 6
uridines) binding NgCYb-558.

137

Calculated dissociation constants were consistent in general with previous experiments
and data (Koslowsky et al., 2004).

Gel shifis were also used to look at the apparent dissociation rate constant (kompp) and
the apparent observed rate constant (kobsapp) for CYbU (Fig. 26A,D and 27A,D, Table 9);
the apparent observed rate constant and the apparent dissociation rate constant were used
to calculate the apparent association rate constant (kon app) (Table 9). These dissociation
gels had 1500 nM CYbU binding to 1 nM gRNA at 27°C for 20 hours at 1 mM
magnesium, then an oligonucleotide, complementary to the short gRNA anchor, was
added at ten times the concentration of the mRNA. Time points were taken and run on a
native gel. The complex dissociation happened quickly with almost all complexes
dissociated by five hours. There was one main band of complex formation, and
dissociation occurred faster in the absence of the U-tail (Fig. 26D). The apparent
dissociation rate constant for CYbU at 1 mM magnesium was ~l x 104 s'l (Table 9) both
with and without the U-tail. Native gels were also used to find the observed rate
constant, so that an association constant could be calculated. Each sample had 1500 nM
CYbU and 1 nM gRN A incubated at 27'C at the time points indicated. A single
exponential equation was used to calculate the CYbU observed rate constant. The
complex associated slowly, and the absence of the U-tail resulted in an even slower
association rate. In addition, the bands of complex formation were very blurry indicating
dissociation was occurring rapidly during electrophoresis. The apparent association rate
constant at 1 mM magnesium was 59 M'ls'l with the U-tail and 38 M'ls’l without the U-

tail (Table 9). The U-tail increased the association rate two fold as observed in the plus

138

ken korr

 

 

 

 

 

 

A D "r“ fi' nut-ryellgo

mm. 1.5 ..M cvsu mm“ 99.9». 1.5 .M mu m

Ilone 1 30 so 12o an 900 15001020 WM“ +olloo 0 1 15I30I00I120I100I240 300mm“

Ii, _l+ _l+ _l+ _l+ _l+ _l+ _l+ _l+ _1+ _1u.uu 1+ _l+ 3+ __1+ _1+ U-hll
" f". - 1.“

;-s

 

 

 

 

E 5,.“ ' é...ollgo
gRNA—AWN!“ WM 500 nM cum; mRNA
" 3° 7° 12° :40 47°19” [15001327mmu' +ongo o 15 so 120 240 no 960 15001020“"‘"“'
'+ -1+ -‘+ -'_+ .'+ .14: .1-0 1+ . 1+ _l+ _1+ _1+ _1+ _1+_+_1+_1+_1+_ n

 

   

C FNA 1 pM complementary long ollgo 7
__ 100 nM CYbPESl! mRNA 100 nM CYbPES3 ,, mRNA

alone 1 H30 00- I120 I240 no oso I1sooI1szoM'W‘“ «Elmo o 30 so 120I 240 no 900 15001620""“‘“‘
lU-Q. (all .14. .l... -l... -lU-hll

   

Figure 26. Electrophoretic mobility shifi assay looking at the binding of the CYb gRNA
to one concentration of the cognate CYb mRNA at 1 mM magnesium in order to
determine the association rate constants and the dissociation rate constants for the
unedited and partially edited mRNAs. Representative autoradiographs of 8%
polyacrylamide native gels. A. Association rate constant gel for CYbU+NgCYb—558. B.
Association rate constant gel for CYbPESl+NgCYb-558. C. Association rate constant
gel ibr CYbPES3+NgCYb-558. D. Dissociation rate constant gel for CYbU+NgCYb-
558. E. Dissociation rate constant gel for CYbPESl+NgCYb558. F. Dissociation rate
constant gel for CYbPES3+NgCYb~558.

139

Figure 27
++UUIIM CID-flu

 

 

 

 

 

 

 

 

 

 

 

 

 

 

      

-¢ 'NeUnlnMCe-pbx
0‘2 +Utaillgm1=1 11(:ll0.07)x1 4s“
5 -U tail 11:0,, =9.99(i1.3)x1d*5 s-1
a l
s \
U \
E 0.1 \
‘ 0
R “LEE
all 50 100 150 200 250 300 350

B CYbPESl E koflCYbPESI
a
'E.
E
O
U
s
E

C CYbPES3
:
'5.
E
O
U
E

   

Time (minutes) Time (minutes)

140

Figure 27. Graphs showing the observed rate constant and the dissociation rate constant
analyses of CYb gRNA/mRNA complexes at 1 mM magnesium. These rate constants
can then be used to calculate the association rate constant. The x-axis represents time in
minutes, while the y-axis represents the amount of mRNA/gRN A complex formation in
fmols/ul (nM). The solid line indicates the curve for the complexes containing NgCYb-
558, while the dashed line indicates the complexes containing NgCYb-558sU (no U-tail).
The error bars indicate the standard deviation in complex formation observed between
gels with the calculated error in apparent dissociation rate constants shown in
parentheses. A. CYbU (unedited mRNA) binding NgCYb-558. B. CYbPESl (partially
edited mRN A, first editing site contains two uridines) binding NgCYb558. C.
CYbPE83 (partially edited mRNA, first three editing sites contain six uridines) binding
NgCYb-558. Notice that the CYbU anchor contains 13 bps, the CYbPESl anchor
contains 16 bps, and the CYbPES3 anchor has doubled in length to 26 bps.

 

 

mRNA gRNA KD.".(M) kc... ..1, (sec") (“gig km ...p(s")
S’CYbUT NgCYb-558 4.6(:l:O.6)x10'7 2.09(i0.5)x10'4 59.1 1.11(¢o.07)x10“
NgCYb558sU l.3(:l:0.l)x10'6 l.62(:l:0.4)x10'4 37.5 9.99(:tl.3)x10'5

S’CYbPESlT NgCYb-SSS l.2(:l:0.2)x10'7 l.58(d=0.3)x10'4 2.47m+2 1.68(io.1)x10‘5
NgCYb-SSSSU l.3(:l:0.3)x10'7 9.55(:l:0.1)x10'5 1.49m+2 l.05(i:0.08)x10'5
5’CYbPES3T NgCYb—558 6.0(:l:0.3)x10'8 5.48(i0.4)x1075 4.15x10+2 l.77(t0.2)x10'5
NgCYb558sU 6.0(:l:0.4)x10'8 5.46(d:0.4)x10'5 5.68x10+2 l.9l(d:0.2)x10'5

 

Table 9. Table of rate constants and dissociation constants calculated from gel band shifi
data. Column 1: mRNA, Column 2: gRNA, Column 3: apparent dissociation equilibrium
constants, Column 4: apparent observed rate constants, Column 5: apparent association
rate constants, and Column 6: apparent dissociation rate constants fi'om gels run at 1 mM
magnesium.

141

U-tail lanes of the association gel shifts that formed more complex than the minus U-tail
lanes (Fig. 26A).

This trend of binding was confirmed by using real time kinetics to find the
dissociation rate constant (km), the association rate constant (km), and then calculating
the dissociation equilibrium constant (Kn) at 2 mM magnesium. Using surface plasmon
resonance (SPR), for CYbU+NgCYb—558 (Table 10) a dissociation equilibrium constant
(KB) of 4.7 uM (4.7 x 1046 M) and 26 uM (2.6 x 10'5 M) without the U—tail (Fig. 28A,
Table 11) were observed. The same trend was observed in the gel shifts where the
presence of the U-tail resulted in a higher afl'mity of binding between the CYb mRN A
and gRN A. These numbers were calculated fiom separate line fits of the association and
dissociation curves from the Biacore sensograms (Fig. 28A). The association rate
constant (k...) for CYbU+NgCYb-558 is 570 M's" (5.7 x 102 M's") and this is two fold
faster than without the U-tail with 240 M's" (2.4 x 102 M's"). The dissociation rate
constant (keg) for CYbU is 2.7 x 10'3 s'1 with and 6.1 x 10'3 8'1 without the U-tail. The U-
tail has been predicted to slow dissociation; interestingly, for the CYb substrate it
increased NgCYb—558 association with CYbU.

The NgCYb558/5’CYbPESlT Interaction

Surprisingly, with the addition of two uridylates fi'om the first editing event (a three bp
increase in the size of the anchor duplex), there was a large increase in mRN A/ gRN A
complex stability at all concentrations of magnesium (Fig. 24B). Similar to the unedited

transcript, the 5’CYbPESlT/NgCYb-558 interaction resulted in the formation of one

142

main complex at all magnesium concentrations (Fig. 243). The major complex formed

was similar in mobility to that formed by the S’CYbUT, unedited substrate.

 

 

Analyte No.
RNA or Sequence of
Ligand nts.

CYbU+BigSK Ligand S’GGUUAUAAAUUUUAUAUAAAAGCGG
AGAAAAAAGAAAGGGUCUUUUAAUGUC
AGGUUGUUUAUAUAGAAUAUAUGGAUC 100
atatggaagtaattagaagga-biotinB’

CYbU+Biatagl Ligand S’GGUUAUAAAUUUUAUAUAAAAGCGGA
GAAAAAAGAAAGGGUCUUUUAAUGUCAG
GUUGUUUAUAUAGAAUAUAUGGAUCcacua 98
guucuagagcggcc-biotin3’

CYbPESl+ Ligand S’GGUUAUAAAUUUUAUAUAAAAGCGGAG

CYbBia AAAAAAGAAAGGUUGUCUUUUAAUGUCAG 83
GUUGUUUAUAUAGaatttataaccggg+biotin3’

CYbPES3+ Ligand S’GGUUAUAAAUUUUAUAUAAAAGCGGAG

CYbBia AAAAAAGAAAUUUGUGUUGUCUUUUAAU 87
GUC AGGUUGUUUAUAUAGaatttataaccggg+biotin3 ’

NgCYb-558 Analyte S’GGGAUUAAAAGACAAUGUGAAUUUCUA
GGUGAUAAAGGGAAUAAUUUUUUUUUUU 59
UUUU3 ’

NgCYb—558SU Analyte 5’GGGAUUAAAAGACAAUGUGAAUUUCUA
GGUGAUAAAGGGAAUAAB’ 44

 

Table 10. List of all Biacore substrates used. The biotinylated DNA tags were ligated on
using DNA bridges listed in Table 8 and are indicated by lowercase letters. Column 1:
RNA name. Column 2: Ligands were biotinylated and attached to streptavidin coated SA

chips (Biacore, Uppsala, Sweden) and Analytes were injected in running buffer and
flowed across the surface of the chip to bind the Ligand. Cohlmn 3: Sequence of the

Biacore substrates. Column 4: number of nucleotides.

143

Figure 28
A Plus U-Tail
Association Curve Dissociation Curve

as CYbUSh+NgCYb-558

 

 

S
5 E
. _
g - .. - _-_.
i
K k... I 5.7 (150.9) x 10“ M434 k0,: 2.7 ($0.5) x 10315-1
no 1200 2000 2300 3000
Time (s)
No U-tail
Association Curve Dissociation Curve
CYbUSh+NgCYb-558sU

 

Response (RU)

 

M: 2.4 (11.3) x 10+2 M‘s" k..- 6.1 ($1.5) x 1038"
* ”M

m 1250 1is—o
Time (s)

Figure 28 continued

   

 

 

    
 

 

 

 

 

 

    

 

 

 

 

3 Plus U-Tail
Association Curve Dissociation Curve
CYbPES1+NgCYb~558
A 77 1372
a ’ .
5 “C 1013 al. \
g 32’ 7 ' _ x
a 18' M
8 C
°‘ ° kt..." 5-6 (no.4) x 10*2M'1r‘ k...- 1.0 ($0.7) x 1040-1
1f50 : 1900 2700
Time (s)
No U-tail
W0" CW0 Dissociation Curve
120 CYbPES‘I'I'NgCYb-SSBSU
g 00
g to ’ J...
8
i: 0 I 7.2 (13.3) x 10" it‘s" k“- 4,9 ($0.6) x 194 ,4
0 1000 2000

 

Time (s)

145

Figure 28 continued

 

 

 

 

 

 

 

 

 

 

 

 

 

C Plus U-‘l'ail
Association Curve Dissociation Curve
CYbPES3+NgCYb-558
§ ‘7 7300 oil 1...; 5 4
g 20 4325 nM__‘ : -__ * fghwfiw _
- 197L195... —WA M
& k0,, =3 3.4 (12.4) x 10” ""8“ . k0,,- 1.4 ($0.9) x 10" s"
‘ 1660 * 5666 a800 3600
Time (s)
No U-tail
Association Curve Dissociation Curve
CYbPESS'I'NgCYb-SSBSU
32" W
5 1 fi‘ 7 ”a... v .. fivfiwwmamw
3,10
, 1
°‘ ‘ k” .1 3,5 x «rm-1.4 kg“: 7.5 x 104:-1
500 1500 2060 2500

Time (s)

146

Figure 28. Separate association and dissociation line fits of the CYb Biacore
sensograms. Each sensogram represents injections of gRNA over a channel with
attached mRNA. The concentration of gRN A is indicated on each graph above each line.
A. Sensogram of CYbU + NgCYb-558 and CYbU + NgCYb-558SU (no U-tail) with first
the association line fit and then the dissociation line fit. B. Sensogram of CYbPESl +
NgCYb-558 and CYbPESl + NgCYb-SSSSU (no U-tail) with first the association line fit
and then the dissociation line fit. C. Sensogram of CYbPE83 + NgCYb—558 and
CYbPES3 + NgCYb-5585U (no U-tail) with first the association line fit and then the
dissociation line fit. Each association rate constant (km) and dissociation rate constant
(keg) is listed with the error in parentheses. RU = resonance units and s = seconds. The
CYbPESl+NgCYb-558SU and CYbPES3+NgCYb-558SU experiments need to be
repeated to verify these results.

