University of Southern California Dissertations and Theses

Using cyclotides as a bioscaffold to target intracellular protein-protein interactions

(USC Thesis Other)

Using cyclotides as a bioscaffold to target intracellular protein-protein interactions

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content 1

USING CYCLOTIDES AS A BIOSCAFFOLD TO TARGET
INTRACELLULAR PROTEIN-PROTEIN INTERACTIONS

By
Tao Bi

A Thesis presented to the
Department of Pharmacology and Pharmaceutical Sciences
University of Southern California
In partial fulfillment of the
Requirements for the degree
DOCTOR OF PHILOSOPHY
PHARMACEUTICAL SCIENCES

December 2017
Copyright 2017 Tao Bi
2

DEDICATION
I dedicate my dissertation to my loving family whom I am forever indebted to for the
sacrifices they made to help me pursue my dreams. They have served as a source of
unconditional motivation, strength, and love that no amount of gratitude and appreciation
can ever suffice.

3

ACKNOWLEDGEMENTS
First, I would like to thank my parents, Jingjiang Bi and Lin Xu, for being a role model that
I always look up to, and for their endless love, encouragement, and inspiration that have
supported me all the way through. I also want to thank my fiancé e, Yuting Kuang, for her
unwavering love, support, and company in all these years.

I am immensely thankful to my mentor, Dr. Julio Camarero for his guidance and support.
He has offered me valuable expertise in the field of protein chemistry and engineering
that has been indispensable to my research. I would like to thank him for providing me
various opportunities to expand my knowledge and experience in the doctoral training.

I would like to extend my gratitude to all my colleagues and friends in the Camarero lab. It
has been such a blessing to work with every one of them. I appreciate their friendship and
willingness to always lend a helping hand.

I would also like to thank my committee members, Drs. Bangyan Stiles, Curtis Okamoto,
Matthew Pratt for their time, guidance and support.

I am grateful to USC School of Pharmacy for providing me with great training and learning
opportunities as a graduate student.

4

Table of Contents
DEDICATION ...................................................................................................................................................... 2
ACKNOWLEDGEMENTS .............................................................................................................................. 3
ABSTRACT.......................................................................................................................................................... 6
ABBREVIATIONS ............................................................................................................................................. 7
CHAPTER ONE Introduction .................................................................................................................. 9
1.1 Introduction to cyclotides ....................................................................................................... 11
1.1.1 Discovery and history of cyclotides ...................................................................... 11
1.1.2 Structure of cyclotides .............................................................................................. 12
1.1.3 Synthesis of cyclotides .............................................................................................. 14
1.1.4 Properties and biological activity of cyclotides ................................................ 25
1.1.5 Using cyclotides as scaffolds for grafting biologically active peptides ... 31
1.1.6 Summary ......................................................................................................................... 41
1.2 Targeting the p53-HDM2 interaction for cancer therapy........................................... 41
1.2.1 General introduction of p53 ..................................................................................... 41
1.2.2 Cellular function of p53 ............................................................................................. 42
1.2.3 p53 regulation by the HDM2 and HDMX complex ......................................... 42
1.2.4 Targeting the p53-HDM2 interaction for cancer therapy ............................. 44
CHAPTER TWO Targeting the p53-HDM2 complex using a grafted cyclotide ................ 47
2.1 Design and synthesis of cyclotide MCo-PMI ................................................................. 48
2.1.1 Design of PMI grafting into MCoTI-I scaffold ................................................... 48
2.1.2 Synthesis and folding of MCo-PMI cyclotides ................................................. 50
2.2 In vitro characterization of MCo-PMI cyclotides ........................................................... 54
2.2.1 Direct binding assay ................................................................................................... 54
2.2.2 Inhibition assay ............................................................................................................. 55
2.3 Structural characterization of the MCo-PMI by NMR ....................................... 56
2.4 Serum stability of cyclotide MCo-PMI .............................................................................. 60
2.5 Biological activity using cell based assays ..................................................................... 61
2.5.1 Cell availability assay ................................................................................................. 61
2.5.2 Characterization of the mechanism of action of cyclotide MCo-PMI ...... 64
2.5.3 Other cell based assays ........................................................................................... 66
2.6 Characterization using a colon carcinoma mouse xenograft model .................... 67
5

2.7 Optimization of MCo-PMI activity using a molecular constrain approach ......... 70
2.8 Conclusion ................................................................................................................................... 73
2.9 Materials and methods ........................................................................................................... 73
2.9.1 Materials and instrumentation ................................................................................ 73
2.9.2 Molecular cloning and protein expression ......................................................... 74
2.9.3 In vitro characterization assays ............................................................................. 82
2.9.4 Serum stability and serum protein binding assay........................................... 83
2.9.5 Cell based assays ....................................................................................................... 83
2.9.6 Mice xenografts model experiments .................................................................... 85
CHAPTER THREE Design of a high-throughput screening system for the selection of
bioactive cyclotide targeting the p53-HDM2 interaction .................................................................. 87
3.1 Expression of cyclotides expression in E. coli cells using a protein trans-splicing
approach .................................................................................................................................................... 87
3.2 Design of a barnase-based screening system ............................................................. 89
3.3 In vitro characterization of the barnase screening system ...................................... 92
3.4 In cell characterization of the barnase screening system ........................................ 94
3.5 Testing the barnase screening system with a mock plasmid library .................... 99
3.6 Design of a combinatorial MCoTI-I based library ...................................................... 102
3.7 Screening of the MCoTI-I library using the barnase cell-based reporter ......... 104
3.8 Characterization of the sequences selected in the 3
rd
round of enrichment .. 107
3.9 Design of a MCo-PMI based library ................................................................................ 108
3.10 Screening of the MCo-PMI based library ................................................................... 109
3.11 Materials and methods ....................................................................................................... 110
3.11.1 Materials and instrumentation. ........................................................................... 110
3.11.2 Experimental methods .......................................................................................... 110
CHAPTER FOUR Discussion ............................................................................................................. 124
4.1 Discussion of molecular grafting method ...................................................................... 124
4.2 Discussion of the high-throughput barnase-based cell reporter ........................ 126
4.3 Future directions ..................................................................................................................... 129
References ....................................................................................................................................................... 131

6

ABSTRACT
The success of protein-based therapeutics has revolutionized drug development. Unlike
small molecule drugs, peptide and protein-based therapeutics can target defective
protein-protein interactions involved in human diseases with high selectivity and
specificity. Despite their success, however, there are numerous stability and delivery
issues associated with their use as therapeutic agents. As an emerging macromolecule,
cyclotide has excellent pharmacological properties.
Here, we explored the possibility of cyclotide to target protein-protein interactions. The
overexpression of HDM2 and HDMX is a common mechanism used by many tumor cells
to inactive the p53 tumor suppressor pathway promoting cell survival. Targeting HDM2
and HDMX has emerged as a validated therapeutic strategy for treating cancers with
wild-type p53. Small linear peptides mimicking the N-terminal fragment of p53 have been
shown to be potent HDM2/HDMX antagonists. The potential therapeutic use of these
peptides, however, is limited by their poor stability and bioavailability.
In this project, we first report the engineering of the cyclotide MCoTI-I to efficiently
antagonize intracellular p53 degradation. The resulting cyclotide MCo-PMI was able to
bind with low nanomolar affinity to both HDM2 and HDMX, showed high stability in human
serum, and was cytotoxic to wild-type p53 cancer cell lines by activating the p53 tumor
suppressor pathway both in vitro and in vivo.
We also report a cell based high-throughput screening method for selection of
protein-protein interaction inhibitors using a library based on the cyclotide MCoTI-I. This
screening method has been proved able to select p53-HDM2 antagonist peptide both in
vitro and in living E. coli cell. Both mock library and real libraries have been screened by
this split barnase method and some preliminary sequences have been analyzed. These
features make the cyclotide MCoTI-I an optimal scaffold for targeting intracellular
protein−protein interactions.
7

ABBREVIATIONS
AFM atomic force microscopy
AziF p-azidophenylalanine
Boc tert-butyloxycarbonyl
CBD chitin-binding domain
CCK circular Cys-knot
DBCO dibenzocyclooctyl
DMSO dimethyl sulfoxide
DOTA 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid
dsDNA double stranded DNA
EPL expressed protein ligation
ES-MS electrospray mass spectrometry
FACS fluorescence-activated cell sorting
FITC fluorescein isothiocyanate
FMDV foot-and-mouth disease virus)
Fmoc 9-fluorenylmethoxycarbonyl
GSH reduced glutathione
HLE human leukocyte elastase
i.p. intraperitoneal injection
IPTG isopropyl-β-D-thiogalactopyranoside
MACS magnetic-activated cell sorting
MCR4 melanocortin receptor 4
MIC minimal inhibitory concentration
MOG myelin oligodendrocyte glycoprotein
MRM multiple reaction monitoring
MS multiple sclerosis
8

MTT 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide
NCL native chemical ligation
Ni
2+
-NTA Ni
2+
-nitrilotriacetic acid
NMR nuclear magnetic resonance
OD optical density
OmeF p-methoxyphenylalanine
PCAs protein fragment complementation assays
PI propidium iodide
PPIs protein-protein interactions
PTS protein trans-splicing
SFTI-1 sunflower trypsin inhibitor 1
SICLOPPS split intein circular ligation of protein and peptides
SPPS solid-phase peptide synthesis
TCA trichloroacetic acid
TEV tobacco etch virus
TFA trifluoroacetic acid
Txn thioredoxin
VEGF-R2 vascular endothelial growth factor receptor 2

9

CHAPTER ONE Introduction
The success of protein-based therapeutics has revolutionized drug development.
Unlike small molecule drugs, peptide and protein-based therapeutics can target defective
protein-protein interactions involved in human diseases with high selectivity and
specificity, which means peptide-based drugs normally have fewer side effects. Despite
their success, however, there are still numerous stability and delivery issues associated
with their use as therapeutic agents. For example, monoclonal antibodies, one the most
successful protein-based therapeutics with several blockbuster drugs on the market and
much more in clinical development, can only target extracellular molecular targets due to
their inability to cross biological membranes. They are also extremely expensive to
produce and are not orally bioavailable due to their susceptibility to proteolytic
degradation. These issues have led to the exploration of alternative protein scaffolds as a
source for novel types of protein-based therapeutics. In response to this important
challenge, our laboratory is exploring the development of engineered cyclotides to
antagonize specific protein-protein interactions inside cells.
Cyclotides are a new emerging family of large plant-derived backbone-cyclized
polypeptides (≈30 amino acids long) that share a disulfide-stabilized core (three disulfide
bonds) characterized by an unusual knotted structure (Fig. 1.1)(Garcia & Camarero,
2010). They have several characteristics that make them ideal drug development tools
(Andrew Gould & Camarero, 2017). First, they are remarkably stable to chemical, thermal
and proteolytic degradation due to the cyclic topology and cysteine knot structure
(Colgrave & Craik, 2004). Second, they are small, making them readily accessible to
chemical synthesis (Daly, Love, Alewood, & Craik, 1999). Third, they can be encoded
within standard cloning vectors and be expressed in bacterial or mammalian cells (R.
Kimura & Camarero, 2005). Fourth, they are amenable to substantial sequence
variation(Austin, Wang, Puttamadappa, Shekhtman, & Camarero, 2009). These
10

characteristics make them ideal substrates for molecular evolution strategies to enable
generation and selection of compounds with optimal binding and inhibitory characteristics
(Garcia & Camarero, 2010; Andrew Gould & Camarero, 2017; Krishnappa Jagadish &
Camarero, 2010). Finally, cyclotides have been shown to be orally bioavailable. For
example, the first cyclotide to be discovered, kalata B1, is an orally effective uterotonic,
and other cyclotides have been shown to cross eukaryotic cell membranes (Cascales et
al., 2011; Contreras, Elnagar, Hamm-Alvarez, & Camarero, 2011). Cyclotides thus appear
as promising leads or frameworks for peptide drug design (Andrew Gould & Camarero,
2017).
In this thesis, the use of cyclotide MCoTI-I is explored as a scaffold to target intracellular
p53-HDM2 interaction, which has been characterized as a therapeutic target for cancer
treatment(Honda, Tanaka, & Yasuda, 1997; Momand, Zambetti, Olson, George, & Levine,
1992). There are two approaches will be used in this thesis. One involves grafting a
bioactive peptide antagonist into cyclotide scaffold. This novel engineered cyclotide will
target the interaction of p53 with both HDM2 and HDMX. The other approach will involve
design a cyclotide based library and use a high-throughput screening system to select
active compounds from this library. Those two innovative approaches contribute to the
development of a novel type of peptide-based therapeutic able to target intracellular
protein-protein interactions. The use of the cyclotide scaffold will enable these peptide
antagonists to have the required increased stability, cellular membrane penetration, and
serum clearance needed to be considered as viable drug development candidates. The
cell-based technology developed here could be easily adapted to screen for antagonists
for other disease relevant protein-protein interactions, which makes this project important
in protein/peptide based drug development.
11

1.1 Introduction to cyclotides
1.1.1 Discovery and history of cyclotides
The name “cyclotide” was formally coined by Craik in 1999 (Craik, Daly, Bond, & Waine,
1999), in which “cyclo” means head to tail cyclic backbone structure and “tide” stands for
the peptide. This word was used to describe a family of plant-derived backbone cyclized
polypeptides containing six conserved cysteine residues forming a Cys-knot topology. The
first cyclotide was discovered in Africa in early 1970 by Gran (L Gran, 1970). In central
Africa, people use a traditional remedy called “kalata-kalata” to facilitate childbirth for
pregnant women. This medicinal tea is prepared by boiling the above ground part of plant
Oldenlandia affinis in water, which indicates the active compound in this tea is resistant to
degradation to water boiling temperatures. Gran discovered the most abundant and active
component of the tea was a polypeptide that was named “kalata B1”. After that, he
characterized the pharmacological properties of kalata B1, as well as determined the
amino acid composition and the approximate size of this peptide (L. Gran, 1973a, 1973b).
However, due to the lack of modern spectroscopic techniques available in the 1970s for
the analysis of polypeptides, he was only able to establish the peptide composition and its
cyclic or N-capped nature. It was not until 1995, the structure study by Saether unveiled
the backbone-cyclized (or circular) and Cys-knotted structure of kalata B1 (Saether et al.,
1995). Coincidently, in 1990s several other polypeptides with similar circular and
cystine-knotted structure were discovered from a variety of plant species by different
groups (Gustafson et al., 1994; Schö pke, Hasan Agha, Kraft, Otto, & Hiller, 1993;
Witherup et al., 1994). The three-dimensional structure of those peptides showed
similarity to kalata B1 (Craik et al., 1999), indicating that kalata B1 was not the only cyclic
cystine-knotted peptide in the plant. Since then, around 300 different cyclotide sequences
have been found (Kaas & Craik, 2010). Cyclotides have been detected among different
plate tissues and one plant may have different types of cyclotide (Trabi & Craik, 2004;
12

Trabi et al., 2004). Also, cyclotides have been found different plant species in major two
families, the violet family (Violaceae) and in few species of the coffee family (Rubiaceae).
It is estimated that there are more than tens of hundreds of cyclotide sequences (Gruber
et al., 2008), which makes cyclotide one of the potential largest marcoprotein family in the
plant.
1.1.2 Structure of cyclotides
The structure of cyclotides is shown in Figure 1.1. The cyclotide framework is formed by
a circular backbone topology (shown in blue line). This circular structure is further
stabilized by three disulfide bonds (shown in red line) formed by six conserved Cys
residues. The first disulfide bond is formed between CysI and CysIV, the second between
CysII and CysV and the third disulfide bond between CysIII and CysVI. These three
disulfides give rise to six loops (Fig. 1.1)(Garcia & Camarero, 2010)
The cyclotide kalata-B1 is stable and biologically active after being extracted by boiling
water to make the uterotonic tea used to accelerate childbirth by local natives in central
Africa (L Gran, 1970). Proteins are normally denatured when exposed under those
conditions resulting in loss of their bioactivity. The high stability of the cyclotides is due to
circular Cys-knot (CCK) topology. As shown in Fig. 1.1 the Cys-knot topology is formed
by disulfides CysI-CysIV and CysII-CysV, which form a cycle that is penetrated by the
third disulfide bond CysIII-CysVI locking the cyclotide in extremely compact structure
(Goransson & Craik, 2003).
Cyclotides can be classified into three different subfamilies based on sequence
differences, these include the Möbius, bracelet, and trypsin inhibitor subfamilies (Fig. 1.1).
Most cyclotides belong to the Möbius and bracelet subfamilies and there are only a small
number of cyclotide in the trypsin inhibitor subfamily. The Möbius and bracelet subfamily
have a similar structure. The differences between Möbius and bracelet subfamilies are
13

slight sequence differences which also include the presence of cis-proline residue in loop
5 in the Möbius cyclotides while bracelet cyclotides have all the peptide bonds in trans
configuration (C. Jennings, West, Waine, Craik, & Anderson, 2001; Rosengren, Daly,
Plan, Waine, & Craik, 2003).

Figure 1.1. Structural classification of cyclotides. The structures of representative cyclotides from
the bracelet (kalata B1; PDB ID: 1NB1), Möbius (cycloviolacin O1, PDB ID: 1NBJ), and trypsin
inhibitor (MCoTI-II, PDB ID: 1IB9) subfamilies. Disulfide bonds are shown in red line. The blue line
denotes the circular backbone. All these cyclotides are separated into 6 loops by these three
disulfide bonds. The picture was taken from Li (Y . Li, Bi, & Camarero, 2015).
The trypsin inhibitor subfamily differs from the other two subfamilies in the sequence
and size of different loops (Heitz et al., 2001; Hernandez et al., 2000). Unlike the
cyclotides in the other two subfamilies, which have a more rigid structure, the cyclotides
from the trypsin inhibitor subfamily have a longer loop 1 and a relatively more flexible
loop6. According to a recent NMR study of MCoTI-I, the structure of MCoTI-I is well
defined in solution and the flexibility increases a little in loop 1 and loop 6 (Puttamadappa,
Jagadish, Shekhtman, & Camarero, 2010). Consist result was obtained by a later
14

dynamic study of MCoTI-II (Daly et al., 2013).
Despite all cyclotides may have different sequences in different loops, they all adopt
a CCK motif, which indicates the high tolerance of the cyclotide scaffold to sequence
variation. The cyclotide MCoTI-I has been subjected to amino acid scanning at all the
positions except for the Cys residues and loop 6. Only two of these mutants showed
inefficient folding (Austin et al., 2009). Furthermore, the chemical synthesis of a complete
suite of Ala mutants of cyclotide kalata B1 has also been reported. Only two of the 23 Ala
mutants studied lost the ability to fold correctly (Simonsen et al., 2008).
1.1.3 Synthesis of cyclotides
1.1.3.1 Chemical synthesis of cyclotides
Cyclotides are small size cyclic peptides, approximately 30 amino acids long. Peptides
of this size can be chemically synthesized by solid phase peptide synthesis (SPPS)
(Marglin & Merrifield, 1970). Currently, cyclotides from all the three subfamilies have been
chemically synthesized using either Boc- or Fmoc-based chemistries (T. L. Aboye et al.,
2012; Contreras et al., 2011).
As shown in Fig. 1.2, backbone cyclized peptides can be produced by using an
intra-molecular version of the native chemical ligation (NCL) reaction. Intermolecular NCL
has been widely used for the synthesis of long peptides and proteins (Ayers et al., 1999; J.
A. Camarero & Mitchell, 2005; Dawson & Kent, 2000). In this reaction, a peptide
containing a C-terminal -thioester group reacts with a peptide containing an N-terminal
Cys residue, forming a native peptide bond (Fig. 1.2A). When the thioester group and
Cys residue are in the same polypeptide the reaction happens intramolecularly resulting
in a backbone cyclized peptide product (Fig. 1.2B). This cyclization technique has been
widely used for the synthesis of circular peptides and proteins (T. L. Aboye & Camarero,
2012).
15

The most commonly used method to produce -thioester linear peptide precursors is
using the 3-mercaptopropionamide linker (J. Camarero & Muir, 1997). This linker is stable
under the condition used during Boc-based peptides synthesis and provides -thioester
in excellent yield. This approach was first used by Daly and Tam to synthesize cyclotides
kalata B1, circulins A/B and cyclopsychotride to study their folding and biological activity
(Daly et al., 1999; Tam, Lu, Yang, & Chiu, 1999). Despite the Boc-based peptide
synthesis requires repeated use of trifluoroacetic acid (TFA) and highly toxic anhydrous
hydrogen fluoridein for the final cleavage step, this approach is still widely used to make
cyclic peptides. For example, some bioactive cyclotides targeting extracellular target
have been synthesized using this approach (Chan et al., 2011; Eliasen et al., 2012).
Due to the limitations of the Boc-based method, different approaches using Fmoc-based
chemistry for the synthesis of -thioester linear cyclotide precursors have been
developed (J. A. Camarero & Mitchell, 2005). The use of the thioester linker is limited by
its poor stability to the conditions employed during the peptide assembly using the
Fmoc-based approach. Thus, a more stable linker is required for this approach to be used
for the synthesis of linear cyclotide precursors.
The first method to synthesize peptide -thioester by using Fmoc-based SPPS was
reported by Futaki (Futaki et al., 1997). In this approach, the peptide -thioester was
prepared using partially protected precursor. By using Fmoc-based peptide synthesis, the
protected peptide precursor was prepared on a 4-chlorotrityl (4-Cl-Trt) resin. After
cleavage, the C-terminal carboxylic group was thioesterified with the appropriate thiol and
coupling reagent to generate the corresponding C-terminal -thioester. Recently, this
approach was used to synthesize the bracelet cyclotide cycloviolacin O2 (Leta Aboye,
Clark, Craik, & Göransson, 2008). However, it is important to point out that that
epimerization of C-terminal residue may occur during the thioesterification reaction. A
study has shown that the epimerization depends on the coupling reagent used for
thioesterification (von Eggelkraut-Gottanka, Klose, Beck-Sickinger, & Beyermann, 2003).
16

Another approach using an acylsulfonamide liker was reported by Ingenito for the
production of -thioester linear peptides (Ingenito, Bianchi, Fattori, & Pessi, 1999). The
acylsulfonamide liker is highly stable to the basic or nucleophilic conditions used in
Fmoc-based SPPS and can be activated by treatment with trimethylsilydiazomethane
(TMS-CHN2) or iodoacetonitrile to provide an N

-alkyl acylsulfonamide for nucleophilic
attack reaction with appropriate thiols to provide the corresponding -thioester peptide.
This approach was first used by Thongyoo et al to synthesize cyclotide MCoTI-II
(Panumart Thongyoo, Tate, & Leatherbarrow, 2006).
17

Figure 1.2. Backbone cyclization of polypeptides using native chemical ligation (NCL). (a) The
principle of NCL. (b) Intramolecular NCL leads to the formation of a backbone-cyclized
polypeptide.
18

In this paper, the end of loop 3 of MCoTI-II was chosen as ligation site. The same group
has used this approach to synthesize a MCoTI-based cyclotide with potent -tryptase and
human leukocyte elastase inhibitory activity (Gray et al., 2014; P. Thongyoo, Bonomelli,
Leatherbarrow, & Tate, 2009). This approach has also been used by Camarero group for
the synthesis of engineered cyclotides targeting extracellular protein CXCR4 (T. L. Aboye
et al., 2012) or intracellular protein-protein interactions (Ji et al., 2013). In both cases, the
N-terminal region of loop 6 was chosen as cyclization site. The linear peptide -thioester
precursors are produced by thiolysis and then cyclized and folded in a “one-pot”
approach using reduced glutathione (GSH) as redox agent at pH 7.2 in very high yield.
This synthetic scheme has been recently used for the parallel synthesis of a
cyclotide-based amino acid scanning library using a “tea-bag” approach (T. Aboye, Kuang,
Neamati, & Camarero, 2015). This work also developed a very efficient protocol using
active thiol Sepharose-based chromatography for removal of non-folded or partially
folded cyclotide which enables its use in mixtures of cyclotides generated during the
generation of positional scanning libraries(Dooley & Houghten, 1993).
1.1.3.2 Recombinant expression of cyclotides
Despite the fact that the chemical synthesis of cyclotides has been well explored and
a number of different approaches have been developed, the advances in molecular and
protein biology fields have made possible the biosynthesis of cyclotides using standard
expression systems (Y . Li et al., 2015; Sancheti & Camarero, 2009). The biological
synthesis of cyclotides has many advantages over chemical synthesis, like epitope
labeling, or the possibility of screening of genetically-encoded libraries (Austin, Kimura,
Woo, & Camarero, 2010). Different approaches have been described for the biological
synthesis of cyclotides, including the use of expressed protein ligation (EPL) (R. Kimura &
Camarero, 2005), intein-mediated protein trans-splicing (PTS) (K. Jagadish et al., 2013),
thiol-induced selective N->S acyl transfer (Cowper, Craik, & Macmillan, 2013) and
19

enzyme-mediated ligation.
Expressed protein ligation (EPL)
The first method is called expressed protein ligation (EPL). This approach relies on a
mechanism similar to that of native chemical ligation described previously. Backbone
cyclization through native chemical ligation requires a C-terminal -thioester and an
N-terminal cysteine within the same linear precursor, which then react in an
intramolecular fashion to provide a circular polypeptide (Fig. 1.2). In the EPL approach,
an engineered intein is used to introduce a C-terminal a-thioester in a recombinantly
produced polypeptide (see a reference for a recent review (Y. Li et al., 2015)). This
cyclization strategy was first reported in a study of the SH3 domain of the c-Crk protein in
vitro (J. A. Camarero & Muir, 1999). In this work, the SH3 domain was fused to the
N-terminus of a modified Saccharomyces cerevisiae vacuolar membrane ATPase intein.
A MIEGRC motif, which contains a factor Xa protease recognizing sequence was
introduced at the N-terminus of SH3 domain. This fusion protein was expressed in
Escherichia coli. The precursor protein was purified, cleaved with factor Xa protease,
resulting in an N-terminal Cys-containing SH3-intein fusion protein. This fusion protein
reacted in an intramolecular fashion to yield a backbone cyclized SH3 domain protein.
This reaction was performed under physiological conditions and was extremely efficient
and clean. Based on the high-efficiency cyclization of this approach an in vivo study was
also carried out later (J. A. Camarero, Fushman, Cowburn, & Muir, 2001). In this study,
SH3-intein fusion precursor protein was expressed in E. coli cells and expected cyclic
protein was found in the cell lysate, which indicates the success of EPL approach in living
cells.
Later, this approach was also used alter for the recombinant expression of cyclotides
using a standard bacterial expression system. The first cyclotide expressed in bacterial
cells was kalata B1 (R. H. Kimura, Tran, & Camarero, 2006). In this work, several
cyclotide linear precursors were fused to the N-terminus of a modified Saccharomyces
20

cerevisiae (Sce) vacuolar membrane ATPase intein. An N-terminal Met residue was
fused to the N-terminus of the cyclotide precursor. Once this precursor is expressed
inside the cell, the N-terminal Met residue is removed in situ by the endogenous bacterial
methionine aminopeptidase MAP, providing the required N-terminal Cys residue for the
cyclization. Other protease recognition motifs also have been used for the generation of
the required N-terminal Cys residue, including factor Xa (Erlandson et al. 1996), TEV
protease (Tolbert & Wong, 2002). Co-expression of the protease with the cyclotide-intein
precursor removes the leading peptide sequence exposing the N-terminal cysteine
required for backbone cyclization. The recombinant precursor with the N-terminal Cys
can be cyclized and oxidatively folded in a “one-pot” reaction in phosphate buffer at pH7.2
with reduced GSH (Fig.1.3). In this work, different kalata B1 mutants were also made by
using this approach. Different mutants were designed to test the importance of several
residues near the Cys residues in the folding of corresponding cyclotide mutants. As
shown in Fig. 1.4, the “one-pot” cyclization/folding approach was very efficient in all the
mutants with yields ranging from 20% to 60%, which indicates the robustness of this
methodology.
This intein-mediated backbone cyclization technique has been applied to obtain
other cyclotides such as MCoTI-I (Austin et al., 2009). In this work, a more efficient
Mycobacterium xenopi (Mxe) GyrA intein was used. In this study, this EPL approach was
also used to generate a small library based on the cyclotide MCoTI-I scaffold. All residues
except cysteines were replaced by different residues. Most of the mutants were able to
adopt a native cyclotide fold indicating that this cyclotide can be mutated without
significant change in the structure and function. These results open the intriguing
possibility of using cyclotide scaffold to generate large genetically encoded libraries for
in-cell high throughput screening techniques.
Intein-mediated backbone cyclization of polypeptides has also been used for the
biosynthesis of other naturally occurring disulfide-rich backbone cyclized polypeptides,
21

such as Bowman-Birk inhibitor SFTI-1 (Austin et al., 2010), -defensins (Garcia et al.,
2011) and θ-defensins (Conibear et al., 2014; A. Gould et al., 2012), indicating the high
versatility of this approach for the biosynthesis of backbone cyclized polypeptides.

Figure 1.3. Intein-mediated backbone cyclization of a kalata B1-based cyclotide linear precursor.
The cyclotide-intein fusion construct is recombinantly expressed in E. coli cells and can be
cyclized and folded either in vitro (R. H. Kimura et al., 2006) or in vivo (Austin et al., 2009). The
figure is taken from Li (Y. Li et al., 2015).
22

Figure 1.4. In vitro production of different kalata B1-based cyclotide mutants using an
intein-mediated backbone cyclization (R. H. Kimura et al., 2006). (a) The intein kalata B1
constructs were expressed in E. coli, purified, and cyclized/folded using “one-pot’ approach in
aqueous buffer at pH 7.2 containing reduced GSH. (b) Sequences of all the kalata B1 mutants
cyclized using this approach. (c) Analytical high-performance liquid chromatography traces of the
cyclization/folding crude after 18h. The peak corresponding to the folded cyclotide is marked with
an asterisk. The figure is taken from Kimura (R. H. Kimura et al., 2006).

Protein trans-splicing (PTS)
Another method used to generate cyclotides is using protein trans-splicing (PTS). This
technique was first published by Benkovic using the Synechocystis sp PCC6803 DnaE
split intein to generate cyclized polypeptide (Scott, Abel-Santos, Wall, Wahnon, &
23

Benkovic, 1999). In this case, the protein splicing domain or intein is naturally split into
two distinct fragments, called N-intein (IN) and C-intein (IC). These two fragments are
non-functional individual; however, they can interact with each other in a highly specific
way to produce a functional protein splicing domain in trans. When the IN and IC
fragments are fused to the C- and N-terminus of a linear polypeptide, respectively, the
consequent protein trans-splicing reaction generates a backbone cyclized polypeptide
(Fig. 1.5). This method, also called SICLOPPS (split intein circular ligation of protein and
peptides) has been used to generate different cyclic peptides (Tavassoli & Benkovic,
2007) and proteins (Deschuyteneer et al., 2010). Recently, Schulty and co-workers
reported a novel technique using this PTS technology in combination with nonsense
codon suppressor tRNA technology to introduce non-natural amino acid into cyclic
hexapeptides to build a large library which was used to screen HIV protease inhibitors
(Young et al., 2011). These studies demonstrate the possibility of using PTS in the
biosynthesis of cyclotides as well as introducing non-natural amino acid into cyclotides.
The first cyclotide obtained using this approach was reported by our group (K.
Jagadish et al., 2013). In this study, several MCoTI-I-based cyclotides and containing
non-natural amino acids were expressed using both intein based approaches, EPL and
PTS approaches. Two non-natural amino acids, p-methoxyphenylalanine (OmeF) and
p-azidophenylalanine (AziF), were introduced into the wild-type sequence of cyclotide
MCoTI-I. In this work, the highly efficient Nostoc puntiforme (Npu) PCC73102 DnaE split
intein (Iwai, Zuger, Jin, & Tam, 2006) was used for the cyclization instead of Ssp DnaE
split intein. The Npu DnaE intein is one of the most efficient intein reported and it has one
of the highest reported rates of PTS ( 1/2~60s), a high splicing efficiency (80%), and high
sequence tolerance at the extein-intein junctions (Bi, Li, Shekhtman, & Camarero, 2017;
Borra, Dong, Elnagar, Woldemariam, & Camarero, 2012). In cell expression was highly
efficient providing an intracellular concentration in the range of 20-40µM. It is worth to
highlight that the high efficiency of PTS-mediated cyclization combined with the use of
24

non-sense suppressing orthogonal tRNA/synthetase technology enabled for the first time
the in-cell production of cyclotides containing non-natural amino acids.

