S-aroylthiooximes: A facile route to hydrogen sulfide releasing compounds with structure-dependent release kinetics

Article abstract of DOI:10.1021/ol500385a

We report the facile preparation of a family of S-aroylthiooxime (SATO) H₂S donors, which are synthesized via a click reaction analogous to oxime formation between S-aroylthiohydroxylamines (SATHAs) and aldehydes or ketones. Analysis of cysteine-triggered H₂S release revealed structure-dependent release kinetics with half-lives from 8-82 min by substitution of the SATHA ring. The pseudo-first-order rate constants of substituted SATOs fit standard linear free energy relationships (p = 1.05), demonstrating a significant sensitivity to electronic effects.

Full text of DOI:10.1021/ol500385a

Letter
pubs.acs.org/OrgLett
S‑Aroylthiooximes: A Facile Route to Hydrogen Sulfide Releasing
Compounds with Structure-Dependent Release Kinetics
Jeffrey C. Foster, Chadwick R. Powell, Scott C. Radzinski, and John B. Matson*
Department of Chemistry, Virginia Tech, Blacksburg, Virginia 24061, United States
S
* Supporting Information
ABSTRACT: We report the facile preparation of a family of S-
aroylthiooxime (SATO) H2S donors, which are synthesized via a click
reaction analogous to oxime formation between S-aroylthiohydroxyl-
amines (SATHAs) and aldehydes or ketones. Analysis of cysteine-
triggered H2S release revealed structure-dependent release kinetics with
half-lives from 8−82 min by substitution of the SATHA ring. The
pseudo-first-order rate constants of substituted SATOs fit standard linear
free energy relationships (ρ = 1.05), demonstrating a significant
sensitivity to electronic effects.
simplicity and robust nature of oxime and hydrazone-forming
asotransmitters such as hydrogen sulfide (H2S) are
Gendogenous signaling gases that are enzymatically
produced and have specific physiological functions,1 and
delivery of these gasotransmitters can be exploited for
therapeutic applications.2 The therapeutic potential of H2S is
vast, with promising preclinical studies conducted on several
diseases and disorders. For example, H2S exhibits cardiopro-
tection through vasorelaxation,3 suppresses oxidative stress,4
reduces inflammation in the brain,5 and protects the liver in
ischemia-reperfusion events,6 among other effects. Unfortu-
nately, H2S lags behind the other gasotransmitters (NO and
CO) in studies on its physiological role and translational
potential. Versatile H2S-releasing functional groups are needed
to unlock this untapped potential and elevate the status of this
gasotransmitter.
reactions, SATO formation could provide a method for
attaching an H2S-releasing functionality to many types of
compounds under mild conditions. Herein, we report a
synthetic strategy for accessing a variety of H2S-releasing
compounds via a thiooxime click reaction analogous to oxime
formation. Additionally, we explore the reactivity and H2S-
releasing capacity of this unique functional group.
In order to evaluate our hypothesis, we synthesized a series
of SATOs from substituted SATHAs and commercially
available aldehydes and ketones (Scheme 1). SATHAs were
Scheme 1. Synthesis of SATOs
The study of H2S physiology is accomplished through the
use of various H2S organic and electrochemical sensors7 in
conjunction with H2S-releasing compounds (we refer to H2S,
HS−, and S2− collectively here as H2S, although all three species
are in a protonation equilibrium in aqueous solution). Most
studies on H2S physiology and biology have been done with
sulfide salts (Na2S and NaHS). Organic H2S donors have
recently been developed, but most suffer a number of
limitations including structural constraints and uncontrolled
H2S release kinetics.8 The in vivo function of H2S is coupled to
its local concentration;9 therefore, precise control over the rate
of H2S release from donor compounds is paramount.
