An oral redox-sensitive self-immolating prodrug strategy

Article abstract of DOI:10.1039/c5cc00405e

We report a novel oral prodrug approach where a solubilizing polymer conjugated to the drug is designed to be released by the action of an exogenously administered agent in the intestine. A redox-sensitive self-immolating design was implemented, and the reconversion kinetics were studied for three reducible prodrugs.

Full text of DOI:10.1039/c5cc00405e

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
An oral redox-sensitive self-immolating prodrug
strategy†
Cite this: Chem. Commun., 2015,
51, 5721
Tao Sun, Andrea Morger, Bastien Castagner*‡ and Jean-Christophe Leroux*
Received 15th January 2015,
Accepted 20th February 2015
DOI: 10.1039/c5cc00405e
www.rsc.org/chemcomm
We report a novel oral prodrug approach where a solubilizing
polymer conjugated to the drug is designed to be released by the
action of an exogenously administered agent in the intestine.
A redox-sensitive self-immolating design was implemented, and
the reconversion kinetics were studied for three reducible prodrugs.
The oral delivery of poorly water-soluble drugs is a critical
challenge in pharmaceutical sciences. Approximately 40% of
drugs are poorly water-soluble and this number is believed to
be much higher for candidates currently in development.1 Over
the years, a host of formulations, such as solubilizing adjuvants
(co-solvents, surfactants, and complexing agents) or molecular
dispersion, have been developed to increase solubility and
dissolution rates, which are key steps in the absorption process.2
Most reported strategies rely on supramolecular interactions
between drugs and adjuvants, which may suffer from stability
issues such as disassembly and precipitation upon dilution
when encountering the variable and harsh conditions of the
gastro-intestinal (GI) tract.
Scheme 1 The (a) phosphate ester3a and (b) proposed redox-sensitive
self-immolative prodrug strategies for orally administered poorly water-
soluble drugs.
Prodrugs are commonly used to overcome a drug’s unfavour-
able intrinsic physicochemical properties.3 Surprisingly, only
few prodrug strategies have been described to enhance the oral
bioavailability of poorly water soluble drugs. The most advanced of the intact prodrug in the feces.4 The release kinetics greatly
approach is the phosphorylation of a hydroxyl group of the depends on the structure of the drug and there is inter-patient
drug.3a The resulting phosphate ester undergoes reconversion variability in alkaline phosphatase levels in the GI tract.5 Further-
by the action of alkaline phosphatase at the intestinal brush more, this strategy is limited to hydroxyl group-containing drugs.
border membrane, followed by absorption of the released hydro- More recently, redox-sensitive sulfenamide prodrugs of amide-
phobic compound (Scheme 1a).4 This strategy requires a careful containing drugs have been described.6 However, due to the
control over the release rate of the free drug at the intestinal requirement of enterocyte surface-based reconversion, this
barrier since fast reconversion will lead to precipitation of the approach relies on endogenous free-thiol-containing protein
free drug, while slow reconversion will result in the elimination extending from the apical surface of the cell, which will vary
between patients. Ideally, an oral prodrug strategy for poorly
Institute of Pharmaceutical Sciences, Department of Chemistry and Applied
Biosciences, ETH Zurich, Vladimir-Prelog-Weg 3, 8093 Zurich, Switzerland.
E-mail: jleroux@ethz.ch; Fax: +41 44 633 13 14
soluble drugs would allow control over the reconversion rates in
the GI tract.
Here, we report a redox-sensitive oral prodrug platform
designed to undergo reconversion under the action of a
co-administered agent (Scheme 1b). Though the redox-triggered
release in cytosol based on disulphide bonds is a prevalent
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5cc00405e
‡ Present address: Department of Pharmacology & Therapeutics, McGill Univer-
sity, 3655 Promenade Sir-William-Osler, Montreal (Quebec) H3G 1Y6, Canada.
E-mail: bastien.castagner@mcgill.ca.
