A small molecule inhibits protein disulfide isomerase and triggers the chemosensitization of cancer cells

Article abstract of DOI:10.1002/anie.201406577

Resistance to chemotherapeutic agents represents a major challenge in cancer research. One approach to this problem is combination therapy, the application of a toxic chemotherapeutic drug together with a sensitizing compound that addresses the vulnerability of cancer cells to induce apoptosis. Here we report the discovery of a new compound class (T8) that sensitizes various cancer cells towards etoposide treatment at subtoxic concentrations. Proteomic analysis revealed protein disulfide isomerase (PDI) as the target of the T8 class. In-depth chemical and biological studies such as the synthesis of optimized compounds, molecular docking analyses, cellular imaging, and apoptosis assays confirmed the unique mode of action through reversible PDI inhibition. In the battle against the chemoresistance of cancer cells a screening approach has identified a novel compound class that sensitizes cancer cells in combination with etoposide. Proteomic target discovery revealed the reversible inhibition of protein disulfide isomerase as the molecular mechanism, which was further supported by cellular imaging studies and docking and biochemical assays in various cancer model systems.

Full text of DOI:10.1002/anie.201406577

Angewandte
Chemie
DOI: 10.1002/anie.201406577
Cancer Sensitization
Hot Paper
A Small Molecule Inhibits Protein Disulfide Isomerase and Triggers the
Chemosensitization of Cancer Cells**
Jꢀrgen Eirich, Simone Braig, Liliana Schyschka, Phil Servatius, Judith Hoffmann,
Sabrina Hecht, Simone Fulda, Stefan Zahler, Iris Antes, Uli Kazmaier, Stephan A. Sieber,* and
Angelika M. Vollmar*
Abstract: Resistance to chemotherapeutic agents represents
a major challenge in cancer research. One approach to this
problem is combination therapy, the application of a toxic
chemotherapeutic drug together with a sensitizing compound
that addresses the vulnerability of cancer cells to induce
apoptosis. Here we report the discovery of a new compound
class (T8) that sensitizes various cancer cells towards etoposide
treatment at subtoxic concentrations. Proteomic analysis
revealed protein disulfide isomerase (PDI) as the target of
the T8 class. In-depth chemical and biological studies such as
the synthesis of optimized compounds, molecular docking
analyses, cellular imaging, and apoptosis assays confirmed the
unique mode of action through reversible PDI inhibition.
significant homology. PDI catalyzes thiol–disulfide exchange
reactions of both intra- and intermolecular disulfides.[4] The
role of this enzyme in diseases reaches far beyond cancer and
was recently reviewed in detail.[5] Although several inhibitors
were published in recent years[6–12] the most specific com-
pounds—16F16, RB-11-ca, PACMA 31, and P1—exhibit
a pharmacologically less desired irreversible mode of action.
Here we introduce reversible and highly specific PDI
inhibitors that sensitize tumor cells towards classical chemo-
therapeutic agents.
The screening of a commercial compound library for the
chemosensitization of etoposide-induced apoptosis in various
cancer cell lines revealed T8 as a promising candidate. The
combination of subtoxic concentrations of etoposide (500 nm)
and T8 dose-dependently led to pronounced apoptosis rates
with a minimal concentration of 25 mm T8 in a leukemic
(Jurkat) as well as in a breast cancer cell line (MDA-MB-231)
(Figure 1A). In addition, the long-term survival of Jurkat cells
was synergistically inhibited after treatment with a combina-
tion of etoposide and T8 (Figure 1B). Growth of various
carcinoma cell lines such as LNCAP (prostate cancer),
PancTu1, and L3.6pl (pancreas cancer) was also strongly
affected by the combined treatment with other chemother-
apeutic drugs such as doxorubicin or TRAIL and T8
(Figures S1 and S2). In contrast, noncancerous human
endothelial cells (HUVEC) did not respond to T8 in
combination with etoposide or doxorubicin (Figure S3).
