Why does the type of halogen atom matter for the radiosensitizing properties of 5-halogen substituted 4-thio-20-deoxyuridines?

Article abstract of DOI:10.3390/molecules24152819

Radiosensitizing properties of substituted uridines are of great importance for radiotherapy. Very recently, we confirmed 5-iodo-4-thio-2⁰-deoxyuridine (ISdU) as an efficient agent, increasing the extent of tumor cell killing with ionizing radiation. To our surprise, a similar derivative of 4-thio-2’-deoxyuridine, 5-bromo-4-thio-2⁰-deoxyuridine (BrSdU), does not show radiosensitizing properties at all. In order to explain this remarkable difference, we carried out a radiolytic (stationary and pulse) and quantum chemical studies, which allowed the pathways to all radioproducts to be rationalized. In contrast to ISdU solutions, where radiolysis leads to 4-thio-2’-deoxyuridine and its dimer, no dissociative electron attachment (DEA) products were observed for BrSdU. This observation seems to explain the lack of radiosensitizing properties of BrSdU since the efficient formation of the uridine-5-yl radical, induced by electron attachment to the modified nucleoside, is suggested to be an indispensable attribute of radiosensitizing uridines. A larger activation barrier for DEA in BrSdU, as compared to ISdU, is probably responsible for the closure of DEA channel in the former system. Indeed, besides DEA, the XSdU anions may undergo competitive protonation, which makes the release of X^? kinetically forbidden.

Full text of DOI:10.3390/molecules24152819

molecules
Article
Why Does the Type of Halogen Atom Matter for the
Radiosensitizing Properties of 5-Halogen Substituted
4-Thio-20-Deoxyuridines?
Paulina Spisz 1 , Magdalena Zdrowowicz 1 , Samanta Makurat 1, Witold Kozak 1
,
Konrad Skotnicki 2, Krzysztof Bobrowski 2 and Janusz Rak 1,
*
1
Laboratory of Biological Sensitizers, Faculty of Chemistry, University of Gdan´sk, Wita Stwosza 63,
80-308 Gdan´sk, Poland
Centre of Radiation Research and Technology, Institute of Nuclear Chemistry and Technology, Dorodna 16,
03-195 Warsaw, Poland
2
*
Correspondence: janusz.rak@ug.edu.pl
ꢀꢁꢂꢀꢃꢄꢅꢆꢇ
Academic Editor: Chryssostomos Chatgilialoglu
Received: 17 July 2019; Accepted: 31 July 2019; Published: 2 August 2019
ꢀꢁꢂꢃꢄꢅꢆ
Abstract: Radiosensitizing properties of substituted uridines are of great importance for radiotherapy.
Very recently, we confirmed 5-iodo-4-thio-20-deoxyuridine (ISdU) as an efficient agent, increasing
the extent of tumor cell killing with ionizing radiation. To our surprise, a similar derivative of
4-thio-2’-deoxyuridine, 5-bromo-4-thio-20-deoxyuridine (BrSdU), does not show radiosensitizing
properties at all. In order to explain this remarkable difference, we carried out a radiolytic (stationary
and pulse) and quantum chemical studies, which allowed the pathways to all radioproducts to be
rationalized. In contrast to ISdU solutions, where radiolysis leads to 4-thio-2’-deoxyuridine and its
dimer, no dissociative electron attachment (DEA) products were observed for BrSdU. This observation
seems to explain the lack of radiosensitizing properties of BrSdU since the efficient formation of the
uridine-5-yl radical, induced by electron attachment to the modified nucleoside, is suggested to be an
indispensable attribute of radiosensitizing uridines. A larger activation barrier for DEA in BrSdU,
as compared to ISdU, is probably responsible for the closure of DEA channel in the former system.
Indeed, besides DEA, the XSdU anions may undergo competitive protonation, which makes the
release of X− kinetically forbidden.
Keywords: radiosensitizers; stationary radiolysis; pulse radiolysis; modified nucleosides;
cellular response
1. Introduction
Trojan horse radiotherapy employs a nucleoside radiosensitizer, a “Trojan horse”, that is activated
only due to DNA exposure to ionizing radiation [1]. Such radiosensitizers are usually electrophilic
nucleosides, incorporated into DNA during replication and repair, and undergoing efficient dissociative
electron attachment (DEA) that leaves behind a nucleoside radical, which in secondary reactions
is able to produce damage to the biopolymer (frequently a strand break) [
derivatives of nucleosides were proposed as potential radiosensitizers [
examples comprise uridines substituted at the C5 position. This is because thymidine kinase accepts a
broad set of modified uridines [ ], which after phosphorylation may be incorporated into DNA [ ].
The modified DNA sensitivity to hydrated electron attachment is especially important for radiotherapy,
since cells of solid tumors (80% of cases [ ]) are hypoxic, which make them resistant to hydroxyl
radicals, a major damaging agent of native DNA produced during radiotherapy [ ]. In the normoxic
cells, damage produced by the •OH radicals becomes “fixed” due to reaction with oxygen, while under
1]. Although the purine
2–5], most of the reported
6
7
8
9
Molecules 2019, 24, 2819; doi:10.3390/molecules24152819
www.mdpi.com/journal/molecules

Molecules 2019, 24, 2819
2 of 17
hypoxia, naturally occurring radioprotectors like cysteine or glutathione can restore DNA through
hydrogen donation [10]. Moreover, hydrated electrons, the second most abundant product of water
radiolysis, are not harmful to the natural DNA [11,12]. The situation becomes quite different when
DNA is labeled with nucleosides undergoing efficient DEA. Indeed, Sanche et al. [13] demonstrated
efficient formation of single strand breaks in DNA oligonucleotides labeled with BrdU, when an
aqueous solutions of these biopolymers were irradiated with X-rays in the presence of a hydroxyl
radical scavenger. Similarly, a radiolysis of a solution containing TXT oligonucleotides (where X stands
for 5-bromo-20-deoxyuridine (BrdU), 5-iodo-20-deoxyuridine (IdU), 5-bromo-20-deoxycitidine (BrdC),
5-iodo-20-deoxycitidine (IdC), 8-bromo-20-deoxyadenosine (BrdA), or 8-bromo-20-deoxyguanosine
(BrdG)) and t-butyl alcohol (t-BuOH) as •OH scavenger led to strand breaks besides other types of
DNA damage [11,12].
BrdU and IdU are well-known radiosensitizers, which are phosphorylated in cytoplasm
forming the respective 50-triphosphates and then incorporated into the cellular DNA by human
DNA polymerases [
7
]. Their promising radiosensitizing properties were investigated in numerous
17] studies and even in clinical trials [18]. In one of the most extensive
in vitro 14 16] and in vivo [
[
–
clinical studies on brain tumor patients, no positive effects were observed in patients exposed to the
specific doses of BrdU besides radiotherapy [18]. To this end, it is worth emphasizing that a swift and
efficient metabolism (most radiosensitizers, as other chemotherapeutics, are applied systemically) of
a sensitizer, may lead to its lower cellular concentration in vivo than in vitro. This, at least partially,
explains a high radiosensitizing activity of BrdU in vitro, and practically, the lack of such activity in
the clinical studies.
This situation calls for new radiosensitizers of superior pharmacokinetic and/or better
radiosensitizing properties. Recently, we proposed several new C5-pyrimidine derivatives that have
not been studied in animal models or in clinic to date [19–23]. In terms of electron-induced degradation
yields, they are all more prone to dissociative electron attachment (DEA) than BrdU. These compounds
comprise 5-thiocyanato-20-deoxyuridine (SCNdU), 5-selenocyanato-20-deoxyuridine (SeCNdU),
5-selenocyanatouracil (SeCNU), and 5-trifluoromethanesulfonyl-20-deoxyuridine (OTfdU). Other
promising candidates of this type of radiosensitizers seem to be derivatives of 4-thio-20-deoxyuridine.
The latter compound, similarly to BrdU and IdU, is incorporated into genomic DNA by the cellular
enzymatic machinery [24]. It is worthwhile to note that BrdU works both as a DNA radio- and
photosensitizer [
1]. On the other hand, the photosensitizing properties of BrSdU and ISdU were proven
in the past [25
,
26]. By the same token, one may conclude that 5-haloderivatives of 4-thio-20-deoxyuridine
could also work as radiosensitizers. Indeed, we recently demonstrated the radiosensitizing properties
of ISdU [27]. Under the same conditions, the yield of damage produced by 140 Gy of X-ray was 1.5-fold
larger than that assayed in the irradiated BrdU aqueous solutions. Simultaneously, in vitro studies
demonstrated a significant increase of the mortality in cells treated with ISdU after irradiation.
In the current paper, studies on BrSdU–similar to those shown in [27] on ISdU–are described.
To our surprise, BrSdU does not possess increased radiosensitizing properties. It is decomposed during
radiolysis by X-ray, but the stable products resulted only from the reactions between the compound
studied and H2O2 or radicals forming in the reaction between t-BuOH and the •OH radicals. We did
not observe the characteristic pattern of DEA, i.e., the formation of 4-thio-20-deoxyuridine, in this case.
In accordance with this finding, the clonogenic assay does not differentiate the cells that were grown
with and without BrSdU. We explain this striking difference between ISdU and BrSdU with the height
of activation barrier for DEA, which is almost twice as much as in BrSdU.
2. Results and Discussion
A radiosensitizing nucleoside working under hypoxia must be sensitive to hydrated electrons,
which are the second most abundant product of water radiolysis. In order to assess the radiosensitizing
potential of a nucleoside, one must expose its aqueous, deoxygenated solution to ionizing radiation.
If hydroxyl radicals are scavenged during irradiation, only the reaction between hydrated electrons

