Communications
primary amines, but can be cleaved selectively by a thiol
moiety via formation of a Meisenheimer complex.[13–15] In this
regard, the combination of glutathione (GSH) and glutathione-
S-transferase (GST) in the cytoplasm has the potential to elicit
the highly selective cleavage of the 2-nitrobenzenesulfonamide
group in the cell. In general, the redox potential drastically dif-
fers across cellular membranes; the concentration of GSH,
which plays a prime role in upregulating redox potential, is
present at 1–10 mm in the cell, whereas that its concentration
in the extracellular milieu is 1–20 mm.[16] In addition, GST, which
resides in the cell, works as a metabolic enzyme for catalyzing
nucleophilic attack of the thiol moiety of the cysteine residue
in GSH, toward facilitated cleavage of the 2-nitrobenzenesul-
fonamide group.[17–18] To demonstrate the utility of the 2-nitro-
benzenesulfonamide group as a chemical linker, in the present
study, siRNA was conjugated with poly(ethylene glycol) (PEG).
PEG can inhibit siRNA recognition by the gene-silencing-relat-
ed proteins due to its steric hindrance effect, leading to com-
promised gene-silencing activity.[19] Therefore, conjugation of
siRNA with PEG would work as a good indicator of the intracel-
lular cleavage of the chemical linker.
siRNA, in a mock extracellular environment (Figure 2c,d), sug-
gesting the stability of the 2-nitrobenzenesulfonamide linker
under extracellular reductive conditions. Of note, the siRNA re-
lease from PEG–sul–siRNA by the increased GSH (1 mm GSH,
pH 7.4, 378C) was modest (20%, 72 h), but was effectively fa-
cilitated by the presence of GST at 0.019 mgmLÀ1, correspond-
ing to intracellular GST concentrations[13] (89%, 72 h incuba-
tion; Figure 2c,d), indicating the pivotal role of GST on the effi-
cient cleavage of the 2-nitrobenzenesulfonamide group. This
result strongly supports the notion that PEG–sul–siRNA enables
efficient siRNA release selectively in the cell as well as minimal
siRNA release in the extracellular milieu, considering the exclu-
sive presence of the GSH/GST combination in the cell. Appreci-
able siRNA release was not detected for the PEG–car–siRNA
system under the aforementioned conditions, suggesting that
a carboxylic amide linker would work as a non-cleavable con-
trol in subsequent cell experiments (SI Figure S16).
To examine the intracellular siRNA release from PEG–sul–
siRNA, a gene silencing assay was performed with human cer-
vical cancer cells stably expressing luciferase (HeLa-Luc), using
Lipofectamine RNAiMAX. Luciferase-targeting (siLuc) and
scrambled siRNA (siScr) sequences were used to confirm se-
quence-specific gene silencing. Treatment with PEG–sul–siLuc
at 50 nm suppressed ~40% of luciferase expression, whereas
treatment with PEG–car–siLuc resulted in ~20% luciferase si-
lencing activity (Figure 3a). The significantly higher luciferase
silencing activity observed with the PEG–sul–siLuc system over
PEG–car–siLuc (p<0.01) strongly suggests the intracellular
cleavage of the 2-nitrobenzenesulfonamide group toward
siRNA release. Interestingly, treatment with PEG–disulfide–siLuc
resulted in ~20% luciferase silencing activity. The significantly
lower gene silencing efficacy (p<0.01) with the PEG–disulfide–
siLuc system relative to PEG–sul–siLuc might be attributed to
unexpected siRNA degradation outside the cell; moderate
cleavage of the disulfide linkage in the extracellular milieu re-
leases siRNA from PEG for siRNA exposure to RNase activity in
the culture medium. This result supports the utility of the 2-ni-
trobenezenesulfonamide group as a chemical linker, in order
to avoid premature cleavage before entering the cell. Note
that treatment with the siRNA series did not induce cell death
and luciferase silencing by the siScr series, confirming negligi-
ble cytotoxicity and sequence-specific gene silencing, respec-
tively (Figure 3b).
Firstly, PEG (Mr = ~40000 Da) was conjugated with siRNA
(Mr = ~13000 Da) via a 2-nitrobenzenesulfonamide linker to
produce PEG–sul–siRNA (Supporting Information (SI)
Schemes S1 and S2). Briefly, methoxy-PEG-NH2 was reacted
with methyl 4-(chlorosulfonyl)-3-nitrobenzoate to introduce
a 2-nitrobenzensulfonamide moiety at the PEG terminus (SI
Figure S2), followed by DBCO introduction to produce PEG–
sul–DBCO (SI Figures S4 and S8). The obtained PEG–sul–DBCO
was reacted with azide-siRNA via copper-free click chemis-
try,[20,21] and the resulting raw product was purified by ion-ex-
change chromatography, in order to remove unreacted PEG–
sul–DBCO and azide-siRNA (SI Figures S11 and S15). In a similar
manner, PEG–siRNA conjugates containing a non-cleavable car-
boxylic amide linker (PEG–car–siRNA; SI Scheme S3) or a con-
ventional redox-sensitive disulfide linker (PEG–disulfide–siRNA;
SI Schemes S4 and S5) were also prepared as controls (SI Fig-
ures S5–S7, S9, S10, S12, S13, S16, and S17). In agarose gel
electrophoresis, all PEG–siRNA series migrated shorter distan-
ces than unconjugated siRNA because of the increased molec-
ular weight of conjugated PEG, confirming successful synthesis
(SI Figure S14).
The obtained PEG–siRNA conjugates were subjected to
siRNA release evaluation under conditions mimicking extracel-
lular or intracellular environments. siRNA release was analyzed
by size-exclusion chromatography. In a mock extracellular envi-
ronment (20 mm GSH, pH 7.4, 378C),[16] PEG–disulfide–siRNA ex-
hibited a considerable amount of siRNA release in a time-de-
pendent manner (38%, 72 h; Figure 2a,b and SI Figure S18),
suggesting the undesired cleavage of the conventional disul-
fide linkage under extracellular conditions. The increased GSH
concentration (1 mm GSH, pH 7.4, 378C), corresponding to in-
tracellular environment,[16] induced complete siRNA release
from PEG–disulfide–siRNA within 6 h (Figure 2b), possibly lead-
ing to immediate cleavage of disulfide linkage after entering
the cell. Meanwhile, PEG–sul–siRNA showed dramatically sup-
pressed siRNA release (8%, 72 h) relative to PEG–disulfide–
The present study successfully demonstrated that the 2-ni-
trobenzenesulfonamide group can serve as a more stable
chemical linker than the conventional redox-sensitive disulfide
group, under extracellular reductive conditions. In contrast, the
combination of GSH and GST, which resides exclusively in the
cell, triggers efficient cleavage of the 2-nitrobenzenesulfona-
mide group. In this regard, GST activity is overexpressed in
tumor tissues[22] relative to normal tissues, which could lead to
even greater facilitated cleavage of the 2-nitrobenzenesulfona-
mide group in cancer cells. Ultimately, PEG–sul–siRNA induced
enhanced gene silencing activity in assays with cultured HeLa-
Luc cells. Notably, the 2-nitrobenzenesulfonamide group is
more stable upon exposure to light than the disulfide group
(SI Figures S20 and S21). In general, disulfide groups can be
&
ChemMedChem 2016, 11, 1 – 5
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!
Huang, Chih Hao
