In situ activation and monitoring of the evolution of the intracellular caspase family† †Electronic supplementary information (ESI) available: Experimental details and supplementary figures. See DOI: 10.1039/c5sc00471c Click here for additional data file.

An intergrated nano-platform is designed to achieve in situ activation, monitoring and signal feedback of the caspase family evolution from upstream to downstream.


Introduction
Caspases are a family of cysteine-aspartic proteases that are only activated during cell apoptosis, and can be used as feedback markers of cell death. 1 Caspase-controlled apoptosis has a characteristic enzyme cascade, which involves multiple caspases at different stages and pathways. 2 Upstream caspase, such as caspase-9 (casp-9), plays a central role in the induction of apoptosis, while downstream caspase such as caspase-3 (casp-3) is critical for carrying out the nal step of cell apoptosis. Thus the evaluation of the intracellular caspase family is essential to elucidate the cell apoptosis process. Indeed, many uorescent probes have been developed for imaging of caspase activity in living cells and animals, 3 and real-time monitoring of caspase cascade activation by diverse pairs of dyes and corresponding quenchers. 4 However, besides the sensing probes, some additional inducers are usually needed to activate intracellular caspase activity. 4,5 Thus an apoptosis sensor has developed for in situ activation and imaging of intracellular casp-3 using aggregation-induced emission. 6 A platform for in situ activation and monitoring the evolution of caspase family during cell apoptosis is still an urgent need.
Noble metal nanostructures with good biocompatibility have received considerable research interest due to their strong absorption in the near-infrared (NIR) region and high photothermal conversion efficiency. 7 Herein, using gold nanorods (AuNR) as the model of both nanocarrier and matter inducing the cell apoptosis, which was found to be able to quench simultaneously two kinds of dyes at two unique surface plasmon resonance (SPR) absorption wavelengths, a versatile nanoprobe was designed for in situ activation and monitoring of the evolution of the caspase family from upstream to downstream via NIR photothermal treatment.
The nanoprobe was prepared by assembling a uorescein isothiocyanate (FITC)-labelled peptide specic to casp-9 (peptide-9) and cyanine-5.5 (Cy5.5)-labelled peptide specic to casp-3 (peptide-3) as signal switches and recognition elements, and folic acid (FA) as a target specic moiety on AuNR (Scheme 1). Their uorescence was initially quenched via energy transfer from FITC and Cy5.5 to AuNR with transverse and longitudinal SPR absorption, respectively. Upon endocytosis of the nanoprobe and NIR irradiation, cell apoptosis was encouraged by the photothermal effect and thus the peptide could be cleaved successfully by the corresponding activated caspase from upstream casp-9 to downstream casp-3, which released the dyes from the nanocarrier for uorescent imaging. The turn-on signals provided an efficient way for quantication of both caspa-9 and casp-3 activities in cancer cells and monitoring of their evolution in living mice. Since caspase activity is the marker of cell apoptosis, the uorescence response could also be used to monitor therapeutic efficacy in real time.