Molecules 2019, 24, 2819
3 of 17
and potential radiosensitizer may lead to serious damage associated with the formation of radical
products. If the radiolysis proceeds in the cells containing DNA labeled with radiosensitizer, those
radicals will produce DNA damage, hopefully single/double strand breaks, leading to apoptosis as
an ultimate cellular response. This is why the radiolysis of a nucleoside in aqueous solution and the
qualitative and quantitative analysis of radiolytic products is an indispensable step for assessing the
radiosensitizing potential of the derivative under investigations.
2.1. Stationary Radiolysis
The high-performance liquid chromatography (HPLC) traces of radiolytes, originating from X-ray
irradiation of buffered aqueous solution, containing 10−4 M of BrSdU in the presence of a hydroxyl
radical scavenger (t-BuOH, 0.03 M) with the dose of 140 Gy, are depicted in Figure 1.
Figure 1. High-performance liquid chromatography (HPLC) traces for a solution of BrSdU before
(black) and after irradiation (red). The chemical structures of products, as indicated by the liquid
chromatography–mass spectrometry (LC–MS/MS) analysis, are depicted at particular peaks.
As indicated by the comparison of two chromatograms (Figure 1), three main products are
formed due to radiolysis. The liquid chromatography–mass spectrometry (LC–MS) analysis enabled
the identification of all these species. Figure 1 also depicts the chemical structures of the identified
decomposition products, while the MS/MS spectra (shown in Figures S1–S4 in Supplementary Materials)
confirm the assignment of particular structures. These radiolysis products are a dimer and two oxidation
products, whose probable mechanism of formation was modelled for 5-bromo-1-methyl-4-thiouracil
(see the Computational section) at the B3LYP(PCM)/DGDZVP++ level and is shown in Figure 2.
The (BrSU)2 dimer is suggested to be the product of two BrSU• radicals recombination (Figure 2A),
which are created in the reaction of BrSU with •CH2(CH3)2COH (the •CH2(CH3)2COH radicals are
formed in the reaction between t-BuOH, present in the solution as a hydroxyl radical scavenger, and the
•OH radicals–a primary product of water radiolysis). The reaction is associated with a kinetic barrier
of 76.1 kJ/mol and is favorable thermodynamically (Figure 2A). The second reaction, leading to BrSOU
(Figure 2B) due to oxidation of BrSU by H2O2 [28–30] (H2O2 is produced during water radiolysis [31]),
was modelled with three explicitly added water molecules, which is the approach that had been
suggested in the literature [32]. It is worthwhile to note that under the experimental conditions, BrSOU

Molecules 2019, 24, 2819
4 of 17
(BrSOdU) appears to be the least abundant product, which probably results from the fact that it is also
the substrate for the most abundant one, i.e., for BrU (Figure 2C). The latter is formed via a cyclic
oxathiirane followed by the sulfur extrusion reaction [33,34]. In our calculations, we were unable to
obtain the stable oxathiirane intermediate. During the optimization, the ring opened to give the BrOSU
structure shown in Figure 2C. Also the sulfur extrusion does not show any intermediates. After the TS
structure is achieved (4.4 kJ/mol barrier), one of the sulfur atoms attaches to the other one and the S-O
bond breaks leading to BrU and S2 (Figure 2).
Figure 2. The radiolysis products. (A) (BrSU)2, (B) BrSOU, and (C) BrU formation, as suggested by
calculations. The optimized reactants in ball and stick representation are shown along with their
kinetic barriers (marked with asterisks) and thermodynamic stimulus (kJ/mol). All the reactions shown
correspondence to the most favorable pathways, as obtained by the IRC procedure. The transition
states structures can be found in Supplementary Materials (Figure S5).
To our surprise, one of the expected products, 4-thio-20-deoxyuridine, that should form due to
DEA to BrSdU, was not detected in the BrSdU radiolytes (see Figure 1). DEA, occurring in many
similar systems including BrdU, IdU, and ISdU, is thought to be the main reason of DNA damage in the
irradiated cells, i.e., it is responsible for the radiosensitizing potential of modified nucleosides [12,35–37].
However, neither 4-thio-20-deoxyuridine nor dimer with the substrate (both were observed in radiolytes
of ISdU) were observed among the radiolysis products.
In order to explain why the radiolysis of ISdU leads to SdU, while this reaction channel is
actually closed for BrSdU under the same experimental conditions, we calculated the respective
DEA profile (Figure 3). We found that the kinetic barrier for the C5-Br bond had breakage as much
as 26.0 kJ/mol–more than two times higher than that for C5-I in ISU (12.6 kJ/mol [27]) dissociation
calculated at the same level of theory. Similarly, the thermodynamic stimulus for the release of the
bromide anion from BrSU•− amounts to only
These differences, especially a significant difference in the height of activation barriers in the two
compared systems, explains formation of SdU only in the latter derivative. The estimated lifetime
of BrSdU•− at the ambient temperature is ca. 200-fold longer than that of ISU•, which results from
−
10.5 kJ/mol as compared to [27].
24.3 kJ/mol for ISU•−
−

