Identification of malate dehydrogenase 2 as a target protein of the HIF-1 inhibitor LW6 using chemical probes

Article abstract of DOI:10.1002/anie.201304987

Tracking the target: To identify a protein that binds directly to the HIF-1α inhibitor LW6, a series of new chemical probes were synthesized with a clickable tag and/or a photoactivatable moiety. LW6 was found to be localized to the mitochondria (see picture) and MDH2 was identified as a target protein of LW6. These results indicate that the HIF-1α inhibitory activity of LW6 is a consequence of MDH2 suppression. Copyright

Full text of DOI:10.1002/anie.201304987

Angewandte
Chemie
DOI: 10.1002/anie.201304987
Target Identification
Identification of Malate Dehydrogenase 2 as a Target Protein of the
HIF-1 Inhibitor LW6 using Chemical Probes**
Kyeong Lee, Hyun Seung Ban, Ravi Naik, Ye Seul Hong, Seohyun Son, Bo-Kyung Kim,
Yan Xia, Kyung Bin Song, Hong-Sub Lee, and Misun Won*
Hypoxia-inducible factor (HIF) regulates tumor angiogenesis
and metastasis in response to low oxygen tension.[1] In the
presence of oxygen, HIF-1a is rapidly degraded through the
ubiquitin–proteasome pathway. In hypoxic conditions, stabi-
lized HIF-1a dimerizes with HIF-1b. The HIF-1a/b hetero-
dimer binds to hypoxia response elements (HRE) in gene
promoters and induces the expression of target genes
involved in angiogenesis, metastasis, glycolysis, cell prolifer-
ation, and resistance to apoptosis.[2] Increased expression of
HIF-1a in many solid tumors correlates with aggressive tumor
growth, therapeutic resistance, and a poor clinical outcome.
HIF-1a shifts the metabolism from oxidative phosphorylation
to anaerobic glycolysis.[3] Therefore, HIF-1a is an important
therapeutic target for cancer.
We previously synthesized and evaluated aryloxyacetyl-
amino benzoic acid analogues.[4] LW6 (1 in Figure 1A)
potently inhibited HIF-1a accumulation by degrading HIF-
1a without affecting the HIF-1a mRNA levels during
Figure 1. Biological activities and cellular localization of a chemical
probe for LW6. A) Formula of 1 and its clickable probe 2. B) Inhibitory
effects of 1 and probe 2 on HIF-1a accumulation were determined by
hypoxia.[4a,b] LW6, which is commercially available, has been
used in various studies as an HIF-1a inhibitor.[5] However, the
immunoblot analysis. C) Localization of probe 2 (3 mm, green) was
detected through a click reaction using azide-linked Alexa Fluor 488 in
HCT116 cells. Mitochondria were selectively stained with the Mito-
Tracker probe (red). Nuclei (blue) were stained with 4,6-diamidino-2-
phenylindole (DAPI). D) Competitive binding of probe 2 (3 mm) to its
target molecules in the presence or absence of 1 (10 mm). Scale
molecular target of LW6 remains unknown.
To identify a drug target, chemical biological methods
such as activity-based probes (ABPs),[6] photoaffinity label-
ing, biotinylation, and click conjugation have been used.[7]
Herein, we identify the molecular target of 1 using chemical
probes. Cellular images and direct protein interactions of
1 were examined in living cells with a series of chemical
probes (2–6), which were designed using the structure–
activity relationship (SAR) of 1.[4a] Synthesis and character-
bars=20 mm.
ization data for these probes are available in the Supporting
Information.
The distribution of drug molecules within subcellular
compartments can provide information about the mechanism
of drug action. The intracellular localization of LW6 was
visualized through click chemistry with probe 2, containing an
acetylene group, in colon cancer HCT116 cells (Figure 1A).
Both 1 and 2 suppressed HIF-1a accumulation (Figure 1B)
and HRE-luciferase activity (Figure 1A; Supporting Infor-
mation, Figure S12). Subsequently, the cellular localization of
probe 2 was determined by a click reaction with an azide-
linked Alexa Fluor 488 molecule. Notably, copper-catalyzed
azide–alkyne cycloadditions (click reactions) are highly
specific and efficient bio-orthogonal reactions[7b] to visualize
intracellular probe distribution. We found that compound 2
was localized primarily in the cytoplasm (Figure 1C). The co-
localization of compound 2 (3 mm) with the mitochondria-
selective probe, MitoTracker (500 nm), indicated that 2 is
specifically localized in the mitochondria, whereas the local-
ization of an adamantyl-free probe 4 was not observed
(Figure S13). The mitochondrial localization of probe 2 was
[*] Dr. H. S. Ban,[+] Y. S. Hong, Dr. B.-K. Kim, Dr. M. Won
Medical Genomics Research Center, Korea Research Institute of
Bioscience and Biotechnology
125 Gwahak-ro, Yuseong-gu, Daejeon 305-806 (Korea)
E-mail: misun@kribb.re.kr
Prof. K. Lee,[+] R. Naik, S. Son, Dr. Y. Xia
College of Pharmacy, Dongguk University-Seoul
26 Pildong, Junggu, Seoul 100-715 (Korea)
Prof. K. B. Song
Department of Food Science and Technology
Chungnam National University
99 Daehak-ro, Yuseong-gu, Daejeon 305-764 (Korea)
Dr. H.-S. Lee
ILDONG Pharmaceutical Co. Ltd.
