disclaimer - seoul national...
TRANSCRIPT
저 시-비 리- 경 지 2.0 한민
는 아래 조건 르는 경 에 한하여 게
l 저 물 복제, 포, 전송, 전시, 공연 송할 수 습니다.
다 과 같 조건 라야 합니다:
l 하는, 저 물 나 포 경 , 저 물에 적 된 허락조건 명확하게 나타내어야 합니다.
l 저 터 허가를 면 러한 조건들 적 되지 않습니다.
저 에 른 리는 내 에 하여 향 지 않습니다.
것 허락규약(Legal Code) 해하 쉽게 약한 것 니다.
Disclaimer
저 시. 하는 원저 를 시하여야 합니다.
비 리. 하는 저 물 리 목적 할 수 없습니다.
경 지. 하는 저 물 개 , 형 또는 가공할 수 없습니다.
공학 사 학 논
Utilization of HSP90 chaperone
machinery for drug screening and
enhancement of HSP90 client protein
solubility
HSP90 샤페 시스 용한 약 스크리닝과
HSP90 클라이언트 단 질 가용 증 연구
2016 8 월
울 학 학원
학생 공학부
태
Utilization of HSP90 chaperone
machinery for drug screening and
enhancement of HSP90 client protein
solubility
by
Tae-Joon Hong
Advisor : Professor Ji-Sook Hahn, Ph.D.
Submitted in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy
in Seoul National University
August, 2016
School of Chemical and Biological Engineering
Graduate School
Seoul National University
i
ABSTRACT
Utilization of HSP90 chaperone machinery for
drug screening and enhancement of
HSP90 client protein solubility
Tae-Joon Hong
School of Chemical and Biological Engineering
The Graduate School
Seoul National University
Hsp90 chaperone machinery performs crucial functions in the folding, activation,
and stabilization of a wide range of proteins called Hsp90 ‘client’ proteins. As
Hsp90 client proteins include various oncoproteins, diverse classes of Hsp90
inhibitors have been developed intensively as anti-cancer drugs. By utilizing both
virtual and experimental screening, two Hsp90 inhibitors with novel scaffolds
could be identified and characterized in this work. The virtual screening was
performed with modified AutoDock program and the binding mode of each
compound was visualized by molecular dynamics simulations. The two compounds
ii
inhibited the ATPase activity of yeast Hsp90 in vitro, and showed antiproliferative
effects on MCF-7 cells. One compound had pyrimidine-2,4,6-trione scaffold, while
the other retained 4H-1,2,4-triazole-3-thiol. Together with the two identified
compounds, additional compounds that share these scaffolds were selected from
the library and tested for their effects on Her2, one of the oncogenic Hsp90 client
proteins. The result demonstrated that most of these compounds induced Her2
degradation in MCF-7 cells, confirming the Hsp90 inhibitory property of the newly
identified scaffolds. Also, by the successful identification of the Hsp90 inhibitors,
the effectiveness of virtual screening for drug discovery has been demonstrated.
Hsp90 chaperone machinery operates through dynamic interactions of Hsp90
with its cochaperones that regulate and assist the functions of Hsp90. Like Hsp90,
several Hsp90 cochaperones such as p23, Cdc37, SGT1, FKBP51, and PP5 have
been implicated in cancer. PP5 belongs to the PPP phosphatase family whose
members share significantly similar catalytic domain structure. As the known,
natural inhibitors of PP5 bind to the catalytic domains of PPP family phosphatases,
they exhibit broad specificity over the phosphatases of PPP family. In an attempt to
identify PP5 inhibitors that act through novel, allosteric mechanisms, drug
screening was performed with the compounds from the library of
pharmacologically active compounds (LOPAC) 1280. As a result, three
compounds were selected as PP5 inhibitors. Unlike other inhibitors, one of these
inhibitors, Ro 90-7501 exhibited TPR-dependent inhibition in the assays performed
with full-length PP5 and the deletion constructs of PP5 which lack either or both of
iii
the regulatory regions, the TPR domain and the C-terminal 13 amino acids. As the
TPR domain is a unique feature of PP5 among the PPP family phosphatases, the
TPR-dependent inhibitor might be a useful tool for both the researches of PP5 and
the drug discovery of PP5-specific inhibitors.
Hsp90 chaperone machinery promotes the folding, activation and stabilization
of its client proteins. NOD1, one of the Hsp90 client protein, is an important
intracellular sensor of innate immunity which is related to a number of
inflammatory diseases. Molecular level researches of NOD1 have been greatly
retarded because expression and purification of soluble, active NOD1 protein have
proven extremely difficult. For the soluble expression of the NOD1-LRR domain
in E. coli, Hsp90 chaperone machinery was utilized. It has been reported that
Hsp90 cochaperone SGT1 was necessary for the activation of NOD1, while Hsp90
was required for the stability of NOD1 in human cell line. Based on these
observations, SGT1 was chosen as the fusion partner of NOD1-LRR. The fusion
protein of human SGT1 and NOD1-LRR exhibited enhanced solubility, but it
suffered severe degradation and low expression level. To overcome this drawback,
Hsp90s from yeast, E. coli, and mouse were tested as coexpression partners. The
coexpression of Hsp90s enhanced the stability of the fusion protein to some extent,
in accordance with the previous report. The comparison between E. coli HtpG,
yeast Hsp82 and Hsc82, mouse Hsp90α demonstrated that the coexpression of
mouse Hsp90α was most effective for stabilizing SGT1NOD1LRR fusion protein.
iv
Keywords : Drug screening, Inhibitor, Fusion protein, Coexpression,
Hsp90, PP5, SGT1, NOD1
Student Number : 2007-30853
v
CONTENTS
Abstract ................................................................................................. i
Contents................................................................................................ v
List of Figures ................................................................................... viii
List of Tables ........................................................................................ x
List of Abbreviations .......................................................................... xi
Chapter 1. Research background and objective ................................. 1
Chapter 2. Literature review ............................................................... 4
2.1. Hsp90 chaperone machinery ............................................................... 5
2.1.1. Hsp90 molecular chaperone ....................................................... 5
2.1.2. Hsp90 as an anti-cancer target .................................................... 6
2.1.3. Hsp90 chaperone cycle .............................................................. 9
2.1.4. Hsp90 cochaperones ................................................................ 12
2.1.5. Asymmetry of Hsp90 chaperone machinery ............................. 13
2.2. Protein phosphatase 5 (PP5) ............................................................. 17
2.2.1. Overview of protein phosphatase 5 .......................................... 17
2.2.2. The structure of PP5 ................................................................ 17
2.2.3. Cellular functions and diseases related to PP5 .......................... 21
2.2.4. Regulators of PP5 phosphatase acitivity ................................... 24
2.3. Nucleotide-binding oligomerization domain-containing 1 (NOD1) ... 27
2.3.1. NOD-like receptors (NLRs) ..................................................... 27
2.3.2. NOD1 protein purification ....................................................... 31
2.3.3. Overview of SGT1 ................................................................... 33
2.3.4. SGT1-Hsp90 complex and NOD1 ! 책갈피가 어 있지 않습니다.
vi
Chapter 3. Materials and methods .....................................................38
3.1. Plasmid DNA vectors for protein expression ..................................... 39
3.2. Protein expression and purification ................................................... 43
3.3. Structure-based virtual screening with modified AutoDock ............... 44
3.4. Molecular dynamics for Hsp90-inhibitor comlexes ........................... 47
3.5. In vitro Hsp90 ATPase assay ............................................................. 48
3.6. Sulforhodamine B (SRB) assay ........................................................ 49
3.7. Cell-based Her2 degradation assay ................................................... 50
3.8. PP5 phosphatase assay ..................................................................... 51
3.9. SGT1NOD1LRR fusion protein expression ...................................... 52
Chapter 4. Virtual and experimental screening to identify new
Hsp90 inhibitor scaffolds .............................................................54
4.1. Introduction...................................................................................... 55
4.2. Virtual screening with modified AutoDock program ......................... 57
4.3. Inhibitory and antiproliferative effects of the compounds .................. 59
4.4. Molecular dynamics simulations of the compounds .......................... 61
4.5. Her2 degradation assay with compounds sharing the scaffolds .......... 65
4.6. Conclusions...................................................................................... 67
Chapter 5. Screening of PP5 inhibitors that act through different
mechanisms ..................................................................................69
5.1. Introduction...................................................................................... 70
5.2. Screening of LOPAC1280 by PP5 phosphatase assay........................ 70
5.3. Confirmation of the primary screening ............................................. 71
5.4. Mechanism study of the screened compounds ................................... 71
5.5. Conclusions...................................................................................... 81
vii
Chapter 6. Enhancing the solubility of NOD1-LRR in E. coli...........82
6.1. Introduction...................................................................................... 83
6.2. Expression of full-length NOD1 protein in E. coli............................. 85
6.3. Expression of NOD1-LRR domain in E. coli ! 책갈피가 어 있지 않습니다.
6.4. Expression of SGT1NOD1LRR fusion protein in E. coli .................. 88
6.5. Conclusions.................................................................................... 102
Chapter 7. Overall discussion and recommendations ..................... 105
Bibliography ...................................................................................... 111
Abstract in Korean ......... ! 책갈피가 어 있지 않습니다.
viii
LIST OF FIGURES
Figure 2.1 Hsp90 chaperone cycle and six hallmarks of cancer ···················· 8
Figure 2.2 Binding sites of Hsp90 cochaperones on Hsp90························15
Figure 2.3 Crystal structure of PP5 ····················································20
Figure 2.4 Structure of PP5 and its auto-regulatory mechanism ··················22
Figure 2.5 Structure of NOD1 and its activation mechanism model ·············30
Figure 2.6 Structure of SGT1 and its function as an adaptor protein ·············37
Figure 4.1 The strategy applied for the discovery of novel Hsp90 inhibitors ···58
Figure 4.2 Structure of compound 1b and its MD trajectory snapshot ···········63
Figure 4.3 Structure of compound 2b and its MD trajectory snapshot ···········64
Figure 4.4 Six compounds that share the scaffolds of 1b and 2b ··················66
Figure 4.5 Her2 degradation of the compounds that share the scaffolds ·········68
Figure 5.1 Primary screening of the compounds from LOPAC1280 ·············75
Figure 5.2 Secondary confirmation assay of the screened compounds ···········76
Figure 5.3 PP5 truncation constructs for mechanism study ························77
Figure 5.4 Application of known inhibitor and activator to PP5 constructs ·····78
Figure 5.5 Application of the selected inhibitors to PP5 constructs ··············79
Figure 5.6 Ro 90-7501 exhibits novel PP5 inhibitory mechanism ················80
Figure 6.1 NOD1 full-length protein expression with chaperones ················87
Figure 6.2 Optimization of IPTG concentration for soluble NOD1 expression ·93
Figure 6.3 Solubility of NOD1-LRR domain coexpressed with chaperones ····94
Figure 6.4 Solubility of SGT1NOD1LRR protein with or without Hsp90 ······95
ix
Figure 6.5 Expression of SGT1NOD1LRR with or without mouse Hsp90 ·····96
Figure 6.6 Effect of antibiotic markers on SGT1NOD1LRR expression ········97
Figure 6.7 Effect of E. coli strains on SGT1NOD1LRR expression ··············98
Figure 6.8 Effect of rare codon tRNAs on SGT1NOD1LRR expression ········99
Figure 6.9 Effect of LRR constructs on SGT1NOD1LRR expression ········· 100
Figure 6.10 Effect of Hsc82 and HtpG on SGT1NOD1LRR expression ······ 101
Figure 6.11 Effect of mHsp90 and Hsp82 on SGT1NOD1LRR expression ··· 103
Figure 6.12 Effect comparison of E. coli HtpG and mouse Hsp90 on
SGT1NOD1LRR expression ··········································· 104
x
LIST OF TABLES
Table 2.1 Hsp90 inhibitors in clinical trials ··········································· 11
Table 2.2 Hsp90 cochaperones and their functions ··································16
Table 2.3 Natural compounds that inhibit PPP-family phosphatases ·············26
Table 2.4 Recombinant protein expression systems ·································34
Table 3.1 Primers for expression vector construction ·······························41
Table 4.1 Inhibitory and antiproliferative effects of 1b and 2b ····················61
xi
LIST OF ABBREVIATIONS
ATP adenosine triphosphate
CARD caspase recruitment domain
CS CHORD-containing protein and SGT1
DAMP danger-associated molecular pattern
DDW distilled deionized water
DMEM Dulbecco's modified eagle medium
DMSO dimethyl sulfoxide
ECL enhanced chemiluminescence
EDTA ethylenediaminetetraacetic acid
FBS fetal bovine serum
FPLC fast protein liquid chromatography
GR glucocorticoid receptor
His histidine
HSP90 heat shock protein 90 kDa
HTS high-throughput screening
IB immunoblotting
IEC ion exchange chromatography
iE-DAP D-glutamyl-meso-diaminopimelic acid
IMAC immobilized-metal affinity chromatography
IPTG isopropyl β-D-1-thiogalactopyranoside
LB Luria-Bertani
LOPAC library of pharmacologically active compounds
LRR leucine-rich repeats
MD molecular dynamics
xii
MDP muramyl dipeptide
NLR NOD-like receptor
NOD nucleotide-binding oligomerization domain
OD optical density
ORF open reading frame
PAGE polyacrylamide gel electrophoresis
PAMP pathogen-associated molecular pattern
PBS phosphate-buffered saline
PCR polymerase chain reaction
PD phosphatase domain
PDB protein data bank
PMSF phenylmethylsulfonyl fluoride
pNPP 4-nitrophenol phosphate
PP5 protein phosphatase 5
PPAR peroxisome proliferator-activated receptor
PPP phosphoprotein phosphatase
PPT1 protein phosphatase T1
Ptase phosphatase
RT room temperature
SDS sodium dodecyl sulfate
SGS SGT1-specific domain
SGT1 suppressor of G2 allele of skp1
SRB sulforhodamine B
TCA trichloroacetic acid
TLR toll-like receptor
TPR tetratricopeptide repeat
Chapter 1.
Research background and objective
2
Heat shock protein 90 kDa (Hsp90) serves as a molecular chaperone that plays
important roles in the folding, activation and stabilization of its ‘client’ proteins. As
many Hsp90 client proteins are in the categories of oncoproteins that support all of
the six hallmarks of cancer [1], Hsp90 has been highlighted as a potential
therapeutic target for anti-cancer drug development. In the beginning, natural
products such as geldanamycin and radicicol were proven to inhibit Hsp90, but
these compounds had drawbacks including toxicity, low solubility and instability
[2]. The structural modifications were introduced to these compounds to improve
their properties, which were successful to some extent. But there were still
drawbacks and limits that resulted from the intrinsic properties of the structural
scaffolds of these compounds. The efforts to find Hsp90 inhibitors with totally new
scaffolds have been made to overcome the intrinsic limitations of the natural
compounds, which have been successful considering the fact that most of the
Hsp90 inhibitors that are currently under development or clinical trials are based on
scaffolds that are distinct from the natural products geldanamycin and radicicol [2,
3]. In the same line, the objective of the first part of this research is to identify new
Hsp90 inhibitor scaffolds which might serve as lead compounds for anti-cancer
drug development. The virtual screening with modified AutoDock program was
applied for efficient screening process and to prove its usefulness in drug discovery.
Hsp90 cochaperones have proven to play crucial roles for the functionality of
Hsp90 chaperone machinery. One of the Hsp90 cochaperone, protein phosphatase
5 (PP5) has been regarded as a potential drug target for its implications in
3
Alzheimer’s disease, cancer and obesity [4-16]. Although there are natural
phosphatase inhibitors that potently suppress PP5 activity, these compounds lack
the specificity toward PP5 [17]. The objective of the second part of this research is
to discover PP5 inhibitors that might downregulate the phosphatase activity of PP5
specifically, through the mechanisms of action that are different from that of the
known PP5 inhibitors. Also, as the compounds subjected to the screening assays
were from the library of pharmacologically active compounds (LOPAC) 1280, the
screened PP5 inhibitors can be candidates for drug repositioning [18, 19].
One of the Hsp90 client proteins, Nucleotide binding oligomerization domain
containing 1 (NOD1) is an intracellular sensor of innate immunity that detects
peptidoglycan fragments of bacteria and activates proinflammatory signaling
pathways such as NF-κB. NOD1 is implicated in inflammatory diseases such as
asthma, inflammatory bowel diseases, atopic eczema, and cancer [20, 21].
Although NOD1 has been spotlighted by its significant roles in health and diseases,
molecular level researches of NOD1 which require sufficient amount of purified
NOD1 protein are limited because NOD1 has been known notoriously difficult to
express and purify [22]. The aim of the third part of this study is to express the
LRR domain of NOD1 as a soluble form in E. coli with the aid of Hsp90 chaperone
machinery. Hsp90 cochaperone SGT1 and Hsp90 were chosen as a fusion partner
and a coexpression partner of NOD1-LRR, respectively, since it has been reported
that SGT1 was necessary for NOD1 activation, while Hsp90 was required for the
stability of NOD1 in human cell line [23-26].
Chapter 2.
Literature review
5
2.1. Hsp90 chaperone machinery
2.1.1. Hsp90 molecular chaperone
A living cell requires numerous proteins to survive and to proliferate. Some
proteins are simultaneously folded and activated after translation, but others require
specific proteins, called molecular chaperones, that aid their proper folding and
subsequent activation. Chaperones are also required to stabilize cellular proteins
under certain stressful conditions that denature the proteins. There are various
classes of molecular chaperones with diverse sizes, oligomeric states, subcellular
localizations and cellular functions [27, 28].
Hsp90 (heat shock protein 90 kDa) is an abundant, conserved and ubiquitous
molecular chaperone, which is involved in a wide spectrum of crucial cellular
functions. It comprises 1~2 % of total cellular proteins in most cells, revealing its
significance and versatility in cellular functions [29]. Mammalian cells have two
major isoforms Hsp90α, and Hsp90β in the cytoplasm. Hsp90α is inducible upon
stressful conditions, while Hsp90β is expressed constitutively. Also, in other
mammalian cellular compartments, Hsp90 analogues exist. These are TRAP-1 in
the mitochondria, and Grp94 in the endoplasmic reticulum [30]. It has been shown
that Hsp90 even exist on the outside of the cell, referred as extracellular Hsp90
[31]. In yeast, inducible form of Hsp90 exist as Hsp82, while the constitutive form
exist as Hsc82. In bacteria such as Escherichia coli, Hsp90 is called HtpG. While
Hsp90 is essential for eukaryotic cells, bacterial Hsp90 is not, and its major
6
function is still elusive [30]. In archaea, no Hsp90 gene has been identified [32].
Hsp90 consists of three domains; the N-terminal, middle, and the C-terminal
domain [29, 33]. The N-terminal domain has an ATP-binding pocket and dimerizes
upon ATP-binding [34]. This domain is connected to the middle domain with a
linker which is absent in HtpG and TRAP1 [30]. Middle domain is known to
possess hydrophobic surface that might bind to the unfolded protein which is
exposing hydrophobic amino acid residues [35]. The C-terminal domain forms a
dimer and also contains MEEVD peptide sequence that binds to the TPR domain of
the TPR-containing proteins (Figure 2.1A)[36]. The Hsp90 homodimer proceeds
through a cycle of open and closed conformational change induced by the binding
and hydrolysis of ATP, depicted as a ‘molecular clamp’ (Figure 2.1B)[34].
A subset of proteins that are aided by Hsp90 molecular chaperone are called
the ‘client’ proteins of Hsp90. These include a wide range of proteins with various
cellular functions. A genomic and proteomic study which used genome-wide yeast
two-hybrid and microarray-based chemical-genetic screens identified 198 putative
physical interactions and 451 putative genetic interactions related to Hsp90 [37].
An updated list of Hsp90 client proteins is maintained by the Picard laboratory [38].
2.1.2. Hsp90 as an anti-cancer target
The client proteins of Hsp90 contain many oncogenic proteins including
Her2/ErbB2, Akt, Raf-1, Cdk4, Hif-1α, MET, hTERT, mutant p53, surviving, and
IP6K2. These proteins are involved in all of the six hallmarks of cancer that are
7
critical in the development and progression of cancer. Degradation of these
oncoproteins by ubiquitin-proteasome pathway is achieved by the inhibition of
Hsp90, which makes Hsp90 a promising target for the development of anti-cancer
agent (Figure 2.1B). Hsp90 inhibitors such as 17-AAG, BIIB021, NVP-
AUY922/VER-52296 have entered clinical trials and other novel classes of Hsp90
inhibitors are under extensive development (Table 2.1 [2]).
8
Figure 2.1 Hsp90 chaperone cycle and six hallmarks of cancer
A. The structure of Hsp90 (red: lid segment, purple: N-terminal stand).
B. The Hsp90 chaperone cycle and the six hallmarks of cancer. ATP binding induces
the closure of the lid segment (red) over the mouth of the ATP binding pocket (I1)
and swapping of the N-terminal stand of each Hsp90 monomer (I2). After the
formation of the compact and twisted closed conformation (Closed), ATP
hydrolysis restores the open conformation of Hsp90 (Open) [30].
