Tabun scavengers based on hydroxamic acid containing cyclodextrins

Article abstract of DOI:10.1039/c3cc41290c

Arrangement of several hydroxamic acid-derived substituents along the cavity of a cyclodextrin ring leads to compounds that detoxify tabun in TRIS-HCl buffer at physiological pH and 37.0 °C with half-times as low as 3 min.

Full text of DOI:10.1039/c3cc41290c

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Tabun scavengers based on hydroxamic acid
containing cyclodextrins†
Florian Brandhuber,a Michael Zengerle,b Luzian Porwol,b Anne Bierwisch,a
Marianne Koller,a Georg Reiter,a Franz Woreka and Stefan Kubik*b
Cite this: Chem. Commun., 2013,
49, 3425
Received 19th February 2013,
Accepted 6th March 2013
DOI: 10.1039/c3cc41290c
www.rsc.org/chemcomm
Arrangement of several hydroxamic acid-derived substituents development of low molecular weight scavengers are cyclodextrins
along the cavity of a cyclodextrin ring leads to compounds that because of their non-toxicity and well-known ability to mediate the
detoxify tabun in TRIS-HCl buffer at physiological pH and 37.0 8C hydrolysis of esters including those of phosphoric acid.6–8 Indeed,
with half-times as low as 3 min.
b-cyclodextrin derivatives with oxime-derived substituents along the
cavity have been shown to cause rapid detoxification of cyclosarin
Tabun (ethyl dimethylphosphoramidocyanidate, GA) is a highly (within seconds under the conditions of the assay).9 The activity is
toxic organophosphate developed in 1936 and manufactured believed to benefit from inclusion of the cyclohexyl residue of the OP
although never used as a chemical warfare agent during World into the cyclodextrin cavity, which brings the cyclosarin phosphor-
War II. The toxic effects of GA are mainly related to the inhibition of ous atom in close proximity to the appended nucleophilic group.
acetylcholinesterase (AChE), the enzyme hydrolysing the neuro-
The effects of substituted cyclodextrin derivatives on GA degrada-
transmitter acetylcholine, by covalent modification of a serine tion are usually smaller. A b-cyclodextrin containing one 3-carboxyl-4-
residue in the protein’s active site.1 The mode of action of GA is iodosobenzyloxy residue in the 2-position of a glucose subunit
thus closely related to that of other organophosphorus-derived detoxifies GA in TRIS-HCl buffer (pH 7.40) at 37 1C with a half-time
nerve agents such as sarin, soman, or VX.2
of 8.4 min, for example.8 Under similar conditions, a substituted
Poisonings with organophosph(on)ates (OP) are typically treated b-cyclodextrin derivative containing an oxime group along the
by administration of atropine and certain oximes.3 Atropine narrow opening has been shown by us to degrade GA with a half-
antagonises the action of acetylcholine at muscarinic receptors time of 10.3 min.10 With a ca. six-fold acceleration over the rate of
while the oximes reactivate inhibited AChE by cleaving the ester spontaneous GA degradation the activity of this compound is far too
on the serine hydroxyl group. Because phosphonates formed low to render its use as a scavenger for the treatment of GA
with, for example, sarin or VX are more easily cleaved by oximes poisonings a realistic prospect.
than the phosphoramidate formed by GA, therapeutic effects of
oximes are smaller in the case of GA poisonings.2
Here we show that hydroxamic acids detoxify GA more efficiently
than oximes. Although their ability to degrade OPs is known for
some time,11 hydroxamic acids have, to the best of our knowledge,
not been used as nucleophilic groups in cyclodextrin-derived OP
scavengers so far. According to our results these groups can lead to
scavengers that detoxify GA at pH 7.40 and 37.0 1C with a half-time
of ca. 3.0 min when appropriately arranged along the b-cyclodextrin
cavity, which is a substantial improvement over previous systems.
While screening for cyclodextrin-based OP scavengers we came
across b-cyclodextrin derivative b61, which exhibited a stronger effect
on GA degradation than the previously prepared oximes.10 With this
new lead structure in hand we performed structural variations by
arranging the hydroxamic acid residue on the secondary side of the
b-cyclodextrin ring (b21, b31),‡ increasing the number of substituents
(b612, b617), varying the cyclodextrin diameter (a616, c618, see ESI† for
structures), changing the structure of the hydroxamic acid (b62), or
introducing methyl groups on one or both heteroatoms of the
–NHOH group (b63, b64, b65, b652, b657). In addition, glucose
derivatives G61 and G65 have also been synthesised to elucidate
An alternative strategy to treat OP poisonings comprises degrada-
tion of the nerve agent prior to its reaction with AChE. In this
context, scavengers on the basis of hydrolytically active enzymes
have been developed4 as well as those based on low molecular
weight compounds.5 Particular interesting lead structures for the
a
¨
Institut fur Pharmakologie und Toxikologie der Bundeswehr, Neuherbergstraße 11,
¨
D-80937 Munchen, Germany
b
¨
Fachbereich Chemie – Organische Chemie, Technische Universitat Kaiserslautern,
¨
Erwin-Schrodinger-Straße, D-67663 Kaiserslautern, Germany.