 

mRNA gRNA K1, (M) k... (M"s") k.n(s")

 

CYbU NgCYb—558 4.7 x 10*5 5.7 (1:09) x 10+2 2.7 (no.5) x10'3
NgCYb-558sU 2.6 x 10'5 2.4 (:13) x 10+2 6.1 (:15) x 10'3

CYbPESl NgCYb-558 1.8 x 10" 5.6 (no.4) x 10+2 1.0 (£07) x 10‘3
NgCYb—SSSSU 6.9 x 10'7 7.2 (1.3.3) x 10*2 4.9 (too) x 10'3

CYbPES3 NgCYb-558 4.0 x 10'7 3.4 (:24) x 10+2 1.4 ($0.9) x 10"
NgCYb-SSSSU 2.1 x10'7 3.6 x1072 7.5 x10'5

 

Table 11. Table of rate constants and dissociation constants calculated fiom CYb
Biacore data. Column 1: mRNA, Column 2: gRNA, Cohimn 3: dissociation equilibrium
constants, Column 4: association rate constants, and Column 5: dissociation rate
constants from surface plasmon resonance experiments nm at 2 mM magnesium. Only
one experiment was used for the CYbPES3+NgCYb—558SU experiment, so no error is
reported and the experiment should be repeated to verify this result.

147

However, in contrast to the S’CYbUT substrate, increasing the magnesium levels to 10
mM, did not cause a significant change in the appearance or mobility of the main
complex (Fig. 24A, B). At 10 mM magnesium, a second distinct complex of faster
mobility was observed at low mRN A concentrations (Fig. 24B, lanes 3-11). In addition,
with increasing concentrations of magnesium, there was increased complex formation
and lowering of the apparent affinity constant. In the absence of magnesium, the
presence of the U-tail significantly decreased the apparent affinity constant (+U-tail Kn
.p, = 0.17 M (1.7 x 10'7 M), -U-tail Kn.pp = 0.26 1111 (2.6 x 10'7 M)) (Fig. 2513).
Surprisingly however, at 1 mM and 2 mM magnesium, there was no statistically
significant effect of the U-tail on complex formation. The apparent affinity constants (Kn
m, for 1 mM and 2 mM magnesium were 0.12 uM (1.2 x 10'7 M) (Fig. 2513). The KD....
fiirther decreased another two fold to 0.04 uM (4 x 10'8 M) (Fig. ZSB) with 10 mM
magnesium.

Gel shifts were also used to look at the apparent dissociation rate constant (kompp) and
the apparent observed rate constant (kobsapp) for CYbPESl (Fig. 263, E and 27B, B,
Table 9); the apparent observed rate constant and the apparent dissociation rate constant
were used to calculate the apparent association rate constant (kon app) (Table 9) at 1 mM
magnesium. These dissociation gels had 500 nM CYbPESl binding to 1 nM gRNA at
27°C for 20 hours at 1 mM magnesium. An oligonucleotide complementary to the short
gRN A anchor was then added at ten times the concentration of the mRN A. Time points
were taken and run on a native gel. The CYbPESl mRNA/gRNA complex dissociated
more slowly than CYbU requiring 27 hours before most of the complex dissociated.

There were multiple bands of complex forumtion with a minor, weaker band that

148

migrated very slowly and only in the presence of the U-tail. In addition, there appeared
to be equal amounts of dissociation plus and minus the U-tail (Fig. 26E). The apparent
dissociation rate constant for CYbPESl was 1.7 x 10.5 s'1 and without the U-tail it was

1.1 x 10‘5 3'1 (Table 9). The dissociation rate was 10 fold slower than CYbU. Native gels
were also used to find the observed rate constant, and the association rate constant was
calculated. Each sample had 500 nM CYbPESl and 1 nM gRNA incubated at 27°C at
the time points indicated. A double exponential equation was used to extract the
CYbPESl observed rate constant. The CYbPESl mRN A/gRN A complex also associated
slowly, but had clearer bands of complex formation indicating less dissociation was
occurring during electrophoresis. Multiple conformations were forming (~3), but they
were difficult to differentiate. In addition, there was a small fraction of complex
formation that occurred by one minute that appeared to occur with a faster association
rate that represented the first rate constant that was too fast to measure (Fig. 26C). The
apparent association rate constant at 1 mM magnesium was 250 M'ls’l with the U-tail and
150 M"s'l without the U-tail (Table 9). The association rate was slightly faster in the
presence of the U-tail. Also, the CYbPESl association rate was four fold faster than
CYbU.

At 2 mM magnesium, this trend of binding was confirmed using SPR (Fig. 288, Table
11. The dissociation rate constant (keg) with the U-tail was 1.0 x 10‘3 s'1 and 4.9 x 10‘3 3'1
without the U-tail (Fig. 278, Table 11). The association rate constant (ken) was 560 M"s'
‘ (5.6 x 102 M‘s") with the U-tail and 720 M's" (7.2 x 102 M‘s") without the U-tail
(Table 11). There did not appear to be a statistically significant difference in the

association or dissociation rate between plus or minus the U-tail for CYbPESl at 2 mM

149

magnesium. For CYbPESl+NgCYb-558 (Table 10), the dissociation equilibrium
constant (K0) was calculated using the association and dissociation rate constants, and the
K1) is 1.8 uM (1.8 x 1045 M) with and 6.9 uM (6.9 x 1045 M) without the U-tail (Table 11).
The Biacore rate constants showed the same trend as the EMSA derived rate constants,
and the differences between plus and minus the U-tail were not statistically significant
(Table 11). In addition, the CYbPESl-gCYb—SS8 complex had a three fold higher
affinity of binding than CYbU that was the result of a two fold decrease in the
dissociation rate.

It was evident that the first editing event decreased the affinity constant for the CYb
mRNA/gRNA binding, and this may be because the terminal loop now included part of
the ABS and appeared to greatly increase the efficiency of binding. Also, it appeared that
at 2 mM magnesium the three additional base pairs in the anchor helix decreased the
dissociation rate and thereby increased the efficiency of anchor target binding.

The N gCYb-558/5’CYbPES3T Interaction

Additional editing adds four more uridylates into the next two editing sites and
effectively doubled the anchor binding site. EMSA analyses indicated that these
sequence changes resulted in a fithher two fold decrease in the apparent affinity constant
to ~ 0.05 M (5 x 10, M) (Fig. 25C). With the doubling ofthe anchor duplex (26 bps),
stable complexes were observed even in the absence of magnesium and no U-tail
contribution to complex formation was observed. In contrast to S’CYbUT and
S’CYbPBSlT, the addition of magnesium did not significantly increase the apparent
affinity constant between the mRN A and gRNA. In addition, the 5’CYbPES3T substrate

formed multiple complexes of different mobilities at all four magnesium levels tested.

150

There were two slower complexes and a faster complex in all gels; however, in the
absence of magnesium, the higher mobility complex appeared to predominate.

EMSA experiments were also used to find the apparent dissociation rate constant (km)
and the apparent observed rate constant (kobs) for CYbPES3 (Fig. 26C, F and 27C, F,
Table 9); the apparent observed rate constant and the apparent dissociation rate constant
were used to calculate the apparent association rate constant (kw) (Table 9) at 1 mM
magnesium. These dissociation rate gels had 100 nM CYbPES3 binding to 1 nM gRNA
at 27°C for 20 hours at 1 mM magnesium, then an oligonucleotide, complementary to the
long gRNA anchor, was added at ten times the concentration of the mRNA. Time points
were taken and run on a native gel. The dissociation of the CYbPES3 mRNA/gRNA
complex happened very slowly with one main band of complex formation. The
CYbPES3 dissociation rate required a much smaller concentration of mRN A and 27
hours of dissociation time. There was also a minor weaker band observed in the presence
of the U-tail whose migration was much slower (Fig. 26F). CYbPES3+NgCYb—558 had
an apparent dissociation rate constant (kog) of ~1 .8 x 10'5 s’1 with and without the U-taiL
This was similar to the CYbPESl dissociation rate constant and was ~10 fold slower than
CYbU. Native gels were also used to find the observed rate constant, so that an
association constant could be calculated. Each sample had 100 nM CYbPES3 and 1 nM
gRNA incubated at 27°C at the time points indicated. A double exponential equation was
used to calculate the CYbPES3 observed rate constant. The CYbPES3 mRNA/gRN A
complex association occurred slowly; however, a small fiaction associated by one minute
with a faster association that represented the first rate constant that was too fast to

measure. The bands of complex formation were much more distinct and there was much

151

less blurriness indicating even less dissociation was occurring. There were three main
bands of complex formation (Fig. 26C). The apparent association rate constant (ken) was
~500 M‘sl (5.0 x 102 M‘s“) with and without the U-tail. This was two fold faster than
the CYbPESl association rate constant and ~10 fold faster than CYbU (Table 9).

This trend of binding was confirmed using SPR. The dissociation rate constant (keg)
is 1.4 x 10“ s'1 with and 7.5 x 10'5 s" without the U-tail (Fig. 28C, Table 11). This was a
nineteen fold slower dissociation rate than CYbU and seven fold slower than CYbPESl.
The association rate constant (la...) was 340 M‘s" (3.4 x 102 M‘s") with and 360 M‘s"
(3.6 x 102 M's") without the U-tail (Fig 28C, Table 11). There did not appear to be a
statistically significant difference in the association rate between CYbU, CYbPESl and
CYbPES3 according to the SPR data at 2 mM magnesium. The equilibrium dissociation
constant (K1,) was 0.4 M (4.0 x 10'7 M) with and 0.2 11M (2.1 x 10'7 M) without the U-
tail (Table 11). There was a four fold decrease of KD, resulting fiom 3 more uridylates
added to the next editing site, when comparing 5’CYbPESlT to 5’CYbPES3T binding to
NgCYb-558.
Photoaffinity crosslinking and mapping of U5 and U10 of the CYb U-tail

Whereas it has been shown that the mRN A structure influences gRNA target binding,
it was unclear where the U-tail was binding during editing. The EMSA and SPR studies
indicated that the presence of the U—tail resulted in an increased association rate for the
CYbU/NgCYb558 complex.

Photoaffinity crosslinking was one way to observe direct interactions between
molecules such as the editing complex, and incorporation of the crosslinking agents did

not change the nucleotide characteristics enough to interrupt most structures. In an

152

effort to discover where the U-tail was interacting in the CYb mRNA/gRNA complex, a
4-thio-uridine (4sU) was placed at the fifth (U 5) or tenth (U10) position in a U—tail
(Table 12) (Dharmacon, Boulder, CO). The 4-thio uridine moiety is a zero-distance
crosslinking agent. The U-tails were 5' end labeled and ligated to NgCYb-558sU (no U-
tail) using an oligonucleotide bridge (gCYb558(sU) bridge, Table 13) to join the two
halves (Leung & Koslowsky, 2001b). P-azido—phenacyl bromide (APA) was added to
half of the sample to increase the reach of the crosslink upon irradiation by converting
the azido group into a nitrene that crosslinks to nearby residues (has a reach of up to 9 A
away). The gRNAs were annealed to either 5’CYbUT or 5’CYbPES3T and then the
mRNA/gRNA complex was UV crosslinked. The crosslinks were isolated on a 6%
polyacrylamide gel (Fig. 29). Positions of the crosslinks were then mapped by using
primer extension analysis with reverse transcriptase (RT) and RNase H analysis (Fig.
30,31).

There were two crosslinked bands with different migration patterns in the gel for all
four pairs of gRNA/mRNA complex and with either U-tail that were labeled band 1 and 2
(Fig. 29). The reactions without APA had a lower percentage of crosslinking. With
APA, there was a much higher percentage for the slowest crosslink band, which was
labeled band 1 (B1), than for the faster band 2 (B2). The control lanes consisted of a no
UV control and a no mRNA control. The only crosslinks of interest were mRNA/gRNA
interactions.