T
he resistance of tumor cells to drugs results from numerous
genetic and epigenetic changes.[1] Cancer cells by nature
require increased protein synthesis and thus respond to
endoplasmic reticulum (ER) stress by activating the unfolded
protein response (UPR) which is mediated by ER chaperones
such as protein disulfide isomerase (PDI).[1c,2]
As ER chaperones maintain ER homeostasis and support
cancer cell survival, interest has emerged in targeting these
proteins to fight chemoresistance. In this respect PDI has
received increasing attention and the crystal structures of the
human full-length protein have recently been published.[3]
The isomerase is organized in four distinct domains (a, a’ and
b, b’). The a and a’ domains are catalytically active and share
[*] Dr. J. Eirich,[$] [+] Prof. Dr. I. Antes, Prof. Dr. S. A. Sieber
Center for Integrated Protein Science Munich CIPSM
Department of Chemistry, Institute of Advanced Studies IAS
Technische Universitꢀt Mꢁnchen
S. Hecht, Prof. Dr. I. Antes
Department of Life Sciences, Technische Universitꢀt Mꢁnchen
Emil-Erlenmeyer-Forum 8, 85354 Freising-Weihenstephan
(Germany)
[$] Current address: Department of Oncology/Pathology
Lichtenbergstrasse 4, 85747 Garching (Germany)
E-mail: stephan.sieber@tum.de
Dr. S. Braig,[+] Dr. L. Schyschka, Prof. Dr. S. Zahler,
Prof. Dr. A. M. Vollmar
Cancer Proteomics Mass Spectrometry
SciLifeLab Stockholm, Karolinska Institutet
Tomtebodavꢀgen 23, 17165 Solna (Sweden)
[+] These authors contributed equally to this work.
Department of Pharmacy, Center for Drug Research
Pharmaceutical Biology, Ludwig-Maximilians-Universitꢀt Mꢁnchen
Butenandtstrasse 5–13, 81377 Mꢁnchen (Germany)
E-mail: angelika.vollmar@cup.uni-muenchen.de
[**] We thank Mona Wolff, Katja Bꢀuml, and Burghard Cordes for
excellent scientific support, Dr. Stuppner and Dr. Langner (Univer-
sity of Innsbruck) for collaboration. S.A.S. was supported by the
Deutsche Forschungsgemeinschaft (SFB749, SFB1035, and
FOR1406), an ERC starting grant, and the Center for Integrated
Protein Science Munich CIPSM, J.E. thanks the TUM Graduate
School and A.M.V. thanks the DFG and the Wilhelm Sander
Foundation.
P. Servatius, J. Hoffmann, Prof. Dr. U. Kazmaier
Institute for Organic Chemistry, Saarland University
Im Stadtwald, Geb. C4.2, 66123 Saarbrꢁcken (Germany)
Prof. Dr. S. Fulda
Institute for Experimental Cancer Research in Pediatrics
University Hospital Frankfurt
Supporting information for this article (synthesis and character-
ization of compounds, bioassays, cell biology as well as proteome
preparation, labeling, and mass spectrometry) is available on the
WWW under http://dx.doi.org/10.1002/anie.201406577.
Komturstrasse 3a, 60528 Frankfurt a. M. (Germany)
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
Figure 1. T8 sensitizes cancer cell lines towards etoposide-induced cell death. A) Apoptosis assay by FACS analysis; chemical structure of T8.
B) Long-term growth analysis of Jurkat cells after treatment. C) Activation of caspases (western blot analysis). D) Effect of caspase inhibition by
zVADfmk. E) Effect on CAM-based tumor growth (L3.6pl pancreatic carcinoma cells). ***p<0.001; *p<0.05; # synergistic.
Corresponding western blot analysis illustrated increased
hallmarks of apoptosis such as PARP cleavage and caspase-9
as well as caspase-3 activity (Figure 1C). Moreover, the
sensitizing effect was dependent on caspase activation as
pretreatment of cells with the pan-caspase inhibitor
zVADfmk as well as with the specific caspase-9 inhibitor
abrogated the effect (Figure 1D and Figure S4). T8 synerg-
istically suppressed the growth of pancreatic carcinoma cells
(L3.6pl cells) seeded on the chorioallantoic membrane
(CAM) of chicken embryos and treated with etoposide
(Figure 1E).
To identify the cellular targets of T8, we applied activity-
based protein profiling (ABPP)[13] and equipped the molec-
ular scaffold with an alkyne handle as well as a photoreactive
group (JP04-042, Figure 2A). Surprisingly, these modifica-
tions even increased the chemosensitizing potency of JP04-
042 (Figure 2A and Figure S4C).
targets on SDS-PAGE and fluorescent scanning (Figure 2C).
Different concentrations of JP04-042 (Figure 2D) and irra-
diation times (Figure 2E, right panel) were tested. Of note,
one dominant protein band appeared at about 63 kDa
emphasizing that the probe almost exclusively addressed
a single cellular target. To investigate whether the photo
probe and the parent T8 molecule bind to the same target,
competitive labeling with a constant concentration of JP04-
042 versus varying concentrations of competitor T8 was
conducted: a 1.5-fold excess of T8 (30 mm) over JP04-042
(20 mm) was already sufficient to decrease labeling intensity
(Figure 2E, left panel).