E-mail: kubik@chemie.uni-kl.de; Fax: +49-631-205-3921
† Electronic supplementary information (ESI) available: Synthetic procedures and
experimental details and results of the activity and NMR assays. See DOI: 10.1039/
c3cc41290c
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 3425--3427 3425

View Article Online
Communication
ChemComm
Not surprisingly, increasing the number of substituents on the
b-cyclodextrin ring to 2 in b612 and 7 in b617 also yields active
scavengers as does appending hydroxamic acid 1 to a- and
g-cyclodextrin as in a616 and c618. Compound b62 decomposes GA
only slowly indicating that the direct attachment of the hydroxamic
acid group to the triazole moiety is crucial for activity. Methylation of
the nitrogen and the oxygen atom of the –NHOH group (b63) or of
only the oxygen atom (b64) yields completely inactive compounds
clearly demonstrating that the OH group of the hydroxamic acid is
involved in the degradation mechanism. Derivatives with a methyl
group on the hydroxamic acid nitrogen atom (b65, b652, b657) exhibit
activity, however. Finally, it is important to point out that the glucose
derivatives G61 and G65 are also rather active, suggesting that the
cyclodextrin ring is not essential for GA degradation.
These results provided qualitative information about the
activities of the whole set of investigated compounds showing
that many of them possess promising properties. Exact rates of
GA degradation were subsequently determined for compounds
that exhibited restoration of more than 75% of AChE activity
after 30 min of incubation in the qualitative assay. Quantitative
measurements involved incubation of GA (1.0 mM in aqueous
TRIS-HCl buffer at pH 7.40 and 37.0 1C) with a hydroxamic acid
(500 mM) and determination of the inhibitory activity of this
the effect of the cyclodextrin subunit on the degradation rate. All solution on AChE at defined intervals. After referencing the
compounds were prepared from known azides and respective results to a control experiment, in which GA was incubated in
terminal alkynes by using the copper(I)-catalysed Huisgen cyclo- the absence of the scavenger, pseudo first order rate constants
addition and were isolated in analytically pure form (see ESI†).
The initial test to assess the effect on GA degradation involved constants and half times thus obtained are shown in Table 1.
incubation of the OP with an excess of a potential scavenger in Table 1 shows that all hydroxamic acids accelerate GA degradation
kobs were calculated from the resulting decay curves.12 Rate
TRIS-HCl buffer (pH 7.40) at 37.0 1C and estimation after 0, 30, and significantly beyond the rate of spontaneous hydrolysis. With a half
60 min to what extent aliquots of these incubated samples inhibit time of 8.6 min b61 is almost 2 min more active than our best oxime-
AChE.9,10,12 The inhibitory activity was related to the activity of a containing b-cyclodextrin.10 Moving the hydroxamic acid to the
GA solution incubated in the absence of the test compounds. The secondary face of the ring causes a reduction in efficiency. Since
resulting parameter Dk1 correlates with scavenger activity: higher there are no indications that the activity of b61 is positively influenced
values for Dk1 indicate higher activity. The results obtained are by specific interactions between the cyclodextrin ring and the GA
summarised in Fig. 1.
molecule because glucose derivative G61 is practically as active as b61,
Fig. 1 shows that native b-cyclodextrin does not detoxify GA we attribute the lower activities of b21 and b31 to a more difficult
within the time frame of the assay. In contrast, all derivatives accessibility of the hydroxamic acid group in these compounds.
containing hydroxamic acid 1 are active, almost completely
Compound b612 is more active than the monosubstituted
degrading the OP within 30 min. The position of the substituent analogue, consistent with the fact that the number of nucleo-
on the cyclodextrin ring has obviously no large effect on degradation philic centres in solution increases by replacing b61 with an
efficiency: compounds b61, b21, and b31 are all similarly active.
