Reactive Oxygen Species-Triggered Tunable Hydrogen Sulfide Release

Article abstract of DOI:10.1021/acs.orglett.8b01356

A series of carbamothioates with tunable release of H₂S after activation by reactive oxygen species are reported. The half-lives of H₂S release could be tuned from 24 to 203 min by varying the basicity of the amine.

Full text of DOI:10.1021/acs.orglett.8b01356

Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
Reactive Oxygen Species-Triggered Tunable Hydrogen Sulfide
Release
Preeti Chauhan,‡ Swetha Jos,‡ and Harinath Chakrapani*
Department of Chemistry, Indian Institute of Science Education and Research Pune, Dr. Homi Bhabha Road, Pashan Pune 411 008,
Maharashtra, India
S
* Supporting Information
ABSTRACT: A series of carbamothioates with tunable release of
H2S after activation by reactive oxygen species are reported. The half-
lives of H2S release could be tuned from 24 to 203 min by varying
the basicity of the amine.
ydrogen sulfide (H2S) is an important gaseous cellular
known and some utilize physiologically relevant stimuli to
Hantioxidant and signaling agent that mediates numerous release H2S.16 However, none to our knowledge display
processes.1−3 Diminished H2S has been associated with various
stimuli-responsiveness as well as tunability by using a
physiologically relevant trigger.17 Here, we report arylboronate
diseases like neurological disorders, cardiovascular diseases,
gastrointestinal diseases, inflammation, etc.1−3 Most of these
ester-based carbamothioates that are triggerable by ROS and
after activation, the release of H2S can be varied by modifying
the basicity of the amine.
conditions are associated with an increased level of reactive
oxygen species (ROS).4,5 Hydrogen sulfide (H2S) is known to
stimulate the cellular antioxidant machinery as a cytoprotective
measure.4 Treatment with H2S induces translocation of nuclear
factor NRF2 to enhance its binding with anti-inflammatory
response element (ARE) and in turn increase the production
of glutathione, a free-radical scavenger. Thus, delivery of
hydrogen sulfide to areas associated with inflammation is of
therapeutic interest.6−8 Left untreated, inflammation can
accentuate tissue damage and accelerate further degeneration.
The major problem with delivery of hydrogen sulfide is that
the antioxidant properties of hydrogen sulfide are largely
dependent on the location and the rate of release. The rate of
release can be modulated by using different classes of donors.
For example, NaSH, an inorganic source of H2S, can be used
for fast release, while GYY4137 is commonly used for
prolonged release of hydrogen sulfide. They are found to
have contrasting effects on inflammation. In a study by Moore
and co-workers, NaSH was found to exacerbate the condition
by increasing the levels of pro-inflammatory cytokines like
TNF-α or IL-6,9,10while GYY4137 attenuated the levels of
these pro-inflammatory cytokines.11,12 In yet another study by
Wang and co-workers, the superior cardioprotective effects of
diallyl trisulfide-based nanoparticle over NaSH was demon-
strated in a myocardial ischemia reperfusion model.13 These
observations underscore the importance of modulating release
of exogenous hydrogen sulfide. While a number of nano-
particle-based methods for slowing down release of hydrogen
sulfide are known, it is desirable to have a class of small
molecules where tunable release is possible.14,15 Next, the
other problem is to localize delivery of hydrogen sulfide. Here,
a number of classes of H2S donors which localize H2S are
Recently, ROS-activated carbonyl sulfide (COS) donors
were reported by Pluth and co-workers.18,19 This donor is
triggered by hydrogen peroxide, a relatively stable ROS to
generate COS, which undergoes hydrolysis to produce
hydrogen sulfide. Hydrolysis of COS is accelerated by catalytic
carbonic anhydrase (CA), a widely prevalent enzyme.20 This
donor has the distinct advantage of being triggered by a
metabolite (H2O2) that is frequently associated with
inflammation.1−3 The mechanism of COS release involves
the generation of a phenolate I, which undergoes self-
immolation to produce the anion II, which fragments to
produce COS and an amine (Figure 1a).The penultimate step
