Reinventing Hsp90 Inhibitors: Blocking C-Terminal Binding Events to Hsp90 by Using Dimerized Inhibitors

Article abstract of DOI:10.1002/chem.201603464

Heat shock protein 90 (Hsp90) is a molecular chaperone (90 kDa) that functions as a dimer. This protein facilitates the folding, assembly, and stabilization of more than 400 proteins that are responsible for cancer development and progression. Inhibiting Hsp90’s function will shut down multiple cancer-driven pathways simultaneously because oncogenic clients rely heavily on Hsp90, which makes this chaperone a promising anticancer target. Classical inhibitors that block the binding of adenine triphosphate (ATP) to the N-terminus of Hsp90 are highly toxic to cells and trigger a resistance mechanism within cells. This resistance mechanism comprises a large increase in prosurvival proteins, namely, heat shock protein 70 (Hsp70), heat shock protein 27 (Hsp27), and heat shock factor 1 (HSF-1). Molecules that modulate the C-terminus of Hsp90 are effective at inducing cancer-cell death without activating the resistance mechanism. Herein, we describe the design, synthesis, and biological binding affinity for a series of dimerized C-terminal Hsp90 modulators. We show that dimers of these C-terminal modulators synergistically inhibit Hsp90 relative to monomers.

Full text of DOI:10.1002/chem.201603464

DOI: 10.1002/chem.201603464
Full Paper
&
Antitumor Agents
Reinventing Hsp90 Inhibitors: Blocking C-Terminal Binding Events
to Hsp90 by Using Dimerized Inhibitors**
Yen Chin Koay,[a] Hendra Wahyudi,[b] and Shelli R. McAlpine*[a]
Abstract: Heat shock protein 90 (Hsp90) is a molecular chap-
erone (90 kDa) that functions as a dimer. This protein facili-
tates the folding, assembly, and stabilization of more than
400 proteins that are responsible for cancer development
and progression. Inhibiting Hsp90’s function will shut down
multiple cancer-driven pathways simultaneously because on-
cogenic clients rely heavily on Hsp90, which makes this
chaperone a promising anticancer target. Classical inhibitors
that block the binding of adenine triphosphate (ATP) to the
N-terminus of Hsp90 are highly toxic to cells and trigger a re-
sistance mechanism within cells. This resistance mechanism
comprises a large increase in prosurvival proteins, namely,
heat shock protein 70 (Hsp70), heat shock protein 27
(Hsp27), and heat shock factor 1 (HSF-1). Molecules that
modulate the C-terminus of Hsp90 are effective at inducing
cancer-cell death without activating the resistance mecha-
nism. Herein, we describe the design, synthesis, and biologi-
cal binding affinity for a series of dimerized C-terminal
Hsp90 modulators. We show that dimers of these C-terminal
modulators synergistically inhibit Hsp90 relative to mono-
mers.
Introduction
lar-stress responses from promoting drug resistance and evad-
ing apoptosis.
Molecular-targeted chemotherapies, which inhibit a single on-
coprotein, are critical in the fight against cancer. Between 2005
and 2016, the Food and Drug Administration (FDA) approved
more than 50 different molecular-targeted anticancer agents
for use against more than 20 different types of cancer (http://
www.centerwatch.com). Heat shock protein 90 (Hsp90) is
a highly conserved molecular chaperone that participates in
the folding, stabilization, and activation of more than 400 pro-
teins involved in the 10 hallmarks of cancer.[1] Hsp90 makes up
1–2% of all cellular proteins in normal cells, whereas Hsp90
makes up 3–6% of all proteins in cancerous cells. The increase
in Hsp90 protects mutated and overexpressed oncoproteins
from degradation, thereby facilitating cancer-cell survival.[2]
Hsp90 also plays a regulatory role in cellular-stress response
pathways, thus increasing the prosurvival proteins heat shock
protein 70 (Hsp70), heat shock protein 27 (Hsp27), and heat
shock factor 1 (HSF-1), which facilitate drug resistance.[3] Thus,
inhibiting the cellular function of Hsp90 blocks multiple roles
of oncoproteins role in cancer-cell growth and prevents cellu-
Hsp90 consists of three domains: an amino (N) domain
(25 kDa), a middle (M) domain (35 kDa), and a carboxy (C)
domain (10 kDa).[4] Classical Hsp90 inhibitors bind to the N-ter-
minal domain inside the ATP binding pocket. Although there
have been significant efforts to produce molecules that target
this chaperone (www.clinicaltrials.gov), the use of these mole-
cules in clinical applications has been hampered by the induc-
tion of the side effects of the prosurvival proteins. Side effects
of the drugs include a dramatic induction of HSF-1, Hsp70, and
Hsp27. These proteins protect the cell from apoptosis and pro-
mote antiapoptotic pathways, thereby facilitating cancer
growth.[5] In addition to these classical Hsp90 inhibitors, two
other categories of inhibitor have been reported: 1) direct C-
terminal inhibitors and 2) C-terminal modulators, which allos-
terically inhibit Hsp90 function by disrupting the interactions
between Hsp90 and cochaperones that bind to the C-termi-
nus.
C-terminal inhibitors of Hsp90 based on novobiocin have
been developed.[6] Our group also recently developed a C-ter-
minal inhibitor that targets the last five residues (MEEVD) at
the C-terminal end of Hsp90.[7] C-terminal modulators (or allo-
steric C-terminal inhibitors) have been extensively studied.[8]
Unlike classical inhibitors, both direct C-terminal inhibitors and
allosteric inhibitors do not trigger a cell protection mecha-
nism.[6,8] Instead, the C-terminal inhibitors and modulators act
through a mechanism that appears to be more effective than
that of the classical inhibitors. C-terminal modulators devel-
oped by our group bind between the amino and middle do-
mains of Hsp90 and allosterically modulate binding between
Hsp90 and the C-terminal clients or cochaperones.[8b–f,9] Four
[a] Y. C. Koay, Prof. S. R. McAlpine
School of Chemistry, University of New South Wales
Sydney, NSW 2052 (Australia)
E-mail: s.mcalpine@unsw.edu.au
[b] Dr. H. Wahyudi
Department of Bioengineering, University of Utah
Salt Lake City, UT 84112 (USA)
[**] Hsp90=heat-shock protein 90.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201603464.
Chem. Eur. J. 2016, 22, 1 – 12
1
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Figure 1. Structures of SM122, SM145, GMD-4c (1), EC5 (2), coumermycin A1 (3), and the SM122-PEG dimer (4).
Figure 2. a) Structures of the SM253 molecules and their PEG-biotinylated analogues (SM253 tagged at positions II, III, and IV). b) Structures of SM258 and its
PEG-biotinylated analogues (SM258 tagged at positions II and III).
analogues from this allosteric series are highly effective:
SM122, SM145, SM253, and SM258 (Figures 1 and 2).[8a–d,9d,e] All
four molecules bind to Hsp90 and block the binding of tetratri-
copeptide repeat (TPR) cochaperones, which interact at the
Hsp90 C-terminus. These TPR proteins regulate pathways in-
volved in cancer-cell growth, including hormone-receptor pro-
duction and antiapoptosis. These C-terminal Hsp90 modula-
tors, termed SM molecules, also block the cellular-stress re-
sponse pathways of by inhibiting TPR protein access to Hsp90.
Hsp90 functions as a dimer, therefore dimerization of an
Hsp90 inhibitor is one useful method in the generation of syn-
ergistic inhibitors. Zheng et al. employed a dimerization strat-
egy in which the classical inhibitor geldanamycin (GM) was di-
merized through several alkylamino chain lengths; in particular,
the GMD-4c dimer (1; Figure 1) exhibited significant antiproli-
ferative activity against tumor cell lines (GI50 ꢀ20–100 nm).[10]
A
subsequent study by Yin et al. showed that dimerization of
a classical ansamycin-based compound to produce EC5 (2;
Figure 1) was effective at inhibiting the tumor growth of head
and neck squamous cell carcinoma (HNSCC) cell lines (GI50 <
200 nm) and was twofold more potent than the well-known
classical inhibitor 17-AAG (GI50 ꢀ500 nm).[11]
The C-terminal inhibitor coumermycin A1 was dimerized by
using several linker lengths. Courmermycin A1 (3) exhibited
&
&
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
a GI50 value of 200 nm against SKBr3 cell lines (Figure 1.[12]
A
to inhibit Hsp90 activity.[9e] Our previously reported dimer 4
placed the PEG linker at position IV on SM122 because the
pulldown assays that used Hsp90 and a biotinylated-tagged
variant of SM122 had shown that this molecule was the most
effective at pulling down Hsp90. However, our pulldown
assays had also shown that the placement of a biotin tag at
position III on SM122 was also very effective at binding Hsp90.
Thus, the dimer can be generated by placing a linker at posi-
tion III. Previous pulldown assays with tagged variants of
SM145 had shown that SM145-Tag-II was highly effective at
pulling down Hsp90.[8e] Thus, SM145 the linker would be
placed at position II.
report on the crystal structure of the Hsp90–Hsp organizing
protein (HOP) complex revealed that the distance between
two N-domains of Hsp90 in its open conformation is approxi-
mately 80 ꢁ.[13] Thus, the appropriate linker length for C-termi-
nal modulators that bind between the N-middle domain was
estimated to be approximately 40–50 ꢁ. Dimerized variants of
SM122, a C-terminal modulator, with a poly(ethylene glycol)
(PEG) chain (length of (PEG)9: ꢀ46 ꢁ) successfully blocked the
protein-folding function of Hsp90 and disrupted the interac-
tion between Hsp90 and TPR-containing proteins.[14] Dimerized
SM122 (Figure 1) synergistically inhibited the binding interac-
tion between Hsp90 and the TPR-containing proteins FKBP52
and HOP, in which the dimer was more than twice as effective
than a single SM122 molecule.[14]
It was not clear which position of SM253 or SM258 would
be the most effective for pulling down Hsp90. Thus, we syn-
thesized all three tagged variants of SM253 (Figure 2), thus
leaving the biphenyl residue (position I) and the thiazolyl
moiety (position V) untouched. Both the biphenyl and thiazole
residues have been established as critical for biological activity
in structure–activity relationship (SAR) studies.[8b,15] We also
synthesized two tagged variants of SM258, in which we main-
tained the thiazole (IV), phenyl (V), and biphenyl (I) residues by
placing the pulldown tag at positions II and III (Figure 2).