Molecules 2019, 24, 2819
5 of 17
the above-mentioned activation barriers and transition state theory [38]. It is probably sufficiently
long to allow the anion to be protonated, which prevents completion of the DEA process [39], and in
consequence, SdU is not formed.
Figure 3. Calculated dissociative electron attachment (DEA) profile for BrSU. After the initial electron
attachment to BrSU, the anion radical BrSU•− is formed, and subsequently dissociates via a transition
state giving SU• and Br−. The thermodynamic and kinetic characteristics [kJ/mol], shown in green,
were calculated as the difference between the given state and the previous stable one, the transition
state barrier marked with asterisk.
2.2. Pulse Radiolysis
The hypothesis, explaining different behavior of BrSdU and ISdU based on the computational
results and discussed in the previous section, is confirmed by the results of our pulse radiolysis
experiments. Figures 4 and 5 depict the transient spectra, as well as the respective decays and growths
in microsecond-time domain for ISdU and BrSdU, respectively.
Figure 4.
(10 mM, pH = 7.0) ISdU solution (5
(
A
) Transient absorption spectra recorded in deoxygenated and buffered with phosphate
10−5 M), in the presence of 0.5 M t-BuOH, after 2
s ( ) and 120
) Short-time profiles representing the growth at = 305 nm ( ) and the
) of transient absorptions and their least-square fits to the first order formation
·
µ
ꢀ
µs
(
#
) after the electron pulse. (
B
λ
ꢀ
decay at
λ
= 720 nm (
#
and decay, respectively. (
C
) Long-time profiles representing the growth at
λ
= 320 nm (
#
) and decays
at
λ
= 300 nm (
4) and
λ
= 385 nm (
ꢀ) of transient absorptions and their least-square fits to the first
order formation and decays, respectively.
The resulting transient spectrum, obtained 2 µs after the electron pulse, exhibits a rising absorption
toward 300 nm with no defined maximum and negative absorption in the wavelength range 320–380 nm
(Figure 4A). This time delay for spectra recording was chosen on purpose in order to get rid of
participation of hydrated electrons in the spectrum. Nonetheless, the registration of “pure” spectrum
of this transient product was not possible due to the bleaching related to the consumption of ISdU.
The decay at
with the pseudo-first order rate constant k720 = 1.8
the formation of a transient product with the pseudo-first order rate constant
(Figure 4B). Since these rate constants are very similar, this species could be a direct product of the
hydrated electron attachment to ISdU if the lifetime of ISdU•− was long enough and included the
microsecond-time domain. However, as indicated by the B3LYP/DGDZVP++ barrier height, its lifetime
λ
= 720 nm represents the decay of hydrated electrons in the presence of 5
·
10−5 M ISdU
·
106 s−1. In turn, the growth at
λ = 300 nm represents
k
300 = 2.0 · 106 s−1

Molecules 2019, 24, 2819
6 of 17
should be very short and is rather in the nanosecond-time domain (see the previous section). Therefore,
we probably observed the product of DEA to ISdU, i.e., the SdU• radical formed via Reaction (1).
e−aq + ISdU → ISdU•− → SdU•+ I−
(1)
Indeed, the calculated UV spectrum for SdU• is characterized by
λmax located at 300 nm. With the
time elapsed, the absorption spectrum underwent further changes and 120
is characterized by a transient absorption band with max = 320 nm, which can be assigned to a new
product (Figure 4A). The growth at = 320 nm is mono-exponential and occurs with the pseudo-first
order rate constant k320 = 7.0 = 300 nm and 385 nm are
104 s−1. Interestingly, the decays at
104 s−1 and
µs after the electron pulse
λ
λ
·
λ
also mono-exponential, with the respective pseudo-first order rate constants k300 = 8.2
·
k385 = 7.1 · 104 s−1, which are reasonably close to k320 (Figure 4C). Since these decays seem to represent
the decay of the SdU• radical, the growth observed at
λ = 320 nm can be tentatively assigned to the
formation of SdU via reaction of SdU• with hydrogen atom donor, which is t-BuOH present in the
system in a large excess (Reaction 2).
SdU•+ t-BuOH → SdU + •CH2(CH3)2COH
(2)
The calculated (and measured) UV spectrum of 4-thiouracil strongly supports this assignment.
Moreover, the absorption at = 320 nm is stable in the time window of our experiment (up to
λ
1.5 ms) with no signs of disappearance, which might suggest, however not directly, a high stability of
the product.
Figure 5. Transient absorption spectra recorded in deoxygenated and buffered with phosphate (10 mM,
pH = 7.0) BrSdU solution (5
·
10−5 M), in the presence of 0.5 M t-BuOH (
A
) after 12
) Short-time profiles representing the growth at
) of transient absorptions and their least-square fits to
) Long-time profiles representing the growth at
µs (ꢀ) and 120 µs
(
#
) and (
B
) (
ꢀ
) 12
µ
s after the electron pulse. (
C
λ = 300 nm (#) and the decay at
λ = 720 nm (ꢀ
the first order formation and decay, respectively. (
D
λ = 335 nm (ꢀ) and decays at λ = 300 nm (#) of transient absorptions and their least-square fits to the
second order formation and decays, respectively.

Molecules 2019, 24, 2819
7 of 17
The main difference between spectral features observed during pulse radiolysis of ISdU and
BrSdU aqueous solutions is the lack of the transient absorption band with = 320 nm for the latter
λ
max
system (compare Figure 5A with Figure 4A). This finding clearly shows that the formation of SdU does
not occur in BrSdU via analogous reaction 2, which requires the presence of SdU• radical. This fact
remains in a good accordance with the results of stationary γ-radiolysis, where we did not observe
the product of the bromide anion (Br−) release from BrSdU•−. In case of ISdU, we also observed an
increase in the integrated XIC signal of the iodide anions (a background signal of iodide anions is
observed due to synthesis-related contamination, unavoidable degradation of the sample in an aqueous
solution, as well as fragmentation of the studied nucleosides in the MS source [40,41]) due to stationary
radiolysis. Indeed, the ratio of the signals after and before irradiation is equal to 2.041 ± 0.016. When
it comes to the BrSdU sample, the integrated XIC signal of Br− anions is the same before and after
irradiation, 0.987 ± 0.0791, within the error bar. Using the calibration curve (not shown), we found out
that the measured increase in the concentration of I− due to irradiation means that ca. 27% of ISdU
decay occurs in the DEA pathway. Thus, the observed changes in the concentration of halogen anions
support that DEA process is operative only in ISdU solutions. This fact can be rationalized by the
slower dissociation of BrSdU•− (Reaction 3) in comparison to dissociation of ISdU•−, which allows
BrSdU•− to be involved in another competitive process, for instance, its protonation by water or
phosphate anions (Reaction 4):
BrSdU•− → SdU•+ Br−
(3)
H2O, H2PO−
4
BrSdU•−
→
BrSdUH•
(4)
The resulting absorption spectrum, obtained 12
µs after the electron pulse, exhibits a rising
absorption toward 300 nm with no defined maximum and negative absorption in the wavelength range
320–380 nm (Figure 5A). This time delay for spectra recording was again chosen on purpose in order to
get rid of participation of the most hydrated electrons in the spectrum. Since the absorption spectrum
recorded 120
µs after the electron pulse does not exhibit the absorption band with
λmax = 320 nm (in
contrast to ISdU), the absorption spectrum recorded 2
µ
s after the electron pulse cannot consequently
be assigned to SdU• radical. Moreover, it is worthwhile to note that the pseudo-first rate constant
of the decay of hydrated electrons in the presence of 5
·
10−5 M BrSdU (
k
720 = 1.4 · 106 s−1) is nearly
two-fold lower than the pseudo-first order rate constant of the formation of the transient measured
at λ = 300 nm (k300 = 2.4
106 s−1) (Figure 5C). This observation suggests that this transient product
·
cannot be formed in an analogous Reaction (2), as observed for ISdU, but might result from the
protonation of BrSdU•− (Reaction 4) (vide supra). Therefore, the absorption spectrum recorded after
12
possesses the absorption bands with
µ
s can be tentatively assigned to BrSdUH•. The UV-VIS spectrum of BrSdUH• (calculated by us)
λ
max
= 280, 350, and 480 nm. Therefore, for better visualization of
the experimental spectrum assigned by us to BrSdUH• radical, the spectrum recorded 12 µs after the
pulse was zoomed to see expected spectral features (Figure 5B). The transient absorption depicted in
Figure 5B shows the maxima below 300 nm and at 450 nm. The maximum at 350 nm is concealed by
the “negative” absorption of BrSdU, but a tail at 380 nm is quite clear.
Interestingly, the decay at
can be fitted by the second-order kinetics and occur with the respective second-order rate constants
2k300 = 1.2
109 M−1s−1 and 2k335 = 1.0 109 M−1s−1, which are very similar (Figure 5D). One of the
λ = 300 nm and the formation at λ = 335 nm within 1.5 ms time domain
·
·
processes, which can be described by the second-order kinetics is disproportionation reaction involving
BrSdUH• radicals, should lead to the partial recovery of the substrate (BrdSU) (Reaction 5).
BrSdUH• + BrSdUH•
→
BrSdU + BrSdUH2
(5)
Thus, the decay observed at
observed at
λ
= 300 nm represents the decay of BrSdUH• radicals and the growth
λ
= 335 nm represents recovery of BrSdU substrate. This mechanism can, at least in part,
explain lower consumption of BrSdU compared to ISdU using the same dose delivered by X-rays.