S-Aroylthiohydroxylamines (SATHAs) have been used in the
synthesis of H2S-releasing N-(benzoylthio)benzamides.8d As
sulfur-containing analogs to acylhydrazides and acylhydroxyl-
amines, thiohydroxylamines are reported to undergo con-
densation reactions with both aldehydes and ketones to form
thiooximes, a functional group that has received little attention
in the literature.10 We hypothesized that S-aroylthiooximes
(SATOs) might react with cysteine to generate H2S in a similar
manner to previously reported compounds.8d Given the
prepared by previously described methods with a range of
substituents.11 Thiooxime formation reactions were conducted
at equimolar concentrations of SATHA and aldehyde/ketone in
CH2Cl2 in the presence of a catalytic quantity of trifluoroacetic
acid and molecular sieves.
The versatility of thiooxime formation was demonstrated by
mixing unsubstituted SATHA (R2 = H) with several aldehydes
and ketones (Table 1) to generate compounds 1a−1j.
Aldehyde-derived SATOs (1a−1g) were synthesized with
high conversions and high isolated yields. Ketone-derived
SATOs (1h−1j) were somewhat more difficult to make.
Condensation of SATHA with a number of aliphatic aldehydes
and ketones was attempted; however, only one aliphatic SATO,
derived from pivalaldehyde (1g), could be isolated due to rapid
Received: December 16, 2013
Published: February 27, 2014
© 2014 American Chemical Society
1558
dx.doi.org/10.1021/ol500385a | Org. Lett. 2014, 16, 1558−1561

Organic Letters
Letter
breakdown or formation of a tetrahedral intermediate at the
aroyl carbon.
Table 1. Conversion and Hydrolysis Kinetics of Substituted
SATOs
To achieve H2S production via Pathway A, the rate of
hydrolysis (Pathway B) must be slow relative to the rate of H2S
release from reaction with cysteine. Hydrolysis of compounds
1a−1j was evaluated using absorbance spectroscopy (Figure
S6). In these experiments, the concentration of the SATOs
relative to their hydrolysis products (SATHA and ketone or
aldehyde) was determined over multiple half-lives. In general,
SATOs hydrolyzed with half-lives in the time scale of days for
compounds 1a−1j excluding 1g (Table 1).
The hydrolysis of SATOs followed first-order kinetics but did
not conform to the expected linear free energy trend.14 The
unsubstituted compound (1a) hydrolyzed faster than the −F,
−COOH, and −OCH3 functionalized compounds (1b, 1c, and
1d). This type of behavior has been observed in other
systems.15 The ketone-derived SATOs hydrolyzed more slowly
and showed a stability order of 1i > 1h > 1j.
a
b
compd
R1, Y
% conversion
t1/2 hydrol. (h)
1a
1b
1c
1d
1e
1f
Ph, H
>99
>99
>99
90
44
p-Ph-F, H
76
p-Ph-COOH, H
p-Ph-OCH3, H
furanyl, H
58
57
>99
>99
74
221
118
1.0
199
266
194
cinnanmyl, H
C(CH3)3, H
Ph, CH3
1g
1h
1i
66
p-Ph-F, CH3
p-Ph-OCH3, CH3
87
1j
37
a
Conversion to SATO as determined by relative integration of
b
To evaluate our ability to control the rate of H2S release from
SATOs, we measured H2S release using two different methods:
(1) a microelectrode-based method that gives real-time data on
H2S concentration and (2) a colorimetric method that is better
suited for longer release periods and was used to determine the
half-life of H2S release for specific compounds (Figure 1).
thiooxime to aldehyde/ketone signals by 1H NMR. Hydrolysis
conducted in 20% (v/v) ACN in PBS (pH = 7.4) at rt.
hydrolysis. We also prepared SATOs with substituents (R2) on
the SATHA ring at near complete conversion.
For the set of ketones studied (1h−1j), conversion increases
with increasing electron withdrawal. These results are
consistent with well-known linear free energy relationships.12
The maximum conversion attained for several substituted
SATOs is summarized in Table 1.