Table 1 Kinetic constants kobs and half-times t1/2 of GA degradation mediated
by selected cyclodextrin and glucose derivativesa
kobs  102/minÀ1
t1/2/min
Spontaneous hydrolysis
1.15 Æ 0.04
8.04 Æ 0.11
6.28 Æ 0.14
6.62 Æ 0.30
13.55 Æ 0.33
22.76 Æ 1.08
17.88 Æ 0.34
16.05 Æ 1.06
5.63 Æ 0.04
9.06 Æ 0.29
8.03 Æ 0.13
8.10 Æ 0.25
60.3 Æ 0.2
8.6 Æ 0.1
11.0 Æ 0.3
10.5 Æ 0.5
5.1 Æ 0.1
3.0 Æ 0.1
3.9 Æ 0.1
4.3 Æ 0.3
12.3 Æ 0.1
7.6 Æ 0.3
8.6 Æ 0.1
8.6 Æ 0.3
b61
b21
b31
b612
b617
a616
c618
b65
Fig. 1 Bar chart indicating the ability of the investigated hydroxamic acids to
mediate GA degradation. Large columns indicate high activity. For reference the
results obtained for native b-cyclodextrin (b-CD) are also shown. All results are
means of n = 3 experiments. Standard deviation of a series of measurements with
the same batch of AChE is o10%. Because the results can vary more strongly if
different enzyme preparations are used we assume an overall error of 20%.
b657
G61
G65
a
In aqueous TRIS-HCl buffer at pH 7.40 and 37.0 1C, c(GA) = 1.0 mM,
c(scavenger) = 500 mM.
c
3426 Chem. Commun., 2013, 49, 3425--3427
This journal is The Royal Society of Chemistry 2013

View Article Online
ChemComm
Communication
equimolar amount of b612. Raising the number of substituents on structure analogous to B is most likely the stable end product in
the ring even further causes an additional improvement in activity this reaction.
although the extent to which the activity of b617 increases does not
In conclusion, we have shown that appending hydroxamic acid-
correlate linearly with the number of substituents. Since an approx- derived substituents to cyclodextrin or glucose units affords potent
imate seven-fold increase in reaction rate was observed when the scavengers for GA. Cyclodextrins with one hydroxamic acid residue
concentration of the glucose derivative G61 was increased from do not possess higher activity than corresponding glucose deriva-
500 mM to 3500 mM in the assay (see ESI†) it seems that not all of tives, presumably because the cyclodextrin ring does not coopera-
the seven substituents in b617 are able to participate in GA degrada- tively contribute to the mode of action. Cyclodextrins represent,
tion. Interestingly, the a- and g-cyclodextrin derivatives a616 and c618 however, valuable scaffolds allowing introduction of several hydro-
are somewhat less active than b617 although the number of nucleo- xamic acid units. This strategy allows improvement of degradation
philic centres is even higher in the case of c618. The substituents are efficiency while keeping the scavenger concentration constant. The
therefore presumably better arranged for reaction with GA in b617 results also indicate that our hydroxamic acids are irreversibly
than in the larger or smaller cyclodextrin analogues. While the modified when reacting with GA with the course of the reaction
N-methylated glucose derivative G65 is as active as G61, the corres- depending on the actual nature of the acid. Work to elucidate the
ponding cyclodextrin derivatives b65 and b657 are less active than the underlying mechanisms is ongoing to obtain information on how to
non-methylated analogues.
improve the activity of these promising scavengers even further.
This research was supported by a contract of the German
GA degradation mediated by b617 was additionally followed
by using a GC-MS assay, which provided information about the Armed Forces (E/UR3G/9G127/9A803). The generous funding is
enantioselectivity of the reaction.13 This experiment yielded gratefully acknowledged. We also thank Wacker Chemie AG,
further evidence that interactions between the cyclodextrin ring Burghausen, Germany, for a gift of unsubstituted cyclodextrins.
and GA play no large role in the mode of action of b617 because
Notes and references
the two enantiomers of GA decompose with practically the
same rate as observed for oxime-containing cyclodextrins.10
‡ Note that the substituent is located in a mannose unit in b21 and in
The half time of 2.0 Æ 0.1 min obtained in the GC-MS assay an altrose unit in b31. The presence of subunits in these compounds
for the b617-mediated GA degradation is in reasonable agree-
ment with the result of the enzymatic assay.
differing from glucose is due to the synthetic approach used for
preparation.
Repeated treatment of b61 or b65 with an excess of GA revealed
1 K. Tuovinen, Toxicology, 2004, 196, 31–39; T. C. Marrs, Pharmacol.
Ther., 1993, 58, 51–66; T. C. Marrs, R. L. Maynard and F. R. Sidell,
that both hydroxamic acids lose activity after the first round of
Chemical Warfare Agents: Toxicology and Treatment, John Wiley &
reaction.10 To obtain information about the nature of the products
Sons, Chichester, 2007.