Results and discussion
Characterizations of the nanoprobe Gold nanocarriers were synthesized according to a typical method of seed-mediated growth and well characterized (see ESI,Fig. S1A and S2 †). 8 For efficient preparation of the nanoprobe, HS-poly(ethylene glycol)-NH 2 was used to protect the nanorod from aggregation and bind efficiently N-hydroxysuccinimide (NHS)-functionalized FA via a typical amide reaction (see ESI,Fig. S3 †). 9 Aer functionalization with dye-labelled peptides and PEG, a slight red shi of the characteristic absorption at 787 nm was observed in the UV-vis spectra ( Fig. S2A †), while the binding of NHS-FA to the PEG produced a characteristic absorption peak of FA at 280 nm. 10 The functionalization did not change their surface prole (Fig. S1B †). In addition, compared with raw nanorods, the nanoprobe showed an Au-Br Raman peak with the shi from 180 cm À1 to 261 cm À1 (Fig. S2B †), and the surface changed to a negative z potential ( Fig. S2C †). These results indicated that the nanoprobe was synthesized successfully with good-dispersibility and excellent optical properties for uorescent detection and imaging.
In vitro detection of casp-9 and casp-3 activities To test the validity of the nanoprobe to caspase, in vitro enzymatic assays were performed with recombinant caspase proteins. Since the emission of FITC and Cy5.5 overlapped with the transverse and longitudinal SPR absorption of the gold nanorod (see ESI, Fig. S4 †), their uorescence (FL) was quenched via FRET (see ESI, Fig. S5 †), respectively. Aer incubating the mixture of the nanoprobe and recombinant caspase proteins in caspase assay buffer at 37 C, the FL was signicantly enhanced at 517 and 694 nm, respectively, indicating that the enzymatic reaction released the dyes from the gold nanostructure, which could be inhibited by the specic inhibitors of casp-9 or casp-3 ( Fig. 1A). At the optimized reaction time of 80 min (Fig. 1B), the FL intensity increased linearly with the enhancing concentration of caspase ( Fig. 1C and D). The cleavage reaction showed good specicity to caspase against other interferents ( Fig. 1E and F), leading to a method for the detection of caspase activity.
The amounts of peptide-9 and peptide-3 loaded on the nanocarrier were determined to be 1.01 Â 10 4 and 9.73 Â 10 3 , respectively (see ESI,Fig. S6 †). The kinetic analysis of cleavage reactions was carried out by incubating casp-9 or casp-3 with the increasing concentration of nanoprobe at 37 C. The Michaelis constants, K M , of casp-9 and casp-3 were calculated to be 6.70 AE 0.4 and 5.58 AE 0.3 mM, and the kinetic constants, k cat , were 1.69 and 1.62 s À1 , respectively ( Fig. 2A-D). The k cat value for casp-3 was comparable to the value reported with another uorescent probe, 11 and the K M value was lower than the 12.7 mM of a commercial substrate for casp-3, 12 indicating the better affinity between the nanoprobe and caspase.
Owing to the essential roles of caspases in cell apoptosis, the selectivity of nanoprobes in monitoring the activity of caspase in complex cellular samples was examined. HeLa cell lysates were collected aer treatment with a commonly used cell apoptosis inducer (staurosporine, 2 mM) to activate the caspase. Time-dependent uorescence at both 517 and 694 nm with excitation wavelengths of 490 and 675 nm was detected during the incubation of the lysate with nanoprobe, respectively ( Fig. 2E and F), which showed the acceptable selectivity of the nanoprobe for intracellular activated casp-9 and casp-3.

Monitoring the evolution of the intracellular caspase family
Prior to intracellular usage, the cytotoxicity of nanoprobe and NIR irradiation (808 nm) was examined with an MTT assay (see ESI,Fig. S7 †). Aer incubation with different amounts of nanoprobe for 3 h or treatment with NIR for 50 min, HeLa cells still maintained a high viability, indicating the good biocompatibility of nanoprobe and the low cytotoxicity of NIR irradiation itself. Compared with HaCat normal cells, the nanoprobetransfected HeLa cells showed obvious apoptosis aer NIR irradiation at 4 W cm À2 for 10 min. The nanoprobe could enter HeLa cells via FA receptor-mediated endocytosis and accumulate in the cytoplasm outside of cell nucleus (see ESI, Fig. S8 and S9 †), which was also veried by TEM image (see ESI, Fig. S10 †). The weak uorescence of FITC could be observed in MCF-7 and HeLa cells, while no change was observed in A549 cells, con-rming FA receptor-mediated internalization of the nanoprobe.
To employ the probe to monitor the evolution of the caspase family from upstream to downstream during therapy, HeLa cells were seeded in a confocal dish for 24 h. Aer incubation with the nanoprobe for 3 h and then treatment with NIR irradiation for different times, the HeLa cells were then studied with confocal uorescence imaging (Fig. 3). Little uorescence was observed in the nanoprobe transfected HeLa cells before NIR irradiation. Aer NIR irradiation for 3 min the uorescence of FITC (green) in HeLa cells rst appeared, and then increased gradually with the progression of cell apoptosis induced by the photothermal effect of the gold nanocarrier, while the uorescence of Cy5.5 (red) was observed aer NIR irradiation for 10 min, indicating that casp-9 was activated in the initial stage of cell apoptosis, which thus played the role of initiator in the caspase family. With the deepening of the degree of cell apoptosis, casp-3 was activated to show the uorescence of Cy5.5. Aer irradiation for 30 min, apoptotic HeLa cells showed shrinkage with maximum uorescence intensity in both the green and red channels.