For primer extension analysis (Fig 30A), a 5’ end labeled primer (BigSK, Table 13)
was annealed to 2-5ng of crosslink. AMV reverse transcriptase (RT) (Seikugaku) was

used to extend the primer along the mRNA (Leung & Koslowsky, 1999). When the RT

153

 

Name Sequence Length(nts.)
UlS-thioUS 5’-UUUUUUUUU—4sU-UUUUU 15
UlS-thioUlO 5’-UUUU-4sU-UUUUUUUUUU 15

 

 

Table 12. Oligoribonucleotides ordered from Dharmacon 4sU = 4-thio-uridine.

 

 

Name Sequence Length(nts.)
T7 22mer 5'-AATTTAATACGACTCACTATAG—3’ 22
NgCYb558(sU) 5 ’—'1'TA'1TCCC1'1'1'ATCACCTAGAAAT 65

TCACATTGTCI I l IAATCCCTATAGT
GAGTCGTATI‘AAATT-T

NgCYb-558 5’-AAAAAAAAAAAAAAATTATTCCCIT 80
TATCACCTAGAAATTCACATTGTCTTIT
AATCCCTATAGTGAGTCGTATTAAATT-3 ’

T7CYbShort 5’-AATT1‘AATACGACTCACTATAGGGT 32
TATAAAT-3’

CYb(-BigSK) 5’-GATCCATATATTCTATATAAACAA 33
CCTGACATT-3’

CYbU(-) S'CTATATAAACAACCTGACATTAAAAG 30
ACCC-3'

3’BigSK-biotin 5’CACTAGTTCTAGAGCGGCC-biotin—3’ 19

CYbbigskBIAbridge 5’-GCTCTAGAACTAGTGGATCCATATAT 30
TCTA-3’

Biatagl 5’-ATATGAGTAGGTAATTAGTA—biotin- 20
3’

BiataglCYbbridge 5’-ATTACCTACTCATATGATCCATAT 3o
ATTCTA-3’

BigSK 5’-GGCCGCTCTAGAACTAGTGG-3’ 20

gCYb558(sU) bridge 5’-AAAAAAAAAAAAAAATTATTCCC 32
TTTATCACC-3’

CYb-l 5’-AT'I'TATAACC-3’ 10

New CYbH-Z 5’-ACCATTATTCT—3’ ll

5’CYbH-1 5’-CAACCTGACATT-3’ 12

 

Table 13. List of Oligodeoxyribonucleotides ordered fi'om Integrated DNA Technologies.

154

' 29
“8“” A cvo mRNAs xllnked to NgCYb-558U5
gCYbUT S'CYDUT‘P 5'CYbPES31’ 5’CYI1PE33T+

No N: NONo Nouo
(Liam x1Cuv9§uACuvl§mCuv

 
  
 

 

 

03% *9‘ 631° 0.5% 2 85 / BM 1
. , " ~ 0
“-29% it! 1.16% . 1.94% 2‘“ / W 2
1i
NgCYb-‘Ffi ANCHOR uuuu4suuuuuuuuuuu

I-B CYb mans xilnked to NQCWW
G’CYWT VOW“ G'CYDPESST mm-
»A APA

    
  

No No No No No No No No
uv g5 Cruv W W
0.37% Q “-671

0691,“ 1.6% .

 

NgCYb—“R ANCHOR UUUUUUUUU4sUUUUUU

155

Figure 29. The crosslinks of 5’CYbUT+NgCYb-558 and 5’CYbPES3T+NgCYb-558
run out on gels. A. 5’CYbUT+NgCYb-558 and 5’CYbPES3T+NgCYb—558 crosslinks
with the 4-thio U in the fifth position. A diagram of the gRNA is below the gel B.
5’CYbUT+NgCYb—558 and 5’CYbPES3T+NgCYb-558 crosslinks with the 4-thio U in
the tenth position. A diagram of the gRNA is below the geL Notice that each set of
crosslinks has a No UV control and a no gRNA control for each crosslinking reaction. C
= crosslink lane. There is a middle crosslink that appears to be mRN A specific that is
faintly visible in the no gRNA control. Bands 1 and 2 are indicated with arrows. The
percentage of crosslink is indicated next to each crosslink band.

156

Figure 30 Primer

3‘ ANCHOR E
A MRNA 51_ _ . * 3’

i3

 

 

 

: ' 5’cYbUT flog.

 

ibis

”E
z

1': .3 ,
W
«”9
01m
“1
1“

mil- 2
i

15

s]

 

on as

w v mm:-

 

 

158

 

Figure 30 continued

 

 

 

 

159

Figure 30 continued

 

 

E “MAM“

on
on ‘W‘

 

ants, 1
3' ., Aw , ,

160

Figure 30. Primer extension analysis of the crosslink bands. A. Diagram of the primer
extension analysis. The primer represents the BigSK primer extended by reverse
transcriptase. The stars represent covalent crosslinks between the U-tail of the gRNA and
the mRN A where the reverse transcriptase will fall off and leave a hard stop on the
primer extension gel B. Primer extension analysis of 5’CYbUT+NgCYb-558 U5
crosslink. C. Primer extension analysis of 5’CYbPES3T+NgCYb-558 U5 crosslink. D.
Primer extension analysis of 5’CYbUT+NgCYb-558 U10 crosslink. E. Primer extension
analysis of 5’CYbPES3T+NgCYb-558 U10 crosslink Bl = band 1, B2 = B2, ABl =
APA+band1, AB2 = APA+band2. The dideoxy sequencing lanes are represented by
ddC, ddA, ddT, and ddG. S’CYbUT alone and 5’CYbPES3T alone are control lanes
showing primer extension of the 5’CYbUT or 5’CYbPES3T mRNA alone. A bar is
placed vertically to represent the anchor binding site on the mRN A. Stars on the gels
represent hard stops where the reverse transcriptase encountered a crosslink. The
5’CYbPES3T gels have a hard stop in the anchor region, but this stop is believed to be
caused by secondary structural elements of the mRNA/ gRN A anchor sequence
interaction. This strong binding causes the strong stop, because this stop is the same for
the U5 and U10 crosslinks. The RT falls off the template one nucleotide (3’) before the
crosslink.

161

 

Figure 31 A

BMW“

Mflmmbormoemcum
an“ m

 

9* '41s '7'?“

 

fits. names

c Wflwciunm

+1”. 2 land 2 “Id 2

Boos-MA C—MMA cm

”é .s‘his .* ‘5'!" " am ..1.
$11 , " a. I.»

g“

3%

  

Figure 31. RNase H analysis of the U10 crosslinks. A. Diagram of the oligonucleotides
used to check the position of the crosslink. The oligonucleotide binds the crosslinked
molecules and RNase H digests the DNA/RNA hybrid. The position of the crosslink can
be placed based on the digestion pattern of two oligonucleotides separately and together.
Oligonucleotide A = New CYbH-Z, Oligonucleotide B = CYb-l, and Oligonucleotide C
= 5’CYbH-l. B. RNase H analysis of the U10 band 1 crosslinks. Notice that digestion
with both oligonucleotides results in the smallest product suggesting that the crosslink
position of band 1 is between the two oligonucleotides. C. RNase H analysis of the U10
band 2 crosslinks. Notice that digestion with oligonucleotide A and with both
oligonucleotides results in the same size product suggesting the position of the band 2
product is before oligonucleotide A. Note: There was not enough of S’CYbUT band 1
without APA for these experiments, so that sample was excluded. Also, sometimes the
oligonucleotide/mRNA hybrids are not fillly digested by RNase H, so when both
oligonucleotides are used there may be extra bands fi'om partial digestion from one or the
other oligonucleotide. The majority of the product should be a double digestion.

163

hit the crosslink, it fell off the template. This created a strong stop one nucleotide (3’)
before the crosslink that was observed when the primer extension analysis was run on a
denaturing polyacrylamide gel beside a sequencing reaction of the mRN A.

The main crosslinks for 5’CYbUT/NgCYb-558U5 Bl included A60-G66 (Fig. 3GB,
323); this was also true for 5’CYbPES3T/NgCYb—558U5 Bl (now A60-G69) (Fig. 30C,
32E). The APA group seemed to increase the spread of crosslinks by one or two
nucleotides for BI U5 substrates, but there were still crosslinks near the first few editing
sites. S’CYbUT/NgCYb-558U5 for B2 had a crosslink spread mainly of A20-U25
(+APA=G22-U25, -APA=A20-A23) (Fig. 29B, 323); the same was true for
5’CYbPES3T/NgCYb—558U5 the crosslinks included A20-A35 (Fig. 30C, 32E). APA
again increased the spread of the crosslinks.

The crosslinks from 5’CYbUT/N gCYb-558U10 for Bl without APA included A42-
A44, with APA C51-A54 (Fig. 30D, 32C) were observed. 5’CYbPES3T/NgCYb—
558U10 Bl crosslinks were one to two nucleotides upstream of 5’CYbUT/NgCYb—
558U10 B1 crosslinks (-APA U43-U45) (+APA GSO-G53) (Fig. 30E,32F). Crosslinks
for 5’CYbUT/N gCYb-558U10 B2 included G17-U18 without APA and G17-A20 with
APA (Fig. 30D,32C); 5’CYbPES3T/NgCYb-558U10 BZ appeared to once again have the
same spread but a few nucleotides upstream (-APA Cl6-Gl7) (+APA C1 5-U19) (Fig.
30E,32F).

When there was more than one stop, there may have been multiple crosslinks
indicating secondary and tertiary interactions. There could have been multiple

interactions caused by structural breathing. It appeared that the U-tail was binding and

164

Figure 32

 

 

A 5’CYbUT B 5’CYbUT+NgCYb-558U5
a i l
‘3 A G G 4UAUUG" 2°
. -¢A65 G IIAlllfilfl
i —-«>A-u ‘8 ’2 3° mi ‘°
3 -"°'© ”00
”A‘U 40040
10A_U 4" 3'
5:3,, He. 1flu } o if;
A 0 5444474015 56 “(so \\ e
A a
U 3 U/U 67°00 \\\ “0’6
:GA-UG 944 A65 66656 ,e “I?
.. :cG-c“ 5° UUUUUA g 610
‘5‘ c-(;6 AU 20 Au
'5 soc—u A \AAG UG
2 A-U G 404 : U
1: A—U c \ J!
2 A-U G \ c
3 ”A—U AG \ U
-:>U-A g A
wA-U U GG
U—A 4G u”
C 5’CYbUT+NgCYb-558U1O
20 u
44 uauugocununcmuueccAuccguuwicccccs'
4
”a
30404”
Wgcfi 608:3
oil/[4: °°°0€\ \655
3' ”Oval/é; 00,644 A“ 6666 %°°\\;;F;\>°F°
\SPGW
U A
A A
AU\4A2;UeU 53am! 1
403 : 1‘1", 1 +APA Band 1
G \
44G \ ((1: Enand 2
4U 3 fl +APA Band 2
Acus‘:

165

Figure 32 continued

D 5’CYbPES3T

US crosslinks

1.110 crosslinks

 

E

5’CYbPES3T+NgCYb-558U5

51518388 38

u”%% \
11 U ., u UgoGUMUMGAgiouucccaucccuuuccccc5-
—>A @U ‘60 1°
jig-3% £1900 .
- d A -ii “7’0 / 600%
A [U] 4" at" \
soA- u 1'“ “00 \\\ 0
AA- UAA “PO °°o\\\\y
A e i at)" \\ or“ 10
U 04 I i d) \\
"WA-MG / ”We °°
—>G—(C (’00 4 p0 \\
"G 3 002/ is “309° 2"
-sc-c o c
A- U c 50000 4,, sou
A- U o 00\\ ‘70 A6
A - U 0 \
A — u o z ‘0
:5 U - A 4 4
9L “A- U
—9U - A
F 5’CYbPES3T+NgCYb-558U10
“access
0 uccununeAgol-IUGCCAUGGSW
2 ’° 3’3
0‘5
(I ‘36 \
‘° 40,?! 0079009“
”‘4 4 o°°c\,‘\\;c\’ 10
59“ "I 0°°\\\ °
/ \ .949
0062 /4 Ari’xgsti“ 3"
80
0(2’ $5 6 a Band 1
o \
“00¢ 44 so: l-i-APA Bend 1
000 00 4a A
0” \4 40 8, Bend 2
(’4 40 fl +APA Bend 2

166

Figure 32. Diagram of the U5 and U10 crosslinks on the mRNA alone and the
mRNA/gRNA complex. A. 5’CYbUT alone with the band 1 U5 and U10 crosslink
positions indicated by vertical bars. Notice that the U-tail appears to bind the loop and
bulge regions. B. 5’CYbUT+NgCYb-558 U5 crosslinks. C. 5’CYbUT+NgCYb-558
U10 crosslinks. D. 5’CYbPES3T alone with the band 1 U5 and U10 crosslink positions
indicated by vertical bars. E. 5’CYbPES3T+NgCYb-558 U5 crosslinks. F.
5’CYbPES3T+NgCYb-558 U10 crosslinks. Arrows indicate the bases where covalent
crosslinks are detected. The thickness of the arrow indicates intensity of the RT band.