in this process, wherein the loss of COS and RNH2 occurred,
was found to be the step with the highest barrier (Figure 1a).19
We hypothesized that a similar intermediate was possible by
starting from the carbamothioate (Figure 1b). The thiocarba-
mate anion V, which is generated by self-immolation of IV, is
expected to undergo fragmentation to produce COS. This
process involves the protonation of V, which we proposed
would depend on the basicity of the amine. Our laboratory had
previously reported esterase-sensitive COS/H2S donors where
we found some evidence for the dependence of rate of
hydrogen sulfide generation on the basicity of the amine. For
example, the rate of hydrogen sulfide generation from the
benzylamine derivative (0.014 min−1; pKa of amine, 9.34) was
significantly slower when compared with the aniline derivative
(0.032 min−1, pKa 5.34).21 We therefore envisaged that the
Received: April 28, 2018
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.8b01356
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 1. Synthesis of Thiol 4
p-nitrophenylcarbamate 5a was synthesized using a reported
protocol (Table 1, entry 1). This carbamate was reacted with
thiol 4 to give carbamothioate 1a (Table 1, entry
1).Compound 2 (Figure 2) was synthesized as a negative
Figure 2. Structures of compounds 2, 6, 7, and mesalamine.
Figure 1. Arylboronate esters of thiocarbamates are activated by
hydrogen peroxide to produce COS: (a) reported donors that
generate hydrogen sulfide; (b) design of tunable COS/H2S donors.
control from the reaction of 3 with 5c (see the Supporting
Information). Compound 2 should cleave in the presence of
H2O2 but will not produce H2S (Figure 2).
The hydroxyethyl derivative 5b was next synthesized in four
steps using a reported protocol.23This amine was chosen due
to its improved aqueous solubility. Similarly, carbamates 5c−g
were converted to the corresponding carbamothioate using a
similar methodology (Table 1, entries 2−7). Under these
reaction conditions, the tert-butylamine carbamate 5h
decomposed, and we were unable to synthesize the
corresponding carmabothioate. In an effort to synthesize
secondary amine-based derivatives, the carbamate of pyrroli-
dine 5i was synthesized. This compound, however, did not
react with 4 possibly due to the low reactivity of carbamates
derived from secondary amines.
We first tested the ability of compounds to produce H2S by
using a methylene blue formation assay, which is frequently
used for the detection of H2S. Briefly, the donors pretreated
with CA were independently incubated with 10 equiv of H2O2
in phosphate-buffered saline (PBS, pH = 7.4, 50 mM) at 37
°C, the resulting mixture was exposed to the rest of the
methylene blue reagents, and absorbance was recorded (see
the Supporting Information).
basicity of the amine would have a significant impact on the
release rate after activation. Although thiocarbamates have
been similarly evaluated, the range of pKa’s was small (1−5),
and correspondingly, the range of rates of release was also
somewhat limited.19 Furthermore, these thiocarbamates also
undergo decomposition by pathways that did not depend on
ROS: thiocarbamates can undergo deprotonation and
subsequent rearrangement to produce an isothiocyanate,
which is known to undergo hydrolysis in buffer to produce
hydrogen sulfide (Figure 1a).
In order to test our hypothesis, we chose amines with a
range of pKa values of roughly 7 units (see Table 1).22 The
compounds were synthesized in four steps starting from (4-
(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)methanol
3 to give the thiol 4, with 82% overall yield (Scheme 1). Next,
Table 1. Synthesis of Carbamothioates
The formation of methylene blue complex was confirmed by
measuring absorbance at 676 nm, and the donors showed
varying levels of H2S release (30 min, Figure 3a; 2 h, see Figure
S1b). The absorbance profile of 1g is shown and was found to
be similar to that of authentic Na2S (see Figure S1a). As
expected, no H2S formation from compound 2 was observed as
it lacks the ability to form COS (Figure 3a). Having
established the H2S release from the donors, we next studied
the effect of the basicity of the amine (pKa) on H2S release.
