Angewandte
Communications
Chemie
bound to Singlet Oxygen Sensor Greenꢀ (SOSG)[3f] through
its most reactive carboxylic group.[4b] SOSG is an anthracene–
fluorescein dyad in which the fluorescence is quenched by
photoinduced intramolecular electron transfer between the
two moieties. Upon anthracene endoperoxidation in the
to the proteins kinetically competing with SOSG for 1O2
molecules. Moreover, they were able to show that SOSG
could in fact be internalized by HeLa cells in protein-free
culture medium.[4b] However, intracellular SOSG shows addi-
1
tional problems that further detract from its use as a O2
1
presence of O2, electron transfer is blocked, thereby restor-
reporter: its fluorescence spectrum is red-shifted compared to
aqueous solutions, intense fluorescence is still observed prior
ing the intrinsic fluorescence of the fluorescein.[4b] According
to the data supplied by the manufacturer, SOSG is a probe
with a high specificity towards singlet oxygen,[13] and there is
no chemical rationale for expecting alteration of its specificity
in the nanoparticle because we linked it through formation of
an amide bond at a site that is rather remote from the reactive
anthracene moiety.[14] Aqueous solutions containing
2 mgmLÀ1 of these nanoprobes and 1 mm of new methylene
blue (NMB) as 1O2 source were irradiated and the probe
fluorescence changes observed over time. The response of the
nanoprobe was poor when compared to the free probe; only
a modest 10% fluorescence increase was observed (Fig-
ure S1C,D in the Supporting Information). Even in deute-
rium oxide (D2O), where the 20-fold longer 1O2 lifetime
facilitates the reaction,[1] the fluorescence enhancement was
below 40%. In order to rule out any effects of light scattering
by the NPs, molecular SOSG solutions were irradiated in the
presence of 1 mm NMB and increasing amounts of NPs
(Figure S2). The results show that for concentrations up to
5 mgmLÀ1, SOSG behaves almost equivalently in the pres-
ence and absence of added free NPs.
1
to O2 exposure, and it is difficult to obtain systematic and
reproducible results.[4b] As shown below, the performance of
the NanoSOSG probe is not affected by such shortcomings.
Interaction with proteins, which strongly affects the
fluorescence of SOSG,[16] causes only marginal fluorescence
quenching in NanoSOSG at the highest BSA concentration
(Figure S3), thus indicating that interaction of the nanoprobe
with proteins is prevented by the NP scaffold. We show in the
Supporting Information that SOSG actually forms 1:1 and 2:1
complexes with BSA, each with distinct spectra and binding
constants (Figure S4). Similar situations are likely to occur in
protein-rich environments such as in cells. This evidence
suggests NanoSOSG to be more suited than molecular SOSG
1
as a fluorescent probe for O2 in biological media.
1
The ability of NanoSOSG to respond to intracellular O2
was then assessed through a series of assays using wild type
(wt) E. coli and a genetically modified E. coli strain that
expresses miniSOG, a flavin-binding fluorescent protein with
1
a strong capacity to sensitize O2 production inside cells.[17]
Figure 1 presents the behavior of the free SOSG probe in
both types of cells.
The poor performance of this early nanoprobe was
ascribed to non-specific interactions occurring between the
probe and the polyacrylamide matrix, leading to decreased
reactivity of the SOSG molecule bound to the polymer
network. In addition, it is worth noting that the initial
fluorescence of the SOSG-NPs was substantially higher than
that of free SOSG, which suggests that the SOSG micro-
environment in the nanoprobe impairs efficient electron-
transfer quenching.
Figure 1A shows that the fluorescence spectrum of SOSG
in the presence of wt E. coli cells matches that in PBS, thus
indicating that it is not bound to proteins. The effects of
irradiation at 420 Æ 20 nm (see Figure S5 for details) are
shown in Figures 1B,C. In wt cells, the fluorescence of SOSG
increases linearly as a result of its well-known self-sensitized
photooxidation.[3d] In miniSOG-expressing cells, the rate of
fluorescence increase shows two distinct regions: up to 15 min
irradiation, the fluorescence increases at a rate similar to that
observed in wt cells, whereas after that, the rate increases by
approximately 2.5-fold compared to the wt cells. The viability
of wt cells is not compromised by irradiation of SOSG,
whereas cells expressing miniSOG are killed very effectively,
with more than 90% being killed after just 15 min (Fig-
ure 1C). Taken together, these results indicate that SOSG and
miniSOG are not in close proximity at the early stages of
irradiation, that is, SOSG is not internalized by E. coli cells.
The fluorescence increase observed in both cells at these early
times is due solely to SOSG self-sensitized 1O2 production in
the aqueous medium and is therefore independent of intra-
To address this problem, spacers of different lengths were
introduced to separate the polyacrylamide scaffold and the
probe. To this end, alkyne-functionalized polyacrylamide
particles were prepared to enable orthogonal copper-cata-
lyzed azide–alkyne cycloaddition (CuAAC) coupling to the
amino-azide-functionalized linkers.[15] Following CuAAC, the
SOSG probe was attached to the free end of the linker
through amide bond formation (Scheme 1B). Linkers of
different lengths were tested (S, M, and L). The inclusion of
linkers improved the performance of the nanoprobes. The
1
best O2 trapping efficiency was observed for the nanoprobe
with the medium-sized linker (M, 7.1 ꢁ), which showed
a fluorescence enhancement of up to 3.2-fold in D2O after
50 min irradiation. This nanoprobe (henceforth called Nano-
SOSG) was selected for further experiments. Its surface was
further functionalized with cationic groups to facilitate
cellular uptake (Scheme 1C).
1
1
cellular O2. In miniSOG-expressing cells, intracellular O2
damages the cell from the inside,[18] and eventually miniSOG
is released to the external medium where it enhances the rate
of SOSG photooxidation. An alternative explanation, such as
photochemical internalization of SOSG,[19] can be ruled out
because no spectral changes are observed after irradiation
(Figure 1A).
SOSG is described as cell-impermeant,[13] which detracts
1
from its usefulness as an O2 probe in cells. Gollmer et al.
concluded that this property is the result of extensive protein
binding in the culture medium. Indeed the presence of bovine
serum albumin (BSA) resulted in red-shifted fluorescence
and lower responsiveness to 1O2, which the authors attributed
The corresponding results for NanoSOSG are also shown
in Figure 1. As expected, NanoSOSG does not interact with
cell proteins either before or after extensive irradiation
(Figure 1D). The rate of fluorescence increase in wt and
2
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 5
These are not the final page numbers!