This journal is ©The Royal Society of Chemistry 2015
Chem. Commun., 2015, 51, 5721--5724 | 5721

View Article Online
Communication
ChemComm
strategy adopted in cancer therapy,7 such an approach is still presence of 0.5 mM NAC. The reconversion was found to be
relatively novel for the design of oral prodrugs. In the GI tract, the slower at pH 4.5, requiring 10 mM NAC for complete reconver-
redox potential is not buffered and most thiol reducing agents sion in 90 min, compared to 0.5 mM NAC at pH 6.8. This is most
come from our diet.8 Thus, we chose a disulphide trigger in order likely due to the slow thiol exchange reaction at acidic pH. This
to allow cleavage by the safe dietary supplement N-acetyl-cysteine is desirable, to avoid prodrug reconversion in the acidic environ-
(NAC). The use of an exogenous reconversion agent should allow ment of the stomach, where little drug absorption is expected to
more control over the reconversion rates and reduce inter-patient take place. These results show that the reconversion should
variability. A poly(ethylene glycol) (PEG) moiety provides water mostly occur under pH conditions encountered in the small
solubility and its cleavage in the small-intestine allows the intestine and that it is modulated by NAC. Prodrug surrogate 1
release and concomitant absorption of the free drug. The design was found to be unstable at pH 1.2 (pH value of empty stomach,
and synthesis of the prodrug are presented here, as well as the Fig. S3, ESI†). Because thiol exchange reactions are greatly
reconversion kinetics of the prodrug as a function of NAC under suppressed at this pH, this reconversion is most likely due to
conditions mimicking the GI tract. To explore the scope of the direct hydrolysis of the activated carbonate group.17 The
this strategy, we synthesized the prodrugs of 7-ethyl-10-hydroxy- spectrophotometry results were also supported by 1H NMR data
camptothecin (SN-38)9 and mitomycin C (MMC),10 both anti- by monitoring the peaks of the released pNP, where a similar
cancer agents, as well as phenytoin, an anticonvulsant.11
trend was obtained (t1/2 = 66 min with 10 mM NAC at pH 4.5
We started by selecting a linker with rapid self-immolation (38 min by UV), and 17 min with 0.2 mM NAC at pH 6.8 (32 min
kinetics since the success of this strategy relies on a rate limiting by UV), Fig. S4, ESI†).
disulphide cleavage. The self-immolation process is generally
With NAC-dependent and tunable reconversion kinetics of
driven by cyclisation,12 elimination reactions,13 or a combination the prodrug surrogate 1, we turned to potential drug candi-
of both.14 Compared with the cyclisation strategy, the 1,6-benzyl dates. Drugs with poor solubility but high permeability (class II
elimination has the advantage of being more independent and drugs in the biopharmaceutical classification system) are
less sensitive to the drug structure. Besides, the self-immolation suited for an oral application, because they can be readily
is rapid in aqueous solution.13a This linker strategy should also absorbed once released into the GI tract.3a In order to probe
be applicable to a variety of hydroxyl and amine-containing the scope of our strategy, we selected drugs with a phenol
drugs via a carbonate or carbamate moiety. PEG was chosen as (SN-38), an aliphatic alcohol (pro-phenytoin), and an amine
the solubilizing group because it is safe, inert, and commercially functional group (MMC). The key intermediate 3 containing the
available. A PEG chain of 5000 Da has a sufficiently large size to disulphide bond was synthesized by a thiol exchange reaction
improve the solubility of common hydrophobic drugs.
of 4-(mercaptophenyl)methanol and 2,20-dithiobispyridine.
We started by preparing model 1, where p-nitrophenol ( pNP) All prodrugs were then easily obtained by coupling the drug
was used as a drug surrogate for rapid reconversion kinetic to alcohol 3 via a carbonate or carbamate bond, followed by
measurement in aqueous buffer solutions using UV absorbance substitution of the 2-thiopyridyl group with the PEG-thiol
(Fig. 1, Fig. S1 and S2, ESI†). The kinetic study was carried out solubilizing group (Scheme 2). Alternatively, carbamate 7 could
in buffers at pH 4.5 (pH value of postprandial stomach)15 and also be obtained in high yield (95%) by direct coupling of MMC
6.8 (pH value of proximal small-intestine).16 At the intestinal pH, with activated carbonate 4, thus avoiding the use of phosgene.
the prodrug was relatively stable in the absence of NAC (ca. 20% The final prodrugs were purified by precipitation in diethyl
release over 4 h) but was completely cleaved in 2 h in the ether followed by silica gel chromatography.
We first measured the solubility of the three prodrugs in
water. As anticipated, the PEGylated prodrug increased the
water solubility of SN-38 from 0.028–0.082 mM18 to 17.6 mM
(216–629 folds), MMC from 2.7 mM19 to 28 mM (410 folds),
and phenytoin from 0.079 mM20 to 21 mM (4260 folds).
The reconversion kinetics of the SN-38 prodrug 6 was first
studied in buffered solution at pH 4.5 and 6.8 (Fig. 2a and b).
At both pH, the reconversion rates were correlated with the
concentration of NAC in the medium. As expected from the
prodrug surrogate 1, the reconversion was significantly slower
at pH 4.5.