167

dissociating from the purine rich upstream area of the mRN A that base pairs with the
anchor binding site.

RNase H analysis was used to verify the primer extension results and showed that
stops were caused by crosslinks and not structural obstacles or other artifacts. Twenty
pmols of DNA oligonucleotides (Table 13) were used to surround the crosslinks that
were observed on the RT analyses (Fig. 31). RNase H (Takara) was used to digest these
areas of RNA duplexed to DNA. DNA oligonucleotides were selected to bracket 3
crosslink band, and when the digest was run on a gel, a migration pattern was observed
that ran faster when the oligos bind closer to the crosslink. If multiple digested bands
were observed, there was more than one crosslink. The RNaseH results verified the
mRNA/gRNA crosslinks (example shown in Fig. 31).

In comparing crosslinks using 5’CYbUT, the unedited transcript, and 5’CYbPES3T,
the partially edited transcript, the same U-tail interactions were found even as editing
progressed supporting the previous work done with a 3’ terminal crosslinker (Leung &
Koslowsky, 1999). The Bl crosslinks were very interesting. The fifth residue of the U-
tail was binding the loop of the stem-loop that was on one side of the anchor binding site
(G34-G38) (Fig. 32A, D). The tenthU residue was binding the bulge region on the other
side of the anchor binding site (C23-A26) (Fig. 32A, D). The U5 crosslinks appeared to
be 5-10 nucleotides downstream of the U10 crosslinks; this made sense as the
crosslinking agent was moved downstream too. The U5 crosslinks appeared to have a
wider spread of crosslinks than the U10 tail; this correlated well with the secondary
structure predictions as the U5 crosslinking agent was much closer to the sites for

editing and this area was expected to be more dynamic or have more activity and

168

movement. Kinetic binding studies showed that the U-tail was helping the gRNA to
bind the unedited CYb mRNA. These Bl crosslinking positions indicated that the U-
tail was binding the bulge regions surrounding the anchor binding site of the unedited
mRNA. The U-tail appeared to be helping disrupt the stem-loop of the CYb mRN A by
base pairing with the complement to the ABS. This U-tail action may be critical for
CYb editing as down regulation of the protein that adds U-tails to the gRNAs resulted in
ablation of CYb editing (N ebohacova et al., 2004). However, the presence of
mitochondrial proteins such as a helicase or annealing factor may alter this U-tail
activity.

The BZ crosslinks were less easy to understand and these crosslinks may represent
tertiary interactions with down stream mRNA that occurred after the mRNA/gRNA
anchor helix is fillly base-paired. However, these crosslinks were interacting with vector
sequence on the mRN A, and these crosslinks could also be experimental artifacts.

These mRNA/gRNA crosslinks appeared to have a more compact structure as the BZ
crosslinks were the faster crosslink band in the gel (Fig. 29).
DISCUSSION

One of the interesting results of these experiments is that the U-tail appears to increase
the association rate of the gRNA with the unedited CYb mRN A two fold in the absence
of protein co-factors. This trend is observed both in the EMSA gels and in the Biacore
data. The crosslinking data shows that the U-tail is binding and releasing the nts of the
bulge region and loop region surrounding the anchor binding site. The U-tail may be
acting as an evolutionary molecular tool that assists in opening mRN A stem-loops to

allow anchor helices to form. The CYb anchor binding site sequence is U/C rich, so the

169

U-tail will not bind the anchor binding site. The sequence base paired to the anchor
binding site is NO rich, so the U-tail can bind this sequence easily. This makes the U-
tail the perfect tool to help the gRN A invade the stem-loop to bind the ABS. Also, the A-
U and G-U base pairs are weaker interactions with only 2 hydrogen bonds holding the
base pairing interaction together; perhaps this makes the U-tail more flexrble as it can
make and then break these interactions as needed to maintain structures during editing
such as the gRNA stem-loop. However, the U-tail may behave differently in the presence
. of proteins such as helicases or annealing factors that affect mRNA/gRNA binding.

The effects of editing on the stability of the mRN A/gRN A complex are interesting.
The introduction of two uridylates into the first editing site causes a significant decrease
in the affinity constant (Kn app) (Fig. 253) of the S’CYbPESlT/NgCYb-558 complex.
When binding a single stranded region versus loop regions, the efficiency of the
interaction depends on RNA structure and accessibility to the target sequence (Lima et
al., 1992). The ABS helix must be disrupted in order for gRNA binding to occur (Franch
et a1, 1999) so that editing can begin. In the unedited mRNA, disruption of the ABS
helix region is unfavorable as only four of the twelve nts on the ABS are accessible for
anchor target binding, but this increases to seven accessible nts after one editing event.
The next two editing events add four additional uridylates to the terminal loop (Fig. 23C)
and again lower the affinity constant (Fig. 25C). However, the gRNA forms a stable
stem-loop structure that is composed of a section of this CYbPES3 anchor. In fact, some
of the newly single stranded anchor complement in the gRNA is part of the gRNA stem-
loop and is unavailable to bind the newly single stranded ABS. The

5’CYbPES3T/NgCYb—558 complex has the lowest dissociation constant, and this may be

170

because the ABS target site is the most accessible with the least disruption of the target
structure necessary for binding (Lima et al., 1992). However, the secondary structure of
the gRNA may compete with the CYbPES3 anchor target binding resulting in no increase
of association rate or possibly a cancellation of the effect of an increased ABS size.
However, once the gRNA binds the interaction is more stable, because the anchor helix
length has doubled. This effect is shown in the slower dissociation rate for the
CYbPES3-gCYb-558 complex. Additionally, there are differences at 1 mM magnesium
versus 2 mM magnesium. The increased magnesium may increase the stability of the
mRN A and gRNA stem-loops resulting in a smaller association rate of the mRNA
binding the gRNA despite the longer ABS. This sequestering of the gRNA long anchor
(complement to partially edited ABS) may be an evolutionary gRNA strategy that keeps
free gRNAs fiom competing with the gRNAs bound to the mRNAs once editing has
begun.

The unedited CYb mRNA/gRNA interaction appears to be regulated kinetically and
requires a chaperone or RNA annealing factor for gRN A target binding. However, such a
protein does not co-purify with the main editing complex responsible for RNA editing
(Simpson et aL, 2004; Stuart et a1., 2005). This implies that once editing has begun, this
CYb initiating co-factor may be released from the complex. The fact that the very first
editing event improves the binding interaction of the CYb mRN A/gRN A complex may
be important for developmental regulation. With editing of the first site, the barrier to
begin editing is greatly reduced. Once editing begins with the help of the cofactor, the

mRNA/gRNA pair may need to be able to proceed through the editing process in the

171

absence of a co-factor. The kinetics of the interaction improve with each editing event;
this may be an important part of editing progression.

An interesting result of the gel shift assays are the multiple bands of different
mobility. In the presence of magnesium there appear to be alternate conformations of the
mRNA/gRNA complex that are kinetically trapped into stable intermediates. While
complexes of more than one mRNA/gRNA cannot be ruled out, it is unlikely as there is
no sequence complementarity in the rest of the molecules. It is more likely that different
band shift mobilities represent differing amounts of base pair formation along with target
structure disruption (Lima et al., 1992). Increasing the magnesium concentration may
stabilize multiple folding intermediates of the mRNA/gRNA complex, and these alternate
structures of the complex would travel through the gel with different mobilities.

One puzzling aspect of these experiments is that the SPR obtained rate constants show
slightly different results than the gel shift obtained rate constants (T able 14). The SPR
rate constants show less variation as editing increases versus the gel shift data. The
presence of the U-tail increases the association rate of the unedited CYb complex by two V
fold using both methods. Also, the rate constants from the gel shift data show a faster
association rate as editing progresses along with a ten fold decrease in dissociation only
after the first editing event. However, the SPR data shows a progressive decrease in
dissociation rate with each editing event with no significant increase of association rate
observed with increased editing (Table 14). The dissociation constants (Kn) for the gel
shift data show the same trend as the Biacore data; however, the gel shift Ko’s range from
10*5 to 10‘“, while the Biacore Kp’s range fi'om 10'5 to 10'7 (Table 14). While the trends

of binding are the same, the numbers are not. Another difference is that only a slower

172

Table 14. Comparison between gel shift data and SPR data. Column 1: mRN A, Column
2: gRNA, Column 3: dissociation equilibrium constants, Column 4: association rate
constants, Column 5: dissociation rate constants.

 

mRNA gRNA Kn...(M) (331;) 1.... ...0")
S’CYbUT NgCYb-558 4.6(a0.6)xlo'7 59.1 1.11(:l:0.07)x10'4
NgCYb-SSSsU 1.3(:l:0.1)x10'6 37.5 9.99(al.3)xlO'5
S’CYbPESlT NgCYb-558 1.2(:l=0.2)xlO'7 2.47x10“2 1.611(10.1)1110'5
NgCYb-558sU 1.3(i0.3)x10'7 1.49x10+2 1.05(:l:0.08)x10'5
5’CYbPES3T NgCYb-558 6.0(i0.3)x10”8 4.15xlo+2 1.77(i0.2)x10'5
NgCYb-558sU 6.0(i0.4)x10'8 5.681110+2 l.9l(:i:0.2)x10'5

 

Table of rate constants and dissociation constants calculated from gel band shift data.
from gels run at 1 mM magnesium.

 

mRNA gRNA Kn (M) k... (M484) ken (8'1)

CYbU NgCYb-558 4.7 x 10*5 5.7 (:09) x 10"2 2.7 (50.5) x 10'3
NgCYb-558sU 2.6 x 10'5 2.4 (11.3) x 10+2 6.1 ($1.5) x 10'3

CYbPESl NgCYb-558 1.8 x 10‘ 5.6 (20.4) x 10+2 1.0 (10.7) x 10'3
NgCYb-558sU 6.9 x 10'7 7.2 (13.3) x 10“2 4.9 (:06) x 10‘3

CYbPES3 NgCYb-558 4.0 x 10'7 3.4 (a2.4) x 10+2 1.4 (10.9) x 10“
gemssssu 2.111107 3.6 x10+2 7.5 x10’5

 

Table of rate constants and dissociation constants calculated from CYb Biacore data run
at 2 mM magnesium. Only one experiment was used for the CYbPES3+NgCYb-558sU
experiment, so no error is reported and the experiment should be repeated to verify this
result.

173

*4.“—
l

observed rate constant was detectable using the gel shifts, while the Biacore experiments
did not detect or separate two rate constants. However, there is also evidence that
hybridization kinetics using techniques such as SPR where one molecule is bound can
result in altered hybridization kinetics when compared to experiments done in solution
(Sekar et al., 2005). In addition, immobilization of the mRNA may alter the
thermodynamics of mRNA/gRNA complex formation by constraining the molecule’s
rotational and diffusional freedom (Nair et al., 2000). Also, when comparing different
RNA-RNA interactions, some interactions measured using SPR have lower association
rate constants when compared to other methods measuring the same interaction (S lagter-
Jager, 2003). The CYb association rate constant is at the detection limit of the Biacore
machine; however, this range may be extended under favorable conditions (Myszka,
1997). This made these experiments difficult requiring very high concentrations of
gRNA in order to observe binding causing possible mass transport problems. However,
the variation between experiments using SPR was much lower than with the gel shift
method indicating that the surface plasmon resonance method is more accurate and
precise than the gel shift method. The gel shift method is a powerful tool to observe
binding interactions, but the variation between experiments is relatively high. In
addition, in order for the gel shifts to come to equilibrium in a reasonable amount of time,
the mRNA and gRNA had to be heated and cooled slowly together. This method of
annealing probably acts as a chaperone protein such as a helicase that would allow the
gRNA to bind the mRN A much easier when the ABS mRN A stem-loop is melted or

disrupted. This method appears to result in a higher affinity of binding as observed by

174

the variation in our dissociation constants between the gel shifts and the SPR
experiments. The Biacore data reveals a similar trend, but provide more reliable kinetic
data showing how the CYb mRNA/gRNA complex comes together in the absence of any
chaperones or annealing factors, while the gel shift data show how this interaction may
change in the presence of accessory factors. Although, there are differences, similar
trends of binding are confirmed using both techniques.

Each editing event appears to increase the affinity of the mRNA for the gRNA. This

1..

indicates that once editing begins, even if a required protein co-factor is released or is

A

displaced by the editosome, the CYb mRNA/gRNA complex will continue in the editing
process, and with each editing event, the mRNA becomes more likely to finish editing.
Progressivity may be an important part of the editing process, and the increased affmity
of the RNA-RNA interaction added with each editing event could be important for
editing continuity. In addition, the secondary structure of the gRNA (gRN A stem-loop)
may keep free gRNAs fi'om competing with gRNAs already bound in mRNA/gRNA
partially edited complexes.
MATERIALS AND METHODS
Oligodeoxyribonucleotides
All Oligodeoxyribonucleotides (Table 13) were ordered from Integrated DNA
Technologies, Inc. (Coralville, IA).
RNA Synthesis and Labeling

The templates and sequences for 5’CYbUT, 5’CYbPESlT, and 5’CYbPES3T (Fig.
22) have been previously described (Koslowsky et a1, 1992; Koslowsky et al., 1996).