Two pathways can be envisioned for generating H2S from
SATOs (Scheme 2). Pathway A involves addition of the
Scheme 2. Possible Routes of H2S Generation
Figure 1. (A) Representative H2S release curve of 1a measured by the
microelctrode method and (B) kinetics as measured by the methylene
blue spectrophotometric method (10−200 min with intensity
increasing over time). Inset shows the kinetics plot derived from
absorbances measured at 676 nm. Experiments were conducted in
triplicate.
cysteine thiol to the SATO acyl group followed by rapid S→N
acyl transfer in a step similar to native chemical ligation.13 In
this case, the arylidenethiooxime would form, which could
decompose to generate the original ketone or aldehyde along
with H2S and NH3, as has been previously noted for similar
substrates.8d Pathway B describes an initial hydrolysis step,
which would generate the SATHA and the ketone or aldehyde
used to make the SATO. (In preliminary experiments,
unsubstituted SATHA released H2S in the presence of cysteine,
leading to the same products as proposed for Pathway A.)
While both pathways lead to H2S release, Pathway A would
be more desirable for functionalizing substrates to make
prodrugs, where H2S would be released simultaneously with
the drug. Considering the tunable hydrolysis kinetics of the
analogous oxime functional group, Pathway B could be
discouraged by controlling steric and electronic factors of the
SATO bond. Additionally, the rate of H2S release could be
precisely controlled by manipulating the electronics of the
SATHA ring because a likely rate-limiting step would be the
Additionally, the colorimetric method allowed for determi-
nation of H2S release kinetics from compounds that were
insoluble in the conditions required for the microelectrode
experiments (1l−1n).
For the microelectrode experiments, SATOs 1a−1j, 1k, and
1o−1p were dissolved in THF and added to buffered solutions
containing 1 mM cysteine. Upon addition of the SATO, H2S
release was monitored continuously with an H2S-selective
microelectrode (Figure 1A). The H2S concentration reached a
maximum value, which is defined as the peaking time (Tables 2
and 3), after which it began to decrease as the result of
volatilization and oxidation by O2.7
In general, substitution at the para position of the
arylaldehyde/ketone ring (1a−1d) has little effect on H2S
release rate (Table 2). Although the rate of H2S release does
not exhibit a strong dependence on the electronics of the
substituent on the arylaldehyde ring, the rate of hydrolysis is
influenced by substitution at this position. Taken together,
1559
dx.doi.org/10.1021/ol500385a | Org. Lett. 2014, 16, 1558−1561

Organic Letters
Letter
Half-lives for the pseudo-first-order kinetic plots were
determined (Figure 1B inset) and found to be on the order
of 8−82 min. In general, results were similar to the
electrochemical method, with H2S release half-lives trending
based on SATHA ring substituent electronics. A Hammett plot
was constructed to quantify the effect of substituents on the
rate of H2S release (Figure 2). The Hammett plot revealed a
Table 2. H2S Release from Substituted SATOs
a
a
compd
peaking time (min)
compd
peaking time (min)
1a
1b
1c
1d
1e
24
20
18
22
2
2
1
3
1f
1g
1h
1i
26
27
34
34
3
2
2
1
14.8 0.3
1j
33.5 0.9
a
Time after which [H2S] reached a maximum value using an H2S
sensitive microelectrode. Studies were conducted at 40 μM SATO in
0.1% THF in PBS buffer (pH = 7.4) at rt in the presence of 1 mM
cysteine.
Table 3. H2S Release from Substituted SATOs
Figure 2. Hammett plot of the rate of H2S release relative to 1a vs the
σ value of the p-substituent (ρ = 1.05, R2 = 0.986).
strong dependence of the rate of H2S release on the σ values of
the SATHA substituents (ρ = 1.05). It follows from Figure 2
that the H2S release kinetics of SATOs can be finely tuned. The
strong correlation between H2S release rate and σ suggests that
the rate of H2S release for other SATOs could be predicted
from the Hammett plot. These data represent an unprece-
dented level of control over the rate of H2S release.