Figure 1.5. In cell expression of natively folded cyclotide using intein-mediated PTS (K. Jagadish
et al., 2013). The linear cyclotide are fused to the IN and IC fragments are, the protein trans-splicing
reaction generates a backbone cyclized polypeptide.
25

The azido containing amino acids, which can react with dibenzocyclooctyl
(DBCO)-containing fluorescence probes, was used to facilitate in-cell production of
fluorescently-labeled cyclotides. These findings open the possibility for in vitro and
potentially in cell screening of genetically encoded libraries using high-throughput
cell-based optical screening approaches (K. Jagadish et al., 2013). Recently, this method
has been used for biosynthesis of other cyclic peptide in E. coli, like SFTI-I (Harris et al.,
2016) and RTD-1 (Bi et al., 2017) and the production of cyclotide in eukaryotic
microorganism S. cerevisiae yeast (K. Jagadish et al., 2015).
1.1.4 Properties and biological activity of cyclotides
1.1.4.1 Properties of cyclotides
The first discovered cyclotide, kalata B1, was used as an oral uterotonic traditional
remedy in central Africa (L. Gran, 1973a), therefore indicating for the first time the high
stability of these type of peptides. After the structure was solved (Figure 1-1), The highly
compact cyclic cystine-knot (CCK) motif seems to be responsible for the high stability of
these peptides. The thermal, chemical and enzymatic stability of kalata B1 was first
reported by the Craik group (Colgrave & Craik, 2004). In this study, the stability of the
prototypic cyclotide kalata B1 was compared to several related peptides, including the
natural variant kalata B2, acyclic mutants in which backbone was linearized at different
locations in the sequence, and a mutant variant with only two disulfide bonds. The natively
folded form of kalata B1 was stable in aqueous solution at temperatures around the boiling
point of water, to the high concentration of chaotropic agents, and to enzymatic
degradation, as well as to acid hydrolysis with aqueous 0.5 M HCl. On the other hand, the
fully reduced and linearized cyclotides showed decreased stability and showed
significantly higher susceptibility to enzymatic degradation. These findings on the stability
26

of kalata cyclotides confirm the role of the CCK topology on the stability of the cyclotide
scaffold. This was also confirmed in different studies using MCoTI-based cyclotides (T. L.
Aboye et al., 2012; Ji et al., 2013). In these studies, natively folded MCoTI-cyclotides were
shown to have higher ex-vivo stability in serum ( 1/2~40h) when compared to their
linearized fully reduced versions ( 1/2<1h), which indicates that the cyclization and the
Cys-knot are critical for the cyclotide stability. The stability of the trypsin inhibitor subfamily
cyclotide MCoTI-II has been also shown to be very stable in serum with a half-life bigger
than 24 h (Y. H. Huang, Chaousis, Cheneval, Craik, & Henriques, 2015). A recent study on
the biodistribution of cyclotide MCoTI-II distribution after intravenous injection was also
recently reported (Wang, Stalmans, De Spiegeleer, & Craik, 2016). The cyclotide MCoTI-II
has been shown better pharmacokinetic profile (less elimination) compared to linear
peptide drug dermorphin in this study, therefore indicating the robustness of cyclotide in
pharmaceutic applications.
Some cyclotides have also shown to be able to be uptaken by mammalian cells. MCoTI-II
was the first cyclotide shown able to be internalized by cells (Greenwood, Daly, Brown,
Stow, & Craik, 2007). In this paper, fluorescent labeled MCoTI-II was incubated with a
macropinocytosis marker tetramethylrhodamine-dextran. Colocalization of MCoTI-II and
dextran was detected by fluorescence microscopy on fixed-cells indicating that the major
endocytosis pathway for MCoTI-II was micropinocytosis. More recently, cyclotide
MCoTI-I was also shown to enter mammalian cells (Contreras et al., 2011). In this study,
the cellular uptake of cyclotide MCoTI-I was monitored by real-time confocal fluorescence
microscopy imaging in live HeLa cells. The cellular uptake process was shown to be
temperature-dependent and could be reversibly inhibited at 4 ° C, thereby proving the
internalization step follows an active mechanism. Internalized cyclotide was shown to be
colocalized with the fluid-phase endocytic marker dextran. Disruption of cellular actin by
Latrunculin B did not completely impair MCoTI-I uptake in HeLa cells. As actin is critical
for cell micropinocytosis, this result indicates the macropinocytosis may not be the only
27

pathway for MCoTI-I uptake in HeLa cells and other pathways could also be involved in
the cyclotide cellular uptake.
The Mö bius cyclotide kalata B1 has been also shown to cross the cellular membrane
(Cascales et al., 2011; Henriques et al., 2015). The mechanism of kalata B1 and MCoTI-II
cell uptake was recently studied (Cascales et al., 2011). In this study, the interaction
between both cyclotides and different model lipid membranes was studied. Kalata B1 was
shown to interact directly with the membrane by binding to the phosphatidylethanolamine
(PE) phospholipid leading to membrane bending and vesicle formation. On the other hand,
MCoTI-II interacted with negatively charged phosphoinositide and entered cells via
macropinocytosis. A recent study has shown that cyclotide kalata B1 can enter
mammalian cells through both active endocytosis and passive direct membrane
translocation (Henriques et al., 2015). It has also been shown that the binding ability of
kalata B1 to phosphatidylethanolamine phospholipid was correlated to its membrane
permeability. The orientation of cyclotide kalata B1 embedded in cell membrane was
illustrated by solid-state NMR recently (Grage et al., 2017). However, it is important to
point out that kalata B1 has significant higher cytotoxicity when compared to that of
cyclotide MCoTI-II (Greenwood et al., 2007). MCoTI-II caused <10% cytotoxicity in
RAW264.7 cell line at 50 µ M concentration. In contrast, kalata B1 was highly cytotoxic at a
concentration of 20 µ M. This could limit the use of kalata B1 in pharmaceutical
applications. The mechanism of cyclotide membrane permeability was summarized in a
recent review as shown in Fig. 1.6 (Troeira Henriques & Craik, 2017).
More than 300 different natural cyclotide sequences have been found thus far. The
large diversity of cyclotide sequence with similar cyclic cystine-knot fold demonstrates the
tolerance to sequence variation in the cyclotide loops. Amino acid scanning of cyclotide
MCoTI-I (Austin et al., 2009) and kalata B1 (Simonsen et al., 2008) have been performed
showing that the folding of both cyclotides was only interfered by a couple of mutants
28

while the rest of mutants were still able to adopt a native cyclotide fold. The high
tolerance of the loops to sequence variation enables the introduction of other bioactive
peptide sequences into the cyclotide backbone opening the possibility of using this
scaffold for the design of novel bioactive peptides with advantageous pharmacological
properties.

Figure 1.6. The mechanism for the internalization of cyclotides. (a) Internalization mechanism for
Mö bius and bracelet cyclotides illustrated with kalata B1. (b) Internalization of trypsin inhibitor
cyclotides. The figure is taken from (Troeira Henriques & Craik, 2017).
1.1.4.2 Biological activity of cyclotides
Cyclotides are naturally found in many plants from the violet family (Violaceae) and f the
coffee family (Rubiaceae) (Trabi & Craik, 2004; Trabi et al., 2004). It is well accepted that
the main biological function of cyclotides in the Mö bius and bracelet subfamilies is plant
defense. The pesticide activity of several cyclotides have been reported (C. Jennings et
al., 2001; Mylne, Wang, van der Weerden, & Craik, 2010). Kalata B1 cyclotide was fed at
the same concentration as expressed naturally in plant leaf to the larvae of Helicoverpa
29

armigera, a common cotton and corn insect pest (C. Jennings et al., 2001). Caterpillars
fed with a diet containing kalata B1 showed remarkable slow growth and development
compared to those larvae fed the control diet. A later study confirmed that the toxicity of
cyclotide kalata B1 to the larvae is caused by the disruption of the microvilli, which causes
swelling, and ultimately ruptures the epithelial cells in the gut of the insect (Mylne et al.,
2010). Other cyclotides have been also shown to possess insecticidal activity, including
kalata B2 (C. V. Jennings et al., 2005) and Cter M (Poth, Colgrave, Lyons, Daly, & Craik,
2011). The toxicity of other Mö bius and bracelet cyclotides to other larval insects has
been reported as well. For example, a novel cyclotide, parigidin-br1, isolated from
Palicourea rigida has been shown to have potent insecticidal activity against the neonate
larvae of Diatraea saccharalis (Pinto et al., 2012). Cyclotides also have shown toxic
activity against non-insect pests like the Golden Apple snail, Pomacea canaliculate (Plan,
Saska, Cagauan, & Craik, 2008).
Besides host defense properties, other pharmaceutical activities have been reported for
native Mö bius and bracelet cyclotides, including antimicrobial activity against human
pathogens, anti-HIV activity, and anti-cancer activity. The antimicrobial activity of
cyclotides was first reported in 1999 (Tam et al., 1999). This study claimed that several
cyclotides had promising activity against different human pathogens including bacteria
and fungi. In vitro, the antimicrobial activity of cyclotides is dependent on the ionic
strength of the buffer used in the assays. The activity decreases substantially under
physiological conditions, which limits its potent. In another study reported recently,
several cyclotides from C. ternatea have been shown activity against some
Gram-negative bacteria with minimal inhibitory concentration (MIC) in the low micromolar
range (G. K. Nguyen et al., 2011). A more recent study by the same group tested the
antimicrobial activity of several new discovered cyclotides from the same plant (K. N.
Nguyen et al., 2016). Different new cyclotides, named cliotides T7-T21 were tested
against E. coli and shown a salt-dependent antimicrobial activity. More importantly, based
30

on atomic force microscopy (AFM) imaging results, the authors suggested that cyclotides
possess bactericidal activity by disrupting the outer bacterial membrane (K. N. Nguyen et
al., 2016). Another study by a different group reported that cyclotide cycloviolacin O2
shows cytotoxicity to human red blood cells at the concentration that is microbicidal
against E. coli (Henriques et al., 2012) which may limit the use of this cyclotide as peptide
antibiotic.
The anti-HIV activity of cyclotide was first reported by one of the National Cancer Institute
natural products screening programs (Gustafson et al., 1994). In this study, two bracelet
cyclotides, circulin A and circulin B, were found having a dose-dependent cytoprotective
effect against HIV. The activity was found to be unrelated to HIV reverse transcriptase
activity. After this finding, different other cyclotides were reported having similar anti-HIV
activities (B. Chen et al., 2005; Daly, Clark, Plan, & Craik, 2006; Ireland, Wang, Wilson,
Gustafson, & Craik, 2008; Wang et al., 2008). The mechanism of anti-HIV activities was
proposed in a recent study (Henriques et al., 2011). In this study, the authors proposed
that kalata B1 targets the HIV membrane via peptide-lipid interactions, which was tested
by comparing the anti-HIV activities of wild-type kalata B1 and different kalata B1 mutants
with reduced membrane binding affinity. The results indicated that wild-type kalata B1
decreased the infectivity of two different HIV strains. In contrast, the membrane-inactive
analog V25K had almost no anti-HIV activity. These findings were further supported by
another study using cyclotides kalata B2 and tcA (Henriques et al., 2012).
The anti-cancer activity of different cyclotides was first evaluated in 2002, where three
cyclotides (varv A, varv F, and cycloviolacin O2) were reported having in vitro cytotoxicity
toward 10 different cancer cell lines (Lindholm et al., 2002). Later, other native cyclotides
in the Mö bius and bracelet subfamilies have been also reported having cytotoxicity
against numerous cancer cell lines (Gerlach, Burman, Bohlin, Mondal, & Goransson,
2010; Gerlach, Rathinakumar, et al., 2010; He et al., 2011; Svangard et al., 2004; J. Tang
31

et al., 2010). The first in vivo study using a xenograft mice model was conducted to test
the anticancer activity of cyclotide cycloviolacin O2 (Burman et al., 2010). However, no
significant antitumor activity was observed after daily dosing of the cyclotide at 0.5 mg/kg
by intravenous injection, indicating that cyclotide cycloviolacin O2 may not have
significant anticancer activity in vivo. More importantly, a recent publication pointed out
that the anticancer activity of some cyclotides may have limited selectivity toward tumor
cell lines and non-tumor cell lines (Troeira Henriques, Huang, Chaousis, Wang, & Craik,
2014). In this study, cytotoxicity of cyclotide kalata B1 and several of its mutants were
tested against two adherent human cancer cell lines (HeLa and MM96L) and the
noncancerous HHF-1 cell line. No significant differences in cell growth suppression were
observed against all the tested cell lines. This study also found that cyclotides possess
their cytotoxic activity through the disruption of the cell membrane. This lacking selectivity
strongly limits the application of natural cyclotides as effective cytotoxic agents.
Different from Mö bius and bracelet subfamily cyclotides, for trypsin inhibitor subfamily
cyclotides, there are no other bioactivities were reported than trypsin inhibition activity. As
noted earlier, trypsin inhibitor cyclotides have different properties from Mö bius and
bracelet cyclotides as they lack the ability to bind and disrupt cell membranes. This is
consistent with their low cytotoxicity (Cascales et al., 2011). The lack of cytotoxicity,
combining with their cell membrane permeability and tolerance to sequence substitution
in many loops, provide a great advantage for the potential use of trypsin inhibitor
cyclotides as molecular frameworks for the design of novel peptide-based therapeutics
(Andrew Gould & Camarero, 2017).
1.1.5 Using cyclotides as scaffolds for grafting biologically active peptides
As noted before, amino acid scanning of kalata B1 (Simonsen et al., 2008) and MCoTI-I
(Austin et al., 2009) has been reported to illustrate the amenability of cyclotides to
32

sequence variation. These studies demonstrated that introducing an external sequence
may not disturb the CCK motif, which is the foundation for the cyclotide grafting concept
(Fig. 1.7). Table I summarizes several examples of grafted cyclotides with novel
biological activities. Most of the cyclotides used for grafting purposes belong to the
Mö bius and trypsin inhibitor subfamilies. This is due to the difficulties associated for in
vitro folding of bracelet cyclotides (Andrew Gould & Camarero, 2017).

33

Table I Examples of grafted cyclotides
1
B = 2-naphthylalanine, X = citrulline, p = D-proline
2
Z = L-2,3-diaminopropionic acid
3
This is the deletion of cyclotide loop 6 sequence instead of insertion.
4
J = amino isobutyric acid
*
extracellular protease
#
extracellular membrane protein
+
Intracellular targets
SCAFFOLD CYCLOTIDE
NAME
GRAFTED
LOOPS
GRAFTED SEQUENCE TARGET REFERENCE
KALATA B1 Cpr3 2,3 or 5 RRKRRR VEGFR
receptor
#

(Gunasekera et al., 2008)
kB1[GHFRWG;22-
28]
6 GHFRWG MCR receptor
#
(Eliasen et al., 2012)
ckb-kal 6 KRPPGFSPL bradykinin B1
receptor
#

(Wong et al., 2012)
MOG16 5 and 6 RSPFSRV & LYRNGK Act as
autoantigen
(Wang et al., 2014)
THR-5 6 IDGGRLM Thrombin
*
(Getz, Rice, & Daugherty,
2011)
N2.1 3, 5 and 6 RTF & KPLR & KAPRMVR NRP1 receptor
#
(Getz, Cheneval, Craik, &
Daugherty, 2013)
MCOTI-I MCo-CVX-5c 6 YRXCRGpRRBCYXK
1
CXCR4 receptor

#

(T. L. Aboye et al., 2012)
MCo-PMI 6 GASKAPTSFAEYWNLLSA P53-HDM2
+
(Ji et al., 2013)
MCo-AT1-7 6 ZRVYIE
2
MAS1 receptor
#
(T. Aboye et al., 2016)
MCOTI-II MCoTI-II [AKQ] 1 AKQ FMDV 3C
protease
*

(P. Thongyoo, Roque-Rosell,
Leatherbarrow, & Tate, 2008)
Δ[SDGG] 6 Deletion SDGG
3
Tryptase
*
(P. Thongyoo et al., 2009)
MCo-OPN 6 SVVYGLR α9β1 integrin
#
(Chan et al., 2011)
R 02 1 RSLARTDLDHLRGR αvβ4 integrin
#
(Richard H Kimura et al.,
2012)
MCOG2 6 GASKAPASJLRKLJKRLLRDA
4
SET
+
(D'Souza et al., 2016)

34

Figure 1.7. The concept for the design of novel cyclotides with new biological activities using
molecular grafting. The loops 6 of cyclotide MCoTI-I is used as an example for the molecular
grafting of biologically active peptide sequences into the backbone of the cyclotide.
1.1.5.1 Grafting using the Mö bius cyclotide kalata B1 as scaffold
1.1.5.1.1 To target extracellular receptors
The first grafted cyclotide was designed by grafting an antiangiogenic peptide sequence
(RRKRRR) into the kalata B1 framework (Gunasekera et al., 2008). The grafted poly-Arg
sequence has been shown able to antagonize vascular endothelial growth factor receptor
2 (VEGF-R2). This sequence was grafted onto loops 2, 3, 5, or 6 of kalata B1. The
cyclotide cpr3 (grafted in loop 3) was the most active with an IC 50 of ≈12 µ M in a
BAF3-VEGFR2 competition assay. Interestingly, the grafted cyclotides showed better
stability in serum when compared to that of the linear poly-Arg peptide, therefore,
demonstrating the grafting of cyclotides may help display and stabilize bioactive peptides.
Recently, another grafted kalata B1-based cyclotide (Eliasen et al., 2012) was
engineered as a melanocortin receptor 4 (MCR4) agonist. Since activation of the MCR4
receptor has been shown to reduce food take and accelerate metabolism in mice (A. S.
Chen et al., 2000), it was hypothesized that stable cyclotide-based MCR4 agonist could
lead to the design of novel anti-obesity therapeutic agents. In this study, the tetrapeptide
HFRW epitope from the melanocyte stimulating hormone ( -MSH) was grafted onto loop
6 of kalata B1. All the grafted cyclotides were shown adopt a native cyclotide fold. One of
35

the most potent grafted peptides was found to have a lower IC 50 value than -MSH in the
radiolabeled binding assay with a calculated? Ki value of ≈29 nM. This grafted cyclotide
was 107 or 314 times more selective over MCR4 than MCR1 or MCR5, respectively.
Again, the grafted cyclotides were shown to have significantly higher stability to
chymotrypsin compared to -MSH. This study validates the use of cyclotide as scaffolds
to display bioactive peptide.
Another example of molecular grafting using kalata B1 was reported targeting
bradykinin B1 receptor which is involved in inflammatory pain (Wong et al., 2012). During
inflammatory pain, bradykinin is released and activates several GPCR receptors inducing
pain responses. Two bradykinin B1 receptor peptide antagonists, kallidin and kinestatin,
were grafted onto loop 6 of kalata B1. The grafted cyclotides were shown to adopt a
native cyclotide fold and be resistant to degradation in human serum. The grafted
cyclotides specifically blocked the bradykinin B1 receptor but not the B2 receptor. Most
importantly, significant inhibition of pain response was observed in an animal model
under i.p. administration of both linear antagonist peptides and grafted cyclotides, in
contrast to, under oral administration, only cyclic analogs but not linear peptides were
able to inhibit the writhing action. The authors claimed that both grafted cyclotides have
significant oral bioactivity although no oral bioavailability data was reported in the paper.
The grafted kalata B1 cyclotides mentioned earlier only involve single loop substitution. A
recent study grafted bioactive epitope peptides into multiple loops of kalata B1 (Wang et
al., 2014). In this study, partial sequences of myelin oligodendrocyte glycoprotein (MOG)
were grafted onto the cyclotide kalata B1 for the treatment of multiple sclerosis (MS).
Since the MOG-based peptide was around 20 amino acids long, more than one loop of
kalata B1 was used for grafting. Among different designed peptides, MOG15 and 16 with
grafted sequences onto loops 5 and 6 were shown adopt the same folding as kalata B1,
which proved the concept of multiple-loop grafting onto the cyclotide scaffold. Some
36

grafted cyclotides showed activity in vivo in an experimental autoimmune
encephalomyelitis mouse model suggesting that there is potential to develop safe and
effective treatments for MS using cyclic peptide scaffolds. It is worth noting that several
grafted cyclotides, like MOG7 (9 amino acids grafted) and MOG12 (10 amino acids
grafted), with long sequences introduced, did not adopt a well-defined structure,
suggesting that cyclotide kalata B1 may have some loops with length requirements for
proper folding of the resulting grafted cyclotide.
1.1.5.1.2 Kalata B1 based molecular evolution
The grafted cyclotides discussed so far are all based on rational design approach, where
bioactive peptide epitopes are grafted onto the kalata B1 cyclotide scaffold for displaying
of bioactive peptides. Different strategies for developing cyclotides able to bind to
different protein targets involve the generation and screening of cyclotide libraries
containing a large number of different sequences. The first randomized kalata B1 library
was reported by Getz et al (Getz et al., 2011). In this study, a library on a linearized
version of kalata B1 was used in combination with a bacterial display approach. All seven
residues in kalata B1 loop 6 were randomized using all 20 genetically encoded amino
acids using an NNR degenerate codon approach. This cyclotide library was successfully
screened using one round of magnetic-activated cell sorting (MACS) followed by three
cycles of fluorescence-activated cell sorting (FACS) to yield a diverse set of thrombin
binding sequences with consensus motif. The binding affinity Kd to thrombin of two
sequence THR-5 and THR-29 was ≈500 nM and ≈330 nM, respectively.
This approach was lately used to screen cyclotides able to target the receptor
neuropilin-1(Getz et al., 2013). Neuropilins interact with growth factors to promote
angiogenesis and tumor progression, including VEGF and HGF. Different from the
previous study, ligands from first generation library (loop6 randomization) were subjected
to one cycle of affinity maturation by fix loop 6 consensus sequence and randomize loop
37

5 sequence. This ended up with yield acyclic peptides with affinities of 40−60 nM toward
the receptors neuropilin-1. Most importantly, the cyclic version of one sequence N2.1
retained high affinity toward neuropilin-1, exhibited increased protease resistance, and
conferred improved potency in in vitro cell migration inhibition assay. This study
demonstrated that potent cyclotide based antagonists could be created by evolutionary
design.
So far, kalata B1 has been widely used for displaying bioactive peptide epitopes.
However, even though kalata B1 can cross cell membranes, no kalata B1 derived
cyclotides targeting intracellular target have been reported yet. This could be related to
the mechanism of kalata B1 cell permeability, which disrupts the cell membrane leading
to high cytotoxicity. Thus, kalata B1 has only been used to graft sequence target
extracellular proteases or receptors.
1.1.5.2 Molecular grafting using trypsin inhibitor cyclotides as scaffold
1.1.5.2.1 To target extracellular proteases
Besides the Mö bius cyclotide subfamily, the trypsin inhibitor subfamily of cyclotides has
been also widely used for grafting purposes. Since the trypsin inhibitor cyclotides have
intrinsic protein inhibition activity, the early applications were focused on exploring the
use of these cyclotides to target other proteases. The first application involved the
mutation of several amino acids at loop 1 of the trypsin inhibitor cyclotide MCoTI-II to
target the foot-and-mouth disease virus (FMDV) 3C protease (P. Thongyoo et al., 2008).
In this study, MCoTI-II K10F was shown to be able to inhibit chymotrypsin. The single
mutants MCoTI-II K10Q and the cyclotide grafted with the peptide segment AKQ into loop
1 were shown µM activity against FMDV 3C protease. By using this approach, the same
group generated cyclotides targeting other serine proteases including β-tryptase, which is
a mediator of allergic responses, and human leukocyte elastase (HLE), which is a
38

secreted during inflammation by macrophages (P. Thongyoo et al., 2009). Replacement
of residue Lys10 by Ala or Val provided selective low nM inhibitors against HLE. Deletion
of the segment SDGG in loop 6 of cyclotide MCoTI-II increased the inhibition activity for
β-tryptase by 60 times. These studies have shown that the MCoTI-II scaffold can be
mutated without affecting the cyclization and folding. Furthermore, MCoTI-II can be
re-engineered to target other proteases like β-tryptase, HLE, or chymotrypsin-like
protease with excellent selectivity.
1.1.5.2.2 To target extracellular receptors
Trypsin inhibitor cyclotides have also been used to display bioactive peptide epitopes
able to target extracellular receptors. Chen et al have used cyclotide MCoTI-II to graft
several linear pro-angiogenesis peptides (LAM, OPN, and QK) (Chan et al., 2011). In
these cases, the loop 6 of MCoTI-II was used for grafting. The grafting generated several
cyclotides with similar bioactivity but more stability. Another study of MCoTI-II grafting
was reported by Kimura et al (Richard H Kimura et al., 2012). In this study, several
peptide sequences bind to integrin v 6 were grafted onto loop 1 of MCoTI-II for imaging
purpose. It needs to point out that the backbone of those cyclotides was opened at loop 6
and the N-terminal amide was used for DOTA-NHS
(1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid) labeling. The acyclic labeled
peptides were shown able to bind to v 6 in vitro and could be used for tumor detection.
Recently, our group reported the production of a CXCR4 antagonist cyclotide by grafting
CVX15-based peptides into the MCoTI-I scaffold (T. L. Aboye et al., 2012). By using the
crystal structure of CXCR4 bound to peptide CVX-15, eight different grafted cyclotides
were designed.
1
H-NMR studies revealed that the grafting of CVX15 based peptides on
the loop 6 did not affect the native folding of cyclotide scaffold. The most potent
compound produced in this work, MCo-CVX-5c, was able to antagonize CXCR4 with an
IC 50 value of ≈19 nM which is around three times better than that of the CVX-15 peptide,
39

and able to block HIV-1 cell entry with IC 50 of ≈ 2 nM. These improvements were
attributed to the potential interaction between CXCR4 and other loops of cyclotide
scaffold. It is worth pointing out that cyclotide MCo-CVX-5c was shown remarkable serum
stability with 1/2 of around 62 h compared to the linear peptide CVX15 Gln6Cit with 1/2 of
≈21 h. A model of cyclotide MCo-CVX-5c bound to CXCR4 was built based on the crystal
structure of CVC15-CXCR4 and the NMR structure of cyclotide MCoTI-II. According to
this model, loops 2 and 5 are in proximity to the receptor surface. A follow-up study trying
to improve MCo-CVX-5c activity by introducing mutations on loop 2 was recently reported
(T. Aboye et al., 2015). Total 15 analogs of MCo-CVX-5c containing different amino acids
on position 12 at loop 2 were efficiently synthesized using an innovative “tea-bag”
approach. Even though none of those mutations showed a significant increase in
antagonistic activity, this study opened the possibility of using this approach for the
production of positional scanning cyclotide-based libraries for improving activity.
1.1.5.2.3 Isopeptide bond grafting of trypsin inhibitor cyclotides
All the grafted cyclotides mentioned so far use a rationale design to graft a known
bioactive peptide sequence into the cyclotide scaffold. In some cases, the N-terminal
-amino and C-terminal carboxylate groups may be required for bioactivity. A recent
study has overcome this limitation by using isopeptide bonds to graft the corresponding
biologically active peptide to the cyclotide backbone (T. Aboye et al., 2016). In this
study, the β-amino group in the side chain of L-2,3-diaminopropionic acid and the
-carboxylate group of glutamic acid were used to form two isopeptide bonds to the
cyclotide backbone, thus preserving the -amino and C-terminal carboxylate groups of an
analog of peptide angiotensin 1-7, which are required for bioactivity. The grafted cyclotide
MCo-AT1-7 has been shown to be a potent MAS1 receptor agonist with very similar
activity to the peptide AT1-7. It is also worth noting that the cyclotide MCo-AT1-7 showed
remarkable resistance to biological degradation in human serum with an ex-vivo 1/2 of
40

≈39 h, which is highly significant when compared to that of angiotensin 1-7 with a 1/2 ~20
min under the same conditions. The native folding of cyclotide backbone was confirmed
by
1
H-NMR, confirming again the high the tolerance to sequence variation of loop 6 in
MCoTI-cyclotides.
1.1.5.2.4 To target intracellular proteins
The cyclotide MCoTI-I has been shown to enter cells (Contreras et al., 2011). The cell
permeability of this cyclotide combined with its high serum stability and high tolerance for
grafting biologically active peptide sequences open the possibility for using this scaffold
to target intracellular proteins. The first grafted cyclotide targeting intracellular
protein-protein interaction was recently reported by the Camarero group (Ji et al., 2013).
In this study, a p53-based -helical peptide (peptide PMI) (Marzena Pazgier et al., 2009)
was grafted onto loop 6 of cyclotide MCoTI-I. The resulting cyclotide MCo-PMI was able
to bind to both HDM2 and HDMX with low nM affinity. MCo-PMI was also cytotoxic to
wild-type p53 cancer cell lines and could activate the p53 tumor suppressor pathway both
in vitro and in xenograft mouse model (Ji et al., 2013). These results indicate that
engineered MCoTI cyclotides can be used for targeting intracellular protein-protein
interactions and makes cyclotide grafting a valuable tool in the design of novel
peptide-based therapeutics. This approach was recently used by another group to target
the intracellular protein SET (D'Souza et al., 2016). A potent SET antagonist peptide
COG, which is derived from apolipoprotein E (apoE), was grafted onto loop 6 of MCoTI-II
using the exactly same approach of MCo-PMI. The resulting cyclotide MCOG1 and
MCOG2 have been shown cytotoxic to K562 cancer cell line and stable in human serum
with an ex-vivo 1/2 > 24 h. These studies confirmed that the MCoTI scaffold can be
engineered to stabilize -helical peptides and deliver it into cells, where it can modulate
protein-protein interactions.
41

1.1.6 Summary
Cyclotides are a new family of large backbone-cyclized polypeptides that share a
disulfide-stabilized core (three disulfide bonds) characterized by an unusual Cys-knotted
structure. As described above, they are remarkably stable to chemical, thermal and
proteolytic degradation due to the cyclic topology and cystine knot structure. Mö bius and
trypsin inhibitor subfamily cyclotides have been shown cell permeability. Additionally,
cyclotides are amenable to substantial sequence variation. All these properties make
cyclotide promising bio-scaffold for pharmaceutical applications.

1.2 Targeting the p53-HDM2 interaction for cancer therapy
1.2.1 General introduction of p53
TP53, as a well-known gene which encodes the protein p53, was discovered more than
30 years ago (DeLeo et al., 1979; Lane & Crawford, 1979; Linzer & Levine, 1979).
Around 10 years later, the protein p53 was characterized as a tumor suppressor (Baker et
al., 1989; Finlay, Hinds, & Levine, 1989). In normal cells, p53 plays an important role in
response to a broad range of stress responses, like DNA damage and oncogenic stress,
among others [refs]. It has been named as the “cellular gatekeeper” (Lane, 1992) or “the
guardian of genome” (Levine, 1997) to emphasize its important function in maintaining
cellular function. After activation by stress, p53 works as a central node that can organize
proper stress responses, including autophagy, apoptosis, cell cycle arrest, senescence,
DNA repairing or cell metabolism change (Vousden & Prives, 2009). The protein p53
regulates stress response genes by transcriptional activity (B. Vogelstein, Lane, & Levine,
2000; Vousden & Lane, 2007). The transcriptional activity of p53 is regulated by
numerous posttranslational modifications during stress-induced responses. There are
more than 30 different posttranslational modification sites within p53 that have been
42

reported, which include phosphorylation, acetylation, and ubiquitination (Kruse & Gu,
2009b). The classical model of p53 activation basically includes three steps:
stress-induced p53 stabilization, DNA binding and gene transcriptional activation (Kruse
& Gu, 2009a).
1.2.2 Cellular function of p53
Genome instability and mutation have been listed as one of the ten cancer hallmarks.
P53 plays an important role in detecting gene DNA damage and triggers cell cycle arrest
or apoptosis as a response (Hanahan & Weinberg, 2011). Impaired p53 function leads to
accumulation of DNA damage and mutations, which will eventually develop a cancer
phenotype. Importantly, it has been reported that in more than half human cancer cases
there are mutations that disturb p53 function, indicating that p53 has a tumor suppressor
function (Hainaut & Hollstein, 2000). Transfection of mutant p53 into p53-null cell lines
causes tumor formation in mice (Wolf, Harris, & Rotter, 1984). However, induction of
wild-type p53 facilitates apoptosis in myeloid leukemia cells (Yonish-Rouach et al., 1991).
1.2.3 p53 regulation by the HDM2 and HDMX complex
Activation of p53 leads to an increase in the expression of different downstream genes
important for cell cycle control, apoptosis and senescence. Therefore, a tight control of
p53 activity is crucial for maintaining normal cell function. The p53 intracellular level is
regulated by ubiquitin-mediated proteasome degradation (Michael & Oren, 2003). Murine
double minute protein 2 (HDM2) was identified as a ubiquitin E3-ligase with high
specificity for the p53 protein, indicating the role of HDM2 in p53 regulation (Haupt, Maya,
Kazaz, & Oren, 1997; Honda et al., 1997; Kubbutat, Jones, & Vousden, 1997). The
protein HDM2 is not the only E3-ligase for p53 because in HDM2 null mice, p53 still can
be degraded by the proteasome in a ubiquitin-dependent fashion (Ringshausen, O'Shea,
Finch, Swigart, & Evan, 2006). Even though other E3-ligase like COP1 (Dornan et al.,
43

2004), Pirh2 (Leng et al., 2003) and Arf-B1 (D. Chen et al., 2005) have been shown to
control also the cellular level of p53, the major regulator of p53 is HDM2.
The central role of HDM2 is controlling p53 protein level. The protein HDMX has
been also shown to be key for regulating p53 cellular concentration (de Graaf et al., 2003;
Linares, Hengstermann, Ciechanover, Muller, & Scheffner, 2003). HDMX, also known as
HDM4, is a structural homologue of HDM2. However, unlike HDM2, HDMX does not have
E3 ligase activity. HDMX can form a heterodimer with HDM2 via their C-terminal RING
domains. By forming a heterodimer, HDMX promotes HDM2 E3-ligase activity and
destabilizes both HDM2 and p53 (Danovi et al., 2004; Linares et al., 2003; Poyurovsky et
al., 2007; Uldrijan, Pannekoek, & Vousden, 2007).
The basic mechanism of how HDM2/HDMX regulate the cellular level of p53 is shown in
Fig. 1.9. HDM2 and p53 forms a negative feedback loop where HDMX promotes HDM2
function. The HDM2 expression is controlled by p53 transcriptionally. Increased
expression of HDM2 facilitates p53 ubiquitination and degradation, leading to decreased
in p53 level which causes a decrease in the rate of HDM2 expression (Picksley & Lane,
1993).
In addition, not only HDM2 and HDMX can control p53 degradation, but also binding of
both HDM2 (J. Chen, Marechal, & Levine, 1993) and HDMX (Marine & Jochemsen, 2005)
can repress p53 mediated transcription activation. A second p53 binding site has been
discovered in HDM2 which may indicate its additional function as transcription
suppressor (Shimizu et al., 2002; Yu et al., 2006). There is evidence showing that both
HDM2 and HDMX can be recruited to the p53 promoter region to suppress p53
transcriptional activity (Y. Tang, Zhao, Chen, Zhao, & Gu, 2008). Furthermore, HDM2 also
facilitates mono-ubiquitination, therefore, regulating the nuclear concentration of p53 by
enhancing its nuclear export (M. Li et al., 2003; Lohrum, Woods, Ludwig, Balint, &
Vousden, 2001; Nie, Sasaki, & Maki, 2007).