Molecules 2019, 24, 2819
8 of 17
2.3. Biological Assessments
2.3.1. Incorporation of BrSdU and ISdU into Genomic DNA
According to the concept of the Trojan horse therapy, a radiosensitizer should easily incorporate
into DNA. For this reason the incorporation of BrSdU and ISdU into genomic DNA was assessed.
The MCF-7 cells treated with BrSdU and ISdU at the concentration of 10−4 M were incubated for 48 h.
Purified DNA was enzymatically digested and analyzed by HPLC (Figure S6) and LC-MS method.
The results of LC-MS analysis shows that both derivatives incorporate to DNA (see extracted-ion
chromatograms, MS and MS/MS spectra for BrSdU/ISdU in Figures S7–S10 in Supplementary Materials),
but the efficiency of this process is very low, which has been already observed by others [24]. To our
surprise, BrdU and IdU beside BrSdU and ISdU were also observed in the digested material, which
suggests an enzymatic degradation of BrSdU/ISdU to BrdU/IdU in the cell with the rate similar to that
of BrSdU/ISdU incorporation into DNA.
2.3.2. Clonogenic Assay
In order to determine the simultaneous effect of BrSdU and ionizing radiation on the survival
and proliferation of cancer cells, the clonogenic assay (based on the ability of a single cell to grow into
a colony) [42] was performed. The test was carried out on human breast cancer cells (MCF-7 line)
treated with BrSdU at the concentration of 0, 10, and 100 µM and/or ionizing radiation (IR) in four
doses of 0.5, 1, 2, and 3 Gy. Figure 6 shows that the studied compound does not affect the survival
of cancer cells. We only observed reduction of survival fraction caused by IR. For example, in case
of dose of 2 Gy, survival was 29.9
±
4.6%, 32.7
±
6.3%, and 28.2
±
2.8% for 0, 10, and 100 µM BrSdU
pretreatment, respectively. This colony formation assay demonstrates that the BrSdU does not sensitize
the MCF-7 cells to X-ray. In our previous studies, we demonstrated a significant radiosensitizing effect
of ISdU [27]. The addition of the latter derivative to cell culture resulted in a significant decrease (about
20%) of their survival after irradiation, even with doses as low as 0.5 Gy.
100
0 µM
90
10 µM
80
70
60
50
40
30
20
10
0
100 µM
0
1
2
3
dose [Gy]
Figure 6. Dose response curves of MCF-7 cells treated with (10
µM or 100
µM solutions of BrSdU) or
without BrSdU. The average plating efficiencies for the controls with and without pretreatment are
equal to 28.16% (0 M), 25.63% (10 M), and 24.78% (100 M). Experiments were performed at least in
two independent experiments in duplicate and the results are expressed as mean standard deviation.
µ
µ
µ
±

Molecules 2019, 24, 2819
9 of 17
2.3.3. Cytotoxicity Assay
One of the properties that good radiosensitizer should possess is low cytotoxicity without IR
treatment. To identify cytotoxicity of BrSdU toward MCF-7 (cancer cells) and HDFa (normal cells) line,
MTT assay [43] was carried out (Figure 7). BrSdU was tested at six concentrations (0, 10−8, 10−7, 10−6
,
10−5, 10−4, 5
significant reduction of viability up to 93% (for 48 h incubation) for MCF-7 line only in case of the
highest tested concentration equal to 5
10−4 M. For lower concentrations, the decrease in vitality was
·
10−4M) and two time variants (24 and 48 h of incubation). Figure 7Ib shows statistically
·
not statistically significant. We also did not observe a meaningful difference between the studied cell
lines. These results show that the cytotoxicity of BrSdU is very low both for normal human dermal
fibroblasts and human breast cancer cells.
Figure 7. The viability of MCF-7 (
in a range of concentrations from 0 to 5
experiments performed in triplicate. *statistically significant difference is present between treated
I
) and HDFa (II) cells after 24 (
a
) and 48 h treatment (
b) with BrSdU
·
10−4 M. Results are shown as mean
±
SD of three independent
culture compared with control (untreated culture).
2.3.4. Analysis of Histone H2A.X Phosphorylation and Cell Death
One of the most common types of DNA damage related to radiosensitization is double-strand
breaks formation. Phosphorylation of histone γH2A.X is the marker of such a damage [44]. The assay
was performed for human breast cancer cells treated with BrSdU at concentration of 10−4 M and/or
irradiated with a dose of 0, 1, or 2 Gy. Analysis of H2A.X phosphorylation was carried out 1 h after
irradiation (this time was optimized in previous experiments). The cells were fixed and analyzed by
flow cytometry. Our studies show that treatment with BrSdU results in a tiny increase in the population
of
BrSdU pretreatment and irradiation with the dose equal to 1 Gy, the level of
from 26.48% (nontreated cells) to 27.64 0.34%). The exposure of treated cells to the dose of 2 Gy
results in a slight enhancement of the γH2A.X fraction from 31.84 ± 3.74% to 36.15 ± 1.25%.
γ
H2A.X positive cells after irradiation with the doses of 1 and 2 Gy (Figure 8 and Figure S11). After
γ
H2A.X was changed
±

Molecules 2019, 24, 2819
10 of 17
Figure 8. Flow cytometric analysis of H2A.X phosphorylation.
irradiation. Results are shown as the mean standard deviation of at least two independent flow
cytometry experiments.
γH2A.X was measured 1 h after
±
Additionally, we performed the multidimensional test to quantify the number of viable,
early apoptotic, late apoptotic, and dead cells. Our results (Figures S12 and S13 in Supplementary
Materials) confirm that pretreatment with BrSdU does not affect the sensitivity of MCF-7 cells to
ionizing radiation. A significant influence of the studies analog on population of viable, early apoptotic,
late apoptotic, and dead cells was not observed. In the case of ISdU, the situation was completely
different. The histone H2A.X phosphorylation test showed that ISdU sensitized breast cancer cells to
ionizing radiation, at least in part, by formation of DSBs, while the cell death assay confirmed that
pretreatment of culture with ISdU led to the IR-induced reduction of cell viability and increase in the
population of early apoptotic cells [27].
3. Materials and Methods
3.1. Chemicals
5-bromo-20-deoxyuridine, acetic anhydride, P2S5, 1,4-dioxane, and sodium hydride were
commercially available from Sigma–Aldrich (Saint Louis, MO, USA). Nuclear magnetic resonance
(NMR) spectrum was recorded on a Bruker AVANCE III (Bruker, Billerica, MA, USA), 500 MHz
spectrometer. Chemical shifts are reported in ppm relative to the residual solvent peak (DMSO-d6
1
2.49 ppm for H and 39.5 ppm for 13C). Column chromatography was performed using silica gel
NORMASIL 60 (40–63 mesh, VWR Chemicals, Gdan´sk, Poland). Preparative thin-layer chromatography
was performed with silica gel plates, 60G, F254 (Sigma–Aldrich).
3.2. Synthesis of 30,50-di-O-acetyl-5-bromo-20-deoxyuridine
A solution of 5-bromo-20-deoxyuridine (500 mg, 1.63 mmol) in pyridine (6 mL) was stirred at room
temperature with acetic anhydride (339 µL, 3.59 mmol) for 24 h. The sirupus residue was co-evaporated
with three portions of aqueous ethanol (5 mL) and n-heptane to remove the pyridine residue. The raw
30,50-di-O-acetyl derivative (580 mg) was obtained in a 91% yield.
3.3. Synthesis of 30,50-di-O-acetyl-5-bromo-4-thio-20-deoxyuridine
30,50-di-O-acetyl-5-bromo-20-deoxyuridine (580 mg, 1.48 mmol) was dissolved in 1,4-dioxane
(20 mL) and P2S5 (990 mg, 4.45 mmol) was added. The mixture was refluxed until thin-layer
chromatography (TLC) analysis (CHCl3:CH3OH, 30:1) showed complete disappearance of the substrate
(2–3 h). Solvent was removed under reduced pressure and the residue was treated several times with
CHCl3. The combined chloroform extracts were evaporated, and the residue was separated on silica