9
2.1.3. Hsp90 chaperone cycle
Hsp90 forms a dimer through its C-terminal domain and its N-terminal
dimerization is driven by the binding of ATP into the ATP-binding pocket of the
N-terminal domain. High dependence on the nucleotide binding for the
conformational change of Hsp90 dimer was shown by the studies of E. coli HtpG
[39, 40]. As Hsp90 is conserved throughout most species, it was thought that this
could be a conserved mechanism of Hsp90 chaperone cycle. But recent studies
with single-molecule fluorescence resonance energy transfer (smFRET) revealed
that eukaryotic yeast Hsp90 adopts quite different and more complicated
mechanism compared to E. coli HtpG. It was shown that the assumption that the C-
terminal domain of Hsp90 forms a stable dimer was not true for yeast Hsp90. Not
only the N-terminal, but also the C-terminal dimer opens and closes in the yeast
Hsp90 dimer [41]. Also, nucleotides could bind to both open and closed states
without driving specific conformational changes [42, 43]. While the mechanism of
prokaryotic E. coli HtpG chaperone cycle is referred as a ‘ratchet’, the mechanism
of eukaryotic yeast Hsp90 is referred as a ‘random (or thermal) fluctuation’ [44].
The comparison of these two mechanisms showed that eukaryotic Hsp90 adopts
more diverse conformations than prokaryotic Hsp90 [44]. Also, it was shown that
the conformational changes of Hsp90 dimer can be regulated by its cochaperones,
Sti1 (inhibition) and Aha1 (acceleration) [45]. More recently, a four-color smFRET
experiment demonstrated that eukaryotic yeast Hsp90 requires not only ATP, but
also the cochaperone p23 for directional progression of the chaperone cycle [46]. It
10
can be speculated that the Hsp90 of prokaryotes such as E. coli which do not
possess Hsp90 cochaperones, adopts simple and strict, nucleotide-dependent
conformations which are sufficient for their relatively simple cellular functions,
while the Hsp90 of eukaryotes adopts more complex and less nucleotide-dependent,
flexible conformations to support more diverse cellular functions [44]. And Hsp90
cochaperones of the eukaryotes enables the progression of the chaperone cycle,
adding even more diversity and functional flexibility to the Hsp90 chaperone
machinery.
11
Table 2.1 Hsp90 inhibitors in clinical trials
Inhibitor Class Phase Route Company
Retaspimycin hydrochloride (IPI-504)
Hydroquinone derivative II/III i.v Infinity
CNF2024/BIIB021 Purine-scaffold agent I
Terminated Oral Biogen Idec
MPC-3100 Purine scaffold agent I/II
Completed Oral Myrexix Inc.
Debio 0932 (CUDC-305) Purine-scaffold derivative I/II
Recruiting Oral Debiopharm Group
NVP-AUY922 Resorcinol-Isoxazole I/II i.v Novartis
SNX-5422 Indazol-4-one I Completed
Oral Serenex
STA-9090 Resorcinol-Triazole I/II Active
i.v Synta Pharmaceuticals
KW-2478 Resorsinol containing synthetic agent
I/II Recruiting
i.v. Kyowa Hakko Kirin
AT13387 Resorsinol containing synthetic agent
I/II Recruiting
Oral, i.v.
Astex Therapeutics
DS-2248 Other chemotype I Oral Daiichi Sankyo Inc.
XL-888 Other chemotype I Terminated
Oral Exelixis
PU-H71 Purine-scaffold agent I Recruiting
PET imaging
i.v. Memorial Sloan-Kettering Cancer Center
12
2.1.4. Hsp90 cochaperones
Hsp90 cochaperones are the proteins that regulate and support the chaperone
functions of Hsp90. They bind to specific regions of the Hsp90 dimer to modulate
its activity. There are more than 20 cochaperones identified in eukaryotic cells [32].
Hsp90 cochaperones dynamically associate with the Hsp90 dimer during the
progression of the chaperone cycle. The cochaperones regulate the ATPase activity
of Hsp90, induce or stabilize certain conformations of Hsp90 dimer, recruit certain
subsets of client proteins [32].
Some cochaperones such as Hop/Sti1, Cdc37, p23/Sba1, Aha1, SGT1 have
been studied intensively, while other cochaperones such as Tah1p, Pih1p are
identified relatively recently by the abovementioned genomic and proteomic study
that revealed genetic and physical interactions of Hsp90 [37].
Hsp organizing protein (Hop) and its yeast homologue, Stress-inducible
protein 1 (Sti1) has one Hsp90 binding site (TPR2A) and two Hsp70 binding sites
(TPR1, TPR2B) [47]. Through these binding sites, Hop/Sti1 binds simultaneously
to Hsp70 and Hsp90, coupling the two chaperones and enabling efficient client
protein handover from Hsp70 to Hsp90 [47-49]. Hop/Sti1 binds to the C-terminal
MEEVD motif of Hsp90 and inhibits the ATPase activity of Hsp90 by preventing
the N-terminal dimerization [50].
Cdc37 recruits mainly kinases as client proteins to Hsp90. The N-terminal
domain of Cdc37 interacts with the kinase, while the C-terminal domain associates
13
with the lid segment of Hsp90 in its open conformation, which inhibits the N-
terminal dimerization of the Hsp90 dimer and the ATPase activity [51]. In this way,
Cdc37 arrests the progression of the Hsp90 chaperone cycle and might provide
enough time for the loading of the client kinase proteins [30].
p23/Sba1 stabilizes the closed conformation of Hsp90 and facilitates the
maturation of client proteins, while inhibiting the ATPase activity of Hsp90 [52].
The binding sites of p23/Sba1 range over both N-terminal domains of the closed
Hsp90 dimer, and one middle domain of a monomer [30, 53]. Interestingly,
p23/Sba1 is known to possess its own chaperone activity [54].
Aha1, unlike the cochaperones that inhibit the ATPase activity of Hsp90, is the
most notable activator of the ATPase activity of Hsp90. Aha1 binds to the middle
domain of Hsp90, and also interacts with the N-terminal domain [55]. By these
interactions, Aha1 can induce the closed conformation of Hsp90 dimer, even
without nucleotide-binding to the N-terminal domain of Hsp90 [54].
The binding sites of Hsp90 cochaperones lie on all three domains of Hsp90,
demonstrating the variety of their structures and functions (Figure 2.2, Table 2.2
[30, 54]).
2.1.5. Asymmetry of Hsp90 chaperone machinery
As Hsp90 forms a dimer through its C-terminal domains and the N-terminal
domains also dimerize as the chaperone cycle proceeds, it could be considered that
the Hsp90 chaperone machinery works in a symmetrical manner. But the
14
asymmetric nature of Hsp90 chaperone system and Hsp90 itself have been revealed
by intensive researches. Recently, the asymmetric structure of the closed state
mitochondrial Hsp90 has been reported [56]. Also, the cochaperone binding to
Hsp90 dimer [57, 58] and the client protein binding in complex with a cochaperone
[59], post-translational modification such as SUMOylation of Hsp90 is
demonstrated to be asymmetric [60]. Another interesting asymmetric feature of
Hsp90 dimer is the ATP hydrolysis which may occur in only one subunit [61].
15
Figure 2.2 Binding sites of Hsp90 cochaperones on Hsp90
The binding sites of each Hsp90 cochaperone has been shown based on the structural
informations from PDB (Hsp90 and p23: 2CG9, Sgt1: 2JKI, Cdc37: 2K5B, Aha1:
1USV, PP5: 2BUG). The images were processed with Pymol.
16
Table 2.2 Hsp90 cochaperones and their functions
ef. [30, 54]
Mammal Yeast Function
Cochaperones with TPR
Hop Sti1 Scaffold for Hsp90/Hsp70 interaction, inhibition of Hsp90 ATPase
Fkbp52 - Peptidy-prolyl-isomerase, chaperone
Fkbp51 - Peptidy-prolyl-isomerase, chaperone
Cyp40 Cpr6/Cpr7 Peptidy-prolyl-isomerase, chaperone
AIP - Complex with AhR, PPARα, Hbx
CHIP - Ubiquitin ligase, tagging protein for degradation
PP5 Ppt1 Phosphatase
Tpr2 - Retrograde transfer of substrates from Hsp90 to Hsp70
SGT1 SGT1 Activate NLR receptors in plant/animal innate immunity, chaperone
Unc45 She4 Assembly of myosin fibers
Ttc4 Cns1 Nuclear transport protein
Tom70 Tom70p Mitochondrial protein import
Tah1 Tah1 Forms complex with Pih1 and Hsp90
Cochaperones without TPR
Aha1 Aha1 Stimulates ATPase activity, induces conformation changes in Hsp90
p23 Sba1 Client protein maturation, inhibition of Hsp90 ATPase, chaperone
Cdc37 Cdc37 Kinase-specific co-chaperone, inhibition of Hsp90 ATPase, chaperone
Chp-1/Melusin - Centrosome duplication, chaperone / cardiac protective
NudC NudC CS domain, Multiple roles in mitosis and cytokinesis, chaperone
17
2.2. Protein phosphatase 5 (PP5)
2.2.1. Overview of protein phosphatase 5
The phosphorylation of proteins is a reversible post-translational modification that
plays an important role in signaling and regulating various cellular responses. Over
99 % of the phosphorylation takes place on serine or threonine residues. Thus,
serine/threonine phosphatases are crucial for the dephosphorylation of phosphory-
lated protein substrates [62].
Protein phosphatase 5 (PP5) is a member of the PPP family of serine/threonine
phosphatases which has been identified relatively late compared to the other
members [63]. This was because PP5 is less abundant than other PP5 family
members and its basal activity is extraordinarily low, which made PP5
undiscovered for decades [62, 63]. PP5 is ubiquitous in all eukaryotic cells, from
yeast to humans. Its localization is both nuclear and cytoplasmic in the cell, and the
tissue expression level is high in the brain [62].
2.2.2. The structure of PP5
Many serine/threonine phosphatases consist of catalytic subunits and regulatory
subunits which control the localization, activity, and specificity of the phosphatase
[62]. PP5 is a unique member of the PPP family because it contains catalytic,
regulatory and localization targeting functions in a single protein [62].
The C-terminal catalytic domain of PP5 is highly similar to the catalytic
18
subunits of other PPP family members such as, PP1, PP2A and PP2B [62]. Due to
this similarity, PP5 is inhibited by low concentrations of okadaic acid, like PP1 and
PP2A. The N-terminal tetratricopeptide repeat (TPR) domain has both interaction
targeting and regulatory functions [62]. TPR domain of PP5 consists of a series of
antiparallel amphipathic α-helices which pack together through hydrophobic
interactions [64]. Generally, TPR domains mediate protein-protein interactions and
the TPR domain of PP5 also mediates the interactions of PP5. It belongs to a
family of Hsp90-binding TPR domains and PP5 binds to the C-terminal MEEVD
sequence of Hsp90 through its TPR domain [65]. At the C-terminal of PP5, next to
the phosphatase domain, a subdomain of 20 amino acids terminating in a short two-
turn α-helix (αJ) resides [65]. With this C-terminal subdomain, TPR domain forms
the auto-inhibited conformation of PP5 [65] (Figure 2.3). Proteolytic cleavage of
the auto-inhibitory TPR domain by trypsin stimulated the phosphatase activity of
PP5 toward its substrates, casein and myelin basic protein. The increase of the PP5
activity was also observed by the polyunsaturated fatty acids, such as arachidonic
acid, linoleic acid, and oleic acid which bind to the TPR domain [66]. Another
study reported that the removal of both TPR domain and the C-terminus 10
residues by subtilisin resulted in even greater activation than trypsin-treated PP5
[67]. It should be noted that the catalytic fragments generated by proteolysis were
no more stimulated by lipid [67, 68]. These results indicate that the C-terminal
subdomain and the N-terminal TPR domain are involved in both the auto-inhibition
and lipid activation of PP5 [67]. The proteolytic activation of PP5 has been shown
19
to be induced by arachidonic acid or nocodazole, implying an in vivo regulation
mechanism of PP5 [69]. Although both Hsp90 and arachidonic acid stimulate PP5
by binding to the auto-inhibitory TPR domain, the mechanisms differ from each
other. Hsp90 activates PP5 by interfering TPR-phosphatase domain interaction
through its C-terminal MEEVD sequence, thereby increases substrate accessibility
to the constitutively active phosphatase domain [65]. In contrast, arachidonic acid
induces an alternate conformation of the TPR domain that destabilizes the TPR-
phosphatase domain interface [65] (Figure 2.4).
20
Figure 2.3 Crystal structure of PP5
The pale blue region represents the N-terminal TPR domain, while the purple
region is the phosphatase domain of PP5. The deep blue region is the C-terminal
subdomain, including the αJ helix. Metal ions (Mn2+) are shown as deep purple
spheres [PDB: 1WAO]. The figure was processed with Pymol.
21
2.2.3. Cellular functions and diseases related to PP5
PP5 is involved in a wide spectrum of cellular functions. These include growth and
differentiation by MAPK pathway, cell cycle arrest and DNA damage repair
through p53 and ATM/ATR pathways, ion channel regulation by the membrane
receptor for atrial natriuretic peptide, heat shock response mediated by heat shock
transcription factor, and steroid receptor signaling by regulating the
phosphorylation state of glucocorticoid receptor (GR) [63]. Even though PP5
participates in these diverse signaling pathways, the development of a viable PP5-
deficient mice [70] suggests that PP5 is not essential [63].
The sequence of PP5 was noted for the TPR domain which was closely related
to that of the Hsp90-binding proteins in steroid receptor heterocomplexes [62]. It
has been demonstrated that PP5 binds to Hsp90 through its TPR domain and
participates in the GR-Hsp90 heterocomplexes as a major component [71]. In the
GR-Hsp90 hetero complex, several TPR-containing immunophilins and PP5 have
been shown to control the hormone binding function of GR [72].
In Saccharomyces cerevisiae, deletion of the PP5 homologue, PPT1 [73]
showed no obvious phenotypical change [74]. Later, it has been shown that PPT1
dephosphorylates Hsp90 [75], Hsf1 [76] and is related to the defective stress
response upon ethanol stress [76, 77].
22
Figure 2.4 Structure of PP5 and its auto-regulatory mechanism
A. The domain structure of PP5. PP5 consists of N-terminal TPR domain,
phosphatase domain, and C-terminal subdomain. The TPR domain mediates
PP5-Hsp90 interaction.
B. The auto-regulatory mechanism of PP5. The TPR domain and the C-terminal
subdomain participate in both auto-inhibition and activation response of
PP5. Okadaic acid is a representative phosphatase inhibitor which binds to
the catalytic region of the phosphatase domain. On the other hand,
arachidonic acid, an activator of PP5, binds to the auto-inhibitory TPR
domain and alters its secondary structure, which releases the phosphatase
domain and activates PP5.
23
In mammals, PP5 has been implicated in cancers such as hepatocellular
carcinoma [15], colorectal cancer [7] and gynecological cancers such as breast
cancer [8, 9, 11] and ovarian cancer [6]. The knockdown of PP5 in these cancers
resulted in cell cycle arrest, inhibiting the proliferation of cancer cells [16].
PP5 is highly expressed in the mammalian brain, implying its neuronal
function. In the neocortex of Alzheimer’s disease, the protein level of PP5 was
observed to be decreased by about 20 %, and PP5 has been shown to
dephosphorylate phospho-tau which is abnormally hyperphosphorylated in
Alzheimer’s disease brain [4, 5]. The neuroprotective role of PP5 against amyloid-
β toxicity has been also demonstrated [78]. These results suggest that PP5 may be a
potential therapeutic target for neurodegenerative disorders such as Alzheimer’s
disease.
In the study of a high-fat diet-fed mouse model, PP5-deficient mice gained less
weight and less visceral fat compared to wild-type mice [12]. The molecular basis
underlying these observations has been suggested that the phosphorylation states of
glucocorticoid receptor-α (GRα) and peroxisome proliferator-activated receptor-γ
(PPARγ) which regulate adipogenesis by adjusting the balance between lipolysis
and lipogenesis, are reciprocally controlled by PP5 [14]. Another mouse model
study demonstrated that PP5-inactivated mice exhibited decreased adipose tissue,
indicating that PP5 has a crucial role in adipose tissue differentiation [13]. These
researches suggest that PP5 can be a potential drug target for obesity treatment.
On the other hand, there are also efforts to characterize and evaluate insect and
24
protozoan PP5 as targets for pesticides [79] and antimalarial drugs [80],
respectively.
2.2.4. Regulators of PP5 phosphatase acitivity
Diverse classes of molecules have been suggested to regulate the phosphatase
activity of PP5. As PP5 is auto-inhibited by its own TPR domain and C-terminal
subdomain, both inhibitors and activators of PP5 have been reported.
Natural compounds from various organisms such as okadaic acid (marine
dinoflagelates, Prorocentrum sp. and Dinophysis sp.), calyculin A, dragmacidins
(marine sponges), microcystins, nodularins (cyanobacteria, Microcystis sp. and
Nodularia sp.), tautomycin, fostriecin (soil bacteria, Streptomyces sp.), and
cantharidin (blister beetles, approx. 1500 species) have proven to be potential
inhibitors of PPP family phosphatases (Table 2.3 [17]). Okadaic acid is the most
widely used, representative inhibitor in the studies of PPP family phosphatases [17].
It was revealed that okadaic acid is the causative compound of diarrhetic shellfish
poisoning and it functions as a potent inhibitor of protein phosphatases PP1 and
PP2A [17].
As PP5 is under the auto-inhibited state in normal, endeavors to identify its
physiological activators have been made. One of the most potent PP5 activators of
polyunsaturated fatty acids was arachidonic acid [66]. But the concentration of
arachidonic acid in the cell is much lower than that required to activate PP5 in vitro.
While arachidonic acid was not regarded as a physiological activator of PP5, long-
25
chain fatty acyl-CoA esters which activated PP5 at physiological concentrations,
have been suggested to be potential physiological activators of PP5 [81].
As the activity, not the protein level of PP5 was observed to be decreased in
Alzheimer’s disease neocortex [4], PP5 activators can be potential therapeutics for
this disease. High-throughput screening (HTS) researches identified PP5 activators
such as chaulmoogric acid [82] and PP5 small-molecule activators (P5SAs) [83].
Since PP5 has been implicated in cancers [6-11], PP5 inhibitors might be
candidates for potential anti-cancer therapeutics. It has been demonstrated that
curcumin inhibited PP2A and PP5, inducing the apoptosis of several types of
human cancer cell lines [10].
There are also several protein interactors of PP5 that have been demonstrated to
regulate the phosphatase activity of PP5. Rac1 GTPase has been shown to promote
PP5 translocation to the plasma membrane and stimulate PP5 [84]. In prostate
cancer cells, highly expressed caveolin-1 which is implicated in the prostate cancer
progression, interacted with and activated PP5 [85]. Other interactors such as S100
proteins [86] and Hsp70 [87] activated PP5, while KLHDC10 [88] and hCRY2 [89]
inhibited PP5. The binding of Hsp90 (MEEVD motif) or Hsp70 (IEEVD motif) to
the TPR domain of PP5 stimulated the phosphatase activity of PP5 [87].
26
Table 2.3 Natural compounds that inhibit PPP-family phosphatases
Ref. [17]
a The IC50 values provided are nanomolars and represent the amount of inhibitor needed to inhibit 50% of the activity in an assay. Because IC50 values are influenced by the concentration of inhibitor, the mode of inhibition, and the amount of both enzyme and substrate employed in an assay, the different values shown likely reflect variations in the assay conditions used by the labs reporting these values [17]. (ND: not determined)
Compound IC50a (nM)
PP1 PP2A PP2B
(calcineurin) PP4 PP5 PP7
Okadaic acid 15–50 0.1–0.3 ~4000 0.1 3.5 > 1000
Microcystin-LR 0.3–1 < 0.1–1 ~1000 0.15 1.0 > 1000
Nodularin 2.4 0.3 > 1000 ND ~4 > 1000
Calyculin A 0.4 0.25 > 1,000 0.4 3 >1,000
Tautomycin 0.23–22 0.94–32 > 1000 0.2 10 ND
Cantharidin 1,100 194 > 10,000 50 600 ND
Fostriecin 45,000-58,000 1.5–5.5 > 100,000 3.0 50,000-70,000 ND
27
2.3. Nucleotide-binding oligomerization domain-containing 1
(NOD1)
2.3.1. NOD-like receptors (NLRs)
NLRs are one class of pattern-recognition receptors, such as toll-like receptors
(TLRs). These receptors recognize pathogen-associated molecular patterns
(PAMPs) and danger-associated molecular patterns (DAMPs), then induce the
downstream signaling pathways. In contrast to TLRs which are transmembrane
receptors, NLRs are cytoplasmic receptors which are crucial for the innate immune
response.