We then sought to study the reconversion kinetics in a
medium that would be more representative of the harsh gut
environment. The reconversion of the prodrug was therefore
assessed in simulated gastric fluid (SGF) consisting of a mixture
of pepsin enzymes in acetate buffer at pH 4.5 (Fig. 2c). Overall,
the release kinetics of 6 was only slightly faster in SGF compared
Fig. 1 (a) Prodrug surrogate 1 reconversion in the presence of NAC.
to the enzyme-free medium, indicating that the prodrug would
survive the gastric environment. We then studied the reconversion
(b) Reconversion kinetics at various NAC concentrations at 37 1C. Values
are represented as mean ꢀ SD (n = 3).
5722 | Chem. Commun., 2015, 51, 5721--5724
This journal is ©The Royal Society of Chemistry 2015

View Article Online
ChemComm
Communication
in simulated intestinal fluid (SIF) containing the pancreatin
enzyme mixture at pH 6.8 (Fig. 2d). The release of SN-38 again
nicely correlated with the concentration of NAC. Taking into
consideration transit time through the small intestine, an ideal
release half-life would be around 1 h, corresponding to 0.5 mM
NAC.21 The background reconversion rate in the absence of NAC
increased in SIF compared to enzyme-free medium to reach
around 50% in 3 h. This is most likely due to hydrolysis of the
carbonate group in SIF. In order to delineate the contributions
of the enzymes present in pancreatin, we studied the stability of
compound 1 that contains an activated carbonate in the presence
of various isolated pancreatin enzymes. The results show that
lipase and trypsin are mostly responsible for the cleavage in SIF
(Fig. S5, ESI†). The NAC-mediated reconversion was still domi-
nant for prodrug 6 at 0.5 mM of NAC, but these results underscore
the need to properly monitor the prodrug stability in SIF.
In order to study the influence of the drug on the release
kinetics, we explored reconversion of MMC prodrug 8 at
selected NAC concentrations (Fig. 3a and b). In the stomach
environment (Fig. 3a), the reconversion was slow, both in the
presence (1 mM) and absence of NAC. In contrast, the reconversion
in the intestinal environment (Fig. 3b) occurred in the presence of
0.1 mM NAC (t1/2 = 60 min). Interestingly, the addition of GI
enzymes did not influence the rate of release in both environments,
in line with the expected enhanced stability of the carbamate,
compared to carbonate towards enzymatic degradation.
Finally, the release kinetics of the primary alcohol-containing
phenytoin prodrug 10 was also studied (Fig. 3c and d). Again,
very little release was observed at pH 4.5 or SGF in the presence
Scheme 2 Synthesis of the SN-38, MMC, and phenytoin prodrugs (6, 8,
and 10). Reagents and conditions: (a) 2,20-dithiobispyridine, EtOH/acetic
acid (20/1 v/v), r.t., 12 h, 82%; (b) 4-nitrophenyl chloroformate, DMAP,
Et3N, dry THF, 0 1C-r.t., 15 h, 97%; (c) 3, phosgene, DMAP, Et3N, dry THF,
0 1C-r.t., 18 h, 33%; (d) mPEG5000-SH, EtOH/acetic acid (20/1, v/v), r.t.,
24 h, 59%; (e) 4, DMAP, dry DMF, 0 1C-r.t., 5 h, 95%; (f) mPEG5000-SH,
EtOH/acetic acid (20/1 v/v), r.t., 24 h, 61%; (g) 3, phosgene, DMAP, Et3N,
dry THF, 0 1C-r.t., 18 h, 30%; and (h) mPEG5000-SH, EtOH/acetic acid
(20/1 v/v), r.t., 24 h, 67%.
Fig. 3 Reconversion kinetics of MMC prodrug 8 (a, b) and phenytoin
prodrug 10 (c, d) at 37 1C determined by HPLC. Values are represented as
mean ꢀ SD (n = 3).
Fig. 2 Reconversion kinetics of SN-38 prodrug 6 at 37 1C determined by
HPLC: (a) at pH 4.5; (b) at pH 6.8; (c) in SGF; and (d) in SIF. Values are
represented as the mean ꢀ SD (n = 3).