Using the methylene blue protocol for monitoring hydrogen
sulfide release, we studied the time course of H2S release (see
Figure S2). Pseudo-first-order rate constants were obtained,
and half-lives were calculated (Table 2). Compound 1c with
ansidine as the leaving group was found to have the fastest rate
of H2S release with half-life of 23.9 min−1 (Table 2).
pK
%
a a
entry
R
RNH2
carbamate prod yield
1
2
3
4
5
6
7
8
9
(3-COOMe)Ph
4-OCH2CH2OHPh
(4-OMe)Ph
(4-NO2)PhCH2
4-(O-propargyl)-PhCH2
PhCH2
4.75
5.03
5.34
8.36
9.18
5a
5b
5c
5d
5e
5f
1a
1b
1c
1d
1e
1f
28
32
27
23
41
21
27
b
b
b
9.34
CH3CH2CH2
10.53
10.45
11.27
4.13
5g
5h
5i
1g
1h
1i
tert-butyl
pyrrolidine
b
10
mesalamine methyl ester
5j
1j
52
On the other hand, compound 1g with propylamine as the
leaving group, turned out to be the slowest H2S donor with the
a
b
pKa values are from ref 22. SciFinder was used.
B
DOI: 10.1021/acs.orglett.8b01356
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
in the plot, we found that the position of 6a (white square,
Figure 4) was nearly identical to the derivative of the amines of
comparable basicity (1b, Table 1). Again, this was a testament
to the predictability of H2S release from carbamothioates.
For a better understanding of the mechanism of
decomposition, we monitored the decomposition of com-
pound 1c with concomitant formation of p-anisidine by HPLC
analysis. Compound 1c when incubated in PBS readily
converted to boronic acid. Upon treatment of 1c with H2O2
we observed complete disappearance of the compound in 30
min, which is in accordance with the previous reports
suggesting fast reactivity of boronic acids with H2O2.24
Formation of a new peak after 30 min suggested the formation
of an intermediate which decomposed over a period of 2 h to
generate p-anisidine (see Figure S4). The rate of p-anisidine
formation (0.036 min−1) was found to be in accordance with
the rate of H2S release (0.029 min−1). A similar pattern was
observed in the case of 1g. Here, 1g was converted to boronic
acid in buffer which upon reaction with H2O2 led to the
formation of an intermediate within 30 min (see Figure S5).
The intermediate formed in the case of 1g gradually
decomposed over a period of 6 h (see Figure S6). Despite
repeated attempts, we were unable to characterize this
intermediate. However, on the basis of its absorbance in the
UV region, it is unlikely to be the V (Figure 1b) as aliphatic
amines have poor absorbance at this wavelength. It is therefore
possible that the intermediate is IV, whose self-immolation rate
is also dependent on the basicity of the amine, and the
observed data would be consistent with this step being the
rate-determining step (see Scheme S1). We are unable to
provide experimental support for this hypothesis at this time.
In order to test the stability of compounds in buffer, 1g was
incubated in phosphate buffer (pH 7.4) at 37 °C for 5 h. We
found nearly quantitative recovery of the compound,
suggesting that the compounds are stable under reported
conditions (see Figures S5 and S6).
It is intriguing that the compounds synthesized and
evaluated by Pluth and co-workers, i.e., 6a and its
thiocarbamate isomer 6b, have distinct rates with nearly 12-
fold difference in their second-order rate constants for
hydrogen sulfide release (Table 2).19 This observation is not
consistent with the decomposition of II, which is proposed as
the common intermediate, being the rate-determining step
since they should produce hydrogen sulfide at the same rate.19
Further work clearly needs to be done to characterize these
differences better. Lastly, the trigger-independent pathways
that may contribute to producing hydrogen sulfide are not
relevant when carbamothioates are used (see Scheme S1).
Thus, due to their predictable mechanism, the use of
carbamothioates (1) reported herein may offer significant
advantages over other hydrogen sulfide donors.
Figure 3. (a) Measurement of H2S release in the presence of H2O2
using methylene blue after 30 min incubation. (b) Representative
absorbance trace of the methylene blue assay conducted with 1g in
the presence of hydrogen peroxide after 4 h. Ctrl represents 1g alone.