44

Figure 1.8. The protein p53 as a node for multiple stress responses. Once p53 is activated by
different stress signal, activates different downstream pathways which result in cell cycle arrest,
apoptosis induction, DNA repair, among others. Picture was taken from Brown (C. J. Brown,
Cheok, Verma, & Lane, 2011)
1.2.4 Targeting the p53-HDM2 interaction for cancer therapy
Due to the importance of HDM2 in the regulation of p53 activity, inhibition of p53 and
HDM2 interaction has been validated as a therapeutic target. Also, HDMX is able to
promote HDM2 function, which makes HDMX another ideal therapeutic biological target
(Shangary & Wang, 2009). Although, it has been reported that in more than half human
cancer cases there are mutations that disturb p53 function, among these cases, more
than half have mutant on p53 pathway rather than p53 itself. Increased expression of
HDM2 (Momand et al., 1992) and/or HDMX (Laurie et al., 2006; Riemenschneider et al.,
1999), deletion or inactivation of p53-HDM2 inhibitor ARF (Esteller et al., 2001; Sherr &
Weber, 2000) all contribute to p53 inactivation. Reactivation of p53 function by interfering
45

HDM2-p53 interaction has been proven of therapeutic value [refs]. Numerous screening
methods have been used to find inhibitors for HDM2-p53 interaction, especially focusing
on p53 binding domain located at N-terminal of HDM2 (Shangary & Wang, 2009).
Currently, there are many drugs in development designed to antagonize the interaction
between p53 and HDM2 or HDMX in order to reactivate p53 function. Among these drugs,
the most promising inhibitor is nutlin (Vassilev et al., 2004) and MI-219 (Shangary et al.,
2008). Some drugs analogs of these compounds have entered clinical trials.

Figure 1.9. Cellular activity of p53 is controlled by HDM2/HDMX complex. HDM2 and HDMX
(HDM4) regulate p53 in three different ways. First, HDM2 can poly-ubiquitinate p53 and cause its
proteasomal degradation. Secondly, binding of HDM2 or HDMX to p53 inhibit the transcriptional
activity of p53. Lastly, low level of HDM2 mono-ubiquitinates p53 and facilitates its nuclear export.
The picture was taken from Jain (Jain & Barton, 2010).
Recently, a high-affinity peptide inhibitor for p53-HDM2/HDMX has been developed,
which was termed PMI (p53-HDM2/HDMX inhibitor) (M. Pazgier et al., 2009). Unlike some
small molecules which can only inhibit either binding of p53 to HDM2 or HDMX, the PMI
peptide has high affinity to the p53 binding domains of both HDM2 and HDMX. The Kd
46

values for these interactions are in the single nM range. The sequence of the PMI peptide
(TSFAEYWNLLSP) shares some of the key residues (Phe, Trp, and Leu) found in the
-terminal region of p53 (ETFSDLWKLLPE) known to bind HDM2 and HDMX
(ETFSDLWKLLPE) HDM2. Despite the improved activity of the PMI peptide when
compared to the wild-type p53-derived polypeptide, in vitro cell viability assays using PMI
show little biological activity. This can be attributed to the low cell membrane permeability
and poor serum stability of this peptide (M. Pazgier et al., 2009). Improving its membrane
permeability and serum stability of PMI could lead to the development of novel
peptide-based therapeutics able to target p53-HDM2/HDMX for cancer therapy.

47

CHAPTER TWO Targeting the p53-HDM2 complex using a grafted cyclotide
The transcription factor p53 plays an important role in protecting cells from malignant
transformation by inducing cell cycle arrest and apoptosis in response to cellular stress
induced by DNA damage (Bert Vogelstein & Kinzler, 1992). As described in the previous
chapter (section 1.2.3) the activity of p53 is regulated by the oncogenic protein named
HDM2 and HDMX. Both of HDM2 and HDMX contain N-terminal p53 binding domains
that can bind to the N-terminal flexible region of p53. In addition, HDM2 is a ubiquitin E3
ligase that can poly-ubiquitinate p53 promoting its proteasomal degradation (Fang,
Jensen, Ludwig, Vousden, & Weissman, 2000) and HDMX can inhibit p53 activity through
protein interaction mediated sequestration (Shvarts et al., 1996). The function of p53 is
prevented in many cancer cells by promoting deletion, mutation, and sequestration of p53,
thus highlighting the importance of p53 as the guardian of the genome (Green & Kroemer,
2009). Therefore, restoration of p53 function is an attractive strategy for the development
of effective cancer therapeutics (Christopher J Brown, Lain, Verma, Fersht, & Lane,
2009).
Based on the structure of the complex between p53 and the p53-binding domains of
HDM2 or HDMX, the interaction is mainly mediated by a 15-residue -helical domain of
p53 inserting into a hydrophobic pocket on the surface of HDM2 or HDMX (Popowicz et
al., 2010). These findings have led to the development of small molecules (Millard,
Pathania, Grande, Xu, & Neamati, 2011) and peptides (Liu et al., 2010; Marzena Pazgier
et al., 2009) that can antagonize the interaction between p53 and HDM2 or HDMX.
Disruption of this interaction in cells expressing wild-type p53 has been shown to reduce
the viability of cancer cells. One short peptide named PMI derived from the
transactivation sequence of p53 that was obtained by phage display selection has been
shown to bind HDM2 and HDMX with high affinity (Marzena Pazgier et al., 2009).
48

However, the use of this peptide a therapeutic agent has been limited by its poor stability
and bioavailability.
Cyclotides are an emerging family of plant-derived peptides that have a unique head to
tail circular knotted structure of three disulfide bonds as described earlier (Section 1.1.2).
Several cyclotides have been used as molecular scaffolds for stabilizing biological active
polypeptides (T. L. Aboye et al., 2012; Eliasen et al., 2012; Richard H Kimura et al., 2012).
The cyclotide framework provides a usefully molecular scaffold with high stability to
physical, chemical and biological degradation (Colgrave & Craik, 2004; Puttamadappa et
al., 2010). MCoTI-cyclotides have been shown to enter human cells mainly through
macropinocytosis (Contreras et al., 2011). Cyclotides are also tolerant to sequence
variation (Garcia & Camarero, 2010), making them excellent bioscaffold for grafting
biologically active peptides.
In this work, we designed for the first time an engineered cyclotide able to target an
intracellular protein-protein interaction. A p53 and HDM2/HDMX antagonist peptide
derived from p53 -helical peptide (Marzena Pazgier et al., 2009) was grafted into one of
the loops of cyclotide MCoTI-I. The resulting polypeptide was shown to adopt a native
cyclotide fold and bind to both HDM2 and HDMX with low nanomolar affinity. Also, this
cyclotide induced cytotoxicity in p53 wild-type human cancer cell lines in a p53-
dependent manner both in vitro and in vivo.
2.1 Design and synthesis of cyclotide MCo-PMI
2.1.1 Design of PMI grafting into MCoTI-I scaffold
To engineer the cyclotide MCoTI-I for antagonizing the p53-HDM2/HDMX interaction, the
phage-display selected -helical peptide PMI (Marzena Pazgier et al., 2009) was chosen
for grafting into MCoTI-I. The PMI peptide shares the three key residues (Phe19, Trp23
49

and Leu26) of p53 required for the interaction with the p53-binding domain of HDM2 and
HDMX with very high affinities (Kd values of 3.4 ± 0.6 nM and 4.2 ± 1.5 nM for HDM2 and
HDMX, respectively) (Marzena Pazgier et al., 2009). The cytotoxic activity of PMI on
p53
+/+
HCT116 cells is much less than the small molecule nutlin-3. This is likely due to
proteolytic degradation, inefficient cellular uptake and endosomal sequestration
(Marzena Pazgier et al., 2009). Grafting of the PMI peptide into cyclotide scaffold may
overcome these problems.
The design of the grafted cyclotides is shown in Figure 2.1. The PMI peptide was grafted
onto the loop 6 of MCoTI-I. This loop has been shown to be more flexible in solution and
amenable to sequence substitution (Austin et al., 2009; Puttamadappa et al., 2010). To
facilitate the grafting of the PMI peptide without disturbing the -helical structure of PMI
and the cyclotide scaffold folding, the N-terminus of the PMI peptide was fused to the
linker peptide sequence Ala-Ser-Lys/Arg-Ala-Pro as shown in Figure 2.1. This linker is
based on the N-terminal sequence of the bee-venom neurotoxin apamin which adopts a
coil-turn- -helix structure. This peptide linker has been used to display p53 helical
peptides recently (Chong Li, Pazgier, Liu, Lu, & Lu, 2009). The peptide MCo-PMI was
designed by grafting this chimeric apamin-PMI peptide onto the residues Ser31 and
Gly33 (Fig 2.1) to minimize any possible steric hindrance between the scaffold and the
grafted peptide. A serial of peptides was designed. The substitution of the tryptophan
residue in the p53 sequence by the unnatural amino acid 6-choloro tryptophan was
shown to improve the binding affinity to HDM2. The replacement of key residue Phe42 by
Ala eliminates PMI activity that was designed as a negative control for all biological
experiments.
50

Figure 2.1. Design of MCo-PMI peptide. Chimeric apamin (green residues) and PMI (red residues)
peptides were grafted onto loop 6 of cyclotide MCoTI-I. The mutations F42A and W46Z
(Z=6-chloro-tryptophan) are shown in blue. MCoTI-I scaffold sequence is shown in black.
Conserved cysteine residues are highlighted in yellow and disulfide connection in red. The circular
backbone structure is shown with the green line.
2.1.2 Synthesis and folding of MCo-PMI cyclotides
MCoTI-based cyclotides can be produced by both chemical synthesis and bacterial
recombinant expression (Y. Li et al., 2015). The chemical synthesis of cyclotides makes
possible the introduction of unnatural amino acids into cyclotide, while recombinant
expression in E. coli allows the production of cyclotides incorporating NMR-active isotopes
like
15
N or
13
C to facilitate their structural analysis by NMR. The backbone cyclization was
performed by intramolecular native chemical ligation (NCL) (J. Camarero & Muir, 1997; J.
A. Camarero, Pavel, & Muir, 1998) using the native Cys located at the beginning of loop 6
to facilitate the cyclization. The recombinant expression of MCo-PMI was performed by
fusing the corresponding linear precursor in the frame to a modified Mxe gyrase A intein
and a TEV protease recognition sequence at the C- and N-termini, respectively, as shown
in Figure 2.2.
51

Figure 2.2. Scheme summarizing the approaches used for the chemical synthesis (top) or
recombinant production (bottom) of MCo-PMI cyclotides. In both cases, the cyclization/folding was
performed using an intramolecular version of Native Chemical Ligation in the presence of reduced
glutathione (GSH) at pH 7.4. Under these conditions, the linear MCo-PMI cyclotide precursors
were able to efficiently cyclize and fold in 24 has shown in the analytical HPLC traces for the crude
cyclization/folding reaction for MCo-PMI.

Once the intein precursor protein was expressed and purified, the N-terminal TEV
protease recognition sequence was removed proteolytically by TEV protease. Backbone
cyclization and oxidative folding were performed with reduced glutathione (GSH) at
physiological pH in one single step (Fig. 2.2). Chemical synthesis of the linear precursor
peptide thioester was accomplished using Fmoc-based solid-phase peptide synthesis
(SPPS) on a sulfonamide resin (Y. Li et al., 2015). After activation and cleavage of the
peptide-resin, the linear thioester precursors were cyclized and oxidatively folded in one
single step with GSH. The cyclization and oxidative folding were remarkable efficient
yielding in both cased the natively folded cyclotide as the major product. After purification
by preparative reversed-phase HPLC, the purity of MCo-PMI peptides was determined by
52

analytical RP-HPLC and electrospray mass spectrometry (ES-MS) as shown in Figure
2.3.

Figure 2.3. Analytical reverse-phase C18-HPLC traces and electrospray mass spectra
(deconvoluted) of purified MCo-PMI cyclotides. HPLC analysis was performed using a linear
gradient of 0-70% solvent B over 30 minutes.
53

Figure 2.3. Analytical reverse-phase C18-HPLC traces and electrospray mass spectra
(deconvoluted) of purified MCo-PMI cyclotides. HPLC analysis was performed using a linear
gradient of 0-70% solvent B over 30 minutes.

54

2.2 In vitro characterization of MCo-PMI cyclotides
2.2.1 Direct binding assay
The biological activity of MCo-PMI grafted cyclotides was first tested by direct
binding assay using fluorescent polarization anisotropy assay. Briefly, FITC-labeled
derivatives of MCo-PMI-K37R, MCo-PMI-6ClW, and MCo-PMI-K37R-F42A were titrated
with the p53 binding domain of HDM2 and HDMX. The FITC was site-specifically
incorporated into loop 1 by reacting with the ε-NH2 group of residue Lys6. The cyclotide
MCo-PMI-K37R shows strong binding affinity toward the p53 binding domain of HDM2
(KD = 2.3 ± 0.1 nM) and HDMX (KD = 9.7 ± 0.9 nM) (Fig. 2.4). These affinities are similar
to the reported for the linear peptide PMI (Marzena Pazgier et al., 2009) which indicates
that the grafted PMI peptide sequence adopts a biologically active conformation.
Intriguingly, the binding affinity of cyclotide MCo-PMI-6ClW for HDM2 (KD=2.6 ± 0.4 nM)
was similar to that of MCo-PMI-K37R, suggesting that the replacement of Trp residue
with 6-choloro tryptophan is not critical for improving the binding affinity to HDM2. As
expected, the negative control cyclotide MCo-PMI-K37R-F42A did not interact with either
HDM2 or HDMX in this dose range.

Figure 2.4. Direct binding of FITC-labeled MCo-PMI peptides to recombinant HDM2 (17−125) and
HDMX (17−116) was measured by fluorescence polarization anisotropy.

55

2.2.2 Inhibition assay
The MCo-PMI cyclotides activity was tested by analyzing their ability to disrupt the
high-affinity complexes between the transactivation domain of p53 and HDM2 or HDMX.
This was accomplished by using a fluorescence resonance energy transfer (FRET)
based reporter formed by the fluorescent proteins YPet and CyPet (A. W. Nguyen &
Daugherty, 2005)[ref] fused to a p53 peptide and the p53 binding domains of
HDM2/HDMX, respectively. The interaction between p53 and the p53 binding domain of
HDM2/HDMX will bring the two fluorescent proteins in close proximity. Thus, disruption of
this interaction can be monitored by measuring the decrease of the FRET signal. The
results are shown in Figure 2.5.

Figure 2.5. FRET-based inhibition assay. (a) SDS-PAGE analysis of purified YPet-p53,
CyPet-HDM2 and CyPet-HDMX. (b). competition experiments of MCo-PMI peptides and nutlin-3
with p53 (15−29) for binding to HDM2 (17−125) and HDMX (17−116). Binding competition
experiments were performed by titrating a solution of YPet−p53 (5 μM) and CyPet−HDM2 (20 nM)
56

or CyPet−HDMX (20 nM) with increasing concentrations of unlabeled inhibitor. The decrease in
FRET signal was measured at 525 nm (YPet) by excitation at 414 nm (CyPet). Data are the mean
± SEM of experiments performed in triplicate.
The cyclotides MCo-PMI and MCo-PMI-K37R were able to antagonize p53-HDM2/HDMX
interaction with similar IC 50 values, which indicates the mutation Lys to Arg did not affect
the folding or the biological activity of the resulting cyclotides. All wild-type PMI grafted
cyclotides showed IC 50 values for the inhibition of the p53-HDMX interaction that were
around 3 times higher than those for the inhibition of p53-HDM2, which is in agreement
with the values reported for the PMI peptides (Marzena Pazgier et al., 2009). The
wild-type PMI grafted cyclotides were about 3 times more active than the HDM2 inhibitor
nutlin-3. As expected, the cyclotides MCoTI-I and MCo-PMI-F42A did not disrupt the
interaction between p53 and HDM2/HDMX. Taken all together, these data suggest that
the PMI grafted cyclotides target both HDM2 and HDMX and exhibit a slight binding
preference for HDM2 over HDMX.
2.3 Structural characterization of the MCo-PMI by NMR
2D-heteronuclear NMR spectroscopy was used to characterize the structure of
free MCo-PMI (Fig. 2.6). Comparison of the HSQC-spectra of
15
N-labeled MCo-PMI and
MCoTI-I cyclotides clearly showed that cyclotide MCo-PMI is adopting a native cyclotide
fold. The changes in chemical shifts are concentrated around the common sequence at
loop 6, where the PMI peptide was grafted. The differences in chemical shifts between
MCo-PMI and MCoTI-I backbone amide protons within loop 1 to loop 5 are well within 0.2
ppm, which indicates there are only minimal changes in the backbone conformation.
These results strongly approved the robustness of the MCoTI-I cyclotide scaffold for large
size peptide sequence grafting in loop 6 (25 residues for MCo-PMI comparing to original
8 residues in MCoTI-I). The NMR analysis of the cyclotide MCo-PMI segment
corresponding to the PMI peptide also reveals that although this segment has a
57

predisposition to adopt α-helical structure, the absence of a typical α-helical nuclear
Overhauser effect (NOE) pattern indicates that it does not adopt a stable helical structure.

Figure 2.6 NMR analysis of MCo-PMI. (a) Overlay of HSQC spectra of MCoTI-I (black) and
MCo-PMI (red) in solution. (b) Changes in the backbone protons (H-Nα (H’) and Hα) and Nα
chemical shifts between the common sequence of MCoTI-I and MCo-PMI, residues 1 through 43.
(c) The probability of secondary structure formation for the apamin-PMI grafted peptide segment
on MCo-PMI (residues 36 through 50). The probabilities are based on the NH backbone chemical
shifts (H-Nα (H’) and
15
N). The positive chemical shift index with respect to random coil implies that
58

PMI loop is likely alpha-helical.
To better understand the molecular interaction between cyclotide MCo-PMI and
HDM2/HDMX, we elucidated the three-dimensional structure of the molecular complex
between MCo-PMI and HDM2 (17-116) (Fig. 2.7). The structure was determined by
heteronuclear NMR using
15
N-labeled MCo-PMI and
13
C,
15
N-labeled HDM2. The solution
structure of HDM2 is in close agreement with the crystal structure of HDM2 in the
complex with PMI peptide (Marzena Pazgier et al., 2009). As shown in Figure 2.7a, the
overall folding of MCo-PMI conserves the Cys-knot topology of MCoTI-I with an extended
loop 6. As expected, the PMI peptide sequence in the loop 6 of MCo-PMI adopts an
-helical conformation, allowing the side chains of residues Phe43, Trp47, and Leu50 to
bury deep into the hydrophobic cleft of the p53 binding domain of HDM2 (Fig. 2.7b).
When compared to the linear PMI, the cyclotide MCo-PMI binds to HDM2 in a very similar
fashion. The side chain conformation of several key residues (Phe42, Trp46, and Leu49)
in MCo-PMI and the same residues in PMI take nearly identical positions in the complex
with p53 binding domain of HDM2 and make the same type of interactions with the
residues of HDM2 (Fig 2.7c). It is worth pointing out that the loop 2 of MCo-PMI is located
within Van der Waals distance to the p53 binding domain of HDM2 and may favorably
contribute to the observed increase binding affinity of MCo-PMI to the p53 binding
domain of HDM2 (Fig. 2.7d). In 9 out of 10 of the lowest energy structures, the negatively
charged residue Asp35 on loop2 of MCo-PMI was located within 2 Å distance from the
positively charged residue Lys116 of HDM2, indicating the potential formation of a salt
bridge between MCo-PMI and HDM2 to further stabilize the complex.

59

Figure 2.7. Solution structure of the MCo-PMI and HDM2 (17−125) complex. (a) Ribbon
representation of MCo-PMI (purple) and HDM2 (cyan blue) complex. The side chains of Phe42,
Trp46, and Leu49 in MCo-PMI and the HDM2 residues shaping the hydrophobic binding pocket
are shown as sticks in purple and cyan blue, respectively. (b) Close-up view of the binding
interface within the HDM2−MCo-PMI complex. The electrostatic potential at the molecular surface
of HDM2 is shown as positive in blue, negative in red, and non-charged in white. (c) Ribbon
representation of the backbone superposition of MCo-PMI (purple) and the PMI peptide (green)
(PDB code: 3EQS) complexed with HDM2 (light blue) and MDM2 (blue), respectively. The key
side chains of Phe, Trp, and Leu in the MCo-PMI cyclotide and PMI peptide are shown as sticks.
The residues lining the cavity of HDM2 and MDM2 are also shown. (d) Close-up view of the
HDM2−MCo-PMI complex reveals an additional salt bridge interaction between Asp35 (MCo-PMI)
and Lys51 (HDM2).
60

2.4 Serum stability of cyclotide MCo-PMI
Naturally occurring MCoTI-cyclotides present a very rigid structure due to the cystine
knot topology, which makes them ultra-stable to physical, chemical and proteolytic
degradation. To test the stability of wild-type and grafted cyclotides, serum was used to
mimic in vivo conditions. Cyclotide MCoTI-I showed a half-life of more than 2 days ( 1/2=
55 ± 5 h) in human serum at 37 °C (Fig. 2-8). The grafted cyclotide MCo-PMI was only
slightly less stable ( 1/2= 30 ± 4 h) in serum than the parent cyclotide. In contrast, the
linearized, reduced, and alkylated version of MCo-PMI was degraded rapidly under the
same conditions ( 1/2= 0.7 ± 0.1 h), which indicates the importance of circular cystine knot
for stability.
The cyclotides MCoTI-I and MCo-PMI were found to be ≈80% and 99% bound to the
serum proteins, respectively, under the conditions used in our study. The association and
dissociation constant rates of MCo-PMI to human serum proteins were measured
indicating that MCo-PMI binds to serum proteins with association and dissociation
constant rates of 2.4 × 10
3
M
−1
s
−1
and 2.2 × 10
−2
s
−1
, respectively (Fig. 2.9). The resulting
dissociation constant was estimated to be around 10 µ M.

Figure 2.8. Serum stability of cyclotides MCo-PMI and MCoTI-I, and a linearized and S-alkylated
version of MCo-PMI at 37° C.
61

Figure 2.9. Binding kinetics of cyclotide MCo-PMI to human serum proteins.

2.5 Biological activity using cell-based assays
2.5.1 Cell availability assay
As described above, one of the most interesting properties of cyclotide is cell
permeability. The ability of the cyclotide MCo-PMI to target intracellular HDM2 and HDMX
prompted us to investigate its effect on cell viability by treating a panel of solid tumor cells
expressing wild-type p53 and different levels of HDM2 and/or HDMX. The cell lines used
here includes the HDM2 and HDMX expressing prostate and colon cancer cell lines,
LNCaP and HCT116 p53
+/+
and the HDMX overexpressing human choriocarcinoma cell
line JEG-3. In order to study the p53 dependence on the cytotoxic activity of MCo-PMI
cyclotides, we used several p53 mutant or null cell lines, including human prostate cancer
line PC3 (bearing P274L and V233F p53 mutation) and DU145 (bearing a base pair
deletion at codon 138 which generates a stop codon at position 179) and a p53 deficient
62

HCT116 p53
-/-
cell line. We also included breast and kidney epithelial cell lines HBL-100
and HEK293T to evaluate the toxicity of MCo-PMI cyclotides to non-tumorigenic cells.
Cytotoxicity assays were performed by treating cultured cells for 48 h with serial
dilutions of nutlin-3, MCo-PMI, MCo-PMI-F42A, and MCoTI-I. Cell viability after treatment
was measured by MTT assay (Fig. 2.10). The cyclotide scaffold MCoTI-I showed no
detectable cytotoxicity to any of the cells tested in this study up to 100 µ M. The lack of
cytotoxicity observed in MCoTI-I is in agreement with previously published data indicating
that MCoTI-cyclotides are nontoxic at a concentration up to 100 µ M (Contreras et al.,
2011), which confirm that this scaffold is suitable for peptide-based therapeutics. The
cyclotide MCo-PMI showed a dose-dependent cytotoxicity in all three cell lines tested
with wild-type p53 phenotypes, suggesting that MCo-PMI can reactivate the p53 pathway
efficiently in the cell lines expressing high level of HDM2, HDMX, or both. The most
sensitive cell line to MCo-PMI tested in this study was HCT116 p53
+/+
(EC 50 ≈ 2 μM),
while the HDMX-overexpressing cell lines LNCaP and JEG3 were about 10-fold less
sensitive. As expected, p53 mutated or deletion cell lines PC3, DU145, and HCT116
p53
-/-
were unaffected by MCo-PMI or Nutlin-3 treatment. Importantly, the cyclotide
MCo-PMI only showed little cytotoxicity to non-tumorigenic HBL100 and HEK293T
epithelial cells (EC 50 > 100 μM). In contrast, nutlin-3 showed moderately cytotoxic to the
normal breast epithelial HBL-100 cell line (EC 50 = 33 ± 5 μM). It is also worth noting that
the mutant cyclotide MCo-PMI-F42A was completely inactive in all the cell lines tested in
this work, indicating the specificity of the biological activity of MCo-PMI.
63

Figure 2.10. Cell viability of cancer and normal cells exposed to MCo-PMI cyclotides. Different
cancer cell lines expressing different levels of HDM2 and HDMX (HCT116 p53
+/+
, LNCaP, and
JEG3), nonfunctional p53 (HCT 116 p53
−/−
, DU145, and PC3), and nontumor cells (HBL100 and
HEK293T) were treated with 0−100 μM, Nutlin-3, MCo-PMI, MCo-PMI-F42A, and MCoTI-I for 48 h.
Cell viability was assessed by using the MTT assay. Data are the mean ± SEM of experiments
performed in triplicate.
64

2.5.2 Characterization of the mechanism of action of cyclotide MCo-PMI
To investigate whether the cytotoxicity activity of MCo-PMI was derived from the
stabilization of p53 pathway, we treated LNCap cells for 48 hs with vehicle (PBS), nutlin-3,
and cyclotides MCoTI-I, MCo-PMI, and MCo-PMI-F42A. The cell lysate was analyzed by
Western blotting to visualize the level of different proteins associated with the p53
pathway, including p53, HDM2, HDMX and p21 (Fig. 2.11). Treatment of LNCaP cells
with nutlin-3 (20 µ M) and cyclotide MCo-PMI (50 µ M) increased the levels of p53. In
contrast, treatment with the parent cyclotide MCoTI-I or the inactive mutant
MCo-PMI-F42A at the same concentration had minimal effect on the p53 protein level
when comparing it to the untreated cells (Fig. 2.11a).
As expected, the level of endogenous HDM2 also increased in cells treated with nutlin-3
or cyclotide MCo-PMI. These results are consistent with an intact p53-HDM2 counter
regulatory mechanism where HDM2 transcription and expression are under the control of
p53 (Fig. 2.11d). Also, as downstream protein transcriptionally regulated by p53, the
cyclin-dependent kinase inhibitor p21 level was up-regulated by nutlin-3 and MCo-PMI
treatment. The levels of HDM2 and p21 were not changed in the cells treated with the
inactive mutant MCo-PMI-F42A when compared to the vehicle, which highlights the
specificity of the cyclotide MCo-PMI to modulate the p53 signaling pathway. As expected,
the intracellular levels of HDMX were down-regulated by both nutlin-3 and cyclotide
MCo-PMI, whereas no effect was observed with the inactive mutant MCo-PMI-F42A and
MCoTI-I. This result is consistent with the high affinity of MCo-PMI for HDM2 and HDMX,
which inhibits binding of endogenous p53 to the HDM2-HDMX complex, preventing its
degradation. The stabilization of p53 up-regulates the expression of HDM2, which then
promote the ubiquitination and degradation of HDMX (de Graaf et al., 2003).
The up-regulation of HDM2 and p53 showed a dose-dependent relationship with EC50
65

values of ~20 μM (p53) and ~15 μM (HDM2) (Fig. 2.11b), which is consistent with the
EC 50 values obtained in the cell viability assay for this cell line. The up-regulation of p21
and down-regulation of HDMX were also dose-dependent with EC 50 values of ≈30 μM. To
investigate the kinetics of p53 activation, we treated LNCap cells with 50 µM MCo-PMI for
1 h and evaluated the p53 level at different times upon treatment (Fig. 2.11c). Cells
exposed to MCo-PMI treatment demonstrated increased p53 levels that peaked at 36-48
hours after treatment. A similar trend was found for the up-regulation of p21 and HDM2
and for the down-regulation of HDMX.

Figure 2.11. Cyclotide MCo-PMI activates the p53 tumor suppressor pathway. (a) LNCaP
cells were exposed to cyclotides (50 µ M) MCo-PMI, MCo-PMI-F42A, MCoTI-I, and
nutlin-3 (50 µ M) for 1 h. After 48 h, the soluble cell extracts were analyzed by SDS−PAGE
and Western blotting for p53, HDM2, HDMX, and p21. (b) LNCaP cells were exposed to
different concentrations of MCo-PMI (0−100 μM) for 48 h. Cell lysates were analyzed for
p53, HDM2, HDMX, and p21 as described above. (c) LNCaP cells were treated with
MCo-PMI (50 μM) for 1 h. The amount of p53, HDM2, HDMX, and p21 was evaluated by
66

Western blot after 0−48 h of treatment. (d) Protein level changes after blocking the
interaction between p53 and HDM2.
2.5.3 Other cell-based assays
First, to determine whether MCo-PMI mediated stabilization of p53 could inhibit cancer
cells by reactivating the apoptotic pathway, we performed a caspase-3/7 assay using
LNCap cells treated with MCo-PMI and MCo-PMI-F42A for 30 h (Fig. 2.12). The result
showed that treatment with either cyclotide MCo-PMI or nutlin-3 could induce
dose-dependent caspase-3 activation. Cells treated with MCo-PMI-F42A did not show
any caspase-3/7 activity.

Figure 2.12. Cyclotide MCo-PMI triggers apoptosis in a dose-dependent fashion. LNCaP cells
were treated with vehicle, MCo-PMI (0-100 μM) or Nutlin-3 (0-50 μM) for 1 h. The amount of
caspase-3/7 was evaluated after 24 h by exposure to Caspase-Glo 3/7 reagent (Promega).
Caspase 3/7 activation was measured by monitoring the luminescence signal upon proteolytic
cleavage of a luciferin-containing caspase-3/7 substrate in the presence of luciferase.