The functions of NLRs are related to inflammasome assembly, signaling
transduction, transcription activation, and autophagy. In accordance with their
functions, NLRs are associated to various diseases related to immunity and
inflammation such as asthma [90], Crohn’s disease [91-93], atopic dermatitis [94,
95], Blau syndrome and gout. There are 22 members of NLRs in humans, and
based on their N-terminal domain, they are categorized into four subfamilies;
NLRA, NLRB, NLRC, and NLRP [96]. The NLRC subfamily consists of six
members; NLRC1 (NOD1), NLRC2 (NOD2), NLRC3 (NOD3), NLRC4 (IPAF),
NLRC5, and NLRX1. These subfamily members have characteristic N-terminal
caspase recruitment domain (CARD) domains, which recruit caspase-1 or kinases
to induce downstream signaling [20].
NOD1 and NOD2 sense conserved motifs of bacterial peptidoglycan through
28
their C-terminal leucine-rich repeats (LRR) domains and activate proinflammatory
and antimicrobial responses [97]. In humans, at least 34 LRR-containing proteins
are related to diseases [98]. The LRR domain of NOD1 recognizes γ-D-glutamyl-
meso-diaminopimelic acid (iE-DAP), a dipeptide present in a peptidoglycan which
is primarily found in Gram-negative bacteria but also in some groups of Gram-
positive bacteria, including Listeria spp. and Bacillus spp [99]. In contrast, NOD2
detects muramyl dipeptide (MDP) that is present in bacterial peptidoglycan
ubiquitously [100]. In the N-terminal, NOD1 contains a single CARD domain,
while NOD2 has tandem CARD domains. These CARD domains recruit
downstream effector proteins such as receptor-interacting protein 2 (RIP2) kinase.
In the middle, NOD1 and NOD2 contains a nucleotide-binding and oligomerization
domain (NOD) domain (also known as a NACHT domain; present in NAIP, CIITA,
HET-E, and TP1) which has ATP-binding activity [101, 102] (Figure 2.5A). Upon
ligand binding and recognition by the LRR domains, NOD proteins are thought to
undergo a conformational change which exposes the NOD domain that was
masked by the LRR domain, and this promotes the nucleotide-binding and
oligomerization of the NOD proteins. It has been suggested by an induced-
proximity model that, this oligomerization, in turn, promotes the recruitment of the
downstream activator, RIP2 into a close proximity through the CARD-CARD
interaction, forming a large signaling platform that eventually activates NF-κB [21]
(Figure 2.5B).
The direct ligand interaction with the LRR domain and the consequent
29
activation by induced-proximity model have been proved experimentally [99, 103,
104]. Also, ligand-induced oligomerization of NOD1 and ATP-binding have been
demonstrated experimentally [101].
30
Figure 2.5 Structure of NOD1 and its activation mechanism model
A. The domain structure of NOD1. NOD1 consists of N-terminal CARD
domain, NOD domain, and C-terminal LRR domain.
B. The activation mechanism model of NOD1. In the absence of the ligand,
the LRR domain prevents the oligomerization and subsequent activation.
Once the ligand is recognized through the LRR domain, the confor-
mational change occurs, which exposes the NOD domain, allowing
oligomerization and subsequent proximity activation of RIP2 kinases that
mediate the downstream signaling.
31
2.3.2. NOD1 protein purification
NOD1 consists of three domains. An N-terminal CARD domain, NOD domain,
and a C-terminal LRR domain (Figure 2.5A). There is no report of the crystal
structure of full-length NOD1 protein, but the crystal structure of the N-terminal
CARD domain of NOD1 has been revealed [105, 106]. The crystal structures of the
NOD domain and the LRR domain of NOD1 do not exist, but the structures of the
LRR domains of other proteins have been reported [107-109].
The LRR domains are known to adopt an arc or horseshoe-like structure
with the concave face of parallel β-strands and the convex face of more variable
secondary structures including helices [98]. Most LRR domains consist of 2 to 45
leucine-rich repeats, and each repeat consists of about 20 to 30 amino acids with a
highly conserved LxxLxLxxN motif [98]. Many LRR family proteins are known to
be hard to produce in sufficient quantities [110]. Also, NOD1 has proven to be
notoriously difficult to express and purify sufficiently in an active, functional form
[22]. This hampered the molecular-scale researches of NOD1 for a long time and
most of the studies have been based on cellular-scale experiments. Some high-
throughput screening researches to identify NOD1 inhibitors were also performed
in a cell-based manner, implying the difficulty in purifying NOD1 protein [22, 111].
NOD1 is a potential therapeutic target which is related to inflammatory
diseases such as asthma, inflammatory bowel disease, and atopic diseases.
Inhibitors of NOD1 could be potential therapeutic drugs for immune or
inflammatory diseases. Also, NOD1 inhibitors might serve as valuable tools for
32
researches on the involvement of NOD1 in inflammatory diseases and normal
immune responses [111]. In vitro mechanistic studies and assays, high-throughput
screenings that are necessary for drug development require respectable amount of
purified protein. To date, NOD1 protein production has been quite limited in the
aspects of cost, amount and technology. There are mainly two studies that utilized
purified NOD1 for in vitro characterization of the protein. In the first research
[103], NOD1 protein was purchased from Novus biologicals [112]. The preparation
method of this NOD1 protein was in vitro wheat germ expression system [112].
Even though the manufacturer claims that this protein has not been tested for any
functionality [112], it has been shown that the protein interacts with its ligand Tri-
DAP through its LRR domain [103]. Also, it was demonstrated that, when NOD1
was prebound to its ligand, the downstream RIP2 kinase (RICK or RIPK) binds to
the NOD1 protein with higher affinity and the phosphorylation activity of RIPK
increased [103]. This work has proved the direct binding of NOD1 ligand to its
LRR domain and subsequent activation of downstream effector kinase RIPK, but
the nucleotide-binding, oligomerization and hydrolysis by the NOD domain were
not shown. The second research which utilized NOD1 protein purified from animal
cells has demonstrated the functions of the NOD domain [101]. In this study,
addition of the NOD1 ligand γ-Tri-DAP induced NOD1 oligomerization. Also,
NOD1 has been shown to exhibit preference to ATP over ADP, AMP, or dATP
[101]. This work actually filled the missing parts of the former study, but it did not
demonstrate the ATP hydrolysis by NOD domain. They utilized NOD1 stable cell
33
line of 293 Freestyle cells to purify NOD1 protein [101]. The 293 Freestyle cells
were infected with lentivirus containing the NOD1 open reading frame (ORF) with
N-terminal 6ÍHis- and FLAG tags under the CMV promoter [101]. To increase
the expression level of NOD1 protein, multiple infections were performed [101].
There are several recombinant protein expression systems that have been
developed (Table 2.4 [113]). Each of them have pros and cons of their own. As for
NOD1 protein, mammalian cell system [101] and cell-free system (wheat-germ)
[103, 112] have been demonstrated by the two researches mentioned above. But
there is no report that utilizes highly productive E. coli system. Like most of the
other proteins, for cost-effective, large-scale, and high-yield production of NOD1
protein, the development of E. coli system for NOD1 expression and purification
could be the next step to facilitate researches such as crystallography and high-
throughput screening that require functional NOD1 protein in a respectable amount.
2.3.3. Overview of SGT1
Suppressor of G2 allele of skp1 (SGT1) is an essential, conserved protein which is
required for kinetochore assembly in both yeast [114, 115] and mammalian cells
[116, 117]. SGT1 was discovered in the budding yeast Saccharomyces cerevisiae
as a high copy suppressor of skp1-4 [118], a temperature-sensitive allele of SKP1
[119]. SGT1 has sequence homology to both TPR and p23 domains which are
found in
34
Table 2.4 Recombinant protein expression systems
E. coli Yeast Insect cells Mammalian cells Cell-free (E. coli)
Cell-free (wheat germ)
Average time of cell division
30 min 90 min 18 hr 24 hr N/A N/A
Cost of expression
Low Low High High High High
Expression level
High Low-High Low-High Low-Moderate Low-High Low-High
Success rate (% soluble)
40-60 50-70 50-70 80-95 Variable Variable
Advantages Simple, low cost, rapid, robust, high yield, easy labeling for structural studies
Simple, low cost
Post-translational modifications
Natural protein configuration, post-translational modifications
High yield, fast, flexible, disulfide-bonded and membrane proteins, easy labeling for structural studies
Fast, flexible, disulfide-bonded and membrane proteins, post-translational modifications
Disadvantages No post-translational modifications, insoluble protein, production of disulfide-bonded and membrane proteins is difficult
Less post-translational modifications, production of membrane proteins is difficult
Slow, higher cost, production of membrane proteins is difficult
Slow, high cost, lower yield
High cost, less post-translational modifications, efficient production requires highly-specialized setup
High cost, lower yield than E. coli cell-free system, efficient production requires highly specialized setup
35
Hsp90-binding proteins. Indeed, SGT1 interacts with Hsp90 in multiple functional
complexes [114, 120-122]. SGT1 participates in diverse, unrelated cellular
functions ranging from kinetochore assembly and cell cycle progression to
activation of ubiquitin ligases, adenylyl cyclase and Polo kinase, plant immunity,
plant hormone signaling, and heat shock tolerance [26, 123].
SGT1 consists of three domains, all of which mediate protein-protein
interactions and confer the functional diversity of SGT1 as an adaptor protein.
These domains are tetratricopeptide repeats (TPR), CHORD-containing protein and
SGT1 (CS) and SGT1-specific (SGS) domains (Figure 2.6A). TPR and CS
domains are known to be globular and stable, while SGS is intrinsically unfolded
[121, 124]. The CS domain which is related to the p23 cochaperone, interacts with
the N-terminal domain of Hsp90 [124]. In contrast, the TPR domain is not involved
in the interactions with either the N-terminal domain or the C-terminal MEEVD of
Hsp90. In yeast, the TPR domain mediates the dimerization of SGT1, which is
crucial for kinetochore assembly and growth of yeast cells [125]. The dimerization
of SGT1 is negatively regulated by casein kinase 2 (CK2) which phosphorylates
the Ser361 of yeast SGT1 [126]. These are the different features of yeast SGT1
compared to human SGT1 which does not self-associate [127]. Although plant
SGT1 has TPR-mediated dimerization property which controls the stability of
SGT1 protein, the TPR domain of plant SGT1 has been shown to be dispensable
for plant immunity and auxin signaling [25]. The SGS domain is the most
conserved part of SGT1. SGT1 seems to bind LRR-containing proteins such as
36
adenylyl cyclase, NLR proteins including NOD1 and plant R proteins, through its
SGS domain [128]. The SGS domain also mediates the interactions of SGT1 with
Hsp70 and S100 calcium-binding proteins [26].
2.3.4. SGT1-Hsp90 complex and NOD1
In plants, Hsp90, SGT1 and RAR1 are essential for the functions of many NLR
sensors and plant disease resistance. [26] The Hsp90-SGT1-RAR1 complex is
implicated for the chaperoning of NLR sensors that are crucial for plant immunity
[23].
In mammals, Hsp90 and SGT1 are also necessary for active NLR sensors such
as NOD1 [24] and NLRP3 [128]. It has been indicated that many mammalian NLR
sensors interact with SGT1-Hsp90 complex [128]. These SGT1-Hsp90-NLR
complexes show an aspect of similarity in innate immunity between plants and
mammals. In the case of NOD1, depletion of SGT1 with siRNA prevented multiple
NOD1-associated cellular responses, even though it did not affect the stability of
NOD1 protein or downstream signaling molecules [24]. On the other hand, the
depletion or inhibition of Hsp90 resulted in decreased stability of NOD1 [24].
These results imply that SGT1 is required for NOD1 activity, while Hsp90 is
required for NOD1 stability (Figure 2.6B).
37
Figure 2.6 Structure of SGT1 and its function as an adaptor protein
A. The domain structure of human SGT1a. SGT1 consists of N-terminal TPR
domain, CS domain, and C-terminal SGS domain.
B. A model of SGT1 function as an adaptor protein. SGT1 acts as an adaptor
protein that delivers LRR-containing protein to Hsp90 chaperone, leading to
the stabilization and activation of the LRR-containing protein. This event
might take place through the concomitant interactions of SGT1(CS)-
Hsp90(N) and SGT1(SGS)-LRR.
Chapter 3.
Materials and methods
39
3.1. Plasmid DNA vectors for protein expression
HSC82, one of the two HSP90s of Saccharomyces cerevisiae, was PCR-amplified
and subcloned into the NdeI and XhoI sites of pET15b vector (Novagen). All PCR
reactions were performed with Mastercycler (Eppendorf).
Homo sapiens PP5 full-length (amino acid residues 1-499) and its domain
constructs, PP5ΔC (1-486), PD (170-499), and PDΔC (170-486) were all
subcloned into the NheI and NotI sites of pET28b vector.
pETDUET-KTM vector which was derived from pETDuet-1 vector (Novagen),
contains kanamycin resistance gene instead of its original ampicillin resistance
gene and a T7 tag sequence, an MauBI site in its MSC2 region. The construction of
pETDUET-KTM vector is as follows. First, 100 μM stock solutions of T7 tag
sequence oligomers were mixed to a final concentration of 4 μM each in annealing
buffer (10 mM Tris-HCl, pH 8.0, 50 mM NaCl, 1 mM EDTA) and annealed by
heating at 95 °C for 5 min and subsequent cooling down to RT. This produced T7
tag sequence fragment with sticky ends of BglII and MfeI sites which was ligated
into the BglII and MfeI (MunI) sites of the MCS2 region of pETDuet-1 vector,
resulting in pETDUET-T7 vector. Second, an MauBI site was introduced by
ligating a PCR fragment into the MluI and AvrII sites of pETDUET-T7 vector,
generating pETDUET-T7-MauBI vector. Finally, PciI/DraIII digested fragment of
pET28b vector (Novagen), containing kanamycin resistance gene, was ligated into
the PciI and DraIII sites of pETDUET-T7-MauBI vector, replacing its ampicillin
40
resistance gene. The constructed pETDUET(KanR)-T7 tag-MauBI site vector was
designated as pETDUET-KTM.
SGT1NOD1LRR fusion genes were generated and amplified by two-step PCR
and subcloned into the NcoI and NotI sites which are in the MCS1 region of
pETDUET-KTM. SGT1 gene for the fusion protein generation was either from H.
sapiens or S. cerevisiae, and the resulting fusion proteins were designated as
HsSGT1NOD1LRR and ScSGT1NOD1LRR, respectively. The LRR domain of
NOD1 was first constructed as an elongated form (641-953). Then, in an attempt to
reduce the proteolytic degradation of the fusion protein, the construct was
shortened (656-953) and subcloned into the pre-existing HindIII site of the first,
elongated construct, remaining three amino acids from it. SGT1 and NOD1-LRR
were connected by 2-step PCR with the linker primer which contains a thrombin
site sequence. A 6ÍHis tag sequence was added to the C-terminal end of NOD1-
LRR by PCR.
For the coexpression of SGT1NOD1LRR fusion protein and HSP90s in
Escherichia coli, S. cerevisiae HSC82 and HSP82, Mus musculus HSP90α, and E.
coli HtpG were subcloned into the XhoI and MauBI sites that are in the MCS2
region of pETDUET-KTM.
41
Table 3.1 Primers for expression vector construction
Primers Sequence (5’-3’)
pETDUET-T7(2nd)-F GATCTCATGGCTAGCATGACTGGTGGACAGCAAATGGGTCTC
pETDUET-T7(2nd)-R AATTGAGACCCATTTGCTGTCCACCAGTCATGCTAGCCATGA
pETDUET-MluI-F TTTACGCGTTGCGCGAGAAGA
pETDUET-MauBI-R TTTCCTAGGTTAATTAACGCGCGCGGCTGCGCTAGTAGACGAG
NOD1LRR-F(1st) GAAAGCTTCAACCAGGTGCAGG
NOD1LRR-F-NcoI TTTCCATGGGCGAAAGCTTCAACCAGGTGCAGG
NOD1LRR-R-6xHis-NotI TTTGCGGCCGCGTGGTGATGATGGTGATGGAAACAGATAATCCGCTTCTCAT
HsSGT1a-F-NcoI TTTCCATGGGCATGGCGGCGGCTGCAGCAGG
HsSGT1a-R GTACTTTTTCCATTCCATATCATC
Linker CCTGCACCTGGTTGAAGCTTTCGCTGCCGCGCGGCACCAGGCCGCTGCTGTACTTTTTCCATTCCATATCATC
NOD1LRR-F-HindIII TTTAAGCTTCCTGCGCTGCATCTACGAGA
EcHtpG-XhoI-F(new) TATGATCTCGAGATGAAAGGACAAGAAACTCGTG
EcHtpG-MauBI-R TATCATCGCGCGCGTCAGGAAACCAGCAGCTG
MmHSP90a-XhoI-F TATGATCTCGAGATGCCTGAGGAAACCCAG
ScHSC82-XhoI-F TATGATCTCGAGATGGCTGGTGAAACTTTTGAATTTC
ScHSC82-MauBI-R TATCATCGCGCGCGTTAATCAACTTCTTCCATCTCGGTG
ScHSP82-XhoI-F TATGATCTCGAGATGGCTAGTGAAACTTTTGAATTTC
ScHSP82-MauBI-R TATCATCGCGCGCGCTAATCTACCTCTTCCATTTCGG
42
Table 3.2 Primers for expression vector construction (Continued)
Primers Sequence (5’-3’)
scSGT1-F-NcoI TTTCCATGGGCATGCCTGTTGAAAAAGATTTAAAAAC
scSGT1-R-(-stop)linker CGGCACCAGGCCGCTGCTCCAATGTTTAGGTTCCATGC
HtpG-Sense-NdeI CGCCATATGAAAGGACAAGAAACTCGTGG
HtpG-Anti-sense-XhoI CCGCTCGAGTCAGGAAACCAGCAGCTGGTTC
MmHSP90a-Sense-NotI ATAAGAATGCGGCCGCATGCCTGAGGAAACCCAGACCC
MmHSP90a-Anti-sense-XhoI
CCGCTCGAGTTAGTCTACTTCTTCCATGCGTG
PP5ΔC13-F-NheI TTTGCTAGCATGGCGATGGCGGAGGGCGAG
PP5ΔC13-R-NotI TTTGCGGCCGCTCAGGGCTTGACGTTGGGATGAGG
PDΔC13-F-NheI TTTGCTAGCATGACCATTGAGGATGAGTACAG
PDΔC13-R-NotI TTTGCGGCCGCTCAGGGCTTGACGTTGGGATGAGG
HSC82-F-NdeI TTACATATGATGGCTGGTGAAACTTTTGAA
HSC82-R-XhoI TTACTCGAGTTAATCAACTTCTTCCATCTC
43
3.2. Protein expression and purification
For Hsc82p (~ 82 kDa) expression, pET15b-HSC82 plasmid DNA was introduced
into chemically competent E. coli strain BL21 star(DE3) (Invitrogen) or Rosetta
gami2(DE3)pLysS (Novagen) by heat-shock transformation. The transformants
were selected by spreading and growing the E. coli cells on an ampicillin plate
over-night in a 37 °C standing incubator. A single colony from the plate was
inoculated into Luria-Bertani (LB) media containing ampicillin, and incubated
over-night in a 37 °C shaking incubator. The first inoculate was secondly
inoculated by 1/100 volume of LB media containing ampicillin and this second
inoculate was incubated in a shaking incubator maintained at 37 °C. The optical
density (OD) at 600 nm was measured by Cary50 conc UV-Vis spectrophotometer
(Varian). At the OD600 of about 0.7~0.8, isopropyl β-D-1-thiogalactopyranoside
(IPTG) was added to the culture to a final concentration of 1 mM. Then, the culture
was moved into a shaking incubator maintained at 30 °C and further incubated for
5 hr. The cells were harvested by centrifugation at 6500 rpm for 15 min with a
high-speed centrifuge (Supra21K, Han-il) and the cell pellets were stored in a deep
freezer (Sanyo).
The frozen cell pellets were thawed on ice and resuspended in the Ni-NTA
binding buffer. Then the cell suspension was lysed by sonication (VCX500, Sonics)
with repeated on (2 s) and off (8 s) cycles for about 20 min. Then, the lysate was
cleared by centrifugation by high-speed centrifuge at 4 °C, 10000 rpm for 30~40
44
min. The supernatant was either mixed with Ni-NTA resin or applied to HisTrap
column (GE healthcare). Each fraction from the immobilized-metal affinity
chromatography (IMAC) was analyzed by polyacrylamide gel electrophoresis
(PAGE) followed by Coomassie staining and the elution fractions were collected
and dialyzed with MonoQ start buffer (20 mM Tris-HCl, pH 8.0). The dialyzed
protein was applied to the MonoQ 5/50 GL column (Amersham Biosciences) and a
linear gradient with the elution buffer (20 mM Tris-HCl, pH 8.0, 1.0 M NaCl) was
applied by ÄKTA-FPLC (Amersham Biosciences) for 20 CVs. Each fraction from
the ion exchange chromatography (IEC) was analyzed by PAGE and the adequate
fractions were collected, dialyzed and stored at -80° C.