The generation of the tagged molecules was accomplished
by using solid-phase synthesis in which a lysine residue was
substituted for the amino acid at the relevant position. The
synthesis of SM253-Tag-IV is a representative example of how
the molecules were generated (Scheme 1). By using the pre-
loaded 2-chlorotrityl-leucine (Leu) resin, Fmoc-protected amino
acids were sequentially coupled to the free amine group of
the leucine residue. The amine group on each amino acid was
deprotected after each coupling reaction. After the fourth
amino acid residue was coupled, the pentapeptide unit was
cleaved from the resin to produce the linear peptide 5
Herein, we describe the design and synthesis of four dimer
molecules based on the C-terminal modulators SM122, SM145,
SM253, and SM258. The linker placement was chosen by using
pulldown assays to identify the optimal position for the linker
of each molecule. The dimers were assessed by using Hsp90
binding assays for their ability to inhibit TPR proteins from
binding to the Hsp90 C-terminus.
Results and Discussion
Determining the optimal linker placement on the mono-
mers
A critical factor in the design of the PEG-linked SM dimer is to
first identify the optimal site for the linker attachment that
connects the two SM monomers together. Previous work on
SM122-based dimer molecules had shown that the position for
incorporating linkers to form dimers could impact their ability
Scheme 1. Solid-phase peptide synthesis followed by macrocyclization to afford the biotinylated macrocycle.
Chem. Eur. J. 2016, 22, 1 – 12 www.chemeurj.org ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3
&
&
These are not the final page numbers! ÞÞ

Full Paper
(Scheme 1). Closing the macrocycle under our reported cycliza-
tion conditions[8b,d,9c,d] produced cyclic peptide 6. Removal of
the amino-protecting group tert-butyloxycarbonyl (Boc) pro-
duced 7 and coupling (PEG)4–biotin in the presence of N,N-dii-
sopropylethylamine (DIPEA) produced the tagged molecule
(i.e., SM253–Tag-IV in this case; Scheme 1).
dimer linker. Evaluation of SM253-Tag-II, -III, and -IV in the
same gel (Figure 3a) demonstrated that SM253-Tag-IV was the
most effective molecule for binding to Hsp90. Thus, position IV
on SM253 was used as the linker site. The comparison of
SM258-Tag-II and -III (Figure 3b) in pulldown assays revealed
that SM258-Tag-III was the most effective at binding to Hsp90
in cell lysate. Thus, position III was chosen to link two SM258
molecules.
By using cell lysate and our reported pulldown assay condi-
tions,[8d,e,9e] we evaluated how effectively the five tagged mole-
cules (SM253-Tag-II, -III, and -IV and SM258-Tag-II and -III)
pulled down Hsp90 (Figure 3). The most effective of each
series would then be used to decide the placement of the
Synthesis of dimerized C-terminal modulators
With the optimal linker position for each monomer identified,
we synthesized four dimers: SM122-Tag-III, SM145-Tag-II,
SM253-Tag-IV, and SM258-Tag-III (Figure 4). Each dimer was
constructed from the corresponding monomers. By starting
from free amine 7 (Figure 4b), which was generated by using
solid-phase synthesis (Scheme 1), the dimer was produced
using a 2:1 ratio of amine 7 to bis-N-succinimidyl-(PEG)9
(BS(PEG)9) in the presence of DIPEA. Upon purification by
means of reversed-phase HPLC, the dimer structures were con-
firmed by using NMR spectroscopy and LC-MS and high-resolu-
tion (HR) MS. Ratios of the PEG peaks to individual protons in
the monomer provided evidence that the compounds had di-
merized.
Ability to block binding between Hsp90 and TPR-containing
proteins HOP and FKBP52
Figure 3. Hsp90 protein pulldowns in HCT116 cell lysates. Pulldown data for
SM253 and the PEG-biotinylated analogues (left). Pulldown data for the
SM258 molecules and their PEG-biotinylated analogues (right). Flow through
the lanes are shown in the Supporting Information.
Hsp90 interacts with 19 different C-terminal cochaperones and
14 of these molecules contain TPR domains.[16] These TPR-con-
Figure 4. a) Structures of SM122, SM145, SM253, and SM258 and their dimerized variants. b) The synthesis of the SM253-Tag-IV dimer, which exemplifies the
synthesis of the four dimers.
&
&
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
4
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
taining proteins include HOP and the peptidyl–prolyl isomer-
ase (PPIase) family members FKBP52, FKBP51, and Cyp40.[17]
The monomers SM122, SM145, SM253, and SM258 block the
interactions between Hsp90 and HOP and between Hsp90 and
FKBP52.[8b–d,14] A structurally similar monomer to SM253 and
SM258 was used as the negative control (SM271; see the Sup-
porting Information) and improved the affinity between Hsp90
and HOP.[8b] We have also demonstrated that the linker (PEG)8
had no impact on the binding affinity of the monomer.[14]
Thus, the improved affinity of the dimer over the monomer
would be purely due to the synergistic effect of the two mono-
mers linked together.
Binding assays between FKBP52 and Hsp90 also showed
that the dimeric compounds were all synergistically able to dis-
rupt the binding event. The most effective dimers were the
SM122-Tag-III and SM253-Tag-IV dimers (Figure 5b). Because
we established that the monomers bind to Hsp90 (Figure 3), it
is most likely that these dimer molecules act by binding to the
Hsp90 N-middle domain and synergistically modifying the con-
formation of Hsp90 to allow the binding of the second SM
molecule on the dimers.
Despite the fact that the dimeric compounds provided
a proof-of-concept demonstration for Hsp90 inhibition, these
dimerized variants failed to show growth inhibition in HCT116
cancer cells (see the Supporting Information). These molecules
were highly soluble in water, which was attributed to the PEG
linker. Thus, the failure was most likely due to the large molec-
ular size of the dimers (MWꢀ2000 gmolÀ1), which were ap-
proximately 2.8-fold larger than the size of the monomers that
inhibited their entry into the cells through passive diffusion.
To evaluate how effective the dimers were relative to the
monomers, we performed pure-protein binding experiments
with twice the concentration of the monomer relative to the
dimer. Previous reports have shown that the use of the SM122-
Tag-IV dimer (10 mm) was relatively effective at blocking the in-
teraction between Hsp90 and HOP or Hsp90 and FKBP52
(blocking 56 and 40% of the binding affinity, respectively).[14]
Thus, we evaluated these four dimers at 15 mm (two C-terminal
modulators linked together) for which the monomer concen-
tration was 30 mm, (e.g., a single C-terminal modulator). Al-
though the same number of C-terminal modulators were pres-
ent in the assay, the dimers were more effective than the mon-
omers for inhibiting the binding between Hsp90 and HOP (Fig-
ure 5a). The two most effective dimers for inhibiting binding
between these two proteins were SM122-Tag-III and SM258-
Tag-III.
Conclusion
The design of dimeric C-terminal modulators has been de-
scribed. The identification of the most effective linker place-
ment for each compound was carried out by using pulldown
assays. The identification was accomplished by quantifying the
amount of Hsp90 that was pulled down by each tagged com-
pound. The linker placement varied depending on the struc-
ture of the compound. The synthesis of four dimers, in which
one molecule was composed of two C-terminal modulator (i.e.,
SM122, SM145, SM253, and SM258) with linkers placed at the
appropriate positions, has been described. Evaluations that
compared the monomer to the dimer in a binding assay by
using Hsp90 and two different C-terminal cochaperones
showed that the dimers synergistically inhibited the binding
events. Although the number of C-terminal modulators were
the same in both the dimer and monomer solutions, the re-
sults indicated that the dimers were more effective at blocking
the Hsp90/cochaperone interactions than the monomers.
These data have indicated that one approach to target Hsp90,
and perhaps other dimeric proteins, would involve the dimeri-
zation of active monomers.
Experimental Section
General information
All moisture-sensitive reactions were performed in nitrogen gas
and were monitored by using thin-layer chromatography (TLC) and
liquid-chromatography mass spectrometry (LC-MS). TLC analysis
was performed on sheets of aluminum silica gel (250 mm; What-
manꢂ (4861-820) with UV light (l=254 nm) as the visualizing
method. The developing agents for TLC analysis included potassi-
um permanganate (general purpose) and ninhydrin (for amine-
group detection). Flash column chromatography on silica gel was
used to purify the crude materials from the synthesis. LC-MS analy-
ses were performed on a LC-MS system connected to a trap run-
ning in positive electrospray ionization (ESI+) mode. The mobile
phase consisted of doubly deionized water with 0.1% (v/v) formic
Figure 5. The inhibitory effects of SM122, SM145, SM253, and SM258 and
the SM122-Tag-III, SM145-Tag-II, SM253-Tag-IV, and SM258-Tag-III dimers on
the binding between Hsp90 and its cochaperones HOP (top) and FKBP52
(bottom). The significant differences between the indicated treatments are
represented with P values (*Pꢁ0.05 and **Pꢁ0.01).
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
acid (solvent A) and HPLC-grade acetonitrile with 0.1% (v/v) formic
acid (solvent B) at a flow rate of 0.5 mLminÀ1, starting at 70% sol-
vent A, 30% solvent B. The gradient elutions were as follows: flow
rate=2 mLminÀ1; initial: 70% solvent A, 30% solvent B held for
35 min; at 35 min: 100% solvent B, held for 18 min; at 53 min:
70% solvent A, 30% solvent B, held for 7 min. 1H and 13C NMR
spectra were obtained and recorded at 258C on Bruker Avance III
500 and 600 MHz spectrometers.
of CH2Cl2. DDLP in solution was added to the bulk solution drop-
wise by using a syringe pump over 2 h. After the addition of all of
the DDLP, the reaction was monitored by using LC-MS after 1–2 h.
The reaction was stirred overnight, if not complete within 2 h.
Upon completion, the crude product was subjected to an acid/
base wash to remove excess DIPEA and the coupling agents. The
resulting crude product was first purified by flash column chroma-
tography, followed by reversed-phase HPLC by using a gradient of
acetonitrile and deionized water with 0.1% trifluoroacetic acid
(TFA) to afford the final pure compounds.
General solid-phase peptide synthesis
Stepwise solid-phase peptide synthesis (SPPS) was performed in
a polypropylene solid-phase extraction cartridge fitted with a poly-
ethylene frit (20 mm) and preloaded 2-chlorotrityl resins with an ap-
proximate loading scale of 0.5 mmolgÀ1 were used. The resin was
weighed, transferred to the cartridge, and swelled in DMF for
30 min prior to peptide coupling in the corresponding sequence.