Next, the target was identified by a SILAC approach with
“heavy” and “medium-weight” isotope-labeled MDA-MB-
231 cells. After UV cross-linking and cell lysis either a biotin-
PEG-azide or a trifunctional linker (TFL)[15] was attached to
the probe by CC. Labeled proteins were enriched by a biotin–
avidin pull-down. Proteins were either analyzed by SDS-
PAGE or prepared for mass spectrometry (MS) by tryptic on-
bead digest directly. All independent experiments revealed
that PDI and some of its isoforms could be highly enriched
Next, MDA-MB-231 and HeLa cells were incubated with
JP04-042 followed by UV cross-linking to its cellular targets
in situ. After cell lysis a rhodamine reporter dye azide was
introduced by click chemistry (CC)[14] to visualize potential
2
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
These are not the final page numbers!

Angewandte
Chemie
Figure 2. Protein labeling, target
identification, and validation.
A) Chemical structure of photo
probe JP04-042. Apoptotic effect
of JP04-042 alone and in combi-
nation. # indicates synergy.
B) Labeling of recombinant PDI
with JP04-042 and T8. FM: Fluo-
rescent marker. C) Workflow for
target identification. D) Concen-
tration-dependent labeling of
MDA-MB-231 cells with JP04-042.
E) Left: Competitive labeling:
20 mm probe, increasing T8 con-
centration; right: time-dependent
labeling (100 mm probe). HeLa
cells. F) PDI staining by antibody
(green; left) or JP04-042 (red;
middle). Merged picture (right)
indicates colocalization of PDI
with JP04-042.
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
Table 1: Selected hits from quantitative full proteome pull-down and gel-based analysis in isotope-labeled MDA-MB-231 cells.[a]
Uniprot ID
Description
Fold enrichment probe/DMSO
gel-free, replicate
MW [kDa]
gel-based
1
2
3
P07237
B3KQT9
protein disulfide isomerase
26
18
56
13
41
11
47
27
57.1
54.1
cDNA PSEC0175 fis, clone OVARC1000169,
highly similar to protein disulfide isomerase A3
protein disulfide isomerase A4
P13667
31
3
72.9
[a] For full lists of proteins see Tables S1–S4 in the Supporting Information.
(see Table 1 and Tables S1–S4 in the Supporting Informa-
tion).
localization of the probe and PDI, cells were incubated with
CC reagents and PDI antibodies, respectively. Hoechst
staining was used to mark the nuclei of the cells and
background rhodamine binding was determined by the sole
addition of CC reagents (Figure S5). Importantly, the PDI-
directed antibody as well as the click-dye-conjugated probe
overlap in their fluorescent staining, suggesting a consolidated
binding to the same target protein (Figure 2F).
In order to explore the structure–activity relationship
(SAR) of the T8-derived N-(1-(2-(R2amino)-2-oxoethyl)-
piperidin-4-yl)R1amide core scaffold, we prepared several
new analogues that exhibited diversity in the substituents of
the two R1 and R2 benzene rings (Figure 3A). Among those,
PS89, a close analogue of T8 in which the fluorine in R1 is
For target validation, labeling as well as inhibition (see
below) of recombinant PDI protein by JP04-042 was con-
firmed. Recombinant protein was incubated with the probe
and the fluorescent signal vanished when the protein had
been pretreated with T8 or thermally denatured prior to
labeling; this suggests that the probe specifically interacts
with the folded and active enzyme (Figure 2B).
To verify PDI as a potential target of JP04-042 in intact
cells, JP04-042-labeled cells were co-stained with a PDI-
specific antibody. MDA-MB-231 cells were incubated with
JP04-042, irradiated to covalently attach the probe to the
target protein, and finally fixed. To visualize the cellular
Figure 3. T8 derivatives, inhibition of recombinant PDI, and apoptosis assays. A) Chemical structures of further T8 derivatives (PS83–PS89) and
corresponding IC50 values for in vitro PDI inhibition. B) Apoptosis analysis of T8 derivatives PS83–PS89 (Jurkat cells) +/- etoposide. C) PS89
concentration-dependent induction of apoptosis +/- etoposide. # indicates synergy.
4
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
These are not the final page numbers!

Angewandte
Chemie
replaced by an azide group, displayed the best
sensitizing activity of all derivatives in the apop-
tosis assay with Jurkat cells (Figures 3B and C).
Moreover, as described for PDI inhibition the
compound is toxic at concentrations above 50 mm.