The templates and sequences for CYbU, CYbPESl, and CYbPES3 (Fig. 22) were similar

175

to the above sequences but had the vector sequence removed. NgCYb-558 and NgCYb-
558sU (no U-tail) (Leung & Koslowsky, 1999, 2001a, b) (Fig. 22) were synthesized
using the Uhlenbeck single-stranded T7 transcription method (Milligan et al., 1987) with
the oligonucleotides described in Table 13. All RNAs were synthesized by T7 RNA
polymerase using a Ribomax Large Scale RNA production kit (Promega) according to
the manufacturer’s directions. RNAs were gel purified as described previously
(Koslowsky et al., 2004). The gRNAs used for the gel shifts were 5’ end labeled by
treating with calf intestinal phosphatase (Invitrogen) followed by T4 kinase (Invitrogen)
with [732P] rATP as described previously (Koslowsky et al., 2004).
Serial Dilutions of mRNA
All mRN A concentrations (Table 8) were measured using an ND-lOOO
spectrophotometer (NanoDrop Technologies, Wilmington, DE).
Gel Band-shift Analysis

The apparent equilibrium constants of the gRNA binding to the mRN A were found
using a direct band-shift electrophoresis assay (Koslowsky et a1, 2004). The mRN As
were serially diluted (Table 8). The serial diluted concentrations of mRNA were mixed
with 50 frnols of 5’end-1abeled gRNA (50 mM Tris pH 7.5, 0.1 mM EDTA, 100 mM
KCl, 3% glycerol, 0.05% xylene cyanol, and MgOAc as indicated) in 10 pl (Koslowsky
et al., 2004). Each sample was heated to 70°C for 2 minutes and allowed to cool to room
temperature for 3 hours before being loaded onto a 6% polyacrylamide native gel (50
mM Tris pH 7.5, 50 mM Hepes pH 7.5, 0.1 mM EDTA, indicated concentration of
magnesium acetate) under current. The gels were electrophoresed, fixed, and scanned on

a phosphorirnager (Molecular Dynamics) as described previously (Koslowsky et al.,

176

2004). The apparent affinity constant of gRNA binding the mRN A was extracted from

data-point fitting using Kaleidagraph 3.5 (Synergy Software) with the equation:

([gRNAfi’imRNAfD
[[gRNAf ] + KB)

fiee gRN A, [mRNAf] equal to the concentrations of unlabeled fiee mRNA, [Complex]

 

[Complex] = , with [gRNAf] equal to the concentration of the

equal to the concentration of mRNA/gRNA complex, and K D equal to the dissociation

constant. The values reported here represent the average of 4 gels and the error is
calculated fi'om the difference in these values.
Dissociation rate gels

The apparent dissociation rate constants of the gRN A binding to the mRN A were
found using a direct band-shift electrophoresis assay (Lima et a1., 1992). The mRN A was
mixed with 10 fmols of 5’end-1abeled gRNA (50 mM Tris pH 7.5, 0.1 mM EDTA, 100
mM KCl, 3% glycerol, 0.05% xylene cyanol, and l X RNase Secure (Ambion, Austin,
TX), and 1 mM magnesium acetate) in 100 ill. Each sample was heated to 60°C for 10
minutes, then held at 27°C for 20 hours before adding an oligonucleotide complementary
(Integrated DNA Technologies, Coralville, IA) (Fig. 22) to the gRNA anchor sequence
(10X the concentration of the mRN A). A short complementary oligonucleotide was
added to the CYbU and CYbPESl samples, while a long complementary oligonucleotide
was added to the CYbPES3 samples (Fig. 22D). This complementary oligonucleotide
does not allow the gRNA to rebind the mRN A. Time points were taken at the indicated
times and snap fi‘ozen on dry ice. The time points were individually thawed and loaded

onto an 8% 19:1 polyacrylamide native gel (50 mM Tris pH 7.5, 50 mM Hepes pH 7.5,

177

«1'-

0.1 mM EDTA, and 1 mM magnesium acetate) under current. The gels were
electrophoresed, fixed, and scanned on a phosphorimager (Molecular Dynamics) as
described previously (Koslowsky et al., 2004). The apparent dissociation rate constant of
gRN A binding the mRNA was extracted from data-point fitting using Kaleidagraph 3.5

(Synergy Software) with the equation:

— k T '
[ComplexTotallz[ComplexTime1]*e ( Off * me).

The [ComplexTot a1 ] represents total binding of mRNA/gRNA complex formation,
[Complexnme 1] is the amount of complex formation at the first time point and k0,, is the

apparent dissociation rate constant (Motulsky & Christopoulos, 2003).
Association rate gels

The apparent association rate constants of the gRN A binding to the mRN A were
found using a direct band-shift electrophoresis assay (Lima et al., 1992). The mRN A was
mixed with 10 frnols of 5’end-labeled gRNA (50 mM Tris pH 7.5, 0.1 mM EDTA, 100
mM KC], 3% glycerol, 0.05% xylene cyanol, l X RNase Secure (Ambion) and 1 mM
magnesium acetate) in 100 pl. Each sample was heated to 60°C for 10 minutes, and then
held at 27°C. Time points were taken at the indicated times and snap frozen on dry ice.
The time points were individually thawed and loaded onto an 8% 19:] polyacrylamide
native gel (50 mM Tris pH 7.5, 50 mM Hepes pH 7.5, 0.1 mM EDTA, 1 mM magnesium
acetate) under current. The gels were electrophoresed, fixed, and scanned on a
phosphorimager (Molecular Dynamics) as described previously (Koslowsky et al., 2004).

The apparent observed rate constant for CYbU was calculated using a single exponential

178

k
on
fit based on the reaction mechanism mRNA + gRNA (.9 gRNIZtRN A where the mRN A

k 017
and gRNA bind to form the mRNA/gRN A complex. For the partially edited mRNAs,
CYbPESl and CYbPES3, the best line fit for the data points was a double exponential fit

using the reaction mechanism:

k
k RN 0772
mRNA * +gRNA—0—nl—9g %RNA 4.) mRNA + gRNA where the mRNA“
k
Off

represents an alternate conformation of the mRNA that allows a significantly faster
association rate with no noticeable dissociation along with a slower mechanism of
binding similar to the unedited CYb reaction. The observed rate constant of gRN A
binding the unedited CYb mRN A was extracted from data-point fitting using

Kaleidagraph 3.5 (Synergy Software) with the equation:

[Complains]= [ComplexToml]1t(1— e(' “0’” " Time») The [Complexsal represents

specific binding of the complex for each time point, [ComplaxTotal] is the amount of

specific binding of the complex at equilibrium for a single exponential fit. The observed
rate constant of gRNA binding the partially edited CYb mRN As was extracted fiom data-

point fitting usingKaleidagraph 3.5 (Synergy Software) with the equation:

[ComplexSBl = ([Complexl 1*(1 — e(-(k0b31 7 Time») + ([CompleXZ ]‘ (I - e(— (1‘0sz 7 nM))D) . The

[Complex SB] represents specific binding of the complex for each time point, km is the

observed rate constant, [Complex,] and [Complexz] represent two amplitudes of binding

for a double exponential fit. The first observed rate constant for CYbPESl and

179

 

CYbPES3 can only be measured by the first data point, and therefore the first observed
rate constant cannot be reliably measured using gel shifts. Since the first binding event is
so rapid, it cannot be accurately measured using this technique, only the second slower
observed rate constant will be reported and used to calculate k0... This first binding event
may represent a small fi‘action of mRNAs that have an open helix due to structural
breathing, or it could represent a very unstable helix nucleation event. The limitations of

this technique prevent further examination of this phenomenon as it is not stable enough

to measure using gel shifts. The apparent association rate constant km was calculated l"
. . (kobs 'kofl 7
from the observed rate constant k0,” usmg the equation: k0,, = —— where k0,,
[mRNATotaI]

represents the apparent association rate constant, kw. is the apparent dissociation rate

constant calculated earlier, and the [mRNAW] is the concentration of mRN A used

(Motulsky & Christopoulos, 2003).
Surface Plasmon Resonance Studies

Measurements of the association and dissociation rate constants were performed in a
BIACORE 2000 instrument (BIACORE, Uppsala, Sweden). All the solutions used in the
binding studies were filtered through a 0.22 pm polyethersulfone membrane (Corning) or
a 0.22 pm Millex-GS membrane (Millipore) and degassed. All Biacore mRN A and
gRNA substrates are listed in Table 10. The CYbU mRNA was ligated to the BigSK-
biotin DNA tag or to Biatagl , a different biotinylated DNA tag. The measurements of
CYbU were similar regardless of the DNA tag that was attached. BigSK-biotin or
Biatagl was annealed to the 3’ end of CYbU using DNA bridging oligonucleotides

(Table 13) as described previously (Y u & Koslowsky, 2006). The bridging

180

oligonucleotides are listed in Table 13. The CYbU mRN As were gel extracted without
phenol and purified using ultra-flee MC membranes (Millipore) and Microcon tubes
(YM-SO, Millipore) according to the manufacturer’s directions. The gRNAs were
transcribed as described above and purified using Ultra-free MC and the microcon tubes
(YM-10,30) (Millipore) as above. The RNA samples were then diluted in running buffer
(20 mM Tris pH 7.5, 0.1 mM EDTA, 2 mM MgC12, and 100 mM KCI); the CYbU I
samples were run using 100 mM Tris pH 7.5, while the partially edited substrates, J
CYbPESl and CYbPES3, were run in 20 mM Tris pH 7.5. The partially edited CYb
mRNAs (CYbPESl and CYbPES3) were transcribed as described above then ligated to
CYbBia, a biotinylated DNA tag, (Table 10) using T4 DNA ligase (Roche). CYbBia is
complementary to the CYb mRN As, and so taking advantage of the stem-loop structure
of the CYb mRN A, it was directly annealed to the mRNA and ligated using T4 DNA
ligase (Roche). The gRNAs were transcribed as described above. The partially edited
CYb mRN As were gel purified using the Ultra-free MC membranes as above. The
partially edited CYb mRNA and gRNA samples were ethanol precipitated and then
dialyzed for six hours in running buffer (20 mM Tris pH 7.5, 0.1 mM EDTA, 2 mM
MgC12, and 100 mM KCl) in a Spectra/For Float-A-Lyzer (5 mM, 300 pl, MWCO 8,000)
(Spectrum Laboratories, Inc.) according to the manufacturer’s directions. All
biotinylated mRN As were diluted to 10 nM and between 250-600 resonance units (RU)
of mRNA was attached at 5 mein to a streptavidin coated SA sensor chip (BIACORE,
Uppsala, Sweden). Two cells were immobilized with mRNA, one was left unmodified to
serve as a reference cell, and one cell was immobilized with the biotinylated DNA tag as

a control cell. Binding studies were carried out running all four cells in series with

181

repetitive cycles of 150-250 pl gRNA injection at 5 mein (association of varying
concentrations of gRNA, ZOO-7000 nM), buffer flow (dissociation of gRNA) 5 pl/min for
15-20 minutes, and regeneration (50 pl injection of regeneration buffer, two 50 pl
injections of running buffer at 50p1/min). These experiments were all run at 27° C. The
regeneration buffer for the partially edited CYb experiments was 1-10 mM EDTA. The
regeneration buffer for CYbU experiments was 8 M Urea. The determination of rate
constants was performed by fitting theoretical curves to the experimental curves obtained
using BIAevaluation 3.0 software (BIACORE, Uppsala, Sweden). The equation used to
calculate the dissociation rate constants for CYbPES3 from the Biacore sensograms was

the 1:1 (Langmuir) dissociation formula describing analyte (gRNA) dissociates fi'om
surface complex (mRNA/gRNA complex): R0 te("‘d '0 "0»+01}Cset , R0 = Ymax =
maximum analyte binding capacity (RU), k, = dissociation rate constant, I = time, 70 =
time at start, Offset = residual response at infinite time (RU). The line fit for CYbU and
CYbPESl mRNA/gRNA complexes did not fit the data well, so the equation was slightly
altered for the CYbU and CYbPESl dissociation rate constant:

(RO- Ofifset)*e(- I‘d " (’ “’0» + omen The equation used to calculate the association rate

constant was the 1:1 (Langmuir) association formula describing analyte (gRNA) binds to

ligand (mRNA):

 

kg ‘ConC‘Rmax x(1-e('(k0 7C0n0+kd )‘(i "0») + R! , k = association rate constant (ken),
(Ira tConc+kd) a

C0“ = "30137 analyte concentration, R1 = bulk refiactive index effect (RU), Rm =
Ym = maximum analyte binding capacity (RU), t = time, ,0 = time at start, k d =

dissociation rate constant (km). The dissociation rate constant was incorporated into the

182

association rate constant line fit. Some of the CYbU and CYbPESl association rate line
fits used the dissociation rate constant calculated from the altered dissociation rate
equation and some used the dissociation constant calculated from the regular dissociation
rate equation. When the line fit the association curve data, the rate constants were
similar. The association rate for CYb was actually slower than can be reliably measured
by the Biacore 2000 according to machine specifications; however, this range may be
extended under favorable conditions (Myszka, 1997). This may account for the
differences observed in association rate between the gel shift data and the surface
plasmon resonance data. Separate fits for each association and dissociation curve were
analyzed globally from each experiment to obtain kon and koff, individually, and the
results were averaged. The errors reported for the rate constants were based on the
variances of all curves obtained (Nordgren et al., 2001). The dissociation equilibrium
constant (K9) was calculated fi'om the averages of the rate constants using the

equation: K D = t—d.
RNA crosslinking and mapping of crosslinks

A Uls-tail containing a 4-thio-uridine (4sU) (Dharmacon, Boulder, CO) either in the fifth
or tenth position (Table 12) was 5’ trace labeled and ligated to NgCYb-558sU with T4
DNA ligase (Roche) using a DNA bridging oligonucleotide, gCYb558(sU) bridge (IDT,
Table 13), as previously described (Leung & Koslowsky, 2001b). Half of the gRNA was
treated with p-azidophenacyl bromide (Sigma) as described previously (Leung &
Koslowsky, 2001b). The mRN A was then annealed to the gRNA, UV crosslinked,

resolved on a 6% acrylamide gel, and recovered as previously described (Leung &

183

Koslowsky, 2001b). Primer extension analysis was done as previously described (Leung

& Koslowsky, 1999) as was the RNase H analysis (Leung & Koslowsky, 2001b).