Synthesis
Experimental methods for SM253-Tag-II
HO-d-Leu-d-3-(4-thiazolyl)-Ala-3,3-diphenyl-d-Ala-Lys(Boc)-N-Me-
Val-NH:
Resin-O-d-Leu-d-3-(4-thiazolyl)-Ala-3,3-diphenyl-d-Ala-
Lys(Boc)-N-Me-Val-NH was synthesized by using resin-O-d-Leu-NH2
(1.0 g, 0.5 mmol, 1.0 equiv) and the subsequent peptide coupling
in the sequence by using an aliquot (1.5 mmol, 3.0 equiv) of each
of the following: Fmoc-d-3-(4-thiazolyl)-Ala-OH (0.59 g), Fmoc-3,3-
diphenyl-d-Ala-OH (0.70 g), Fmoc-Lys(Boc)-OH (0.70 g), and Fmoc-
N-Me-Val-OH (0.53 g). Each peptide coupling was carried out in the
presence of 1-hydroxy-7-aza-benzotriazole (HOAt; 0.20 g, 1.5 mmol,
3.0 equiv) or HOBt (0.20 g, 1.5 mmol, 3.0 equiv), DIC (0.47 mL,
3.0 mmol, 6.0 equiv), and DMF (2.5 mL) to generate a concentration
of 0.20m based on the amino acid. Each coupling reaction was run
for 3 h and a negative ninhydrin test was used to confirm the reac-
tion completion. The reaction mixture was drained to afford the
Fmoc-protected resin-bound pentapeptide and the Fmoc protect-
ing group was removed after completion of each coupling reac-
tion. The linear pentapeptide was cleaved from the resin by using
a solution of TFE (6 mL) and CH2Cl2 (6 mL). The resin-containing so-
lution was filtered and dried in vacuo to yield the DDLP SM253-
Tag-II as a white solid (140 mg, 33%). LC-MS (ESI): m/z calcd for
C44H63N7O8S: 850.46 [M+1]; found: 850.10.
General solid-phase peptide synthesis
Fmoc-protected amino acid coupling reactions were performed in
DMF (0.2m), amino acid (3.0 equiv), 1-hydroxybenzotriazole (HOBt;
3.0 equiv), and diisopropylcarbodiimide (DIC; 6.0 equiv). Coupling
reaction mixture was shaken for a minimum of 4 h on a shaker and
checked by using a ninhydrin test to confirm completion. Once
completed, the coupling reaction mixture was drained, and the
resin was subjected to removal of the Fmoc protecting group.
(Note that 1-hydroxybenzotriazole was replaced with 1-hydroxy-7-
azabenzotriazole for the peptide coupling between Fmoc and the
N-methyl amino terminus and the coupling process was allowed to
run overnight.)
General N-terminal solid-phase amine deprotection
After the peptide coupling process was completed, removal of the
Fmoc protecting group was performed according to the following
steps: DMF (3ꢃ1 min), 20% piperidine/DMF (1ꢃ5 min), 20% piper-
idine/DMF(1ꢃ10 min), DMF (2ꢃ1 min), isopropyl alcohol (IPA; 1ꢃ
1 min), DMF (1ꢃ1 min), IPA (1ꢃ1 min), and DMF (3ꢃ1 min).
cyclo-d-Leu-d-3-(4-thiazoyl)-Ala-3,3-diphenyl-d-Ala-Lys(Boc)-N-
Me-Val: The product was afforded by using DDLP SM253-Tag-II
(0.11 g, 0.13 mmol, 1.0 equiv), TBTU (0.021 g, 0.065 mmol,
0.50 equiv), HATU (0.017 g, 0.065 mmol, 0.50 equiv), DMTMM
(0.019 g, 0.065 mmol, 0.50 equiv), and DIPEA (0.27 mL, 1.55 mmol,
12.0 equiv) in anhydrous CH2Cl2 (129 mL, 0.001m) and following
the macrocyclization procedure. The reaction was stirred overnight
and the reaction was monitored by TLC and LC-MS. Upon comple-
tion, the reaction mixture was subjected to an acid/base wash to
afford the crude product, which was purified by flash column chro-
matography on silica gel with an ethyl acetate/hexane gradient
system as the eluent, followed by purification by HPLC to yield the
Boc-protected macrocycle as a white solid (23 mg, 21.4%). LC-MS
(ESI): m/z calcd for C44H61N7O7S: 832.44 [M+1]; found: 832.20.
Cleavage of linear peptide
The eventual cleavage of a linear pentapeptide from the resin was
carried out by swelling the resin in a solution of 2,2,2-trifluoroetha-
nol (TFE)/CH2Cl2 (1:1 (v/v); 10 mL per gram of dried resin) and was
allowed to stir for 24 h. The suspension was filtered through
a Bꢄchner filter, and the resin was washed repeatedly with addi-
tional CH2Cl2 to fully extract the cleaved peptide. The filtrate was
evaporated and dried in vacuo overnight. The dried solid was
eventually redissolved in CH2Cl2, coevaporated with CH2Cl2 several
times to remove the entrapped TFE residue completely, and dried
in vacuo overnight.
The Boc protecting group of the macrocycle was removed by uti-
lizing a mixture of TFA/CH2Cl2 (1:4, 0.1m) and anisole (2.0 equiv) to
generate free amine groups in the lysine residue. The free amine
compound was used in the subsequent biotinylation reaction with-
out purification. LC-MS (ESI): m/z calcd for C39H53N7O5S: 732.38
[M+1]; found: 732.00.
Macrocyclization procedure (syringe pump)
Macrocyclization of the double-deprotected linear pentapeptide by
using a combination of three coupling agents 4-(4,6-dimethoxy-
1,3,5-triazin-2-yl)-4-methylmorpholin-4-ium (DMTMM), (7-azabenzo-
triazol-1-yl)tetramethyluronium (HATU), and 2-(1H-benzotriazole-1-
yl)-1,1,3,3-tetramethyluronium tetrafluoroborate (TBTU; 0.8 equiv
each) with N,N-diisopropylethylamine (DIPEA; 8.0 equiv) in 75% of
a calculated volume of anhydrous CH2Cl2 to generate an overall
concentration of 0.001m. The crude anhydrous double-deprotect-
ed linear peptide (DDLP) was dissolved in the remaining amount
The biotinylated SM253-Tag-II was afforded by utilizing the depro-
tected macrocycle (23 mg, 0.031 mmol, 1.0 equiv), NHS-(PEG)4-
Biotin (25.9 mg, 0.044 mmol, 1.4 equiv), and DIPEA (43.8 mL,
0.25 mmol, 8.0 equiv) in CH2Cl2 (314 mL). The reaction mixture was
stirred for 4 h and monitored by LC-MS. Upon completion, the
crude product was purified by using preparative HPLC to generate
pure the biotinylated compound as a white solid (12 mg, 33.1%).
&
&
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
6
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
1
Rf =0.41 (EtOAc/MeOH=0.60:0.40); H NMR (600 MHz, MeOD): d=
tinylated compound as a white solid (13 mg, 15.3%). Rf =0.45
(EtOAc/MeOH=1:1); 1H NMR (600 MHz, CDCl3): d=8.69 (d, J=
1.8 Hz, 1H), 8.20 (d, J=5.6 Hz; NH), 7.77 (br; NH), 7.30–7.17 (m,
10H), 6.94 (br; NH), 6.86 (br; NH), 6.42 (m, 1H), 5.68 (br; NH), 5.23
(t, J=10.2 Hz, 1H), 5.16 (dd, J=10.5, 4.9 Hz, 1H), 4.67 (d, J=
10.9 Hz, 1H), 4.57 (br; NH), 4.52 (m, 1H), 4.41 (m, 1H), 4.17 (m, 1H),
3.74 (t, J=6.0 Hz, 2H), 3.64 (m, 12H), 3.55 (t, J=5.0 Hz, 2H), 3.37
(m, 1H), 3.29 (m, 2H), 3.20 (m, 2H), 2.95 (m, 1H), 2.91 (s, 3H), 2.81
(m, 1H), 2.46 (m, 2H), 2.30 (m, 6H), 2.10 (m, 1H), 1.81–1.70 (m,
7H), 1.61–1.43 (m, 8H), 1.31 (m, 2H), 0.94–0.81 ppm (m, 12H);
13C NMR (150 MHz, CDCl3): d=174.2, 173.3, 172.4, 171.9, 171.5,
171.2, 170.5, 163.9, 152.8, 152.6, 141.2, 140.5, 128.7, 128.6, 128.3,
128.2, 126.9, 126.8, 115.9, 70.6, 70.5, 70.4, 70.3, 70.2, 70.1, 70.0,
69.9, 67.4, 62.1, 62.0, 60.1, 57.7, 56.7, 55.9, 55.4, 52.0, 50.8, 49.5,
40.6, 39.6, 39.2, 39.1, 36.9, 36.0, 31.1, 30.6, 28.7, 28.4, 28.2, 25.7,
24.8, 24.7, 23.4, 23.2, 22.6, 22.4, 22.0 ppm; LC-MS (ESI): m/z calcd
for C61H90N10O12S2: 1219.62 [M+1]; found: 1219.00, 610.00 (half
mass); HRMS (ESI-TOF): m/z calcd for C61H90N10O12S2Na: 1241.6079
[M+Na+]; found: 1241.6074.
9.34 (m, 1H), 7.28–7.04 (m, 11H), 5.19–5.16 (m, 1H), 4.54–4.48 (m,
2H), 4.43–4.41 (m, 1H), 4.25–4.19 (m, 2H), 4.07 (s, 1H), 4.02 (s, 1H),
3.26 (t, J=6.4 Hz, 3H), 3.53–3.49 (m, 14H), 3.45–3.42 (m, 3H), 3.28–
3.24 (m, 3H), 2.99–2.95 (m, 2H), 2.82 (s, 1H), 2.75 (s, 3H), 2.35–2.32
(m, 3H), 2.17–2.11 (m, 4H), 1.82–1.76 (m, 1H), 1.68–1.40 (m, 12H),
1.29–1.19 (m, 5H), 0.82–0.72 ppm (m, 12H); 13C NMR (150 MHz,
MeOD): d=174.8, 172.4, 171.6, 170.9, 170.2, 170.0, 161.9, 156.1,
148.3, 140.8, 140.6, 128.6, 128.5, 128.4, 128.3, 128.2, 128.1, 128.0,
127.9, 127.8, 126.9, 126.6, 118.5, 70.6, 70.4, 70.3, 70.2, 70.1, 70.0,
69.9, 69.8, 69.2, 66.9, 63.7, 62.9, 58.0, 56.9, 56.8, 52.7, 48.5, 48.2,
39.0, 38.8, 36.2, 35.3, 34.9, 29.9, 29.1, 28.7, 28.5, 28.4, 28.1, 26.7,
25.5, 25.3, 25.1, 24.3, 23.1, 22.0, 21.2, 21.1, 18.8, 18.7, 17.8 ppm; LC-
MS (ESI): m/z calcd for C60H88N10O12S2: 1205.60 [M+1]; found:
1205.05; HRMS (ESI-TOF): m/z calcd for C60H88N10O12S2: 1204.6025
[M+Na+]; found: 1227.5926.