In contrast, PS88 showed no synergism with
etoposide suggesting that a 3,4-dimethyl substitu-
tion in the R2 ring is less desired. Similar trends
could be observed with other cell lines (Fig-
ures S4D and E). To show that PS89 induces ER
stress in combination with etoposide, the up-
regulation of main proteins involved in ER stress
signaling, namely p-eIF2a, BiP/GRP78, and
CHOP, was analyzed by western blot (Figure S6).
To investigate whether the interaction of T8
and its derivatives with PDI results in the
inhibition of enzymatic activity we analyzed all
compounds in a turbidimetric insulin assay[16]
(Table in Figure 3A). T8 turned out to be only
a weak inhibitor in the reductase assay, whereas
the probe molecule JP04-42 and PS89 revealed
concentration-dependent inhibition with IC50
values down to 15 mm.
Figure 4. Docking experiments. Binding sites at the two catalytic centers in the
An explanation for the differences in the IC50
values was provided by docking studies with
JP04-42, PS89, and T8 for the two catalytic sites
a and a’. Remarkably, we obtained two preferred
binding sites in close proximity (Æ 40 ꢀ) to the
catalytic centers (Figure S8). A closer inspection
of the two sites revealed that they are extended
solvent-exposed grooves, which are located on the
domains a (A) and a’ (B). The two inserted pictures show the back (top insert) and
front binding grooves (bottom insert). Deep pockets are indicated by green circles.
C,D) Bound structures of JP04-042 (green), PS89 (yellow), and T8 (green) in the two
catalytic binding sites, C) domain a, D) domain a’. For the protein its electrostatic
potential surface is shown, the ligands are given in green (JP04-042 and T8) and
yellow (PS89) rod representations colored according to atom type. The graphics
were created with VMD.
front and the back side of the catalytic cysteines (Figures 4A
and B and Figure S9). However, there are distinct differences
in the binding specificities of the JP04-42 and PS89 com-
pounds with respect to T8. Analysis of the ten energetically
best structural clusters (Figure S9) shows that for JP04-42 and
PS89 several poses shield the catalytic sites, which is
a prerequisite of strong inhibition. T8, in contrast, binds
deeply into the two binding grooves adjacent to each catalytic
site (Figure 4C and D). JP04-42 and PS89 carry an azide
substituent, which perfectly fits into electrostatically comple-
mentary deep subpockets within the binding grooves in front
of the a and a’ sites (Figures 4C and D and Figure S10). In
contrast, the diethylphenyl group of T8 preferentially
attaches to predominantly hydrophobic protein interaction
sites within the binding groove. Thus, T8 can be considered an
allosteric inhibitor that impairs substrate binding without
obstruction of the catalytic cysteines. Of note, although T8
exhibits a different binding mode it partially overlaps in the
pocket with JP04-42 and PS89 explaining the observed
competition between the inhibitors (Figure 4C). Moreover,
this T8-specific binding mode may account for the obtained
IC50 differences in the insulin reduction assay (a more
detailed discussion is provided in the Supporting Informa-
tion).
pounds for further pharmaceutical testing. Most of the
previously reported PDI inhibitors are irreversible binders
or lack selectivity, both of which are undesired for medicinal
applications. The great specificity and reversible mode of
action of our T8 derivatives and their performance in tumor-
based assays emphasize a suitable pharmacological profile,
which is further substantiated by satisfaction of the Lipinski
rules. Furthermore, in-depth analysis of the PDI binding
mode and work on apoptotic signaling provides a basis for
further exploration of this target in cancer research.
Received: June 30, 2014
Published online: && &&, &&&&
Keywords: cancer · chemotherapeutics · combination therapy ·
.
proteomics · sensitization
[1] a) C. Holohan, S. Van Schaeybroeck, D. B. Longley, P. G.
Johnston, Nat. Rev. Cancer 2013, 13, 714 – 726; b) T. R. Wilson,
P. G. Johnston, D. B. Longley, Curr. Cancer Drug Targets 2009, 9,
307 – 319; c) I. Kim, W. Xu, J. C. Reed, Nat. Rev. Drug Discovery
2008, 7, 1013 – 1030.
[2] B. Luo, A. S. Lee, Oncogene 2013, 32, 805 – 818.
[3] C. Wang, W. Li, J. Ren, J. Fang, H. Ke, W. Gong, W. Feng, C. C.
Wang, Antioxid. Redox Signaling 2013, 19, 36 – 45.
In conclusion, the unique PDI target selectivity, the
unprecedented mode of inhibition, and its potent sensitiza-
tion of cancer cells in various experimental settings place T8
and optimized derivatives among the most promising com-
[4] a) T. E. Creighton, BioEssays 1992, 14, 195 – 199; b) P. T. Var-
andani, M. A. Nafz, M. L. Chandler, Biochemistry 1975, 14,
2115 – 2120; c) N. A. Murkofsky, M. E. Lamm, J. Biol. Chem.