We also evaluated cell cycle arrest in LNCap cells treated with MCo-PMI, MCo-PMI-F42A,
parent cyclotide MCoTI-I, and nutlin-3 for 24 hours by using the propidium iodide (PI)
cytometric assay (Fig. 2.13). Both MCo-PMI and nutlin-3 induced cell cycle arrest,
resulting in a depression of S-Phase fraction. The reduction of S-phase was associated
67

with accumulation of cells in the G0/G1 phase, suggesting that the treatment of MCo-PMI
and nutlin-3 impedes the cell cycle progression at the G1/S checkpoint, which is in
agreement with the upregulation of p21 observed in cells treated with MCo-PMI and
nutlin-3 (Fig. 2.11a). Treatment of MCo-PMI-F42A did not induce cell cycle arrest.

Figure 2.13. Cyclotide MCo-PMI triggers cell cycle arrest. LNCaP cells were treated with vehicle,
MCo-PMI (50 μM), MCo-PMI-F42A (50 μM) and Nutlin-3 (50 μM) for 1 h. Cell cycle progression
was monitored 24 h after treatment by propidium iodide staining and fluorescence-activated cell
sorting.

Altogether, these data confirm that in-cell disruption of p53-HDM2 and p53-HDMX
complexes by MCo-PMI in LNCaP cells leads to the up-regulation of p53 transcriptional
targets (HDM2 and p21), caspase-3 activation, and induction of cell cycle arrest at G1/S
checkpoint.
2.6 Characterization using a colon carcinoma mouse xenograft model
Based on the results obtained in vitro, the cyclotide MCo-PMI was also tested to evaluate
if it could also modulate the p53 tumor suppressor pathway in vivo and therefore inhibit
tumor growth. This experiment was accomplished by using a murine xenograft model
68

using the human colon carcinoma HCT116 cell line. HCT116 p53
+/+
xenografts were
established by injecting 0.5 × 10
6
cells subcutaneously into the teat right flanks of female
nude mice (nu/nu). When the tumor reached an average volume of ≈100 mm
3
as
determined by caliper measurements, cohorts (n=3) were treated intravenously with
vehicle (5% dextrose in water), MCo-PMI (40 mg/kg, 7.6mmole/kg), or nutlin-3 (10
mg/kg,17.2mmol/kg) daily for up to 37 days. The tumor size was measured every day by
caliper measurement (Fig. 2.14). Treatment with MCo-PMI significantly suppressed
tumor growth when compared to animals treated only with vehicle (≈85% reduction at day
31, p=0.019) and nutlin-3 (≈75% reduction at day 31, p=0.022). In contrast, animals
treated with nutlin-3 only showed a moderated reduction in tumor growth (≈40% reduction
at day 31, p=0.223).

Figure 2.14. Cyclotide MCo-PMI activates the p53 tumor suppressor pathway and blocks
tumor growth in vivo. Cohorts (N = 3) of HCT116 p53
+/+
xenografts mice were treated with
vehicle (5% dextrose in water), MCo-PMI (40 mg/kg, 7.6 mmol/kg), or nutlin-3 (10 mg/kg,
17.2 mmol/kg) by intravenous injection daily for up to 38 days. Tumor volume was
monitored by caliper measurement. Data are the mean ± SEM (day 31: MCo-PMI/vehicle,
p = 0.019, MCo-PMI/Nutlin-3, p = 0.022, and Nutlin-3/vehicle, p = 0.223).

At the end of the treatment, the animals were sacrificed and the tumors excised. Snap
frozen tumor samples were analyzed by qRT-PCR using HDM2 and p21 primer sets.
69

Analysis of the tumor tissue from animals treated with MCo-PMI and nutlin-3 showed a
statistically significant transcriptional activation of HDM2 (≈2.5 times) and p21 (≈4 times)
when compared to animals treated with just vehicle (Fig. 2.15a). Protein expression
levels of HDM2 and p21 in tumors treated with MCo-PMI peptide were measured by
western blotting and immunohistochemical staining. Tumors treated with cyclotide
MCo-PMI and nutlin-3 showed a marked increase in p53, HDM2, and p21 expression
when compared to vehicle-treated tumors (Fig. 2.15b-c).

Figure 2-15. Cyclotide MCo-PMI activates the p53 tumor suppressor pathway in vivo. (a) Tumors
were excised on day 31 (vehicle), day 36 (Nutlin-3), and day 38 (MCo-PMI), and the level of p53
transcriptional targets HDM2 and P21 was measured after RNA extraction by qRT-PCR
(MCo-PMI/vehicle, p = 0.019, MCo-PMI/Nutlin-3, p = 0.022, and Nutlin-3/vehicle, p = 0.223). Data
are the mean ± SEM. (b) Tumors samples were also subjected to SDS−PAGE and analyzed by
Western blotting for p53, HDM2, and p21. (c) The expression level of p53, HDM2, and p21 was
also assessed by immunohistochemical staining. GAPDH was used as loading control in Western
blots.
70

2.7 Optimization of MCo-PMI activity using a molecular constraint approach
With the purpose of further enhancing the affinity of the grafted cyclotide MCo-PMI
for the p53 binding domains of HDM2 and HDMX, an extra disulfide bond was introduced
into the cyclotide structure by replacing two original amino acids with Cys residues. In
order to constrain the -helical structure of PMI, the pair of residues 2/20 and 4/17, which
are in close proximity to each other on the structure of MCo-PMI, were selected.
Therefore, residues Gly2/Ser20 and Gly4/Asn17 were replaced with cysteines to give
cyclotides MCo-PMI-C2-20 and MCo-PMI-C4-17, respectively (Fig. 2.16). Therefore,
cyclotides MCo-PMI-C2-20 and MCo-PMI-C4-17 contain an extra disulfide bond at the C-
and N-termini of the grafted PMI helical structure, respectively. The hypothesis is that the
introduction of such constraint will stabilize the a-helix structure of the PMI peptide
segment improving its biological activity.

Figure 2.16. Design and preparation of the Cys-mutant MCo-PMI cyclotides. To increase the
binding affinity of the cyclotide MCo-PMI to HDM2 and HDMX, a pair of cysteines were introduced
into the original structure, which is shown in yellow.

71

The DNA encoding MCo-PMI-C2-20 and MCo-PMI-C4-17 were cloned into an intein
expression vector as previously described for MCo-PMI (section 2.1.2). Once the intein
precursor protein was expressed and purified, the N-terminal TEV protease recognition
sequence was removed by TEV protease. The backbone cyclization and oxidative folding
were performed with reduced glutathione (GSH) at pH 7.2 in one single step same as
previously described for MCo-PMI (section 2.1.2). After purification by preparative
reversed-phase HPLC, the purity of the MCo-PMI-derived peptides was determined by
analytical RP-HPLC and electrospray mass spectrometry (ES-MS) as shown in Figure
2.17.

Figure 2.17. Analytical C18-reversed-phase HPLC traces and electrospray mass spectra of
purified MCo-PMI cyclotides. HPLC analysis was performed using a linear gradient of 0-70%
solvent B over 30 min. The MCo-PMI and MCo-PMI-C2-20 samples showed a clear single peak on
HPLC and single mass. The MCo-PMI-C4-17 sample showed two peaks on HPLC with the same
mass, which indicates they are isomers.
72

Competition inhibition assays with the wild-type MCo-PMI C2-20 and C4-17 were
conducted by using a FRET reporter formed by fusing the fluorescent proteins YPet and
CyPet to PMI and HDM2 respectively. As shown in Figure 2.18a, the interaction between
PMI and HDM2 will bring fluorescence proteins YPet and CyPet in close in proximity,
resulting in a FRET. When there is inhibitor present, the two fusion proteins YPet-PMI and
CyPet-HDM2 will disassociate causing a decrease of the FRET signal. By measuring the
FRET signal, we can quantitatively monitor the inhibition of the PMI-HDM2 interaction by
MCo-PMI cyclotides (Fig. 2.18b). The isomer mixture of MCo-PMI C4-17 showed slight
lower IC 50 (64.5 ± 10.3 nM) comparing to the wild-type sample (81.3 ± 12.0 nM), while the
IC 50 of MCo-PMI C2-20 (249.4 ± 27.2 nM) was significantly higher than the wild-type
MCo-PMI. This result suggested to further investigate the components present in the
oxidative folding of MCo-PMI C4-17 worth the further research.

73

Figure 2.18. FRET-based inhibition assay. (a) The principle of FRET-based reporter system.
Briefly, the presence of inhibitor will dissociate two FRET proteins. The change in FRET signal is
measured at 525 nm (YPet-PMI) by excitation at 414 nm (CyPet-HDM2). (b) competition
experiments of MCo-PMI peptides. Binding competition experiments were performed by titrating a
solution of YPet−PMI (50 nM) and CyPet−HDM2 (100 nM) with increasing concentrations of
unlabeled inhibitor. The decrease in FRET signal was measured at 525 nm (YPet) by excitation at
414 nm (CyPet). Data are the mean ± SEM of experiments performed in triplicate.

2.8 Conclusion
In summary, we have shown that the cyclotide MCoTI-I can be successfully engineered
to display and stabilize -helical peptide fragments and that the resulting cyclotide can be
efficiently used both in vitro and in vivo to target intracellular interactions. These
properties make engineered MCo-cyclotides superior in many aspects to other alternative
peptide- or mini protein-based scaffolds to modulate intracellular protein−protein
interactions, and therefore opening new approaches for in the design of novel
peptide-based therapeutic leads.
2.9 Materials and methods
2.9.1 Materials and instrumentation
Analytical HPLC was performed on an HP1100 series instrument with 220 nm and 280
nm detection using a Vydac C18 column (5 μm, 4.6 x 150 mm) at a flow rate of 1 mL/min.
Semipreparative HPLC was performed on a Waters Delta Prep system fitted with a
Waters 2487 Ultraviolet-Visible (UV-vis) detector using a Vydac C18 column (15-20 μm,
10 x 250 mm) at a flow rate of 5 mL/min. All runs used linear gradients of 0.1% aqueous
trifluoroacetic acid (TFA, solvent A) vs. 0.1% TFA, 90% acetonitrile in H2O (solvent B).
UV-vis spectroscopy was carried out on an Agilent 8453 diode array spectrophotometer,
and fluorescence analysis on a Flurolog-3 spectrofluorimeter (Horiba Scientific).
Electrospray mass spectrometry (ES-MS) analysis was routinely applied to all cyclized
74

peptides. ES-MS was performed on an Applied Biosystems API 3000 triple quadrupole
electrospray mass spectrometer using Analyst 1.4.2. LC-MS was performed on an
HP1100 HPLC/API-3000 system using a multiple reaction monitoring (MRM) mode.
Calculated masses were obtained by using ProMac v1.5.3. Protein samples were
analyzed by SDS-PAGE. Samples were run on Invitrogen (Carlsbad) 4-20% Tris-Glycine
Gels. The gels were then stained with Pierce (Rockford) Gelcode Blue,
photographed/digitized using a Kodak (Rochester) EDAS 290, and quantified using NIH
Image-J software (http://rsb.info.nih.gov/ij/). DNA sequencing was performed by the DNA
Sequencing and Genetic Analysis Core Facility at the University of Southern California
using an ABI 3730 DNA sequencer, and the sequence data were analyzed with DNAStar
Lasergene v5.5.2. All chemicals were obtained from Sigma-Aldrich unless otherwise
indicated. Nutlin-3 was purchased from sellechchem.com (www.sellechchem.com).
2.9.2 Molecular cloning and protein expression
Construction of cyclotide expressing plasmids. Plasmids expressing the MCo-PMI
cyclotides were constructed using the pTXB1 expression plasmid (New England Biolabs),
which contain an engineered Mxe GyrA intein, respectively, and a chitin-binding domain
(CBD). Oligonucleotides coding for the different MCo-PMI variants were synthesized,
phosphorylated and PAGE purified by IDT DNA. Complementary strands were annealed
in 20 mM sodium phosphate, 300 mM NaCl and the resulting double-stranded DNA
(dsDNA) was purified using Qiagen’s (Valencia, CA) miniprep column and buffer PN.
pTXB1 plasmids were double digested with NdeI and SapI (NEB). The linearized vectors
and the MCoTI-I encoding dsDNA fragments were ligated at 16° C overnight using T4
DNA Ligase (New England Biolabs). The ligated plasmids were transformed into DH5α
cells (Invitrogen) and plated on Luria Broth (LB)-agar containing ampicillin. Positive
colonies were grown in 5 mL LB containing ampicillin at 37° C overnight and the
corresponding plasmids purified using a Miniprep Kit (Qiagen). Plasmids expressing the
75

MCo-PMI cyclotide precursors with an N-terminal TEV recognition sequence
(Met-Glu-Asn-Leu-Tyr-Phe-Gln) were cloned as follows. The DNA encoding TEV
N-terminal recognition sequence was generated by PCR using the corresponding
MCo-PMIpTXB1 plasmid. The 5' primer (5'- AAA CAT ATG GAA AAC CTG TAC TTC
CAG TGC GGT TCT GGT TCT GG-3’) encoded a Nde I restriction site. The 3'
oligonucleotide (5'-GAT TGC CAT GCC GGT CAA GG-3’) introduced a Spe I restriction
site during the PCR reaction. The PCR amplified product was purified, digested
simultaneously with Nde I and Spe I and then ligated into a Nde I- and Spe I-treated
plasmid pTXB-1 (New England Biolabs). The linearized vectors and the TEV-MCo-PMI
encoding dsDNA fragments were ligated at 16° C overnight as described above. The
ligated plasmids were transformed into DH5α cells and screened as described above.
The DNA sequence of all the plasmids was confirmed by sequencing.
Cloning and expression of fluorescent protein YPet-p53. The DNA encoding the
fluorescent protein YPet was isolated by PCR using the plasmid pBAD-6 as a template.
The forward (5’- AAA AGG ATC CGA TGT CTA AAG GTG-3’) and reverse (5’-TTT TGA
GCT CTT TGT ACA ATT CAT TC-3’) primers contained a BamHI and SacI restriction site,
respectively. The resulting amplicon was purified using Qiagen’s PCR purification kit,
digested and ligated into BamHI and SacI-treated plasmid pRSF-DUET-1 (Novagen) to
give T7 expression vector pRSF-YPet. The resulting plasmid was sequenced and shown
to be free of mutations. 5’-Phosphorylatedsynthetic DNA oligos (IDT) (5’-C GGT GGT
TCT GGT GGT TCT GGT GGT TCT GGT GGT TCT GGT GGT TCT CTG CAG AGT
CAG GAA ACA TTT TCA GAC CTA TGG AAA CTA CTT CCT GAA AAC TAA G-3’ and
5’-TC GAC TTA GTT TTC AGG AAG TAG TTT CCA TAG GTC TGA AAA TGT TTC CTG
ACT CTG CAG AGA ACC ACC AGA ACC ACC AGA ACC ACC AGA ACC ACC AGA
ACC ACC GAG CT-3’) encoding a flexible linker (Gly-Gly-Ser) (Richard H Kimura,
Steenblock, & Camarero, 2007) fused in frame to the DNA encoding human p53 (15-29
aa) were annealed and ligated into pRSF-YPet using the Sac I and Sal I restriction sites
76

to give pRSF-YPet-p53. The resulting plasmid was sequenced and shown to be free of
mutations. BL21(DE3) cells (Novagen) (1L) transformed with pRSF-YPet-p53 plasmid
were grown to mid-log phase (OD at 600 nm ≈ 0.6) in LB medium containing kanamycin
(34 μg/L) at 37° C and then induced with 1 mM IPTG at 30° C for 4 hours. Cells were
harvested, lysed and YPet-p53 purified by Ni-affinity chromatography as described above.
YPet-p53 protein was eluted with 50 mM sodium phosphate, 250 mM imidazole, 300 mM
NaCl buffer at pH 8.0 containing 30% glycerol. The purified proteins were immediately
dialyzed against the same buffer with no imidazole and stored at -80° C until use. The
purified protein was characterized by SDS-PAGE and ES-MS.
Cloning and expression of CyPet-HDM2. The DNA encoding the fluorescent protein
CyPet was isolated by PCR using the plasmid pBAD-6 [5] as a template. The forward
(5’-AAA AGG ATC CAA TGT CTA AAG GTG AAG-3’) and reverse (5’- TTT TGA GCT
CTT TGT ACA ATT CAT -3’) primers contained a BamH I and Sac I restriction site,
respectively. The resulting amplicon was purified using Qiagen’s PCR purification kit,
digested and ligated into BamH I and Sac I-treated plasmid pRSF-DUET-1 (Novagen) to
give T7 expression vector pRSF-CyPet. The resulting plasmid was sequenced and
shown to be free of mutations. The DNA encoding the p53 binding domain of the human
homolog of Mdm2 (HDM2, residues 17-125) was isolated by PCR using the cDNA for
HDM2 (accession number: BT007258) as a template. The forward primer (5’-T GCA CTG
CAG TCA CAG ATT CCA GCT TCG GAA C-3’) encoded a Pst I restriction site. The
reverse primer (5’- GC GTCGAC TTA GTT CTC ACT CAC AGA TGT ACC TGA-3’)
encoded a Sal I site and a stop codon. The resulting amplicon was purified using
Qiagen’s PCR purification kit, digested and ligated into Pst I-and Sal I-treated plasmid
pRSFCyPet to give T7 expression vector pRSF-CyPet-HDM2. The resulting plasmid was
sequenced and shown to be free of mutations. BL21(DE3) cells (Novagen) (1L)
transformed with pRSFCyPet-HDM2 plasmid were grown to mid-log phase (OD at 600
nm ≈ 0.6) in LB medium containing kanamycin (34 μg/mL) at 37° C and then induced with
77

1 mM IPTG at 30° C for 4 h. Cells were harvested, lysed and CyPet-HDM2 purified by
Ni
2+
-affinity chromatography and stored as described before for YPet-p53. Purified
CyPet-HDM2 was characterized by SDS-PAGE and ES-MS.
Cloning and expression of CyPet-HDMX. The DNA encoding the p53 binding domain
of the human homolog of MdmX (HDMX, residues 17-116) was isolated by PCR using the
cDNA for HDMX (accession number: BC105106) as a template. The forward primer (5’- T
GCA CTGCAG TGC AGG ATC TCT C-3’) encoded a Pst I restriction site. The reverse
primer (5’- GC GTC GAC TTA AGC ATC TGT AGT AGC AGT-3’) encoded a Sal I site
and a stop codon. The resulting amplicon was purified using Qiagen’s PCR purification kit,
digested and ligated into PstI- and SalI-treated plasmid pRSF-CyPet to give T7
expression vector pRSF-CyPet-HDMX. Rosetta(DE3) cells (Novagen) (1L) transformed
with pRSF-CyPet-HDMX plasmid were grown to mid-log phase (OD at 600 nm ≈ 0.6) in
LB medium containing kanamycin (34 mg/L) and chloramphenicol (34 μg/mL) at 37° C
and then induced with 1 mM IPTG at 30° C for 4 h. Cells were harvested, lysed and
CyPet-HDMX purified by Ni-affinity chromatography and stored as described above for
YPet-p53. Purified CyPet-HDM2 was characterized by SDS-PAGE and ESMS.
Expression and purification of recombinant MCo-PMI cyclotides. BL21(DE3)
(Novagen) were transformed with MCo-PMI encoding plasmids (see above). Expression
was carried out in LB medium (1 L) containing ampicillin (100 μg/mL) at 30°C for 4 h
respectively. Briefly, 5 mL of an overnight starter culture derived from either a single clone
or single plate were used to inoculate 1 L of LB media. Cells were grown to an OD at 600
nm of ≈ 0.6 at 37° C, and expression was induced by the addition of
isopropyl-β-D-thiogalactopyranoside (IPTG) to a final concentration of 0.3 mM at 30° C
for 4 h. The cells were then harvested by centrifugation. For fusion protein purification,
the cells were resuspended in 30 mL of lysis buffer (0.1 mM EDTA, 1 mM PMSF, 50 mM
sodium phosphate, 250 mM NaCl buffer at pH 7.2 containing 5% glycerol) and lysed by
sonication. The lysate was clarified by centrifugation at 15,000 rpm in a Sorval SS-34
78

rotor for 30 min. The clarified supernatant was incubated with chitin-beads (2 mL beads/L
cells) (New England Biolabs), previously equilibrated with column buffer (0.1 mM EDTA,
50 mM sodium phosphate, 250 mM NaCl buffer at pH 7.2) at 4° C for 1 h with gentle
rocking. The beads were extensively washed with 50 bead-volumes of column buffer
containing
0.1% Triton X100 and then rinsed and equilibrated with 50 bead-volumes of column
buffer. For the purification of TEV-MCo-PMI-intein-CBD fusion proteins, the beads were
washed with 50 bead-volumes of TEV reaction buffer (50mM Tris-HCl, 0.5mM EDTA pH
8.0). Proteolytic cleavage of the TEV sequence was performed on the column by
complementing the buffer with 3 mM reduced GSH and adding TEV protease to a final
concentration of ≈ 0.1 mg/mL. The proteolytic reaction was kept at 4° C overnight with
gentle rocking. Once the proteolytic step was completed, the column was then washed
with 50-bead volumes of column buffer. Chitin beads containing the different purified
MCo-PMI-Intein-CBD fusion proteins were cleaved with 50 mM GSH in degassed column
buffer. The cleavage reactions were kept for up to 1-2 days at 25° C with gentle rocking.
Once the cleavage reaction was complete, the supernatant of the cleavage reaction was
separated by filtration and the beads were washed with additional column buffer to reach
a final concentration of 5 mM GSH, and the folding was allowed to proceed with gently
rocking at 4° C for 48 h. Folded MCo-PMI cyclotides were purified by semipreparative
HPLC using a linear gradient of 25-45% solvent B over 30 min. Purified MCo-PMI
cyclotides were characterized by C18-RP-HPLC and ES-MS, and quantified by UV-vis
spectroscopy.
Refolding of TEV-MCo-PMI-intein-CBD constructs from inclusion bodies. Inclusion
bodies were first washed with column buffer containing 0.2% Triton X (50 mL) and then
just column buffer (3 x 50 mL). The pellet was dissolved in 50 mM sodium phosphate and
250 mM NaCl, 8 M urea buffer at pH 7.2 (10 mL). After centrifugation at 15,000 rpm in a
Sorval SS-34 rotor the supernatant was slowly flash diluted in 0.1mM EDTA, 50 mM
79

sodium phosphate and 250 mM NaCl, 0.5 M Arg•HCl buffer at pH 7.2. This solution was
dialyzed against column buffer (2 L) at 4° C for 2 days. The dialyzed solution was
centrifuged at 15,000 rpm for 20 min in a Sorval SS-34 rotor and the supernatant was
purified by affinity chromatography on chitin beads. TEV-MCo-PMI-intein-CBD constructs
were treated as described above to remove the TEV-leading signal and induce backbone
cyclization/folding of the MCo-PMI cyclotides.
Expression of
15
N-labeled MCo-PMI. Expression was carried out using BL21(DE3) cells
as described above except grown in M9 minimal medium containing 0.1%
15
NH4Cl as the
nitrogen source. Cyclization and folding were performed in solution as described above.
15
N-labeled MCo-PMI was purified by semipreparative HPLC as before. Purified products
were characterized by HPLC and ES-MS.
Preparation of FITC-labeled MCo-PMI cyclotides. MCo-PMI-K37R and
MCo-PMI-K37R-F42A were prepared either by chemical synthesis or recombinant
expression as described above. The pTXB1-TEV-MCo-PMI-K37R and
pTXB1-TEV-MCo-PMI-K37R-F42A plasmids were prepared by mutagenesis using
pTXB1-TEV-MCo-PMI or pTXB1-TEV-MCo-PMI- F42A plasmids as template,
respectively, and the forward primer (5’ - T GGT GCT TCT CGT GCT CCG ACC TC -3’)
and reverse primer (5’- G AGG TCG GAG CAC GAG AAG CAC CA - 3’) in both cases.
MCoTI-PMI-K37R and MCo-PMI-K37R-F42A were purified and characterized as
described before. MCo-PMI cyclotides (100 μg) were mixed with 5 times excess FITC
(molar ratio) in 0.1 M sodium bicarbonate buffer at pH 9.0. The reaction was carried out at
room temperature in the dark for 2 h. The labeling reaction was quenched with diluted
AcOH and the labeled cyclotide purified by C18 Semi-prep HPLC using a linear gradient
of 10-45% solvent B over 30 min. The pure labeled peptide was characterized by HPLC
and ES-MS.
TEV protease expression and purification. BL21(DE3) cells were transformed plasmid
pRK793, which encodes His-tagged TEV protease (Addgene). Expression was carried
80

out in 1 L of LB medium containing ampicillin (100 μg/mL) and chloramphenicol (34
μg/mL) at 30°C for 4 h. Briefly, 5 mL of an overnight starter culture derived from a single
clone was used to inoculate 1 L of LB media. Cells were grown to an OD at 600 nm of ≈
0.6 at 37° C, and expression was induced by the addition of IPTG to a final concentration
of 1 mM at 30° C overnight. The cells were harvested by centrifugation, resuspended in
30 mL of lysis buffer (0.1 mM PMSF, 10 mM imidazole, 50 mM sodium phosphate, 300
mM NaCl buffer at pH 8.0 containing 5% glycerol) and lysed by sonication. The lysate
was clarified by centrifugation at 15,000 rpm in a Sorval SS-34 rotor for 30 minutes. The
clarified supernatant was incubated with 1 mL of Ni-NTA agarose beads (Qiagen)
previously equilibrated with Ni-NTA column buffer (20 mM imidazole, 50 mM sodium
phosphate, 300 mM NaCl buffer at pH 8.0) at 4° C for 1 hour with gentle rocking. The
Ni-NTA agarose beads were washed sequentially with Ni-NTA column buffer (2 x 100
mL). The fusion protein was eluted with Ni-NTA elution buffer (50 mM sodium phosphate,
250mM imidazole, 300 mM NaCl, buffer at pH 8) and immediately dialyzed in
TEV-protease storage buffer (1 mM EDTA, 5 mM DTT, 50mM Tris•HCl buffer at pH7.5
containing 50% (v/v) glycerol and 0.1% (w/v) Triton X-100). The purity of the TEV
protease was checked by SDS-PAGE.
Cloning and expression of HDM2 (17-125). The DNA encoding the p53 binding domain
of HDM2 (residues 17-125) was isolated by PCR using the cDNA for HDM2 (accession
number: BT007258) as a template. The forward (5’-AAA ACA TAT GTC ACA GAT TCC
AGC TTC G-3’) and reverse (5’-AAA AGG ATC CTT AGT TCT CAC TCA CAG ATG -3’)
primers contained a Nde I and BamH I restriction site, respectively. The resulting
amplicon was purified using Qiagen’s PCR purification kit, digested and ligated into Nde
I- and BamH I-treated plasmid pET28a (Novagen) to give T7 expression vector
pET28-HDM2. The resulting plasmid was sequenced and shown to be free of mutations.
BL21(DE3) cells (1L) transformed with pET28-HDM2 plasmid were grown to mid-log
phase (OD600 ≈ 0.6) in LB medium containing kanamycin (34 μg/mL) at 37° C and then
81

induced with 1 mM IPTG at 30° C for 4 h. Cells were lysed and the protein purified by
Ni-affinity chromatography as described above. HDM2 (17-125) was dialyzed against
PBS buffer (50 mM sodium phosphate and 150 mM NaCl pH 7.2 containing 30% glycerol)
and immediately used.
Cloning and expression of HDMX (17-116). The DNA encoding the p53 binding domain
of the human homolog of MDMX (HDMX, residues 17-116) was isolated by PCR using
the cDNA for HDMX (accession number: BC105106) as a template. The forward (5’-ATT
AGG ATC CTG CAG GAT CTC TCC TGG ACA AAT C-3’) and reverse (5’- ATT AAA
GCT TCT ACT AAG CAT CTG TAG TAG CAG TGG CTA AAG TG -3’) primers contained
a BamH I and Hind III restriction site, respectively. The resulting amplicon was purified
using Qiagen’s PCR purification kit, digested and ligated into BamH I- and Hind III-treated
plasmid pET28a (Novagen) to give T7 expression vector pET28-HDMX. The resulting
plasmid was sequenced and shown to be free of mutations. BL21(DE3) cells (1L)
transformed with pET28-HDMX plasmid were grown to mid-log phase (OD at 600 nm ≈
0.6) in LB medium containing kanamycin (34 μg/mL) at 37° C and then induced with 1
mM IPTG at 30° C for 4 h. Cells were lysed and the protein purified by Ni-affinity
chromatography as described above. HDMX (17-116) was dialyzed against PBS buffer
(50mM sodium phosphate and 150mM NaCl pH 7.2) and further purified by gel filtration
chromatography on a Superdex-75 column in PBS buffer. Pure fractions were pooled,
re-concentrated in 50mM sodium phosphate and 150mM NaCl pH 7.2 and immediately
used.
Expression of
15
N,
13
C-labeled HDM2 (17-125). Expression was carried out using
BL21(DE3) cells as described above except grown in M9 minimal medium containing 0.1%
15
NH4Cl and 0.2%
13
C6-D-glucose as the nitrogen and carbon sources, respectively.
Protein purification was performed as described above. Protein was characterized by
ES-MS.
82

2.9.3 In vitro characterization assays
Fluorescence polarization binding assays. Fluorescence polarization of FITC-labeled
MCo-PMI cyclotides upon addition of either HDM2 (17-125) or HDMX (17-116) was
measured at 22° C using a Spex Fluorolog 3 spectrofluorometer (Horiba Scientific) with
the excitation bandwidth set at 1 nm and emission at 5 nm. The excitation wavelength for
fluorescein was set at 495 nm and emission was monitored at 521 nm. The equilibrium
dissociation constant (KD) were obtained by titrating a fixed concentration of FITC-labeled
MCo-PMI cyclotide (5 nM) with increasing concentrations of either HDM2 (17-125) or
HDMX (17-116) in 50 mM sodium phosphate, 150 mM NaCl buffer at pH7.2 by assuming
formation of a 1:1 complex and using the Prism (GraphPad) software package.
In vitro inhibition p53-HDM2/HDMX competition experiments. In vitro IC50 values
were measured by inhibition competition experiments using the FRET-based reporter
formed by CyPet-HDM2/CyPet-HDMX and YPet-p53. Briefly, a solution of
CyPet-HDM2/HDMX (20 nM) and YPet-p53 (5 μM) in 10 mM phosphate buffer, 150 mM
NaCl buffer at pH 7.2 was titrated with increasing amounts of inhibitor (ranging from 0 to
100 μM). The decrease in fluorescence signal at 525 nm (excited at 414 nm) was
measured and plotted against the concentration of free inhibitor. The resulting plot was
fitted to a single binding site competition curve using the Prism (GraphPad) software
package.
In vitro inhibition PMI-HDM2 competition experiments. The FRET reporter mixture
was made by mixing CyPet-HDM2 and YPet-PMI in assay buffer (50 mM sodium
phosphate 150 mM NaCl and 1 mM EDTA pH 7.2) to a final concentration 50 nM and 100
nM respectively. Then the different concentration of inhibitor peptides (final concentration
from 0.1 nM to 1 µ M) was added to 96 well plates containing FRET protein mixture. A
mixture of FRET reporter and inhibitors were incubated at room temperature with gentle
shaking for 30 min. Then FRET signal was measured by EnVision plate reader (Perkin
Elmer) with excitation wavelength at 405 nm and emission wavelength at 535 nm. Data
83

were analyzed using Prism GraphPad.
2.9.4 Serum stability and serum protein binding assay
Human serum stability. Peptides (150 μg dissolved in 50 μL PBS) were mixed with 500
μL human serum and incubated at 37° C. Aliquot samples (50 μL) were taken at different
time points (0-120 h) and precipitated with 20% trichloroacetic acid (TCA). After
centrifugation, the pellet was dissolved in 200 μL of 8 M GdmCl. Both the supernatant
and solubilized pellet fractions were analyzed by HPLC and LC-MS/MS. Each experiment
was done in triplicate.
Human serum binding kinetics. Binding kinetics were carried out at 25 °C on a BLItz™
instrument (ForteBIO), using biotinylated MCo-PMI immobilized onto a streptavidin
coated biosensor tip. MCo-PMI (1 mg, 190 nmol) was conjugated with three-fold molar
excess of NHS-PEG4-biotin in 0.1 M sodium phosphate buffer (1.9 mL) at pH 7.4 for 1 h.
The reaction was quenched by adding 2% TFA until pH ≈ 4. Purification and desalting of
biotinylated MCo-PMI were performed on a Zeba spin desalting columns (Thermo
Scientific). Binding of MCo-PMI to human serum proteins was performed at 1/100 and
1/200 serum dilutions in 20 mM sodium phosphate, 100 mM NaCl buffer at pH 7.2, which
correspond to a concentration of 25 μM and 50 μM human serum albumin, respectively.
Serum proteins were allowed to bind to the MCo-PMI coated biosensor tip for 2 minutes
followed by a dissociation step of 2 minutes. Nonlinear regression analysis was
performed using Prism (GraphPad Software) to calculate the association (k on) and
dissociation (k off) rates, and corresponding KD value.
2.9.5 Cell-based assays
Cell viability assay. LnCaP, HCT116 p53
+/+
, HCT116 p53
-/-
, JEG3, DU145, PC3,
HEK293T and HBL100 cell lines were cultured in RPMI 1640 medium supplemented with
10% fetal calf serum, penicillin (50 IU/mL) and streptomycin (50 μg/mL) at 37° C in 5%
84