Experimental methods for SM253-Tag-III
HO-Leu-N-Me-Lys(Boc)-d-Leu-d-3-(4-thiazolyl)-Ala-3,3-diphenyl-
d-Ala-NH2: Resin-O-Leu-N-Me-Lys(Boc)-d-Leu-d-3-(4-thiazolyl)-Ala-
3,3-diphenyl-d-Ala-NH2 was synthesized by using resin-O-Leu-NH2
(1.0 g, 0.5 mmol, 1.0 equiv) and the subsequent peptide coupling
in the sequence by using an aliquot (1.5 mmol, 3.0 equiv) of each
of the following: Fmoc-N-Me-Lys(Boc)-OH (0.72 g), Fmoc-d-Leu-OH
(0.53 g), Fmoc-d-3-(4-thiazolyl)-Ala-OH (0.59 g), and Fmoc-3,3-di-
phenyl-d-Ala-OH (0.70 g). Each peptide coupling was carried out in
the presence of HOAt (0.20 g, 1.5 mmol, 3.0 equiv) or HOBt (0.20 g,
1.5 mmol, 3.0 equiv), DIC (0.47 mL, 3.0 mmol, 6.0 equiv), and DMF
(2.5 mL) to generate a concentration of 0.20m based on the amino
acid. Each coupling reaction was run for 3 h and a negative ninhy-
drin test was used to confirm the reaction completion. The resin-
bound linear pentapeptide was cleaved from the resin by using
a solution of TFE (6 mL) and CH2Cl2 (6 mL) The resin-containing so-
lution was filtered and dried in vacuo to yield the DDLP SM253-
Tag-III as a white solid (352 mg, 81.5%). LC-MS (ESI): m/z calcd for
C45H65N7O8S: 864.46 [M+1]; found: 864.20.
Experimental methods for SM253-Tag-IV
HO-Leu-N-Me-Val-d-Lys(Boc)-d-3-(4-thiazolyl)-Ala-3,3-diphenyl-d-
Ala-NH2: Resin-O-Leu-N-Me-Val-d-Lys(Boc)-d-3-(4-thiazolyl)-Ala-3,3-
diphenyl-d-Ala-NH2 was synthesized by using resin-O-Leu-NH2
(1.0 g, 0.50 mmol, 1.0 equiv) and the subsequent peptide coupling
in the sequence with an aliquot (1.5 mmol, 3.0 equiv) of the follow-
ing: Fmoc-N-Me-Val-OH (0.53 g), Fmoc-d-Lys(Boc)-OH (0.53 g),
Fmoc-d-3-(4-thiazolyl)-Ala-OH (0.59 g), and Fmoc-3,3-diphenyl-d-
Ala-OH (0.70 g). Each peptide coupling was carried out in the pres-
ence of HOAt (0.20 g, 1.5 mmol, 3.0 equiv) or HOBt (0.20 g,
1.5 mmol, 3.0 equiv), DIC (0.47 mL, 3.0 mmol, 6.0 equiv) and DMF
(2.5 mL) to generate a concentration of 0.20m based on the amino
acid. Each coupling reaction was carried out for 3 h, and a negative
ninhydrin test was used to confirm the reaction completion. The
DDLP SM253-Tag-IV was cleaved from the resin by using a solution
of TFE (6 mL) and CH2Cl2 (6 mL). The resin-containing solution was
filtered and dried in vacuo to yield the DDLP SM253-Tag-IV as
a white solid (430 mg, quantitative). LC-MS (ESI): m/z calcd for
C44H63N7O8S: 850.45 [M+1]; found: 850.70.
cyclo-Leu-N-Me-Lys(Boc)-d-Leu-d-3-(4-thiazolyl)-Ala-3,3-diphenyl-
d-Ala: The product was afforded by using DDLP SM253-Tag-III
(0.35 g, 0.41 mmol, 1.0 equiv), TBTU (0.092 g, 0.29 mmol,
0.70 equiv), HATU (0.15 g, 0.41 mmol, 1.0 equiv), DMTMM (0.079 g,
0.29 mmol, 0.70 equiv), DIPEA (0.57 mL, 3.26 mmol, 8.0 equiv) in
anhydrous CH2Cl2 (407 mL, 0.001m) and following the macrocycli-
zation procedure. The reaction was stirred overnight and the reac-
tion was monitored by TLC and LC-MS. Upon completion, the reac-
tion mixture was subjected to an acid/base wash to afford the
crude product, which was purified by flash column chromatogra-
phy on silica gel with an ethyl acetate/hexane gradient system as
the eluent, followed by purification by HPLC to yield the Boc-pro-
tected macrocycle as a white solid (105 mg, 30.5%). LC-MS (ESI):
m/z calcd for C45H63N7O7S: 846.45 [M+1]; found: 846.50.
cyclo-Leu-N-Me-Val-d-Lys(Boc)-d-3-(4-thiazolyl)-Ala-3,3-diphenyl-
d-Ala: The product was generated from DDLP SM253-Tag-IV
(0.43 g, 0.51 mmol, 1.0 equiv) with TBTU (0.11 g, 0.35 mmol, 0.70
equiv), HATU (0.19 g, 0.51 mmol, 1.0 equiv), DMTMM (0.10 g,
0.35 mmol, 0.70 equiv), and DIPEA (0.70 mL, 4.05 mmol, 8.0 equiv)
in anhydrous CH2Cl2 (506 mL, 0.001m) and following the macrocyc-
lization procedure. The reaction was stirred overnight and the reac-
tion was monitored by TLC and LC-MS. Upon completion, the reac-
tion mixture was subjected to an acid/base wash to afford the
crude product, which was purified by flash column chromatogra-
phy on silica gel with an ethyl acetate/hexane gradient system as
the eluent, followed by purification by HPLC to yield the Boc-pro-
tected macrocycle as a white solid (124 mg, 29.5%). LC-MS (ESI):
m/z calcd for C44H61N7O7S: 832.44 [M+1]; found: 831.95.
The Boc protecting group of the macrocycle was removed by uti-
lizing a mixture of TFA/CH2Cl2 (1:4, 0.1m) and anisole (2.0 equiv) to
generate free amine groups in the lysine residue. The free amine
compound was used in the subsequent biotinylation reaction with-
out purification. LC-MS (ESI): m/z calcd for C40H55N7O5S: 746.40
[M+1]; found: 746.50.
The Boc protecting group of the macrocycle was removed by uti-
lizing a mixture of TFA/CH2Cl2 (1:4, 0.1m) and anisole (2.0 equiv) to
generate free amine groups in the lysine residue. The free amine
compound was used in the subsequent biotinylation reaction with-
out purification. LC-MS (ESI): m/z calcd for C39H53N7O5S: 732.38
[M+1]; found: 732.15.
The biotinylated SM253-Tag-III was afforded by utilizing (52 mg,
0.070 mmol, 1.0 equiv) of the deprotected macrocycle, NHS-(PEG)4-
Biotin (57.4 mg, 0.098 mmol, 1.4 equiv), DIPEA (97.1 mL, 0.56 mmol,
8.0 equiv) in CH2Cl2 (697 mL). The reaction mixture was stirred for
4 h and monitored by LC-MS. Upon completion, the crude product
was purified by using preparative HPLC to generate the pure bio-
The biotinylated SM253-Tag-IV was afforded by using utilizing the
deprotected macrocycle (86.1 mg, 0.12 mmol, 1.0 equiv), NHS-
(PEG)4-Biotin (96.9 mg, 0.16 mmol, 1.4 equiv), and DIPEA (163.9 mL,
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
7
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
0.94 mmol, 8.0 equiv) in CH2Cl2 (1.18 mL). The reaction mixture was
stirred for 4 h and monitored by LC-MS. Upon completion, the
crude product was purified by using preparative HPLC to generate
pure the biotinylated compound as a white solid (25 mg, 17.3%).
cyclo-Lys(Boc)-N-Me-Val-d-3-(4-thiazolyl)-Ala-d-Phe-3,3-diphenyl-
d-Ala: The product was generated by using DDLP SM258-Tag-II
(0.40 g, 0.45 mmol, 1.0 equiv), TBTU (0.10 g, 0.32 mmol, 0.70 equiv),
HATU (0.21 g, 0.54 mmol, 1.20 equiv), DMTMM (0.063 g, 0.23 mmol,
0.50 equiv), and DIPEA (0.63 mL, 3.62 mmol, 8.0 equiv) in anhy-
drous CH2Cl2 (452 mL, 0.001m) and following the macrocyclization
procedure. The reaction was stirred overnight and the reaction was
monitored by TLC and LC-MS. Upon completion, the reaction mix-
ture was subjected to an acid/base wash to afford the crude prod-
uct, which was purified by flash column chromatography on silica
gel with an ethyl acetate/hexane gradient system as the eluent,
followed by purification by HPLC to yield the Boc-protected macro-
cycle as a white solid (82 mg, 20.9%). LC-MS (ESI): m/z calcd for
C47H59N7O7S: 866.42 [M+1]; found: 866.05.