Molecules 2019, 24, 2819
11 of 17
gel column, which was eluted with CHCl3:CH3OH, 20:1. After evaporation, the desired product was
obtained as a yellow solid (567 mg, 94%).
3.4. Synthesis of 5-bromo-4-thio-20-deoxyuridine
30,50-di-O-acetyl-5-bromo-4-thio-20-deoxyuridine (567 mg, 1.39 mmol) was dissolved in methanol
(10 mL) and stirred at 0 ◦C. A methanolic sodium methoxide solution (111 mg, 2.78 mmol),
freshly prepared from NaH and anhydrous methanol, was added in portions. The mixture was
stirred at room temperature until TLC analysis showed complete disappearance of the substrate
(10 min). The mixture was purified on silica gel column, which was eluted with CHCl3:CH3OH, 20:1.
The final product, 5-bromo-4-thio-20-deoxyuridine, was obtained as a yellow solid (185 mg, 41%).
1H NMR (Bruker AVANCE III, 500 MHz, DMSO),
δ
: 13.09 (s, 1H), 8.53 (s, 1H), 6.03 (t, 1H), 4.25 (q,
1H), 3.82 (q, 1H), 3.55–3.69 (m, 2H), 2.14–2.26 (m, 2H); 13C NMR (125 MHz, DMSO),
δ: 186.8, 147.6,
137.7, 107.1, 88.2, 86.1, 69.9, 60.8, 40.8. HRMS (TripleTOF 5600+, SCIEX), m/z: [M–H]− calculated for
C9H11BrN2O4S 323.1636, found 322.9918; UV spectrum (water), λmax: 345 nm.
3.5. Stationary Radiolysis
A mixture of BrSdU (10−4 M), 0.03 M t-BuOH (used as a scavenger of the •OH radicals), and
phosphate buffer (10 mM, pH = 7.0) was purging with argon for ca. 3 min in order to remove oxygen
from the solution. The dose absorbed by all samples during irradiation was 140 Gy (4.14 Gy ,
min−1
·
130.0 kV, 5.0 mA). The studied samples were analyzed in triplicate. Radiolysis was performed in a
Cellrad X-ray cabinet (Faxitron X-ray Corporation, Tucscon, AZ, USA).
3.5.1. HPLC Analysis
Irradiated and non-irradiated samples of BrSdU were analyzed with reversed-phase HPLC
method. For the separation of analytes a C18 column (Wakopak Handy ODS, 4.6
particle size and 100 Å in pore size), gradient elution with 80% ACN and 0.1% HCOOH (from 0 to
50% ACN in 30 min), flow rate 1 mL
min−1 were used. The HPLC analysis was performed on a
·
150 mm, 5 µm in
·
DionexUltiMate 3000 System (Dionex Corporation, Sunnyvale, CA, USA) with a Diode Array Detector
set at 260 nm.
3.5.2. LC-MS and LC-MS/MS Analysis
The solution of BrSdU containing t-BuOH and phosphate buffer was analyzed by LC-MS
and LC-MS/MS methods, before and after irradiation. Conditions of separations: Kinetex column
(Phenomenex, 1.7
µ
m, C18, 100 Å, 2.1
×
150 mm); flow rate 0.3 mL
min−1; a gradient elution with
·
80% ACN and 0.1% HCOOH (from 0 to 50% acetonitrile); the oven temperature was maintained at
25 ◦C. The effluent was diverted to waste for 2 min after injection. Conditions for MS and MS/MS
analysis: the spray voltage was 4.5 kV, the nebulizer gas (N2) pressure was 25 psi, the flow rate
−
was 11 L · min−1, and the source temperature was 300 ◦C. Each spectrum was obtained by averaging
three scans and the time of each scan was 0.25 s. The LC-MS and LC-MS/MS analysis was performed
TripleTOF 5600+ (SCIEX) mass spectrometer (operated in negative mode) coupled with Ultra-High
Performance Liquid Chromatography (UHPLC) system Nexera X2 (Shimadzu, Canby, OR, USA).
3.6. Pulse Radiolysis
Pulse radiolysis experiments were performed with the INCT LAE 10 MeV linear electron accelerator
with a typical pulse length of 8 ns. A detailed description of the experimental setup can be found in [45],
along with the basic details of the equipment and the data collection system. Absorbed doses per pulse
were on the order of 20 Gy (1 Gy = 1 J
kg−1). Dosimetry was based on the solutions saturated with
·
nitrous oxide, containing 10−2 M KSCN, taking a radiation chemical yield of G = 0.635 µmol · J−1 and
•−
a molar absorption coefficient of 7580 M−1
·
cm−1 at 472 nm for the (SCN)2 radical [45]. Experiments

Molecules 2019, 24, 2819
12 of 17
were performed with a continuous flow of sample solutions at room temperature (~23 ◦C). All solutions
were made with triply distilled water provided by a Millipore Direct-Q 3-UV system. The typical
concentration of BrSdU and ISdU was 5
10−5 M, and 0.5 M of t-BuOH was used as a hydroxyl radical
·
scavenger. Solutions were deoxygenated by purging with high purity argon, and 10 mM of phosphate
buffer was added to maintain pH = 7.
3.7. Clonogenic Assay
Adherent cell line MCF-7 (human breast cancer cells obtained from Cell Line Service–CLS,
Eppelheim, Germany), which was treated with BrSdU in concentration of 10−4 and 10−5, respectively,
◦
was plated on 60 mm dishes in a density of 106 cells per dish. After 48 h incubation under 37 C
and 5% CO2, the cells were exposed to 0.5, 1, 2, and 3 Gy, respectively (1.27 Gy
min−1, 130.0 kV,
·
5.0 mA). After 6 h, the cells were trypsinized and plated on 100 mm dishes in a density of 800 cells
per dish. After 16 days, formed colonies were fixed with 6.0% (v/v) glutaraldehyde and 0.5% crystal
violet. Stained colonies were counted manually, and colony size was rated using inverted fluorescence
microscope (Olympus, IX73, Tokyo, Japan). The cells were grown in the RPMI medium supplemented
with 10% FBS (fetal bovine serum) and with antibiotics (streptomycin and penicillin) at concentration of
100 U · mL−1. Irradiation has been performed in a Cellrad X-ray cabinet (Faxitron X-ray Corporation).
Plating efficiencies are shown in Table S1 in Supplementary Materials.
3.8. Cytotoxicity Assay
The MTT assay was used to identify the cytotoxic activity. Adherent MCF-7 and HDFa cell lines
were seeded into 96-well plate in a density of 4
103 per well and incubated under 37 ◦C and 5% CO2,
·
overnight. After that, the medium was replaced to fresh and the cells were treated with BrSdU at
concentration of 0 (control), 10−4, 10−5, 10−6, 10−7, 10−8 M. Plates with cells were incubated (under the
same conditions) with compound at 24 and 48 h. After this time, the aqueous solution of MTT salt
at concentration 4 mg
dimethylsulfoxide was added to each well in 200
·
mL−1 was added and incubate for 4 h. Then, the medium was removed and
L volume. The absorbance was measured at 570 nm
µ
(and 660 nm, reference wavelength). Absorbance measurement has been performed with use of EnSpire
microplate reader (PerkinElmer, Waltham, MA, USA). The liveliness of control was taken as 100%.
The results were analyzed with the use of GraphPad Prism software. The statistical evaluation of treated
samples and untreated control was calculated using one-way analysis of variance (ANOVA) followed
by Dunnett’s multiple comparison test. The data was obtained from three independent experiments
and each treatment condition assayed in triplicate. The differences were considered significant at
p < 0.05. The cells were grown in the RPMI (MCF-7)/DMEM (HDFa) medium supplemented with
10% FBS (fetal bovine serum) and with antibiotics (streptomycin and penicillin) at concentration of
100 U · mL−1
.
3.9. Incorporation of BrSdU and ISdU into Genomic DNA
The MCF-7 cell line was seeded into plate and incubated under 37 ◦C and 5% CO2 overnight.
After that, the medium was replaced with fresh one and the cells were treated with BrSdU and ISdU at
concentration of 0 (control) and 10−4 M. Plates with cells were incubated (under the same conditions)
with compound for 48 h. After this time, the cells were pulled from the plates and isolation was
carried out according to the protocol provided by the manufacturer (GeneMATRIX Cell Culture DNA
Purification Kit, EURX, Gdan´sk, Poland). After that, the purified DNA was enzymatically digested
by the simultaneous action of DNase I, snake venom phosphodiesterase (SVP) and bacterial alkaline
phosphatase (BAP).
3.9.1. HPLC Analysis
The mixture of nucleoside (dC, dA, dG, dT, BrdU, BrSdU, and ISdU), nontreated and treated
with BrSdU/ISdU samples, were analyzed with reversed-phase HPLC method. For the separation of