2 F. Worek, H. Thiermann, L. Szinicz and P. Eyer, Biochem. Pharma-
col., 2004, 68, 2237–2248.
formed we followed GA degradation in TRIS-HCl buffer in the
absence and presence of G61 and G65 using 31P NMR spectroscopy
and mass spectrometry. These measurements showed that sponta-
neous hydrolysis of GA in TRIS-HCl buffer practically exclusively
yields hydroxytabun A. When 2 equiv. of G61 are present the OP not
only disappears significantly more rapidly, no unreacted GA could
be detected in the first NMR spectrum recorded ca. 30 min after
sample preparation, but also the outcome of the reaction is different.
Under these conditions, formation of a single product was observed,
which represents phosphate diester B according to mass spectro-
metry lacking the CN and the NMe2 group of GA. This product is
expected to be relatively resistant towards further hydrolysis because
of its anionic nature at physiological pH. Thus, reaction between G61
and GA results in irreversible covalent modification of the hydro-
xamic acid explaining why one scavenger molecule can induce
degradation of only a single molecule of GA.
3 J. A. Vale, P. Rice and T. C. Marrs, in Chemical Warfare Agents:
Toxicology and Treatment, ed. T. C. Marrs, R. L. Maynard and
F. R. Sidell, John Wiley & Sons, Chichester, 2007, pp. 249–260.
4 D. E. Lenz, D. Yeung, J. R. Smith, R. E. Sweeney, L. A. Lumley and
D. M. Cerasoli, Toxicology, 2007, 233, 31–39; P. Masson, F. Nachon,
C. A. Broomfield, D. E. Lenz, L. Verdier, L. M. Schopfer and
O. Lockridge, Chem.-Biol. Interact., 2008, 175, 273–280.
5 H. Morales-Rojas and R. A. Moss, Chem. Rev., 2002, 102, 2497–2522;
B. M. Smith, Chem. Soc. Rev., 2008, 37, 470–478.
6 L. Marinescu and M. Bols, Trends Glycosci. Glycotechnol., 2009, 21,
309–323; R. Breslow and S. D. Dong, Chem. Rev., 1998, 98, 1997–2011.
`
7 N. Masurier, F. Estour, M. T. Froment, B. Lefevre, J.-C. Debouzy,
B. Brasme, P. Masson and O. Lafont, Eur. J. Med. Chem., 2005, 40,
615–623; S. Mu¨ller, M. Koller, R. Le Provost, O. Lafont, F. Estour,
T. Wille, H. Thiermann, F. Worek and G. Reiter, Toxicol. Lett., 2011,
200, 53–58; R. Le Provost, T. Wille, L. Louise, N. Masurier, S. Mu¨ller,
G. Reiter, P.-Y. Renard, O. Lafont, F. Worek and F. Estour, Org.
Biomol. Chem., 2011, 9, 3026–3032.
8 T. Wille, O. Tenberken, G. Reiter, S. Mu¨ller, R. Le Provost, O. Lafont,
F. Estour, H. Thiermann and F. Worek, Toxicology, 2009, 265, 96–100.
9 M. Zengerle, F. Brandhuber, C. Schneider, F. Worek, G. Reiter and
S. Kubik, Beilstein J. Org. Chem., 2011, 7, 1543–1554.
10 F. Brandhuber, M. Zengerle, L. Porwol, O. Tenberken, H. Thiermann,
F. Worek, S. Kubik and G. Reiter, Toxicology, 2012, 302, 163–171.
11 B. E. Hackley Jr, R. Plapinger, M. Stolberg and T. Wagner-Jauregg,
J. Am. Chem. Soc., 1955, 77, 3651–3653; R. Swidler and G. M. Steinberg,
ˇ
J. Am. Chem. Soc., 1956, 78, 3594–3598; L. Louise-Leriche, E. Paunescu,
´
G. Saint-Andre, R. Baati, A. Romieu, A. Wagner and P.-Y. Renard,
Chem.–Eur. J., 2010, 16, 3510–3523.
The course of the reaction between G65 and GA is somewhat 12 T. Wille, H. Thiermann and F. Worek, Toxicol. In Vitro, 2010, 24,
1026–1031.
more complex as it involves several transitional species (see ESI†).
Although not all aspects of the underlying transformations have
13 O. Tenberken, J. Mikler, I. Hill, K. Weatherby, H. Thiermann, F. Worek
and G. Reiter, Toxicol. Lett., 2010, 198, 177–181; O. Tenberken, F. Worek,
been fully elucidated yet our results indicate that diester C with a
H. Thiermann and G. Reiter, J. Chromatogr., B, 2010, 878, 1290–1296.
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 3425--3427 3427