Fig. 2
Plots of (F À F 0 )/F 0 vs. concentration of (A) peptide-9 and (B) peptide-3 loaded on nanoprobes, where F 0 and F are the fluorescence intensity of nanoprobe and the mixture of nanoprobe with target caspase (10 Unit mL À1 ) after incubation for 80 min at 37 C, respectively. Caspase enzymatic kinetics assay of (C) casp-9 (10 Unit mL À1 ) and (D) casp-3 (10 Unit mL À1 ) with increasing substrate concentration. Timedependence of the fluorescent response of nanoprobe (10 mL) in apoptotic HeLa cell lysate at (E) 517 nm and (F) 694 nm.
to release FITC from the nanoprobe. Contrarily, aer the probetransfected HeLa cells were treated with casp-9 inhibitor, both the green uorescence and the following red uorescence disappeared. Thus the activation of casp-3 depended on casp-9, which demonstrated the involvement of the mitochondrial apoptotic pathway. 13 The evolution was also validated by ow cytometric assays, which showed the increasing uorescence of FITC and Cy5.5 (Fig. 5A) as the apoptosis percentage rose (Fig. 5B).
The mitochondrial pathway of apoptosis was conrmed using Rhodamine 123 staining (see ESI, Fig. S11 †), which is readily sequestered by living mitochondria in cells undergoing apoptosis. 14 Furthermore, both the apoptotic detection kit with ow cytometry (Fig. 5B) and uorescence imaging of HeLa cells stained with 4 0 ,6-diamidino-2-phenylindole (DAPI) dye, specic for the cell nucleus (see ESI, Fig. S12 †), veried the caspasedependent early apoptosis through mitochondrial pathway. 15 To verify further the application of the designed nanoprobe in monitoring the evolution of the caspase family, nanoprobe-9 and nanoprobe-3 were also synthesized (see ESI †). Aer incubating HeLa cells with these probes for 3 h at 37 C, the cells were treated with NIR irradiation for different times. Consistent with the appearance shown in Fig. 3, the nanoprobe-9 transfected HeLa cells showed the green uorescence at 3 min postirradiation followed with casp-9 activation (Fig. 6A), while the nanoprobe-3 transfected cells showed red uorescence from 10 min due to the activation of casp-3 (Fig. 6B). The specicity of the intracellular cleavage reaction was demonstrated by  immunouorescence imaging (see ESI, Fig. S13 †). An excellent overlap was observed both between the green uorescence of nanoprobe-9 and the immunouorescence signal of casp-9, and the red uorescence of nanoprobe-3 and immunouorescence signal of casp-3, suggesting intracellular caspase-specic activation and imaging.

Quantication of two intracellular caspases
The proposed probe could be used to quantify the activities of two intracellular caspases. To obtain the calibration curve, HeLa cells (5.0 Â 10 4 ) were incubated with the nanoprobe for 3 h and then treated with NIR irradiation with increasing time to obtain the confocal uorescence images. The FL intensity was measured in the cell area with Leica soware (Fig. 7A-C). The corresponding casp-9 and casp-3 activities of the treated HeLa cells were detected via in vitro casp-9 and casp-3 kit analysis of the cell extracts using their standard curves (Fig. 7D-G). The obtained caspase activities (c) were then used to obtaining the calibration curve for quantication of casp-9 and casp-3 activities in a single cell from the FL intensity ( Fig. 7H and I). The plots of FL intensity (FI) vs. c (10 À7 Unit) in single cell followed the linear regression equations of FI ¼ 5.67 + 2.10 Â c for casp-9 and FI ¼ 5.33 + 2.33 Â c for casp-3. The average activity of casp-9 and casp-3 in a single HeLa cell was 4.25 and 5.09 Â 10 À7 Unit aer the therapy-induced apoptosis. Therefore, the proposed strategy possessed the applicability to monitor the change of intracellular caspase activity.