CO2. Cell viability was performed using the MTT assay. Briefly, ≈ 2 x 10
3
cells were
seeded in 96-well microtiter plates in 100 μL RPMI 1640 in the presence of 10% calf
serum. After 24 h incubation, the cells were washed with PBS and treated with 30 μL/well
of PBS or RPMI 1640 media containing the peptides or nutlin-3 at the indicated doses for
1 h at 37° C in 5% CO2. After 1 h, 210 μL/well of full complemented media was added and
the cells were grown for 48 h and then treated with 20 μL of a solution of
3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT, 5 mg/mL) for 2 h. The
medium was discarded and DMSO (100 μL/well) was added to each well and incubated
with gentle shaking for 20 min at room temperature. The absorbance at 595 nm of the
solution was analyzed using a Tecan Genios Multifunctional Microplate Reader (Tecan
System Inc) and the background at 670 nm subtracted.
Cell Cycle Assay. LnCaP cells were seeded at a density of 8 x 10
4
cells per well in
12-well plates and grown in RPMI 1640 in the presence of 10% calf serum for 24 h. Cells
were washed with PBS and incubated with 300 μL/well of PBS or RPMI 1640 containing
the peptides or Nutlin-3 at the indicated doses for 1 h at 37° C in 5% CO 2. After 1 h, 2.1
mL of full complemented media was added per well and the cells were grown for 24 h,
washed twice with ice-cold PBS, trypsinized and resuspended in 0.3 mL of ice-cold PBS
at a density of ≈ 2 x 10
5
cells/mL. To this solution, 0.7 mL of 100% EtOH was added
dropwise with gentle vortexing and stored at 4° C overnight. Cells were washed once with
ice-cold PBS, resuspended in PBS containing propidium iodide (10 μg/mL) and
DNAse-free RNAse A (100 μg/mL); and incubated at 37° C for 15 min. The suspension
was then analyzed in a FACSAria II (BD Biosciences).
Caspase-3/7 activation assay. LNCaP or HCT116 p53
+/+
cells were seeded in 96-well
plates at a density of ≈ 2 x 10
3
cells and treated with vehicle, Nutlin 3 (EMD Chemicals)
(20 μM), MCo-PMI (50 μM) or MCo-PMI-F42A (50 μM) as described before for the cell
cycle analysis assay. After 24 h of treatment, caspase-3/7 activity was measured by
addition of Caspase-Glo 3/7 chemiluminescence reagent (Promega) according to the
85

manufacturer’s protocol and the luminescence was measured using a Synergy H1 Hybrid
Multi-Mode Microplate Reader (BioTek).
2.9.6 Mice xenografts model experiments
Mice xenografts studies. HCT116 p53
+/+
xenografts were established by injecting 100
uL suspension of basal RPMI containing 0.5 x 10
6
cells into the rear right flanks of female
nude mice (nu/nu) mice (Simonsen Laboratories). When tumors reached an average
volume of ≈100 mm
3
, cohorts (n=3) were treated with vehicle (5% dextrose in water,
D5W), MCo-PMI (40 mg/kg, 7.6 mmol/kg), or Nutlin-3 (10 mg/kg, 17.2 mmol/kg) (EMD
Chemicals) once daily for up to 38 days by intravenous injection (MCo-PMI and vehicle)
or by intraperitoneal administration (Nutlin-3). Compounds were prepared in a D5W at a
final volume of 50 μL (peptide) or 100 μL (Nutlin-3). Health checks were performed daily
to observe parameters such as body conditioning score, overall appearance, and
cleanliness, the strength of grip, skin color, and tone, mobility, gait and activity level as
indicators of potential drug-related toxicities. Individual weights were recorded thrice
weekly, comparing the control and treatment groups as an additional indicator of
tolerance of drug treatment as well as providing the average weight for calculation of drug
dosing. Tumor size was measured with calipers. Tumor volume was calculated by
measuring tumor size in two dimensions and applying those measurements to the
calculation V = d
2
x D/2, where d and D equal to the smaller and larger of the two
measurements, respectively. Mice bearing tumors larger than 2.4 cm3 were removed
from the study, sacrificed and necropsies performed to gather tumor and organ samples
for histological analysis. Tumor and tissue samples were perfused with PBS, after which
portions were either fixed in 10% formalin overnight then transferred to 100% EtOH or
snap-frozen in liquid nitrogen and stored at -70° C until analysis.
Quantitative RT-PCR. HCT116 p53
+/+
subcutaneous tumors were excised, flash frozen
and the RNA extracted using the RNeasy Mini kit (QIAGEN). Total RNA was reverse
86

transcribed to cDNA using M-MLV reverse transcriptase (Promega). The generated
cDNA was amplified with power SYBR Green PCR master mix (Applied Biosystems) on a
96 well plate and measured the relative transcript levels by qRT-PCR on an ABI 7900HT
Fast Real-Time PCR System (Applied Biosystems). Specific primers for HDM2, p21 and
the β-actin control were used. The amplification reactions were done in triplicate in
96-well optical plates. Threshold-cycle (Ct) values were automatically calculated for each
replicate and used to determine the relative expression of the gene of interest relative to
β-actin.

87

CHAPTER THREE Design of a high-throughput screening system for the
selection of bioactive cyclotide targeting the p53-HDM2 interaction
Cyclotides are an emerging family of plant-derived peptides that have a unique head to
tail circular knotted structure of three disulfide bonds as described earlier (Section 1.1.2).
Cyclotides have been used as molecular scaffolds for stabilizing biological active
polypeptides (T. L. Aboye et al., 2012; Eliasen et al., 2012; Richard H Kimura et al., 2012)
and delivery of polypeptide into cytoplasm for targeting intracellular proteins (D'Souza et
al., 2016; Ji et al., 2013). Cyclotides are also tolerant to sequence variation (Garcia &
Camarero, 2010), making them excellent bioscaffold for generating genetically encode
library.
Here, we reported the generation of a genetically-encoded MCoTI-based cyclotide library
in E. coli cells interfaced with a high-throughput split barnase screening approach. This
should enable us to perform the selection of novel cyclotide based inhibitors for
protein-protein interaction in living cells efficiently, making this approach extremely
valuable in the design of novel peptide-based therapeutic leads.
3.1 Expression of cyclotides expression in E. coli cells using a protein
trans-splicing approach
Cyclotides based peptides can be made by both chemically or recombinantly in vitro.
However, for in cell expression of cyclotides, the use of intein-mediated protein
trans-splicing has been recently reported to produce higher expression yields (Harris et
al., 2016; Krishnappa Jagadish et al., 2013). MCo-PMI was expressed using
intein-mediated trans-splicing approach. The DNA oligos encoding MCo-PMI sequence
were cloned into a protein trans-splicing plasmid and expressed in E. coli cells. Spliced
intein was characterized by SDS-PAGE and quantified by UV spectroscopy (Fig. 3.1).
88

The soluble cell lysate fraction pulled down by Ni
2+
-NTA-sepharose beads shows that
more than 90% of the precursor protein was spliced in the cell during expression,
indicating the high efficiency of intein splicing. The total expression of the precursor
cyclotide-intein was estimated to be around 90 mg per L of cell culture, as quantified by
UV spectroscopy.

Figure 3.1. SDS-PAGE analysis of expression of cyclotide MCo-PMI using backbone cyclization
mediated by protein trans-splicing in E. coli cells.

The folded MCo-PMI level was quantified using LC/MS-MS (Fig. 3.2a). Briefly, a known
amount of pure MCo-PMI was injected as standard. Cell lysates spiked with purified
MCo-PMI (7.5 µg or 15 µg) were used as a control to quantify the loss during the
solid-phase extraction process. The area under the curve (AUC) of the corresponding
peaks was measured and used for quantification of MCo-PMI expression level (Fig. 3.2b).
The total amount of fully folded cyclotide MCo-PMI was estimated to be ≈124.2 µg per
liter of cell culture. The intracellular MCo-PMI concentration was calculated using the
average E. coli volume of 0.7 µm
3
, where 1 optical density unity at 600 nm (OD 600) is
equal to ≈10
9
cells. The calculated intracellular concentration of MCo-PMI is ≈20 µM.

89

Figure 3.2. Quantification of in cell concentration of MCo-PMI using LC-MS/MS. (a) LC-MS/MS
traces of standard samples with known concentration, spiked controls, and unknown samples. (b)
The Linear correlation between the MCo-PMI peptide amount and areas of the peaks obtained by
LC-MS/MS.
3.2 Design of a barnase-based screening system
The development of high-throughput screening for inhibitors of specific
90

protein-protein interactions (PPIs) is an important tool for the rapid discovery of novel
therapeutic molecules. Numerous research studies have developed the concept of
protein fragment complementation assays (PCAs) or split-protein reporters (Piehler, 2005;
Remy & Michnick, 2007; Shekhawat & Ghosh, 2011), which are useful for in vivo
characterization of PPIs in real time, intracellular and membrane protein localization, and
functions of cellular proteins and protein networks. The detection of PPIs relies on the
complementation of the split-protein, in which each protein fragment is genetically fused
to a protein of interest (Fig. 3.3a). This technique has been used in various organisms,
including E. coli, yeast, mammalian cells, and animal models, and is useful for studying
cellular networks such as signal transduction pathways, and for optimizing
pharmaceuticals that can interfere with protein-protein interactions.
A split-toxin reporter can efficiently and rapidly screen for inhibitors of PPIs based
on cell death or survival. Only a specific inhibitor of the PPI will allow for cell survival while
without an inhibitor, the split-toxin will complement and result in cell death (Fig. 3.3a).
Here, we have designed a cell-based reporter assay using the cytotoxic ribonuclease
barnase. Barnase is a small protein (110 amino acids) produced and secreted by the
bacteria Bacillus amyloliquefaciens. B. amyloliquefaciens. Barnase has no disulfide
bonds and does not require post-translational modification, divalent cations, or other
non-peptide components for its function, making it an ideal protein for recombinant
expression in E. coli. The dissection of barnase has previously been used to study protein
stability and folding processes (Neira, Vázquez, & Fersht, 2000) and for engineering traits
in wheat (Kempe, Rubtsova, & Gils, 2009). These studies have shown that split-barnase
can complement or reconstitute to form a native-like structure (Fig. 3.3b).
The complementation of split-barnase in vitro has previously been reported,
demonstrating the formation of native-like complexes (Burgess et al., 2002; Neira et al.,
2000; Sancho & Fersht, 1992). The cleavage site between Lys66 and Ser67 has been
shown to refold to form the most stable complex among several pairs tested (Butler,
91

Mitrea, Mitrousis, Cingolani, & Loh, 2009). Using this cleavage site, we designed
constructs to express each barnase fragment genetically linked to either HDM2 or p53
using a (GGS)3 linker. N-barnase D8R (1-66) is linked to HDM2 at its C-terminus
(NBnHDM2), and C-barnase (67-110) is linked to p53 at its N-terminus (p53CBn),
respectively. The D8R mutation was introduced to reduce background activity of split
barnase. When mixed together, the interaction between HDM2 and p53 will bring two
fused proteins close in distance. The N-barnase and C-barnase fragments reform active
barnase (Fig. 3.3b). When there is inhibitor present, the interaction between p53 and
HDM2 was interfered, resulting in dissociation of barnase complex and loss of barnase
activity. Therefore, inhibitors are able to be selected by detecting barnase activity.

Figure 3.3. Design of a split-barnase cell-based reporter for high throughput screening of
p53/HDM2 inhibitors. (a) A split-toxin reporter can efficiently and rapidly screen for inhibitors of
PPIs based on cell death or survival. Only a specific inhibitor of the PPI will allow for cell survival
while without an inhibitor, the split-toxin will complement and result in cell death. (b) The principle
of split barnase based reporter is illustrated. The barnase split between Lys66 and Ser67 residues
92

could reform active barnase when they are close in distance. When NBn and CBn were fused to
HDM2 and p53 respectively, the interaction between HDM2 and p53 cause the formation of active
barnase. Presentation of inhibitor will dissociate two fragment results in loss of barnase activity.
3.3 In vitro characterization of the barnase screening system
The split-barnase system was examined first in vitro by monitoring its RNase activity.
Fragment proteins NBnHDM2 and p53CBn were expressed in BL21(DE3) cells and
purified using Ni
2+
-NTA-sepharose beads. The p53CBn protein expressed in insoluble
form and was purified using Ni
2+
-affinity chromatography under denaturing conditions
using 6 M GdmCl. Purified proteins were characterized using SDS-PAGE and ES-MS
(Fig. 3.4). Expected mass was found for both proteins.

Figure 3.4. Analysis of purified NBnHDM2 and p53CBn by SDS-PAGE (a) and ES-MS (b). P:
Insoluble cell lysate fraction; S: soluble cell lysate fraction; B: fraction captured on Ni2+-NTA
sepharose beads; FT: flow through of affinity capture; E: protein eluted.
93

To test whether the split-barnase fragments were able to reform the active barnase, a
fluorophore-quencher based RNA substrate was used to monitor RNase activity (Fig.
3.5a). Hydrolysis of this RNA substrate, results in the cleavage of the fluorophore from its
quencher, resulting in an increase of the fluorescence signal. The barnase fragments
NBnHDM2 (20 nM) and p53CBn (20 nM) were mixed and the RNase activity of protein
mixture was measured using a FRET-quenched RNA-based substrate (Fig. 3.5b). The
protein mixture showed RNase activity as the increasing of fluorescent was detected. All
the control samples, containing only reaction buffer or one fragment of barnase, did not
show any RNase activity. This result proves the principle of split-barnase reporter design,
that two fragments without RNase activity could form an active form of barnase.
To test whether the reformation of active barnase is caused by the interaction between
HDM2 and p53 fragments, the inhibitors nutlin-3 and linear PMI peptide were
incorporated into the reaction mixture. A decrease in split barnase activity was observed
as the concentration of the inhibitor increased (Fig. 3.5c) indicating that the inhibitor for
the HDM2-p53 interaction could dissociate NBnHDM2 and p53CBn complex eliminating
the barnase activity. The enzymatic reaction rate constant K was plotted against the
logarithm of the corresponding inhibitor to obtain the different IC50 values, 181.4 ± 22.7
nM (nutlin-3) and 63.4 ± 12.0 nM (PMI peptide). These values are similar to those
previously published (Dahl, Chuchalin, Gor, Yoxall, & Sharma, 2006; Phan et al., 2010;
Vassilev et al., 2004), demonstrating the split-barnase proteins worked in vitro.
94

Figure 3.5. In vitro activities of split-barnase linked to HDM2 and p53. (a) The principle of RNase
substrate used in this assay. (b) NBnHDM2 and p53CBn were incubated with substrate together or
separately. The fluorescence signal was measured over time 0-30 min. (c) NBnHDM2 and
p53CBn complex was incubated with increasing concentrations of an HDM2 inhibitor, nutlin-3. The
residual RNase activity was measured by adding RNase substrate. (d) The kinetic rates are
plotted against log concentration of inhibitor. Error bar shows standard deviation (n=3).
3.4 In cell characterization of the barnase screening system
The split-barnase system was then tested in E. coli cells. A polycistronic dual
expression vector was constructed by cloning NBnHDM2 and p53CBn in the pBAD
plasmid, which contains a tightly regulated arabinose-inducible promoter. The dual
expression plasmid was made by the incorporation of two ribosomal binding sites after
the arabinose-inducible promoter bi-cistronic. Peptide inhibitor PMI was fused to
thioredoxin (Txn) for higher expression level in E. coli. The inhibitor Txn-PMI and
95

MCo-PMI were cloned into the pASK-IBA35plus vector, which has a tetracycline promoter
for controllable protein expression (Fig. 3.6).

Figure 3.6. Scheme showing the two expression plasmids used to during the in-cell split-barnase
cell-based reporter characterization. The split barnase fragments were cloned into pBAD33 and
inhibitors (Txn-PMI or MCo-PMI) were cloned into pASK plasmid. The expression plasmids are
completely compatible and independent of each other as they contain a different origin of
replication, antibiotic resistant gene, and different inducible promoter.
The plasmid encoding NBnHDM2 and p53CBn was co-transformed with empty
pASK-IBA35plus into Origami2(DE3) cells and plated onto LB-agar containing either no
inducer, 1% arabinose, 2 ng/ml anhydrotetracycline, or both. In the absence of arabinose,
numerous colonies were able to grow. On the other hand, when the same experiment
was performed under the same conditions but in the presence of arabinose, no colonies
were observed indicating that the split-barnase was able to associate in trans providing
96

cellular toxicity (Fig. 3, 1
st
column). Co-transformation of the split-barnase plasmid with
Txn-PMI in the pASK-IBA35plus vector rescued the formation of colonies (Fig. 3.6, 2
nd

column), while a Txn-PMI construct (Txn-PMI WA) with a PMI mutation (W46A) known to
inhibit its binding to HDM2, did not produce any colony (Fig. 3.6, 3
rd
column) suggesting
the PMI Txn-fusion protein inhibited the interaction between HDM2 and p53, which also
prevented the complementation of the split-barnase fragments. Co-expression of the
p53-HDM2-based split-barnase reporter with MCo-PMI (an engineered cyclotide with the
PMI peptide grafted into loop 6, see section 3.1), also resulted in the recovery of colony
formation, while cells expressing an inactive mutated version (W46A mutation) of
MCo-PMI (MCo-PMI WA) did not form colonies.

Figure 3.7. The in-cell activity of the p53-HDM2-based split-barnase reporter. A plasmid encoding
the split-barnase fragments NBnHDM2 and p53CBn were co-transformed with a plasmid encoding
a PMI-based inhibitor. The same amount of cells were plated on all the plates shown in the figure.
Plates were stained with Coomassie blue to improve the visibility of bacterial colonies
97

It is worth pointing out that co-expression of the HDM2-p53 split-barnase reporter
with either TxnPMI or MCo-PMI in the absence of anhydrous tetracycline (aTet), which
induces their expression, also resulted in colony formation on those plates (Fig. 3.6, 3
rd

row). This result could be explained due to the leaky expression of pASK that is under the
control of Tet promoter, which still produces some inhibitor (MCo-PMI or Txn-PMI) even in
the absence of the corresponding inducer that is enough to inhibit barnase–induced
toxicity and recover colony formation.
The expression level of split barnase protein was quantified using Western blotting.
The colonies containing inhibitor plasmid TxnPMI or MCo-PMI were collected from both
barnase induced and suppressed plates. The number of bacteria cells was quantified by
measuring optical density at 600 nm to ensure the same amount of sample was loaded in
every lane for Western blotting (Fig. 3.8). Expression of both NBnHDM2 and p53CBn
were detected when induced with 1% arabinose. A significant less amount of split
barnase was detected from expression-suppressed colonies.
The expression level of NBnHDM2 was estimated to be ≈766 nM and ≈520 nM under
induced (with 1% arabinose) and suppressed (with 0.5% glucose) conditions, respectively.
The expression level of p53CBn was estimated to be ≈179nM and ≈57nM under induced
and suppressed conditions, respectively.

Figure 3.8. Western blotting to detect split barnase protein expression. The colonies containing
inhibitor plasmid Txn-PMI or MCo-PMI was collected from both 1% arabinose (induced) and 0.5%
glucose (suppressed) plates. Protein expression level was quantified using Western blotting with
anti-FLAG and anti-p53 antibodies. Purified proteins were used to estimate the cellular
concentration.
98

To further test the dynamic range of the split-barnase reporter in live cells, we made
several MCo-PMI mutants with different degrees of antagonistic activity (C. Li et al.,
2010). An inactive MCo-PMI-W46A and different cyclotides with different levels of affinity
to HDM2, including less active MCo-PMI-Y45A (Kd ≈ 610 nM) and more active
MCo-PMI-N47A (Kd ≈ 0.5 nM) were encoded in pASK plasmids prepared by site-directed
mutagenesis. All the MCo-PM-based cyclotides were co-expressed with the split-barnase
p53-Hmd2 reporter under different induction conditions from 0% to 1% arabinose (Fig.
3.9). Cyclotides MCo-PMI and mutant MCo-PMI-N47A were able to recover
barnase-induced toxicity even under high induction conditions (1.0% arabinose) (Fig. 3.9).
As expected, inactive mutant MCo-PMI-W46A was only able to form colonies in the
absence of arabinose. The less active mutant MCo-PMI-Y45A was able to recovery split
barnase toxicity only when inducing conditions involved the use of arabinose
concentrations ≤0.05%. When 0.1% arabinose was used for induction of the split barnase
reporter, MCo-PMI-Y45A was not able to recover the toxicity, resulting in no colony
formation. These results indicate that by using a different amount of arabinose the
amount of expressed split barnase HDM2-p53-reporter can be modulated to screen and
select inhibitors with different activities.
99

Figure 3.9. Plating of E. coli cells co-expressing the split barnase reporter and different
MCo-PMI cyclotides. The mutations were selected based on the alanine scanning of
linear PMI peptide, which provides PMI mutants with Kd from 0.5 nM to 610 nM (C. Li et
al., 2010). Different concentrations of arabinose (0%~1%) were used to induce the
expression of split barnase reporter. Mutants with different binding affinity showed the
correlated ability to recovery toxicity from split barnase.
3.5 Testing the barnase screening system with a mock plasmid library
To test whether the split barnase reporter can be used for high throughput screening in
live cells, a mock library was made by mixing plasmid encoding MCo-PMI (active) and
MCo-PMI-W46A (inactive) in a ratio of 1 to 10
5
(Fig. 3.10). The mock library was used to
mimic a real library with a hit rate in the order of 10 per million. The mock library was
transformed into competent E. coli cells previously transformed with split barnase plasmid,
and different amounts of transformed cells were plated on plates containing 1%
arabinose (induced) or 0.5% glucose (suppressed). Theoretically, only cells with a
plasmid encoding the active MCo-PMI cyclotide could form colonies on arabinose treated
100

plates.

Figure 3.10. Scheme depicting the screening/selection process using a mock library containing
plasmids encoding MCo-PMI (green) and MCo-PMI-W46A (red) in a ratio of 1/10
5
. The mock
library was transformed into E. coli cells with split barnase plasmid. Transformed cells were plated
on LB agar containing 1% arabinose to allow induction of the barnase reporter. Only cells
containing the plasmid expressing MCo-PMI are able to form colonies on the plate (green dot).
When barnase expression was suppressed by growing the cells in the presence of 0.5%
glucose, cells bearing plasmids encoding pASK, MCo-PMI, MCo-PMI-W46A, and mock
plasmid library all formed colonies (Fig. 3.11). However, when barnase was induced by 1%
arabinose, only cells containing the plasmid encoding MCo-PMI were able to form
colonies. Colony number on plates was counted and is listed in Table 3.2. The number of
colonies on the glucose plate was used as a reference for quantifying enrichment.
For empty vector pASK, no colony was detected on 1% arabinose plates even when
1000 times the number of cells (≈ 2 x 10
5
cells) plated on the control plate containing
glucose was used (plate labeled as X1000, Fig. 3.11). For cells containing the vector
encoding cyclotide MCo-PMI, colonies were observed on plates containing 1% arabinose.
For the cells expressing the cyclotide MCo-PMI-W46A, no colony was formed on
101

arabinose plates containing up to 100 times the number of cells plated on the control
plates containing glucose. However, when the number of plated cells was incremented
1000 times (≈ 1.2 x 10
6
cells), 3 colonies were observed. This result indicates the
background level of the cell-based screening system, which is around 1 in 4 x 10
5
cells.
This means that when ≈4 x 10
5
cells are plated on an arabinose plate, a colony will
appear due likely to an inactivating mutation in the barnase reporter. For cells containing
the mock library, no colony was formed on arabinose plates when plating up to 100 times
the number of cells plated in the glucose control plate (≈ 1 x 10
5
cells). When more cells
were plated, colonies started to become visible on the arabinose plate (Fig. 3.11). When
≈1,000 times the number of cells plated on the glucose plate were analyzed (≈1.4 x 10
6

cells) 29 colonies were observed. Twenty colonies were selected and their corresponding
plasmids isolated was and sequenced. DNA sequencing results showed that 13 out of 20
sequences (65%) encoded the MCo-PMI sequence, while the rest (35%) encoded the
inactive cyclotide MCo-PMI-W46A, which was considered normal given the expected
background of the barnase-based reporter. For the screening of ≈1.4 x 10
6
cells, the level
of expected false positives should be around 4, which is very close the observed value.
Table 3.2. Results obtained for the screening of the mock library containing plasmids
encoding MCo-PMI and MCo-PMI-W46A using the p53-HDM2 split-barnase reporter.
Plate Volume
pASK MCo-PMI
MCo-PMI-W4
6A
Mock Lib
Plated
cells
Colony
Plated
cells
Colony
Plated
cells
Colony
Plated
cells
Colony
0.5%
Glucose
1X 173 173 1,238 1,238 1,190 1,190 1,443 1,443
1%
Arabinose
1X 173 0 1,238 Full 1,190 0 1,443 0
10X
1.73x
10
3

0
1.24x
10
4

Full
1.19x
10
4

0
1.44x
10
4

0
100X
1.73x
10
4

0
1.24x
10
5

Full
1.19x
10
5

0
1.44x
10
5

4
1,000X
1.73x
10
5

0
1.24x
10
6

Full
1.19x
10
6

3
1.44x
10
6

29
102

Figure 3.11. Results obtained for the screening of the mock library containing plasmids encoding
MCo-PMI and MCo-PMI-W46A using the p53-HDM2 split-barnase reporter. When barnase
expression was suppressed by growing the cells in the presence of 0.5% glucose, cells bearing
plasmids encoding pASK, MCo-PMI, MCo-PMI-W46A, and mock plasmid library all formed
colonies. However, when barnase was induced by 1% arabinose, only cells containing the plasmid
encoding MCo-PMI were able to form colonies.
These findings show that the split-barnase reporter can be used to rapidly screen for
inhibitors of protein-protein interactions from large libraries in E. coli containing hit ratios
of 1 to 10
6
. This level of hit ratios should provide an acceptable level of false positives,
that could be excluded by further in vitro screening.
3.6 Design of a combinatorial MCoTI-I based library
Recombinant biosynthesis of cyclotides allows us to generate large combinatorial
libraries of based on the cyclotide scaffold. By using the protein trans-splicing method,
103

intracellular concentrations in the µM range of natively folded cyclotide can be
accomplished in E. coli cells. Recombinant libraries using E. coli cells can easily reach
the complexities of ≈10
9
. Unlike display technology, the cyclotide library will not be fused
or displayed by any protein but will stay inside the live cell, which allows linking the
cellular phenotype (i.e. death or survival) to the DNA code of the corresponding member
of the library.
The first genetically encoded cyclotide MCoTI-I-based library was reported by the
Camarero group (Austin et al., 2009). This library contained a complete suite of amino
acid mutants and was recombinantly expressed in E. coli cells.
Here we designed a large genetically encoded MCoTI-I based library with the theoretical
complexity of ≈10
11
(Fig. 3.12). The library was designed by randomizing loops 1 and 5.
These two loops are on the same side of MCoTI-I molecule, which should allow the
generation of a larger complementary binding surface (Fig. 3.12). A total of nine residues,
among which five residues are in loop 1 and four residues are in loop 5, were randomized.
Reside Pro3 located in loop 1, which is critical for cyclotide folding, was not mutated. In
addition, residue Tyr26 in loop 5, which is at the extein-IN junction and has been shown
previously (Ji et al., 2013) to provide very good cyclization yield was also not changed
(Fig. 3.12). The DNA encoding this library was synthesized using a codon scheme
optimized for E. coli expression which encodes 19 amino acid codons on all nine mutated
positions, in which cysteine codon was eliminated to avoid interference with folding. The
library was cloned into pASK-TS plasmid and 60 clones were selected for
characterization. The diversity of the library was calculated using the Weblogo software
(Fig. 3.12). The theoretical size of this library is 19
9
, which is equal to 3.2 X 10
11
. The size
of the cell library was calculated based on the colony-forming unit (CFU) which reached
10
9
.
104

Figure 3.12. Design of an MCoTI-I based genetically-encoded library. Loops 1 and 5 were used to
generate the genetically encoded library. A total of nine residues were randomized (five from loop
1 and four from loop 5). The DNA sequences obtained from 60 clones were used to generate the
WebLogo representation of the library (Crooks, Hon, Chandonia, & Brenner, 2004)
3.7 Screening of the MCoTI-I library using the barnase cell-based reporter
The MCoTI-based library was screened using the split-barnase in-cell described earlier
(Section 3.2). As described above, the split barnase system has a background ratio of 1
to 4 x 10
5
. To overcome this background issue, an enrichment protocol similar to those
used in display technologies was used (Fig. 3.13). The plasmid library was first
transformed into E. coli cells containing the split-barnase expression vector (Fig. 3.6)
using electroporation. Cells were plated on LB agar plates containing 0.5% glucose for
counting the transformation efficiency and around 5 x 10
8
cells were plated on plates
containing 1% arabinose for selecting active sequences (Fig. 3.13, Step 1). Colonies
selected on 1% arabinose plates were picked and grown to extract the library plasmid
(Fig. 3.13, Steps 2 and 3). Because the background colonies, the first-time screening will
get false sequences mixed with real active sequences. To eliminate the false sequences,
the plasmids obtained from the first screen (≈300 different sequences) were transformed
105

a second time on fresh competent cells bearing the split-barnase plasmid. These steps
were repeated a second time to enrich active sequences. Enriched DNA plasmids were
extracted individually and send for DNA sequencing (Fig. 3.13, Step 4).

Figure 3.13. Experimental flow for the screening/selection process of the MCoTI-based library
using a split barnase cell-based reporter. Library screening was repeated twice to enrich active
sequences. The plasmid library was first transformed into E. coli cells containing the split-barnase
expression vector (step 1). Colonies selected on 1% arabinose plates were picked (step 2) and
grown to extract the library plasmid (step 3). To eliminate the false sequences, the plasmids
obtained from the first screen were transformed a second time on fresh competent cells bearing
the split-barnase plasmid. These steps were repeated a second time to enrich active sequences.
Enriched DNA plasmids were extracted individually and send for DNA sequencing (step 4).
The enrichment process was quantified by calculating the ratio of colonies in plates
containing 1% arabinose versus 0.5% glucose plate (Fig. 3.14). For the plasmid encoding
the cyclotide MCo-PMI, which was used as a positive control, the observed ratio was as
expected around, while for the plasmid encoding MCoTI-I the ration was ≈ 1 to 4 x 10
5

which is the background signal for the split barnase cell reporter in E. coli cells. The
observed ratio for the library during the first screening enrichment step was very close to
106

that the one observed with the MCoTI-I plasmid, which indicates that most sequences in
the MCoTI-based library was not active. After the first enrichment step, the ratio for the
second round increased to 1 to 20, which means one sequence among 20 sequences
seemed to be active. After the second enrichment, the ratio for third one almost reached a
value of 1. At this point, 60 colonies were selected and sent for DNA sequencing.

Figure 3.14. Library screening results using the split barnase cell-based reporter system. The cfu
ratio between the number of colonies in plates with 1% arabinose and 0.5% glucose indicates the
level enrichment of the library. At the beginning, the arabinose to glucose cfu ratio for the library
was close to that of the negative control MCoTI-I plasmid. After two rounds of enrichment, the
enriched library gave a similar arabinose to glucose cfu ratio when compared to that of the positive
control MCo-PMI plasmid. The numbers indicated on the different panels are the cfu on plates.

107

3.8 Characterization of the sequences selected in the 3
rd
round of enrichment
Table 3.3. Summary characterizing the sequence, production, folding ability and
HDM2/p53 antagonistic activity of the different clones selected in the 3
rd
round of
enrichment.
1
Cleavage of cyclotide from intein precursor could be induced by thiols, like reduced glutathione
(GSH) or Mesna.
2
Cyclotide folding was carried out in phosphate buffer (1 mM EDTA, 2mM GSH, 50 mM Sodium
Phosphate pH7.2, 150 mM NaCl) with or without 20% isopropanol (ISP).