PP5 and its domain constructs subcloned into pET28b were transformed,
expressed in Rosetta gami2(DE3)pLysS. Harvested cells were lysed by sonication
and the proteins were purified by Ni-NTA IMAC with HisTrap FF column or
HisPur Ni-NTA resin (Thermo scientific), in a similar manner to Hsc82p
purification.
3.3. Structure-based virtual screening with modified
AutoDock
Tthe AutoDock program [129] was utilized in the structure-based virtual screening
of Hsp90 inhibitors due to the outperformance of its scoring function over those of
the others had been shown in several target proteins [130]. The 3-D coordinates in
45
the crystal structure of Hsp90 in complex with a benzenesulfonamide inhibitor
(PDB code: 2BZ5) [131] were selected as the receptor model in the virtual
screening with docking simulations. After removing the ligand and solvent
molecules, hydrogen atoms were added to each protein atom. A special attention
was paid to assign the protonation states of the ionizable Asp, Glu, His, and Lys
residues. The side chains of Asp and Glu residues were assumed to be neutral if
one of their carboxylate oxygens pointed toward a hydrogen-bond accepting group
including the backbone aminocarbonyl oxygen at a distance within 3.5 Å, a
generally accepted distance limit for a hydrogen bond with moderate strength [132].
Similarly, the lysine side chains were protonated unless the NZ atom was in
proximity of a hydrogen-bond donating group. The same procedure was applied to
determine the protonation states of ND and NE atoms in His residues.
The docking library of about 85,000 compounds for Hsp90 was constructed
from the latest version of InterBioScreen database [133] containing approximately
30,000 natural and 290,000 synthetic compounds. This selection was based on
drug-like filters that adopt only the compounds with physicochemical properties of
potential drug candidates [134] and without reactive functional group(s). All of the
compounds included in the docking library were then subjected to the Corina
program to generate their 3-D coordinates, followed by the assignment of
Gasteiger-Marsilli atomic charges [135]. Docking simulations were then carried
out in the active site of Hsp90 to score and rank the compounds according to the
calculated binding free energy of the most populated binding configurations.
46
In the actual docking simulation of the compounds in the docking library, the
empirical scoring function improved by the implementation of a new solvation
model for a compound was applied. The modified scoring function has the
following form:
å ååå
åååå
= >
-
= >
= >= >
÷÷
ø
ö
çç
è
æ-+++
÷÷
ø
ö
çç
è
æ-+
÷÷
ø
ö
çç
è
æ-=D
1
2max
1
11012
1612
2
2
)(
)(
i ij
r
jiisoltortori ij ijij
ji
elec
i ij ij
ij
ij
ij
hbondi ij ij
ij
ij
ij
vdWaqbind
ij
eVOccSWNWrr
qqW
r
D
r
CtEW
r
B
r
AWG
s
e
(1)
where WvdW, Whbond, Welec, Wtor, and Wsol are the weighting factors of van der Waals,
hydrogen bond, electrostatic interactions, torsional term, and desolvation energy of
an inhibitor, respectively. rij represents the interatomic distance, and Aij, Bij, Cij, and
Dij are related to the depths of the potential energy well and the equilibrium
separations between the two atoms. The hydrogen bond term has an additional
weighting factor, E(t), representing the angle-dependent directionality. Cubic
equation approach was applied to obtain the dielectric constant required in
computing the interatomic electrostatic interactions between Hsp90 and a ligand
molecule [136]. In the entropic term, Ntor is the number of sp3 bonds in the ligand.
In the desolvation term, Si and Vi are the solvation parameter and the fragmental
volume of atom i [137], respectively, while Occimax stands for the maximum atomic
occupancy. In the calculation of molecular solvation free energy term in Eq. (1),
47
the atomic parameters recently developed by Kang et al. [138] were used because
those of the atoms other than carbon were unavailable in the current version of
AutoDock. This modification of the solvation free energy term is expected to
increase the accuracy in virtual screening, because the underestimation of ligand
solvation often leads to the overestimation of the binding affinity of a ligand with
many polar atoms [139].
3.4. Molecular dynamics for Hsp90-inhibitor comlexes
To obtain better binding configurations for Hsp90-inhibitor complexes, molecular
dynamics (MD) were performed in aqueous solution. The most stable structures of
Hsp90-inhibitor complexes obtained from docking simulation were equilibrated in
solution through 1 nanosecond MD simulation with AMBER program, which had
been successful in modeling the structures of proteins [140] and nucleic acids [141]
in solution. This equilibration procedure started with the addition sodium ions as
the counterion to neutralize the total charge of the all-atom model of Hsp90. The
system was then immersed in a rectangular solvent box containing about 8000
TIP3P water molecules. After 1000 cycles of energy minimization to remove bad
van der Waals contacts, the system was equilibrated beginning with 20 ps
equilibration dynamics of the solvent molecules at 300 K. The next step involved
equilibration of the solute with a fixed configuration of the solvent molecules for
10 ps at 10, 50, 100, 150, 200, 250, and 300 K. Then, the equilibration dynamics of
48
the entire system was performed at 300 K for 500 ps using the periodic boundary
condition. The SHAKE algorithm [142] was applied to fix all bond lengths
involving hydrogen atom. A time step of 1.5 fs and a nonbond-interaction cutoff
radius of 12 Å were used.
3.5. In vitro Hsp90 ATPase assay
The Hsp90 used in the assay was ScHsc82 protein purified by Ni-NTA IMAC and
subsequent MonoQ IEC. As a colorimetric assay, the assay utilizes the malachite
green reagent (BioAssay Systems). To a 96-well plate, distilled water, 5Í assay
buffer (500 mM Tris-HCl, pH 7.4, 100 mM KCl, 30 mM MgCl2), 10 mM
compound were added. As the compounds were dissolved in DMSO, 0.4 % DMSO
was used as a negative control. Then 0.5 μg of purified ScHsc82 protein was added
to each well and finally, 10 mM ATP stock was added to the final concentration of
1 mM to complete the 25 μl reaction mixture. The plate was shaken for 3 min,
1200 rpm in a 96-well plate reader (Thermo scientific), sealed with parafilm and
incubated at 37° C for 3 hr. After the incubation, 6.25 μl of malachite green reagent
was added to each well and the plate was shaken in the plate reader for 10 min,
1200 rpm, then the optical density at 620 nm was measured [143].
The assay for IC50 estimation was essentially identical to the abovementioned
assay protocol. The compounds were added to the concentrations of 0, 10, 20, 40,
60 μM, while geldanamycin was added to the concentrations of 0, 1, 2, 4, 6, 10 μM.
49
3.6. Sulforhodamine B (SRB) assay
Human breast cancer cell line MCF-7 cells were grown and maintained in the
DMEM media (89 % DMEM, 1 % penicillin/streptomycin, 10 % FBS). To
determine the GI50 values, 180 μl of 5Í104 cells/ml concentration of MCF-7 cells
were seeded in a 96-well plate and grown for 24 hr in a CO2 incubator (Sanyo), 37°
C. Each compound was diluted to give desired concentrations (1~100 μM) and 20
μl of each diluent was added to the cells in the 96-well plate. The plate was
incubated for 4 days in the CO2 incubator. After the incubation, 50 μl of cold 50 %
TCA (Sigma) was added carefully and after 2 min, the plate was incubated at 4 ° C
for fixation. Then, the fixed cells were washed with distilled water five times and
100 μl of 0.4 % SRB (Sigma) solution was added and incubated for 1 hr, RT. After
the staining, cells were washed with 1 % acetic acid for five times and set dry for
10 min at RT. The SRB dye was then melted with 150 μl of 10 mM Tris buffer and
the optical density at 550 nm was measured by the plate reader [144].
50
3.7. Cell-based Her2 degradation assay
MCF-7 cells were seeded into each well of a 6-well plate at a density of 5Í105
cells/well. The cells were incubated in a CO2 incubator, 37° C for about 30 hr to
reach the confluency of ~ 60 %. Each compound was added into wells with the
cells to final concentrations of 20 and 40 μM. As a positive control, 17-AAG was
added to a final concentration of 0.5 μM. After incubating the plate for 24 hr in the
CO2 incubator, the cells were harvested with the medium, centrifuged for 10 min,
800Íg at 4° C. The cell pellets were washed with 1 ml of ice-cold PBS and stored
at -80° C until use. The cells were lysed by resuspension in 70 μl of TNES lysis
buffer (50 mM Tris-HCl, pH 7.4, 1 % NP-40, 2 mM EDTA, 100 mM NaCl,
supplemented with 1 mM PMSF, 0.1 % protease inhibitor cocktail from
Calbiochem) and incubation on ice for 1 hr, with vortexing once during the
incubation. Then the lysed cells were centrifuged at 13,200 rpm, 20 min, at 4° C.
The supernatants were moved to a new micro tube and the protein concentrations
of each sample were measured by Bradford assay. 50 μg of each sample was
subjected to PAGE (Bio-RAD mini-protean 3) and the gel was transferred to a
nitrocellulose membrane (Amersham) by Trans-blot semi-dry transfer cell (Bio-
RAD). The western blot analysis was performed with Her2 antibody (sc-284, Santa
Cruz biotech.), Hsp70 antibody (SPA-840, Stressgen), and β-tubulin antibody (sc-
9104, Santa Cruz biotech.).
51
3.8. PP5 phosphatase assay
The screening for PP5 modulators was executed in a 96-well plate format with a
total reaction volume of 100 μl. Each compound of LOPAC1280 (The Library of
Pharmacologically Active Compounds, Sigma-Aldrich) was dissolved in 100 %
DMSO as a 10 mM stock. The compounds were diluted with deionized-distilled
water (DDW) and then applied to a 96-well plate to give a final concentration of 20
μM. As the compounds were dissolved in DMSO, final concentration of 0.2 %
DMSO was used as a negative control. A negative control mix and an enzyme mix
were constituted with DDW, 5Íphosphatase assay buffer (200 mM HEPES-NaOH,
pH 7.5, 2.5 mM MnCl2) and the dialysis buffer (20 mM Tris-HCl, pH 8.0, 150 mM
NaCl) or the purified proteins. Purified PP5 protein was added to the reaction
mixture to a final concentration of 50 nM and the plate was shaken for 15 s in a
plate reader (Multiskan GO, Thermo scientific). After sealed with rubber bands, the
plate was incubated in a 30 ° C water-bath (WB-03, JEIO tech.) for 10 min with the
artificial substrate 4-nitrophenol phosphate (pNPP, N4645, Sigma). After the
incubation, pNPP was added to each well of the plate with automatic dispensing
multi-pipette (Eppedorf Xplorer / Sartorius). The plate was shaken for 15 s and
read at 405 nm in the plate reader for the time zero point. The plate was sealed with
the rubber bands, incubated in the 30 ° C water-bath for 1 hr. Then 50 μl of stop
solution (1 M potassium phosphate, dibasic, adjusted to pH 10 with NaOH) was
added to each well [145]. The plate was shaken for 15 s and read at 405 nm in the
52
plate reader.
3.9. SGT1NOD1LRR fusion protein expression
The constructed vectors were transformed into chemically competent Rosetta
gami2(DE3)pLysS cells which were grown in LB media containing
chloramphenicol to retain the pRARE2LysS vector. The transformed cells were
grown in LB media containing chloramphenicol and kanamycin in a shaking
incubator at 37° C to the point when optical density at 600 nm reached ~0.8. Then
the expression of the proteins was induced by adding IPTG to a final concentration
of 1 mM. The flasks were moved to a shaking incubator at 30° C and samples were
collected each hour. The collected samples were centrifuged and stored at -20° C or
-80 ° C as cell pellets. The pellets were resuspended in diluted 5Í SDS-PAGE
sample buffer (60mM Tris-HCl, pH 6.8, 25 % glycerol, 2 % SDS, 5 % β-
mercaptoethanol, 0.1 % bromophenol blue) which was diluted with distilled water
to give 1.5 ~ 2.5Í concentration. The samples were boiled, centrifuged and
applied to SDS-PAGE and western blot analysis. ScSGT1NOD1LRR fusion
protein was detected with His antibody (sc-8036, Santa Cruz biotech.).
HsSGT1NOD1LRR protein was probed with either SGT1 antibody (612104, BD
bioscience) or His antibody. HSP90s were detected by T7 tag antibody (69522,
Merck). Enhanced chemiluminescence (ECL) solutions used for western blot
detection were SuperSignal West Pico (and Dura) substrate (Thermo scientific).
53
The western blot images were taken with G:box iChemi-XL image analyzer
(Syngene). Restore™ western blot stripping buffer (Thermo scientific) was used
for the stripping of nitrocellulose membrane.
Chapter 4.
Virtual and experimental screening to
identify new Hsp90 inhibitor scaffolds
55
4.1. Introduction
Hsp90 molecular chaperone supports numerous proteins for their folding,
stabilization and activation, serving them as client proteins of Hsp90. The Hsp90
client proteins include many oncoproteins which participate in the maintenance of
the six hallmarks of cancer such as self-sufficiency in growth signal, evading
apoptosis, sustained angiogenesis, limitless replicative potential, tissue invasion
and metastasis, and insensitivity to anti-growth signals (Figure 2.1B). It has been
proposed that Hsp90 inhibition leads to the degradation of its client proteins by the
ubiquitin-proteasome pathway [146] and this also applies to the fate of the
oncoproteins that are dependent on Hsp90 as its client proteins. As anti-cancer drug
development has been focused on a single target such as Her2, Hif-1, Hsp90 has
been highlighted as a promising new target for anti-cancer therapeutics that can
suppress not only one but all of the six hallmarks of cancer cell [2].
Initially, natural products geldanamycin and radicicol were discovered by their
antibiotic activities. Later, the molecular target of these compounds was found to
be Hsp90 [147]. These compounds has several major drawbacks as therapeutic
drug candidates such as hepatotoxicity, low solubility and instability/reactivity [1,
148]. In spite of these limits, they provided chemical basis for the developments of
derivatives that inhibit Hsp90 and retain appropriate properties as drugs [149].
Geldanamycin is a benzoquinone ansamycin, while radicicol is a resorcylic macro-
lactone. The developments of the derivatives based on these scaffolds could
56
overcome the drawbacks of the original compounds, while retaining the inhibitory
activity on Hsp90. Finally, some of the derivatives could enter clinical trials [150-
152]. 17-allylamino-17-demethoxy-geldanamycin (17-AAG) was the first geldana-
mycin derivative that proceeded to clinical trial [153, 154]. Another geldanamycin
derivative, 17-dimethylaminoethylamino-17-desmethoxy-geldana-mycin (17-DMA
G) also entered clinical testing [152]. Although the results from the clinical trials
demonstrated successful targeting of Hsp90 by these compounds, they had
limitations such as intrinsic liver toxicity problem due to its benzoquinone moiety
[152]. Oxime-derivatives of radicicol with less liver toxicity were developed by
Kyowa Hakko [155], and some cyclopropyl analogs of radicicol have been
synthesized to eliminate the reactive epoxide moiety [156, 157], but none of them
have advanced to clinical tests [2]. Instead, the resorcinol unit of radicicol that
binds to Hsp90 was introduced into many small-molecule Hsp90 inhibitors [1]. To
overcome the limitations of the scaffolds derived from the natural products
geldanamycin and radicicol, Hsp90 inhibitors with new scaffolds have been
discovered through intensive researches and developments, leading to the
discovery of diverse classes of Hsp90 inhibitors such as purines, resorcinols,
pyrimidines and azoles [158].
As most of the Hsp90 inhibitors target the ATP-binding pocket that resides in
the N-terminal domain of Hsp90 and compete with ATP, Hsp90 inhibitors with
purine scaffold have been rationally designed [159]. The resulting prototype PU3
led to the clinical candidates BIIB021 [160], BIIB028 [161], and PU-H71 [162].
57
Other compounds that contain the resorcinol unit of radicicol, such as NVP-
AUY922, KW-2478 [163], AT13387 [164], and STA-9090 [165] also are under
clinical tests [166].
Aforementioned state of Hsp90 inhibitor development clearly shows that the
discoveries of novel scaffolds are quite useful to overcome the intrinsic limits of
the known scaffolds. Also, novel scaffolds can serve as a functional parts for
fragment-based drug design. The aim of this work is to find Hsp90 inhibitors that
possess new scaffolds. The screening process was facilitated by in silico virtual
screening which enabled the coverage of a large number of compounds, and in
vitro Hsp90 ATPase assay and cell-based Her2 degradation assay were performed,
subsequently.
4.2. Virtual screening with modified AutoDock program
To date, most of the Hsp90 inhibitors have been identified by 1) the modifications
of natural or pre-existing Hsp90 inhibitors, 2) high-throughput screening for new
scaffolds, 3) rational design, or the combination of these. New classes of Hsp90
inhibitors such as 1-(2-phenol)-2-naphthol [131] and 5-aminoimidazole-4-
carboxamide-1-b-D-ribofuranoside (AICAR) [167] were discovered by structure-
based virtual screening, proving that this method could be a useful tool for the
discovery of lead compounds.
The schematic procedures executed in the process of identifying the Hsp90
58
inhibitors characterized in this chapter are demonstrated in Figure 4.1 [168].
Approximately 320,000 compounds (30,000 natural, 290,000 synthetic compounds)
from the InterBioScreen database [133] were narrowed down to 85,000 compounds
by drug-like filters that adopt the compounds with physicochemical properties of
potential drug candidates (Lipinski’s rules of five) [134] and without reactive
functional group(s).
Figure 4.1 The strategy applied for the discovery of novel Hsp90 inhibitors
Computational and experimental methods were integrated into the strategy that
identified novel Hsp90 inhibitors investigated in this work. The asterisk represents the
compounds of InterBio Screen database.
59
The modified version of the AutoDock program enabled the selection of 300
top-scored compounds out of the 85,000 compounds, and 285 compounds of the
selected 300 compounds were available from the compound supplier. These
compounds were subjected to experimental screening with in vitro Hsp90 ATPase
assay. This secondary screening identified 41 compounds that showed more than
70 % inhibition at the concentration of 50 μM. Subsequent cell-based Her2
degradation assay revealed 5 compounds that reduced Her2 protein levels more
than 50 % compared to the vehicle control (DMSO). Out of these 5 compounds, 3
compounds shared a common scaffold, 3-phenyl-2-styryl-3H-quinazolin-4-one.
Previously, these compounds had been characterized and confirmed as Hsp90
inhibitors [168].
In this chapter, the other 2 compounds from the formerly executed screening
processes are investigated. These compounds have distinct scaffolds with the
abovementioned 3 compounds, and are also different with each other. One of these
compounds has pyrimidine-2,4,6-trione scaffold (1b), while the other has 4H-1,2,4-
triazole-3-thiol scaffold (2b).
4.3. Inhibitory and antiproliferative effects of the compounds
The inhibitory effect of each compound on purified Saccharomyces cerevisiae
Hsp90 (Hsc82) ATPase activity was assessed by ATPase assay with malachite
green reagent utilizing colorimetric detection of the free inorganic phosphate which
60
results from ATP hydrolysis by Hsp90 [143]. The phosphate-malachite green
complex can be detected by measuring the optical density at 620 nm. The
compounds 1b and 2b showed the IC50 values of 24.6 μM and 15.4 μM,
respectively. Geldanamycin, as a positive control, showed the IC50 value of 2.3 μM
(Table 4.1). The anti-proliferation activities of 1b and 2b were evaluated by
sulforhodamine B (SRB) assay with MCF-7 breast cancer cell line [144]. The GI50
values of 1b and 2b were 18.1 μM and 44.6 μM, respectively. The GI50 of
geldanamycin was 9.6 μM (Table 4.1). Although 1b was less effective than 2b in
inhibition of Hsp90 ATPase activity, 1b was more effective in growth inhibition of
MCF-7 cells, which could be due to the differences in the cell membrane
permeability or stability of the compounds in the medium. Also, these compounds
might have different inhibitory effects on yeast Hsp90 and human Hsp90.
To obtain some structural insights into the mechanisms of inhibition by these
compounds, their binding modes into the ATP-binding pocket of Hsp90 were
calculated with AutoDock program. Based on the structure of the protein-inhibitor
complexes obtained with the docking simulations, improved binding configurations
of Hsp90-inhibitor complexes were acquired by molecular dynamics (MD)
simulations of the complexes in aqueous solution. simulations of the complexes in
aqueous solution.
61
4.4. Molecular dynamics simulations of the compounds
Due to the structural flexibility of Hsp90, the calculated binding mode of inhibitors
to Hsp90 should be dependent on the selected conformation of the Hsp90 protein.