The fast reaction between SATOs and cysteine to generate
H2S (t1/2 ≈ 8−82 min) compared with their hydrolysis rate
(t1/2 ≈ 45−250 h) rules out pathway B as the operative H2S
release mechanism for most SATOs under the conditions
tested. To study the mechanism of H2S release, the products of
the reaction between 1a and cysteine were analyzed. Based on
the compounds isolated by HPLC, we propose that H2S may
be generated according to Scheme 3. First, reversible thiol
a
b
R2
peaking time (min)
t1/2 H2S release (min)
H (1a)
24
21
2
2
34
29
18
11
8
6
2
1
1
F (1k)
c
Cl (1l)
−
c
CF3 (1m)
CN (1n)
CH3 (1o)
OCH3 (1p)
−
c
−
1
33.7 0.4
52.9 1.4
82 11
36
3
a
Time after which [H2S] reached a maximum value detected by an
H2S sensitive microelectrode. Studies were conducted at 40 μM SATO
in 0.1% THF in PBS buffer (pH = 7.4) at rt in the presence of 1 mM
cysteine. Half-lives of release were measured by using the methylene
blue assay at 100 μM SATO in 30% THF in PBS buffer (pH = 7.4) at
b
c
rt in the presence of 1 mM cysteine. Samples were not soluble in the
media for the microelectrode probe assay.
a
Scheme 3. Proposed Mechanism of H2S Release
these data suggest that the stability of SATOs can be tuned
without drastically altering the H2S release profile.
We also tested several other molecules as triggers for H2S
release (Figure S2). Glutathione was shown to trigger H2S
release at levels similar to the case of cysteine, while N-
acetylcysteine was found to trigger H2S generation at a slower
rate than cysteine. However, neither lysine nor serine triggered
H2S release, nor was H2S generation observed in aqueous
solutions of SATO in the absence of thiol functionality.
Additionally, H2S release was evaluated at different cysteine
concentrations. In general, higher concentrations of cysteine
resulted in an increased H2S release rate (Figure S3). H2S
generation was also observed in plasma (Figure S5).
The colorimetric method of measuring the concentration of
H2S involves the H2S-promoted conversion of N,N-dimethyl-p-
phenylenediamine into methylene blue,16 which can be
measured by its peak absorbance at 676 nm (Figure 1B). As
real-time detection is not possible with the methylene blue
method, time points are taken at various intervals for each
compound. We note that the methylene blue method can
overestimate H2S levels, so we make no attempt here to
translate our measurements to actual H2S concentrations.17
a
Boxed products were either isolated from the reaction mixture or
detected indirectly.
exchange between the SATO and cysteine gives the arylidene-
thiooxime along with N-benzoylated cysteine after a S → N acyl
transfer step. Reaction of a second equivalent of cysteine with
the arylidenethiooxime generates thiocysteine, ammonia, and
the original aldehyde. Thiocysteine then reacts with another
equivalent of cysteine to generate cystine and H2S.18 Further
mechanistic work is needed to evaluate whether this is the
1560
dx.doi.org/10.1021/ol500385a | Org. Lett. 2014, 16, 1558−1561

Organic Letters
Letter
(8) (a) Kodela, R.; Chattopadhyay, M.; Kashfi, K. ACS Med. Chem.
Lett. 2012, 3, 257−262. (b) Devarie-Baez, N. O.; Bagdon, P. E.; Peng,
B.; Zhao, Y.; Park, C.-M.; Xian, M. Org. Lett. 2013, 15, 2786−2789.
(c) 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, e21077. (d) Zhao, Y.;
Wang, H.; Xian, M. J. Am. Chem. Soc. 2011, 133, 15−17.
(9) Wallace, J. L. Trends Pharmacol. Sci. 2007, 28, 501−505.
(10) Barton, D. H. R.; Magnus, P. D.; Pennanen, S. I. J. Chem. Soc.,
Chem. Commun. 1974, 1007.
(11) Raasch, M. S. J. Org. Chem. 1972, 37, 3820−3823.
(12) Hammett, L. P. J. Am. Chem. Soc. 1937, 59, 96−103.
(13) Dawson, P. E. Israel J. Chem. 2011, 51, 862−867.
(14) Gregory, B. J.; Moodie, R. B. J. Chem. Soc. B 1970, 862−866.
(15) Stein, A. R.; Tencer, M.; Moffatt, E. A.; Dawe, R.; Sweet, J. J.