The 6 sequences selected by the split barnase screening were further characterized. The
sequences were cloned into an expression vector using an intein-mediated cyclization
approach (Fig. 1.3). Among the different selected sequences, only two of them could be
cyclized and folded in vitro correctly (L1Seq3 and L3Seq1, Table 3.3). The other four
sequences did not fold well in buffer with or without isopropanol, which is commonly used
for improving cyclotide folding. The two folded sequences were further purified and test
using an in vitro FRET-based inhibition assay as described above in section 2.7 (Fig.
2.18). None of them showed inhibition activity concluding that more sequences would
need to be selected from the library.
Loop1 Loop 5 Repeats Intein cleavage
1
Folding
2
Activity
PKILQR RGNGY GSH Mesna PBS 20%
ISP
Inhibition
L1Seq 1 PSEATS PKPPY 19 OK Good - - -
L1Seq 2 PGQILP SVSNY 5 Bad Good + - -
L1Seq 3 PSANQT EIFPY 3 OK Good + +++ Not Active
L1Seq 4 PWTPRW RRWFY 1 Bad Good - - -
L3Seq 1 PFPQYR YWSWY 11 Good + Not Active
L4Seq 1 PTEGSY DWWPY 2 Good - -
108

3.9 Design of a MCo-PMI based library
MCoTI-based cyclotides have six loops and most of those residues are tolerant to
mutations. As described in Chapter 2, loop 6 of MCoTI-I was used to graft a bioactive PMI
peptide resulting in cyclotide MCo-PMI that possessed good activity against p53 active
tumor cells. To further improve the activity of MCo-PMI, split the barnase based molecular
evolution strategy was used. Molecular evolution is widely used for improving protein
binding affinity toward targets. By introducing random mutations into key position and
using high throughput screening system, the more active compound could be selected.
Figure 3.15. Design of a MCoTI-I based library using the structure of the MCo-PMI/HDM2
complex (Ji et al., 2013). Six residues, which includes residues 21-23, 36 and 40-41 were chosen
to generate a genetically encoded library. Those residues are close in distance to HDM2 surface
based on NMR complex structure. The NNK (N = A, T, G, and C and K = T and G) codon was used
for randomization at DNA level on these 6 positions.
MCo-PMI based library was designed based on the NMR structure of the complex (Fig.
2.7). The MCo-PMI residues that are in close distance to the molecular surface HDM2
were selected for molecular evolution. These included residues 21 to 23 on loop 6,
109

residue 36 on loop 2, and residues 40 to 41 on loop 3 (Fig. 3.15). The NNK (N = A, T, G,
and C and K = T and G) codon was used for randomization at DNA level to minimize stop
codons. Because grafted PMI peptide was optimized by phage display, none of PMI
residues was selected was changed. By introducing mutations into these six positions, an
MCo-PMI-based library was generated with the theoretical complexity of 6.4 x 10
7
. This
library was cloned into the trans-splicing expression vector and screened using the
split-barnase cell-based reporter system.
3.10 Screening of the MCo-PMI based library
The MCo-PMI library was screened using the split-barnase system. For the selection of
the most active compounds, different concentration of arabinose (from 0.5% to 10%)
were used. Increasing the concentration of the inducer arabinose should also increase
the expression level of the split barnase reporter, resulting in the selection of the colonies
expressing the most active compounds (Fig. 3.16). Cells transformed with the MCo-PMI
plasmid produce significantly fewer colonies on plates containing 5% arabinose, and no
colonies were observed when plated on 10% arabinose. Cells transformed with a library
plasmid encoding MCo-PMI mutants produced fewer colonies on 2% and 5% arabinose
plate (Fig. 3.16). However, there were some bigger colonies than the observed when
transforming with MCo-PMI plasmid, which indicates that under the same condition for
barnase expression, cells containing these library sequences grow faster and healthier.
This could be a sign that those sequences may have a more active cyclotide when
compared to the original MCo-PMI sequence. These colonies were selected for further
analysis.
110

Figure 3.16. In-cell screening and selection of MCo-PMI plasmid library in E. coli cells using the
split-barnase complementary assay. Around 2 x 10
3
cells were plated in 0.5% glucose or different
arabinose concentrations (0.5% to 10%). Plasmids encoding MCo-PMI and MCoTI-I were used as
positive and negative controls.
3.11 Materials and methods
3.11.1 Materials and instrumentation.
The materials and instrumentation are same as Section 2.9.1 except the DNA
sequencing service was provided by Retrogene Inc.

3.11.2 Experimental methods
Cloning of MCo-PMI trans-splicing expression vector. A plasmid expressing the
MCo-PMI cyclotide trans-splicing precursor protein was constructed using the
pASK-TS-MCoTI vector as a template (Harris et al., 2016). Briefly, DNA oligos encoding
MCo-PMI were amplified by PCR from template DNA pTXB1-MCo-PMI using primers
5’-TTC CAT AGC TTC GAA CTG CGG TTC TGG TTC TGG TG and 5’-TCT ACT GTC
111

AAG ATC TCC GTT TCA TAT GAT AAA CAG AAA CCA CGG GTG CAG ATG C. PCR
product and vector pASK-TS-MCoTI were sequentially digested with BglII and BstBI. The
linearized vector and the MCo-PMI encoding dsDNA fragments were ligated at 16 ° C
overnight using T4 DNA ligase (New England Lab). The ligated plasmid was transformed
into DH5α cells (Invitrogen) and plated on Luria Broth (LB) agar containing 100 μg/mL
ampicillin. Positive colonies were grown in 5mL LB containing ampicillin at 37 ° C
overnight and the corresponding plasmids were purified using a Miniprep kit (Qiagen).
The DNA sequence was confirmed by sequencing.

Expression of MCo-PMI using trans-splicing method. Origami2 (DE3) cells (Novagen)
were transformed with plasmid pASK-TS-MCo-PMI. Expression was carried out in 2XYT
medium (1L) containing 100 μg/mL ampicillin at room temperature. Briefly, 5 mL of an
overnight starter culture derived from a single clone were used to inoculate 1L of 2XYT
media. Cells were grown to an OD600 nm of ~ 0.6 at 37 ° C. Protein expression was
induced by addition of anhydrotetracycline (ATC) to a final concentration of 200 μg/L at
room temperature for overnight. The cells were harvested by centrifugation. For spliced
intein purification, the cells were resuspended in 40 mL of lysis buffer (1 mM PMSF, 10
mM imidazole, 50 mM sodium phosphate, 250 mM NaCl buffer at pH 8.0 containing 5%
glycerol) and then lysed by sonication. The lysate was clarified by centrifugation at
15,000 rpm in a Sorval SS-34 rotor for 30 min. The clarified supernatant was incubated
with Ni-NTA beads (4 mL/L cell culture, Thermo Scientific), previous equilibrated with Ni
2+

wash buffer (20 mM imidazole, 50 mM sodium phosphate, 250 mM NaCl buffer at pH 8.0)
at 4 ° C for 1 h with gentle rocking. The beads were extensively washed with 50
bed-volumes of Ni wash buffer and then eluded with Ni elution buffer (20mM imidazole,
50 mM sodium phosphate, 250 mM NaCl buffer at pH 8.0). The spliced intein was
quantified spectrophotometrically using an extinction coefficient at 280 nm of 18,545
M
-1
cm
-1
. The expression level for spliced intein was ≈ 84 mg/L.
112

Quantification of MCo-PMI in cell expression level. Origami (DE3) cells (Novagen)
were transformed with the pASK-TS-MCo-PMI plasmid. Protein was expressed as
previously described. The clarified supernatant was acidified with 5% acetic acid, and
then clarified again by centrifugation at 15,000 rpm in a Sorval SS-34 rotor for 30 min.
Total 5 mL of 40 mL supernatant was desalted and concentrated by solid-phase
extraction using a C18 sep-pak (Waters) and then lyophilized. An Origami cell lysate with
empty pASK plasmid was spiked with known amount pure MCo-PMI to quantify the loss
during the solid-phase extraction step. The spiked cell lysate was processed the same
way as described above. The lyophilized samples were reconstituted with 0.1% TFA in
H2O. Insoluble particles were removed by centrifugation. The soluble fraction was
analyzed by LC/MS/MS using the +4 state (1316.1) of MCo-PMI. The standard curve for
quantification was obtained by injecting a different known amount of pure MCo-PMI into
LC-MS/MS and quantifying the area of peaks. The loss of peptide during purification
process was ≈55% and the in-cell concentration of MCo-PMI was ≈18.7μM.

Cloning of split-barnase constructs. Plasmid pRSF-NBnHDM2 was constructed by
ligating fragments encoding N-barnase, a (GGS)3 linker, and p53 binding domain of
HDM2 (residues 17-125) into pRSF-duet vector MCS1. Briefly, DNA oligos encoding the
N-barnase and (GGS)3 linker (Table 3.4) were synthesized (Integrated DNA
Technologies), annealed and ligated into pRSF-duet1 MCS1 using BamHI/EcoRI and
EcoRI/PstI sites respectively. The DNA oligos encoding the p53 binding domain of HDM2
were amplified by PCR using template pRSFduet-CyPetHDM2 and the set of primers
HDM2 forward and reverse (Table 3.4), and then ligated into PstI/HindIII site to obtain the
plasmid encoding NBn-GGS3-HDM2 protein. The final plasmid, pRSF-NBnHDM2, was
confirmed by sequencing free of mutants.
Plasmid pRSF-p53CBn was constructed by ligating fragments encoding p53, (GGS)3
113

linker and C-barnase into the MCS2 of the pRSF-duet vector. Briefly, forward and reverse
DNA oligos encoding p53-GGS linker and C-barnase (Table 3.4) were annealed and
ligated by T4 ligase. The ligated DNA oligo encoding full length of p53-GGS3 -CBarnase
was ligated into pRSF-duet1 vector MCS2 NdeI and XhoI site. The final plasmid,
pRSF-p53CBn, was confirmed by sequencing free of mutants.

Table 3.4. Oligonucleotides used for the cloning of plasmids pRSF-NBnHDM2 and
pRSF-p53CBn which encode proteins NBnHDM2 and p53CBn
Name Sequence
NBn
Forward
GAT CCT GCA CAG GTT ATC AAC ACG TTT CGC GGG GTT GCG
GAT TAT CTT CAG ACC TAT CAT AAG CTT CCT GAT AAT TAC ATT
ACA AAG TCA GAA GCG CAG GCC CTC GGC TGG GTT GCA TCA
AAG GGG AAT CTT GCA GAC GTC GCT CCG GGG AAA AGC ATC
GGC GGA GAC ATC TTC TCA AAC AGG GAA GGC AAA CTC CCT
GCC AAA GCG
NBn
Reverse
AAT TCG CTT TGG CAG GGA GTT TGC CTT CCC TGT TTG AGA
AGA TGT CTC CGC CGA TGC TTT TCC CCG GAG CGA CGT CTG
CAA GAT TCC CCT TTG ATG CAA CCC AGC CGA GGG CCT GCG
CTT CTG ACT TTG TAA TGT AAT TAT CAG GAA GCT TAT GAT AGG
TCT GAA GAT AAT CCG CAA CCC CGC GAA ACG TGT TGA TAA
CCT GTG CAG
GGS
Forward
AAT TCA GGC GGA AGT GGC GGA AGT GGC GGA AGT CTG CA
GGS
Reverse
GAC TTC CGC CAC TTC CGC CAC TTC CGC CTG
HDM2
Forward
AAA AAA AAC TGC AGT CAC AGA TTC CAG CTT CGG
114

HDM2
Reverse
AAA AAA AAA AGC TTA GTT CTC ACT CAC AGA TG
P53GGS
Forward
TAT GGG CAG CAG CCA TCA CCA TCA TCA CCA CAG CAG TCA
GGA AAC ATT TTC AGA CCT ATG GAA ACT ACT TCC TGA AAA CAT
CGG CGG AAG TGG CGG AAG TGG CGG
P53GGS
Reverse
ACT TCC GCC ACT TCC GCC ACT TCC GCC GAT GTT TTC AGG
AAG TAG TTT CCA TAG GTC TGA AAA TGT TTC CTG ACT GCT GTG
GTG ATG ATG GTG ATG GCT GCT GCC CA
CBn
Forward
AAG TAG CGG ACG GAC GTG GCG TGA AGC GGA TAT TAA CTA
TAC ATC AGG CTT CAG AAA TTC AGA CCG GAT TCT TTA CTC AAG
CGA CTG GCT GAT TTA TAA GAC GAC AGA TCA TTA TAA AAC CTT
TAC AAA AAT CAG ATA AC
CBn
Reverse
ACT TCC GCC ACT TCC GCC ACT TCC GCC GAT GTT TTC AGG
AAG TAG TTT CCA TAG GTC TGA AAA TGT TTC CTG ACT GCT GTG
GTG ATG ATG GTG ATG GCT GCT GCC CA

Expression of split-barnase proteins. Plasmids pRSF-NBnHDM2 and pRSF-p53CBn
were transformed to BL21(DE3) (Invitrogen). Expression was carried out in LB medium
(1L) containing 34 μg/mL kanamycin at 30° C. Briefly, 5 mL of an overnight starter culture
derived from a single clone were used to inoculate 1L of LB media. Cells were grown to
an OD 600 nm of ~ 0.6 at 37° C. Protein expression was induced by addition of IPTG to a
final concentration of 0.3 mM 30° C for 4 h. The cells were harvested by centrifugation.
The cells were then resuspended in 40 mL of lysis buffer (1 mM PMSF, 10mM imidazole,
50 mM sodium phosphate, 250 mM NaCl buffer at pH 8.0 containing 5% glycerol) and
then lysed by sonication. The lysate was clarified by centrifugation at 15,000 rpm in a
Sorval SS-34 rotor for 30min. The clarified supernatant was incubated with Ni
2+
-NTA
sepharose beads (1 mL/L cell culture, Thermo Scientific), previous equilibrated with Ni
2+

115

wash buffer (20 mM imidazole, 50 mM sodium phosphate, 250 mM NaCl buffer at pH 8.0)
at 4° C for 1 h with gentle rocking. The beads were extensively washed with 50
column-volumes of Ni
2+
wash buffer and then eluded with Ni
2+
elusion buffer (20 mM
imidazole, 50 mM sodium phosphate, 250 mM NaCl buffer at pH 8.0). The purified
proteins were analyzed by SDS-PAGE and ES-MS and dialyzed against 50mM sodium
phosphate, 250mM NaCl, 0.1mM EDTA, pH7.2 overnight at 4° C (see Fig. 3.4)
The p53CBn protein was extracted and purified from insoluble inclusion bodies. Briefly,
the insoluble cell lysate fraction was washed once with 30 mL 0.5% Triton X100, 50 mM
sodium phosphate, 250 mM NaCl buffer at pH 8.0 followed by 3 washes of 30 mL 50 mM
sodium phosphate, 250 mM NaCl buffer at pH 8.0. The insoluble fraction was solubilized
with 20 mL of freshly prepared denaturing buffer (100mM sodium phosphate, 150mM
NaCl, and 8 M urea at pH 8.0) and clarified by centrifugation at 15,000 rpm. The clarified
supernatant was incubated with 1 mL Ni
2+
-NTA beads, previously equilibrated with
denaturing buffer at 4° C for 2 h with gentle rocking. The sepharose beads were washed
with 50 column volumes of wash buffer (100 mM sodium phosphate, 150 mM NaCl and 8
M urea at pH 6.5) then eluted with 2 mL elution buffer (100 mM sodium phosphate, 150
mM NaCl and 8M urea at pH4.5). The purified proteins were analyzed by SDS-PAGE and
ES-MS and desalted by sep-pak (Waters) and lyophilized for storage (Fig. 3.4).

Split barnase activity assay. 20 nM of protein NBnHDM2 or p53CBn or both were
prepared in barnase reaction buffer (50 mM sodium phosphate, 250 mM NaCl, 0.1 mM
EDTA, pH 8.0). Then a fluorescence-quenched RNA substrate (RNase alert, Integrated
DNA Technologies) was added to a final concentration of 33 nM. The amount of
fluorescence produced after nuclease digestion was followed over time (0-30 min) using
fluorescence plate reader (2103 EnVision Multilabel Plate Reader, Perkin Elmer) with
excitation at 485 nm and emission at 520 nm.

116

Split barnase inhibition assay. For split barnase inhibition assay, 20 nM of NBnHDM2
in barnase reaction buffer was first incubated with various concentrations of inhibitors,
Nutlin-3 or PMI peptide, for 30 min at room temperature to form equilibration. Then
p53CBn protein was added to a final concentration of 20 nM and cleavage of the RNA
substrate was measured by fluorescence as described above. The resulting plot was
fitted to a single binding site competition curve using the Prism (GraphPad) software
package.

Cloning of polycistronic co-expression plasmid for NBnHDM2 and p53CBn. The
NBnHDM2 and p53CBn genes were cloned into the pBAD33 vector, which was modified
for co-expression of two proteins by the insertion of two ribosomal binding sites. Briefly,
DNA oligos encoding NBnHDM2 with ribosome binding site sequence (RBS) was
obtained by PCR using the set of primers NBnHDM2-F and NBnHDM2-R (Table 3.5) and
pRSF-NBnHDM2 as a template. The PCR product was double digested and ligated into
SacI and KpnI of pBAD33 vector to obtain pBAD-NBnHDM2. The DNA oligos encoding
p53CBn with RBS was obtained by PCR using primer pair P53CBn-F and P53CBn-R
(Table 3.5) and pRSF-p53CBn as a template. The PCR product was double digested and
ligated into XbaI and XphI of pBAD-NBnHDM2 to obtain a polycistronic co-expression
plasmid for NBnHDM2 and p53CBn. The final plasmid was confirmed by sequencing free
of mutants.
Table 3.5. Oligonucleotides used for cloning polycistronic plasmid for proteins NBnHDM2
and p53CBn
Name Sequence
NBnHDM2-F AAA AAA AAG AGC TCA GGA GGA CAG CTA TGG C
NBnHDM2-R AAA AAA AAG GTA CCT TAG TTC TCA CTC ACA GAT G
P53CBn-F AAA AAA AAT CTA GAG TTA AGT ATA AGA GGA GG
P53CBn-R AAA AAA AAG CAT GCT TAT CTG ATT TTT GTA AAG G
117

Cloning of pASK-Txn-PMI. PMI sequence was first cloned into pET32a vector which
has thioredoxin tag in it. Briefly, DNA oligonucleotides encoding PMI sequence (Table 3.6)
(Integrated DNA Technologies), annealed and ligated into double digested pET32a
vector using the NcoI and EcoRI restriction sites. The entire Txn-PMI sequence was
subcloned into the pASK-IBA35plus vector using the XbaI and SalI restriction sites to
obtain pASK-Txn-PMI plasmid. The final plasmid was confirmed by sequencing free of
mutants.
Table 3.6. Oligonucleotides for cloning pASK-Txn-PMI which encodes protein Txn-PMI
Name Sequence
PMI forward CAT GGC TAC CAG CTT TGC GGA ATA TGC GAA CCT GCT
GAG CCC GTA AG
PMI reverse AAT TCT TAC GGG CTC AGC AGG TTC GCA TAT TCC GCA AAG
CTG GTA GC
Cloning of Txn-PMI-WA and MCo-PMI-mutants. The MCo-PMI and Txn-PMI single
mutants were generated by using the Quikchange Lighting Multi Site-directed
Mutagenesis kit (Agilent) following the manufacturer protocol. Briefly, mutagenesis PCR
reactions were done by using primers listed in Table 3.7 to introduce different mutations
into MCo-PMI (Y45A, W46A, and N47A) or Txn-PMI (WA). The PCR product was
digested by DpnI, transformed into XL10 cells and then plated on LB agar containing 100
μg/mL carbenicillin. Colonies were grown in LB overnight and the corresponding plasmid
DNA extracted. The final plasmid was confirmed by sequencing to be free of mutations.
Table 3.7. Oligonucleotides used to produce plasmids pASK-TS-MCo-PMI-Y45A,
pASK-TS-MCo-PMI-W46A, and pASK-TS-MCo-PMI-N47A.
Name Sequence
Y45A TTT CGC TGA AGC TTG GAA CCT GC
W46A CGC GGA ATA TGC GAA CCT GCT GTC
N47A GAA TAC TGG GCT CTG CTG TCT GC
118

Split barnase plating assay. Origami2 (DE3) competent cells containing plasmid
pBAD-SplitBarnase were made for electroporation. Briefly, Origami2(DE3) competent cell
(Invitrogen) was transformed with plasmid pBAD-SplitBarnase and plated on LB agar with
60 µ g/mL chloramphenicol and 0.5% glucose. After overnight incubation at 37° C, 10
colonies were picked and grown in 5 mL LB with 60 µ g/mL chloramphenicol and 0.5%
glucose overnight. 1L LB broth containing 60 µ g/mL chloramphenicol and 0.5% glucose
was inoculated with overnight starter culture and incubated at 37° C with 250 rpm shaking
until culture OD 600 reach 0.5. The cells were chilled on ice for 15 min and harvested by
centrifugation using GSA3 rotor 6,000rpm 10 min at 4° C. The cells were washed with
ice-cold ddH2O twice and re-suspended in small volume ice-cold ddH2O to a final OD 600
around 200~250.
100 ng inhibitor plasmids were electroporated into pBAD-SplitBarnase origami2 cells
using 1,800 V electroporator (Eppendorf 2510). After 1 h recovery in 1 mL SOC broth, 0.1
µ L transformed cells were plated on plates containing different inducer for protein
expression (2 ng/mL ATC for pASK expression, 1.0% arabinose for pBAD expression and
0.5% glucose for pBAD suppression). Plates were incubated at 37° C 24 hours and
stained with Coomassie blue. Plate pictures were taken using a Kodak EDAS290 camera
system.

Western Blotting for barnase expression. Colonies grown on plates were collected
and protein expression level was analyzed using western blotting. Briefly, 1 mL of
OD 600=0.5 cells were collected, pelleted down, resuspended in 100 µ L SDS loading buffer
(50 mM Tris pH 6.8 containing 0.1% bromophenol blue, 2% SDS, 5% 2-mercaptoethanol,
and 10% glycerol), and lysed by heating up at 95° C for 5 min. 10µ L samples were loaded
for western blotting together with different amounts of purified split barnase proteins.
Anti-FLAG (Sigma) and anti-p53 (Santa Cruz) antibodies were used to detect the
expression level of NBnHDM2 and p53CBn.
119

Mock Library test. The mock library plasmid was made by mixing 1 μL of 100 ng/μL
pASK-TS-MCo-PMI-W46A plasmid with 1 uL of 1 pg/uL pASK-TS-MCo-PMI plasmid. The
control plasmids pASK, pASK-TS-MCo-PMI, pASK-TS-MCo-PMI-W46A and the mock
plasmid library were used to transform electrocompetent Origami2(DE3) cells containing
the pBAD-SplitBarnase plasmid as described above. 0.1 μL of each transformed cell
were plated on LB agar plates containing 200 μg/mL carbenicillin, 60 μg/mL
chloramphenicol, and 0.5% glucose to check transformation efficiency. Different amount
(0.1μL~100μL) of cells were plated on LB agar plates containing 200 μg/mL carbenicillin,
60 μg/mL chloramphenicol, and 1% arabinose for the selection of positive clones. After
incubation at 37° C for 24 h, 20 colonies from the mock library plate were picked and
grown in LB media containing 200 μg/mL carbenicillin, 60 μg/mL chloramphenicol, and
0.5% glucose at 37° C overnight. Plasmids were extracted and sent for DNA sequencing.

MCoTI library construction. MCoTI-I based library template was synthesized
(ThermoFisher) and amplified using the pair of primers library-forward and library-reverse
listed in Table 3.8. To yield a large number of library sequences, 500 μg of plasmid
pASK-TS-MCoTI and 60 μg of PCR amplicon were digested sequentially with BglII
followed by BstBI. 100 μg of double digested plasmid and 15 µ g of PCR product were
used for the DNA ligation reaction (1:6 molar ratio of vector and insert fragment). Ligated
DNA was purified and used to transform electrocompetent DH5  cells. The library size
was estimated by plating a small portion of cells and counting the number of colonies. A
total of 60 clones were used to estimate the complexity using the Weblogo tool (Fig. 3.12)
(Crooks et al., 2004).
Table 3.8. Oligonucleotides used to produce MCoTI-I library plasmid
Name Sequence
MCoTI-I
library
TT GTA AAA CGA CGG CCA GTG AGC GCG ACG TAA TAC GAC
TCA CTA TAG GGC GAA TTG GCG GAA GGC CGT CAA GGC
120

template
sequence
CAC GTG TCT TGT CCA GGC GCG CCA GCT TCG AAT TGT
GGT AGC GGT AGT GAT GGT GGT GTT TGC CCA NNN NNN
NNN NNN NNN TGC CGT CGC GAT AGC GAT TGT CCG GGT G
CA TGT ATT TGC NNN NNN NNN NNN TAC TGC CTG AGC TAT
GAA ACC GAG ATC TAT TAA TTA ATG GAG CAC AAG ACT GGC
CTC ATG GGC CTT CCG CTC ACT GCC CGC TTT CCA GTC
GGG AAA CCT GTC GTG CCA GCT GCA TTA ACA TGG TCA
TAG CTG TTT CC
Library
Forward
TAG CTT CGA ACT GCG GAT CCG GTT CTG ACG GTG GTG
TTT G
Library
Reverse
CCA TAT TCT ACT GTC AAG ATC TCC GTT TCA TAT GAT AAA
CAG TA

Screening of the MCoTI-I library. The MCoTI-I library plasmid was screened as
described above. Briefly, 100 ng of library plasmid was electroporated into 100 μL of
Origami2(DE3) cells (OD600 ≈ 200) containing pBAD-SplitBarnase plasmid. After 1 h
recovery at 37° C, 0.1μL cells were plated on LB agar containing 200 μg/mL carbenicillin,
60 μg/mL chloramphenicol, and 0.5% glucose for colony counting. The rest of the cells
were plated on LB agar containing 200 μg/mL carbenicillin, 60 μg/mL chloramphenicol,
and 1% arabinose for selecting of positive clones. Colonies grown on arabinose plates
were picked and grown in LB media with 200 μg/mL carbenicillin, 60 μg/mL
chloramphenicol, and 0.5% glucose at 37° C overnight and then the plasmid was
extracted. The split barnase plasmid copurified was removed transforming the plasmid
mixture into chemically competent DH5  cells, plated on LB agar containing only 200
μg/mL carbenicillin. The colonies on plates were collected and library plasmids were
purified. The selection process was repeated several times. The final plasmids were sent
for DNA sequencing.
121

Construction of pTXB1 expression plasmid for selected sequences. Selected
sequences were cloned into pTXB1 for expression. Briefly, the selected sequences were
amplified by PCR using primers listed in Table 3.9. The PCR products were double
digested and ligated into pTXB1 NdeI and SapI site. The final plasmids were confirmed
by DNA sequencing.
Table 3.9 Oligonucleotides for cloning pTXB1-Hits which encodes protein precursors for
selected library hit sequences
Name Sequence
Universal forward primer AAA AAA AAC ATA TGT GTG GTA GCG GTA GTG
ATG
Reverse L1Seqt1 AAA AAA AAG CTC TTC CGC AGT ACG GCG GTT
TCG G
Reverse L1Seq2 AAA AAA AAG CTC TTC CGC AGT AAT TAG AAA
CAG AGC AAA TAC ATG CAC
Reverse L1Seq3 AAA AAA AAG CTC TTC CGC AGT ACG GAA ATA TTT
CGC AAA TAC ATG C
Reverse L1Seq4 AAA AAA AAG CTC TTC CGC AGT AAA ACC AAC
GAC GGC AAA TAC
Reverse L3Seq1 AAA AAA AAG CTC TTC CGC AGT ACC AAG ACC
AGT AGC
Reverse L4Seq1 AAA AAA AAG CTC TTC CGC AGT ACG GCC ACC AG

Expression of selected sequences. BL21(DE3) (Novagen) cells were transformed with
plasmids pTXB1-Hits encoding precursor for cyclotide sequences selected. Expression
was carried out in LB medium (1 L) containing ampicillin (100 µ g/mL) at 30° C for 4 h
respectively. Briefly, 5 mL of an overnight starter culture derived from either a single clone
122

or single plate were used to inoculate 1 L of LB media. Cells were grown to an OD at
600 nm of ≈0.6 at 37° C, and expression was induced by the addition of isopropyl- β
-D-thiogalactopyranoside (IPTG) to a final concentration of 0.3 mM at 30° C for 4 h. The
cells were then harvested by centrifugation. For fusion protein purification, the cells
were resuspended in 30 mL of lysis buffer (0.1 mM EDTA, 1 mM PMSF, 50 mM sodium
phosphate, 250 mM NaCl buffer at pH 7.2 containing 5% glycerol) and lysed by
sonication. The lysate was clarified by centrifugation at 15,000 rpm in a Sorval SS-34
rotor for 30 min. The clarified supernatant was incubated with chitin-beads (2 mL beads/L
cells) (New England Biolabs), previously equilibrated with column buffer (0.1 mM EDTA,
50 mM sodium phosphate, 250 mM NaCl buffer at pH 7.2) at 4° C for 1 h with gentle
rocking. The beads were extensively washed with 50 bead-volumes of column buffer
containing 0.1% Triton X100 and then rinsed and equilibrated with 50 bead-volumes of
column buffer. Cyclotide-intein precursor proteins were cleaved with either 50 mM GSH
or 100mM MENSA in degassed column buffer. The cleavage reactions were kept for up
to 1-2 days at 25° C with gentle rocking. For GSH cleavage, the supernatant of the
cleavage reaction was separated by filtration and the beads were washed with additional
column buffer to reach a final concentration of 5 mM GSH, and the folding was allowed to
proceed with gently rocking at 4° C for 48 h. For MENSA cleavage, the cleaved and
cyclized cyclotides were desalted using sep-pak C18 column (Waters), lyophilized, and
reconstituted in folding buffer (0.1 mM EDTA, 2 mM GSH, 50 mM sodium phosphate, 250
mM NaCl buffer at pH 7.2). Folded cyclotides were purified by semipreparative. Purified
cyclotides were characterized by C18-RP-HPLC and ES-MS and quantified by UV-vis
spectroscopy.