23-9 Seogwoo-dong, Hwaseong, Kyungi-do 445-811 (Korea)
[+] These authors contributed equally to this work.
[**] This work was supported in part by the KRIBB Initiative Program,
NRF-2011-0031690, and NRF-2012M3A9C1053532.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201304987.
Angew. Chem. Int. Ed. 2013, 52, 1 – 5
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
attenuated by competition from unlabeled
1 (10 mm) (Figure 1D), confirming that
1 also localizes in the mitochondria.
Mitochondrial respiration consumes
cellular oxygen and controls intracellular
oxygen tension through the electron trans-
port chain (ETC). Pharmacological inhib-
ition of mitochondrial respiration increases
intracellular oxygen concentrations, induc-
ing degradation of HIF-1a.[8] The mito-
chondrial localization of probe 2 suggests
that 1 may inhibit mitochondrial respira-
tion. Accordingly, we examined the effects
of 1 on cellular oxygen consumption with
an XF24 Extracellular Flux Analyzer.
Real-time oxygen consumption rates
(OCR) for the basal and maximal respira-
tions were determined in the presence of
1 (Figure 2A). The basal and maximal
respirations were measured by treatment
with oligomycin, the ATP synthase inhib-
itor, and trifluorocarbonylcyanide phenyl-
Figure 2. Inhibition of mitochondrial respiration by 1. A) OCR was determined at the
indicated concentrations of 1 in HCT116 cells. The basal and maximal OCR were measured
by adding oligomycin (1 mm), FCCP (0.5 mm), and rotenone (1 mm)/antimycin A (1 mm).
B) The total oxygen consumption was determined by calculating the area under the curve
(AUC). Values represent the mean of three samples. Vertical bars indicate standard deviation
(S.D.). Statistical significance: *P<0.05, **P<0.01; compared to untreated control cells.
hydrazone (FCCP), which decouples ATP synthesis, respec-
tively. As a result, OCR and total oxygen consumption were
inhibited by 1 in a concentration-dependent manner during
respiration (Figure 2A,B). Cellular ATP production was
reduced by 1, presumably because respiration was inhibited
(Figure S14). These results suggest that 1 promotes HIF-1a
degradation by reducing the mitochondrial oxygen consump-
tion and increasing the intracellular oxygen during hypoxia.
Subsequently, we added trifluoromethyl diazirine to
probe 2, generating the photoactivatable probe 3 (Fig-
ure 3A). Trifluoromethyl diazirine has several valuable
characteristics, including photoactivity at long wavelengths,
good chemical stability, rapid photolysis, and generation of
a highly reactive carbene.[9] A cell-based HRE reporter assay
and immunoblot analysis demonstrated that probe 3 (10 mm)
suppressed hypoxia-induced HIF-1a accumulation, while the
adamantyl-free probe 4 (10 mm), a negative control, did not
(Figure 3B).
Photoaffinity labeling was performed in the HCT116 cells
using ultraviolet irradiation and click conjugation with
a fluorescent dye (Cy5 for probe 3 and Cy3 for probe 4).
Cellular proteins were separated by sodium dodecyl sulfate-
polyacrylamide gel electrophoresis (SDS-PAGE) and visual-
ized by in-gel fluorescence scanning. A Cy5-specific band was
detected at 30–37 kDa (Figure 3C). Competition between
probes 3 and 1 reduced the Cy5 fluorescence intensity
(Figure 3D).
Figure 3. Target identification with multifunctional chemical probes.
A) Multifunctional chemical probes for photoaffinity labeling and click
conjugation. B) Effects of the probes on transcriptional activity (HRE-
Luc) and accumulation of HIF-1a. C) Proteins conjugated to probe 3–
Cy5 (50 mm) or probe 4–Cy3 (50 mm) were detected by in-gel fluores-
cence after SDS-PAGE. D) Competitive binding of probe 3–Cy5 to the
target protein in the presence of 1 (100 mm). Coomassie brilliant blue
(CBB) staining of the gel is shown.
Two-dimensional gel electrophoresis (2DE) was per-
formed to separate the proteins bound to probes 3 and 4
(Figure 4; Figure S18), followed by in-gel trypsin digestion
and mass spectrometry. We expected that the protein
specifically bound to probe 3–Cy5 would appear red in an
overlaid gel and be localized to the mitochondria. From an
analysis of the red fluorescence, malate dehydrogenase 2
(MDH2) was identified as a protein bound to probe 3–Cy5
(Table S1). MDH2 is a 36 kDa mitochondrial enzyme that
catalyzes the malate/NAD+ conversion to oxaloacetate/
NADH during the tricarboxylic acid cycle (TCA cycle) and
functions as a respiratory substrate in the mitochondrial
ETC.[10]
To confirm the binding of 1 to MDH2, biotin affinity
probes were synthesized (Figure 5A). In the affinity pull-
down assays, MDH2 bound to biotinylated 1 (6), compared
with a negative control the adamantyl-free biotinylated probe
(5; Figure 5B). Binding of biotinylated 1 to MDH2 was
2
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Figure 4. Two-dimensional electrophoresis and imaging. Photoaffinity
labeling of cellular proteins was followed by click conjugation with
a fluorescent dye (Cy5 for probe 3 and Cy3 for probe 4). The cellular
proteins conjugated to probe 3–Cy5 or probe 4–Cy3 were separated by
two-dimensional electrophoresis and visualized by in-gel fluorescence
scanning.
Figure 6. Inhibition of MDH2 activity by 1 and 7. A) In vitro binding of
3 (10 mm) to human recombinant MDH2 and competition assay with
1 (100 mm). B) Effects of 1 and 7 on the activity of isolated MDH2
from HCT116 cells. C) Formula of l-thyroxine (7). D) Inhibition of HIF-
1a accumulation by 7. E) Inhibition of oxygen consumption by
1 (20 mm) and 7 (80 mm) in HCT116 cells. Values are the mean
averages of three samples, vertical bars indicate S.D. **P<0.01,
compared with untreated control cells.
ure S17) and accumulation (Figure 6D) of HIF-1a. However,
1 (20 mm) demonstrated a much stronger effect on MDH2
inhibition (Figure 6B) and oxygen consumption than 7
(80 mm; Figure 6E). Thus, we suggest that LW6 binds to
MDH2 to block the TCA cycle, thereby inhibiting mitochon-
drial respiration and increasing the local oxygen tension. This
result explains our previous finding that 1 promotes HIF-1a
degradation through proteasome-dependent degradation.[4b]
This is the first report that MDH2 regulates HIF-1a
accumulation under hypoxia.
A high level of MDH2 expression was associated with
shorter, relapse-free survival and chemoresistance of prostate
cancer patients.[12] The knockdown of MDH2 enhanced
docetaxel sensitivity and induced metabolic inefficiency.
Future studies should examine the relevance of MDH2 with
respect to HIF-1 regulation in cancer and the potential
benefits of MDH2 inhibition with cancer therapeutics.
Our chances of isolating MDH2 as a target of 1 were
improved by using a photoactivatable trifluoromethyl diazir-
ine and a fluorescent probe in living cells. These modifications
overcame the common limitations of conventional methods
(e.g., poor detection or low affinity of target molecules). Non-
specific binding proteins were excluded by double fluorescent
labeling and 2DE.
Figure 5. Pull-down assay of the biotinylated probes. A) Formula of the
biotinylated probes. B) Affinity pull-down assay. HCT116 cell lysates
were incubated with biotinylated probe (5 mm) and avidin resin. MDH2
was detected by immunoblot analysis after SDS-PAGE. C) Competition
assay. The binding of MDH2 to biotinylated probe 6 (5 mm) was
inhibited by the presence of 1 (50 mm).
attenuated by addition of non-biotinylated 1 (Figure 5C).
These results clearly support an interaction between 1 and
MDH2.
In addition, fluorescent probe 3 bound to purified
recombinant human MDH2 (rh-MDH2; Figure 6A, middle
lane) and its binding was reduced in the presence of 1 (right
lane), indicating that probe 3 bound directly to MDH2. Next,
we examined the effect of 1 and probes on the MDH2 activity
using MDH2 isolated from HCT116 cells. We found that
1 and 3, but not 4, inhibited the MDH2 activity in a concen-
tration-dependent manner (Figure 6B; Figure S15) without
affecting the MDH2 expression (Figure S16). Then, we found
the MDH2 inhibitor l-thyroxine 7[11] (Figure 6C) inhibited
transcriptional activity (HRE-Luc IC50 = 26.3 Æ 0.97 mm; Fig-
In summary, we designed and synthesized a series of LW6
(1)-derived chemical probes by installing a clickable tag and
a photoactivatable moiety. We observed the mitochondrial
localization of 1 and identified MDH2 as a target protein. In
parallel with the MDH2 inhibition, 1 and MDH2 inhibitor 7
suppressed hypoxia-induced HIF-1a accumulation by inhib-
iting mitochondrial respiration. In conclusion, LW6 reduces
Angew. Chem. Int. Ed. 2013, 52, 1 – 5
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
One 2012, 7, e39292; b) L. Wang, Y. O. Son, S. Ding, X. Wang,
J. A. Hitron, A. Budhraja, J. C. Lee, Q. Lin, P. Poyil, Z. Zhang, J.
Luo, X. Shi, Toxicol. Sci. 2012, 130, 269 – 280.
HIF-1a accumulation by inhibiting MDH2 activity and
suppressing mitochondrial respiration.
[6] a) J. Park, M. Koh, S. B. Park, Mol. Biosyst. 2013, 9, 544 – 550;
b) U. Rix, G. Superti-Furga, Nat. Chem. Biol. 2009, 5, 616 – 624;
c) M. Schenone, V. Dancik, B. K. Wagner, P. A. Clemons, Nat.
Chem. Biol. 2013, 9, 232 – 240.
[7] a) F. Kotzyba-Hibert, I. Kapfer, M. Goeldner, Angew. Chem.
1995, 107, 1391 – 1408; Angew. Chem. Int. Ed. Engl. 1995, 34,
1296 – 1312; b) H. C. Kolb, M. G. Finn, K. B. Sharpless, Angew.
Chem. 2001, 113, 2056 – 2075; Angew. Chem. Int. Ed. 2001, 40,
2004 – 2021; c) V. V. Rostovtsev, L. G. Green, V. V. Fokin, K. B.
Sharpless, Angew. Chem. 2002, 114, 2708 – 2711; Angew. Chem.
Int. Ed. 2002, 41, 2596 – 2599.
Received: June 10, 2013
Published online: && &&, &&&&
Keywords: click chemistry · fluorescent probes · HIF inhibitors ·
.
malate dehydrogenase 2 · photoaffinity labeling
[1] G. L. Wang, B. H. Jiang, E. A. Rue, G. L. Semenza, Proc. Natl.
Acad. Sci. USA 1995, 92, 5510 – 5514.
[2] G. L. Semenza, Nat. Rev. Cancer 2003, 3, 721 – 732.
[3] G. L. Semenza, Curr. Opin. Genet. Dev. 2010, 20, 51 – 56.
[4] a) K. Lee, J. H. Lee, S. K. Boovanahalli, Y. Jin, M. Lee, X. Jin,
J. H. Kim, Y. S. Hong, J. J. Lee, J. Med. Chem. 2007, 50, 1675 –
1684; b) K. Lee, J. E. Kang, S. K. Park, Y. Jin, K. S. Chung, H. M.
Kim, K. Lee, M. R. Kang, M. K. Lee, K. B. Song, E. G. Yang, J. J.
Lee, M. Won, Biochem. Pharmacol. 2010, 80, 982 – 989; c) R.
Naik, M. Won, B. K. Kim, Y. Xia, H. K. Choi, G. Jin, Y. Jung,
H. M. Kim, K. Lee, J. Med. Chem. 2012, 55, 10564 – 10571.
[5] a) Y. Hashimoto, H. Tazawa, F. Teraishi, T. Kojima, Y. Wata-
nabe, F. Uno, S. Yano, Y. Urata, S. Kagawa, T. Fujiwara, PLoS
[8] Y. Chua, E. Dufour, E. Dassa, P. Rustin, H. Jacobs, C. Taylor, T.
Hagen, J. Biol. Chem. 2010, 285, 31277 – 31284.
[9] M. Hashimoto, Y. Hatanaka, Eur. J. Org. Chem. 2008, 2513 –
2523.
[10] C. R. Goward, D. J. Nicholls, Protein Sci. 1994, 3, 1883 – 1888.
[11] S. Varrone, E. Consiglio, I. Covelli, Eur. J. Biochem. 1970, 13,
305 – 312.
[12] Q. Liu, C. T. Harvey, H. Geng, C. Xue, V. Chen, T. M. Beer, D. Z.
Qian, Prostate 2013, 73, 1028 – 1037.
4
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Communications
Target Identification
Tracking the target: To identify a protein
that binds directly to the HIF-1a inhibitor
LW6, a series of new chemical probes
were synthesized with a clickable tag and/
or a photoactivatable moiety. LW6 was
found to be localized to the mitochondria
(see picture) and MDH2 was identified as
a target protein of LW6. These results
indicate that the HIF-1a inhibitory activity
of LW6 is a consequence of MDH2
suppression.
K. Lee, H. S. Ban, R. Naik, Y. S. Hong,
S. Son, B.-K. Kim, Y. Xia, K. B. Song,
H.-S. Lee, M. Won*
&&&& — &&&&
Identification of Malate Dehydrogenase 2
as a Target Protein of the HIF-1 Inhibitor
LW6 using Chemical Probes
Angew. Chem. Int. Ed. 2013, 52, 1 – 5
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!