ACKNOWLEDGEMENTS

This work was supported by National Institutes of Health Grant A134155 to D.K. We’d
like to thank Larissa Reifur, Dr. Ron Patterson, Dr. John Wang, Dr. Charles Hoogstraten,
and all other members of the MSU RNA Journal Club for critical reading of the
manuscript and helpful discussions. We’d like to thank Dr. Robert Hausinger,
Microbiology and Molecular Genetics MSU, and Dr. Zachary Burton, Biochemistry and
Molecular Biology MSU, for their help with Kaleidagraph and rate constant equation
discussions. We’d like to thank Dr. Hoogstraten for his help with analyzing the Biacore
data. Also, we’d like to thank Dr. Karen Friderici and her lab, Microbiology and
Molecular Genetics MSU, for the use of their nanodrop spectrophotometer. We would
like to thank Dr. Joseph Leykam and the members of the Macromolecular Structure
Facility in the Biochemistry and Molecular Biology Department, MSU, for the use the
Biacore machine. Many thanks to Remy Brim and Andrea Hingst for many excellent

solutions made.

184

Literature Cited

Adler BK, Hajduk SL. 1997. Guide RNA requirement for editing-site-specific
endonucleolytic cleavage of preedited mRN A by mitochondrial ribonucleoprotein
particles in Trypanosoma brucei. Mol Cell Biol 1 7 :5377-5385.

Blum B, Bakalara N, Simpson L. 1990. A model for RNA editing in kinetoplastid
mitochondria: "guide" RNA molecules transcribed from maxicircle DNA provide
the edited information. Cell 60:189-198.

Blum B, Simpson L. 1990. Guide RNAs in kinetoplastid mitochondria have a
nonencoded 3' oligo(U) tail involved in recognition of the preedited region. Cell
62:391-397.

F eagin JE, Jasmer DP, Stuart K. 1985. Apocytochrome b and other mitochondrial DNA
sequences are differentially expressed during the life cycle of Trypanosoma
brucei. Nucleic Acids Res 13:4577-4596.

Franch T, Petersen M, Wagner EG, Jacobsen JP, Gerdes K. 1999. Antisense RNA
regulation in prokaryotes: rapid RNA/RN A interaction facilitated by a general U-
turn loop structure. J Mol Biol 294:1115-1125.

Koslowsky DJ, Goringer HU, Morales TH, Stuart K. 1992. In vitro guide RN A/mRN A
chimaera formation in Trypanosoma brucei RNA editing. Nature 356:807-809.

Koslowsky DJ, Kutas SM, Stuart K. 1996. Distinct differences in the requirements for
ribonucleoprotein complex formation on differentially regulated pre-edited
mRN As in Trypanosoma brucei. Mol Biochem Parasitol 80: 1-14.

Koslowsky DJ, Reifur L, Yu LE, Chen W. 2004. Evidence for U-Tail Stabilization of
gRNA/mRN A Interactions in Kinetoplastid RNA Editing. RNA Biology 1:28-34.

Leung SS, Koslowsky DJ. 1999. Mapping contacts between gRN A and mRNA in
trypanosome RNA editing. Nucleic Acids Res 27:778-787.

Leung SS, Koslowsky DJ. 2001a. Interactions of mRNAs and gRNAs involved in
trypanosome mitochondrial RNA editing: structure probing of an mRN A bound to
its cognate gRNA. RNA 7:1803-1816.

Leung SS, Koslowsky DJ. 2001b. RNA editing in Trypanosoma brucei: characterization
of gRN A U-tail interactions with partially edited mRN A substrates. Nucleic Acids
Res 292703-709.

Lewis BP, Burge CB, Bartel DP. 2005. Conserved seed pairing, often flanked by
adenosines, indicates that thousands of human genes are microRNA targets. Cell
120:15-20.

185

Lima WF, Monia BP, Ecker DJ, Freier SM. 1992. Implication of RNA structure on
antisense oligonucleotide hybridization kinetics. Biochemistry 31:12055-12061.

Milligan JF, Groebe DR, Witherell GW, Uhlenbeck CC. 1987. Oligoribonucleotide
synthesis using T7 RNA polymerase and synthetic DNA templates. Nucleic Acids
Res 15:8783-8798.

Moore MJ, Sharp PA. 1992. Site-specific modification of pre-mRN A: the 2'-hydroxyl
groups at the splice sites. Science 256:992-997.

Motulsky HJ, Christopoulos A. 2003. Fitting models to biological data using linear and
nonlinear regression. A practical guide to curve fitting. San Diego: Graphpad
Software Inc., www.graphpad.com.

Myszka DG. 1997. Kinetic analysis of macromolecular interactions using surface
plasmon resonance biosensors. Curr Opin Biotechnol 8:50-57.

Nair TM, Myszka DG, Davis DR. 2000. Surface plasmon resonance kinetic studies of the
HIV TAR RNA kissing hairpin complex and its stabilization by 2-thiouridine
modification. Nucleic Acids Res 28:1935-1940.

Nebohacova M, Maslov DA, Falick AM, Simpson L. 2004. The effect of RNA
interference Down-regulation of RNA editing 3'-terminal uridylyl transferase
(TUTase) 1 on mitochondrial de novo protein synthesis and stability of
respiratory complexes in Trypanosoma brucei. J Biol Chem 279:7819-7825.

Nordgren S, Slagter-Jager JG, Wagner GH. 2001. Real time kinetic studies of the

interaction between folded antisense and target RNAs using surface plasmon
resonance. J Mol Biol 310:1 125-1 134.

Reifur L, Cruz-Reyes J, vanHartesvelt M, Koslowsky DJ. 2006. Anchor binding site
accessibility and RNA editing in Trypanosoma brucei. manuscript in prep.

Seiwert SD, Stuart K. 1994. RNA editing: transfer of genetic information from gRNA to
precursor mRNA in vitro. Science 266:114-117.

Sekar MM, Bloch W, St John PM. 2005. Comparative study of sequence-dependent
hybridization kinetics in solution and on microspheres. Nucleic Acids Res 33 :366-
375.

Simpson L, Aphasizhev R, Gao G, Kang X. 2004. Mitochondrial proteins and complexes
in Leishmania and Trypanosoma involved in U-insertion/deletion RNA editing.
RNA 10:159-170.

Slagter-Jager JG. 2003. CopA and CopT: The Perfect RNA Couple. Department of Cell
and Molecular Biolog. Uppsala, Sweden: Uppsala University. pp 1-47.

186

Stuart KD, Schnaufer A, Ernst NL, Panigrahi AK 2005. Complex management: RNA
editing in trypanosomes. Trends Biochem Sci 30297-105.

Young S, Wagner RW. 1991. Hybridization and dissociation rates of phosphodiester or
modified oligodeoxynucleotides with RNA at near-physiological conditions.
Nucleic Acids Res 19:2463-2470.

Yu LE, Koslowsky DJ. 2006. Interactions of mRNAs and gRNAs involved in
trypanosome mitochondrial RNA editing: structure probing of an gRN A bound to
its cognate mRNA. RNA 12:1050-1060.

187

CHAPTER 5

CONCLUSIONS AND FUTURE RESEARCH

188

Summary

The trypanosome is able to effect a rapid change in its development, morphology, and
energy metabolism in response to instantaneous differences in temperature (insect 27°C,
mammal 37°C), cellular environment, and immune response that is brought on by a
change of host (Brown & Neva, 1983). Many of the proteins for energy metabolism that
are made in the mitochondria are thought to be developmentally regulated by RNA
editing. Developmental regulation of RNA editing may control mitochondrial biogenesis
by only allowing production of respiratory proteins during the correct life cycle.
However, very little is known concerning how editing is regulated. The gRNA
transcripts, the mRN A transcripts, and the editosome protein complexes are always
present (Hajduk & Sabatini, 1998). Additionally, very little is known about how the
editing complex is assembled onto specific RNAs. There are hundreds of different
mRNA/gRNA pairs, but no conserved sequence domains have been found in mRN As.
The only conserved sequence domain shared between the gRN As is the U-tail. Previous
work in this lab suggests that different mRNA/gRNA pairs can form similar structures
composed of three helices. Structure recognition may be important for efficient
editosome assembly with the mRN A/gRN A pair. Discovering the structure of an
mRNA/gRNA pair such as CYb may be the first step in unraveling what RNA structure
attracts the editosome, and how the RNA structure changes when in contact with the
editosome. This lab is interested in gRNA targeting of the mRN A for binding and how
this affects editing efficiency. Specifically, discovering how the mRNA/gRNA complex
interaction is occurring and understanding how the anchor sequence, U-tail, and guiding

region interact with the mRNA during RNA editing was investigated. The objective of

189

this study has been to characterize the structure of the CYb gRNA/mRNA complex as
well as what role the structure of the CYb mRNA plays in the binding affinity of the
gRNA to the mRNA. The efl‘ects partial editing of the mRNA have on the gRNA/mRNA
complex structure were studied as well as the kinetics of its formation. Additionally, the
role the U-tail has in the CYb mRNA/gRNA complex interaction was investigated.
Anchor binding site structure influences editing

Expression of the CYb mRNA is regulated through RNA editing (Feagin et al., 1988)
and is needed for mitochondrial respiration in the insect host (V ickerman, 1965). The
unedited CYb mRN A forms a stable stem-loop structure, with the anchor binding site
(ABS) base paired within the helical stem (Leung & Koslowsky, 20013). It is clear that
the structure of the immediate editing domain on the mRN A can profoundly affect the
efficiency of the interaction between the gRNA and the target mRN A (Koslowsky et al.,
2004). The unedited CYb mRNA/gRNA binding interaction was compared with two
different unedited mRN A/gRN A pairs that have single stranded predicted anchor binding
sites (ABS). The ATPase 6 (A6) mRNA/gRNA pair and the NADH dehydrogenase 7
(ND7) mRNA/gRNA pair had a 2000 fold and 8400 fold higher aflinity anchor target
binding respectively than the CYb mRNA/gRNA pair. The association rate constants and
dissociation rate constants show that this is because the A6 pair has a 20 fold faster
association rate and a 20 fold slower dissociation rate than the CYb pair. The ND7
results are similar showing that the ND7 mRNA/gRNA pair has a 90 fold faster
association rate than the CYbU pair and a similar 20 fold slower dissociation rate
comparable to A6. Clearly, the mRNA structure around the immediate editing domain

can strongly affect the gRN A anchor target binding and appears to be regulating the CYb

190

editing kinetically with a slow on-rate and a fast off-rate. The mND7550 substrate and
the A6UENDSh substrate are not controlled kinetically as they have a relatively fast on-
rate and a very slow off-rate. The mRNA/gRNA complex stability appears to be
controlled by the slow off-rate. This slow off-rate may be necessary to allow editing to
occur. Both A6 and ND7 mRN As are predicted to have single stranded anchor binding
sites that are easily accessible by the gRNA. However, the ND7 mRNA/gRNA complex
is not edited in vitro as the constitutively edited A6 pair is, suggesting that an alternative
method of editing regulation exists. The editing of the mRNA/gRNA complexes may
also be controlled structurally, so that only RNA complexes with the correct tertiary
structures are edited. Accessory factors such as chaperones that alter and stabilize
structure may be required for editing to occur. Perhaps A6 is the only mRN A/gRN A
capable of correct folding in the absence of these factors.
CYb mRNA/gRNA complex structure during editing