This can be overcome by choosing the most probable structure from the molecular
dynamics simulations which involve the motions of the protein atoms as well as the
ligand atoms. The representative MD trajectory snapshot of Hsp90-1b complex
shows that a stable hydrogen bond is established between one of the carbonyl
oxygens of 1b and the site chain of Lys112 (Figure 4.2). The inhibitor can be
further stabilized in the ATP-binding site by an additional hydrogen bond between
one of the carbonyl oxygens and the structural water molecule which is in turn
hydrogen-bonded to the side chain aminocarbonyl group of Asn51. This
exemplifies the involvement of a solvent-mediated hydrogen bond, one of the
Table 4.1 Inhibitory and antiproliferative effects of 1b and 2b
Compound Geldanamycin 1b 2b
IC50 (mM)a 2.3 24.6 15.4
GI50 (mM)a 9.6 18.1 44.6
a Values correspond to n=2.
IC50, The concentration inhibiting Hsp90 ATPase activity by 50%.
GI50, The concentration inhibiting MCF-7 cell growth by 50%.
62
significant binding forces in protein-ligand interactions in aqueous solution [169],
in stabilizing the inhibitor in the Hsp90 ATP-binding pocket. The hydrophobic
interactions between the nonpolar groups of 1b and the side chains of Met30,
Leu107, Val136, and Phe138 seem to be also a significant binding force stabilizing
1b in the ATP-binding site.
The representative MD trajectory snapshot for Hsp90–2b complex
demonstrates that one of the oxygen atoms on the terminal dioxolane ring forms a
hydrogen bond with the side chain hydroxyl group of Thr115 (Figure 4.3). It is a
common feature in the binding modes of 2b and 1b that a solvent-mediated
hydrogen bonds participate in the stabilization of the inhibitors in the ATP binding
site of Hsp90. The presence of these solvent-mediated hydrogen bonds in the
Hsp90-inhibitor complexes indicates that the role of structural water molecules
should be considered for the designing of new potent Hsp90 inhibitors. As in the
Hsp90-1b complex, hydrophobic interactions should also be a significant binding
force for stabilizing 2b in the ATP-binding site, because its nonpolar groups form
van der Waals contacts with the side chains of Ala55, Met98, Leu107, Val136, and
Phe138.
63
Figure 4.2 Structure of compound 1b and its MD trajectory snapshot
The binding mode of a pyrimidine-2,4,6-trione compound 1b to the ATP-binding
pocket of Hsp90 is visualized. Carbon atoms of the protein and the compound are
indicated in green and cyan, respectively. Each dotted line indicates a hydrogen bond.
64
Figure 4.3 Structure of compound 2b and its MD trajectory snapshot
The binding mode of a 4H-1,2,4-triazole-3-thiol compound 2b to the ATP-binding
pocket of Hsp90 is visualized. Carbon atoms of the protein and the compound are
indicated in green and cyan, respectively. Each dotted line indicates a hydrogen bond.
65
4.5. Her2 degradation assay with compounds sharing the
scaffolds
As the previous screening identified 3 compounds that share the same 3-phenyl-2-
styryl-3H-quinazolin-4-one scaffold, some other compounds that share the
scaffolds of 1b and 2b might be potential Hsp90 inhibitors. To test the potential of
the scaffolds of 1b and 2b as lead compounds for novel Hsp90 inhibitors, two
compounds (1a and 1c) that share pyrimidine-2,4,6-trione scaffold with 1b, and
two compounds (2a and 2c) sharing 4H-1,2,4-triazole-3-thiol scaffold with 2b
(Figure 4.4), were selected from the compounds obtained from the original virtual
screening and their inhibitory effects on Hsp90 were examined by Her2
degradation assay. MCF-7 cells were treated with each compound or 17-AAG as a
positive control for 24 h, and the levels of Her2, an Hsp90 client protein, were
detected by western blot analysis. All compounds except 1a caused degradation of
Her2 to some extent, confirming the sharing scaffolds possess the properties of
Hsp90 inhibitors (Figure 4.5). Compound 1b had the most potent effect on Her2
degradation and also induced the expression of Hsp70, in agreement with its
inhibitory effect on Hsp90. As the heat shock transcription factor 1 (Hsf1) is
suppressed by Hsp90, the inhibition of Hsp90 activates Hsf1 and induces the
expression of Hsf1 target genes including HSP70 [170]. Other compounds did not
show obvious effect on Hsp70 induction, probably due to their relatively low
efficacy or distinct inhibitory mechanism on Hsp90. Taken together, these results
66
suggest that these two newly identified scaffolds could contribute to the
development of novel Hsp90 inhibitors with improved properties.
Figure 4.4 Six compounds that share the scaffolds of 1b and 2b
Other compounds that share the scaffolds of 1b and 2b were selected from the library.
Compounds 1a, 1b, and 1c share pyrimidine-2,4,6-trione scaffold, while 2a, 2b, and
2c share 4H-1,2,4-triazole-3-thiol scaffold.
67
4.6. Conclusions
By utilizing structure-based virtual screening and consecutive experimental assays,
two Hsp90 inhibitors with novel scaffolds were identified and characterized. The
compounds that share these pyrimidine-2,4,6-trione and 4H-1,2,4-triazole-3-thiol
scaffolds are also demonstrated to be Hsp90 inhibitors. The virtual screening with
modified AutoDock program could minimize the amount of compounds for actual
experimental screening, while the experimental screening proved the usefulness of
the virtual screening. Both the approach applied to this research and the new
classes of Hsp90 inhibitors identified by this approach could be valuable to drug
discovery.
68
Figure 4.5 Her2 degradation of the compounds that share the scaffolds
MCF-7 cells were treated with indicated concentrations of 17-AAG or each three of
pyrimidine-2,4,6-trione (1a, 1b, and 1c) and 4H-1,2,4-triazole-3-thiol (2a, 2b, and 2c)
compounds. DMSO was treated as a negative control (C), while 17-AAG was used as
a positive control. The protein levels of Her2 and Hsp70 were detected by western blot
analysis. β-tubulin was detected as a loading control.
Chapter 5.
Screening of PP5 inhibitors that act
through different mechanisms
70
5.1. Introduction
Phosphorylation is a crucial molecular switch that is involved in a wide spectrum
of signaling pathways. Protein phosphatases and protein kinases are the key players
that regulate the phosphorylation states of the substrates. Protein phosphatase 5
(PP5) is a unique phosphatase of PPP family, that contains auto-inhibitory N-
terminal TPR domain and C-terminal subdomain. PP5 has been implicated in
cancer, Alzheimer’s disease and obesity, which highlights it as a potential drug
target. The aim of this research was to identify PP5 inhibitors which act in novel
mechanisms, distinct from that of the existing PP5 inhibitors.
5.2. Screening of LOPAC1280 by PP5 phosphatase assay
PP5 phosphatase assay system for the screening was set by the conditions
described in materials and methods. As PP5 has been implicated in cancer, the
screening of LOPAC1280 library was performed primarily based on the data of the
compounds that inhibited BT-474 growth (data not shown). All of the compounds
from several racks containing some BT-474 growth inhibitors were inspected by
the assay. These were total 5 racks, approximately 400 compounds from the
LOPAC1280. The primary screening identified several compounds that
downregulated the phosphatase activity of PP5 for the artificial substrate pNPP. In
cotrast, some compounds upregulated PP5. There were known inhibitors of PP5
such as cantharidin, cantharidic acid and known activator such as oleic acid. This
71
proved the system was valid and working well. After the primary screening
covered over half of the BT-474 growth inhibitors, all of the compounds assayed
were subjected to the selection which was based on more than 50 % decrease or
increase of the activity compared to DMSO control (Figure 5.1). The resulting
compounds are depicted in Figure 5.2.
5.3. Confirmation of the primary screening
As the primary screening was performed in a singlet manner, the efficacies of the
compounds identified by the primary screening were confirm by secondary
screening executed in a triplicate manner (Figure 5.2). Cantharidin and oleic acid
were included as a known inhibitor and an activator of PP5, respectively. Some
compounds such as leflunomide and SCH-202676 hydrobromide were shown to be
false positive, while other compounds were confirmed for their modulatory effects
on PP5 phosphatase activity. Ro 90-7501, aurothioglucose, and N-oleoyldopamine
were identified as PP5 inhibitors. Palmitoyl-DL-carnitine chloride was shown to be
an activator of PP5, probably due to its palmitoyl moiety. Among the finally
confirmed inhibitors, aurothioglucose was included in the BT-474 growth inhibitor
list, showing enriched hit rate to some extent (approx. 0.75 % vs. 4.2 %).
5.4. Mechanism study of the screened compounds
To elucidate the mode of action, three inhibitors identified by the screening were
72
subjected to the assay with several PP5 constructs (Figure 5.3). PP5, the full-length
protein contains all of the three domains, the regulatory TPR domain, the C-
terminal subdomain, and the phosphatase domain. PP5ΔC is a construct absent of
the C-terminal 13 amino acids which participates in the auto-inhibition with the
TPR domain and is responsible for lipid activation [68]. PD is the phosphatase
domain, a construct absent of the auto-inhibitory TPR domain, while it retains the
C-terminal subdomain containing the αJ helix. Finally, PDΔC is a construct absent
of both regulatory regions, TPR domain and C-terminal 13 amino acids. First, to
test the assay system of the PP5 constructs, cantharidin which is the known
inhibitor of PP5, and oleic acid which is the known activator of PP5 were applied
to the assay system. As an inhibitor that targets the catalytic domain, cantharidin
inhibited all four PP5 constructs which contain the catalytic domain. In contrast,
oleic acid activated only the full-length PP5 (Figure 5.4). As the auto-inhibitory
TPR domain and C-terminal subdomain are also responsible for activation response
[68], loss of either part results in no response upon activating signal. As the assay
system of the PP5 constructs could identify the regulatory mechanisms of
cantharidin and oleic acid, the three inhibitors identified from LOPAC1280 library
were investigated by the assay system (Figure 5.5). Two of the three inhibitors,
aurothioglucose, and N-oleoyldopamine, showed similar inhibition trends on each
PP5 construct, demonstrating that these compounds exert their inhibitory effect on
the phosphatase domain itself. In contrast, the remaining one, Ro 90-7501,
exhibited a different manner of inhibition. This compound required the TPR
73
domain for efficient inhibition of PP5 phosphatase activity. Further experiment to
observe whether the compounds show concentration- dependent and TPR-
dependent inhibition on PP5 or PD was performed with Ro-90-7501 and N-
oleoyldopamine which had stronger inhibitory effect on all of the four PP5
constructs than aurothioglucose (Figure 5.5). N-oleoyldopamine inhibited both PP5
and PD in a concentration-dependent manner (Figure 5.6, square). In contrast,
while Ro 90-7501 inhibited PP5 concentration-dependently (Figure 5.6, closed
circle), it could not inhibit PD efficiently even with increasing concentrations
(Figure 5.6, open circle). These results show that these two compounds inhibit PP5
in different mechanisms, revealing a novel PP5 inhibitory mechanism of Ro 90-
7501.
74
Figure 5.1 Application of known inhibitor and activator to PP5 constructs
To test the assay system of the PP5 constructs, a known phosphatase inhibitor
cantharidin and oleic acid which has been demonstrated to activate PP5 were
applied to each construct of PP5. The values are normalized to the OD405 values of
the DMSO-treated controls of each construct.
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
- + - + - + - + - + - + - + - +
PP5 PP5ΔC PD PDΔC PP5 PP5ΔC PD PDΔC
Cantharidin Oleic acid
Fo
ld a
cti
vit
y
75
Figure 5.2 Primary screening of the compounds from LOPAC1280
Approximately 400 compounds were tested for their modulatory effects on the
phosphatase activity of PP5. The concentration of each compound was 20 μM.
Fold activity was calculated by normalizing the OD405 values of each sample to
that of DMSO-treated controls placed in each column of the 96-well plate.
Compounds which showed more than 50 % variation of PP5 activity were selected
for the secondary confirmation assay.
0
50
100
150
200
250
0 50 100 150 200 250 300 350 400
Fo
ld a
cti
vit
y (
%)
Compound number
76
Figure 5.3 Secondary confirmation assay of the screened compounds
Compounds that showed more than 50 % variation of PP5 activity were subjected
to the secondary confirmation screening. The rack numbers, well positions from
the LOPAC1280 library and the identities of each compound are shown below.
0
50
100
150
200
250
300
Control R4A9 R9A10 R9C11 R9E6 R14B9 R14D9 R12B4 R12G9
Fo
ld a
cti
vit
y (
%)
R4A9 R9A10 R9C11 R9E6 R14B9 R14D9 R12B4 R12G9
Cantharidin Leflunomide Ro 90-7501Aurothio-glucose
N-oleoyl-dopamine
SCH-202676 hydrobromide
Oleic AcidPalmitoyl-DL-
Carnitine chloride
77
Figure 5.4 PP5 truncation constructs for mechanism study
The four constructs of human PP5 (~ 59.3 kDa) were prepared to identify the
inhibition mechanisms of the selected compounds. PP5 is the full-length protein
and PP5ΔC (~ 57.9 kDa) represents the construct which the C-terminal 13 amino
acids are deleted. PD (~ 40.1 kDa) is the phosphatase domain which is absent of
the TPR domain. PDΔC (~ 38.6 kDa) represents the construct that the C-terminal
13 amino acids are deleted from the PD. The purified PP5 constructs are visualized
by western blot analysis.
78
Figure 5.5 Application of known inhibitor and activator to PP5 constructs
To test the assay system of the PP5 constructs, a known phosphatase inhibitor
cantharidin and oleic acid which has been demonstrated to activate PP5 were
applied to each construct of PP5. The values are normalized to the OD405 values of
the DMSO-treated controls of each construct.
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
- + - + - + - + - + - + - + - +
PP5 PP5ΔC PD PDΔC PP5 PP5ΔC PD PDΔC
Cantharidin Oleic acid
Fo
ld a
cti
vit
y
79
Figure 5.6 Application of the selected inhibitors to PP5 constructs
The three PP5 inhibitors identified by the screening process were applied to each
PP5 construct. These three compounds are Ro 90-7501, aurothioglucose, and N-
oleoyldopamine. The values are normalized to the OD405 values of the DMSO-
treated controls of each construct.
0
0.2
0.4
0.6
0.8
1
1.2
- + - + - + - + - + - + - + - + - + - + - + - +
PP5 PP5ΔC PD PDΔC PP5 PP5ΔC PD PDΔC PP5 PP5ΔC PD PDΔC
Ro 90-7501 Aurothioglucose N-oleoyldopamine
Fo
ld a
cti
vit
y
80
Figure 5.7 Ro 90-7501 exhibits novel PP5 inhibitory mechanism
The two PP5 inhibitors, Ro 90-7501 and N-oleoyldopamine which showed
different inhibitory aspects on the PP5 constructs (Figure 5.5) were further
examined for their concentration-dependent, TPR-dependent effects on PP5 and
PD.
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
0 20 40 60
Acti
vit
y (
OD
40
5)
Concentration (μM)
Ro 90-7501 / PP5
Ro 90-7501 / PD
N-oleoyldopamine / PP5
N-oleoyldopamine / PD
81
5.5. Conclusions
In an attempt to find PP5 inhibitors within the pharmacologically active
compounds led to the discovery of two PP5 inhibitors which inhibit PP5 in a
similar manner with the known phosphatase inhibitors such as cantharidin. In
contrast, one PP5 inhibitor, Ro 90-7501 inhibited PP5 in a different manner
compared to the other inhibitors. Unlike other inhibitors that target the phosphatase
domain of PP5 directly, Ro 90-7501 showed TPR domain-dependent inhibition on
PP5 phosphatase activity. These results demonstrate that Ro 90-7501 acts through a
novel TPR-dependent inhibitory mechanism on PP5, which implies a possibility
that Ro 90-7501 might be able to inhibit the basal activity of PP5 specifically over
the phosphatases of PPP family.
Chapter 6.
Enhancing the solubility of NOD1-LRR in
E. coli
83
6.1. Introduction
NOD1 is an intracellular innate immunity sensor whose polymorphisms are related
to several immune-related, inflammatory diseases. This spotlights NOD1 as a
potential drug target, but the purification of the protein itself has been proven to be
notoriously difficult [22]. This significantly limited signaling mechanism or HTS
researches of NOD1 into the scale of cell-based system for a long time. The
productions of NOD1 protein by cell-free synthesis or mammalian cell expression
enabled the molecular level confirmations of the signaling mechanism model of
NOD1 derived from cell-based researches [101, 103] (Figure 2.5B). But these
protein expression systems are not cost-effective to produce sufficient amount of
the protein for crystallography or drug screening researches (Table 2.4). The HTS
assays developed for NOD1 inhibitors still remain as cell-based systems [22, 111].
To date, there have been mainly two successful cases of NOD1 purification.
First, N-terminally GST-tagged full-length NOD1 protein have been produced with
in vitro wheat germ expression system by Novus biologicals, who claims that their
protein has not been tested for any functionality [112]. But the research performed
with their protein demonstrated that the protein retains the abilities of ligand
binding and recruiting downstream signaling kinase, RIP2 [103]. Another NOD1
production was achieved by suspension culture of NOD1 animal cell line. This
approach required multiple infections to increase the copy number of NOD1 gene,
thus improving the expression level of NOD1 protein. Several properties of NOD1
84
that were not demonstrated in the first case, such as ligand (γ-Tri-DAP)-stimulated
oligomerization, nucleotide (ATP)-binding, interaction with Bid, could be shown
with the purified protein [101].
Although NOD1 protein could be expressed and purified utilizing cell-free
expression system and mammalian expression system, there has been no report of
soluble NOD1 protein expression in E. coli system. Compared to the other two
methods mentioned above, E. coli system has unsurpassed advantages in cost,
productivity and robustness (Table 2.4). But, in the preliminary research of NOD1
expression in E. coli, full-length NOD1 protein was shown to be extremely
unstable and aggregation-prone. Conditional optimizations or chaperone
coexpressions could not increase the solubility of the full-length protein
significantly. So, the target region of NOD1 protein for soluble expression in E.
coli was reduced to the LRR domain of NOD1, because this domain is known to be
the ligand-binding domain which is essential for the study of NOD1 activation
mechanism and NOD1 inhibitor screening. Unfortunately, NOD1-LRR domain
expressed in E. coli also showed aggregation-prone propensity like the full-length
protein. A new strategy had to be considered for the soluble expression of NOD1-
LRR domain in E. coli. Single amino acid substitutions are known to have
profound impact on the protein solubility [171, 172], but this approach was not
considered because amino acid substitutions might compromise the ligand-binding
activity of NOD1-LRR domain. In the case of TLRs, it has been reported that the
hybrid LRR technique which fuse the LRRNT or LRRCT modules of the hagfish
85
VLRs to the TLRs, could improve the solubility and crystallization of TLRs [108].
But unlike TLRs, intracellular NOD1 and NOD2 are known to be absent of
LRRNT or LRRCT. As the unnatural fusion of LRRNT or LRRCT to NOD1-LRR
domain could also lead to the alteration of its structural integrity and functionality,
this method was also ruled out. For a similar reason, refolding of the aggregated
protein [91] was not chosen. Instead, an N-terminally extended construct of NOD1-
LRR domain which includes the N-terminal region that might function as an
LRRNT, was adopted and tested for its solubility and expression level. In general,
some commonly used fusion partners, such as MBP, GST, NusA, are tested for
their ability to enhance the solubility of aggregation-prone protein of interest [173].
But the choice for these fusion partners requires a trial and error approach. Instead,
SGT1, the natural interacting partner of NOD1, and stabilizing chaperone Hsp90
were chosen as the fusion and coexpression partners of NOD1-LRR, respectively.
The rationale of this approach is based on the report that SGT1 was required for the
activation of NOD1, while Hsp90 was needed for the stability of NOD1 in human
cell line [24]. Also, it is supported by the researches on the function of Hsp90-
SGT1 chaperone complex which has been demonstrated to be crucial for NLRs of
both plants and mammals [128, 174].
6.2. Expression of full-length NOD1 protein in E. coli
At first, the expression of full-length NOD1 protein in E. coli was attempted. But
full-length NOD1 protein exhibited aggregation and degradation-prone expression
86
properties (Figure 6.1). The coexpression of HtpG, the E. coli Hsp90, had limited
effect on the solubility of NOD1 protein, while the additional coexpression of E.
coli chaperones GroEL, GroES, DnaK, DnaJ, GrpE (pG-KJE8 vector; Takara) with
HtpG showed small increase of NOD1 protein in the soluble fraction (Figure 6.1).
As a general approach, optimization of IPTG concentration was tried, but this also
showed limited effect on the solubility of full-length NOD1 protein (Figure 6.2).
6.3. Expression of NOD1-LRR domain in E. coli
As the solubility of full-length NOD1 protein was not increased significantly by the
coexpression of chaperones, the target construct for expression was narrowed down
to the LRR domain which is the ligand binding region of the NOD1 protein. The
coexpression of either E. coli Hsp90 (HtpG) or mouse Hsp90 (MmHsp90α) and
additional E. coli chaperones with the NOD1-LRR domain had limited effect on
the solubility of NOD1-LRR domain (Figure 6.3).