1
Rf =0.38 (EtOAc/MeOH=1:1); H NMR (600 MHz, CDCl3): d=8.89 (s,
1H), 8.25 (br; NH), 7.57 (d, J=8.6 Hz; NH), 7.44 (d, J=7.5 Hz; NH),
7.36–7.14 (m, 10H), 7.07 (br; NH), 6.90 (m; NH), 6.76 (m; NH), 5.21
(t, J=10.3 Hz, 1H), 4.83 (d, J=10.6 Hz, 1H), 4.73 (d, J=10.3 Hz,
1H), 4.59–4.51 (m, 3H), 4.40 (m, 2H), 4.02 (m, 1H), 3.74 (m, 3H),
3.64–3.61 (m, 13H), 3.42 (m, 3H), 3.27 (m, 2H), 3.20 (m, 3H), 3.12
(m, 2H), 2.96 (m, 1H), 2.83 (m, 5H), 2.47 (m, 2H), 2.32 (m, 1H), 2.24
(m, 2H), 1.72–1.61 (m, 6H), 1.50 (m, 4H), 1.29 (m, 1H), 0.83–
0.72 ppm (m, 12H); 13C NMR (150 MHz, CDCl3): d=173.3, 171.4,
170.9, 170.5, 170.1, 169.5, 164.4, 152.7, 152.5, 140.9, 140.7, 128.8,
128.6, 128.2, 127.9, 127.1, 127.0, 115.9, 70.6, 70.5, 70.4, 70.3, 70.2,
70.1, 69.8, 67.5, 63.4, 62.2, 62.1, 60.2, 58.4, 57.5, 55.9, 55.6, 51.9,
50.5, 49.5, 40.8, 39.3, 37.1, 35.9, 32.1, 31.8, 30.9, 29.1, 28.2, 26.0,
25.6, 24.8, 23.0, 22.8, 22.5, 22.4, 19.8, 19.1 ppm; LC-MS (ESI): m/z
calcd for C60H88N10O12S2: 1205.60 [M+1]; found: 1205.00 [M+1],
603.00 (half mass); HRMS (ESI-TOF): m/z calcd for C60H88N10O12S2Na:
1227.5922 [M+Na+]; found: 1227.5918.
The Boc protecting group of the macrocycle was removed by uti-
lizing a mixture of TFA/CH2Cl2 (1:4, 0.1m) and anisole (2.0 equiv) to
generate free amine groups in the lysine residue. The free amine
compound was used in the subsequent biotinylation reaction with-
out purification. LC-MS (ESI): m/z calcd for C42H51N7O5S: 766.37
[M+1]; found: 766.00.
The biotinylated SM258-Tag-II was afforded by using the depro-
tected macrocycle (40 mg, 0.052 mmol, 1.0 equiv), NHS-(PEG)4-
Biotin (43 mg, 0.073 mmol, 1.4 equiv), and DIPEA (72.8 mL,
0.42 mmol, 8.0 equiv) in CH2Cl2 (522 mL). The reaction mixture was
stirred for 4 h and monitored by LC-MS. Upon completion, the
crude product was purified by using preparative HPLC to generate
pure biotinylated compound as a white solid (7.8 mg, 12.1%). Rf =
0.38 (EtOAc/MeOH=1:1); 1H NMR (600 MHz, DMSO): d=8.96 (d,
J=1.9 Hz, 1H), 8.27 (d, J=8.2 Hz; 2NH), 7.84 (t, J=5.6 Hz; NH),
7.76 (m; 2NH), 7.74 (t, J=5.4 Hz; NH), 7.23–7.15 (m, 13H), 6.88 (m,
2H), 6.42 (s, 1H), 6.36 (s, 1H), 5.15 (m, 1H), 4.99 (dd, J=15.0,
8.0 Hz, 1H), 4.52 (d, J=11.3 Hz, 1H), 4.31 (m, 2H), 4.13 (m, 1H),
4.05 (m, 1H), 3.59 (t, J=6.4 Hz, 2H), 3.50 (m, 14H), 3.40 (t, J=
5.9 Hz, 2H), 3.19 (dd, J=11.7, 5.8 Hz, 1H), 3.10 (m, 1H), 2.91 (m,
4H), 2.82 (dd, J=12.4, 5.1 Hz, 1H), 2.69 (dd, J=14.0, 6.3 Hz, 1H),
2.62 (m, 3H), 2.59 (m, 2H), 2.29 (t, J=6.7 Hz, 2H), 2.07 (t, J=7.5 Hz,
2H), 1.61 (m, 2H), 1.51 (m, 4H), 1.31 (m, 4H), 0.92 (m, 2H), 0.81 (d,
J=6.4 Hz, 3H), 0.59 ppm (d, J=6.6 Hz, 3H); 13C NMR (150 MHz,
DMSO): d=172.6, 170.6, 170.2, 170.1, 169.9, 169.8, 168.9, 163.2,
154.0, 153.5, 141.7, 141.1, 137.7, 129.2, 129.1, 129.0, 128.8, 128.6,
128.4, 127.1, 126.9, 126.8, 115.4, 70.3, 70.2, 70.1, 70.0, 69.9, 69.6,
67.4, 67.3, 62.9, 61.5, 60.2, 59.7, 57.3, 56.7, 55.9, 53.9, 53.7, 49.4,
40.5, 40.3, 38.9, 37.7, 36.6, 35.6, 33.5, 30.2, 29.2, 28.7, 28.5, 25.7,
25.3, 23.6, 21.2, 20.0, 19.8 ppm; LC-MS (ESI): m/z calcd for
C63H86N10O12S2: 1239.59 [M+1] [M+1]; found: 606.00 (half mass)
[M+1]; HRMS (ESI-TOF): m/z calcd for C63H86N10O12S2Na: 1261.5766
[M+Na+]; found: 1261.5763.
Experimental methods for SM253-Tag-IV dimer
SM253-Tag-IV dimer: The product was synthesized by utilizing
cyclo-Leu-N-Me-Val-d-Lys-d-3-(4-thiazolyl)-Ala-3,3-diphenyl-d-Ala
(31.2 mg, 0.043 mmol, 1.0 equiv), BS(PEG)9 (15.1 mg, 0.021 mmol,
0.5 equiv), and DIPEA (59.4 mL, 8.0 equiv) dissolved in CH2Cl2
(426 mL). The crude material was purified by using HPLC to obtain
the product (5 mg, 6%). 1H NMR (600 MHz, DMSO): d=8.93 (m,
2H), 8.39 (m; 2NH), 7.56–7.16 (m, 24H), 6.87 (m, 2H), 6.67 (m;
2NH), 5.06 (m; 2NH), 4.87 (m; 2NH), 4.62 (m; 2NH), 4.46 (m, 2H),
4.21 (m; 2NH), 3.64 (m, 34H), 3.22 (m, 6H), 2.80 (m, 6H), 2.49 (m,
8H), 1.48 (m, 10H), 1.25 (m, 14H), 0.99–0.74 ppm (m, 24H);
13C NMR (150 MHz, DMSO): d=171.6, 171.5, 171.4, 171.3, 170.8,
170.6, 152.2, 141.1, 140.5, 128.9, 128.8, 128.6, 128.2, 127.1, 126.9,
116.5, 70.6, 70.5, 70.4, 70.3, 70.2, 67.4, 59.5, 58.1, 49.4, 39.1, 39.0,
38.2, 37.0, 31.9, 31.6, 31.2, 29.8, 28.8, 26.1, 22.8, 22.7, 19.8,
19.7 ppm; LC-MS (ESI): m/z calcd for C100H144N14O21S2: 1942.01 [M+
1]; found: 993.00 (half mass); HRMS (ESI-TOF): m/z calcd for
C100H144N14O21S2 1942.0072 [M+H+]; found: 1943.0162.
Experimental methods for SM258-Tag-II
HO-Lys(Boc)-N-Me-Val-d-3-(4-thiazolyl)-Ala-d-Phe-3,3-diphenyl-d-
Ala-NH2: Resin-O-Lys(Boc)-N-Me-Val-d-3-(4-thiazolyl)-Ala-d-Phe-3,3-
diphenyl-d-Ala-NH2 was synthesized by using resin-O-Lys(Boc)-NH2
(1.0 g, 0.5 mmol, 1.0 equiv), and the subsequent peptide coupling
in the sequence was carried out by using an aliquot (1.5 mmol,
3.0 equiv) of each of the following: Fmoc-N-Me-Val-OH (0.53 g),
Fmoc-d-3-(4-thiazolyl)-Ala-OH (0.59 g), Fmoc-d-Phe-OH (0.58 g),
and Fmoc-3,3-diphenyl-d-Ala-OH (0.70 g). Each peptide coupling
was carried out in the presence of HOAt (0.20 g, 1.5 mmol,
3.0 equiv) or HOBt (0.20 g, 1.5 mmol, 3.0 equiv), DIC (0.47 mL,
3.0 mmol, 6.0 equiv), and DMF (2.5 mL) to generate a concentration
of 0.20m based on the amino acid. Each coupling reaction was car-
ried out for 3 h, and a negative ninhydrin test was used to confirm
the reaction completion. The Fmoc protecting group was removed
after the completion of each coupling reaction. The DDLP SM258-
Tag-II was cleaved from the resin by using a solution of TFE (6 mL)
and CH2Cl2 (6 mL). The resin-containing solution was filtered and
dried in vacuo to yield the DDLP SM258-Tag-II as a white solid
(400 mg, 90.5%). LC-MS (ESI): m/z calcd for C47H61N7O8S: 884.43
[M+1]; found: 884.10.
Experimental methods for SM258-Tag-III
HO-Leu-N-Me-Lys(Boc)-d-3-(4-thiazolyl)-Ala-d-Phe-3,3-diphenyl-
d-Ala-NH2: Resin-O-Leu-N-Me-Lys(Boc)-d-3-(4-thiazolyl)-Ala-d-Phe-
3,3-diphenyl-d-Ala-NH2 was synthesized by using resin-O-Leu-NH2
(1.0 g, 0.5 mmol, 1.0 equiv) and the subsequent peptide coupling
in the sequence by using an aliquot (1.5 mmol, 3.0 equiv) of each
of the following: Fmoc-N-Me-Lys(Boc)-OH (0.72 g), Fmoc-d-3-(4-
thiazolyl)-Ala-OH (0.59 g), Fmoc-d-Phe-OH (0.58 g), and Fmoc-3,3-
diphenyl-d-Ala-OH (0.70 g). Each peptide coupling was carried out
in the presence of HOAt (0.20 g, 1.5 mmol, 3.0 equiv) or HOBt
(0.20 g, 1.5 mmol, 3.0 equiv), DIC (0.47 mL, 3.0 mmol, 6.0 equiv),
and DMF (2.5 mL) to generate a concentration of 0.20m based on
the amino acid. Each coupling reaction was carried out for 3 h, and
a negative ninhydrin test was used to confirm the reaction comple-
&
&
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
8
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
tion. The Fmoc protecting group was removed after the comple-
tion of each coupling reaction. HO-Leu-N-Me-Lys(Boc)-d-3-(4-thia-
zolyl)-Ala-d-Phe-3,3-diphenyl-d-Ala-NH2 was obtained from the
resin cleavage by using a solution of TFE (6 mL) and CH2Cl2 (6 mL).