Molecules 2019, 24, 2819
13 of 17
analytes, a C18 column (Wakopak Handy ODS, 4.6
×
150 mm, 5
µ
m in particle size and 100 Å in pore
size), gradient elution with 80% ACN, and 0.1% HCOOH (from 0 to 50% ACN in 30 min), flow rate
1 mL · min−1 were used. The HPLC analysis was performed on the DionexUltiMate 3000 System with
a Diode Array Detector (Dionex Corporation, Sunnyvale, CA, USA) set at 260 nm.
3.9.2. LC-MS Analysis
DNA samples, isolated from the ISdU/BrSdU-treated culture, after enzymatic digestion were
analyzed by LC-MS and LC-MS/MS methods. Conditions of separations: Kinetex column (Phenomenex,
1.7
µ
m, C18, 100 Å, 2.1
×
150 mm); flow rate 0.3 mL
min−1; a gradient elution with 80% ACN and
·
0.1% HCOOH (from 0 to 50% acetonitrile); the oven temperature was maintained at 25 ◦C. The effluent
was diverted to waste for 2 min after injection. Conditions for MS and MS/MS analysis: the spray
voltage was
4.5 kV and the source temperature was 300 ◦C. The LC-MS and LC-MS/MS analyses
−
were performed with the use of TripleTOF 5600+ (SCIEX) mass spectrometer (operated in negative
mode) coupled with Ultra High Performance Liquid Chromatography (UHPLC) system Nexera X2.
3.10. Flow Cytometry Analysis of Histone H2A.X Phosphorylation and CellDdeath
3.10.1. Analysis of Histone H2A.X Phosphorylation
Human cells of breast cancer MCF-7 were grown in RPMI medium supplemented by the 10% FBS
and antibiotics (streptomycin and penicillin) at a concentration of 100 U
mL−1. Cells, at a density
·
of 0.2
106 per plate, were incubated for 24 h (37 ◦C, 5% CO2). After this time, the cells were treated
·
with BrSdU at a concentration of 10−4 M and incubated for next 48 h under the same conditions. Then
plate cultures were irradiated (Cellrad X-ray cabinet, Faxitron X-ray Corporation) with 1 and 2 Gy
doses (1.27 Gy
were dissociated with 1x Accutase solution, fixed, permeabilized, and stained. The last step was
cytometric analysis (Guava easyCyte 12, Merck, Hayward, CA, USA). Fixation, permeabilization,
and staining were carried out according to the manufacturer’s protocol (FlowCellect Histone H2A.X
·
min−1, 130.0 kV, 5.0 mA) and incubated for 1 h. After this time, the MCF-7 cells
™
™
Phosphorylation Assay Kit, Merck). The experiment was carried out in duplicate. Nontreated cultures
were used as controls.
3.10.2. Cell Death
After 24 h incubation, cells were treated with BrSdU at a concentration of 10−4 M and again
incubated for 48 h (37 ◦C, 5% CO2). Then, the cells were irradiated (Cellrad X-ray cabinet, Faxitron
X-ray Corporation) with a dose of 5 Gy (1.27 Gy
under the same conditions. After this time, the cells were dissociated with 1x Accutase solutions and
analyzed by flow cytometry (Guava easyCyte 12, Merck, Warsaw, Poland) using the manufacturer’s
·
min−1, 130.0 kV, 5.0 mA) and incubated for 24 h
™
protocol (FlowCellect™ MitoDamage Kit, Merck).
3.11. Computational
All calculations were performed with the B3LYP functional and DGDZVP++ basis set using the
Gaussian09 package. Water environment was simulated with the polarizable continuum model (PCM)
implemented therein, and in one of the reactions, three water molecules were included explicitly.
In order to reduce the cost of calculations, 20-deoxyribose moiety in nucleoside was substituted with
the methyl group in the computational model (i.e., we used 5-bromo-1-methyl-4-thiouracil (BrSU),
see Figures 2 and 3) The sugar moiety does not take part in any of the considered reactions, hence the
conversion of the studied system to BrSU is not expected to affect the computational results. The minima,
as well as transition states (TSs), were proven with the vibrational frequency calculations, and the IRC
procedure was applied to show that particular TSs are connected to the appropriate minima.

