Angewandte
Communications
Chemie
there might be another way to prevent siRNA loading or
activation through the RNA interference silencing complex.
We then tracked the location of native siRNA (SL/AL) and
vitE-siRNA (SL/VAL) after lipo-transfection. A Cy3-labeled
sense strand RNA analog (Cy3-SL) was used for hybrid-
ization with AL or VAL. After co-transfection, the cells were
washed and cultured for an additional 6 and 12 hours. The
cells were then imaged (Figure S2). After 6 hours, cells
transfected with Cy3-SL/AL or Cy3-SL/VAL contained
bright red fluorescence spots of siRNA/lipo complexes.
After 12 hours, a significant difference was observed since
cells with Cy3-SL/AL had few red fluorescence spots left,
while cells co-transfected with vitE-siRNA (Cy3-SL/VAL)
still contained many red fluorescence spots in the cytoplasm.
These data indicate that siRNAwas released from the siRNA/
lipo complexes in Cy3-SL/AL-transfected cells, yet vitE-
siRNA/lipo complexes were defective in this process and
vitE-siRNA was not released. According to other reports,
vitE can interact with proteins such as a-tocopherol transport
protein (a-TTP).[8] To confirm this, we first tested different
inhibitors based on different mechanisms of cellular uptake.
As shown in Table S1 and Figure S3, chlorpromazine, ami-
loride, methyl-b-cyclodextrin, and genestein almost did not
inhibit the entrance of vitE-siRNA/lipo complexes into cells,
while vitE and low temperature displayed clear blocking of
vitE-siRNA entrance. These results indicated that the cellular
uptake of vitE-siRNA/lipo complexes was most possibly
mediated by vitE receptor proteins. Further vitE concentra-
tion dependence experiments indicated that the competition
of vitE binding caused the less efficient delivery of vitE-
siRNA/lipo complexes with increasing vitE concentrations
(Figure S4). Protein binding at siRNA terminals would
introduce much bulkier groups, which might lead to difficulty
in complex dissociation and/or the interaction of vitE-siRNA
duplexes with the RNAi cellular machinery.
Figure 2. Dose effect on photomodulation of luciferase activity with
caged vitE-siRNAs (SL/VpAL). The concentration of PC siRNA was
fixed at 5 nm.
photocontrol using firefly luciferase activity as a read-out
(Figure 2 and Figure S6). The same dual reporter firefly/
renilla luciferase assay in HEK293 cells was used with the
following doses of caged siRNAs: 2.5 nm, 5.0 nm, 10.0 nm, and
20.0 nm. For cells treated with caged siRNA duplexes, slight
inhibition of firefly luciferase expression was observed for
only 20 nm siRNAs (approximately 80% for SL/VpAL and
92% for VpSL/VpAL).
Light activation restored siRNA activity, and a dose
dependency for gene silencing was observed. The photo-
modulation ratios for different concentrations of SL/VpAL
siRNA duplexes were up to 12.5-fold for 10 nm and 18.6-fold
for 20 nm, which represent the best known photomodulation
of caged siRNAs to date. Similar results were also achieved
with a photomodulation efficiency of up to 7.3-fold for 20-nm
VpSL/VpAL (Figure S5).
Photomodulation of firefly luciferase activity of siRNA
duplexes with one caged vitE-RNA and one caged or non-
caged complementary vitE-RNA was also investigated. Light
activation of VpSL/VAL and VSL/VpAL produced SL/VAL
and VSL/AL siRNA duplexes. As expected, the non-caged
VpSL/VAL and VSL/VpAL in cells showed the same levels of
firefly luciferase activity as that of direct co-transfection of
SL/VAL and VSL/AL in cells (Figure 1). For the caged
siRNA (VpSL/VpAL), light irradiation partially recovered its
knockdown of firefly luciferase with only a 4.3-fold enhance-
ment of siRNA gene silencing activity. As indicated in
Figure S1, brief irradiation may remove one vitE moiety
from VpSL/VpAL to form VpSL/VAL or VSL/VpAL, which
showed almost no gene silencing activity. Complete removal
of both vitE moieties could fully restore siRNA function.
Different light exposure times (1 to 7 min) were then applied
to cells co-transfected with the caged VpSL/VpAL duplex. As
shown in Figure S5, prolonged irradiation time recovered the
full silencing ability of the caged VpSL/VpAL duplex.
Maximum photomodulation (up to 10.5-fold) of siRNA
activity was achieved with 6 min of light exposure.
Based on the observations above, photolabile siRNA with
only a single vitE modification at the 5’ terminal of antisense
strand RNA could be used to efficiently photoregulate gene
expression. Another gene, GFP, was then tested to confirm
the generality of gene silencing with our caged vitE-siRNAs
(Figure 3A). Antisense strands of GFP-targeting siRNA
labeled with only a vitE moiety or vitE and a photolinker at
the 5’ terminal were then synthesized using similar protocols.
HEK293 cells were then cotransfected for 6 hours with
pEGFP-N1 and pDsRed2-N2 (RFP, as the internal control),
as well as SG/AG, SG/VAG, or SG/VpAG. Two sets of
experiments were conducted with or without brief light
activation. After another 42 hours of incubation, the cells
were imaged, and GFP/RFP expression was quantified by
flow cytometry (Figure 3B). The amount of cells with both
GFP and RFP expression was then normalized to cells with
RFP expression. As expected, light irradiation had no effect
on GFP and RFP expression in both negative and positive
control experiments. Cells treated with GFP siRNA (SG/AG)
showed 91% knockdown of GFP expression. Although cells
treated with non-caged vitE-siRNA (SG/VAG) still displayed
strong GFP fluorescence, light irradiation had little effect on
GFP expression, which is similar to gene silencing of firefly
luciferase with non-caged siRNA (SL/VAL). In the presence
Because siRNAs (SL/VpAL and VpSL/VpAL) were fully
inactive at 5 nm without light activation, we further evaluated
the dose dependency of the above caged vitE-siRNAs on
2154
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 2152 –2156
Ji, Yuzhuo