The resin-containing solution was filtered and dried in vacuo to
yield the DDLP SM258-Tag-III as a white solid (352 mg, 78.4%). LC-
MS (ESI): m/z calcd for C48H63N7O8S: 898.45 [M+1]; found: 898.10.
the product (2 mg, 7.6%). 1H NMR (600 MHz, DMSO): d=9.03 (m,
2H), 8.36 (m; 2NH), 8.06 (m; 2NH), 7.91(d, J=8.6 Hz; 2NH), 7.78
(m; 2NH), 7.52 (d, J=6.2 Hz; 2NH), 7.29–7.13 (m, 28H), 6.95 (d, J=
5.8 Hz, 4H), 5.06 (t, J=10.4 Hz, 2H), 4.85 (dd, J=14.1, 7.1 Hz, 2H),
4.81 (dd, J=10.2, 5.7 Hz, 2H), 4.37 (d, J=11.6 Hz, 4H), 3.97 (dd, J=
14.1, 7.9 Hz, 2H), 3.72 (m, 2H), 3.58 (m, 8H), 3.50 (m, 26H), 3.24
(dd, J=14.2, 8.3 Hz, 2H), 3.05 (dd, J=14.2, 6.0 Hz, 2H), 2.97 (m,
6H), 2.72 (m, 2H), 2.58 (s, 6H), 2.29 (m, 4H), 1.71 (m, 2H), 1.56 (m,
2H), 1.48 (m, 2H), 1.31 (m, 9H), 1.11 (m, 2H), 0.93 (m, 3H), 0.67 (d,
J=6.5 Hz, 6H), 0.58 ppm (d, J=6.5 Hz, 6H); 13C NMR (150 MHz,
DMSO): d=175.2, 173.5, 172.3, 171.7, 170.3, 170.0, 159.9, 153.6,
142.0, 141.5, 138.6, 129.8, 129.4, 128.9, 128.6, 127.3, 127.1, 126.8,
126.7, 125.0, 116.1, 72.0, 70.2, 67.4, 59.4, 57.4, 56.9, 52.8, 50.4, 38.9,
37.0, 36.7, 36.3, 33.2, 30.0, 29.1, 26.9, 25.5, 24.8, 23.2, 23.0 ppm; LC-
MS (ESI): m/z calcd for C108H144N14O21S2: 2038.01 [M+1]; found:
1020.00; HRMS (ESI-TOF): m/z calcd for C108H144N14O21S2: 2059.9970
[M+Na+]; found: 2060.9998.
cyclo-Leu-N-Me-Lys(Boc)-d-3-(4-thiazolyl)-Ala-d-Phe-3,3-diphen-
yl-d-Ala: The product was synthesized by using DDLP SM258-Tag-
III (0.35 g, 0.39 mmol, 1.0 equiv), TBTU (0.088 g, 0.27 mmol,
0.70 equiv), HATU (0.15 g, 0.39 mmol, 1.0 equiv), DMTMM (0.076 g,
0.27 mmol, 0.70 equiv), and DIPEA (0.55 mL, 3.13 mmol, 8.0 equiv)
in anhydrous CH2Cl2 (392 mL, 0.001m) and following the macrocyc-
lization procedure. The reaction was stirred overnight and the reac-
tion was monitored by TLC and LC-MS. Upon completion, the reac-
tion mixture was subjected to an acid/base wash to afford the
crude product, which was purified by flash column chromatogra-
phy on silica gel with an ethyl acetate/hexane gradient system as
the eluent, followed by purification by HPLC to yield the Boc-pro-
tected macrocycle as a white solid (150 mg, 43.7%). LC-MS (ESI):
m/z calcd for C48H61N7O7S: 880.44 [M+1]; found: 880.10.
Experimental methods for SM145-Tag-II dimer
HO-d-Phe-racemic-b-OH-Phe-Lys(Boc)-N-Me-Val-d-Leu-NH2:
Resin-O-d-Phe-racemic-b-OH-Phe-Lys(Boc)-N-Me-Val-d-Leu-NH2 was
synthesized by using resin-O-d-Phe-NH2 (1.5 g, 0.74 mmol,
1.0 equiv) and the subsequent peptide coupling in the sequence
by using an aliquot (2.2 mmol, 3.0 equiv) of each of the following:
Fmoc-(2S, 3R)/(2R, 3S)-b-OH-Phe-OH (0.89 g), Fmoc-Lys(Boc)-OH
(1.03 g), Fmoc-N-Me-Val-OH (0.78 g), and Fmoc-d-Leu-OH (0.78 g).
Each peptide coupling was carried out in the presence of HOAt
(0.30 g, 2.2 mmol, 3.0 equiv) or HOBt (0.30 g, 2.2 mmol, 3.0 equiv),
DIC (0.69 mL, 4.4 mmol, 6.0 equiv), and DMF (3.68 mL) to generate
a concentration of 0.20m based on the amino acid. Each coupling
reaction was carried out for 3 h, and a negative ninhydrin test was
used to confirm the reaction completion. The Fmoc protecting
group was removed after completion of each coupling reaction.
The DDLP SM145-Tag-II was generated by using a solution of TFE
(7 mL) and CH2Cl2 (7 mL). The resin-containing solution was filtered
and dried in vacuo to yield the DDLP as a white solid (263 mg,
45.4%). LC-MS (ESI): m/z calcd for C41H62N6O9: 783.46 [M+1];
found: 783.20.
The Boc protecting group of the macrocycle was removed by uti-
lizing a mixture of TFA/CH2Cl2 (1:4, 0.1m) and anisole (2.0 equiv) to
generate free amine groups in the lysine residue. The free amine
compound was used in the subsequent biotinylation reaction with-
out purification. LC-MS (ESI): m/z calcd for C43H53N7O5S: 780.38
[M+1]; found: 780.45.
The biotinylated SM258-Tag-III was afforded by using the depro-
tected SM258-Tag-III (52 mg, 0.067 mmol, 1.0 equiv), NHS-(PEG)4-
Biotin (54.9 mg, 0.093 mmol, 1.4 equiv), and DIPEA (92.9 mL,
0.25 mmol, 8.0 equiv) in CH2Cl2 (522 mL). The reaction mixture was
stirred for 4 h and monitored by LC-MS. Upon completion, the
crude product was purified by using preparative HPLC to generate
pure biotinylated compound as a white solid (18 mg, 21.4%). Rf =
1
0.35 (EtOAc/MeOH=0.30:0.70); H NMR (600 MHz, DMSO): d=8.99
(d, J=1.9 Hz, 1H), 8.35 (d, J=6.7 Hz; NH), 8.14 (br; NH), 7.98 (d, J=
8.9 Hz; NH), 7.85 (t, J=5.6 Hz; NH), 7.76 (m; NH), 7.52 (d, J=6.4 Hz;
NH), 7.26–7.15 (m, 15H), 6.95 (d, J=5.7 Hz, 1H), 6.80 (s, 1H), 6.70
(s, 1H), 5.06 (m, 1H), 4.85 (dd, J=14.6, 6.9 Hz, 1H), 4.81 (dd, J=
10.3, 5.7 Hz, 1H), 4.46 (m, 1H), 4.34 (m, 2H), 3.97 (dd, J=14.6,
8.2 Hz, 1H), 3.71 (dd, J=14.2, 6.7 Hz, 1H), 3.59 (m, 2H), 3.50 (m,
14H), 3.40 (t, J=5.9 Hz, 2H), 3.35 (dd, J=12.8, 1.7 Hz, 1H), 3.24 (m,
1H), 3.19 (m, 2H), 3.04 (dd, J=14.1, 6.0 Hz, 1H), 2.97 (m, 2H), 2.93
(m, 1H), 2.69 (dd, J=13.7, 8.7 Hz, 1H), 2.58 (s, 3H), 2.29 (m, 2H),
2.09 (t, J=7.4 Hz, 1H), 1.73 (m, 2H), 1.55 (m, 2H), 1.35 (m, 8H),
0.93 (m, 2H), 0.69 (m, 3H), 0.57 ppm (m, 3H); 13C NMR (150 MHz,
DMSO): d=172.6, 171.3, 170.8, 170.7, 170.3, 170.2, 170.1, 161.7,
153.9, 153.5, 141.8, 141.3, 138.2, 129.3, 128.8, 128.7, 128.6, 128.5,
128.4, 128.1, 126.9, 126.7, 116.0, 70.5, 70.4, 70.3, 70.2, 70.1, 70.0,
69.9, 69.6, 67.3, 59.1, 57.5, 57.4, 57.1, 56.3, 53.2, 52.9, 52.8, 50.4,
38.9, 38.7, 37.1, 36.6, 35.4, 33.2, 30.2, 29.1, 27.1, 26.9, 25.6, 25.5,
24.7, 24.2, 23.1, 23.0, 22.4, 22.1 ppm; LC-MS (ESI): m/z calcd for
C64H88N10O12S2: 1253.60 [M+1]; found: 1253.00; HRMS (ESI-TOF):
m/z calcd for C64H88N10O12S2Na: 1275.5922 [M+Na+]; found:
1275.5920.
cyclo-d-Phe-racemic-b-OH-Phe-Lys(Boc)-N-Me-Val-d-Leu:
The
product was generated by using DDLP SM145-Tag-II (0.26 g,
0.34 mmol, 1.0 equiv), TBTU (0.086 g, 0.27 mmol, 0.80 equiv), HATU
(0.10 g, 0.27 mmol, 0.8 equiv), DMTMM (0.074 g, 0.27 mmol,
0.80 equiv), and DIPEA (0.47 mL, 2.69 mmol, 8.0 equiv) in anhy-
drous CH2Cl2 (336 mL, 0.001m) and following the macrocyclization
procedure. The reaction was stirred overnight and the reaction was
monitored by TLC and LC-MS. Upon completion, the reaction mix-
ture was subjected to an acid/base wash to afford the crude prod-
uct, which was purified by flash column chromatography on silica
gel with an ethyl acetate/hexane gradient system as the eluent,
followed by purification by HPLC to yield SM145-Tag-II as a white
solid (150 mg, 43.7%). LC-MS (ESI): m/z calcd for C41H60N6O8: 765.45
[M+1]; found: 765.15.
cyclo-d-Phe-b-benzoxy-Phe-Lys(Boc)-N-Me-Val-d-Leu: The prod-
uct was synthesized following the benzylation procedure by using
a mixture of the macrocycle (84.4 mg, 0.11 mmol, 1.0 equiv), NaH
(60% in mineral oil; 7.9 mg, 0.33 mmol, 3.0 equiv), BnBr (37.7 mL,
0.22 mmol, 2.0 equiv), and anhydrous THF/DMF (1:1; 112 mL, 0.1m).