87
Figure 6.1 NOD1 full-length protein expression with chaperones
NOD1 full-length protein (~ 109 kDa, pET15b) was expressed in E. coli strain
BL21 star(DE3) either alone, or with HtpG (E. coli Hsp90, ~ 71 kDa, pET24ma), or
with HtpG and E. coli chaperones GroEL, GroES (GroELS), DnaK, DnaJ, and
GrpE (DnaKJE) (pG-KJE8). The solubility of the protein was analyzed by western
blotting.
T: total protein, S: soluble fraction, I: insoluble fraction
88
6.4. Expression of SGT1NOD1LRR fusion protein in E. coli
SGT1 is an adaptor protein that recruit and deliver LRR-containing proteins to
Hsp90 [119]. The coexpression of interacting partners of aggregation-prone target
proteins has been demonstrated to help solubilizing the protein [175, 176]. Also, it
is known that the expression properties of a fusion protein is dictated by the N-
terminally fused protein [177, 178]. Finally, there is a report that NOD1 requires
SGT1 for its activity and Hsp90 for its stability in human cell line [24]. So, as a
binding partner of NOD1-LRR domain, SGT1 was selected for the fusion partner.
Both human and yeast SGT1 were chosen as the N-terminal fusion part, because
these proteins exhibited expression levels and solubilities which were high enough
for purification (data not shown). The linker between SGT1 and NOD1-LRR
domain contained a thrombin site sequence, and a 6ÍHis tag sequence was
attached to the C-terminal end of the fusion protein for the purification of NOD1-
LRR domain, if possible. The expression and the solubility of the SGT1NOD1LRR
fusion protein in E. coli were observed by western blot analysis. Also, as
mentioned above, Hsp90 which NOD1 protein requires for its stability in human
cell line was coexpressed by utilizing pETDuet-1 vector.
The result shows that HsSGT1NOD1LRR fusion protein, the fusion of human
SGT1 to NOD1-LRR domain led to a significant increase in the ratio of the fusion
protein in the soluble fraction to that in the insoluble fraction (Figure 6.4A). The
ratios of the soluble band to the insoluble band of Figure 6.3, 6.4A, and 6.9 were
89
calculated by ImageJ program and the enhancement of the solubility by this factor
was estimated to be about 4 to more than 50 fold, depending on the samples chosen
for comparison. In contrast, ScSGT1NOD1LRR fusion protein, the fusion of yeast
SGT1 to NOD1-LRR domain resulted in mostly insoluble expression of the fusion
protein (Figure 6.4B). This is probably due to the dimerization of yeast SGT1
through its TPR domain [127] which might put the highly aggregating NOD1LRR
domains into close proximity, leading the aggregation of the whole fusion protein.
The coexpression of prokaryotic Hsp90, E. coli HtpG with ScSGT1NOD1LRR
fusion protein could not lead the soluble expression of the fusion protein (Figure
6.4B). On the other hand, the coexpression of eukaryotic mouse Hsp90α with
HsSGT1NOD1LRR fusion protein stabilized the fusion protein (Figure 6.4A, 6.5),
in accordance with the previous report that Hsp90 is required for the stability of
NOD1 protein in human cell line [24].
Even though the solubility of NOD1-LRR domain was enhanced by the fusion
with HsSGT1 and the stability of the fusion protein could be improved by the
coexpression of MmHsp90α, the fusion protein showed relatively low expression
level and highly degrading propensities (Figure 6.4A, 6.5). These drawbacks were
considered to result from the plasmid instability, mRNA instability, translation pre-
termination by rare codons or structural vulnerability of the fusion protein to
proteolysis. To overcome these problems, each of the expected causes was
inspected. As the ampicillin resistance gene product, β-lactamase can be secreted
into the culture medium and degrade the ampicillin which confers selection
90
pressure for the expression vector, causing the plasmid instability, the antibiotic
marker of the pETDuet-1 vector was exchanged to kanamycin resistance gene. The
result shows that the exchange of the selection marker has some positive effect on
the protein expression level, but the fusion protein still exhibited the same
degrading pattern (Figure 6.6). To test whether the cause of this degradation of the
fusion protein could be mRNA instability, E. coli strain BL21 star(DE3) was used
for the experiment. This strain has a mutated RNase E which is the major endo-
ribonuclease responsible for mRNA turnover in E. coli [113]. The expression level
of the fusion protein was lower than that of the Rosetta gami2(DE3)pLysS strain
and exhibited similar degrading pattern (Figure 6.7). This implies that mRNA
instability might not be the reason for the degradation of the fusion protein. Next,
the effect of rare codon tRNA supplementation was tested by the addition of
chloramphenicol, the selection marker of pRARE2LysS vector in the Rosetta
gami2(DE3)pLysS strain. This approach increased the expression level of the
fusion protein significantly, while the protein still suffered the same degradation
(Figure 6.8). Finally, the extended NOD1-LRR construct (first construct) was
thought to be the reason of structural vulnerability and degradation by proteolysis.
The NOD1-LRR construct used for the fusion with SGT1 has about 15 more amino
acids than the original NOD1-LRR construct (second construct). These amino
acids are fused to the linker, probably extending the proteolytically vulnerable
region. So the NOD1-LRR construct in the fusion protein was reduced to the
original construct. This second construct, similar to the first one, was mostly
91
soluble and its stability was enhanced by coexpressing MmHsp90α, but the
degradation pattern was essentially similar to the first construct (Figure 6.9). By the
overall efforts, the expression level of the HsSGT1NOD1LRR fusion protein under
the coexpression of mouse Hsp90 could be enhanced to some extent, but the major
proteolytic degradation could not be alleviated.
The Hsp90 is a well-conserved protein and the domain structures of Hsp90s
from E. coli, yeast, mammal are essentially identical, but the mechanisms by which
each Hsp90 proceeds through the chaperone cycle seem to be quite different [44].
Eukaryotic Hsp90s are known to require cochaperones to advance through the
chaperone cycle. Also, mammalian Hsp90 exhibits low ATPase activity without
the cochaperone Aha1 [179]. In contrast, E. coli HtpG is highly ATP-dependent
and does not require cochaperones (which do not exist in E. coli) for the
progression through the chaperone cycle. While, E. coli HtpG is dispensable and its
exact function is still enigmatic, eukaryotic Hsp90s are essential and participate in
various critical cellular functions. There is a possibility that eukaryotic Hsp90s may
not function properly in E. coli environment which is absent of Hsp90
cochaperones. On the other hand, as HsSGT1NOD1LRR fusion protein is a fusion
of two human proteins, eukaryotic Hsp90 or mammalian Hsp90 might stabilize the
fusion protein more efficiently. The difference in the expression levels of each
Hsp90s in E. coli could also influence their efficacies. Finally, there could be some
differences between the Hsp90 isoforms for their abilities to activate certain client
proteins [180].
92
To investigate the effects of several Hsp90s from different species and different
paralogs, each of E. coli HtpG, mouse Hsp90α, yeast Hsp82 and Hsc82 was
coexpressed respectively with the HsSGT1NOD1LRR fusion protein containing
the second construct of NOD1-LRR. To detect the expression levels of each
Hsp90s, T7 tag was added to the N-terminal end of each Hsp90 construct by
utilizing pETDUET-KTM vector (Materials and methods). The effect of each
Hsp90 coexpression on the expression level of HsSGT1NOD1LRR protein at each
time point were compared in a pairwise manner by western blotting on a single blot.
The coexpression of HtpG showed higher expression level of the
HsSGT1NOD1LRR fusion protein than that of Hsc82 coexpression (Figure 6.10),
while the coexpression of mouse Hsp90 exhibited enhanced expression level of the
fusion protein compared to that of Hsp82 coexpression (Figure 6.11). Finally, in
the comparison of the fusion protein level between the coexpression of HtpG and
mouse Hsp90 revealed that mouse Hsp90 stabilizes the HsSGT1NOD1LRR fusion
protein most efficiently (Figure 6.12).
93
Figure 6.2 Optimization of IPTG concentration for soluble NOD1 expression
NOD1 full-length protein expression was induced by varying concentrations of
IPTG. HtpG and E. coli chaperones (GroEL, GroES, DnaK, DnaJ, GrpE) were
coexpressed with NOD1. The solubility of the protein was analyzed by western
blotting.
T: total protein, S: soluble fraction, I: insoluble fraction.
94
Figure 6.3 Solubility of NOD1-LRR domain coexpressed with chaperones
The NOD1-LRR domain (pET15b) was coexpressed in BL21 star(DE3) with either
(A) E. coli HtpG or (B) mouse Hsp90α (pET24ma) and E. coli chaperones (pG-
KJE8) together. The solubility of the protein was analyzed by western blotting.
T: total protein, S: soluble fraction, I: insoluble fraction
95
Figure 6.4 Solubility of SGT1NOD1LRR protein with or without Hsp90
The solubility of SGT1NOD1LRR fusion protein, either fused with (A) human
SGT1 (pETDuet-1) or (B) yeast SGT1 (pETDUET-KTM), expressed with or
without mouse Hsp90 or E. coli Hsp90 (HtpG) was analyzed by immunoblotting.
T: total protein, S: soluble fraction, I: insoluble fraction
96
Figure 6.5 Expression of SGT1NOD1LRR with or without mouse Hsp90
The SGT1NOD1LRR fusion protein which is the NOD1-LRR domain N-
terminally fused with human SGT1, was expressed with or without MmHsp90α
(pETDuet-1).
A. The expression level of HsSGT1NOD1LRR at each time point indicated.
B. The expression level of HsSGT1NOD1LRR coexpressed with mouse Hsp90
at each time point indicated. Upper arrowhead on the side of Coomassie
Brilliant Blue (CBB) stained gel indicates the size of mHsp90, while the
lower arrowhead indicates the size of SGT1NOD1LRR.
97
Figure 6.6 Effect of antibiotic markers on SGT1NOD1LRR expression
The antibiotic marker of pETDuet-1 vector (Amp) was replaced with kanamycin
resistance gene (Kan) from pET28b, generating pETDUET(KanR). The effect of the
marker exchange on the HsSGT1NOD1LRR protein level under the coexpression of
mouse Hsp90 was observed at each time point indicated. The left part (Amp) of this
figure is the identical blot of figure 6.5B.
98
Figure 6.7 Effect of E. coli strains on SGT1NOD1LRR expression
In an attempt to assess the possible mRNA instability of HsSGT1NOD1LRR, BL21
star(DE3) which has a mutated RNase E, was applied to the expression of the fusion
protein under the coexpression of mouse Hsp90 (pETDuet-1) and was compared
with that of Rosetta gami2(DE3)pLysS at each time point indicated.
99
Figure 6.8 Effect of rare codon tRNAs on SGT1NOD1LRR expression
The addition of chloramphenicol for the maintenance of pRARE2LysS vector
which supplies rare codon tRNAs, enhanced the expression level of
HsSGT1NOD1LRR fusion protein under the coexpression of mouse Hsp90
(pETDUET(KanR)) at each time point indicated.
.
100
Figure 6.9 Effect of LRR constructs on SGT1NOD1LRR expression
The fusion protein with the second NOD1-LRR construct (2nd: 656-953, amino acid
residues of NOD1) was soluble and stabilized by coexpressed mouse Hsp90, similar
to the fusion protein containing the first (1st: 641-953) NOD1-LRR construct.
101
Figure 6.10 Effect of Hsc82 and HtpG on SGT1NOD1LRR expression
The expression levels of HsSGT1NOD1LRR fusion protein under the coexpression
of Hsp90s were compared in a single blot at each time point indicated. Coexpressed
Hsp90s were detected by their N-terminal T7 tags (pETDUET-KTM).
102
6.5. Conclusions
The solubility of NOD1-LRR domain was enhanced by N-terminal fusion of
human SGT1. Even though the severe proteolytic degradation and the low
expression level of the HsSGT1NOD1LRR fusion protein could not be fully
alleviated by either effort such as antibiotic marker exchange, rare codon tRNAs
supplementation, the exchange of LRR construct and Hsp90s coexpression, the
stability and the expression level of the fusion protein could be enhanced to some
extent. Among the coexpressed Hsp90s, mouse Hsp90 was the most efficient
stabilizer for the HsSGT1NOD1LRR fusion protein.
103
Figure 6.11 Effect of mHsp90 and Hsp82 on SGT1NOD1LRR expression
The expression levels of HsSGT1NOD1LRR fusion protein under the coexpression
of Hsp90s were compared in a single blot at each time point indicated. Coexpressed
Hsp90s were detected by their N-terminal T7 tags (pETDUET-KTM).
104
Figure 6.12 Effect comparison of E. coli HtpG and mouse Hsp90 on
SGT1NOD1LRR expression
The expression levels of HsSGT1NOD1LRR fusion protein under the coexpression
of Hsp90s were compared in a single blot at each time point indicated. Coexpressed
Hsp90s were detected by their N-terminal T7 tags (pETDUET-KTM).
Chapter 7.
Overall discussion and recommendations
106
In this dissertation, three aspects of the Hsp90 chaperone machinery were
investigated. These are (1) Hsp90 chaperone, (2) Hsp90 cochaperone, (3) Hsp90
client protein. The aims of the first two parts were the inhibitor screening of Hsp90
chaperone and Hsp90 cochaperone PP5, while the aim of the last part was to utilize
Hsp90 and Hsp90 cochaperone SGT1 for soluble expression of the LRR domain of
Hsp90 client protein NOD1 in E. coli.
In the first part of this dissertation, Hsp90 inhibitors with novel scaffolds were
identified by sequential virtual screening and experimental screening. The virtual
screening significantly narrowed down the scale of candidate compounds library
from HTS-scale into lab-scale, which enabled manual screening of the library,
without HTS setup. As a result, two novel classes of Hsp90 inhibitors containing
scaffolds of pyrimidine-2,4,6-trione and 4H-1,2,4-triazole-3-thiol, respectively,
were identified by this combination of virtual and experimental screening. Other
compounds in the library that share these scaffolds of the newly identified
compounds were also shown to lead Her2 degradation in MCF-7 cells. These
results demonstrate the Hsp90-inhibitory efficacies of the two new scaffolds which
might serve as lead compounds for the discovery of potent Hsp90 inhibitors with
favorable properties for anti-cancer drug. Also, with the aid of virtual screening,
drug discovery can be performed in a more cost-effective, efficient way.
In the second part, PP5 inhibitors were identified by screening the compounds
of LOPAC1280 library which consists of pharmacologically active compounds
including marketed drugs and pharmaceutically relevant structures with known
107
biological activities. The catalytic regions of the PPP family phosphatases,
including PP5, share significant similarity in structure and catalytic mechanism.
This result in the broad selectivity of the known natural PP5 inhibitors over the
PPP family phosphatases (Table 2.3). The broad selectivity of an inhibitor is an
undesirable attribute for drug candidates because it can cause unwanted side-effects.
These natural PP5 inhibitors target the catalytic site of PP5 which is structurally
similar to those of other phosphatases of PPP family [17, 181-184]. There are also
structural similarities between auto-inhibited PP5 and other inhibitor-bound PPP
family members [65]. These structural similarities of the catalytic sites of PPP
family phosphatases imply that selective PP5 inhibitors might need to be allosteric
or target the unique feature of PP5, the auto-inhibitory TPR domain. In an attempt
to identify PP5-specific inhibitors, screening assays were performed with the
compounds from LOPAC1280. Three compounds were identified as PP5 inhibitors
while one of these inhibitors, Ro 90-7501 exhibited TPR-dependent inhibition on
PP5 which could be a prerequisite property for PP5-specific phosphatase inhibitor.
It has been suggested that the domain interface of the TPR domain and the
phosphatase domain of PP5 retains flexibility [65]. This compound might exert
its inhibitory effect on PP5 by binding to the TPR domain and tightening the auto-
inhibited conformation of PP5. Although Ro 90-7501 showed its inhibitory effect
on PP5 in a TPR-dependent manner, the specificity toward PP5 needs to be verified
by observing its inhibitory effects on other phosphatase members of the PPP family.
It has been reported that Ro 90-7501 efficiently retarded mature fibril formation of
108
amyloid-β which is considered responsible for Alzheimer’s disease [185]. Another
research showed that Ro 90-7501 enhanced TLR/RLR agonist induced antiviral
response [186]. They suggested that this process is likely via selective activation of
p38 MAPK pathway. It is interesting that PP5 is implicated in both Alzheimer’s
disease and MAPK pathway [4, 5, 187]. It has been shown that PP5 can be
activated by fatty acids, such as arachidonic acid and oleic acid which bind to the
TPR domain and liberate the phosphatase domain from the TPR-bound, auto-
inhibited state [66, 188]. There are HTS reports which identified PP5 activators
that have different mechanisms of action from that of arachidonic acid [82, 83]. In
contrast to the PP5 activators, PP5 inhibitors that act through distinct mechanisms
from that of the known inhibitors have not been reported. To my knowledge, Ro
90-7501 is the first PP5 inhibitor which acts in a new mechanism, TPR-dependent
inhibition. The identifications of these PP5 modulators that act through novel
mechanisms, imply that there can be compounds that regulate PP5 activity through
diverse mechanisms. If the specificity of Ro 90-7501 on PP5 could be confirmed, it
could serve not only as a tool for researches on PP5 but also as a potential
therapeutic agent for cancer and obesity.
In the third part of this dissertation, efforts for soluble expression of NOD1-
LRR domain in E. coli was made. As NOD1 is known to be notoriously hard to
express and purify, both NOD1 full-length protein and the LRR domain were
mainly expressed as insoluble state. Instead of applying conventional solubilizing
fusion tags in a trial-and-error manner, a fusion partner and a coexpression partner
109
were selected based on the research which reported that SGT1 was essential for
NOD1 activation, while Hsp90 was required for NOD1 stability in human cell line
[24]. In accordance with this report, the fusion of human SGT1 to NOD1-LRR
domain led to the soluble expression of HsSGT1NOD1LRR fusion protein, and the
coexpression of Hsp90 enhanced the stability of the fusion protein. The fusion
protein suffered degradation and low expression level, showing it is proteolysis-
prone in E. coli. Coexpression of Hsp90s from different species with the fusion
protein demonstrated that the most related mouse Hsp90 stabilized human
SGT1NOD1LRR fusion protein most efficiently. The next most efficient
stabilizing coexpression partner of the fusion protein was E. coli HtpG which is the
most related to the expression host. In contrast to human SGT1NOD1LRR fusion
protein, yeast SGT1NOD1LRR exhibited low solubility. This is probably due to
the dimerization of yeast SGT1 through its TPR domain. The dimerization of the
fusion protein can lead the LRR domains into a close proximity which may
enhance their aggregation. For further investigation, TPR-deletion construct of
yeast SGT1 may be chosen for the fusion partner of NOD1-LRR. These results
imply that the rationale used in this experiment which is the N-terminal fusion with
SGT1 and the coexpression of Hsp90, can be applied to other NLR proteins that
require SGT1-Hsp90 chaperone complex for their activation and stabilization [24,
128]. Also, this approach can be tested with different expression hosts, because the
expression level of the fusion protein was low in E. coli. The enhanced solubility
and stability of NLRs or LRR-containing proteins, by the fusion with SGT1 and the
110
coexpression of Hsp90, could facilitate the purification and biochemical researches
of the protein of interest.
111
Bibliography
1. Neckers, L. and P. Workman, Hsp90 Molecular Chaperone Inhibitors: Are We There Yet?
Clin Cancer Res, 2012. 18(1): p. 64-76.
2. Katerina, S. and P. Evangelia, HSP90 Inhibitors: Current Development and Potential in
Cancer Therapy. Recent Pat Anticancer Drug Discov, 2014. 9(1): p. 1-20.
3. Tatokoro, M., et al., Heat shock protein 90 targeting therapy: state of the art and future
perspective. EXCLI J, 2015. 14: p. 48-58.
4. Liu, F., et al., Dephosphorylation of Tau by Protein Phosphatase 5. J Biol Chem, 2005.
280(3): p. 1790-1796.
5. Gong, C.-X., et al., Dephosphorylation of microtubule-associated protein tau by protein
phosphatase 5. J Neurochem, 2004. 88(2): p. 298-310.
6. Zheng, X., et al., Knockdown of protein phosphatase 5 inhibits ovarian cancer growth.
Oncol Lett, 2016. 11(1): p. 168-172.
7. Wang, J., et al., Inhibition of protein phosphatase 5 (PP5) suppresses survival and
growth of colorectal cancer cells. Biotechnol Appl Biochem, 2015. 62(5): p. 621-627.
8. Urban, G., et al., Identification of an Estrogen-inducible Phosphatase (PP5) That
Converts MCF-7 Human Breast Carcinoma Cells into an Estrogen-independent
Phenotype when Expressed Constitutively. J Biol Chem, 2001. 276(29): p. 27638-
27646.