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

.
Angewandte
Communications
1979, 254, 12181 – 12184; d) C. C. Wang, C. L. Tsou, FASEB J.
1993, 7, 1515 – 1517.
[13] a) M. J. Evans, A. Saghatelian, E. J. Sorensen, B. F. Cravatt, Nat.
Biotechnol. 2005, 23, 1303 – 1307; b) M. J. Evans, B. F. Cravatt,
´
[5] a) A. M. Benham, Antioxid. Redox Signaling 2012, 16, 781 – 789;
b) S. Xu, S. Sankar, N. Neamati, Drug Discovery Today 2014, 19,
222 – 240.
[6] J. Ge, C. J. Zhang, L. Li, L. M. Chong, X. Wu, P. Hao, S. K. Sze,
S. Q. Yao, ACS Chem. Biol. 2013, 8, 2577 – 2585.
Chem. Rev. 2006, 106, 3279 – 3301; c) M. Fonovic, M. Bogyo,
Expert Rev. Proteomics 2008, 5, 721 – 730; d) B. F. Cravatt, E. J.
Sorensen, Curr. Opin. Chem. Biol. 2000, 4, 663 – 668; e) P. P.
Geurink, L. M. Prely, G. A. van der Marel, R. Bischoff, H. S.
Overkleeft, Top. Curr. Chem. 2012, 324, 85 – 113.
[7] M. M. Khan, S. Simizu, N. S. Lai, M. Kawatani, T. Shimizu, H.
Osada, ACS Chem. Biol. 2011, 6, 245 – 251.
[8] B. G. Hoffstrom, A. Kaplan, R. Letso, R. S. Schmid, G. J. Turmel,
D. C. Lo, B. R. Stockwell, Nat. Chem. Biol. 2010, 6, 900 – 906.
[9] R. Banerjee, N. J. Pace, D. R. Brown, E. Weerapana, J. Am.
Chem. Soc. 2013, 135, 2497 – 2500.
[14] a) V. V. Rostovtsev, L. G. Green, V. V. Fokin, K. B. Sharpless,
Angew. Chem. Int. Ed. 2002, 41, 2596 – 2599; Angew. Chem.
2002, 114, 2708 – 2711; b) Q. Wang, T. R. Chan, R. Hilgraf, V. V.
Fokin, K. B. Sharpless, M. G. Finn, J. Am. Chem. Soc. 2003, 125,
3192 – 3193; c) C. W. Tornøe, C. Christensen, M. Meldal, J. Org.
Chem. 2002, 67, 3057 – 3064.
[10] S. Xu, A. N. Butkevich, R. Yamada, Y. Zhou, B. Debnath, R.
Duncan, E. Zandi, N. A. Petasis, N. Neamati, Proc. Natl. Acad.
Sci. USA 2012, 109, 16348 – 16353.
[15] J. Eirich, J. L. Burkhart, A. Ullrich, G. C. Rudolf, A. Vollmar, S.
Zahler, U. Kazmaier, S. A. Sieber, Mol. Biosyst. 2012, 8, 2067 –
2075.
[11] J. C. Tsibris, L. T. Hunt, G. Ballejo, W. C. Barker, L. J. Toney,
W. N. Spellacy, J. Biol. Chem. 1989, 264, 13967 – 13970.
[12] R. Guthapfel, P. Gueguen, E. Quemeneur, Eur. J. Biochem.
1996, 242, 315 – 319.
[16] A. M. Smith, J. Chan, D. Oksenberg, R. Urfer, D. S. Wexler, A.
Ow, L. Gao, A. McAlorum, S. G. Huang, J. Biomol. Screening
2004, 9, 614 – 620.
6
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
These are not the final page numbers!

Angewandte
Chemie
Communications
Cancer Sensitization
J. Eirich, S. Braig, L. Schyschka,
P. Servatius, J. Hoffmann, S. Hecht,
S. Fulda, S. Zahler, I. Antes, U. Kazmaier,
S. A. Sieber,*
A. M. Vollmar*
&&&& — &&&&
In the battle against the chemoresistance
of cancer cells a screening approach has
identified a novel compound class that
sensitizes cancer cells in combination
with etoposide. Proteomic target discov-
ery revealed the reversible inhibition of
protein disulfide isomerase as the
molecular mechanism, which was further
supported by cellular imaging studies
and docking and biochemical assays in
various cancer model systems.
A Small Molecule Inhibits Protein
Disulfide Isomerase and Triggers the
Chemosensitization of Cancer Cells
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7
These are not the final page numbers!