Molecules 2019, 24, 2819
14 of 17
4. Conclusions
Radiotherapy is one of the most common modalities employed against cancer diseases.
Unfortunately, its efficacy is seriously impaired due to hypoxia of solid tumors. Therefore, to be efficient,
radiotherapy should be combined with radiosensitizers–compounds that are able to sensitize cells to
ionizing radiation. The modified uridines belong to radiosensitizers, which produce radiosensitization
by incorporation into DNA. Results of numerous studies, mainly on BrdU and IdU, suggest that
dissociative electron attachment to the modified uridines incorporated into DNA is responsible for the
increased level of damage.
Our studies seem to confirm this view. Using enzymatic digestion of genomic DNA extracted from
the cells incubated with BrSdU and ISdU, which was followed by the HPLC and LC-MS analysis of the
lysates, we demonstrated that both compounds are incorporated into the DNA of the studied cell lines.
Although BrSdU/ISdU are nontoxic, as shown by the results of the MTT test, their radiosensitizing
efficacy turned out to be very different. Thus, clonogenic assay, the method of choice to determine
cell reproductive death after treatment with ionizing radiation, shows significant activity of ISdU and
practical lack of radiosensitization in the case of BrSdU. Similarly, the number of viable, early apoptotic,
late apoptotic, and dead cells was not influenced by the pre-incubation with BrSdU, while the incubation
with ISdU significantly increased the population of apoptotic and dead cells. Finally, the level of
double-stand breaks which are associated with the cell death, assayed with the phosphorylation of
histone γH2A.X test, clearly increased only in case of incubation with ISdU.
The significant variation in radiosensitizing activity of such similar derivatives has been explained
with our radiolysis and computational studies. The stationary radiolysis of BrSdU demonstrated
that although the compound was decomposed, no DEA products were detected. On the other
hand, the radiolysis of ISdU solutions led to 4-thio-2’-deoxyuridine and its dimers. A similar
conclusion was drawn from pulse radiolysis. Namely, the main dissimilarity between two studied
4-thio-2’-deoxyuridines lies in the fact that the transient absorption of 4SdU• radical was observed
only in the solutions of ISdU.
All of the above-mentioned observations are well-explained by difference in the activation barrier
that accompanies the release of the halogen anion from the XSdU•− anion radical. Namely, the barrier
for BrSdU•− is more than two-fold larger than that characteristic for DEA in ISdU•−. As a consequence,
the C–X dissociation process is ca. 200-fold slower in the BrSdU anion. Hence, the lifetime of the anion
is sufficiently long for BrdU•− to be protonated forming BrSdUH•, which prevents the DEA process.
Thus, our studies confirm the crucial role of DEA leading to uracil-5-yl radical in the radiosensitization
mechanism of modified uracils. Moreover, they show that the relatively low barrier of ca. 26 kJ/mol is
able to inhibit DEA in the studied class of molecules. Our studies demonstrate that even such a modest
barrier makes the potential radiosensitizer completely inactive. This finding is of highly valuable for
the computational search of radiosensitizing molecules.
Supplementary Materials: The following are available online, LC-MS/MS identification of radiolysis products
(Figures S1–S4); Transition state geometries for the reactions identified during stationary radiolysis (Figure S5);
Incorporation yield of BrSdU and ISdU into the genomic DNA determined with its digestion and HPLC and
LC-Ms analyses (Figures S6–S10); Flow cytometry analysis of histone H2A.X phosphorylation and cell death
(Figures S11–S13).
Author Contributions: Conceptualization, J.R. and K.B.; methodology, M.Z., P.S., K.S., W.K., S.M.; investigation,
P.S., M.Z., K.S., W.K., S.M.; writing—original draft preparation, P.S., M.Z., S.M., W.K., K.S., J.R. and K.B.;
writing—review and editing, P.S., M.Z., S.M., W.K., K.S., J.R. and K.B., supervision, J.R.; funding acquisition, J.R.
Funding: This work was supported by the Polish National Science Centre under Grants No. 2014/14/A/ST4/00405
(J.R.) and 2018/28/C/ST4/00479 (K.S.). Some of the calculations were performed in the Wroclaw Center for
Networking and Supercomputing, grant No. 209.
Conflicts of Interest: The authors declare no conflict of interest.

Molecules 2019, 24, 2819
15 of 17
References
˙
1.
Rak, J.; Chomicz, L.; Wiczk, J.; Westphal, K.; Zdrowowicz, M.; Wityk, P.; Zyndul, M.; Makurat, S.; Golon, Ł.
Mechanisms of damage to DNA labeled with electrophilic nucleobases induced by ionizing or UV radiation.
J. Phys. Chem. B 2016, 119, 8227–8238. [CrossRef] [PubMed]
2.
3.
4.
Chomicz, L.; Rak, J.; Storoniak, P. Electron-induced elimination of the bromide anion from brominated
nucleobases. A computational study. J. Phys. Chem. B 2012, 116, 5612–5619. [CrossRef] [PubMed]
Wieczór, M.; Wityk, P.; Czub, J.; Chomicz, L.; Rak, J. A first-principles study of electron attachment to the
fully hydrated bromonucleobases. Chem. Phys. Lett. 2014, 595–596, 133–137. [CrossRef]
Park, Y.; Polska, K.; Rak, J.; Wagner, J.R.; Sanche, L. Fundamental mechanisms of DNA radiosensitization:
Damage induced by low-energy electrons in brominated oligonucleotide trimers. J. Phys. Chem. B 2012, 116,
9676–9682. [CrossRef] [PubMed]
5.
Polska, K.; Rak, J.; Bass, A.D.; Cloutier, P.; Sanche, L. Electron stimulated desorption of anions from native
and brominated single stranded oligonucleotide trimers. J. Chem. Phys. 2012, 136, 075101. [CrossRef]
[PubMed]
6.
7.
8.
9.
Jagiello, K.; Makurat, S.; Perec´, S.; Rak, J.; Puzyn, T. Molecular features of thymidine analogues governing
the activity of human thymidine kinase. Struct. Chem. 2018, 29, 1367–1374. [CrossRef]
Goz, B. The Effects of Incorporation of 5-halogenated deoxyuridines into the DNA of eukaryotic cells.
Pharmacol. Rev. 1977, 29, 249–272.
Visvader, J.E.; Lindeman, G.J. Cancer stem cells in solid tumors: Accumulating evidence and unresolved
questions. Nat. Rev. Cancer 2008, 8, 755–768. [CrossRef]
Rockwell, S.; Dobrucki, I.T.; Kim, E.Y.; Marrison, S.T.; Vu, V.T. Hypoxia and radiation therapy: Past history,
ongoing research and future promise. Curr. Mol. Med. 2009, 9, 442–458. [CrossRef]
10. Oronsky, B.T.; Knox, S.J.; Scicinski, J. Six degrees of separation: The oxygen effect in the development of
radiosensitizers. Trans. Oncol. 2011, 4, 189–198. [CrossRef]
11. Westphal, K.; Skotnicki, K.; Bobrowski, K.; Rak, J. Radiation damage to single stranded oligonucleotide
trimers labelled with 5-iodopyrimidines. Org. Biomol. Chem. 2016, 14, 9331–9337. [CrossRef]
12. Westphal, K.; Wiczk, J.; Miloch, J.; Kciuk, G.; Bobrowski, K.; Rak, J. Irreversible electron attachment—A
key to DNA damage by solvated electrons in aqueous solution. Org. Biomol. Chem. 2015, 13, 10362–10369.
[CrossRef] [PubMed]
13. Cecchini, S.; Girouard, S.; Huels, M.A.; Sanche, L.; Hunting, D.J. Interstrand cross-links: A new type of γ-ray
damage in bromodeoxyuridine-substituted DNA. Biochemistry 2005, 44, 1932–1940. [CrossRef]
14. Ling, L.L.; Ward, J.F. Radiosensitization of Chinese hamster V79 cells by bromodeoxyuridine substitution of
thymidine: Enhancement of radiation-induced toxicity and DNA strand break production by monofilar
andbifilar substitution. Radiat. Res. 1990, 121, 76–83. [CrossRef]
15. Miller, E.M.; Fowler, J.F.; Kinsella, T.J. Linear-quadratic analysis of radiosen-sitization by halogenated
pyrimidines. I.Radiosensitization of humancolon cancer cells by iododeoxyuridine. Radiat. Res. 1992, 131,
81–89. [CrossRef]
16. Miller, E.M.; Fowler, J.F.; Kinsella, T.J. Linear-quadratic analysis of radiosen-sitization by halogenated
pyrimidines. II. Radiosensitization of humancolon cancer cells by bromodeoxyuridine. Radiat. Res. 1992
131, 90–97. [CrossRef]
,
17. McGinn, C.J.; Shewach, D.S.; Lawrence, T.S. Radiosensitizing nucleosides. J. Natl. Cancer Inst. 1996, 88,
1193–1203. [CrossRef] [PubMed]
18. Phillips, T.L.; Scott, C.B.; Leibel, S.A.; Rotman, M.; Weigensberg, I.J. Results of a randomized comparison of
radiotherapy and bromodeoxyuridine with radiotherapy alone for brain metastases: Report of RTOG trial
89-05. Int. J. Radiat. Oncol. Biol. Phys. 1995, 33, 339–348. [CrossRef]
˙
19. Zdrowowicz, M.; Chomicz, L.; Zyndul, M.; Wityk, P.; Rak, J.; Wiegand, T.J.; Hanson, C.G.; Adhikary, A.;
Sevilla, M.D. 5-Thiocyanato-20-deoxyuridine as a possible radiosensitizer: Electron-induced formation of
uracil-C5-thiyl radical and its dimerization. Phys. Chem. Chem. Phys. 2015, 17, 16907–16916. [CrossRef]
[PubMed]
20. Sosnowska, M.; Makurat, S.; Zdrowowicz, M.; Rak, J. 5 selenocyanatouracil: A potential hypoxic
radiosensitizer. electron attachment induced formation of selenium centered radical. J. Phys. Chem.
B 2017, 121, 6139–6147. [CrossRef]

Molecules 2019, 24, 2819
16 of 17
21. Makurat, S.; Zdrowowicz, M.; Chomicz-Man´ka, L.; Kozak, W.; Serdiuk, I.E.; Wityk, P.; Kawecka, A.;
Sosnowska, M.; Rak, J. 5-Selenocyanato and 5-trifluoromethanesulfonyl derivatives of 20-deoxyuridine:
Synthesis, radiation and computational chemistry as well as cytotoxicity. RSC Adv. 2018, 8, 21378–21388.
[CrossRef]
22. Ameixa, J.; Arthur-Baidoo, E.; Meißner, R.; Makurat, S.; Kozak, W.; Butowska, K.; Ferreira da Silva, F.; Rak, J.;
Denifl, S. Low–energy electron–induced decomposition of 5–trifluoromethanesulfonyl–uracil: A potential
radiosensitizer. J. Chem. Phys. 2018, 149, 164307. [CrossRef] [PubMed]
23. Meißner, R.; Makurat, S.; Kozak, W.; Limão-Vieira, P.; Rak, J.; Denifl, S. Electron-induced dissociation of the
potential radiosensitizer 5-selenocyanato-20-deoxyuridine. J. Phys. Chem. B 2019, 123, 1274–1282. [CrossRef]
[PubMed]
24. Brem, R.; Zhang, X.; Xu, Y.Z.; Karran, P. UVA photoactivation of DNA containing halogenated thiopyrimidines
induces cytotoxic DNA lesions. J. Photochem. Photobiol. B Biol. 2015, 145, 1–10. [CrossRef] [PubMed]
25. Xu, Y.Z.; Zhang, X.; Wu, H.C.; Massey, A.; Karran, P. 4-Thio-5-bromo-20-deoxyuridine: Chemical synthesis
and therapeutic potential of UVA-induced DNA damage. Bioorg. Med. Chem. Lett. 2004, 14, 995–997.
[CrossRef] [PubMed]
26. Brem, R.; Guven, M.; Karran, P. Oxidatively-generated damage to DNA and proteins mediated by
photosensitized UVA. Free Rad. Biol. Med. 2017, 107, 101–109. [CrossRef] [PubMed]
27. Makurat, S.; Spisz, P.; Kozak, W.; Rak, J.; Zdrowowicz, M. 5-iodo-4-thio-20-deoxyuridine as a sensitizer of
X-ray induced cancer cell killing. Int. J. Mol. Sci. 2019, 20, 1308. [CrossRef]
28. Zeida, A.; Babbush, R.; Gonzalez-Lebrero, M.C.; Trujillo, M.; Radi, R.; Estrin, D.A. Molecular basis of the
mechanism of thiol oxidation by hydrogen peroxide in aqueous solution: Challenging the SN2 paradigm.
Chem. Res. Toxicol. 2012, 25, 741–746. [CrossRef]
29. Van Bergen, L.A.; Roos, G.; De Proft, F. From thiol to sulfonic acid: Modeling the oxidation pathway of
protein thiols by hydrogen peroxide. J. Phys. Chem. A 2014, 118, 6078–6084. [CrossRef]
30. Bahrami, K.; Khodaei, M.; Tajik, M. Trimethylsilyl chloride promoted selective desulfurization of thiocarbonyls
to carbonyls with hydrogen peroxide. Synthesis 2010, 24, 4282–4286. [CrossRef]
31. Sonntag, C. Free-Radical-Induced DNA Damage and Its Repair: A Chemical Perspective; Springer Science &
Business Media: Heidelberg, Germany, 2010.
32. Chu, J.W.; Trout, B.L. On the mechanisms of oxidation of organic sulfides by H2O2 in aqueous solutions. J.
Am. Chem. Soc. 2004, 126, 900–908. [CrossRef] [PubMed]
33. McCaw, P.G.; Buckley, N.M.; Collins, S.G.; Maguire, A.R. Generation, reactivity and uses of sulfines in organic
synthesis. Eur. J. Org. Chem. 2016, 9, 1630–1650. [CrossRef]
34. Adam, W.; Bargon, R.M. Synthesis of thiiranes by direct sulfur transfer: The challenge of developing effective
sulfur donors and metal catalysts. Chem. Rev. 2004, 104, 251–262. [CrossRef] [PubMed]
35. Chomicz, L.; Zdrowowicz, M.; Kasprzykowski, F.; Rak, J.; Buonaugurio, A.; Wang, Y.; Bowen, K.H. How to
find out whether a 5-substituted uracil could be a potential dna radiosensitizer. J. Phys. Chem. Lett. 2013, 4,
2853–2857. [CrossRef]
36. Wetmore, S.D.; Boyd, R.J.; Eriksson, L.A. A theoretical study of 5-halouracils: Electron affinities, ionization
potentials and dissociation of the related anions. Chem. Phys. Lett. 2001, 343, 151–158. [CrossRef]
37. Li, X.; Sanche, L.; Sevilla, M.D. Dehalogenation of 5- halouracils after low energy electron attachment:
A density functional theory investigation. J. Phys. Chem. A 2002, 106, 11248–11253. [CrossRef]
38. Eyring, H. The Activated Complex in Chemical Reactions. J. Chem. Phys. 1935, 3, 107. [CrossRef]
39. McAllister, M.; Smyth, M.; Gu, B.; Tribello, G.A.; Kohanoff, J. Understanding the Interaction between
Low-Energy Electrons and DNA Nucleotides in Aqueous Solution. J. Phys. Chem. Lett. 2015, 6, 3091–3097.
[CrossRef] [PubMed]
40. Putschew, A.; Jekel, M. Induced in-source fragmentation for the selective detection of organic bound iodine by
liquid chromatography/electrospray mass spectrometry. Rapid Commun. Mass Spectrom. 2003, 17, 2279–2282.
[CrossRef]
41. Hütteroth, A.; Putschew, A.; Jekel, M. Selective detection of unknown organic bromine compounds and
quantification potentiality by negative-ion electrospray ionization mass spectrometry with induced in-source
fragmentation. Int. J. Environ. Anal. Chem. 2007, 87, 415–424. [CrossRef]
42. Rafehi, H.; Orlowski, C.; Georgiadis, G.T.; Ververis, K.; El-Osta, A.; Karagiannis, T.C. Clonogenic assay:
Adherent cells. J. Vis. Exp. 2011, 49, 2573–2576. [CrossRef] [PubMed]

Molecules 2019, 24, 2819
17 of 17
43. Mosmann, T. Rapid colorimetric assay for cellular growth and survival: Application to proliferation and
cytotoxicity assays. J. Immunol. Methods. 1983, 65, 55–63. [CrossRef]
44. Taneja, N.; Davis, M.; Choy, J.S.; Beckett, M.A.; Singh, R.; Kron, S.J.; Weichselbaum, R.R. Histone H2AX
phosphorylation as a predictor of radiosensitivity and target for radiotherapy. J. Biol. Chem. 2004, 279,
2273–2280. [CrossRef] [PubMed]
45. Bobrowski, K. Free radicals in chemistry, biology and medicine: contribution of radiation chemistry.
Nukleonika 2005, 50, S67.
Sample Availability: Samples of the compounds are available from the authors.
©
2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).