9. Golden, T., et al., Elevated levels of Ser/Thr protein phosphatase 5 (PP5) in human
breast cancer. Biochim Biophys Acta - Mol Basis Dis, 2008. 1782(4): p. 259-270.
10. Han, X., et al., Curcumin inhibits protein phosphatases 2A and 5, leading to activation
of mitogen-activated protein kinases and death in tumor cells. Carcinogenesis, 2012.
33(4): p. 868-875.
11. Golden, T., et al., Constitutive over expression of serine/threonine protein phosphatase 5
(PP5) augments estrogen-dependent tumor growth in mice. Cancer Lett, 2004. 215(1): p.
95-100.
12. Grankvist, N., et al., Genetic disruption of protein phosphatase 5 in mice prevents high-
fat diet feeding-induced weight gain. FEBS Lett, 2013. 587(23): p. 3869-3874.
13. Jacob, W., et al., Decreased adipogenesis and adipose tissue in mice with inactivated
protein phosphatase 5. Biochem J, 2015. 466(1): p. 163-176.
14. Hinds, T.D., et al., Protein Phosphatase 5 Mediates Lipid Metabolism through
Reciprocal Control of Glucocorticoid Receptor and Peroxisome Proliferator-activated
Receptor-γ (PPARγ). J Biol Chem, 2011. 286(50): p. 42911-42922.
112
15. Feng, L., et al., Knockdown of PPP5C Inhibits Growth of Hepatocellular Carcinoma
Cells In Vitro. Appl Biochem Biotechnol, 2015. 175(1): p. 526-534.
16. Zuo, Z., et al., Ser/Thr Protein Phosphatase Type 5 (PP5) Is a Negative Regulator of
Glucocorticoid Receptor-Mediated Growth Arrest. Biochemistry, 1999. 38(28): p. 8849-
8857.
17. Swingle, M., L. Ni, and R. Honkanen, Small-Molecule Inhibitors of Ser/Thr Protein
Phosphatases, in Protein Phosphatase Protocols, G. Moorhead, Editor. 2007, Springer
New York. p. 23-38.
18. Sleigh, S.H. and C.L. Barton, Repurposing Strategies for Therapeutics. Pharm Med,
2010. 24(3): p. 151-159.
19. Würth, R., et al., Drug-repositioning opportunities for cancer therapy: novel molecular
targets for known compounds. Drug Discov Today, 2016. 21(1): p. 190-199.
20. Motta, V., et al., NOD-Like Receptors: Versatile Cytosolic Sentinels. Physiol Rev., 2015.
95(1): p. 149-178.
21. Philpott, D.J., et al., NOD proteins: regulators of inflammation in health and disease.
Nat Rev Immunol, 2014. 14(1): p. 9-23.
22. Rickard, D.J., et al., Identification of Selective Small Molecule Inhibitors of the
Nucleotide-Binding Oligomerization Domain 1 (NOD1) Signaling Pathway. PLoS One,
2014. 9(5): p. e96737.
23. Zhang, M., et al., Structural Basis for Assembly of Hsp90-Sgt1-CHORD Protein
Complexes: Implications for Chaperoning of NLR Innate Immunity Receptors. Mol
Cell, 2010. 39(2): p. 269-281.
24. da Silva Correia, J., et al., SGT1 is essential for Nod1 activation. Proc Natl Acad Sci
USA, 2007. 104(16): p. 6764-6769.
25. Azevedo, C., et al., Role of SGT1 in resistance protein accumulation in plant immunity.
EMBO J, 2006. 25(9): p. 2007-2016.
26. Kadota, Y., K. Shirasu, and R. Guerois, NLR sensors meet at the SGT1-HSP90
crossroad. Trends Biochem Sci, 2010. 35(4): p. 199-207.
27. Chang, H.-C., et al., SnapShot: Molecular Chaperones, Part I. Cell, 2007. 128(1): p.
212.e1-212.e2.
28. Tang, Y.-C., et al., SnapShot: Molecular Chaperones, Part II. Cell, 2007. 128(2): p.
412.e1-412.e2.
29. Csermely, P., et al., The 90-kDa molecular chaperone family: structure, function, and
clinical applications. A comprehensive review. Pharmacol Ther, 1998. 79(2): p. 129-168.
30. Hahn, J.S., The Hsp90 chaperone machinery: from structure to drug development. BMB
Rep, 2009. 42(10): p. 623-630.
113
31. Tsutsumi, S. and L. Neckers, Extracellular heat shock protein 90: a role for a molecular
chaperone in cell motility and cancer metastasis. Cancer Sci, 2007. 98(10): p. 1536-
1539.
32. Li, J. and J. Buchner, Structure, function and regulation of the hsp90 machinery.
Biomed J, 2013. 36(3): p. 106-17.
33. Pearl, L.H. and C. Prodromou, Structure and mechanism of the Hsp90 molecular
chaperone machinery. Annu Rev Biochem, 2006. 75: p. 271-94.
34. Prodromou, C., et al., The ATPase cycle of Hsp90 drives a molecular 'clamp' via
transient dimerization of the N-terminal domains. EMBO J, 2000. 19(16): p. 4383-4392.
35. Street, Timothy O., Laura A. Lavery, and David A. Agard, Substrate Binding Drives
Large-Scale Conformational Changes in the Hsp90 Molecular Chaperone. Mol Cell,
2011. 42(1): p. 96-105.
36. D'Andrea, L.D. and L. Regan, TPR proteins: the versatile helix. Trends Biochem Sci,
2003. 28(12): p. 655-662.
37. Zhao, R., et al., Navigating the chaperone network: an integrative map of physical and
genetic interactions mediated by the hsp90 chaperone. Cell, 2005. 120(5): p. 715-27.
38. Picard lab, https://www.picard.ch/downloads/Hsp90interactors.pdf.
39. Graf, C., et al., Spatially and kinetically resolved changes in the conformational
dynamics of the Hsp90 chaperone machine. EMBO J, 2009. 28(5): p. 602-13.
40. Shiau, A.K., et al., Structural Analysis of E. coli hsp90 reveals dramatic nucleotide-
dependent conformational rearrangements. Cell, 2006. 127(2): p. 329-40.
41. Ratzke, C., et al., Dynamics of heat shock protein 90 C-terminal dimerization is an
important part of its conformational cycle. Proc Natl Acad Sci USA, 2010. 107(37): p.
16101-16106.
42. Ratzke, C., F. Berkemeier, and T. Hugel, Heat shock protein 90’s mechanochemical
cycle is dominated by thermal fluctuations. Proc Natl Acad Sci USA, 2012. 109(1): p.
161-166.
43. Mickler, M., et al., The large conformational changes of Hsp90 are only weakly coupled
to ATP hydrolysis. Nat Struct Mol Biol, 2009. 16(3): p. 281-286.
44. Ratzke, C., et al., From a Ratchet Mechanism to Random Fluctuations—Evolution of
Hsp90's Mechanochemical Cycle. J Mol Biol, 2012. 423(3): p. 462–471.
45. Hessling, M., K. Richter, and J. Buchner, Dissection of the ATP-induced conformational
cycle of the molecular chaperone Hsp90. Nat Struct Mol Biol, 2009. 16(3): p. 287-93.
46. Ratzke, C., B. Hellenkamp, and T. Hugel, Four-colour FRET reveals directionality in
the Hsp90 multicomponent machinery. Nat Commun, 2014. 5.
114
47. Scheufler, C., et al., Structure of TPR Domain–Peptide Complexes: Critical Elements in
the Assembly of the Hsp70–Hsp90 Multichaperone Machine. Cell, 2000. 101(2): p.
199-210.
48. Pearl, L.H., C. Prodromou, and P. Workman, The Hsp90 molecular chaperone: an open
and shut case for treatment. Biochem J, 2008. 410(3): p. 439-453.
49. Rohl, A., et al., Hsp90 regulates the dynamics of its cochaperone Sti1 and the transfer of
Hsp70 between modules. Nat Commun, 2015. 6.
50. Richter, K., et al., Sti1 Is a Non-competitive Inhibitor of the Hsp90 ATPase: BINDING
PREVENTS THE N-TERMINAL DIMERIZATION REACTION DURING THE
ATPASE CYCLE. J Biol Chem, 2003. 278(12): p. 10328-10333.
51. Roe, S.M., et al., The Mechanism of Hsp90 regulation by the protein kinase-specific
cochaperone p50(cdc37). Cell, 2004. 116(1): p. 87-98.
52. McLaughlin, S.H., et al., The co-chaperone p23 arrests the Hsp90 ATPase cycle to trap
client proteins. J Mol Biol, 2006. 356(3): p. 746-58.
53. Ali, M.M., et al., Crystal structure of an Hsp90-nucleotide-p23/Sba1 closed chaperone
complex. Nature, 2006. 440(7087): p. 1013-7.
54. Li, J., J. Soroka, and J. Buchner, The Hsp90 chaperone machinery: Conformational
dynamics and regulation by co-chaperones. Biochim Biophys Acta - Mol Cell Res,
2012. 1823(3): p. 624-635.
55. Meyer, P., et al., Structural basis for recruitment of the ATPase activator Aha1 to the
Hsp90 chaperone machinery. EMBO J, 2004. 23(6): p. 1402-10.
56. Lavery, Laura A., et al., Structural Asymmetry in the Closed State of Mitochondrial
Hsp90 (TRAP1) Supports a Two-Step ATP Hydrolysis Mechanism. Mol Cell, 2014.
53(2): p. 330-343.
57. Li, J., K. Richter, and J. Buchner, Mixed Hsp90–cochaperone complexes are important
for the progression of the reaction cycle. Nat Struct Mol Biol, 2011. 18(1): p. 61-66.
58. Retzlaff, M., et al., Asymmetric Activation of the Hsp90 Dimer by Its Cochaperone
Aha1. Mol Cell, 2010. 37(3): p. 344-354.
59. Vaughan, C.K., et al., Structure of an Hsp90-Cdc37-Cdk4 Complex. Mol Cell, 2006.
23(5): p. 697-707.
60. Mollapour, M., et al., Asymmetric Hsp90 N Domain SUMOylation Recruits Aha1 and
ATP-Competitive Inhibitors. Mol Cell, 2014. 53(2): p. 317-329.
61. Mishra, P. and Daniel N.A. Bolon, Designed Hsp90 Heterodimers Reveal an
Asymmetric ATPase-Driven Mechanism In Vivo. Mol Cell, 2014. 53(2): p. 344-350.
62. Chinkers, M., Protein phosphatase 5 in signal transduction. Trends Endocrinol Metab,
2001. 12(1): p. 28-32.
115
63.Hinds, T.D., Jr. and E.R. Sanchez, Protein phosphatase 5. Int J Biochem Cell Biol, 2008.
40(11): p. 2358-2362.
64. Das, A.K., P.W. Cohen, and D. Barford, The structure of the tetratricopeptide repeats of
protein phosphatase 5: implications for TPR-mediated protein-protein interactions.
EMBO J, 1998. 17(5): p. 1192-1199.
65. Yang, J., et al., Molecular basis for TPR domain-mediated regulation of protein
phosphatase 5. EMBO J, 2005. 24(1): p. 1-10.
66. Chen, M.X. and P.T.W. Cohen, Activation of protein phosphatase 5 by limited
proteolysis or the binding of polyunsaturated fatty acids to the TPR domain. FEBS Lett,
1997. 400(1): p. 136-140.
67. Sinclair, C., et al., The Tetratricopeptide Repeat Domain and a C-terminal Region
Control the Activity of Ser/Thr Protein Phosphatase 5. J Biol Chem, 1999. 274(33): p.
23666-23672.
68. Kang, H., et al., Identification of Amino Acids in the Tetratricopeptide Repeat and C-
Terminal Domains of Protein Phosphatase 5 Involved in Autoinhibition and Lipid
Activation. Biochemistry, 2001. 40(35): p. 10485-10490.
69. Zeke, T., et al., Human protein phosphatase 5 dissociates from heat-shock proteins and
is proteolytically activated in response to arachidonic acid and the microtubule-
depolymerizing drug nocodazole. Biochem J, 2005. 385(1): p. 45-56.
70. Yong, W., et al., Mice Lacking Protein Phosphatase 5 Are Defective in Ataxia
Telangiectasia Mutated (ATM)-mediated Cell Cycle Arrest. J Biol Chem, 2007. 282(20):
p. 14690-14694.
71. Silverstein, A.M., et al., Protein Phosphatase 5 Is a Major Component of Glucocorticoid
Receptor-hsp90 Complexes with Properties of an FK506-binding Immunophilin. J Biol
Chem, 1997. 272(26): p. 16224-16230.
72. Davies, T.H., Y.-M. Ning, and E.R. Sanchez, Differential Control of Glucocorticoid
Receptor Hormone-Binding Function by Tetratricopeptide Repeat (TPR) Proteins and
the Immunosuppressive Ligand FK506. Biochemistry, 2005. 44(6): p. 2030-2038.
73. Jeong, J.Y., et al., Characterization of Saccharomyces cerevisiae protein Ser/Thr
phosphatase T1 and comparison to its mammalian homolog PP5. BMC Cell Biol, 2003.
4: p. 3.
74. Chen, M.X., et al., A novel human protein serine/threonine phosphatase, which
possesses four tetratricopeptide repeat motifs and localizes to the nucleus. EMBO J,
1994. 13(18): p. 4278-4290.
75. Wandinger, S.K., et al., The phosphatase Ppt1 is a dedicated regulator of the molecular
chaperone Hsp90. EMBO J, 2006. 25(2): p. 367-376.
116
76. Cho, B.-R., P. Lee, and J.-S. Hahn, CK2-dependent inhibitory phosphorylation is
relieved by Ppt1 phosphatase for the ethanol stress-specific activation of Hsf1 in
Saccharomyces cerevisiae. Mol Microbiol, 2014. 93(2): p. 306-316.
77. Noguchi, C., et al., Association of Constitutive Hyperphosphorylation of Hsf1p with a
Defective Ethanol Stress Response in Saccharomyces cerevisiae Sake Yeast Strains.
Appl Environ Microbiol, 2012. 78(2): p. 385-392.
78. Sanchez-Ortiz, E., et al., Protein phosphatase 5 protects neurons against amyloid-beta
toxicity. J Neurochem, 2009. 111(2): p. 391-402.
79. Chen, X., S. Lu, and Y. Zhang, Characterization of protein phosphatase 5 from three
lepidopteran insects: Helicoverpa armigera, Mythimna separata and Plutella xylostella.
PLoS One, 2014. 9(5): p. e97437.
80. Gupta, S., et al., Exploration of new drug-like inhibitors for serine/threonine protein
phosphatase 5 of Plasmodium falciparum: a docking and simulation study. J Biomol
Struct Dyn, 2015. 33(11): p. 2421-41.
81. Ramsey, A.J. and M. Chinkers, Identification of Potential Physiological Activators of
Protein Phosphatase 5. Biochemistry, 2002. 41(17): p. 5625-5632.
82. Cher, C., et al., Identification of Chaulmoogric Acid as a Small Molecule Activator of
Protein Phosphatase 5. Appl Biochem Biotechnol, 2009.
83. Haslbeck, V., et al., Selective activators of protein phosphatase 5 target the auto-
inhibitory mechanism. Biosci Rep, 2015. 35(3).
84. Chatterjee, A., et al., Activated Rac1 GTPase Translocates Protein Phosphatase 5 to the
Cell Membrane and Stimulates Phosphatase Activity in Vitro. J Biol Chem, 2010.
285(6): p. 3872-3882.
85. Taira, J. and Y. Higashimoto, Caveolin-1 interacts with protein phosphatase 5 and
modulates its activity in prostate cancer cells. Biochem Biophys Res Commun, 2013(0).
86. Yamaguchi, F., et al., S100 proteins modulate protein phosphatase 5 function: a link
between CA2+ signal transduction and protein dephosphorylation. J Biol Chem, 2012.
87. Connarn, J.N., et al., The Molecular Chaperone Hsp70 Activates Protein Phosphatase 5
(PP5) By Binding the Tetratricopeptide Repeat (TPR) Domain. J Biol Chem, 2013.
88. Sekine, Y., et al., The Kelch Repeat Protein KLHDC10 Regulates Oxidative Stress-
Induced ASK1 Activation by Suppressing PP5. Mol Cell, 2012(0).
89. Zhao, S. and A. Sancar, Human blue-light photoreceptor hCRY2 specifically interacts
with protein serine/threonine phosphatase 5 and modulates its activity. Photochem
Photobiol, 1997. 66(5): p. 727-31.
90. Hysi, P., et al., NOD1 variation, immunoglobulin E and asthma. Hum Mol Genet, 2005.
14(7): p. 935-41.
117
91. Perez, L.-H., et al., Direct Bacterial Killing In Vitro by Recombinant Nod2 Is
Compromised by Crohn's Disease-Associated Mutations. PLoS One, 2010. 5(6): p.
e10915.
92. Hugot, J.-P., et al., Association of NOD2 leucine-rich repeat variants with susceptibility
to Crohn's disease. Nature, 2001. 411(6837): p. 599-603.
93. Lauro, M.L., J.M. Burch, and C.L. Grimes, The effect of NOD2 on the microbiota in
Crohn's disease. Curr Opin Biotechnol, 2016. 40: p. 97-102.
94. Macaluso, F., et al., Polymorphisms in NACHT-LRR (NLR) genes in atopic dermatitis.
Exp Dermatol, 2007. 16(8): p. 692-698.
95. Weidinger, S., et al., Association of NOD1 polymorphisms with atopic eczema and
related phenotypes. J Allergy Clin Immunol, 2005. 116(1): p. 177-184.
96. Kim, Y.K., J.-S. Shin, and M.H. Nahm, NOD-Like Receptors in Infection, Immunity,
and Diseases. Yonsei Med J, 2016. 57(1): p. 5-14.
97. Caruso, R., et al., NOD1 and NOD2: Signaling, Host Defense, and Inflammatory
Disease. Immunity, 2014. 41(6): p. 898-908.
98. Ng, A. and R.J. Xavier, Leucine-rich repeat (LRR) proteins: Integrators of pattern
recognition and signaling in immunity. Autophagy, 2011. 7(9): p. 0--1.
99. Monie, T.P., NLR activation takes a direct route. Trends Biochem Sci, 2013. 38(3): p.
131-139.
100. Girardin, S.E., et al., Peptidoglycan Molecular Requirements Allowing Detection by
Nod1 and Nod2. J Biol Chem, 2003. 278(43): p. 41702-41708.
101. Askari, N., et al., Expression, purification, and characterization of recombinant NOD1
(NLRC1): A NLR family member. J Biotechnol, 2012. 157(1): p. 75-81.
102. Strober, W., et al., Signalling pathways and molecular interactions of NOD1 and
NOD2. Nat Rev Immunol, 2006. 6(1): p. 9-20.
103. Laroui, H., et al., l-Ala-γ-d-Glu-meso-diaminopimelic Acid (DAP) Interacts Directly
with Leucine-rich Region Domain of Nucleotide-binding Oligomerization Domain 1,
Increasing Phosphorylation Activity of Receptor-interacting Serine/Threonine-protein
Kinase 2 and Its Interaction with Nucleotide-binding Oligomerization Domain 1. J
Biol Chem, 2011. 286(35): p. 31003-31013.
104. Grimes, C.L., et al., The Innate Immune Protein Nod2 Binds Directly to MDP, a
Bacterial Cell Wall Fragment. J Am Chem Soc, 2012. 134(33): p. 13535-13537.
105. Coussens, N.P., et al., Crystal structure of the Nod1 caspase activation and recruitment
domain. Biochem Biophys Res Commun, 2007. 353(1): p. 1-5.
106. Ver Heul, A.M., et al., Crystal Structure of a Complex of NOD1 CARD and Ubiquitin.
PLoS One, 2014. 9(8): p. e104017.
118
107. Reubold, T.F., et al., Crystal structure of the leucine-rich repeat domain of the NOD-
like receptor NLRP1: Implications for binding of muramyl dipeptide. FEBS Lett, 2014.
588(18): p. 3327-3332.
108. Kim, H.M., et al., Crystal Structure of the TLR4-MD-2 Complex with Bound
Endotoxin Antagonist Eritoran. Cell, 2007. 130(5): p. 906-917.
109. Stumpp, M.T., et al., Designing repeat proteins: modular leucine-rich repeat protein
libraries based on the mammalian ribonuclease inhibitor family. J Mol Biol, 2003.
332(2): p. 471-87.
110. Jin, M.S. and J.O. Lee, Application of hybrid LRR technique to protein crystallization.
BMB Rep, 2008. 41(5): p. 353-7.
111. Magnuson, G., et al., High Throughput Screening Assays for NOD1 Inhibitors. Probe
Reports from the NIH Molecular Libraries Program, 2010.
112. Novus biologicals, http://www.novusbio.com/NOD1-Recombinant-Protein_H0001039
2-P01.html.