Evaluation of therapeutic efficiency in real-time
Interesting, the unique caspase-responsive uorescence of the nanoprobe could be used to evaluate the therapeutic efficiency. To verify this capability, the uorescent (FITC and Cy5.5) and morphological changes of nanoprobe transfected HeLa cells were tracked in real time under NIR irradiation by confocal uorescence imaging (see ESI, Fig. S14 † and Fig. 8), which showed the increasing luminescence with cell apoptosis. This  result proved that the functionalized nanoprobe not only had the potential for in situ activation and detection of intracellular caspase, but also could be used for real-time monitoring of the therapeutic effect of a targeted cancer cell, providing a novel tool to evaluate the therapeutic response.

Monitoring caspase activity in living mouse
This design could also be applied to monitor caspase activity in living mouse activated by the treatment effect of the gold nanostructure. A HeLa tumor was subcutaneously implanted on the ank of the nude mice. The tumor-bearing mice were then  intravenously injected with the nanoprobe. At 24 h post-injection, tumor accumulation of the nanoprobe was found to be highly efficient (see ESI, Fig. S15 †). Aerward, the mice were irradiated with a NIR laser for 30 min to perform the therapy. The therapeutic efficiency was assessed by monitoring the tumor volume aer treatment, which showed that the tumor growth was signicantly inhibited at 24 h post-irradiation (see ESI, Fig. S16 †). The evaluation of excised tissues further demonstrated that strong uorescence occurred in the cancer tissue in comparison with other organs such as the liver and kidneys (see ESI, Fig. S17 †), indicating the cleavage of peptide-9 and/or peptide-3 by the corresponding active caspases. With reduced tumor volume, the uorescence acquired at both 680-800 nm for Cy5.5 (Fig. 9A) and 500-620 nm for FITC (Fig. 9B) from the irradiated tumor increased gradually aer laser irradiation. The more obvious background in Fig. 9B could be attributed to auto-uorescence at the excitation wavelength of 455 nm, which was greatly reduced under NIR excitation (Fig. 9A). Furthermore, the uorescence of FITC for casp-9 showed an earlier and greater change (Fig. 9B) than that of Cy5.5 for casp-3 (Fig. 9A), which indicated that casp-9 was activated prior to casp-3, and was in accord with the above cellular experiments. As expected, the un-irradiated mouse showed negligible change in the uorescence (the le in Fig. 9A) and growing tumor volume. Therefore, the nanoprobe was efficient for the in situ activation and monitoring of caspase family activity for therapeutic feedback in living mice.

Conclusions
This work designs a protocol to monitor in situ the evolution of intracellular caspase from upstream to downstream activated by a nanocarrier with a therapeutic effect to inducing cell apoptosis. The evolution is performed by sequentially lightingup the uorescence of dyes labelled to peptides assembled on the nanocarrier via caspase-catalytic cleavage. The uorescence signal can be used for not only in situ quantication of both caspa-9 and casp-3 activities in cancer cells but also as a selffeedback for the therapeutic response of cancer cells, which leads to a signicant method for monitoring therapeutic effect in vivo. This methodology is applicable for other nanocarriers with a related effect to monitor caspase-dependent apoptosis. The strategy not only offers a new insight for real-time monitoring the evolution of the intracellular caspase family and evaluating the therapeutic efficacy but also accelerates the uncovering of the biological roles of caspases in cancer apoptosis.