Table 2. Kinetics of H2S Release
a
compd
pKa
k
t1/2 (min)
relative rate
1a
1b
1c
1d
1e
1f
1g
1j
6a
4.75
5.03
5.34
8.36
9.18
9.34
10.53
4.13
4.66
4.66
0.026
0.017
0.029
0.004
26.7
40.8
23.9
0.90
0.59
1.00
0.14
0.16
0.17
0.12
0.50
0.55
n.d.
173.3
154.0
141.4
203.8
54.8
0.0045
0.0049
0.0034
0.0145
b
c
0.10
43.3
b
6b
1.16
n.d.
a
b
All values are in min−1 unless otherwise mentioned. Values are in
c
M−1 s−1. Calculated from a pseudo-first-order rate constant of 0.016
min−1.19
half-life of 203.8 min−1. Consistent with our hypothesis, we
observed a fast rate of H2S release from aniline-based
derivatives with pKa values in the range 4.75 to 5.34 compared
to derivatives with pKa values in the range of 8.36−10.53
(Table 2). Relative rates were determined, and an 8-fold
difference in the rate of H2S release from our donors was
recorded.
Linear regression analysis of relative rates and pKa of the
corresponding amine showed a good correlation, suggesting
that it is possible to modulate pKa to change hydrogen sulfide
release rate (Figure 4). While this paper was in preparation, 4-
fluoroaniline derivative 6 was synthesized and found to
generate H2S when exposed to hydrogen peroxide.19 The
pKa of 4-fluoroaniline was 4.66, and the rate constant was 0.01
M−1 s−1. The pseudo-first-order rate constant reported by
them was 0.016 min−1 (data is from Figure S3, Supporting
Information, from ref 19). When this data point was included
Lastly, hydrogen sulfide has been used for the treatment of
gastrointestinal disorders associated with inflammation.25,26 In
a study by Wallace and co-workers, H2S contributed to gastric
mucosal defense by resisting injury induced by exogenous
substances such as NSAIDs (corrodes the stomach lining),
stress, or ischemia reperfusion.29,30 The underlying cytopro-
tective activity of H2S was attributed to its ability to inhibit
leukocyte adherence to the vascular endothelium, markedly
caused due to the pathogenesis of NSAIDs. Wallace and co-
workers reported H2S−NSAID hybrids (mesalamine, Figure 2,
or naproxen)27 which retained the anti-inflammatory proper-
ties of the NSAID alone28 and also significantly reduced the
Figure 4. Linear regression analysis of relative rates of hydrogen
sulfide release upon treatment of the compound with hydrogen
peroxide with the pKa of the amine. The data for this plot are available
in Table 2. The white square indicates the relative rate data for the 4-
fluoroaniline derivative 6a.19
C
DOI: 10.1021/acs.orglett.8b01356
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
gastric mucosal damage caused by them. Several H2S-releasing
NSAIDs have been previously synthesized and evaluated.29,30
However, using ROS as a trigger, to our knowledge, a hybrid
compound has not been reported. We synthesized 1j, which is
a derivative of mesalamine, a clinically used NSAID for the
treatment of colitis (Table 1, entry 10).28 This compound was
found to release H2S (Figure 5a), and its release profile
ACKNOWLEDGMENTS
■
We thank the Department of Science and Technology (DST,
Grant No. EMR/2015/000668), Department of Biotechnol-
ogy, India (BT/PR15848/MED/29/1025/2016), for financial
support for our research. We also thank the Council for
Scientific and Industrial Research (CSIR) and the Department
of Science and Technology − Innovation in Science Pursuit for
Inspired Research (DST-INSPIRE) for fellowships. We thank
Shraddha Bhurkunde, IISER Pune, for her help with the TOC
graphic.
REFERENCES
■
(1) Wang, R. Antioxid. Redox Signaling 2003, 5 (4), 493−501.
(2) Gadalla, M. M.; Snyder, S. H. J. Neurochem. 2010, 113 (1), 14−
26.
(3) Wang, R. Physiol. Rev. 2012, 92 (2), 791−896.