113. Francis, D.M. and R. Page, Strategies to Optimize Protein Expression in E. coli. Curr
Protoc Protein Sci, 2001.
114. Bansal, P.K., R. Abdulle, and K. Kitagawa, Sgt1 associates with Hsp90: an initial step
of assembly of the core kinetochore complex. Mol Cell Biol, 2004. 24(18): p. 8069-79.
115. Rodrigo-Brenni, M.C., et al., Sgt1p and Skp1p Modulate the Assembly and Turnover
of CBF3 Complexes Required for Proper Kinetochore Function. Mol Biol Cell, 2004.
15(7): p. 3366-3378.
116. Steensgaard, P., et al., Sgt1 is required for human kinetochore assembly. EMBO Rep,
2004. 5(6): p. 626-631.
117. Davies, A.E. and K.B. Kaplan, Hsp90-Sgt1 and Skp1 target human Mis12 complexes
to ensure efficient formation of kinetochore-microtubule binding sites. J Cell Biol,
2010. 189(2): p. 261-74.
118. Kitagawa, K., et al., SGT1 encodes an essential component of the yeast kinetochore
assembly pathway and a novel subunit of the SCF ubiquitin ligase complex. Mol Cell,
1999. 4(1): p. 21-33.
119. Catlett, M.G. and K.B. Kaplan, Sgt1p Is a Unique Co-chaperone That Acts as a Client
Adaptor to Link Hsp90 to Skp1p. J Biol Chem, 2006. 281(44): p. 33739-33748.
120. Takahashi, A., et al., HSP90 interacts with RAR1 and SGT1 and is essential for RPS2-
mediated disease resistance in Arabidopsis. Proc Natl Acad Sci USA, 2003. 100(20): p.
11777-82.
121. Lee, Y.T., et al., Human Sgt1 binds HSP90 through the CHORD-Sgt1 domain and not
the tetratricopeptide repeat domain. J Biol Chem, 2004. 279(16): p. 16511-7.
119
122. Lingelbach, L.B. and K.B. Kaplan, The interaction between Sgt1p and Skp1p is
regulated by HSP90 chaperones and is required for proper CBF3 assembly. Mol Cell
Biol, 2004. 24(20): p. 8938-50.
123. Stuttmann, J., J.E. Parker, and L.D. Noel, Staying in the fold: The SGT1/chaperone
machinery in maintenance and evolution of leucine-rich repeat proteins. Plant Signal
Behav, 2008. 3(5): p. 283-5.
124. Taube, M., et al., Low-Resolution Structure of the Full-Length Barley (Hordeum
vulgare) SGT1 Protein in Solution, Obtained Using Small-Angle X-Ray Scattering.
PLoS One, 2014. 9(4): p. e93313.
125. Bansal, P.K., et al., Sgt1 Dimerization Is Required for Yeast Kinetochore Assembly. J
Biol Chem, 2009. 284(6): p. 3586-3592.
126. Bansal, P.K., et al., Sgt1 Dimerization Is Negatively Regulated by Protein Kinase
CK2-mediated Phosphorylation at Ser361. J Biol Chem, 2009. 284(28): p. 18692-
18698.
127. Nyarko, A., et al., TPR-Mediated Self-Association of Plant SGT1†. Biochemistry,
2007. 46(40): p. 11331-11341.
128. Mayor, A., et al., A crucial function of SGT1 and HSP90 in inflammasome activity
links mammalian and plant innate immune responses. Nat Immunol, 2007. 8(5): p.
497-503.
129. Morris, G.M., et al., Automated docking using a Lamarckian genetic algorithm and an
empirical binding free energy function. J Comput Chem, 1998. 19(14): p. 1639-1662.
130. Park, H., J. Lee, and S. Lee, Critical assessment of the automated AutoDock as a new
docking tool for virtual screening. Proteins, 2006. 65(3): p. 549-54.
131. Barril, X., et al., Structure-based discovery of a new class of Hsp90 inhibitors. Bioorg
Med Chem Lett, 2005. 15(23): p. 5187-5191.
132. Jeffrey, G.A., An Introduction to Hydrogen bonding. 1997, Oxford: Oxford University
Press.
133. Interbioscreen Database, Moscow, http://www.ibscreen.com.
134. Lipinski, C.A., et al., Experimental and computational approaches to estimate
solubility and permeability in drug discovery and development settings. Adv Drug
Deliv Rev, 1997. 23(1-3): p. 3-25.
135. Gasteiger, J. and M. Marsilli, Iterative partial equalization of orbital
electronegativity—a rapid access to atomic charges. Tetrahedron, 1980. 36(22): p.
3219-3228.
120
136. Park, H. and Y.H. Jeon, Cubic equation governing the outer-region dielectric constant
of globular proteins. Phys Rev E Stat Nonlin Soft Matter Phys, 2007. 75(2 Pt 1): p.
021916.
137. Stouten, P.F.W., et al., An Effective Solvation Term Based on Atomic Occupancies for
Use in Protein Simulations. Mol. Simul, 1993. 10: p. 97-120.
138. Kang, H., H. Choi, and H. Park, Prediction of molecular solvation free energy based
on the optimization of atomic solvation parameters with genetic algorithm. J Chem Inf
Model, 2007. 47(2): p. 509-514.
139. Shoichet, B.K., A.R. Leach, and I.D. Kuntz, Ligand solvation in molecular docking.
Proteins, 1999. 34(1): p. 4-16.
140. Case, D.A.C., T. E.III; Darden, T.; Gohlke, H.; Luo, R.; Merz, K. M., Jr.; Onufriev, A.;
Simmerling, C.; Wang, B.; Woods, R. J., The Amber biomolecular simulation
programs. J Comput Chem, 2005. 26(16): p. 1668-1688.
141. Park, H. and S. Lee, Role of Solvent Dynamics in Stabilizing the Transition State of
RNA Hydrolysis by Hairpin Ribozyme. J Chem Theory Comput, 2006. 2(3): p. 858-
862.
142. Ryckaert, J.-P., G. Ciccotti, and H.J.C. Berendsen, Numerical integration of the
cartesian equations of motion of a system with constraints: molecular dynamics of n-
alkanes. J Comput Phys, 1977. 23(3): p. 327-341.
143. Rowlands, M.G., et al., High-throughput screening assay for inhibitors of heat-shock
protein 90 ATPase activity. Anal Biochem, 2004. 327(2): p. 176-183.
144.Skehan, P., et al., New Colorimetric Cytotoxicity Assay for Anticancer-Drug Screening.
J Natl Cancer Inst, 1990. 82(13): p. 1107-1112.
145. Swingle, M.R. and R.E. Honkanen, Development and Validation of a Robust and
Sensitive Assay for the Discovery of Selective Inhibitors for Serine/Threonine Protein
Phosphatases PP1α (PPP1C) and PP5 (PPP5C). Assay Drug Dev Technol, 2014. 12(8):
p. 481-496.
146. Burger, A.M., Highlights in experimental therapeutics. Cancer Lett, 2007. 245(1-2): p.
11-21.
147. Yoshimasa, U., Natural Product Origins of Hsp90 Inhibitors. Curr Cancer Drug
Targets, 2003. 3(5): p. 325-330.
148. Jilani, K., S.M. Qadri, and F. Lang, Geldanamycin-Induced Phosphatidylserine
Translocation in the Erythrocyte Membrane. Cell Physiol Biochem, 2013. 32(6): p.
1600-1609.
149. Day, J.E., et al., Targeting the Hsp90 Chaperone: Synthesis of Novel Resorcylic Acid
Macrolactone Inhibitors of Hsp90. Chemistry, 2010. 16(9): p. 2758-63.
121
150. Banerji, U., Heat Shock Protein 90 as a Drug Target: Some Like It Hot. Clin Cancer
Res, 2009. 15(1): p. 9-14.
151. Heath, E.I., et al., A Phase II Trial of 17-Allylamino-17-Demethoxygeldanamycin in
Patients with Hormone-Refractory Metastatic Prostate Cancer. Clin Cancer Res, 2008.
14(23): p. 7940-7946.
152. Taldone, T., et al., Targeting Hsp90: small-molecule inhibitors and their clinical
development. Curr Opin Pharmacol, 2008. 8(4): p. 370-374.
153. Solit, D.B. and G. Chiosis, Development and application of Hsp90 inhibitors. Drug
Discov Today, 2008. 13(1-2): p. 38-43.
154. Workman, P., et al., Drugging the Cancer Chaperone HSP90. Ann N Y Acad Sci, 2007.
1113(1): p. 202-216.
155.Shiro, S., et al., Development of Radicicol Analogues. Curr Cancer Drug Targets, 2003.
3(5): p. 359-369.
156. Yamamoto, K., et al., Total Synthesis as a Resource in the Discovery of Potentially
Valuable Antitumor Agents: Cycloproparadicicol. Angew Chem Int Ed, 2003. 42(11):
p. 1280-1284.
157. Yang, Z.-Q., et al., New Efficient Synthesis of Resorcinylic Macrolides via Ynolides:
Establishment of Cycloproparadicicol as Synthetically Feasible Preclinical Anticancer
Agent Based on Hsp90 as the Target. J Am Chem Soc, 2004. 126(25): p. 7881-7889.
158. Messaoudi, S., et al., Heat-shock protein 90 inhibitors as antitumor agents: a survey of
the literature from 2005 to 2010. Expert Opin Ther Pat, 2011. 21(10): p. 1501-1542.
159. Chiosis, G., et al., A small molecule designed to bind to the adenine nucleotide pocket
of Hsp90 causes Her2 degradation and the growth arrest and differentiation of breast
cancer cells. Chem Biol, 2001. 8(3): p. 289-299.
160. Saif, M.W., et al., A Phase 1, Dose-Escalation, Pharmacokinetic and
Pharmacodynamic Study of BIIB021 Administered Orally in Patients with Advanced
Solid Tumors. Clin Cancer Res, 2014. 20(2): p. 445-455.
161.Hong, D., et al., Phase I Study of BIIB028, a Selective Heat Shock Protein 90 Inhibitor,
in Patients with Refractory Metastatic or Locally Advanced Solid Tumors. Clin Cancer
Res, 2013. 19(17): p. 4824-4831.
162. Trendowski, M., PU-H71: An improvement on nature's solutions to oncogenic Hsp90
addiction. Pharmacol Res, 2015. 99: p. 202-216.
163. Nakashima, T., et al., New Molecular and Biological Mechanism of Antitumor
Activities of KW-2478, a Novel Nonansamycin Heat Shock Protein 90 Inhibitor, in
Multiple Myeloma Cells. Clin Cancer Res, 2010. 16(10): p. 2792-2802.
122
164. Woodhead, A.J., et al., Discovery of (2,4-Dihydroxy-5-isopropylphenyl)-[5-(4-
methylpiperazin-1-ylmethyl)-1,3-dihydroisoindol-2-yl]methanone (AT13387), a Novel
Inhibitor of the Molecular Chaperone Hsp90 by Fragment Based Drug Design. J Med
Chem, 2010. 53(16): p. 5956-5969.
165. Goyal, L., et al., A phase I and pharmacokinetic study of ganetespib (STA-9090) in
advanced hepatocellular carcinoma. Invest New Drugs, 2015. 33(1): p. 128-137.
166. Neckers, L. and J.B. Trepel, Stressing the Development of Small Molecules Targeting
HSP90. Clin Cancer Res, 2014. 20(2): p. 275-277.
167. Meli, M., et al., Small-molecule targeting of heat shock protein 90 chaperone function:
rational identification of a new anticancer lead. J Med Chem, 2006. 49(26): p. 7721-30.
168. Park, H., Y.J. Kim, and J.S. Hahn, A novel class of Hsp90 inhibitors isolated by
structure-based virtual screening. Bioorg Med Chem Lett, 2007. 17(22): p. 6345-9.
169. Verdonk, M.L., et al., Modeling water molecules in protein-ligand docking using
GOLD. J Med Chem, 2005. 48(20): p. 6504-15.
170. Zou, J., et al., Repression of Heat Shock Transcription Factor HSF1 Activation by
HSP90 (HSP90 Complex) that Forms a Stress-Sensitive Complex with HSF1. Cell,
1998. 94(4): p. 471-480.
171. Fox, J.D., R.B. Kapust, and D.S. Waugh, Single amino acid substitutions on the
surface of Escherichia coli maltose-binding protein can have a profound impact on the
solubility of fusion proteins. Protein Sci, 2001. 10(3): p. 622-630.
172. Tam, M.F., et al., Sickle Cell Hemoglobin with Mutation at alphaHis-50 Has Improved
Solubility. J Biol Chem, 2015. 290(35): p. 21762-72.
173. Esposito, D. and D.K. Chatterjee, Enhancement of soluble protein expression through
the use of fusion tags. Curr Opin Biotechnol, 2006. 17(4): p. 353-358.
174. Shirasu, K., The HSP90-SGT1 chaperone complex for NLR immune sensors. Annu
Rev Plant Biol, 2009. 60: p. 139-64.
175. Tolia, N.H. and L. Joshua-Tor, Strategies for protein coexpression in Escherichia coli.
Nat Meth, 2006. 3(1): p. 55-64.
176. Wang, H. and S. Chong, Visualization of coupled protein folding and binding in
bacteria and purification of the heterodimeric complex. Proc Natl Acad Sci USA, 2003.
100(2): p. 478-483.
177. Kim, C.W., et al., N-terminal domains of native multidomain proteins have the
potential to assist de novo folding of their downstream domains in vivo by acting as
solubility enhancers. Protein Sci, 2007. 16(4): p. 635-643.
178. Costa, S., et al., Fusion tags for protein solubility, purification and immunogenicity in
Escherichia coli: the novel Fh8 system. Front Microbiol, 2014. 5: p. 63.
123
179. Panaretou, B., et al., Activation of the ATPase activity of hsp90 by the stress-regulated
cochaperone aha1. Mol Cell, 2002. 10(6): p. 1307-18.
180. Millson, S.H., et al., Expressed as the sole Hsp90 of yeast, the α and β isoforms of
human Hsp90 differ with regard to their capacities for activation of certain client
proteins, whereas only Hsp90β generates sensitivity to the Hsp90 inhibitor radicicol.
FEBS J, 2007. 274(17): p. 4453-4463.
181. Swingle, M.R., R.E. Honkanen, and E.M. Ciszak, Structural Basis for the Catalytic
Activity of Human Serine/Threonine Protein Phosphatase-5. J Biol Chem, 2004.
279(32): p. 33992-33999.
182. Shi, Y., Serine/Threonine Phosphatases: Mechanism through Structure. Cell, 2009.
139(3): p. 468-484.
183. Honkanen, R.E. and T. Golden, Regulators of Serine / Threonine Protein Phosphatases
at the Dawn of a Clinical Era? Curr Med Chem, 2002. 9(22): p. 2055-2075.
184. Chattopadhyay, D., et al., Crystal structures and mutagenesis of PPP-family ser/thr
protein phosphatases elucidate the selectivity of cantharidin and novel norcantharidin-
based inhibitors of PP5C. Biochem Pharmacol, 2016. 109: p. 14-26.
185. Bohrmann, B., et al., Self-Assembly of β-Amyloid 42 Is Retarded by Small Molecular
Ligands at the Stage of Structural Intermediates. J Struct Biol, 2000. 130(2–3): p. 232-
246.
186. Guo, F., et al., RO 90-7501 enhances TLR3 and RLR agonist induced antiviral
response. PLoS One, 2012. 7(10): p. e42583.
187. Morita, K., et al., Negative feedback regulation of ASK1 by protein phosphatase 5
(PP5) in response to oxidative stress. EMBO J, 2001. 20(21): p. 6028-36.
188. Skinner, J., et al., Purification of a Fatty Acid-stimulated Protein-serine/threonine
Phosphatase from Bovine Brain and Its Identification as a Homolog of Protein
Phosphatase 5. J Biol Chem, 1997. 272(36): p. 22464-22471.
124
국
Hsp90 샤페 시스 Hsp90 클라이언트 (client) 단 질
분 는 다양한 단 질들 힘과 , 안 에 있어 요한
역할 행한다. Hsp90 클라이언트 단 질들 에는 여러 암
단 질들 (oncoprotein)이 포함 어 있 에, Hsp90 해
작용할 있는 다양한 골격 (scaffold) 합 들이 항암
심도있게 연구, 개 어 다. 본 연구에 는 가상 스크리닝과 실험
스크리닝 용하여 굴 2 종 새 운 골격 갖는 Hsp90
해 들 특 규명하 다. 가상 스크리닝 AutoDock
프 그램 행 었 며, 각 합 결합 식 분자동역학
시뮬 이 통해 가시 었다. 이 2 종 합 들 in vitro 상에
효모 Hsp90 ATP 가 분해 해하 며, MCF-7 암
포주에 해 생장억 효과를 나타내었다. 이들 합 들 구 하는
골격들 각각 pyrimidine-2,4,6-trione 과 4H-1,2,4-triazole-3-thiol 이었다.
라이 러리 상에 이 골격들 공 하는 합 들 추가
하여, 이 합 들이 Hsp90 클라이언트 단 질이며 암
단 질인 Her2 안 에 미 는 향 살펴보았다. 결과 ,
6 종 에 5 종 합 들이 Her2 단 질 감소를
하 며, 이 결과를 통해 새롭게 굴 2 종 골격이 Hsp90
해 특 나타냄 인하 다. 또한 이 같 Hsp90
해 들 공 인 굴 통해, 가상 스크리닝이 신약개 에
용하게 용 있 시하 다.
Hsp90 샤페 시스 Hsp90 이합체 Hsp90 능 조 하고
보조하는 Hsp90 코샤페 (cochaperone)들 구 다. Hsp90 뿐아니라
p23, Cdc37, SGT1, FKBP51, PP5 등과 같 몇몇 Hsp90 코샤페 들 또한
125
암과 연 이 었다. 이 PP5 는 PPP family 에 속한
탈인산 효소이며, 이 PPP family 에 속한 탈인산 효소들 구조
매우 사한 매 도 인 (catalytic domain) 지닌다. 재 지 알 진
천연 PP5 해 들 이 매 도 인에 결합하므 , PPP family 에
속한 여러 탈인산 효소들에 해 범 한 택 나타낸다. 본
연구에 는 이러한 존 특이 인 PP5 해 들과는 다른,
새 운 작 작용하는 PP5 해 를 탐색하 해 LOPAC
(Library Of Pharmacologically Active Compounds) 1280 라이 러리에 포함
합 들 스크리닝 행하 다. 그 결과, 3 종 PP5 해 를
굴할 있었다. 이 해 들 작용 작 규명하 해, PP5
체 이 (full-length) 단 질, 그리고 탈인산 자가억 하는
PP5 TPR 도 인과 C 말단 13 개 아미노산이 각각, 또는 모 결손
단 질들 에 이 합 들이 미 는 향 어 이 인하 다.
이를 통해 존 PP5 해 들과는 다른, PP5 에 해 TPR- 존
해 작 나타내는 1 종 해 Ro 90-7501 인할
있었다. TPR 도 인 PPP family 에 속한 탈인산 효소들 에
PP5 만이 지니는 고 한 도 인이 에 TPR- 존 인 PP5
해 는 PP5-특이 인 해 작용할 가능 이 있 며,
신약개 이나 PP5 연구에 용하게 용 있 것
사료 다.
Hsp90 샤페 시스 클라이언트 단 질들 힘, , 그리고
안 능 행한다. NOD1 Hsp90 클라이언트 단 질
하나이며, 염증 질병들과 연 포 내 천 면역 이다.
NOD1 단 질 분리 가 매우 어 운 것 알 있 며,
이는 NOD1 에 한 분자 연구에 걸림돌이 어 다. 본
연구에 는 Hsp90 샤페 시스 용하여 NOD1 단 질 LRR
126
도 인 장균 내에 가용 상태 하고자 하 다. 인간
암 포주 행 존 연구보고에 르면 Hsp90 코샤페 SGT1
NOD1 에 필요하며, Hsp90 는 NOD1 안 에 여한다. 이를
SGT1 NOD1 단 질 LRR 도 인과 합할 트 단 질
하 다. 인간 SGT1 NOD1-LRR N 말단에 합한 합단 질
HsSGT1NOD1LRR 향상 가용 나타내었 나, 극심한 분해
낮 량 나타내었다. 이를 극복하 해 효모, 장균, 마우스
래 Hsp90 들 공 단 질 하여 그 효과를 인하 다.
존 보고 일 결과 , Hsp90 공 합단 질
안 향상시켰다. 장균 HtpG, 효모 Hsp82 Hsc82, 마우스
Hsp90α 등 다양한 래 Hsp90 들 공 효과를 하 ,
마우스 Hsp90α 가 공 시 가장 효과 SGT1NOD1LRR
합단 질 안 시킴 찰하 다.
주요어 : 약 스크리닝, 해 , 합단 질, 공 , Hsp90, PP5, SGT1, NOD1
학 번 : 2007-30853