In this study, the structure of NgCYb-558 alone and its structure when paired with
its cognate unedited mRNA or partially edited mRNA were also investigated. Solution
structure probing of the CYb mRNA/gRNA complex reveals a 3 helix structure in the
unedited complex that, although modified by editing, is maintained through the third
editing site. The changes brought about by partial editing include an anchor duplex
doubled in length (Leung & Koslowsky, 2001b; Yu & Koslowsky, 2006). As a
consequence of the longer anchor, the stem-loop of the gRNA appears to begin
incorporating the U-tail; this results in decreased U-tail interaction with the mRNA (Yu
& Koslowsky, 2006). The gRNA stemoloop may be an important structural component

of the initial editing complex. Previous work using gel shift analyses indicates that the

191

U-tail is very important for stabilization of the interaction of some mRN A/ gRN A pairs
(Koslowsky et al., 2004). RNAi studies have also shown that when the RNA editing
terminal uridylyl transferase (RETl), responsible for adding the U-tail to gRNAs is down
regulated, there is a decrease in edited mRNAs and inhibited growth of the trypanosome
(Aphasizhev et al., 2002) suggesting that the U-tail is necessary for in viva editing (Gott,
2003; Nebohacova et al., 2004). Guide RNAs appear to be able to take advantage of U-
tail flexibility through the ability of uridines to base pair with both purine bases. By
employing an uridylate tail, the gRNA may increase mRNA/gRNA complex stability,
without hampering the U-tail migration needed within the complex during the editing
process.
Kinetics of anchor target binding during editing

Using gel band shift assays and surface plasmon resonance, the effects of the changes
in structure in the partially edited CYb mRNA/gRNA complex were investigated to
discover how the binding interaction changes between the mRN A and gRNA during
editing. In this study, the addition of two uridines to the anchor binding site (ABS) of the
mRN A decreases the dissociation constant (KD) significantly. It appears that these three
bases increase the size of the terminal loop of the mRNA and make it more accessible.
This change in structure leads to an increase in formation of the mRNA/gRNA complex
that appears to be caused by a decrease in the dissociation rate. There is an additional
decrease in KD when four more uridylates are added to the next two editing sites. These
next two editing events have doubled the size of the mRN A ABS extending through the
terminal loop. In addition, the gRNA forms a stable stem-loop structure composed of

part of the newly accessible ABS and this section of the gRN A anchor is unavailable for

192

gRNA binding. This may explain why a large increase of association rate is not observed
for this substrate. This may be an evolutionary strategy to keep free gRN As from
disrupting partially edited mRNA/gRNA complexes. This alteration in mRNA structure
through editing appears to further increase the stability of the complex by decreasing the
dissociation rate. With editing of the first site, the barrier to begin editing is greatly
reduced, and the CYb mRNA/gRNA complex becomes more stable with increased
editing. The longer anchor helix results in a slower dissociation rate. Once editing
begins with the help of cofactors, the mRNA/gRN A pair may need to proceed through
the editing process in the absence of a co-factor. The kinetics of the interaction improve
with each editing event; this may be an important part of editing progression.

The U-tail has been predicted to slow dissociation; however, one of the interesting
results of these experiments is that the U—tail appears to help the gRN A to associate with
the unedited CYb mRNA. This trend was observed both in the gel shifts and in the
Biacore data. This is the first evidence that shows the U-tail contributes to anchor target
binding. However, in the presence of a mitochondrial helicase or an anchor armealing
factor this U-tail function may be altered.

Where the U-tail is interacting in the CYb mRNA/gRNA complex was also studied.
A 4-thio-uridine (4sU) crosslinking agent was placed in the fifth and then tenth position
of the U-tail on NgCYb-558. This gRN A was annealed to either the unedited CYb
mRNA or a partially edited substrate and UV crosslinked. These U-tail positions were
mapped and the main crosslink band was found to interact with the loop (U 5) and bulge
region (U 10) that surround the anchor binding site. This suggests that the U-tail is

helping the CYb gRNA/mRNA complex form. It appears to do this through disruption of

193

the mRN A stem-loop, by binding one side of the helix that hides the anchor binding site.
The CYb anchor binding site sequence is U/C rich, so the U-tail will not bind the anchor
binding site. The sequence base paired to the anchor binding site is A/G rich, so the U-
tail can bind this sequence easily. This makes the U-tail the perfect tool to help disrupt
the stem-loop. Also, the A-U and G-U base pairs are weaker interactions with only 2
hydrogen bonds holding the base pairing interaction together; perhaps this nukes the U-
tail more flexible so it is able to continue its migration during editing to become part of
the gRNA stem-loop.

The U-tail may not be necessary for the A6 and ND7 mRNA/gRN A interactions, or
the U-tail function in these interactions may be different than it is for CYb. Interestingly,
the anchor binding site sequence for these mRNAs is U/C rich, so the U-tail will not
interfere with anchor helix formation. It will be interesting to see if the anchor binding
sites for most mRNA/gRNA pairs have evolved to be U/C rich to avoid U-tail
interference.

Much is known about the protein components of the editosome. A growing list of
proteins is now known to be necessary for RNA editing. However, while much is known
about the proteins, very little is known about how gRNAs target mRN As for binding and
how the editosome proteins target the mRNA/gRNA complexes for editing. There is no
sequence homology between the hundreds of different pairs other than the U-tail on the
gRNA. This lab is interested in discovering how these protein complexes recognize the
mRN A/gRN A complexes, and we hypothesize that the editosome protein complexes
recognize elements of structure instead of sequence. This work has made a significant

contribution to discovering how the gRN A targets its mRNA and what type of binding

194

affinity different gRNA/mRN A pairs have. By comparing the double stranded CYb
mRN A to other single stranded mRN As, it was found that the structure of the immediate
editing domain is important for gRNA target binding. Also, the editing process appears
to improve the mRNA/gRNA interaction, so that mRNA and gRNA pairs that have
started editing are more likely to finish with each editing event; this may be important for
editing progression. Additionally, the versatility of the U-tail and its role in gRN A
binding association along with its role in maintaining important structures such as the
gRN A stem-loop provide new insights into its biological function and purpose. This
work helps illuminate how RNA hybridization during editing occurs and provides deeper
insights into the biological function of this RNA-RNA interaction. This has laid the
groundwork for future studies that include mitochondrial proteins to see how the RNA-
RNA interactions change in the presence of mitochondrial proteins.

The solution structure probing data show that the structure of the mRNA/ gRN A
complex appears to form a three way helical junction. RNA helical junctions are
versatile structural elements that are able to perform multiple biological functions using
differing structural requirements. Some ribozymes containing three way junctions can
exist in inactive extended conformations for long periods of time and may require
regulatory ligands to form the correct structure (Goldschmidt et al., 2002). The
mRNA/gRNA complexes may require multiple chaperones for editing to occur. One
example of a functional RNA requiring chaperones is the group I self-splicing intron.
For group I introns, CYT-18 appears to stabilize group I intron structures, while CYT-l9,
an ATP dependent nucleic acid helix-destabilizer, uses ATP hydrolysis to unfold

kinetically trapped intermediates to allow correct folding. Both proteins appear to be

195

necessary for self-splicing of this group I intron (Lorsch, 2002; Mohr et al., 2002).
Chaperones similar to these may be required specifically for the CYb mRNA/gRNA
complex to destabilize the CYb mRN A stem-loop and stabilize the mRNA/gRN A
complex. In addition, other RNA chaperones involved in RNA-RNA matchmaking
(Muller et aL, 2001) or involved in strand exchange (Arthur et aL, 2003) may be
necessary to promote faster anchor binding in vivo for other mRNA/gRNA pairs. The
ND7 and A6 mRN As are both predicted to have similar secondary structures (three way
junctions) and have similar kinetics of binding, but A6 is the only mRNA/gRNA pair
capable of in vitro editing. Perhaps the ND7 complex forms the incorrect tertiary
structure and the editosome can not recognize it, so editing can not occur. This occurs in
the hepatitis delta virus where ADARl, an RNA editing enzyme in humans, only
performs adenosine to inosine editing on the RNA genomes with branched structures but
not unbranched structures (Linnstaedt et al., 2006). Perhaps editing in trypanosomes is
controlled structurally and mRNA/gRNA complexes other than A6 require chaperones to
form the correct tertiary structures as well as stabilize this structure in order to be
recognized by the editosome.
Editosome recruitment

An objective of this study was to discover the structure and kinetics of the CYb
mRNA/gRNA interaction in order to discover how the editosome is recruited to the
mRNA/gRNA complex. One possibility for editosome recruitment is that proteins with
RNA binding sites are needed to recognize different structural elements of the
mRNA/gRNA complex. Solution structure probing of additional mRNA/gRNA pairs

may help prove that structure recognition may be how the editosome targets

I96

mRNA/gRNA complexes. One element of structure the editosome may recognize is the
mRN A/gRN A anchor helix. The anchor helix is the first element of structure formed
during editing, and the gRNA anchor binding the mRNA ABS is thought to be the first
step in editing. The endonucleases of the editosome have double-stranded RNA binding
domains that may target the anchor helix for binding to cleave the mRN A strand
(Panigrahi et al., 2006). This anchor binding step needs to be extremely accurate in order
to eliminate incorrect editing that could result in a non-functional protein.

The anchor helix may not be the only element of RNA structure necessary for editing.
The local structure at the editing site may determine which protein editing complex is
recruited. Different editosome complexes (U—deletion or U-insertion) must bind during
editing, so there must be additional structural elements necessary for the correct complex
to be recruited. One of the editosome proteins that binds RNA, KREPA4, appears to bind
gRNA U-tails or U—rich sequence with a putative Sl motif similar to a cold shock domain
(Salavati et al., 2006). Proteins that prefer U-rich sequence could function as sensors to
detect U’s needing deletion. Perhaps the key factor in deciding which editing complex is
recruited is an mRN A bulge detector. The editosome may recognize the 3-helix
mRNA/gRNA structure and bind. If the single stranded bulge is U-rich next to the
anchor helix, a U-deletion complex is recruited. If the single stranded bulge next to the
anchor helix is not U-rich, a U-insertion complex is recruited. In addition, there may be
proteins of the editosome that bind various elements of RNA structure such as the gRN A
stem-loop in order to correctly position the editosome on the complex.

One method of characterizing mRN A/gRN A structure and protein positioning would

be using crosslinking agents in various positions in the gRNA and mRNA, annealing the

197

 

two RNA molecules, and adding editosome complexes. Once this has been done, the
RNA/protein complexes would be UV crosslinked and resolved on SDS page gels.
Antibodies against the various proteins could be used to identify which proteins are
binding in which positions. Additionally, changing the sequence of the mRNA and
gRN A from U-insertion to U-deletion and vice versa would allow differentiation of
different proteins involved as bulge sensors for U-insertion and U—deletion. If a specific
protein always binds the same element of structure such as the gRN A stem-loop, anchor
helix, editing bulge, or U-tail, these elements of structure could serve as points of
orientation that each protein binds to correctly position the editosome for editing.
Removal of these elements of structure could disrupt editing.

In addition to crosslinking studies, monitoring conformational changes in the
mRNA/gRNA secondary and tertiary structure through time resolved fluorescence
resonance energy transfer (FRET) would give detailed information about secondary and
tertiary structure changes of the mRNAs and gRNAs separately and together. FRET
describes experiments where donor and acceptor dyes are attached to two sites on the
molecule to measure the distance between two regions of a molecule. The donor
fluorophore donates its energy to the acceptor fluorophore in a distance dependent
manner, so that the change in fluorescence between two dyes indicates a change in
conformation. Using FRET would enable study of single-molecule conformation
changes, conformation changes caused by protein binding, and thermodynamics of
protein binding (Ha et al., 1999). In addition, FRET measurements would be able to

detect faster association events that SPR and gel shifts cannot detect. Investigating how

198

each new accessory factor affects the mRNA/gRNA complex conformation would
provide much new information.
Editing Accessory Factors

Although a core editosome complex has been identified, this set of proteins is not
sufficient to generate editing in vitro with any substrate other than the ATPase 6 subunit.
Additional accessory factors have been identified that affect editing. Most of these have
not been proven to directly affect the editing process. A few of these proteins seem to
have direct effects on CYb editing. These proteins include MRP1 and 2 as well as
RBP16.
MRP1 and MRP2

The mitochondrial RNA-binding proteins (MRP), MRP1 and MRP2, are accessory
factors to the editing process that are arginine rich and bind to gRNAs (Koller et al.,
1997; Blom et al., 2001). Both MRP proteins co-purify in a protein complex, and when
both proteins are knocked down in RNAi experiments, there are very low amounts of
CYb mRNA editing (Vondruskova et al., 2005). MRP1 has an RNA annealing activity
(Muller et al., 2001) that is thought to reduce the electrostatic repulsion between the
mRN A and gRNA anchor. This is thought to make anchor hybridization favorable. The
formation of the anchor helix is thought to release MRP1 fi'om the gRN A, because MRP1
has a low affmity for dsRNA (Muller & Goringer, 2002). The crystal structure of the
MRP1/MRP2 complex reveals a heterotetramer with MRP2 binding the gRNA stem-loop

and MRP1 holding the anchor sequence in a single stranded confornmtion. MRP1

' presents the anchor sequence with the bases exposed to the solvent ready to bind the pre-

mRN A. The beta sheet of the MRP] protein is electropositive and thought to provide

199

charge neutralization for the anchor phosphate backbone (Schumacher et a1., 2006). The
solution structure probing on the CYb gRN A also suggests that the anchor sequence is
presented to the mRN A in single-stranded form (Y 11 & Koslowsky, 2006). It appears that
this annealing factor ensures that the anchor is single stranded. The structure data for the
gRNA in this complex correlates well with our gRNA alone structure for NgCYb-558
providing evidence that the RNA structures are relevant. In addition, these accessory
factors appear to recognize structure not sequence also verifying our hypothesis that the
structure and not the sequence of the mRN A/gRN A complex is important for recognition
by the editosome proteins as MRP1 and MRP2 are transient members of the editosome
complex (Allen et al., 1998). Interestingly, the structure of the gRN A is not altered by
the MRP complex (Hermann et al., 1997; Schumacher et al., 2006), and the gRNA
guiding region stem-loop is maintained providing additional evidence that the gRNA
stem-loop appears to be an important structure for the editing process. The MRP
complex only weakly binds the U-tail (Muller et al., 2001; Schumacher et al., 2006)
suggesting that while the anchor sequence is bound by MRP1 and the gRN A stem-loop is
bound to MRP2 (Schumacher et al., 2006), the U-tail is free to interact with the mRN A.
In addition, both previous and present crosslinking studies show that the U-tail interacts
with a section of mRN A upstream of the editing sites (Leung & Koslowsky, 1999). One
possibility for an U-tail interaction may be formation of Hoogsteen base pairs with the
purine rich mRN A sequence upstream of the first editing site forming a
pyrimidine/purine/pyrimidine triplex that destabilizes the ABS helix and allows the

gRNA anchor to bind the ABS (Rajagopal & Feigon, 1989; Pilch et al., 1990).

200

 

MRP1 and 2 may be important co-factors for CYb editing. The CYb gRNA cannot
bind the mRN A easily because of the double stranded nature of the anchor binding site
(Koslowsky et al., 2004), and it appears that MRP1 and 2 might be involved in the
formation of anchor duplexes. The fact that constitutively edited mRN As such as A6
were unaffected by MRP1 and 2 RNAi knock-down (V ondruskova et al., 2005), shows
that these proteins could be specific annealing factors for CYb editing or that A6 editing
may only be enhanced with this co-factor. Future experiments with the addition of MRP
l and 2 could show that these proteins are important chaperone proteins to help the
gRN A bind the mRN A. Also, if these proteins were developmentally regulated or
involved with proteins that were developmentally regulated, this would allow discovery
of how CYb editing is developmentally regulated.

RBP16

Another possible editing accessory factor is, RBP16, a 16 kDa protein that binds U-
rich sequences such as the gRNA U-tail. It contains an N-terminal cold shock domain
and a C-terminal region rich in arginine and lysine residues (Hayman & Read, 1999). The
in vitro association between RBP16 and gRN A is increased in the presence of p22, a
human p32 homologue (Hayman et al., 2001). RNAi knock-down of RBP16 results in 3
~98% reduction in CYb mRNA editing (Pelletier & Read, 2003). Recent studies using
enhanced gRNAs show that RBP16 may enhance CYb U-insertion editing (Miller et al.,
2006). RBP16 could be important for CYb editing. Additional in vitro studies using
RBP16 with p22 will need to be done to see if RBP16 can improve the unenhanced CYb

gRNA’s interaction with the mRN A through structure stabilization or annealing activity.

201

The studies of these accessory factors are inconclusive. And showing that these
accessory factors exist and proving their function in vivo is very difficult. In addition,
studying these mRN A/gRN A pairs in the presence of these mitochondrial proteins will
show if the mRN A/gRN A complex structure changes in the presence of protein. Adding
the editosome complexes to the mRN A/gRN A complexes after exposure to these
accessory factors could result in more in vitro editing for other mRN As besides A6.

One of the objectives of these studies was to discover the function of the U-tail. The
U-tail appears to have multiple functions during editing in the CYb mRNA/gRNA
complex. It appears to improve association between the gRNA and the mRN A by
binding the complementary strand to the ABS to help disrupt the helix. Additional
mRNAs both with single stranded and double stranded anchor binding sites could be
investigated to see if they have similar kinetics of anchor target binding as well as
association and dissociation rates. The U-tail also maintains the gRNA stem-loop by
feeding into the gRNA stem-loop as the anchor helix grows during editing. It will be

interesting to see if other mRNA/gRNA pairs employ the U-tail for this function.

202

Literature Cited

Allen TE, Heidmann 8, Reed R, Myler PJ, Goringer HU, Stuart KD. 1998. Association of
guide RNA binding protein gBP21 with active RNA editing complexes in
Trypanosoma brucei. Mol Cell Biol 18:6014—6022.

Aphasizhev R, Sbicego S, Peris M, Jang SH, Aphasizheva 1, Simpson AM, Rivlin A,
Simpson L. 2002. Trypanosome mitochondrial 3' terminal uridylyl transferase
(TUTase): the key enzyme in U-insertion/deletion RNA editing. Cell [08:637-
648.

Arthur DC, Ghetu AF, Gubbins MJ, Edwards RA, Frost LS, Glover IN. 2003. FinO is an
RNA chaperone that facilitates sense-antisense RNA interactions. Embo J
22:6346-6355.

Blom D, Burg J, Breek CK, Speijer D, Muijsers AO, Benne R. 2001. Cloning and
characterization of two guide RNA-binding proteins from mitochondria of
Crithidia fasciculata: gBP27, a novel protein, and gBP29, the orthologue of
Trypanosoma brucei gBP21. Nucleic Acids Res 29:2950—2962.

Brown HW, Neva FA. 1983. Ch. 4: Blood and Tissue Protozoa of Man. Basic Clinical
Parasitology. Norwalk, CT: Appleton Century Croft. pp 45-53.

Feagin JE, Shaw JM, Simpson L, Stuart K. 1988. Creation of AUG initiation codons by
addition of uridines within cytochrome b transcripts of kinetoplastids. Proc Natl
Acad Sci U S A 85:539-543.

Goldschmidt V, Rigourd M, Ehresmann C, Le Grice SF, Ehresmann B, Marquet R. 2002.
Direct and indirect contributions of RNA secondary structure elements to the
initiation of HIV -1 reverse transcription. J Biol Chem 277:43233-43242.

Gott JM. 2003. Two distinct roles for terminal uridylyl transferases in RNA editing. Proc
Natl Acad Sci U SA 100.10583—10584.

Ha T, Zhuang X, Kim HD, Orr JW, Williamson JR, Chu S. 1999. Ligand-induced
conformational changes observed in single RNA molecules. Proc Natl Acad Sci U
S A 96:9077-9082.

Hajduk SL, Sabatini RS. 1998. Mitochondrial mRNA Editing in Kinetoplastid Protozoa.
In: Grosjean H, Benne R, eds. Modification and Editing of RNA. Washington,
DC: ASM Press. pp 377-411.

Hayman ML, Miller MM, Chandler DM, Goulah CC, Read LK 2001. The trypanosome
homolog of human p32 interacts with RBP16 and stimulates its gRN A binding
activity. Nucleic Acids Res 29:5216—5225.

203

 

Hayman ML, Read LK. 1999. Trypanosoma brucei RBP16 is a mitochondrial Y-box
family protein with guide RNA binding activity. J Biol Chem 274: 12067-12074.

Hermann T, Schmid B, Heumann H, Goringer HU. 1997 . A three-dimensional working
model for a guide RNA from Trypanosoma brucei. Nucleic Acids Res 25:2311-
2318.

Koller J, Muller UF, Schmid B, Missel A, Kruft V, Stuart K, Goringer HU. 1997.
Trypanosoma brucei gBP21. An arginine-rich mitochondrial protein that binds to
guide RNA with high affmity. J Biol Chem 272:3749-3757.

Koslowsky DJ, Reifin L, Yu LE, Chen W. 2004. Evidence for U-Tail Stabilization of
gRNA/mRNA Interactions in Kinetoplastid RNA Editing. RNA Biology 1:28-34.

Leung SS, Koslowsky DJ. 1999. Mapping contacts between gRNA and mRN A in
trypanosome RNA editing. Nucleic Acids Res 27 :778-787.

Leung SS, Koslowsky DJ. 2001a. Interactions of mRN As and gRN As involved in
trypanosome mitochondrial RNA editing: structure probing of an mRN A bound to
its cognate gRNA. RNA 7:1803-1816.

Leung SS, Koslowsky DJ. 2001b. RNA editing in Trypanosoma brucei: characterization
of gRN A U-tail interactions with partially edited mRN A substrates. Nucleic Acids
Res 29:703-709.

Linnstaedt SD, Kasprzak WK, Shapiro BA, Casey JL. 2006. The role of a metastable
RNA secondary structure in hepatitis delta virus genotype III RNA editing. RNA
12:1521-1533.

Lorsch JR. 2002. RNA chaperones exist and DEAD box proteins get a life. Cell 109:797-
800.

Miller MM, Halbig K, Cruz-Reyes J, Read LK. 2006. RBP16 stimulates trypanosome
RNA editing in vitro at an early step in the editing reaction. RNA 12:1292—1303.

Mohr S, Stryker JM, Lambowitz AM. 2002. A DEAD-box protein functions as an ATP-
dependent RNA chaperone in group I intron splicing. Cell [09:769-779.

Muller UF, Goringer HU. 2002. Mechanism of the gBP21-mediated RNA/RNA
annealing reaction: matchmaking and charge reduction. Nucleic Acids Res
30:447-455.

Muller UF, Lambert L, Goringer HU. 2001. Annealing of RNA editing substrates
facilitated by guide RNA-binding protein gBP21. Embo J 20: 1394-1404.

Nebohacova M, Maslov DA, Falick AM, Simpson L. 2004. The effect of RNA
interference Down-regulation of RNA editing 3'-terminaluridy1yl transferase

204

 

(TUTase) l on mitochondrial de novo protein synthesis and stability of
respiratory complexes in Trypanosoma brucei. J Biol Chem 279:7819-7825.

Panigrahi AK, Ernst NL, Domingo GJ, Fleck M, Salavati R, Stuart KD. 2006.
Compositionally and functionally distinct editosomes in Trypanosoma brucei.
RNA 12:1038-1049.

Pelletier M, Read LK. 2003. RBP16 is a multifunctional gene regulatory protein involved
in editing and stabilization of specific mitochondrial mRN As in Trypanosoma
brucei. RNA 9:457-468.

Pilch DS, Levenson C, Shafer RH. 1990. Structural analysis of the (dA)10.2(d'I)10 triple
helix. Proc Natl Acad Sci U S A 87: 1942-1946.

Rajagopal P, Feigon J. 1989. NMR studies of triple-strand formation from the
homopurine-homopyrimidine deoxyribonucleotides d(GA)4 and d(TC)4.
Biochemistry 28:7859-7870.

Salavati R, Ernst NL, O'Rear J, Gilliam T, Tarun S, Jr., Stuart K. 2006. KREPA4, an
RNA binding protein essential for editosome integrity and survival of
Trypanosoma brucei. RNA 12:819-831.

Schumacher MA, Karamooz E, Zikova A, Trantirek L, Lukes J. 2006. Crystal Structures
of T. brucei MRP1/MRP2 Guide-RNA Binding Complex Reveal RNA
Matchmaking Mechanism. Cell [26:701-711.

Vickerman K. 1965. Polymorphism and mitochondrial activity in sleeping sickness
trypanosomes. Nature 208:762-766.

Vondruskova E, van den Burg J, Zikova A, Ernst NL, Stuart K, Benne R, Lukes J. 2005.
RNA interference analyses suggest a transcript-specific regulatory role for
mitochondrial RNA-binding proteins MRP1 and MRP2 in RNA editing and other
RNA processing in Trypanosoma brucei. J Biol Chem 280:2429-2438.

Yu LE, Koslowsky DJ. 2006. Interactions of mRNAs and gRNAs involved in
trypanosome mitochondrial RNA editing: Structure probing of a gRNA bound to
its cognate mRNA. RNA 12:1050-1060.

205