MCo-PMI library construction. The MCo-PMI-based library template was chemically
synthesized (IDT DNA) and amplified using primers library-forward and library-reverse
(see Table 3.10). In order to produce a large amount of DNA library covering the
123

maximum number of sequences, ≈500 μg of plasmid pASK-TS-MCoTI and 60 μg of PCR
amplicon were digested sequentially with BglII followed by BstBI. 100 μg of double
digested plasmid and 15 µ g of PCR product were used for DNA ligation reaction (1:6
molar ratio of vector and insert fragment). Ligated DNA was purified and transformed into
DH5α electrocompetent cells. The size of the real library was estimated by plating a small
portion of the ligation mixture and counting the number of resulting colonies.
Table 3.10 Oligonucleotides for cloning MCo-PMI library plasmid

MCo-PMI
library
template
sequence
T TCG AAC TGC GGT TCT GGT TCT GGT GCT TCT AAA GCT
CCG ACC TCT TTC GCT GAA TAC TGG AAC CTG CTG TCT NNK
NNK NNK GTT TGC CCG AAA ATC CTG CAG CGT TGC CGT CGT
GAC NNK GAC TGC CCG NNK NNK TGC ATC TGC CGT GGT AAC
GGT TAC TGT TTA TCA TAT GAA ACG GAG ATC T
Library
Forward
TTC ATA GCT TCG AAC TGC GGT TCT GGT TC
Library
Reverse
TAC TGT CAA GAT CTC CGT TTC ATA TGA TAA ACA GTA ACC

124

CHAPTER FOUR Discussion
4.1 Discussion of molecular grafting method
The use of peptide-based drug is highly limited by their poor stability and bioavailability.
Here we have reported a novel approach involving the use of a stable disulfide-rich
backbone-cyclized polypeptide as a bio-scaffold to improve the delivery of bioactive
polypeptides into cells. We successfully used the cyclotide MCoTI-I to display an
-helical polypeptide PMI to improve its stability and cellular uptake properties. The
grafted cyclotide MCo-PMI was able to antagonize p53 binding domain of both HDM2
and HDMX with low nanomolar affinity, showed high stability in human serum and was
toxic to cancer cell lines expressing wild-type p53 both in vitro and in vivo through
activating p53 tumor suppressor pathway.
The design of the grafted cyclotide MCo-PMI was a success. Since both HDM2 and
HDMX bind to p53 through their p53-binding domain, targeting both HDM2 and HDMX is
critical for the reactivation of p53 tumor suppressor pathway in cell lines overexpressing
HDM2 and/or HDMX (Bernal et al., 2010). The novel cyclotide MCo-PMI was able to bind
to both HDM2 and HDMX, reactivate the p53 pathway in different cancer lines expressing
different levels of HDM2 and/or HDMX and not toxic to a non-cancer cell or cell lines with
unfunctional p53 protein. The p53 transcriptional targets were upregulated after MCo-PMI
treatment, indicating that MCo-PMI activates the p53 pathway in live cells. The
intravenous administration of MCo-PMI to mice with a human colon-carcinoma (HCT116)
xenograft tumor-induced up-regulation of the p53 pathway in the tumor tissue significantly
reducing its growth when compared to control animals treated with a saline solution of
nutlin-3. These results are in total agreement with the high serum stability of MCo-PMI
and with the ability to reversible bind serum proteins, which improves its half-life in blood.
125

Before these results were published, other groups have shown that the cyclotides
MCoTI-I/II and kalata B1 can be used for introducing novel biological activities by
molecular grafting (T. L. Aboye et al., 2012; Austin et al., 2009; Chan et al., 2011;
Gunasekera et al., 2008; Y.-H. Huang, Colgrave, Clark, Kotze, & Craik, 2010; P.
Thongyoo et al., 2008). For example, Craik group have grafted peptides from the
extracellular matrix protein laminin, osteopontin, and VEGF into loop 6 of cyclotide
MCoTI-II (Chan et al., 2011). The grafted cyclotides were shown to be more stable in
human serum than the corresponding linear peptides. The same group has also grafted
an Arg-rich peptide epitope into cyclotide kalata B1 resulting in a VEGF-A antagonist
(Gunasekera et al., 2008). These studies revealed that loops 6 and 3 of cyclotide kalata
B1 were the best loop for grafting foreign polypeptide sequences. Unfortunately, the
cyclotide kalata B1 has been reported to have cytotoxic effects of mammalian cells at a
concentration as low as 10 µM, which limits its potential for therapeutic use unless the
intrinsic cytotoxicity is used. Craik has claimed that mutating residue W23Y in the kalata
B1 can reduce significantly its cytotoxicity (Troeira Henriques et al., 2014) to mammalian
offering some hope on the use of this scaffold for potential biotherapeutic applications.
The Tam group has also used the cyclotide kalata B1 to engineer cyclotides able to
antagonize the bradykinin (BK) B1 receptor (Wong et al., 2012). Short BK B1-antagonist
peptides were grafted into the loop 6 of cyclotide kalata B1, resulting in the production of
bioactive cyclotides that were able to specifically block the BK B1 receptor in cell-based
assays. Importantly, the grafted cyclotide has also been shown active in an abdominal
constriction animal model assay and the cyclotides were also active when administered
orally. This result highlights the high stability and oral availability of the circular cysteine
knot cyclotide.
The Camarero group has also used MCoTI-I based grafted cyclotide targeting the
extracellular cytokine receptor CXCR4 (T. L. Aboye et al., 2012). A modified version of
the peptide CVX15 was grafted into loop 6 of the cyclotide MCoTI-I. The cyclotide was a
126

potent CXCR4 antagonist with IC 50 ≈20nM. This grafted cyclotide also has been shown to
possess high stability in serum, which indicates the cyclotide could be used as
bio-scaffold to delivery bioactive compounds.
The MCo-PMI cyclotide reported in this work is the first engineered cyclotide that can
effectively and selectively target an intracellular protein-protein interaction. Before this,
the cyclotide scaffold has been only used to graft small peptides (Chan et al., 2011;
Gunasekera et al., 2008; Y.-H. Huang et al., 2010). In this work, we proved that longer
peptide containing -helical segments can also be grafted into the MCoTI-cyclotide
framework and adopt the correct structure within the cyclotide scaffold. Protein
interactions involving -helical structure are very common in nature and successfully
grafting -helical segments into cyclotide should enable us to target also other
pharmacologically-relevant protein-protein interactions.
The same approach that was used to graft the p53-derived peptide was also recently
employed by another group to target the intracellular protein SET (D'Souza et al., 2016).
A potent SET antagonist peptide COG, which is derived from apolipoprotein E (apoE),
was grafted onto loop 6 of MCoTI-II using exactly same approach used in the design of
MCo-PMI. The resulting cyclotide MCOG1 and MCOG2 showed cytotoxic to a cancer cell
line and was stable in human serum with a 1/2 > 24 h. This study confirms again that the
MCoTI scaffold can be engineered to stabilize -helical peptide segments and deliver it
into cells to modulate intracellular signaling pathways.
Moreover, cyclotide MCo-PMI has five extra loops that could be used also for molecular
evolution to improve the activity and specificity of the parent cyclotide. This approach has
been explored in Chapter 3 using cyclotide MCo-PMI as a model system.
4.2 Discussion of the high-throughput barnase-based cell reporter
Library screening is a widely-used approach to obtain high-affinity peptide/protein
sequences. This approach required the generation of a large library and a screening
127

method to select out the most active compounds. Here we reported the design of
cyclotide MCoTI-I and MCo-PMI based genetically encoded libraries. To screen these
libraries, a creative in-cell high-throughput screening method was designed using the split
barnase protein complementation assay (PCA) approach. This screening method has
been proved able to select p53-HDM2 antagonist peptides both in vitro and in living E.
coli cells. A mock library and two real libraries have been screened using this
split-barnase cell-based reporter assay providing some novel cyclotide sequences that
are being further analyzed.
Amino acid scanning of cyclotide MCoTI-I (Austin et al., 2009) and kalata B1 (Simonsen
et al., 2008) have been already done. The folding of the corresponding cyclotide mutants
was interfered only by a couple of mutations while the rest of mutations did not affect
significantly the ability of the corresponding cyclotide mutants to fold correctly. This
sequence tolerance makes possible the introduction of other bioactive peptide
sequences into the cyclotide backbone providing cyclotide advantage over other type of
peptide scaffolds for the development of novel peptide-based therapeutics.
An acyclic version of cyclotide kalata B1 has been already used to generate a
genetically-encoded library for screening/selection of bioactive compounds using a
bacterial-display approach (Getz et al., 2011). In this study, a library with a variety of
sequences was constructed by first breaking kalata B1 backbone within loop 2 and fusing
the C-terminal glycine to the N-terminus of an engineered bacterial display protein
(eCPX). All seven residues in the loop 6 of kalata B1 were randomized using all the 20
genetically-encoded amino acids. This was accomplished by using NNK degenerate
codons. This cyclotide library was successfully screened using one round of
magnetic-activated cell sorting (MACS) followed by three cycles of FACS to yield a
diverse set of thrombin binding sequences with a consensus motif. The binding affinity
(Kd) to thrombin of two sequences, THR-5 and THR29, was determined to be ≈ 500 nM
and ≈330 nM, respectively.
128

A similar approach was lately used to screen cyclotides able to target the receptor
neuropilin-1 (Getz et al., 2013). Neuropilins interact with growth factors to promote
angiogenesis and tumor progression, including VEGF and HGF. Different from the
previous study, ligands from first generation library (loop6 randomization) were subjected
to one cycle of affinity maturation by fix loop 6 consensus sequence and randomize loop
5 sequence. This ended up with yield acyclic peptides with affinities of 40−60 nM toward
the receptors neuropilin-1. More importantly, the cyclic version of one sequence N2.1
retained high affinity toward neuropilin-1, exhibited increased protease resistance, and
conferred improved potency in in vitro cell migration inhibition assay.
However, there are still several issues about this approach. Even though kalata B1 can
cross the cell membrane, there is no reported kalata B1 derived cyclotides targeting
intracellular target. This could be related to the mechanism of kalata B1 cell permeability,
disrupting the cell membrane, which may cause high cytotoxicity to cells. Another
disadvantage of this approach is that an acyclic version of the cyclotide is used instead,
which may lose the stability comparing to the native cyclic backbone and may favor the
adoption of less stable alternative folds. The development of protein trans-splicing
approach has made possible express cyclized and fully folded cyclotide inside E. coli
cells in high yield providing relative high intracellular concentration (≈40 µ M) to perform
in-cell the screening/selection process (Krishnappa Jagadish et al., 2013).
The split-barnase screening system could be used to screen protein-protein interaction
inhibitors inside living cells, which is superior to display technology. The display approach
only requires two components, the protein of interest and the peptide library. The result
selected peptide sequences may bind to the protein but not inhibit protein-protein
interaction. On the other hand, the split-barnase screening system allows for a functional
screening like inhibition of protein-protein interaction. In addition, using display method,
the binding of protein and peptides is performed under non-physiological conditions,
comparing to physiological condition screening used by split barnase method. It is worth
129

to point out that in-cell screening approaches allow the process to proceed in a highly
crowded molecular environment favoring the selection of highly specific compounds.
Here, we reported the generation of a genetically-encoded MCoTI-based cyclotide library
in E. coli cells interfaced with a high-throughput split barnase screening approach. This
should enable us to perform the selection of novel cyclotide based inhibitors for
protein-protein interaction in living cells efficiently, making this approach extremely
valuable in the design of novel peptide-based therapeutic leads.
4.3 Future directions
Here, we report two approaches using cyclotide targeting intracellular protein-protein
interactions, in combination with a molecular grafting approach or molecular evolution
techniques. The grafting approach requires a known bioactive peptide in combination
with an appropriate design for grafting. The molecular evolution approach does not
require a known active peptide sequence. Instead, using this approach, a new sequence
will be selected from a genetically-encoded cyclotide-based library.
The grafted cyclotide MCo-PMI could be further improved by increase its stability or
cellular uptake. Comparing the serum stability of MCo-PMI and MCoTI-I (Fig. 2.8), the
grafting of PMI peptide into loop 6 slightly decreased its stability indicating the grafted
peptide part may contain a proteolytic soft spot. By analyzing the cleavage sites of
MCo-PMI, more stable versions of this cyclotide could be designed by adding for example
by introducing - or N-methylated residues close to the labile peptide bond. Improving the
cellular uptake of cyclotide could be another aspect for improving the activity of the
cyclotide MCo-PMI. Even though the mechanism of MCoTI-I uptake is not well
understood yet, introducing hydrophobic (Trp) or positively charged (Arg) residues within
the sequence of cyclotide has been shown to improve cellular uptake.
As mention in Section 3.9, the binding affinity of MCo-PMI may be improved by evolving
the sequence of surrounding loops. By introducing randomized mutations or sequences
130

on appropriate positions of MCo-PMI and using high-throughput screening approach,
cyclotide with higher binding affinity could be selected. This approach was described in
Section 3.9 and it is being pursued now in our lab.
So far, no active cyclotide sequences were selected from the MCoTI-I based library used
in this work. There are several things could be tried in future. The library size screened
was much smaller (less than 1%) than the theoretical size of the library, indicating that the
screening process was far from complete. In addition, the MCoTI-I library we designed
only mutated part of loop 1 and 5, leaving other loops untouched. As known, loop 6 was
the most flexible and sequence tolerant. Exploring other library designs that also involve
loop 6 could be another direction to pursue.
Finally, because the p53-HDM2/HDMX interaction is very well studied, it was used as PPI
target mode in our project as a proof of concept. Targeting other disease-related
protein-protein interaction using the same approaches is also being investigated in the
Camarero group to further explore the real potential of the cyclotide scaffold for the
development of a novel type of peptide-based therapeutics.

131

References
Aboye, T., Kuang, Y., Neamati, N., & Camarero, J. A. (2015). Rapid parallel synthesis of
bioactive folded cyclotides by using a tea-bag approach. Chembiochem, 16(5),
827-833. doi:10.1002/cbic.201402691
Aboye, T., Meeks, C. J., Majumder, S., Shekhtman, A., Rodgers, K., & Camarero, J. A.
(2016). Design of a MCoTI-Based Cyclotide with Angiotensin (1-7)-Like Activity.
Molecules, 21(2), 152. doi:10.3390/molecules21020152
Aboye, T. L., & Camarero, J. A. (2012). Biological synthesis of circular polypeptides. J
Biol Chem, 287(32), 27026-27032. doi:10.1074/jbc.R111.305508
Aboye, T. L., Ha, H., Majumder, S., Christ, F., Debyser, Z., Shekhtman, A., . . . Camarero,
J. A. (2012). Design of a novel cyclotide-based CXCR4 antagonist with anti-human
immunodeficiency virus (HIV)-1 activity. J Med Chem, 55(23), 10729-10734.
doi:10.1021/jm301468k
Austin, J., Kimura, R. H., Woo, Y. H., & Camarero, J. A. (2010). In vivo biosynthesis of an
Ala-scan library based on the cyclic peptide SFTI-1. Amino Acids, 38(5),
1313-1322. doi:10.1007/s00726-009-0338-4
Austin, J., Wang, W., Puttamadappa, S., Shekhtman, A., & Camarero, J. A. (2009).
Biosynthesis and biological screening of a genetically encoded library based on the
cyclotide MCoTI-I. Chembiochem, 10(16), 2663-2670.
doi:10.1002/cbic.200900534
Ayers, B., Blaschke, U. K., Camarero, J. A., Cotton, G. J., Holford, M., & Muir, T. W.
(1999). Introduction of unnatural amino acids into proteins using expressed protein
ligation. Biopolymers, 51(5), 343-354.
doi:10.1002/(SICI)1097-0282(1999)51:5<343::AID-BIP4>3.0.CO;2-W
Baker, S. J., Fearon, E. R., Nigro, J. M., Hamilton, S. R., Preisinger, A. C., Jessup, J. M., . . .
Vogelstein, B. (1989). Chromosome 17 deletions and p53 gene mutations in
colorectal carcinomas. Science, 244(4901), 217-221.
Bernal, F., Wade, M., Godes, M., Davis, T. N., Whitehead, D. G., Kung, A. L., . . .
Walensky, L. D. (2010). A stapled p53 helix overcomes HDMX-mediated
suppression of p53. Cancer cell, 18(5), 411-422.
Bi, T., Li, Y., Shekhtman, A., & Camarero, J. A. (2017). In-cell production of a
genetically-encoded library based on the θ-defensin RTD-1 using a bacterial
expression system. Bioorg Med Chem.
Borra, R., Dong, D., Elnagar, A. Y., Woldemariam, G. A., & Camarero, J. A. (2012).
In-cell fluorescence activation and labeling of proteins mediated by
FRET-quenched split inteins. J Am Chem Soc, 134(14), 6344-6353.
doi:10.1021/ja300209u
132

Brown, C. J., Cheok, C. F., Verma, C. S., & Lane, D. P. (2011). Reactivation of p53: from
peptides to small molecules. Trends Pharmacol Sci, 32(1), 53-62.
doi:10.1016/j.tips.2010.11.004
Brown, C. J., Lain, S., Verma, C. S., Fersht, A. R., & Lane, D. P. (2009). Awakening
guardian angels: drugging the p53 pathway. Nature Reviews Cancer, 9(12),
862-873.
Burgess, D. G., Ralston, E. J., Hanson, W. G., Heckert, M., Ho, M., Jenq, T., . . . Gutterson,
N. (2002). A novel, two‐component system for cell lethality and its use in
engineering nuclear male‐sterility in plants. The Plant Journal, 31(1), 113-125.
Burman, R., Svedlund, E., Felth, J., Hassan, S., Herrmann, A., Clark, R. J., . . . Gullbo, J.
(2010). Evaluation of toxicity and antitumor activity of cycloviolacin O2 in mice.
Biopolymers, 94(5), 626-634. doi:10.1002/bip.21408
Butler, J. S., Mitrea, D. M., Mitrousis, G., Cingolani, G., & Loh, S. N. (2009). Structural
and thermodynamic analysis of a conformationally strained circular permutant of
barnase. Biochemistry, 48(15), 3497-3507. doi:10.1021/bi900039e
Camarero, J., & Muir, T. (1997). Chemoselective backbone cyclization of unprotected
peptides. Chemical Communications(15), 1369-1370.
Camarero, J. A., Fushman, D., Cowburn, D., & Muir, T. W. (2001). Peptide chemical
ligation inside living cells: in vivo generation of a circular protein domain. Bioorg
Med Chem, 9(9), 2479-2484.
Camarero, J. A., & Mitchell, A. R. (2005). Synthesis of proteins by native chemical
ligation using Fmoc-based chemistry. Protein Pept Lett, 12(8), 723-728.
Camarero, J. A., & Muir, T. W. (1999). Biosynthesis of a head-to-tail cyclized protein with
improved biological activity. Journal of the American Chemical Society, 121(23),
5597-5598.
Camarero, J. A., Pavel, J., & Muir, T. W. (1998). Chemical synthesis of a circular protein
domain: evidence for folding-assisted cyclization. Angewandte Chemie
International Edition, 37(3), 347-349.
Cascales, L., Henriques, S. T., Kerr, M. C., Huang, Y. H., Sweet, M. J., Daly, N. L., &
Craik, D. J. (2011). Identification and characterization of a new family of
cell-penetrating peptides: cyclic cell-penetrating peptides. J Biol Chem, 286(42),
36932-36943. doi:10.1074/jbc.M111.264424
Chan, L. Y., Gunasekera, S., Henriques, S. T., Worth, N. F., Le, S. J., Clark, R. J., . . . Daly,
N. L. (2011). Engineering pro-angiogenic peptides using stable, disulfide-rich
cyclic scaffolds. Blood, 118(25), 6709-6717. doi:10.1182/blood-2011-06-359141
Chen, A. S., Metzger, J. M., Trumbauer, M. E., Guan, X. M., Yu, H., Frazier, E. G., . . . Van
der Ploeg, L. H. (2000). Role of the melanocortin-4 receptor in metabolic rate and
food intake in mice. Transgenic Res, 9(2), 145-154.
Chen, B., Colgrave, M. L., Daly, N. L., Rosengren, K. J., Gustafson, K. R., & Craik, D. J.
(2005). Isolation and characterization of novel cyclotides from Viola hederaceae:
133

solution structure and anti-HIV activity of vhl-1, a leaf-specific expressed cyclotide.
J Biol Chem, 280(23), 22395-22405. doi:10.1074/jbc.M501737200
Chen, D., Kon, N., Li, M., Zhang, W., Qin, J., & Gu, W. (2005). ARF-BP1/Mule is a
critical mediator of the ARF tumor suppressor. Cell, 121(7), 1071-1083.
doi:10.1016/j.cell.2005.03.037
Chen, J., Marechal, V., & Levine, A. J. (1993). Mapping of the p53 and mdm-2 interaction
domains. Mol Cell Biol, 13(7), 4107-4114.
Colgrave, M. L., & Craik, D. J. (2004). Thermal, chemical, and enzymatic stability of the
cyclotide kalata B1: the importance of the cyclic cystine knot. Biochemistry, 43(20),
5965-5975. doi:10.1021/bi049711q
Conibear, A. C., Wang, C. K., Bi, T., Rosengren, K. J., Camarero, J. A., & Craik, D. J.
(2014). Insights into the molecular flexibility of theta-defensins by NMR relaxation
analysis. J Phys Chem B, 118(49), 14257-14266. doi:10.1021/jp507754c
Contreras, J., Elnagar, A. Y., Hamm-Alvarez, S. F., & Camarero, J. A. (2011). Cellular
uptake of cyclotide MCoTI-I follows multiple endocytic pathways. J Control
Release, 155(2), 134-143. doi:10.1016/j.jconrel.2011.08.030
Cowper, B., Craik, D. J., & Macmillan, D. (2013). Making ends meet: chemically mediated
circularization of recombinant proteins. Chembiochem, 14(7), 809-812.
doi:10.1002/cbic.201300105
Craik, D. J., Daly, N. L., Bond, T., & Waine, C. (1999). Plant cyclotides: A unique family
of cyclic and knotted proteins that defines the cyclic cystine knot structural motif. J
Mol Biol, 294(5), 1327-1336. doi:10.1006/jmbi.1999.3383
Crooks, G. E., Hon, G., Chandonia, J.-M., & Brenner, S. E. (2004). WebLogo: a sequence
logo generator. Genome research, 14(6), 1188-1190.
D'Souza, C., Henriques, S. T., Wang, C. K., Cheneval, O., Chan, L. Y., Bokil, N. J., . . .
Craik, D. J. (2016). Using the MCoTI-II Cyclotide Scaffold To Design a Stable
Cyclic Peptide Antagonist of SET, a Protein Overexpressed in Human Cancer.
Biochemistry, 55(2), 396-405. doi:10.1021/acs.biochem.5b00529
Dahl, R., Chuchalin, A., Gor, D., Yoxall, S., & Sharma, R. (2006). EXCEL: A randomised
trial comparing salmeterol/fluticasone propionate and formoterol/budesonide
combinations in adults with persistent asthma. Respir Med, 100(7), 1152-1162.
doi:10.1016/j.rmed.2006.03.001
Daly, N. L., Clark, R. J., Plan, M. R., & Craik, D. J. (2006). Kalata B8, a novel antiviral
circular protein, exhibits conformational flexibility in the cystine knot motif.
Biochem J, 393(Pt 3), 619-626. doi:10.1042/BJ20051371
Daly, N. L., Love, S., Alewood, P. F., & Craik, D. J. (1999). Chemical synthesis and
folding pathways of large cyclic polypeptides: studies of the cystine knot
polypeptide kalata B1. Biochemistry, 38(32), 10606-10614.
doi:10.1021/bi990605b
Daly, N. L., Thorstholm, L., Greenwood, K. P., King, G. J., Rosengren, K. J., Heras, B., . . .
Craik, D. J. (2013). Structural insights into the role of the cyclic backbone in a
134

squash trypsin inhibitor. J Biol Chem, 288(50), 36141-36148.
doi:10.1074/jbc.M113.528240
Danovi, D., Meulmeester, E., Pasini, D., Migliorini, D., Capra, M., Frenk, R., . . . Marine, J.
C. (2004). Amplification of Mdmx (or Mdm4) directly contributes to tumor
formation by inhibiting p53 tumor suppressor activity. Mol Cell Biol, 24(13),
5835-5843. doi:10.1128/MCB.24.13.5835-5843.2004
Dawson, P. E., & Kent, S. B. (2000). Synthesis of native proteins by chemical ligation.
Annu Rev Biochem, 69, 923-960. doi:10.1146/annurev.biochem.69.1.923
de Graaf, P., Little, N. A., Ramos, Y. F., Meulmeester, E., Letteboer, S. J., & Jochemsen, A.
G. (2003). Hdmx protein stability is regulated by the ubiquitin ligase activity of
Mdm2. Journal of Biological Chemistry, 278(40), 38315-38324.
DeLeo, A. B., Jay, G., Appella, E., Dubois, G. C., Law, L. W., & Old, L. J. (1979).
Detection of a transformation-related antigen in chemically induced sarcomas and
other transformed cells of the mouse. Proc Natl Acad Sci U S A, 76(5), 2420-2424.
Deschuyteneer, G., Garcia, S., Michiels, B., Baudoux, B., Degand, H., Morsomme, P., &
Soumillion, P. (2010). Intein-mediated cyclization of randomized peptides in the
periplasm of Escherichia coli and their extracellular secretion. ACS Chem Biol, 5(7),
691-700. doi:10.1021/cb100072u
Dooley, C. T., & Houghten, R. A. (1993). The use of positional scanning synthetic peptide
combinatorial libraries for the rapid determination of opioid receptor ligands. Life
sciences, 52(18), 1509-1517.
Dornan, D., Wertz, I., Shimizu, H., Arnott, D., Frantz, G. D., Dowd, P., . . . Dixit, V. M.
(2004). The ubiquitin ligase COP1 is a critical negative regulator of p53. Nature,
429(6987), 86-92. doi:10.1038/nature02514
Eliasen, R., Daly, N. L., Wulff, B. S., Andresen, T. L., Conde-Frieboes, K. W., & Craik, D.
J. (2012). Design, synthesis, structural and functional characterization of novel
melanocortin agonists based on the cyclotide kalata B1. J Biol Chem, 287(48),
40493-40501. doi:10.1074/jbc.M112.395442
Esteller, M., Cordon-Cardo, C., Corn, P. G., Meltzer, S. J., Pohar, K. S., Watkins, D. N., . . .
Herman, J. G. (2001). p14ARF silencing by promoter hypermethylation mediates
abnormal intracellular localization of MDM2. Cancer Res, 61(7), 2816-2821.
Fang, S., Jensen, J. P., Ludwig, R. L., Vousden, K. H., & Weissman, A. M. (2000). Mdm2
is a RING finger-dependent ubiquitin protein ligase for itself and p53. Journal of
Biological Chemistry, 275(12), 8945-8951.
Finlay, C. A., Hinds, P. W., & Levine, A. J. (1989). The p53 proto-oncogene can act as a
suppressor of transformation. Cell, 57(7), 1083-1093.
Futaki, S., Sogawa, K., Maruyama, J., Asahara, T., Niwa, M., & Hojo, H. (1997).
Preparation of peptide thioesters using Fmoc-solid-phase peptide synthesis and its
application to the construction of a template-assembled synthetic protein (TASP).
Tetrahedron letters, 38(35), 6237-6240.
135

Garcia, A. E., & Camarero, J. A. (2010). Biological activities of natural and engineered
cyclotides, a novel molecular scaffold for peptide-based therapeutics. Curr Mol
Pharmacol, 3(3), 153-163.
Garcia, A. E., Tai, K. P., Puttamadappa, S. S., Shekhtman, A., Ouellette, A. J., & Camarero,
J. A. (2011). Biosynthesis and antimicrobial evaluation of backbone-cyclized
alpha-defensins. Biochemistry, 50(48), 10508-10519. doi:10.1021/bi201430f
Gerlach, S. L., Burman, R., Bohlin, L., Mondal, D., & Goransson, U. (2010). Isolation,
characterization, and bioactivity of cyclotides from the Micronesian plant
Psychotria leptothyrsa. J Nat Prod, 73(7), 1207-1213. doi:10.1021/np9007365
Gerlach, S. L., Rathinakumar, R., Chakravarty, G., Goransson, U., Wimley, W. C., Darwin,
S. P., & Mondal, D. (2010). Anticancer and chemosensitizing abilities of
cycloviolacin 02 from Viola odorata and psyle cyclotides from Psychotria
leptothyrsa. Biopolymers, 94(5), 617-625. doi:10.1002/bip.21435
Getz, J. A., Cheneval, O., Craik, D. J., & Daugherty, P. S. (2013). Design of a cyclotide
antagonist of neuropilin-1 and -2 that potently inhibits endothelial cell migration.
ACS Chem Biol, 8(6), 1147-1154. doi:10.1021/cb4000585
Getz, J. A., Rice, J. J., & Daugherty, P. S. (2011). Protease-resistant peptide ligands from a
knottin scaffold library. ACS Chem Biol, 6(8), 837-844. doi:10.1021/cb200039s
Goransson, U., & Craik, D. J. (2003). Disulfide mapping of the cyclotide kalata B1.
Chemical proof of the cystic cystine knot motif. J Biol Chem, 278(48),
48188-48196. doi:10.1074/jbc.M308771200
Gould, A., & Camarero, J. A. (2017). Cyclotides: Overview and Biotechnological
Applications. Chembiochem.
Gould, A., Li, Y., Majumder, S., Garcia, A. E., Carlsson, P., Shekhtman, A., & Camarero,
J. A. (2012). Recombinant production of rhesus theta-defensin-1 (RTD-1) using a
bacterial expression system. Mol Biosyst, 8(4), 1359-1365.
doi:10.1039/c2mb05451e
Grage, S. L., Sani, M. A., Cheneval, O., Henriques, S. T., Schalck, C., Heinzmann, R., . . .
Ulrich, A. S. (2017). Orientation and Location of the Cyclotide Kalata B1 in Lipid
Bilayers Revealed by Solid-State NMR. Biophys J, 112(4), 630-642.
doi:10.1016/j.bpj.2016.12.040
Gran, L. (1970). An oxytocic principle found in Oldenlandia affinis DC. Medd Nor Farm
Selsk, 12(173), 80.
Gran, L. (1973a). On the effect of a polypeptide isolated from "Kalata-Kalata"
(Oldenlandia affinis DC) on the oestrogen dominated uterus. Acta Pharmacol
Toxicol (Copenh), 33(5), 400-408.
Gran, L. (1973b). Oxytocic principles of Oldenlandia affinis. Lloydia, 36(2), 174-178.
Gray, K., Elghadban, S., Thongyoo, P., Owen, K. A., Szabo, R., Bugge, T. H., . . . Ellis, V.
(2014). Potent and specific inhibition of the biological activity of the type-II
transmembrane serine protease matriptase by the cyclic microprotein MCoTI-II.
Thromb Haemost, 112(2), 402-411. doi:10.1160/TH13-11-0895
136

Green, D. R., & Kroemer, G. (2009). Cytoplasmic functions of the tumour suppressor p53.
Nature, 458(7242), 1127-1130.
Greenwood, K. P., Daly, N. L., Brown, D. L., Stow, J. L., & Craik, D. J. (2007). The cyclic
cystine knot miniprotein MCoTI-II is internalized into cells by macropinocytosis.
Int J Biochem Cell Biol, 39(12), 2252-2264. doi:10.1016/j.biocel.2007.06.016
Gruber, C. W., Elliott, A. G., Ireland, D. C., Delprete, P. G., Dessein, S., Goransson, U., . . .
Craik, D. J. (2008). Distribution and evolution of circular miniproteins in flowering
plants. Plant Cell, 20(9), 2471-2483. doi:10.1105/tpc.108.062331
Gunasekera, S., Foley, F. M., Clark, R. J., Sando, L., Fabri, L. J., Craik, D. J., & Daly, N. L.
(2008). Engineering stabilized vascular endothelial growth factor-A antagonists:
synthesis, structural characterization, and bioactivity of grafted analogues of
cyclotides. J Med Chem, 51(24), 7697-7704. doi:10.1021/jm800704e
Gustafson, K. R., Sowder, R. C., Henderson, L. E., Parsons, I. C., Kashman, Y., Cardellina,
J. H., . . . Boyd, M. R. (1994). Circulins A and B. Novel human immunodeficiency
virus (HIV)-inhibitory macrocyclic peptides from the tropical tree Chassalia
parvifolia. Journal of the American Chemical Society, 116(20), 9337-9338.
Hainaut, P., & Hollstein, M. (2000). p53 and human cancer: the first ten thousand
mutations. Adv Cancer Res, 77, 81-137.
Hanahan, D., & Weinberg, R. A. (2011). Hallmarks of cancer: the next generation. Cell,
144(5), 646-674. doi:10.1016/j.cell.2011.02.013
Harris, J. M., Maciuca, R., Bradley, M. S., Cabanski, C. R., Scheerens, H., Lim, J., . . .
Kelsen, S. G. (2016). A randomized trial of the efficacy and safety of quilizumab in
adults with inadequately controlled allergic asthma. Respir Res, 17, 29.
doi:10.1186/s12931-016-0347-2
Haupt, Y., Maya, R., Kazaz, A., & Oren, M. (1997). Mdm2 promotes the rapid degradation
of p53. Nature, 387(6630), 296-299. doi:10.1038/387296a0
He, W., Chan, L. Y., Zeng, G., Daly, N. L., Craik, D. J., & Tan, N. (2011). Isolation and
characterization of cytotoxic cyclotides from Viola philippica. Peptides, 32(8),
1719-1723. doi:10.1016/j.peptides.2011.06.016
Heitz, A., Hernandez, J. F., Gagnon, J., Hong, T. T., Pham, T. T., Nguyen, T. M., . . .
Chiche, L. (2001). Solution structure of the squash trypsin inhibitor MCoTI-II. A
new family for cyclic knottins. Biochemistry, 40(27), 7973-7983.
Henriques, S. T., Huang, Y. H., Castanho, M. A., Bagatolli, L. A., Sonza, S., Tachedjian,
G., . . . Craik, D. J. (2012). Phosphatidylethanolamine binding is a conserved
feature of cyclotide-membrane interactions. J Biol Chem, 287(40), 33629-33643.
doi:10.1074/jbc.M112.372011
Henriques, S. T., Huang, Y. H., Chaousis, S., Sani, M. A., Poth, A. G., Separovic, F., &
Craik, D. J. (2015). The Prototypic Cyclotide Kalata B1 Has a Unique Mechanism
of Entering Cells. Chem Biol, 22(8), 1087-1097.
doi:10.1016/j.chembiol.2015.07.012
137