Upon completion, as determined by using LC-MS, the reaction mix-
ture was extracted with deionized water and dichloromethane. The
organic layer was collected, dried, and concentrated in vacuo. The
residue was subjected to flash column chromatography for prelimi-
nary purification. Further purification of the resulting crude prod-
Experimental methods for SM258-Tag-III dimer
SM258-Tag-III dimer: The product was synthesized by utilizing
cyclo-Leu-N-Me-Lys-d-3-(4-thiazolyl)-Ala-d-Phe-3,3-diphenyl-d-Ala
(10 mg, 0.013 mmol, 1.0 equiv), BS(PEG)9 (4.54 mg, 0.0064 mmol,
0.5 equiv), and DIPEA (17.9 mL, 8.0 equiv) dissolved in CH2Cl2
(128 mL). The crude material was purified by using HPLC to yield
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
9
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
uct was performed by RP-HPLC to yield cyclo d-Phe-b-benzoxy-
Phe-Lys(Boc)-N-Me-Val-d-Leu (8 mg, 8.5%).
The Boc protecting group of the macrocycle was removed by uti-
lizing a mixture of TFA/CH2Cl2 (1:4, 0.1m) and anisole (2.0 equiv) to
generate free amine groups in the lysine residue. The free amine
compound was used in the subsequent biotinylation reaction with-
out purification. LC-MS (ESI): m/z calcd for C45H62N7O7: 812.47 [M+
1]; found: 812.00.
The Boc protecting group of the macrocycle was removed by fol-
lowing the Boc-removal procedure, in which a mixture of TFA/
CH2Cl2 (1:4, 0.1m) and anisole (2.0 equiv) were used to generate
free amine groups in the lysine residue. The free amine compound
was used in the subsequent dimerization without purification. LC-
MS (ESI): m/z calcd for C43H53N7O5S: 755.45 [M+1]; found: 755.15.
SM122-Tag-III dimer: The product was synthesized by utilizing
cyclo Phe-d-Phe-N-Me-Lys-Leu-Lys(Cbz) (110 mg, 0.135 mmol,
1.0 equiv), BS(PEG)9 (95.4 mL, 67.3 mmol, 0.5 equiv), and DIPEA
(0.31 mL, 8.0 equiv) dissolved in dichloromethane (2.7 mL). The
crude material was purified by using column chromatography on
silica gel with ethyl acetate/methanol as the eluent to yield the
product (127 mg, 90%). 1H NMR (300 MHz, CDCl3): d=8.16–8.06
(m, 2H; NH), 7.64–7.55 (d, J=8.71 Hz, 2H; NH), 7.37–7.13 (m, 30H),
7.12–7.03 (d, J=10.16 Hz, 2H; NH), 6.94 (m, 2H; NH), 6.76–6.67 (m,
2H; NH), 5.36–5.26 (m, 2H), 5.09 (s, 4H), 5.09 (s, 4H), 4.80–4.68 (m,
2H), 4.74–4.64 (m, 2H), 4.23–4.13 (m, 2H), 3.75 (t, J=12.1, 6.7 Hz,
4H), 3.66 (m, 28H), 3.67–3.57 (m, 2H), 3.28–3.18 (m, 4H), 3.17–2.99
(m, 4H), 3.28–3.17 (m, 2H), 2.96–2.87 (m, 2H), 3.08–2.85 (m, 4H),
2.78 (s, 6H), 2.55–2.45 (t, J=12.1, 6.7 Hz, 4H), 1.82–1.72 (m, 4H),
1.88–1.65 (m, 8H), 1.71–1.61 (m, 2H), 1.82–1.57 (m, 4H), 1.58–1.36
(m, 4H), 1.33-.123 (m, 4H), 0.99–0.93 (d, J=6.4 Hz, 6H), 0.93–0.86
(m, 4H), 0.91–0.85 ppm (d, J=6.4 Hz, 6H); 13C NMR (75 MHz,
CDCl3): d=172.3, 172.1, 172.0, 171.6, 169.6, 156.6, 136.6, 136.5,
136.1, 129.3, 128.8, 128.7, 128.6, 128.5, 128.1, 127.1, 126.9, 70.4,
70.3, 70.2, 67.4, 66.6, 56.6, 55.6, 54.2, 49.5, 40.7, 40.3, 39.1, 38.0,
37.0, 32.6, 31.7, 29.8, 29.1, 28.6, 25.2, 23.2, 22.9, 22.7, 21.3 ppm; LC-
MS (ESI): m/z calcd for C112H162N14O25: 2103.18 [M+2]; found:
1052.00 (half mass); HRMS (ESI-TOF): m/z calcd for
SM145-Tag-II dimer: The product was synthesized by utilizing
cyclo d-Phe-b-benzoxy-Phe-Lys-N-Me-Val-d-Leu (8 mg, 0.011 mmol,
1.0 equiv), BS(PEG)9 (3.75 mg, 0.0053 mmol, 0.5 equiv), and DIPEA
(14.8 mL, 8.0 equiv) dissolved in CH2Cl2 (106 mL). The crude material
1
was purified by HPLC to yield the product (1.5 mg, 6.9%). HNMR
(600 MHz, DMSO): d=7.58 (m; 2NH), 7.50 (m; 2NH), 7.33–7.11 (m,
36H; 2NH), 4.80 (m, 2H; 2NH), 4.62 (m; NH), 4.42 (m, 3H; NH), 4.14
(m, 1H), 3.98 (m, 1H), 3.65 (m, 34H), 3.26–3.02 (m, 10H), 2.96 (m,
2H), 2.80 (m, 4H), 2.46 (m, 4H), 2.35 (m, 2H), 2.20 (m, 2H), 1.85 (m,
12H), 1.57 (m, 4H), 1.42 (m, 4H), 1.30 (m, 2H), 1.03–0.86 ppm (m,
24H); 13CNMR (150 MHz, DMSO): d=174.4, 173.5, 171.9, 171.6,
169.7, 168.3, 138.5, 137.5, 136.8, 130.1, 129.9, 129.3, 129.1, 128.7,
128.4, 128.1, 127.6, 127.3, 79.9, 79.5, 70.5, 70.0, 67.2, 63.5, 61.6,
59.0, 57.4, 48.2, 42.2, 41.8, 41.2, 40.2, 36.9, 30.1, 29.2, 28.9, 26.3,
24.9, 22.8, 21.8, 19.6, 18.8, 17.9 ppm; LC-MS (ESI): m/z calcd for
C108H154N12O23: 1988.12 [M+1]; found: 995.00 (half mass); HRMS
(ESI-TOF): m/z calcd for C108H154N12O23Na: 2011.1100 [M+Na+];
found: 2011.1174.
C
112H161N14O25Na: 2125.17 [M+Na+]; found: 2125.13.
Experimental methods for SM122-Tag-III
HO-Phe-d-Phe-N-Me-Lys(Boc)-Leu-Lys(Cbz)-NH2: Resin-O-Phe-d-
Phe-N-Me-Lys(Boc)-Leu-Lys(Cbz)-NH2 was synthesized by utilizing
resin-O-Phe-NH2 (1.5 g, 0.96 mmol, 1.0 equiv) and the subsequent
peptide coupling in the sequence using an aliquot (2.88 mmol,
3.0 equiv) of each of the following: Fmoc-N-Me-d-Phe-OH
(1.156 g), Fmoc-Lys(Boc)-OH (1.35 g), Fmoc-Leu-OH (1.02 g), and
Fmoc-Lys(Cbz)-OH (1.447 g). Each peptide coupling was carried out
in the presence of HOAt (0.392 g, 2.88 mmol, 3.0 equiv) or HOBt
(0.389 g, 2.88 mmol, 3.0 equiv), DIC (0.90 mL, 3.0 mmol, 6.0 equiv),
and DMF to generate a concentration of 0.20m based on the
amino acid. Each coupling reaction was performed for 3 h, and
a negative ninhydrin test was used to confirm the reaction comple-
tion. The Fmoc protecting group was removed after the comple-
tion of each coupling reaction. HO-Phe-d-Phe-N-Me-Lys(Boc)-Leu-
Lys(Cbz)-NH2 was obtained from the resin cleavage by using a solu-
tion of TFE (10 mL) and CH2Cl2 (10 mL). The resin-containing solu-
tion was filtered and dried in vacuo to yield the DDLP SM122-Tag-
III as a white solid (708 mg, 79%). LC-MS (ESI): m/z calcd for
C50H72N7O10: 930.53 [M+1]; found: 930.00.
Biological methods
Hsp90 pulldown assays: Hsp90 pull-down was completed on Neu-
trAvidin agrose resin (Thermo Scientific Pierce) with biotin-tagged
compounds (400 mm) incubated with crude cell lysate (2.5 mg)
from human colorectal carcinoma cells (HCT 116) for 24 h at 48C.
The unbound supernatant was removed. The beads were washed
four times with wash buffer (20 mm tris(hydroxymethyl)aminome-
thane hydrochloride (Tris–HCl), 100 mm NaCl, 1% Triton X-100,
pH 7.4) and subsequently boiled for 5 min at 958C in 5ꢃLaemmli
sample buffer (62.5 mm Tris–HCl, 2% (w/v) sodium dodecyl sulfate
(SDS), 10% (v/v) glycerol, 0.1% (w/v) bromophenol blue, 100 mm
dithiothreitol). The samples were analyzed by means of 4–20%
SDS polyacrylamide gel electrophoresis (SDS-PAGE), followed by
using the standard Western blot protocol to detect Hsp90. The
amount of Hsp90 was analyzed by Image J and transformed to
give a percentage of Hsp90 bound to each tagged compound,
where 100% was set to the most effective tagged compound.
Each experiment was completed with n=3, and the data shown
are an average of these three experiments.
cyclo-Phe-d-Phe-N-Me-Lys(Boc)-Leu-Lys(Cbz): The product was
synthesized by using DDLP SM122-Tag-III (0.203 g, 0.22 mmol,
1.0 equiv), TBTU (0.056 g, 0.18 mmol, 0.8 equiv), HATU (0.067 g,
0.18 mmol, 0.8 equiv), DMTMM (0.048 g, 0.18 mmol, 0.8 equiv), and
of DIPEA (0.23 mL, 6.0 equiv) in anhydrous CH2Cl2 (220 mL, 0.001m)
and following the macrocyclization procedure. The reaction was
stirred overnight, and the reaction was monitored by TLC and LC-
MS. Upon completion, the reaction mixture was subjected to an
acid/base wash to afford the crude product, which was purified by
flash column chromatography on silica gel with an ethyl acetate/
hexane gradient system as the eluent, followed by purification by
HPLC to yield the Boc-protected macrocycle as a white solid
(110 mg, 55%). LC-MS (ESI): m/z calcd for C50H70N7O9: 912.52 [M+
1]; found: 912.00.