(4) Wallace, J. L.; Wang, R. Nat. Rev. Drug Discovery 2015, 14, 329.
(5) Zheng, Y.; Ji, X.; Ji, K.; Wang, B. Acta Pharm. Sin. B 2015, 5 (5),
367−377.
Figure 5. (a) Representative kinetic plot of H2S release from 1j in the
presence of H2O2 as determined by a methylene blue assay. Curve
fitting yielded a pseudo-first-order rate constant of 0.0145 min−1. (b)
Cell viability assay conducted with MCF-7 breast cancer cell line with
1j during 24 h. Veh is DMSO.
(6) Wallace, J. L. Trends Pharmacol. Sci. 2007, 28 (10), 501−505.
(7) Wallace, J. L. Antioxid. Redox Signaling 2010, 12 (9), 1125−
1133.
(8) Fiorucci, S.; Distrutti, E.; Cirino, G.; Wallace, J. L. Gastro-
enterology 2006, 131 (1), 259−271.
(9) Lee, Z. W.; Zhou, J.; Chen, C.-S.; Zhao, Y.; Tan, C.-H.; Li, L.;
Moore, P. K.; Deng, L.-W. PLoS One 2011, 6 (6), e21077−e21077.
(10) Li, L.; Rossoni, G.; Sparatore, A.; Lee, L. C.; Del Soldato, P.;
Moore, P. K. Free Radical Biol. Med. 2007, 42 (5), 706−719.
(11) Whiteman, M.; Li, L.; Rose, P.; Tan, C.-H.; Parkinson, D. B.;
Moore, P. K. Antioxid. Redox Signaling 2010, 12 (10), 1147−1154.
(12) Li, L.; Fox, B.; Keeble, J.; Salto-Tellez, M.; Winyard, P. G.;
Wood, M. E.; Moore, P. K.; Whiteman, M. J. Cell. Mol. Med. 2013, 17
(3), 365−376.
(pseudo-first-order rate constant, 0.0145 min−1) was com-
parable with that of amine 1b (pKa 5.03). The H2S response
obtained was selective toward H2O2 (see Figure S7).
Decomposition studies showed the formation of the methyl
ester of mesalamine 7 (see Figure S8). The compound formed
could in turn be hydrolyzed to the active NSAID in the
presence of esterase (widely prevalent in cells; see Figure S9).
The hybrid compound was well tolerated by cells as studied by
a standard cell viability assay (Figure 5b), and further studies
are presently underway to characterize the anti-inflammatory
properties. Thus, taken together, we report ROS-activated
COS/H2S31−37 donors whose rates of hydrogen sulfide release
could be modulated by modifying the basicity of the amine.
We also report a novel H2S−NSAID hybrid that decomposes
when exposed to ROS to produce H2S.
(13) Sun, X.; Wang, W.; Dai, J.; Jin, S.; Huang, J.; Guo, C.; Wang,
C.; Pang, L.; Wang, Y. Sci. Rep. 2017, 7 (1), 3541.
(14) Ciocci, M.; Iorio, E.; Carotenuto, F.; Khashoggi, H. A.; Nanni,
F.; Melino, S. Oncotarget 2016, 7 (51), 84338−84358.
(15) Hasegawa, U.; van der Vlies, A. J. Bioconjugate Chem. 2014, 25
(7), 1290−1300.
(16) Zhao, Y.; Biggs, T. D.; Xian, M. Chem. Commun. 2014, 50 (80),
11788−11805.
(17) Xiao, Z.; Bonnard, T.; Shakouri-Motlagh, A.; Wylie, R. A. L.;
Collins, J.; White, J.; Heath, D.; Hagemeyer, C. E.; Connal, L. A.
Chem. - Eur. J. 2017, 23 (47), 11294−11300.
ASSOCIATED CONTENT
(18) Zhao, Y.; Pluth, M. D. Angew. Chem., Int. Ed. 2016, 55 (47),
14638−14642.
■
S
* Supporting Information
(19) Zhao, Y.; Henthorn, H. A.; Pluth, M. D. J. Am. Chem. Soc. 2017,
139 (45), 16365−16376.