Henriques, S. T., Huang, Y. H., Rosengren, K. J., Franquelim, H. G., Carvalho, F. A.,
Johnson, A., . . . Craik, D. J. (2011). Decoding the membrane activity of the
cyclotide kalata B1: the importance of phosphatidylethanolamine phospholipids
and lipid organization on hemolytic and anti-HIV activities. J Biol Chem, 286(27),
24231-24241. doi:10.1074/jbc.M111.253393
Hernandez, J. F., Gagnon, J., Chiche, L., Nguyen, T. M., Andrieu, J. P., Heitz, A., . . . Le
Nguyen, D. (2000). Squash trypsin inhibitors from Momordica cochinchinensis
exhibit an atypical macrocyclic structure. Biochemistry, 39(19), 5722-5730.
Honda, R., Tanaka, H., & Yasuda, H. (1997). Oncoprotein MDM2 is a ubiquitin ligase E3
for tumor suppressor p53. FEBS Lett, 420(1), 25-27.
Huang, Y.-H., Colgrave, M. L., Clark, R. J., Kotze, A. C., & Craik, D. J. (2010).
Lysine-scanning mutagenesis reveals an amendable face of the cyclotide kalata B1
for the optimization of nematocidal activity. Journal of Biological Chemistry,
285(14), 10797-10805.
Huang, Y. H., Chaousis, S., Cheneval, O., Craik, D. J., & Henriques, S. T. (2015).
Optimization of the cyclotide framework to improve cell penetration properties.
Front Pharmacol, 6, 17. doi:10.3389/fphar.2015.00017
Ingenito, R., Bianchi, E., Fattori, D., & Pessi, A. (1999). Solid phase synthesis of peptide
C-terminal thioesters by Fmoc/t-Bu chemistry. Journal of the American Chemical
Society, 121(49), 11369-11374.
Ireland, D. C., Wang, C. K., Wilson, J. A., Gustafson, K. R., & Craik, D. J. (2008).
Cyclotides as natural anti-HIV agents. Biopolymers, 90(1), 51-60.
doi:10.1002/bip.20886
Iwai, H., Zuger, S., Jin, J., & Tam, P. H. (2006). Highly efficient protein trans-splicing by a
naturally split DnaE intein from Nostoc punctiforme. FEBS Lett, 580(7),
1853-1858. doi:10.1016/j.febslet.2006.02.045
Jagadish, K., Borra, R., Lacey, V., Majumder, S., Shekhtman, A., Wang, L., & Camarero, J.
A. (2013). Expression of fluorescent cyclotides using protein trans-splicing for
easy monitoring of cyclotide-protein interactions. Angew Chem Int Ed Engl, 52(11),
3126-3131. doi:10.1002/anie.201209219
Jagadish, K., Borra, R., Lacey, V., Majumder, S., Shekhtman, A., Wang, L., & Camarero, J.
A. (2013). Expression of Fluorescent Cyclotides using Protein Trans‐Splicing for
Easy Monitoring of Cyclotide–Protein Interactions. Angewandte Chemie
International Edition, 52(11), 3126-3131.
Jagadish, K., & Camarero, J. A. (2010). Cyclotides, a promising molecular scaffold for
peptide‐based therapeutics. Peptide Science, 94(5), 611-616.
Jagadish, K., Gould, A., Borra, R., Majumder, S., Mushtaq, Z., Shekhtman, A., &
Camarero, J. A. (2015). Recombinant Expression and Phenotypic Screening of a
Bioactive Cyclotide Against alpha-Synuclein-Induced Cytotoxicity in Baker's
Yeast. Angew Chem Int Ed Engl, 54(29), 8390-8394. doi:10.1002/anie.201501186
138

Jain, A. K., & Barton, M. C. (2010). Making sense of ubiquitin ligases that regulate p53.
Cancer Biol Ther, 10(7), 665-672.
Jennings, C., West, J., Waine, C., Craik, D., & Anderson, M. (2001). Biosynthesis and
insecticidal properties of plant cyclotides: the cyclic knotted proteins from
Oldenlandia affinis. Proc Natl Acad Sci U S A, 98(19), 10614-10619.
doi:10.1073/pnas.191366898
Jennings, C. V., Rosengren, K. J., Daly, N. L., Plan, M., Stevens, J., Scanlon, M. J., . . .
Craik, D. J. (2005). Isolation, solution structure, and insecticidal activity of kalata
B2, a circular protein with a twist: do Mobius strips exist in nature? Biochemistry,
44(3), 851-860. doi:10.1021/bi047837h
Ji, Y., Majumder, S., Millard, M., Borra, R., Bi, T., Elnagar, A. Y., . . . Camarero, J. A.
(2013). In vivo activation of the p53 tumor suppressor pathway by an engineered
cyclotide. J Am Chem Soc, 135(31), 11623-11633. doi:10.1021/ja405108p
Kaas, Q., & Craik, D. J. (2010). Analysis and classification of circular proteins in CyBase.
Biopolymers, 94(5), 584-591. doi:10.1002/bip.21424
Kempe, K., Rubtsova, M., & Gils, M. (2009). Intein-mediated protein assembly in
transgenic wheat: production of active barnase and acetolactate synthase from split
genes. Plant Biotechnol J, 7(3), 283-297. doi:10.1111/j.1467-7652.2008.00399.x
Kimura, R., & Camarero, J. A. (2005). Expressed protein ligation: a new tool for the
biosynthesis of cyclic polypeptides. Protein Pept Lett, 12(8), 789-794.
Kimura, R. H., Steenblock, E. R., & Camarero, J. A. (2007). Development of a cell-based
fluorescence resonance energy transfer reporter for Bacillus anthracis lethal factor
protease. Analytical biochemistry, 369(1), 60-70.
Kimura, R. H., Teed, R., Hackel, B. J., Pysz, M. A., Chuang, C. Z., Sathirachinda, A., . . .
Gambhir, S. S. (2012). Pharmacokinetically stabilized cystine knot peptides that
bind alpha-v-beta-6 integrin with single-digit nanomolar affinities for detection of
pancreatic cancer. Clinical Cancer Research, 18(3), 839-849.
Kimura, R. H., Tran, A. T., & Camarero, J. A. (2006). Biosynthesis of the cyclotide Kalata
B1 by using protein splicing. Angew Chem Int Ed Engl, 45(6), 973-976.
doi:10.1002/anie.200503882
Kruse, J. P., & Gu, W. (2009a). Modes of p53 regulation. Cell, 137(4), 609-622.
doi:10.1016/j.cell.2009.04.050
Kruse, J. P., & Gu, W. (2009b). MSL2 promotes Mdm2-independent cytoplasmic
localization of p53. J Biol Chem, 284(5), 3250-3263. doi:10.1074/jbc.M805658200
Kubbutat, M. H., Jones, S. N., & Vousden, K. H. (1997). Regulation of p53 stability by
Mdm2. Nature, 387(6630), 299-303. doi:10.1038/387299a0
Lane, D. P. (1992). Cancer. p53, guardian of the genome. Nature, 358(6381), 15-16.
doi:10.1038/358015a0
Lane, D. P., & Crawford, L. V. (1979). T antigen is bound to a host protein in
SV40-transformed cells. Nature, 278(5701), 261-263.
139

Laurie, N. A., Donovan, S. L., Shih, C. S., Zhang, J., Mills, N., Fuller, C., . . . Dyer, M. A.
(2006). Inactivation of the p53 pathway in retinoblastoma. Nature, 444(7115),
61-66. doi:10.1038/nature05194
Leng, R. P., Lin, Y., Ma, W., Wu, H., Lemmers, B., Chung, S., . . . Benchimol, S. (2003).
Pirh2, a p53-induced ubiquitin-protein ligase, promotes p53 degradation. Cell,
112(6), 779-791.
Leta Aboye, T., Clark, R. J., Craik, D. J., & Göransson, U. (2008). Ultra‐Stable Peptide
Scaffolds for Protein Engineering—Synthesis and Folding of the Circular Cystine
Knotted Cyclotide Cycloviolacin O2. Chembiochem, 9(1), 103-113.
Levine, A. J. (1997). p53, the cellular gatekeeper for growth and division. Cell, 88(3),
323-331.
Li, C., Pazgier, M., Li, C., Yuan, W., Liu, M., Wei, G., . . . Lu, W. (2010). Systematic
mutational analysis of peptide inhibition of the p53-MDM2/MDMX interactions. J
Mol Biol, 398(2), 200-213. doi:10.1016/j.jmb.2010.03.005
Li, C., Pazgier, M., Liu, M., Lu, W. Y., & Lu, W. (2009). Apamin as a Template for
Structure‐Based Rational Design of Potent Peptide Activators of p53. Angewandte
Chemie International Edition, 48(46), 8712-8715.
Li, M., Brooks, C. L., Wu-Baer, F., Chen, D., Baer, R., & Gu, W. (2003). Mono- versus
polyubiquitination: differential control of p53 fate by Mdm2. Science, 302(5652),
1972-1975. doi:10.1126/science.1091362
Li, Y., Bi, T., & Camarero, J. A. (2015). Chapter Nine-Chemical and Biological
Production of Cyclotides. Advances in botanical research, 76, 271-303.
Linares, L. K., Hengstermann, A., Ciechanover, A., Muller, S., & Scheffner, M. (2003).
HdmX stimulates Hdm2-mediated ubiquitination and degradation of p53. Proc
Natl Acad Sci U S A, 100(21), 12009-12014. doi:10.1073/pnas.2030930100
Lindholm, P., Goransson, U., Johansson, S., Claeson, P., Gullbo, J., Larsson, R., . . .
Backlund, A. (2002). Cyclotides: a novel type of cytotoxic agents. Mol Cancer
Ther, 1(6), 365-369.
Linzer, D. I., & Levine, A. J. (1979). Characterization of a 54K dalton cellular SV40 tumor
antigen present in SV40-transformed cells and uninfected embryonal carcinoma
cells. Cell, 17(1), 43-52.
Liu, M., Li, C., Pazgier, M., Li, C., Mao, Y., Lv, Y., . . . Zhan, C. (2010). D-peptide
inhibitors of the p53–MDM2 interaction for targeted molecular therapy of
malignant neoplasms. Proceedings of the National Academy of Sciences, 107(32),
14321-14326.
Lohrum, M. A., Woods, D. B., Ludwig, R. L., Balint, E., & Vousden, K. H. (2001).
C-terminal ubiquitination of p53 contributes to nuclear export. Mol Cell Biol,
21(24), 8521-8532. doi:10.1128/MCB.21.24.8521-8532.2001
Marglin, A., & Merrifield, R. B. (1970). Chemical synthesis of peptides and proteins. Annu
Rev Biochem, 39, 841-866. doi:10.1146/annurev.bi.39.070170.004205
140

Marine, J. C., & Jochemsen, A. G. (2005). Mdmx as an essential regulator of p53 activity.
Biochem Biophys Res Commun, 331(3), 750-760. doi:10.1016/j.bbrc.2005.03.151
Michael, D., & Oren, M. (2003). The p53-Mdm2 module and the ubiquitin system. Semin
Cancer Biol, 13(1), 49-58.
Millard, M., Pathania, D., Grande, F., Xu, S., & Neamati, N. (2011). Small-molecule
inhibitors of p53-MDM2 interaction: the 2006-2010 update. Current
pharmaceutical design, 17(6), 536-559.
Momand, J., Zambetti, G. P., Olson, D. C., George, D., & Levine, A. J. (1992). The mdm-2
oncogene product forms a complex with the p53 protein and inhibits p53-mediated
transactivation. Cell, 69(7), 1237-1245.
Mylne, J. S., Wang, C. K., van der Weerden, N. L., & Craik, D. J. (2010). Cyclotides are a
component of the innate defense of Oldenlandia affinis. Biopolymers, 94(5),
635-646. doi:10.1002/bip.21419
Neira, J. L., Vá zquez, E., & Fersht, A. R. (2000). Stability and folding of the protein
complexes of barnase. The FEBS Journal, 267(10), 2859-2870.
Nguyen, A. W., & Daugherty, P. S. (2005). Evolutionary optimization of fluorescent
proteins for intracellular FRET. Nature biotechnology, 23(3), 355.
Nguyen, G. K., Zhang, S., Nguyen, N. T., Nguyen, P. Q., Chiu, M. S., Hardjojo, A., & Tam,
J. P. (2011). Discovery and characterization of novel cyclotides originated from
chimeric precursors consisting of albumin-1 chain a and cyclotide domains in the
Fabaceae family. J Biol Chem, 286(27), 24275-24287.
doi:10.1074/jbc.M111.229922
Nguyen, K. N., Nguyen, G. K., Nguyen, P. Q., Ang, K. H., Dedon, P. C., & Tam, J. P.
(2016). Immunostimulating and Gram-negative-specific antibacterial cyclotides
from the butterfly pea (Clitoria ternatea). FEBS J, 283(11), 2067-2090.
doi:10.1111/febs.13720
Nie, L., Sasaki, M., & Maki, C. G. (2007). Regulation of p53 nuclear export through
sequential changes in conformation and ubiquitination. J Biol Chem, 282(19),
14616-14625. doi:10.1074/jbc.M610515200
Pazgier, M., Liu, M., Zou, G., Yuan, W., Li, C., Li, C., . . . Tarasov, S. G. (2009). Structural
basis for high-affinity peptide inhibition of p53 interactions with MDM2 and
MDMX. Proceedings of the National Academy of Sciences, 106(12), 4665-4670.
Pazgier, M., Liu, M., Zou, G., Yuan, W., Li, C., Li, J., . . . Lu, W. (2009). Structural basis
for high-affinity peptide inhibition of p53 interactions with MDM2 and MDMX.
Proc Natl Acad Sci U S A, 106(12), 4665-4670. doi:10.1073/pnas.0900947106
Phan, J., Li, Z., Kasprzak, A., Li, B., Sebti, S., Guida, W., . . . Chen, J. (2010).
Structure-based design of high affinity peptides inhibiting the interaction of p53
with MDM2 and MDMX. J Biol Chem, 285(3), 2174-2183.
doi:10.1074/jbc.M109.073056
141

Picksley, S. M., & Lane, D. P. (1993). The p53-mdm2 autoregulatory feedback loop: a
paradigm for the regulation of growth control by p53? Bioessays, 15(10), 689-690.
doi:10.1002/bies.950151008
Piehler, J. (2005). New methodologies for measuring protein interactions in vivo and in
vitro. Curr Opin Struct Biol, 15(1), 4-14. doi:10.1016/j.sbi.2005.01.008
Pinto, M. F., Fensterseifer, I. C., Migliolo, L., Sousa, D. A., de Capdville, G.,
Arboleda-Valencia, J. W., . . . Franco, O. L. (2012). Identification and structural
characterization of novel cyclotide with activity against an insect pest of sugar cane.
J Biol Chem, 287(1), 134-147. doi:10.1074/jbc.M111.294009
Plan, M. R. R., Saska, I., Cagauan, A. G., & Craik, D. J. (2008). Backbone cyclised
peptides from plants show molluscicidal activity against the rice pest Pomacea
canaliculata (golden apple snail). Journal of agricultural and food chemistry,
56(13), 5237-5241.
Popowicz, G. M., Czarna, A., Wolf, S., Wang, K., Wang, W., Dö mling, A., & Holak, T. A.
(2010). Structures of low molecular weight inhibitors bound to MDMX and MDM2
reveal new approaches for p53-MDMX/MDM2 antagonist drug discovery. Cell
cycle, 9(6), 1104-1111.
Poth, A. G., Colgrave, M. L., Lyons, R. E., Daly, N. L., & Craik, D. J. (2011). Discovery of
an unusual biosynthetic origin for circular proteins in legumes. Proc Natl Acad Sci
U S A, 108(25), 10127-10132. doi:10.1073/pnas.1103660108
Poyurovsky, M. V., Priest, C., Kentsis, A., Borden, K. L., Pan, Z. Q., Pavletich, N., &
Prives, C. (2007). The Mdm2 RING domain C-terminus is required for
supramolecular assembly and ubiquitin ligase activity. EMBO J, 26(1), 90-101.
doi:10.1038/sj.emboj.7601465
Puttamadappa, S. S., Jagadish, K., Shekhtman, A., & Camarero, J. A. (2010). Backbone
dynamics of cyclotide MCoTI-I free and complexed with trypsin. Angew Chem Int
Ed Engl, 49(39), 7030-7034. doi:10.1002/anie.201002906
Remy, I., & Michnick, S. W. (2007). Application of protein-fragment complementation
assays in cell biology. Biotechniques, 42(2), 137, 139, 141 passim.
Riemenschneider, M. J., Buschges, R., Wolter, M., Reifenberger, J., Bostrom, J., Kraus, J.
A., . . . Reifenberger, G. (1999). Amplification and overexpression of the MDM4
(MDMX) gene from 1q32 in a subset of malignant gliomas without TP53 mutation
or MDM2 amplification. Cancer Res, 59(24), 6091-6096.
Ringshausen, I., O'Shea, C. C., Finch, A. J., Swigart, L. B., & Evan, G. I. (2006). Mdm2 is
critically and continuously required to suppress lethal p53 activity in vivo. Cancer
Cell, 10(6), 501-514. doi:10.1016/j.ccr.2006.10.010
Rosengren, K. J., Daly, N. L., Plan, M. R., Waine, C., & Craik, D. J. (2003). Twists, knots,
and rings in proteins. Structural definition of the cyclotide framework. J Biol Chem,
278(10), 8606-8616. doi:10.1074/jbc.M211147200
142

Saether, O., Craik, D. J., Campbell, I. D., Sletten, K., Juul, J., & Norman, D. G. (1995).
Elucidation of the primary and three-dimensional structure of the uterotonic
polypeptide kalata B1. Biochemistry, 34(13), 4147-4158.
Sancheti, H., & Camarero, J. A. (2009). "Splicing up" drug discovery. Cell-based
expression and screening of genetically-encoded libraries of backbone-cyclized
polypeptides. Adv Drug Deliv Rev, 61(11), 908-917.
doi:10.1016/j.addr.2009.07.003
Sancho, J., & Fersht, A. R. (1992). Dissection of an enzyme by protein engineering: The N
and C-terminal fragments of barnase form a native-like complex with restored
enzymic activity. J Mol Biol, 224(3), 741-747.
Schö pke, T., Hasan Agha, M., Kraft, R., Otto, A., & Hiller, K. (1993). Hä molytisch aktive
komponenten aus Viola tricolor L. und Viola arvensis Murray. Scientia
Pharmaceutica, 61, 145-153.
Scott, C. P., Abel-Santos, E., Wall, M., Wahnon, D. C., & Benkovic, S. J. (1999).
Production of cyclic peptides and proteins in vivo. Proc Natl Acad Sci U S A,
96(24), 13638-13643.
Shangary, S., Qin, D., McEachern, D., Liu, M., Miller, R. S., Qiu, S., . . . Wang, S. (2008).
Temporal activation of p53 by a specific MDM2 inhibitor is selectively toxic to
tumors and leads to complete tumor growth inhibition. Proc Natl Acad Sci U S A,
105(10), 3933-3938. doi:10.1073/pnas.0708917105
Shangary, S., & Wang, S. (2009). Small-molecule inhibitors of the MDM2-p53
protein-protein interaction to reactivate p53 function: a novel approach for cancer
therapy. Annu Rev Pharmacol Toxicol, 49, 223-241.
doi:10.1146/annurev.pharmtox.48.113006.094723
Shekhawat, S. S., & Ghosh, I. (2011). Split-protein systems: beyond binary protein-protein
interactions. Curr Opin Chem Biol, 15(6), 789-797.
doi:10.1016/j.cbpa.2011.10.014
Sherr, C. J., & Weber, J. D. (2000). The ARF/p53 pathway. Curr Opin Genet Dev, 10(1),
94-99.
Shimizu, H., Burch, L. R., Smith, A. J., Dornan, D., Wallace, M., Ball, K. L., & Hupp, T. R.
(2002). The conformationally flexible S9-S10 linker region in the core domain of
p53 contains a novel MDM2 binding site whose mutation increases ubiquitination
of p53 in vivo. J Biol Chem, 277(32), 28446-28458. doi:10.1074/jbc.M202296200
Shvarts, A., Steegenga, W., Riteco, N., Van Laar, T., Dekker, P., Bazuine, M., . . . Van der
Eb, A. (1996). MDMX: a novel p53-binding protein with some functional
properties of MDM2. EMBO J, 15(19), 5349.
Simonsen, S. M., Sando, L., Rosengren, K. J., Wang, C. K., Colgrave, M. L., Daly, N. L.,
& Craik, D. J. (2008). Alanine scanning mutagenesis of the prototypic cyclotide
reveals a cluster of residues essential for bioactivity. J Biol Chem, 283(15),
9805-9813. doi:10.1074/jbc.M709303200
143

Svangard, E., Goransson, U., Hocaoglu, Z., Gullbo, J., Larsson, R., Claeson, P., & Bohlin,
L. (2004). Cytotoxic cyclotides from Viola tricolor. J Nat Prod, 67(2), 144-147.
doi:10.1021/np030101l
Tam, J. P., Lu, Y. A., Yang, J. L., & Chiu, K. W. (1999). An unusual structural motif of
antimicrobial peptides containing end-to-end macrocycle and cystine-knot
disulfides. Proc Natl Acad Sci U S A, 96(16), 8913-8918.
Tang, J., Wang, C. K., Pan, X., Yan, H., Zeng, G., Xu, W., . . . Tan, N. (2010). Isolation and
characterization of cytotoxic cyclotides from Viola tricolor. Peptides, 31(8),
1434-1440. doi:10.1016/j.peptides.2010.05.004
Tang, Y., Zhao, W., Chen, Y., Zhao, Y., & Gu, W. (2008). Acetylation is indispensable for
p53 activation. Cell, 133(4), 612-626. doi:10.1016/j.cell.2008.03.025
Tavassoli, A., & Benkovic, S. J. (2007). Split-intein mediated circular ligation used in the
synthesis of cyclic peptide libraries in E. coli. Nat Protoc, 2(5), 1126-1133.
doi:10.1038/nprot.2007.152
Thongyoo, P., Bonomelli, C., Leatherbarrow, R. J., & Tate, E. W. (2009). Potent inhibitors
of beta-tryptase and human leukocyte elastase based on the MCoTI-II scaffold. J
Med Chem, 52(20), 6197-6200. doi:10.1021/jm901233u
Thongyoo, P., Roque-Rosell, N., Leatherbarrow, R. J., & Tate, E. W. (2008). Chemical and
biomimetic total syntheses of natural and engineered MCoTI cyclotides. Org
Biomol Chem, 6(8), 1462-1470. doi:10.1039/b801667d
Thongyoo, P., Tate, E. W., & Leatherbarrow, R. J. (2006). Total synthesis of the
macrocyclic cysteine knot microprotein MCoTI-II. Chemical Communications(27),
2848-2850.
Tolbert, T. J., & Wong, C. H. (2002). New methods for proteomic research: preparation of
proteins with N-terminal cysteines for labeling and conjugation. Angew Chem Int
Ed Engl, 41(12), 2171-2174.
Trabi, M., & Craik, D. J. (2004). Tissue-specific expression of head-to-tail cyclized
miniproteins in Violaceae and structure determination of the root cyclotide Viola
hederacea root cyclotide1. Plant Cell, 16(8), 2204-2216.
doi:10.1105/tpc.104.021790
Trabi, M., Svangard, E., Herrmann, A., Goransson, U., Claeson, P., Craik, D. J., & Bohlin,
L. (2004). Variations in cyclotide expression in viola species. J Nat Prod, 67(5),
806-810. doi:10.1021/np034068e
Troeira Henriques, S., & Craik, D. J. (2017). Cyclotide Structure and Function: The Role
of Membrane Binding and Permeation. Biochemistry, 56(5), 669-682.
doi:10.1021/acs.biochem.6b01212
Troeira Henriques, S., Huang, Y. H., Chaousis, S., Wang, C. K., & Craik, D. J. (2014).
Anticancer and toxic properties of cyclotides are dependent on
phosphatidylethanolamine phospholipid targeting. Chembiochem, 15(13),
1956-1965. doi:10.1002/cbic.201402144
144

Uldrijan, S., Pannekoek, W. J., & Vousden, K. H. (2007). An essential function of the
extreme C-terminus of MDM2 can be provided by MDMX. EMBO J, 26(1),
102-112. doi:10.1038/sj.emboj.7601469
Vassilev, L. T., Vu, B. T., Graves, B., Carvajal, D., Podlaski, F., Filipovic, Z., . . . Liu, E. A.
(2004). In vivo activation of the p53 pathway by small-molecule antagonists of
MDM2. Science, 303(5659), 844-848. doi:10.1126/science.1092472
Vogelstein, B., & Kinzler, K. W. (1992). p53 function and dysfunction. Cell, 70(4),
523-526.
Vogelstein, B., Lane, D., & Levine, A. J. (2000). Surfing the p53 network. Nature,
408(6810), 307-310. doi:10.1038/35042675
von Eggelkraut-Gottanka, R., Klose, A., Beck-Sickinger, A. G., & Beyermann, M. (2003).
Peptide α thioester formation using standard Fmoc-chemistry. Tetrahedron letters,
44(17), 3551-3554.
Vousden, K. H., & Lane, D. P. (2007). p53 in health and disease. Nat Rev Mol Cell Biol,
8(4), 275-283. doi:10.1038/nrm2147
Vousden, K. H., & Prives, C. (2009). Blinded by the Light: The Growing Complexity of
p53. Cell, 137(3), 413-431. doi:10.1016/j.cell.2009.04.037
Wang, C. K., Colgrave, M. L., Gustafson, K. R., Ireland, D. C., Goransson, U., & Craik, D.
J. (2008). Anti-HIV cyclotides from the Chinese medicinal herb Viola yedoensis. J
Nat Prod, 71(1), 47-52. doi:10.1021/np070393g
Wang, C. K., Gruber, C. W., Cemazar, M., Siatskas, C., Tagore, P., Payne, N., . . . Craik, D.
J. (2014). Molecular grafting onto a stable framework yields novel cyclic peptides
for the treatment of multiple sclerosis. ACS Chem Biol, 9(1), 156-163.
doi:10.1021/cb400548s
Wang, C. K., Stalmans, S., De Spiegeleer, B., & Craik, D. J. (2016). Biodistribution of the
cyclotide MCoTI-II, a cyclic disulfide-rich peptide drug scaffold. J Pept Sci, 22(5),
305-310. doi:10.1002/psc.2862
Witherup, K. M., Bogusky, M. J., Anderson, P. S., Ramjit, H., Ransom, R. W., Wood, T.,
& Sardana, M. (1994). Cyclopsychotride A, a biologically active, 31-residue cyclic
peptide isolated from Psychotria longipes. J Nat Prod, 57(12), 1619-1625.
Wolf, D., Harris, N., & Rotter, V. (1984). Reconstitution of p53 expression in a
nonproducer Ab-MuLV-transformed cell line by transfection of a functional p53
gene. Cell, 38(1), 119-126.
Wong, C. T., Rowlands, D. K., Wong, C. H., Lo, T. W., Nguyen, G. K., Li, H. Y., & Tam,
J. P. (2012). Orally active peptidic bradykinin B1 receptor antagonists engineered
from a cyclotide scaffold for inflammatory pain treatment. Angew Chem Int Ed
Engl, 51(23), 5620-5624. doi:10.1002/anie.201200984
Yonish-Rouach, E., Resnitzky, D., Lotem, J., Sachs, L., Kimchi, A., & Oren, M. (1991).
Wild-type p53 induces apoptosis of myeloid leukaemic cells that is inhibited by
interleukin-6. Nature, 352(6333), 345-347. doi:10.1038/352345a0
145

Young, T. S., Young, D. D., Ahmad, I., Louis, J. M., Benkovic, S. J., & Schultz, P. G.
(2011). Evolution of cyclic peptide protease inhibitors. Proc Natl Acad Sci U S A,
108(27), 11052-11056. doi:10.1073/pnas.1108045108
Yu, G. W., Rudiger, S., Veprintsev, D., Freund, S., Fernandez-Fernandez, M. R., & Fersht,
A. R. (2006). The central region of HDM2 provides a second binding site for p53.
Proc Natl Acad Sci U S A, 103(5), 1227-1232. doi:10.1073/pnas.0510343103

Abstract (if available)

Abstract The success of protein-based therapeutics has revolutionized drug development. Unlike small molecule drugs, peptide and protein-based therapeutics can target defective protein-protein interactions involved in human diseases with high selectivity and specificity. Despite their success, however, there are numerous stability and delivery issues associated with their use as therapeutic agents. As an emerging macromolecule, cyclotide has excellent pharmacological properties. ❧ Here, we explored the possibility of cyclotide to target protein-protein interactions. The overexpression of HDM2 and HDMX is a common mechanism used by many tumor cells to inactive the p53 tumor suppressor pathway promoting cell survival. Targeting HDM2 and HDMX has emerged as a validated therapeutic strategy for treating cancers with wild-type p53. Small linear peptides mimicking the N-terminal fragment of p53 have been shown to be potent HDM2/HDMX antagonists. The potential therapeutic use of these peptides, however, is limited by their poor stability and bioavailability. ❧ In this project, we first report the engineering of the cyclotide MCoTI-I to efficiently antagonize intracellular p53 degradation. The resulting cyclotide MCo-PMI was able to bind with low nanomolar affinity to both HDM2 and HDMX, showed high stability in human serum, and was cytotoxic to wild-type p53 cancer cell lines by activating the p53 tumor suppressor pathway both in vitro and in vivo. ❧ We also report a cell based high-throughput screening method for selection of protein-protein interaction inhibitors using a library based on the cyclotide MCoTI-I. This screening method has been proved able to select p53-HDM2 antagonist peptide both in vitro and in living E. coli cell. Both mock library and real libraries have been screened by this split barnase method and some preliminary sequences have been analyzed. These features make the cyclotide MCoTI-I an optimal scaffold for targeting intracellular protein−protein interactions.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Apoptotic proteins ATF3, RUNX3, FOXO3a, DYRK2, HIPK2 are up regulated by inhibiting E3 ligase of Hdm2 and HdmX in human colon cancer cell lines and ovarian cancer cell lines

PDF

Screening of cysteine-rich backbone cyclized peptide libraries for inhibitors against disease-related proteases

PDF

Proinsulin-transferrin recombinant fusion protein: mechanism of activation and potential application in diabetes treatment

PDF

Analysis of endocytic and trafficking pathways of potential candidates for drug delivery in HeLa and lacrimal gland acinar cells

PDF

Protein kinase Bα (AKT1) but not protein kinase Bβ (AKT2) controls pancreatic β cell growth and survival

PDF

A novel in-cell lethality-based molecular screening system using a split barnase

PDF

Developing recombinant single chain Fc-dimer fusion proteins for improved protein drug delivery

PDF

Short term high fat diet (HFD) stimulates β cell proliferation through mTOR while the prolonged treatment induces β cell senescence via p27

PDF

Steatosis induces Wnt/β-catenin pathway to stimulate proliferation of hepatic tumor initiating cells and promote liver cancer development

PDF

Molecular modeling of cyclodextrin interactions with proteins

PDF

Characterization of the interaction of nucleotide-binding oligomerization domain, leucine-rich repeat and pyrin domain-containing protein 12 (Nlrp12) with hematopoietic cell kinase (Hck)

PDF

Flipping the switch on protein activity activity: elastin-like polypeptides assemble into cell switches and vesicles

PDF

Cell penetrating peptide-based drug delivery system for targeting mildly acidic pH

PDF

Towards DNA-directed assembly of pMHC multimers for detection of low-affinity T cells

PDF

Development of novel small molecules targeting mitochondrial and oxidative stress signaling pathways for pancreatic cancer therapy

PDF

Mechanism study of proinsulin-transferrin fusion protein as a long-acting and liver-targeting insulin prodrug

PDF

PTEN loss antagonizes aging through promoting regeneration and prevents oxidative stress induced cell death

PDF

Structure-based computational analysis and prediction of TCR CDR3 loops in the TCR-peptide-MHC complex using solvation parameters and peptide molecular dynamics.

PDF

Development of a high-throughput screening assay to study mPD-1/mPD-L1 interactions

PDF

Proinsulin-transferrin fusion protein as a liver targeting insulin therapeutic in Type 1 diabetes treatment

Asset Metadata

Creator Bi, Tao (author)

Core Title Using cyclotides as a bioscaffold to target intracellular protein-protein interactions

School School of Pharmacy

Degree Doctor of Philosophy

Degree Program Molecular Pharmacology and Toxicology

Publication Date 11/14/2017

Defense Date 10/06/2017

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag cyclotide,high throughput screening,OAI-PMH Harvest,p53-HDM2,protein-protein interaction

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Camarero, Julio (committee chair), Okamoto, Curtis (committee member), Stiles, Bangyan (committee member)

Creator Email taobi@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-454465

Unique identifier UC11263697

Identifier etd-BiTao-5907.pdf (filename),usctheses-c40-454465 (legacy record id)

Legacy Identifier etd-BiTao-5907.pdf

Dmrecord 454465

Document Type Dissertation

Rights Bi, Tao

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA