Synthesis, structure, and biological evaluation of a platinum-carbazole conjugate

Article abstract of DOI:10.1111/cbdd.13062

Cisplatin resistance is caused, in part, by the efficient removal of the helix-distorting cisplatin 1,2-intrastrand cross-links by nucleotide excision repair (NER) machinery. To make a platinum-DNA adduct that causes less helical distortion than the cisplatin 1,2-intrastrand adduct, we designed and synthesized a monofunctional platinum-carbazole conjugate (carbazoplatin). The 2.5?? crystal structure of carbazoplatin-DNA adduct revealed both the monoplatination of the N7 of a guanine (G) base and the intercalation into two G:C base pairs, while causing a minor distortion of the DNA helix. A 50-mer dsDNA containing a single carbazoplatin lesion was poorly processed by UvrABC endonuclease, the prokaryotic NER machinery that detects helical distortion and performs dual incision around the lesion. Our cell viability assay indicated that the cytotoxic pathways of carbazoplatin might be different from those of cisplatin; carbazoplatin was 5–8 times more cytotoxic than cisplatin against PANC-1 and MDA-MB-231 cancer cell lines.

Full text of DOI:10.1111/cbdd.13062

PROF. SEONGMIN LEE (Orcid ID : 0000-0001-6635-8775)
Article type : Research Article
Synthesis, Structure, and Biological Evaluation of a
Platinum-Carbazole Conjugate
Young Cheun,1* Myong-Chul Koag,1* Youssef W. Naguib,2,3 Hala Ouzon-Shubeita,1 Zhengrong Cui,2
Danaya Pakotiprapha,4 and Seongmin Lee1
1Division of Chemical Biology and Medicinal Chemistry, College of Pharmacy, The University of
Texas at Austin, Austin, TX 78712, United States
2Division of Pharmaceutics, College of Pharmacy, The University of Texas at Austin, Austin, TX
78712, United States
3Department of Pharmaceutics, Faculty of Pharmacy, Minia University, Minia, Egypt
4Department of Biochemistry and Center of Excellence in Protein and Enzyme Technology, Faculty
of Science, Mahidol University, Bangkok 10400, Thailand
*These authors equally contributed to the work.
Corresponding Author
Seongmin Lee, 1Division of Chemical Biology and Medicinal Chemistry, College of Pharmacy, The
University of Texas at Austin, Austin, TX 78712, United States
Phone: +1 (512) 471-1785
Email: seongminlee@austin.utexas.edu
This article has been accepted for publication and undergone full peer review but has not
been through the copyediting, typesetting, pagination and proofreading process, which may
lead to differences between this version and the Version of Record. Please cite this article as
doi: 10.1111/cbdd.13062
This article is protected by copyright. All rights reserved.

KEYWORDS: DNA damage, Cisplatin, X-ray crystallography, Nucleotide excision repair, Cancer
therapy
ABSTRACT
Cisplatin resistance is caused, in part, by the efficient removal of the helix-distorting cisplatin
1,2-intrastrand crosslinks by nucleotide excision repair (NER) machinery. To make a platinum-DNA
adduct that causes less helical distortion than the cisplatin 1,2-intrastrand adduct, we designed and
synthesized a monofunctional platinum-carbazole conjugate (carbazoleplatin). The 2.5Å crystal
structure of carbazoleplatin-DNA adduct revealed both the monoplatination of the N7 of a guanine
(G) base and the intercalation into two G:C base pairs while causing a minor distortion of the DNA
helix. A 50-mer dsDNA containing a single carbazoleplatin lesion was poorly processed by UvrABC
endonuclease, the prokaryotic NER machinery that detects helical distortion and performs dual
incision around the lesion. Our cell viability assay indicated that the cytotoxic pathways of
carbazoleplatin might be different from those of cisplatin; carbazoleplatin was 5 to 8 times more
cytotoxic than cisplatin against PANC-1 and MDA-MB-231 cancer cell lines.
INTRODUCTION
cis-Diamminedichloridoplatinum(II), previously cis-Diamminedichloroplatinum(II), or
commonly cisplatin is a classical bifunctional platinum-based anticancer drug, which has been
administered to more than half of chemotherapy patients (Figure 1).[1] While cisplatin targets multiple
cellular targets, it has been widely accepted that DNA is the main target that gives the cytotoxicity.[2,3]
Upon binding to the target DNA, cisplatin forms a complex with the N7 atoms of two guanines in
close proximity, generating 1,2-intrastrand, 1,3-intrastrand, and interstrand crosslink adducts. The
most predominant form (approximately 65%), 1,2-intrastrand crosslink, distorts the DNA helix by
unwinding the helix and exposing the minor groove opposite to the lesion (verified by both NMR[4]
and X-ray[5,6]), thus inhibiting transcription and replication and ultimately inducing apoptosis.[7,8]
This article is protected by copyright. All rights reserved.

Drug resistance is one of the major limitations of cisplatin-based chemotherapy. It has been
reported that relapses occurred on most of the ovarian cancer patients who had been treated with
cisplatin in about 15 months, leading to poor prognosis (15-20% survival in 5 years).[9,10] One
mechanism of cisplatin resistance is the removal of cisplatin-DNA intrastrand crosslink adducts by
nucleotide excision repair (NER) pathway.[8] Cisplatin-sensitive cancer cells were reported to exhibit
low levels of NER proteins (e.g., XPA, ERCC-XRF), whereas increased NER activity was observed
in cisplatin-resistant cancer cells.[7,11-13]
To circumvent cisplatin resistance, researchers explored the potential of new structural motifs
that deviate from the traditional structure-activity relationship (SAR) guidelines.[14,15] Among those
motifs, polynuclear platinum agent, monofunctional platinum agent, and platinum-intercalator
conjugate have shown some promising preclinical results (Figure 1).[16] Originally intended to study
biophysical properties of DNA[17] and to explore the potential synergism between cisplatin and
doxorubicin,[18] platinum-intercalator conjugate is now a viable concept to overcome cisplatin
resistance.[19-21] A platinum-intercalator conjugate is characterized by a platinum core and a DNA
intercalator connected with a covalent linker, and in general, it is capable of inhibiting the replication
while stabilizing the duplex structure of the target DNA.[16,22] Depending on the platinum core, the
DNA intercalator, and the linker, the conjugate can exert different physical and biological outcomes.
The laboratory of Ulrich Bierbach developed a platinum-acridine conjugate connected by a central 2-
aminoethyl thiourea/guanidine linker (called Pt-ACRAMTU) that promotes the intercalation-driven
platination by utilizing a strongly intercalating acridine moiety and a relatively rigid aminoethyl
guanidine.[19] Pt-ACRAMTU is thought to stall DNA processing enzymes by unwinding and
lengthening the DNA helix at the intercalation site while inducing small conformational changes of
DNA,[23,24] and it was shown to be effective against several cancer cell lines.[19] The laboratory of Curt
Davey developed a platinum-naphthylimide conjugate (called PtNAP), and its cis- and trans- isomers
target different DNA sites in the nucleosome.[25] Both of the isomers were shown to be effective
against a cisplatin-resistant cancer cell line.[25]
This article is protected by copyright. All rights reserved.

In conjunction with our efforts to develop a platinum-based anticancer agent that can be
slowly removed by NER, herein we report the synthesis, the structure, and the bioactivity of a novel
platinum-carbazole conjugate connected with an ethylene diamine linker (hereafter called
carbazoleplatin); in contrast to Pt-ACRAMTU or PtNAP, our molecule uses an aromatic heterocycle
that had not been previously used as a DNA intercalator and a more flexible, compact covalent linker.
We evaluated carbazoleplatin’s mode of interaction by obtaining a crystal structure of
carbazoleplatin-DNA adduct using the pol β host-guest system that we had recently developled in our
laboratory.[26] We also tested how a UvrABC complex (the prokaryotic NER) removes a
carbazoleplatin adduct from its DNA substrate and how cancer cells respond to carbazoleplatin.
1. EXPERIMENTAL SECTION
1.1. Chemistry
1.1.1. Preparation of 2-(9H-carbazol-9-yl)ethan-1-ol. Commercially available starting
material carbazole (3.34 g, 20.0 mmol) was dissolved in acetone (40 mL, 0.5M). Finely crushed
potassium hydroxide (KOH, 7.0 g, 6.3 mmol) was added to the solution. The mixture was stirred at
room temperature for 30 minutes, and then 2-bromoethanol (1.4 mL, 1.25 equiv) was added to the
mixture dropwise. The reaction mixture was refluxed at 60ºC under argon gas for 18 hours in the
dark. Once the product formation was determined via thin layer chromatography, the solvent was
removed in vacuo. The mixture was subject to work-up (ethyl acetate/water), and the resulting organic
layer was washed with brine and dried over sodium sulfate. The crude mixture was purified via silica
gel column chromatography (4:1 hexane/ethyl acetate) to afford the desired product as brown solids
(1.34 g, 6.3 mmol, 32%). Cleaner reaction and improved yield could be obtained with the use of
1
1
sodium hydride instead of KOH. in acetone. H NMR (CDCl3) spectrum is available. H NMR
(CDCl , 400 MHz): δ 8.11 (2H, d, J = 7.9 Hz), 7.47 (4H, m), 7.25 (2H, m), 4.50 (2H, t, J = 5.5 Hz),
3
4.09 (2H, m).
This article is protected by copyright. All rights reserved.

1.1.2. Preparation of 9-(2-azidoethyl)-carbazole. The starting material 2-(9H-carbazol-9-
yl)ethan-1-ol (1.8 g, 8.5 mmol) was dissolved in tetrahydrofuran (THF, 85 mL, 0.1 M). Triphenyl
phosphine (4.46 g, 2.0 equiv) and imidazole (2.9 g, 5.0 equiv) were added to the solution. While
stirring the mixture at room temperature, iodine chips (4.5 g, 2.1 equiv) were added to the mixture in
one portion. The mixture was stirred for 30 minutes. Once the product formation was determined via
thin layer chromatography, aqueous sodium thiosulfate was added to the mixture until the color
disappeared. The reaction mixture was subject to work-up (ethyl acetate/water), and the resulting
organic layer was washed with brine and dried over sodium sulfate. The crude mixture was purified
via silica gel column chromatography (8:1 hexane/ethyl acetate) to afford the desired iodide. The
iodide was dissolved in N,N-dimethylformamide (DMF, 50 mL), and sodium azide (5.5 g 10.0 equiv)
was added to the DMF solution. The mixture was refluxed at 70ºC under argon for 1 day. Once the
product formation was determined via thin layer chromatography, the solvent was removed in vacuo.
The reaction mixture was subject to work-up (ethyl acetate/water), and the resulting organic layer was
washed with brine and dried over sodium sulfate. The crude mixture was purified via silica gel
1
column chromatography to afford the desired azide as yellow solids (2.03 g, 8.6 mmol, 100%). H
1
NMR (CDCl3) spectrum is available. H NMR (CDCl3, 400 MHz): δ 8.11 (2H, t, J = 7.9 Hz), 7.47
(4H, m), 7.27 (2H, m), 4.50 (2H, t, J = 6.2 Hz), 3.74 (2H, t, J = 6.16 Hz).
1.1.3. Preparation of 2-(9H-carbazol-9-yl)ethan-1-amine. The starting material 9-(2-
azidoethyl)-carbazole (840 mg, 3.6 mmol) was dissolved in tetrahydrofuran (THF, 35 mL, 0.1 M).
Triphenyl phosphine (2.8 g, 3.0 equiv), and then water (0.64 mL, 10.0 equiv) were added to the
solution, and the mixture was stirred at room temperature for 16 hours. Once the product formation
was determined via thin layer chromatography, the reaction mixture was subject to work-up (ethyl
acetate/water). The resulting organic layer was washed with brine and dried over sodium sulfate. The
crude mixture was purified via silica gel column chromatography (6:1 hexane/ethyl acetate) to afford
1
the desired product as yellow oil (433 mg, 2.1 mmol, 58%). H NMR (CDCl3 and DMSO-d6) spectra
1
are available. H NMR (CDCl3, 400 MHz): δ 8.11 (2H, d, J =7.9 Hz), 7.47 (4H, d, J = 3.8 Hz), 7.25
1
(2H, m), 4.42 (2H, t, J = 6.2 Hz), 3.23 (2H, t, J = 6.2 Hz), 1.52 (2H, br). H NMR (DMSO-d6, 400
This article is protected by copyright. All rights reserved.

MHz): δ 8.12 (2H, d, J = 7.9 Hz), 7.61 (2H, d, J = 8.2 Hz), 7.42 (2H, t, J = 7.2 Hz), 7.16 (2H, t, J =
7.2 Hz), 4.34 (2H, t, J = 6.8 Hz), 2.91 (2H, d, J = 6.8 Hz).
1.1.4. Preparation of carbazoleplatin. Cis-dichlorodiamminoplatinum(II) (619 mg, 1.0
equiv) was dissolved in N,N-dimethylformamide (DMF, 33 mL). Silver nitrate (350 mg, 1.0 equiv)
was added to the solution in the dark, and the mixture was stirred at 55ºC under argon gas for 1 day.
The reaction mixture was filtered though 20 µm nylon filter to remove insolubles. The filter cake was
washed with DMF. The clear and colorless filtrate was transferred to a new flask. To the flask, the
starting material 2-(9H-carbazol-9-yl)ethan-1-amine (433 mg, 2.06 mmol) was added to the mixture.
The new reaction mixture was stirred at 55ºC under argon gas in the dark. After 1 additional day, the
reaction mixture was taken out from the heat source and the solvent was removed in vacuo. The crude
mixture was dissolved in minimum volume of methanol. 10× (by volume) cold diethyl ether was
poured to the methanol. The resulting white solids were collected and further purified with additional
methanol/ether wash. 1H NMR (DMSO-d6), 13C NMR (DMSO-d6), 195Pt NMR (DMSO-d6), and hrMS
(ESI) spectra are available. 1H NMR (DMSO-d6, 400 MHz): δ 8.15 (2H, d, J = 7.5 Hz), 7.65 (2H, d, J
= 7.9 Hz), 7.48 (2H, t, J = 7.2 Hz), 7.20 (2H, t, J = 7.5 Hz), 5.10 (2H, br), 4.63 (2H, t), 4.31 (3H, br),
3.99 (3H, br), 2.92 (2H, m). 13C NMR (DMSO-d6, 150 MHz): δ 139.8, 125.9, 122.2, 120.4, 119.1,
109.0, 43.3, 42.2. 195Pt NMR (DMSO-d6, 129 MHz): δ –2390. hrMS (ESI): chemical formula:
C14H20ClN4Pt+, calculated m/z 474.1024, observed m/z 474.1020.
1.2. Structure Characterization of Carbazoleplatin
1.2.1. Structure Determination of Carbazoleplatin. The crystals grew as colorless plates by
methanol:diethyl ether vapor diffusion. The crystal was cut from a larger fragment and had
approximate dimensions; 0.18 × 0.11 × 0.04 mm. The data were collected on an Agilent
Technologies SuperNova Dual Source diffractometer using a -focus Cu K radiation source ( =
1.5418Å) with collimating mirror monochromators. A total of 989 frames of data were collected
using w-scans with a scan range of 1 and a counting time of 2 seconds per frame with a detector
offset of ± 41.0 and 5 seconds per frame with a detector offset of ± 108.3. The data were collected
This article is protected by copyright. All rights reserved.

at 100 Kusing an Oxford Cryostream low temperature device. Data collection, unit cell refinement
and data reduction were performed using Agilent Technologies CrysAlisPro. The structure was solved
by direct methods using SuperFlip and refined by full-matrix least-squares on F2 with anisotropic
displacement parameters for the non-H atoms using SHELXL-2013. Structure analysis was aided by
use of the programs PLATON98 and WinGX. The hydrogen atoms were calculated in ideal positions
with isotropic displacement parameters set to 1.2 × Ueq of the attached atom.
1.2.2. Structure Determination of Carbazoleplatinated DNA using pol β Host-Guest
Complex. Human DNA polymerase β (pol β) was expressed and purified from Escherichia coli as
described previously,[27,28] 16-mer template (5’-CCG ACG GAG GAG CAG G-3’), 10-mer upstream
(5’-CCT GCT CCT C-3’), and 5-mer downstream oligonucleotides (5’-phosphate/GTC GG-3’) were
purchased from Integrated DNA Technologies (Coralville, IA). Platinated upstream oligo was
prepared as described previously with minor modifications.[29] Carbazoleplatin was aquated by adding
1.2 equivalent of AgNO3 to 3.0 mM of carbazoleplatin in water. The mixture was vortexed in the dark
at room temperature for 15 hours and centrifuged for 20 minutes at 13,000×g to collect the
supernatant. 1.5 mol of the unmodified upstream primer in 10 mM sodium phosphate buffer (pH 6.3)
was mixed with the aquated carbazoleplatin at 1:1.1 molar ratio. Then, the mixture was incubated at
37ºC for 15 hours. The platinated primer was purified using a GE Healthcare Mono Q 5/50 GL ion
exchange column in 10 mM Tris buffer (pH 8.0) with NaCl gradient from 0.1 to 1.0 M.[30] Then the
purified primer was desalted using a Waters Corporation Sep-Pak C18 column (Milford, MA), dried,
and reconstituted in water. The purified upstream primer, the phosphorylated downstream, and the
template were annealed in 20 mM MgCl2, 0.1 M Tris buffer (pH 7.5) to afford single-nucleotide
gapped DNA as described.[27] The pol β-DNA host-guest complex crystals were grown in 50 mM
imidazole, 14-23% PEG3400, and 350 mM sodium acetate at pH 7.5 as previously reported.[30]
Crystals were cryo-protected in mother liquor supplemented with 12% ethylene glycol, and they were
flash-frozen in liquid nitrogen. Diffraction data were collected at 100 K using either a Rigaku
MicroMax-007 HF microfocus X-ray generator with R-Axis IV++imaging plate area detector or the
beamline 5.0.3 Advanced Light Source at Berkeley Center for Structural Biology. The diffraction data
This article is protected by copyright. All rights reserved.

were processed using HKL 2000.[31] The structures of the pol β-DNA host-guest complexes were
solved by molecular replacement with pol β with a single-nucleotide gapped DNA (PDB ID 1BPX) as
the search model. The model was built using COOT[32] and refined using PHENIX software.[33]
Ramachandran plots were generated using MolProbity program.
1.2.3. DNA Torsion Analysis. We extracted the atom coordinates of the upstream
complementary DNA and the corresponding template strand from the PDB file and analyzed
distortion parameters using W3DNA (http://w3dna.rutgers.edu), a web based platform of the
distortion analysis software 3DNA v2.0.[34]
1.3. Biological Properties of Carbazoleplatin
1.3.1. UvrABC Incision Assay. The laboratory of Danaya Pakotiprapha (Mahidol University,
Thailand) has kindly given us the constructs of UvrA, UvrB, and UvrC. Purified UvrA, UvrB, and
UvrC from Bacillus stearothermophilus were prepared as described previously.[35] Fluorescein-labeled
template strand 5’-FAM/ CTC TCT CTT CTC TTT TCC CTC TCT TGX TCC TTC TCT CCC CTC
TCC TCT CC-3’ and its complementary strand 5′-GGA GAG GAG AGG GGA GAG AAG GAY
CAA GAG AGG GAA AAG AGA AGA GAG AG-3′ (X and Y = C and G for the carbazoleplatin
lesion, G and C for the cisplatin lesion) were purchased from Integrated DNA Technologies
(Coralville, IA). Aquated carbazoleplatin and cisplatin were mixed with the corresponding template
strand (platinum: DNA molar ratio = 2:1) in 10 mM sodium phosphate buffer (pH 6.8) and incubated
at 37ºC in the dark for 18 hours. Each lesion-containing ssDNA was purified using a GE Healthcare
Mono Q 5/50 GL ion exchange column in 10 mM Tris buffer (pH 8.0) with 3-step NaCl gradient (0.0
to 0.15 M for 0.5 column volume, 0.15 to 0.5 M for 25 column volume, 0.5 to 1.0 M for 0.5 column
volume). The desired fractions were analyzed via 20% polyacrylamide urea gel electrophoresis. After
the desalting, each lesion-containing ssDNA was annealed with the corresponding complementary
strand by mixing them in 1:1 molar ratio and heating the mixture at 90ºC for 10 minutes followed by
slow cooling. 40 nM of the annealed DNAs were pre-incubated with 500 nM UvrA and 500 nM UvrB
in 50 mM Tris-HCl pH 7.5, 10 mM MgCl2, 100 mM KCl, 0.1 μg/μl BSA, 1 mM ATP and 2%
This article is protected by copyright. All rights reserved.

glycerol at 37 °C for 2 minutes. Then, 500 nM of UvrC was added to each mixture. After 1 hour, each
reaction was quenched with formamide loading buffer (95% formamide, 20 mM EDTA, pH 8.3, 0.1
mg/mL bromophenol blue, and 0.1 mg/mL xylene cyanol). The reaction products were denatured by
heating the mixture at 90 °C for 5 min and separated by molecular weight via 20% polyacrylamide
urea gel electrophoresis. The 6-carboxyfluorescein (6-FAM) dye was visualized using GE Life
Sciences Typhoon FLA 9500 (excitation laser: 473 nm, emission filter: LPB). The intensity was
quantified using ImageJ.[36]
1.3.2. MTT Cytotoxicity Assay. Carbazoleplatin and cisplatin were dissolved in RPMI1640
media (Invitrogen, Carlsbad, CA) to prepare stock solutions, then serial dilutions in media were made
as 100, 10, 1, 0.1, 0.01, and 0.001 µM. Cells (PANC1 and MDA-MB-231, American Type Culture
Collection, Manassas, VA) were seeded as 1500 cells/well for overnight (in 75 µL). On the second
day, carbazoleplatin or cisplatin was added to the wells to final concentrations as aforementioned (75
µL per well, n = 4). After 48 hours of incubation at 37ºC with 5% CO2, 20 µL of MTT reagent (5
mg/mL in PBS) were added, and the cells were incubated for two additional hours. All the aqueous
media were removed, and 100 µL of dimethyl sulfoxide were added to dissolve the formed formazan
crystals. After crystals were dissolved, the visible absorbance of the wells was measured using a plate
reader at 570 nm, while 630 nm was used as a reference wavelength. Cell survival (%) was calculated
using the equation of (A570 – A630) / (control mean) × 100%. Control mean is the average of 10 control
wells (medium only). The data were fitted to the normalized dose-response curve (response = 100 / [1
+ 10^((log10(IC50) – log10(dose)) × hill slope)])) using GraphPad Prism 6.[37]
2. RESULTS AND DISCUSSION
2.1. Design, Synthesis, and Structure of Carbazoleplatin. We looked into the possibility of
using carbazole as part of our new platinum-intercalator conjugate for two reasons: novelty and
toxicity. Carbazole is a polyaromatic heterocycle containing two benzene rings fused on either side of
a five-membered ring with a nitrogen atom. Carbazole has diverse biological activities, including
This article is protected by copyright. All rights reserved.

antitumor activities.[38] It has been speculated that carbazole may intercalate into DNA helix,[38,39] but
carbazole itself has not been used as a DNA intercalating agent in our best knowledge.[40,41] In
addition, carbazole is less toxic (LD50 >5000 mg/kg) than most of the classical intercalators such as
ethidium or acridine.[42] Toxicity is the major limitation of current platinum-intercalator conjugates,[43]
so we thought there might be a potential benefit of making a less toxic molecule.
We synthesized [PtCl(NH3)2(2-(9H-carbazol-9-yl)ethan-1-amine)](NO3), or carbazoleplatin,
from 9H-carbazole in 5 steps with some modifications of the reported synthetic procedures of
monofunctional platinum compounds (Scheme 1A).[44,45] The chemical structure of carbazoleplatin
was unambiguously characterized by single crystal X-ray crystallography (Scheme 1B, Table 1). This
simple design and efficient synthesis route provide the possibility for facile structural optimizations
around the carbazole ring, the linker, and the platinum.
2.2. Crystal Structure of Carbazoleplatin-DNA Adduct. To define the DNA-binding mode
of carbazoleplatin, we determined a crystal structure of carbazoleplatin-DNA adduct using human
DNA polymerase β host-guest complex (pol β-HGC) system, which we had previously used to
determine the structure of N7-methylguanine-containing DNA .[26,46] The system enables the facile
determination of a three-dimensional structure of a lesion-containing DNA in B-DNA conformation
in the absence of a protein contact (Figure 2A).[26,46] A single nucleotide gapped DNA containing N7-
carbazoleplatin-G-adduct at the upstream primer was incubated with pol β to give a binary complex
crystal (Figure 2A). The N7-carbazoleplatin-G binary complex structure, refined at 2.5 Å resolution
(Table 2), revealed that the carbazole-platinum conjugate platinated N7 of a guanine, and the
carbazole moiety intercalated into two base pairs (Figure 2, PDB ID 5HHI). The long axis of the
carbazole moiety was parallel to the long axis of Watson-Crick base pair, thereby making extensive π-
π stacking interactions (Figure 2B). The carbazole moiety of carbazoleplatin was predominantly
intercalated into two G:C base pairs, as indicated by the strong electron density (Figure 2C). Since
platination in general occurs at the N7 of a guanine, the intercalation of carbazole was predicted to
occur between G:C and G:C base pairs or between G:C and A:T base pairs. In our pol β-HGC
experiment setting (5’-TGC-3’, G denotes the platination site), the platinated-G:C base pair was
This article is protected by copyright. All rights reserved.

flanked by A:T and G:C. The intercalation of carbazoleplatin was observed at the interface between
two G:C base pairs rather than between G:C and A:T base pairs. Is it the base pair or the directionality
that primarily influences this intercalation pattern? In the published Pt-ACRAMTU-DNA structure
that contains 5’-CGT-3’ sequence (G denotes the platination site), its acridine moiety intercalated into
5’-CG-3’ rather 5’-GT-3’ base pairs.[23] Thus, carbazoleplatin’s preferential intercalation into 5’-GC-
3’ and 5’-CG-3’ is likely driven by the surrounding base pairs rather than the directionality. It could
be possible that an intercalation into two G:C base pairs is more thermodynamically favorable than an
intercalation into G:C and A:T base pairs.
We compared the crystal structure of the carbazoleplatin-containing DNA in pol β to the
crystal structure of a lesion-free DNA in pol β (Table 2, PDB ID 5HHH), and it appeared that
carbazoleplatin did not induce significant changes in the most of the distortion parameters (Table 3).
2.3. NER Repair Assay of Carbazoleplatin-DNA Adduct. In order to assess the
carbazoleplatin-induced helical distortion in a biological context, we tested how efficiently UvrABC
endonuclease, the prokaryotic NER system, performs the double incision on carbazoleplatin-
containing DNA (Figure 3). We observed that the UvrABC incision rate was approximately 20% per
hour when the substrate contained the carbazoleplatin lesion (16% dual incision + 4% single incision,
land 6 of Figure 3). In the case of cisplatin 1,2-intrastrand lesion, the incision rate was approximately
45% per hour (34% dual incision + 11% single incision, lane 3 of Figure 3).
Both prokaryotic and eukaryotic NER machineries recognize and/or remove the similar range
of lesions through coordinated actions.[47-52] The UvrABC, which is present only in prokaryotes and
archaea, has been used as a model system for studying NER efficiency of bulky and/or helix-
distorting lesions ranging from pyrimidine dimer to cisplatin intrastrand crosslink adducts.[53]
Although prokaryotic NER requires a much less proteins than eukaryotic NER, the both systems
employ a very similar strategy including damage recognition, dual incision, repair and ligation.
Similar to eukaryotic NER, the UvrABC-mediated repair is a multistep process: (1) damage
recognition by UvrA2B complex, (2) formation of the preincision complex, and (3) UvrC binding and
This article is protected by copyright. All rights reserved.

double incision.[54] A thorough comparison study of the UvrABC’s repair efficiency against cisplatin
1,2-intrastrand crosslink to that of 1,3-intrastrand crosslink revealed that the last incision step by
UvrABC is the rate-limiting step, but it was found that the nature of the substrate also affect the steps
before the incision.[54] In our UvrABC assay, the formation of the single incision and the double
incision products of the lane 6 (Figure 3) suggest that carbazoleplatin damage is recognized by
UvrA2B and the preincision complex could be formed on the lesion, but UvrC could not incise the
lesion as efficient as the cisplatin 1,2-intrastrand lesion. We speculate the carbarzoleplatin’s
diminutive distortion of DNA made UvrC difficult to incise the lesion.
2.4. Cell Viability Assay of Carbazoleplatin. We examined antiproliferative activity of
carbazoleplatin against human cancer cell lines via MTT cell viability assay. Carbazoleplatin was
found to be 5 and 9 times more cytotoxic than cisplatin against pancreatic cancer (PANC-1) and triple
negative breast cancer (MDA-MB-231) cell lines, respectively (Table 4). The higher cytotoxicity of
carbazoleplatin might result from carbazoleplatin’s distinctive mode of action, in which the carbazole
intercalation is assisted and stabilized by the platination of monofunctional platinum. The platinum
moiety may provide additional specificity toward DNA via its cationic charge and prodrug-like
platinum-chloride bond.[22,55]
Although the cytotoxicity data should not be taken literally,[56] the greater cytotoxicity of
carbazoleplatin compared to that of cisplatin would serve to warrant the future mechanism of action
studies of carbazoleplatin. The cellular uptake of carbazoleplatin could be different from that of
cisplatin. While cisplatin’s main uptake pathway is passive diffusion,[8] carbazoleplatin bearing the
bulky liphophilic carbazole moiety would not be readily transported into the nucleus by passive
diffusion, which could influence the cytotoxicity and drug resistance of carbazoleplatin. While
cisplatin-DNA adduct formation does not involve intercalation, the N7-carbazoleplatin-dG adduct
formation could be facilitated by intercalation, which can interfere with topoisomerase-mediated
strand scission and religation. The monofunctional intercalating platinum agent Pt-ACRAMTU has
been shown to stall DNA replication, inhibit transcription and induce S-phase arrest.[57] The
mechanism of action of carbazoleplatin, a monofunctional-intercalative platinum agent, could be
This article is protected by copyright. All rights reserved.

similar to that of Pt-ACRAMTU. Assessing the effect of carbazoleplatin-DNA adduct on biological
processes such as DNA replication, transcription, repair, and topoisomerization in NER-proficient and
NER-deficient cells would be required for elucidation of the mechanism of action of carbazoleplatin.
3. CONCLUSIONS
We have reported the synthesis, the structure, and the bioactivity of a platinum-intercalator
conjugate (carbazoleplatin). We visualized its DNA binding mode via X-ray, and we also tested how
its DNA adduct is repaired by a NER complex. Carbazoleplatin modified the N7 of a guanine, and its
carbazole moiety intercalated into the nearby 5’-GC-3’ base pairs. Carbazoleplatin induced minimal
helical distortion upon DNA modification, and its DNA adduct was less efficiently removed by the
UvrABC system than the cisplatin counterpart. These early stage assays enhance our confidence in
carrying out more rigorous biochemical and cellular studies of carbazoleplatin in the future.
ACKNOWLEDGMENTS
The research was supported by grant from the Cancer Prevention Research Institute of Texas
(Grant RP130219) We are grateful to Dr. Arthur Monzingo for technical assistance. Instrumentation
and technical assistance for this work were provided by the Macromolecular Crystallography Facility,
with financial support from the College of Natural Sciences, the Office of the Executive Vice
President and Provost, and the Institute for Cellular and Molecular Biology at the University of Texas
at Austin. The Berkeley Center for Structural Biology is supported in part by the National Institute of
General Medical Sciences of the National Institute of Health. The Advanced Light Source is
supported by the Director, Office of Science, Office of Basic Energy Sciences, of the U.S.
Department of Energy under Contract No. DE-AC02-05CH11231.
This article is protected by copyright. All rights reserved.

CONFLICT OF INTEREST
The authors declare no conflicts of interest.
REFERENCES
[1] M. Galanski, M.A. Jakupec, B.K. Keppler, Curr. Med. Chem. 2005, 12, 2075.
[2] K. Wang, J.F. Lu, R.C. Li, Coord. Chem. Rev. 1996, 151, 53.
[3] E.R. Jamieson, S.J. Lippard, Chem. Rev. 1999, 99, 2467.
[4] P.M. Takahara, A.C. Rosenzweig, C.A. Frederick, S.J. Lippard, Nature 1995, 377, 649.
[5] P.M. Takahara, C.A. Frederick, S.J. Lippard, J. Am. Chem. Soc. 1996, 118, 12309.
[6] R.C. Todd, S.J. Lippard, J. Inorg. Biochem. 2010, 104, 902.
[7] D. Wang, S.J. Lippard, Nat. Rev. Drug Discov. 2005, 4, 307.
[8] Y. Jung, S.J. Lippard, Chem. Rev. 2007, 107, 1387.
[9] R.F. Ozols, Cancer Treat Rev 1991, 18 Suppl A, 77.
[10] Z.H. Siddik, Oncogene 2003, 22, 7265.
[11] B. Koberle, J.R. Masters, J.A. Hartley, R.D. Wood, Curr. Biol. 1999, 9, 273.
[12] C. Welsh, R. Day, C. McGurk, J.R. Masters, R.D. Wood, B. Koberle, Int. J. Cancer 2004, 110,
352.
[13] K.V. Ferry, T.C. Hamilton, S.W. Johnson, Biochem. Pharmacol. 2000, 60, 1305.
[14] M.J. Cleare, J.D. Hoeschele, Bioinorganic Chemistry 1973, 2, 187.
[15] T.C. Johnstone, G.Y. Park, S.J. Lippard, Anticancer Res. 2014, 34, 471.
[16] T.C. Johnstone, K. Suntharalingam, S.J. Lippard, Chem. Rev. 2016, 116, 3436.
[17] M.V. Keck, S.J. Lippard, J. Am. Chem. Soc. 1992, 114, 3386.
[18] F. Zunino, G. Savi, A. Pasini, Cancer Chemother Pharmacol 1986, 18, 180.
[19] E.T. Martins, H. Baruah, J. Kramarczyk, G. Saluta, C.S. Day, G.L. Kucera, U. Bierbach, J. Med.
Chem. 2001, 44, 4492.
[20] M.C. Ackley, C.G. Barry, A.M. Mounce, M.C. Farmer, B.E. Springer, C.S. Day, M.W. Wright,
S.J. Berners-Price, S.M. Hess, U. Bierbach, J. Biol. Inorg. Chem. 2004, 9, 453.
This article is protected by copyright. All rights reserved.

[21] S.M. Hess, A.M. Mounce, R.C. Sequeira, T.M. Augustus, M.C. Ackley, U. Bierbach, Cancer
Chemotherapy and Pharmacology 2005, 56, 337.
[22] H. Baruah, C.G. Barry, U. Bierbach, Curr. Top. Med. Chem. 2004, 4, 1537.
[23] H. Baruah, M.W. Wright, U. Bierbach, Biochemistry 2005, 44, 6059.
[24] S. Dutta, M.J. Snyder, D. Rosile, K.L. Binz, E.H. Roll, J. Suryadi, U. Bierbach, M. Guthold, Cell
Biochem Biophys 2013, 67, 1103.
[25] E.Y. Chua, G.E. Davey, C.F. Chin, P. Droge, W.H. Ang, C.A. Davey, Nucleic Acids Res. 2015,
43, 5284.
[26] Y. Kou, M.C. Koag, S. Lee, J. Am. Chem. Soc. 2015, 137, 14067.
[27] M.R. Sawaya, R. Prasad, S.H. Wilson, J. Kraut, H. Pelletier, Biochemistry 1997, 36, 11205.
[28] M.C. Koag, S. Lee, J. Am. Chem. Soc. 2014, 136, 5709.
[29] A. Tremeau-Bravard, T. Riedl, J.M. Egly, M.E. Dahmus, J. Biol. Chem. 2004, 279, 7751.
[30] V.K. Batra, W.A. Beard, D.D. Shock, J.M. Krahn, L.C. Pedersen, S.H. Wilson, Structure 2006,
14, 757.
[31] Z. Otwinowski, W. Minor, Methods Enzymol. 1997, 276, 307.
[32] P. Emsley, K. Cowtan, Acta Crystallogr D Biol Crystallogr 2004, 60, 2126.
[33] P.D. Adams, R.W. Grosse-Kunstleve, L.W. Hung, T.R. Ioerger, A.J. McCoy, N.W. Moriarty,
R.J. Read, J.C. Sacchettini, N.K. Sauter, T.C. Terwilliger, Acta Crystallogr D Biol Crystallogr 2002,
58, 1948.
[34] X.J. Lu, W.K. Olson, Nucleic Acids Res. 2003, 31, 5108.
[35] D. Pakotiprapha, Y. Inuzuka, B.R. Bowman, G.F. Moolenaar, N. Goosen, D. Jeruzalmi, G.L.
Verdinel, Mol. Cell 2008, 29, 122.
[36] W. Rasband. (1997) ImageJ. US National Institutes of Health, Bethesda, Maryland. ImageJ. US
National Institutes of Health, Bethesda, Maryland. Available.
[37] One-way ANOVA followed by Dunnett’s multiple comparisons test was performed using
GraphPad Prism version 6.00 for Windows. One-way ANOVA followed by Dunnett’s multiple
comparisons test was performed using GraphPad Prism version 6.00 for Windows. GraphPad
Software, La Jolla, California, USA.
[38] M.S. Shaikh, R. Karpoormath, N. Thapliyal, R.A. Rane, M.B. Palkar, A.M. Faya, H.M. Patel,
W.S. Alwan, K. Jain, G.A. Hampannavar, Anti-Cancer Agents in Medicinal Chemistry 2015, 15,
1049.
[39] R. Rajagopalan, T.S. Lin, A.S. Karwa, A.R. Poreddy, B. Asmelash, R.B. Dorshow, ACS Med
Chem Lett 2012, 3, 284.
This article is protected by copyright. All rights reserved.

[40] H.J. Knolker, K.R. Reddy, Chem. Rev. 2002, 102, 4303.
[41] B.L. Staker, M.D. Feese, M. Cushman, Y. Pommier, D. Zembower, L. Stewart, A.B. Burgin, J.
Med. Chem. 2005, 48, 2336.
[42] E. Eagle, A.J. Carlson, J. Pharmacol. Exp. Ther. 1950, 99, 450.
[43] J. Suryadi, U. Bierbach, Chem. Eur. J. 2012, 18, 12926.
[44] L.S. Hollis, A.R. Amundsen, E.W. Stern, J. Med. Chem. 1989, 32, 128.
[45] G.Y. Park, J.J. Wilson, Y. Song, S.J. Lippard, Proc Natl Acad Sci U S A 2012, 109, 11987.
[46] B.R. Bowman, S. Lee, S. Wang, G.L. Verdine, Structure 2008, 16, 1166.
[47] L. Galluzzi, L. Senovilla, I. Vitale, J. Michels, I. Martins, O. Kepp, M. Castedo, G. Kroemer,
Oncogene 2012, 31, 1869.
[48] S.G. Chaney, A. Sancar, J. Natl. Cancer Inst. 1996, 88, 1346.
[49] T. Furuta, T. Ueda, G. Aune, A. Sarasin, K.H. Kraemer, Y. Pommier, Cancer Res. 2002, 62,
4899.
[50] A.F. Fagbemi, B. Orelli, O.D. Scharer, DNA Repair (Amst) 2011, 10, 722.
[51] J.A. Marteijn, H. Lans, W. Vermeulen, J.H. Hoeijmakers, Nat. Rev. Mol. Cell Biol. 2014, 15,
465.
[52] C. Kisker, J. Kuper, B. Van Houten, Cold Spring Harb Perspect Biol 2013, 5, a012591.
[53] S.F. Bellon, J.H. Coleman, S.J. Lippard, Biochemistry 1991, 30, 8026.
[54] R. Visse, A.J. van Gool, G.F. Moolenaar, M. de Ruijter, P. van de Putte, Biochemistry 1994, 33,
1804.
[55] T.C. Johnstone, J.J. Wilson, S.J. Lippard, Inorg. Chem. 2013, 52, 12234.
[56] J. van Meerloo, G.J. Kaspers, J. Cloos, Methods Mol Biol 2011, 731, 237.
[57] J. Suryadi, U. Bierbach, Chem. Eur. J. 2012, 18, 12926
This article is protected by copyright. All rights reserved.

FIGURES
Figure 1. Cisplatin and monofunctional platinum-based anticancer agents.
This article is protected by copyright. All rights reserved.

Figure 2. Structure of carbazoleplatin-G adduct determined by human DNA polymerase β host-guest
complex (pol β-HGC) system. (A) The X-ray structure of pol β-HGC with carbazoleplatin-DNA
adduct showing both the platination and the intercalation of carbazoleplatin (white: pol β, teal:
template, blue: upstream primer, violet: downstream primer, magenta: carbazoleplatin) (B) A top view
showing the platination of guanine N7 by carbazoleplatin. The electron density map (2Fo - Fc) was
contoured at 1.0 σ around carbazoleplatin and its neighboring DNA bases. (C) A side view showing
the intercalation of carbazoleplatin into the platinated G:C and the next G:C base pairs. The electron
density map (2Fo - Fc) was contoured at 1.0 σ around carbazoleplatin and its neighboring DNA bases.
This article is protected by copyright. All rights reserved.

Figure 3. UvrABC dual incision assay of the cisplatin 1,2-intrastrand crosslink adduct and the
carbazoleplatin adduct. 50-mer duplex DNA containing either a cisplatin-GG or a carbazoleplatin-G
adduct were used. Lane descriptions: (1) Unmodified 50-mer duplex DNA. (2) Cisplatin 1,2-
intrastrand crosslink-containing duplex DNA in the absence of UvrABC. (3) Cisplatin 1,2-intrastrand
crosslink-containing duplex DNA in the presence of UvrABC. (4) Unmodified duplex DNA. (5)
Carbazoleplatin-containing duplex DNA in the absence of UvrABC. (6) Carbazoleplatin-containing
DNA in the presence of UvrABC.
This article is protected by copyright. All rights reserved.

SCHEMES
Scheme 1. Synthesis and Single Crystal X-ray Structure of Carbazoleplatinα
α(A) Preparation of carbazoleplatin (i) KOH, 2-bromoethanol, acetone, reflux, dark, 32%; (ii) I2,
imidazole, PPh3, THF; (iii) NaN3, DMF, 70ºC, 58% (over 2 steps); (iv) PPh3, H2O, THF, 58%; (v) the
reaction product of cis-Diamminedichloridoplatinum(II) and AgNO3, DMF, 55ºC, dark, 23%. (B)
Single crystal X-ray structure of carbazoleplatin and its nitrate counter ion.
This article is protected by copyright. All rights reserved.

TABLES
Table 1. Crystallographic Data for Carbazoleplatin
chemical formula
formula weight
temperature
C14 H20 Cl N5 O3 Pt
536.88
100(2) K
wavelength
1.54184 Å
monoclinic
P21/n
crystal system
space group
unit cell dimensions
a = 17.6087(3) Å
b = 5.45090(10) Å
c = 36.3822(5) Å
volume
α = 90°
β = 96.5600(10)°
γ = 90°
3
3469.22(10) Å
8
Z
density (calculated)
3
2.071 Mg/m
absorption coefficient
-1
16.780 mm
F(000)
2080
crystal size
0.180 × 0.110 x 0.040
3
mm
theta range for data
collection
2.445 to 74.271°.
index ranges
–21 ≤ h ≤ 21
–6 ≤ k ≤4
–45 ≤ l ≤ 45
reflections collected
18275
independent reflections
6857 [R(int) = 0.0242]
This article is protected by copyright. All rights reserved.

completeness to theta =
67.684°
99.5 %
absorption correction
Semi-empirical from
equivalents
max. and min.
transmission
1.00 and 0.585
refinement method
Full-matrix least-squares
2
on F
data / restraints /
parameters
6857 / 566 / 476
1.099
2
goodness-of-fit on F
final R indices
[I>2sigma(I)]
R1 = 0.0328, wR2 =
0.0680
R indices (all data)
R1 = 0.0371, wR2 =
0.0690
extinction coefficient
n/a
largest diff. peak and
hole
-3
1.358 and -1.601 e.Å
This article is protected by copyright. All rights reserved.

Table 2. Data Collection and Refinement Statistics for pol β-carbazoleplatin Host-Guest Complex
5HHH
G:C
5HHI
PDB code
Carbazoleplatin
-G:C
Data Collection
Space Group
P21
P21
Cell constants
a (Å)
51.334
84.595
55.523
90.00
54.372
b
80.119
c
54.674
α (º)
90.00
β
γ
107.494
90.00
108.942
90.00
Resolution (Å)a 20-2.36
(2.41-2.36)
0.080
20-2.52
(2.56-2.52)
0.075
Rmergeb (%)
(0.359)
(0.380)
22.0 (2.50)
99.3 (93.4)
<I/σ>
20.7 (2.73)
100 (99.9)
Completeness
(%)
Redundancy
Refinement
3.7 (3.7)
4.8 (4.3)
Rworkc/Rfreed (%) 21.2/26.2
22.4/27.8
14873
Unique
18516
44.7
reflections
Mean B factor
(Å2)
Protein
53.5
This article is protected by copyright. All rights reserved.

Ligand
Solvent
37.5
40.8
64.3
43.1
Ramachandran
plot
Most
(%)
favored 96.6
94.1
5.6
Additional
3.4
allowed (%)
RMSD
Bond
(Å)
lengths 0.005
0.005
0.680
Bond angles (º) 0.710
a Values in parentheses are for the highest resolution shell.
bRmerge = Σ|I-<I>|/ΣI where I is the integrated intensity of a given reflection.
cRwork = Σ|F(obs)-F(calc)|/ ΣF(obs).
dRfree = Σ|F(obs)-F(calc)|/ ΣF(obs), calculated using 5% of the data.
This article is protected by copyright. All rights reserved.

Table 3. DNA Torsion Analysis of the Carbazoleplatin-G Structure
Distortion
parameter
Control
DNA
(5HHH) (º) (5HHI) (º)
Carbazolep Difference
latin-DNA
(º)
Shear
Stretch
Stagger
Buckle
Propeller
Opening
Shift
–0.41
–0.01
0.24
–0.05
–0.15
0.09
0.36
0.14
0.15
4.66
0.05
0.96
0.63
0.63
0.16
4.53
0.08
1.80
1.02
–4.05
–7.1
0.61
–7.05
–2.15
–0.28
0.28
–1.19
0.35
Slide
–0.35
3.19
Rise
3.03
Tilt
1.52
–3.01
4.19
Roll
4.27
Twist
34.13
–1.3
32.33
–0.28
x-
Displaceme
nt
y-
–0.34
–0.03
0.31
Displaceme
nt
Helical rise 3.18
Inclination 7.27
2.96
7.57
5.51
33.08
0.22
0.30
8.08
1.71
Tip
–2.57
Twist
34.79
This article is protected by copyright. All rights reserved.

Table 4. Cytotoxicity Profile of Carbazoleplatin and Cisplatin
Cell line
IC50 (µM)a
(Cancer type)
PANC-1
Carbazoleplatin Cisplatin
5.3
9.3
43.3
47.5
(Pnacreas)
MDA-MB-231
(Breast)
aEach IC50 was calculated by averaging 4 replicates.
SUPPORTING INFORMATION
Spectral data, X-ray crystallography tables, and MTT cytotoxicity graphs are included the
supporting information.
GRAPHICAL TABLE OF CONTENTS
We designed and synthesized a monofunctional platinum-
carbazole conjugate (carbazoleplatin). The crystal structure
of carbazoleplatin-DNA adduct showed that carbazoleplatin
modified N7 of a guanine (G) base and intercalated into two
G:C base pairs without significantly distorting the double
helix. UvrABC removed the carbazoleplatin-containing DNA
less efficiently than the cisplatin 1,2-intrastrand crosslink-
containing DNA. Carbazoleplatin was several-fold more
cytotoxic than cisplatin in PANC-1 and MDA-MB-231 cell lines.
This article is protected by copyright. All rights reserved.