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.8b01356.
(20) Ogawa, T.; Noguchi, K.; Saito, M.; Nagahata, Y.; Kato, H.;
Ohtaki, A.; Nakayama, H.; Dohmae, N.; Matsushita, Y.; Odaka, M.;
et al. J. Am. Chem. Soc. 2013, 135 (10), 3818−3825.
(21) Chauhan, P.; Bora, P.; Ravikumar, G.; Jos, S.; Chakrapani, H.
Org. Lett. 2017, 19 (1), 62−65.
Synthesis, characterization data, and protocols for assays
(PDF)
(22) Hall, H. K. J. Am. Chem. Soc. 1957, 79, 5441.
̈
(23) Shendage, D. M.; Frohlich, R.; Haufe, G. Org. Lett. 2004, 6
(21), 3675−3678.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: harinath@iiserpune.ac.in.
ORCID
(24) Zielonka, J.; Sikora, A.; Hardy, M.; Joseph, J.; Dranka, B. P.;
Kalyanaraman, B. Chem. Res. Toxicol. 2012, 25 (9), 1793−1799.
(25) Fiorucci, S.; Antonelli, E.; Distrutti, E.; Rizzo, G.; Mencarelli,
A.; Orlandi, S.; Zanardo, R.; Renga, B.; Di Sante, M.; Morelli, A.; et al.
Gastroenterology 2005, 129 (4), 1210−1224.
(26) Distrutti, E.; Sediari, L.; Mencarelli, A.; Renga, B.; Orlandi, S.;
Antonelli, E.; Roviezzo, F.; Morelli, A.; Cirino, G.; Wallace, J. L.; et al.
J. Pharmacol. Exp. Ther. 2006, 316 (1), 325.
Harinath Chakrapani: 0000-0002-7267-0906
Author Contributions
‡P.C. and S.J. contributed equally.
(27) Fiorucci, S.; Orlandi, S.; Mencarelli, A.; Caliendo, G.;
Santagada, V.; Distrutti, E.; Santucci, L.; Cirino, G.; Wallace, J. L.
Br. J. Pharmacol. 2007, 150 (8), 996−1002.
Notes
The authors declare no competing financial interest.
D
DOI: 10.1021/acs.orglett.8b01356
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
(28) Karagozian, R.; Burakoff, R. Ther. Clin. Risk Manag. 2007, 3
(5), 893−903.
(29) Fiorucci, S.; Santucci, L. Inflammation Allergy: Drug Targets
2011, 10 (2), 133−140.
(30) Wallace, J. L.; Caliendo, G.; Santagada, V.; Cirino, G. Br. J.
Pharmacol. 2010, 159 (6), 1236−1246.
(31) Sharma, A. K.; Nair, M.; Chauhan, P.; Gupta, K.; Saini, D. K.;
Chakrapani, H. Org. Lett. 2017, 19 (18), 4822−4825.
(32) Steiger, A. K.; Marcatti, M.; Szabo, C.; Szczesny, B.; Pluth, M.
D. ACS Chem. Biol. 2017, 12 (8), 2117−2123.
(33) Zhao, Y.; Bolton, S. G.; Pluth, M. D. Org. Lett. 2017, 19 (9),
2278−2281.
(34) Steiger, A. K.; Yang, Y.; Royzen, M.; Pluth, M. D. Chem.
Commun. 2017, 53 (8), 1378−1380.
(35) Steiger, A. K.; Pardue, S.; Kevil, C. G.; Pluth, M. D. J. Am.
Chem. Soc. 2016, 138 (23), 7256−7259.
(36) Powell, C. R.; Foster, J. C.; Okyere, B.; Theus, M. H.; Matson,
J. B. J. Am. Chem. Soc. 2016, 138 (41), 13477−13480.
(37) Zhao, Y.; Steiger, A. K.; Pluth, M. D. Chem. Commun. 2018, 54,
4951−4954.
E
DOI: 10.1021/acs.orglett.8b01356
Org. Lett. XXXX, XXX, XXX−XXX