Org. Chem. 1980, 45, 3539−3540.
(16) Siegel, L. M. Anal. Biochem. 1965, 11, 126−132.
(17) Olson, K. R. BBA-Bioenergetics. 2009, 1787, 856−863.
(18) Mueller, E. G. Nat. Chem. Biol. 2006, 2, 185−194.
(19) Ali, S.; Khan, A.; Ahmed, I.; Musaddiq, M.; Ahmed, K.; Polasa,
H.; Rao, L.; Habibullah, C.; Sechi, L.; Ahmed, N. Ann. Clin. Microbiol.
Antimicrob. 2005, 4, 1−7.
(20) Cabello, C. M.; Bair Iii, W. B.; Lamore, S. D.; Ley, S.; Bause, A.
S.; Azimian, S.; Wondrak, G. T. Free Radical Biol. Med. 2009, 46, 220−
231.
operative mechanism for H2S release. See the Supporting
Information for further discussion.
Regardless of the pathway of H2S release, the parent
aldehyde/ketone is regenerated in the process. This implies
that SATHA can be conjugated to aldehyde- or ketone-bearing
therapeutic agents such as cinnamaldehyde (as in 1f), known to
possess antimicrobial19 and anticancer20 properties, to impart
tandem physiological activity. However, the requirement for a
ketone or aldehyde limits the types of drugs amenable to this
type of derivatization to those bearing these functionalities. H2S
donor-drug hybrids such as NOSH-aspirin8a and H2S-releasing
derivatives of diclofenac9 have been reported and in many cases
show improved physiological responses compared with their
parent drugs.
In summary, we have synthesized a series of SATOs from
SATHAs and common aldehydes and ketones. We showed that
SATOs are relatively stable in aqueous solution at physiological
pH, and hydrolysis of these compounds can be controlled to a
degree by tuning the steric and electronic factors of the SATO
bond. We demonstrated that the half-life of H2S release could
be varied between 8 and 82 min simply by changing the
substituent on the SATHA ring. Given the ease of installation
of this versatile functional group, we envision that SATOs could
open new avenues for the study of H2S biology and could
enable a new generation of H2S-releasing therapeutics and H2S-
drug conjugates.
ASSOCIATED CONTENT
* Supporting Information
■
S
Synthetic details, characterization data, and kinetics plots. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: jbmatson@vt.edu.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the Institute for Critical
Technology and Applied Science (JFC12-256). We thank
Prof. Richard D. Gandour (Virginia Tech) for helpful
discussions and Prof. Amanda J. Morris (Virginia Tech) and
Prof. Tijana Z. Grove (Virginia Tech) for use of instruments.
REFERENCES
■
(1) Mustafa, A. K.; Gadalla, M. M.; Snyder, S. H. Sci. Signal. 2009, 2,
re2.
(2) (a) Motterlini, R.; Otterbein, L. E. Nat. Rev. Drug. Discov. 2010, 9,
728−743. (b) Carpenter, A. W.; Schoenfisch, M. H. Chem. Soc. Rev.
2012, 41, 3742−3752. (c) Wang, R. Physiol. Rev. 2012, 92, 791−896.
(3) Hosoki, R.; Matsuki, N.; Kimura, H. Biochem. Biophys. Res.
Commun. 1997, 237, 527−531.
(4) Kimura, Y.; Dargush, R.; Schubert, D.; Kimura, H. Antioxid.
Redox Signaling. 2006, 8, 661−670.
(5) Lee, M.; Sparatore, A.; Del Soldato, P.; McGeer, E.; McGeer, P.
L. Glia. 2010, 58, 103−113.
(6) Jha, S.; Calvert, J. W.; Duranski, M. R.; Ramachandran, A.; Lefer,
D. J. Am. J. Physiol. 2008, 295, H801−H806.
(7) Hughes, M. N.; Centelles, M. N.; Moore, K. P. Free Radical Biol.
Med. 2009, 47, 1346−1353.
1561
dx.doi.org/10.1021/ol500385a | Org. Lett. 2014, 16, 1558−1561