This journal is ©The Royal Society of Chemistry 2015
Chem. Commun., 2015, 51, 5721--5724 | 5723

View Article Online
Communication
ChemComm
7, 255; (b) S. D. Clas, R. I. Sanchez and R. Nofsinger, Drug Discovery
Today, 2014, 19, 79; (c) X.-Q. Li, H.-Y. Wen, H.-Q. Dong, W.-M. Xue,
G. M. Pauletti, X.-J. Cai, W.-J. Xia, D. Shi and Y.-Y. Li, Chem.
Commun., 2011, 47, 8647; (d) D. T. Bui, J. Nicolas, A. Maksimenko,
(1 mM) and absence of NAC. Although clearly correlated with the
concentration of NAC, the release in the intestinal environment
was also promoted by the enzymes of the SIF. Indeed, at a low
concentration (0.1 mM), NAC did not accelerate drug release.
Increasing the concentration of NAC to 1 mM yielded faster
reconversion as expected. Thus the background release in SIF
was dramatically increased compared to the other two prodrugs,
most likely due to rapid hydrolysis of the unhindered carbonate
bond in this prodrug. It would be possible to stabilize the
prodrug by introducing steric hindrance around the carbonate.
The ideal drug reconversion half-life in the GI tract is 2–4 h,
thus requiring a NAC concentration of ca. 0.1 mM based on the
3 prodrugs’ kinetic results. This concentration is inferior to
what would be achieved with a commercially available 600 mg
NAC dosage from Zambon Ltd. (Switzerland). Assuming an
average GI fluid volume of 250 mL,22 the resulting NAC
concentration would be ca. 15 mM.
¨
D. Desmaele and P. Couvreur, Chem. Commun., 2014, 50, 5336;
(e) D. Wang, E. Viennois, K. Ji, K. Damera, A. Draganov, Y. Zheng,
C. Dai, D. Merlin and B. Wang, Chem. Commun., 2014, 50, 15890.
4 T. Heimbach, D. M. Oh, L. L. Y. Li, N. Rodriguez-Hornedo, G. Garcia
and D. Fleisher, Int. J. Pharm., 2003, 261, 81.
5 R. S. deJong, N. H. Mulder, D. R. A. Uges, S. Kaul, B. Winograd,
D. T. Sleijfer, H. J. M. Groen, P. H. B. Willemse, W. T. A. vanderGraaf
and E. G. E. deVries, Br. J. Cancer, 1997, 75, 1660.
6 V. R. Guarino, K. Nti-Addae and V. J. Stella, Bioorg. Med. Chem. Lett.,
2011, 21, 172.
7 T. Wang, D. Y. W. Ng, Y. Wu, J. Thomas, T. TamTran and T. Weil,
Chem. Commun., 2014, 50, 1116.
8 M. L. Circu and T. Y. Aw, Free Radical Res., 2011, 45, 1245.
9 Y. Kawato, M. Aonuma, Y. Hirota, H. Kuga and K. Sato, Cancer Res.,
1991, 51, 4187.
10 M. Tomasz, D. Chowdary, R. Lipman, S. Shimotakahara, D. Veiro,
V. Walker and G. L. Verdine, Proc. Natl. Acad. Sci. U. S. A., 1986,
83, 6702.
11 M. A. Rogawski and W. Loscher, Nat. Rev. Neurosci., 2004, 5, 553.
12 (a) S. Bhuniya, M. H. Lee, H. M. Jeon, J. H. Han, J. H. Lee, N. Park,
S. Maiti, C. Kang and J. S. Kim, Chem. Commun., 2013, 49, 7141;
(b) S. Bhuniya, S. Maiti, E.-J. Kim, H. Lee, J. L. Sessler, K. S. Hong and
J. S. Kim, Angew. Chem., Int. Ed., 2014, 53, 4469; (c) G. Butora, N. Qi,
W. Fu, T. Nguyen, H.-C. Huang and I. W. Davies, Angew. Chem., Int.
Ed., 2014, 53, 14046; (d) J. Liu, W. Liu, I. Weitzhandler,
J. Bhattacharyya, X. Li, J. Wang, Y. Qi, S. Bhattacharjee and
A. Chilkoti, Angew. Chem., Int. Ed., 2015, 54, 1002; (e) A. K. Jain,
M. G. Gund, D. C. Desai, N. Borhade, S. P. Senthilkumar,
M. Dhiman, N. K. Mangu, S. V. Mali, N. P. Dubash, S. Halder and
A. Satyam, Bioorg. Chem., 2013, 49, 40; ( f ) X. Yang, I. Dogan,
V. R. Pannala, S. Kootala, J. Hilborn and D. Ossipov, Polym. Chem.,
2013, 4, 4621; (g) A. Kock, K. Zuwala, A. A. A. Smith, P. Ruiz-Sanchis,
B. M. Wohl, M. Tolstrup and A. N. Zelikin, Chem. Commun., 2014,
50, 14498; (h) P. Ruiz-Sanchis, B. M. Wohl, A. A. A. Smith, K. Zuwala,
J. Melchjorsen, M. Tolstrup and A. N. Zelikin, Adv. Healthcare Mater.,
2015, 4, 65.