Protein binding assay: The pure protein binding assay to measure
the binding affinity between Hsp90 and HOP/FKBP52 was complet-
ed by using human native protein Hsp90 (final concentration=
200 nm) and human-recombinant his-tagged HOP/FKBP52 (final
concentration=100 nm). The experiments were carried out with
concentrations of 30 mm for monomeric molecules and 15 mm for
dimeric molecules. The protein pulldown was completed on
TALON metal affinity resin (Clontech; cat. no. 635501), followed by
three washes of the beads in binding buffer and finally boiling the
beads with 5ꢃLaemmli sample buffer. The samples were analyzed
by means of 4–20% SDS-PAGE, followed by using the standard
Western blot protocol to detect Hsp90 and HOP/FKBP52. The re-
&
&
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
10
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
[8] a) Y. Wang, S. R. Mcalpine, Chem. Commun. 2015, 51, 1410–1413;
b) Y. C. Koay, J. R. Mcconnell, Y. Wang, S. R. Mcalpine, RSC Adv. 2015, 5,
59003–59013; c) J. M. Mcconnell, L. D. Alexander, S. R. Mcalpine, Bioorg.
Med. Chem. Lett. 2014, 24, 661–666; d) Y. C. Koay, J. R. Mcconnell, Y.
Wang, S. J. Kim, S. R. Mcalpine, ACS Med. Chem. Lett. 2014, 5, 771–776;
e) V. C. Ardi, L. D. Alexander, V. A. Johnson, S. R. Mcalpine, ACS Chem.
Biol. 2011, 6, 1357–1367; f) R. C. Vasko, R. A. Rodriguez, C. N. Cunning-
ham, V. C. Ardi, D. A. Agard, S. R. Mcalpine, ACS Med. Chem. Lett. 2010,
1, 4–8.
[9] a) Y. C. Koay, N. L. Richardson, S. S. Zaiter, J. Kho, S. Y. Nguyen, D. H. Tran,
K. W. Lee, L. K. Buckton, S. R. Mcalpine, ChemMedChem 2016, 11, 881–
892; b) C. L. Carroll, J. V. C. Johnston, A. Kekec, J. D. Brown, E. Parry, J.
Cajica, I. Medina, K. M. Cook, R. Corral, P.-S. Pan, S. R. Mcalpine, Org. Lett.
2005, 7, 3481–3484; c) T. J. Styers, R. A. Rodriguez, P.-S. Pan, S. R. Mcal-
pine, Tetrahedron Lett. 2006, 47, 515–517; d) R. P. Sellers, L. D. Alexand-
er, V. A. Johnson, C.-C. Lin, J. Savage, R. Corral, J. Moss, T. S. Slugocki,
E. K. Singh, M. R. Davis, S. Ravula, J. E. Spicer, J. L. Oelrich, A. Thornquist,
C.-M. Pan, S. R. Mcalpine, Bioorg. Med. Chem. 2010, 18, 6822–6856;
e) J. B. Kunicki, M. N. Petersen, L. D. Alexander, V. C. Ardi, J. R. Mcconnell,
S. R. Mcalpine, Bioorg. Med. Chem. Lett. 2011, 21, 4716–4719; f) M. R.
Davis, E. K. Singh, H. Wahyudi, L. D. Alexander, J. Kunicki, L. A. Nazarova,
K. A. Fairweather, A. M. Giltrap, K. A. Jolliffe, S. R. Mcalpine, Tetrahedron
2012, 68, 1029–1051; g) L. D. Alexander, J. R. Partridge, D. A. Agard,
S. R. Mcalpine, Bioorg. Med. Chem. Lett. 2011, 21, 7068–7071; h) D.
Ramsey, M. J. Mcconnell, R. L. Alexander, D. K. Tanaka, W. C. Vera, M. S.
Mcalpine, Bioorg. Med. Chem. Lett. 2012, 22, 3287–3290.
spective ratio of Hsp90 to its cochaperones were analyzed by
Image J and transformed to give a percentage of Hsp90 bound to
cochaperone or client proteins. Each experiment was completed
with n=3.
Cytotoxicity assay: HCT116 cells (ECACC) were maintained in Dul-
becco’s modified eagle medium (DMEM) supplemented with 10%
fetal bovine serum (FBS) and 1% penicillin/streptomycin (Invitro-
gen/Life Technologies). The cells were propagated in a humidified
chamber at 378C with 5% CO2 seeded into 96-well plates at
2000 cells per well and allowed to adhere overnight. The cells were
treated with at a concentration of 30 mm, with a constant dimethyl
sulfoxide (DMSO) concentration of below 1% for 72 h. After
72 hours, the media was removed and replaced with DMEM
(100 mL) with the cell-counting kit 8 reagent (Dojindo; 10 mL). The
cells were left in the incubator at 378C with 5% CO2 for 2–4 h. The
absorbance was read according to the manufacturer’s protocol by
using a chromate plate reader.
Acknowledgements
We thank UNSW for support of S.R.M. and Y.-C.K. We also thank
the Australian government for providing the Endeavour schol-
arship fund to Y.-C.K. Finally, we thank the NHMRC for funding
(grant GNT1043561).
[10] F. F. Zheng, S. D. Kuduk, G. Chiosis, P. N. Mꢄnster, L. Sepp-Lorenzino, S. J.
Danishefsky, N. Rosen, Cancer Res. 2000, 60, 2090–2094.
[11] X. Yin, H. Zhang, F. Burrows, L. Zhang, C. G. Shores, Clin. Cancer Res.
2005, 11, 3889–3896.
[12] B. R. Kusuma, L. B. Peterson, H. Zhao, G. Vielhauer, J. Holzberlein, B. S.
Blagg, J. Med. Chem. 2011, 54, 6234–6253.
[13] D. R. Southworth, D. A. Agard, Mol. Cell 2011, 42, 771–781.
[14] H. Wahyudi, Y. Wang, S. R. Mcalpine, Org. Biomol. Chem. 2014, 12, 765–
773.
[15] a) A. Islam, Y. Zhang, Y. Wang, S. R. Mcalpine, Med. Chem. Commun.
2015, 6, 300–305; b) S. J. Kim, C.-C. Lin, C.-M. Pan, D. P. Rananaware,
D. M. Ramsey, S. R. Mcalpine, Med. Chem. Commun. 2013, 4, 406–410.
[16] J. Li, J. Soroka, J. Buchner, Biochim Biophys Acta 2012, 1823, 624–625.
[17] a) F. Pirkl, J. Buchner, J. Mol. Biol. 2001, 308, 795–806; b) P. K. K. Tai, H.
Chang, M. W. Albers, S. L. Schreiber, D. O. Toft, L. E. Faber, Biochemistry
1993, 32, 8842–8847; c) J. L. Johnson, D. O. Toft, J. Biol. Chem. 1994,
269, 24989–24993; d) M. B. Cox, D. L. Riggs, M. Hessling, F. Schumacher,
J. Buchner, D. F. Smith, Mol. Endocrinol. 2007, 21, 2956–2967; e) D.
Riggs, P. Roberts, S. Chirillo, J. Cheung-Flynn, V. Prapapanich, T. Ratajc-
zak, R. Gaber, D. Picard, D. Smith, EMBO J. 2003, 22, 1158–1167.
Keywords: antitumor agents
inhibitors · proteins
·
cancer
·
dimerization
·
[1] a) D. Hanahan, R. A. Weinberg, Cell 2011, 144, 646–674; b) M. Taipale, I.
Krykbaeva, M. Koeva, C. Kayatekin, K. D. Westover, G. I. Karras, S. Lind-
quist, Cell 2012, 150, 987–1001.
[2] G. Chiosis, L. Neckers, ACS Chem. Biol. 2006, 1, 279–284.
[3] a) S. D. Westerheide, R. I. Morimoto, J. Biol. Chem. 2005, 280, 33097–
33100; b) F. Ritossa, Cell Stress Chaperones 1996, 1, 97.
[4] a) M. Maruya, M. Sameshima, T. Nemoto, I. Yahara, J. Mol. Biol. 1999,
285, 903–907; b) T. Nemoto, Y. Ohara-Nemoto, M. Ota, T. Takagi, K. Yo-
koyama, Eur. J. Biochem. 1995, 233, 1–8.
[5] L. Whitesell, S. L. Lindquist; N at. Rev. Cancer 2005, 5, 761–772.
[6] a) B. R. Z. Kusuma, L. Sundstrom, T. Peterson, L. B. Dobrowsky, R. T.
Blagg, J. Med. Chem. 2012, 55, 5797–5812; b) J. D. Eskew, T. Sadikot, P.
Morales, A. Duren, I. Dunwiddie, M. Swink, X. Zhang, S. Hembruff, A.
Donnelly, R. A. Rajewski, B. S. J. Blagg, J. R. Manjarrez, R. L. Matts, J. M.
Holzbeierlein, G. A. Vielhauer, Bio. Med. Central Cancer. 2011, 11, 468.
[7] L. K. Buckton, H. Wahyudi, S. R. Mcalpine, Chem. Commun. 2016, 52,
501–504.
Received: July 21, 2016
Published online on && &&, 0000
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
11
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
FULL PAPER
&
Antitumor Agents
Y. C. Koay, H. Wahyudi, S. R. McAlpine*
&& – &&
Reinventing Hsp90 Inhibitors:
Blocking C-Terminal Binding Events to
Hsp90 by Using Dimerized Inhibitors
Two are better than one: Heat-shock
protein 90 (Hsp90) functions as a dimer
and performs molecular-chaperone
duties, thus stabilizing more than 400
proteins that are responsible for cancer
development and progression. Mole-
cules that modulate the C-terminus of
Hsp90 effectively induce cancer cell
death. Herein, we describe the design,
synthesis, and binding affinity of dimer-
ized C-terminal Hsp90 modulators,
which are synergistically better at bind-
ing to Hsp90 than monomers (see pic-
ture; SM = Hsp90 C-terminal modulator,
TPR=tetratricopeptide repeat.).
&
&
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
12
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!