Oversaturation is beneficial for the intestinal absorption of
BCS class II drugs.23 We followed the aqueous concentration
of the released drug over time upon controlled cleavage by
NAC. Interestingly, a transient oversaturation (1.4–3.5 times
the drugs’ solubilities) lasting over 1 h was observed before
precipitation occurred (Fig. S6–S11, ESI†).
In summary, an oral prodrug strategy for insoluble drugs
with a controllable reconversion kinetics was developed. Three
prodrugs were synthesized via a phenolic or aliphatic carbonate
(SN-38 and phenytoin) or a carbamate (MMC). The prodrug
formation greatly increased water solubility in all cases. The
reconversion at intestinal pH was correlated with the concen-
tration of NAC, suggesting that this exogenous agent could be
used to control the reconversion in the GI tract. Furthermore,
we demonstrated that acidic pH could suppress the prodrug
reconversion in the gastric environment, where little absorp-
tion is expected to take place. Our data also indicate that an
unhindered carbonate prodrug is not sufficiently stable in
the harsh intestinal milieu to be under strict control by the
exogenous reconversion agent, but that a carbamate is. Thus,
we reported here a new oral prodrug concept where an exogenous
agent is used to control reconversion and provide design guide-
lines that will help in developing this strategy.
13 (a) P. D. Senter, W. E. Pearce and R. S. Greenfield, J. Org. Chem.,
1990, 55, 2975; (b) A. N. Dinaut and S. D. Taylor, Chem. Commun.,
`
2001, 1386; (c) R. Labruere, A. Alouane, T. L. Saux, I. Aujard,
P. Pelupessy, A. Gautier, S. Dubruille, F. Schmidt and L. Jullien,
Angew. Chem., Int. Ed., 2012, 51, 9344; (d) S. H. Lee, E. Moroz,
B. Castagner and J.-C. Leroux, J. Am. Chem. Soc., 2014, 136, 12868.
14 K. Haba, M. Popkov, M. Shamis, R. A. Lerner, C. F. Barbas 3rd and
D. Shabat, Angew. Chem., Int. Ed., 2005, 44, 716 (Angew. Chem., 2005,
117, 726).
15 V. Savarino, G. S. Mela, P. Scalabrini, A. Sumberaz, G. Fera and
G. Celle, Dig. Dis. Sci., 1988, 33, 1077.
16 J. Fallingborg, L. A. Christensen, M. Ingeman-Nielsen, B. A.
Jacobsen, K. Abildgaard and H. H. Rasmussen, Aliment. Pharmacol.
Ther., 1989, 3, 605.
This work was supported by a Novartis Fellowship to Dr Tao
Sun. The authors declare no competing financial interest.
17 P. A. Fernandes and M. J. Ramos, Chem. – Eur. J., 2004, 10, 257.
18 J. A. Zhang, T. Xuan, M. Parmar, L. Ma, S. Ugwu, S. Ali and I. Ahmad,
Int. J. Pharm., 2004, 270, 93.
19 T. Kakutani, R. Atsumi, E. Sumimoto and M. Hashida, Chem. Pharm.
Bull., 1987, 35, 4907.
Notes and references
1 H. D. Williams, N. L. Trevaskis, S. A. Charman, R. M. Shanker, 20 S. A. Varia, S. Schuller, K. B. Sloan and V. J. Stella, J. Pharm. Sci.,
W. N. Charman, C. W. Pouton and C. J. Porter, Pharmacol. Rev.,
2013, 65, 315.
1984, 73, 1068.
21 J. Hirtz, Br. J. Clin. Pharmacol., 1985, 19(suppl. 2), 77S.
2 A. K. Shah and S. A. Agnihotri, J. Controlled Release, 2011, 156, 281. 22 J. B. Dressman, G. L. Amidon and D. Fleisher, J. Pharm. Sci., 1985, 74, 588.
3 (a) J. Rautio, H. Kumpulainen, T. Heimbach, R. Oliyai, D. Oh, 23 J. Brouwers, M. E. Brewster and P. Augustijns, J. Pharm. Sci., 2009,
T. Jarvinen and J. Savolainen, Nat. Rev. Drug Discovery, 2008,
98, 2549.
5724 | Chem. Commun., 2015, 51, 5721--5724
This journal is ©The Royal Society of Chemistry 2015