Controllable NO release from Cu1.6S nanoparticle decomposition of S-nitrosoglutathiones following photothermal disintegration of polymersomes to elicit cerebral vasodilatory activity† †Electronic supplementary information (ESI) available: Experimental section and Fig. S1–S10. See DOI: 10.1039/c6sc02

A photo-responsive polymersome was designed to demonstrate the photo-controlled release of NO and its vasodilatory activity on zebrafish.


Introduction
Since its discovery as a an endothelium-derived relaxing factor, 1 nitric oxide (NO) has long been recognized as a unique signaling molecule of therapeutic potential. In particular, the uncharged NO molecule, which can be produced either endogenously from vascular endothelium or exogenously from NO donors, readily diffuses into neighboring smooth muscle cells and causes vascular relaxation. 2 Based on this notion, numerous NO therapies have been developed aiming to improve the blood ow in order to remedy disorders related to pathological vasoconstriction. For instance, nitroglycerin (an NO donor) is administered to patients with coronary artery disease, 3 and inhalation of NO is suggested to treat pulmonary artery hypertension by reversing hypoxic pulmonary vasoconstriction. 4 Moreover, a preclinical study of ischemic stroke shows that NO caused selective dilatation of collateral arterioles and resulted in an improved neurological outcome. 5 Nevertheless, clinical applications of NO therapies remain limited mainly because of concerns over the reduction of arterial pressure and tissue perfusion, and the excessive generation of cytotoxic reactive oxygen/nitrogen species (such as peroxynitrite) aer the systemic administration of NO. In addition, the short half-life of NO in biological uids, which results from its high reactivity to endogenous molecules such as glutathione, albumin hemoglobin and molecular dioxygen, may decrease the bioavailability of NO near the targeted blood vessels. [6][7][8] In this regard, development of a strategy to control the generation of NO with high spatiotemporal precision in physiological settings and to increase the bioavailability of NO is hence critically important for optimization of the therapeutic effect of NO. Herein, we present the rst example of delivery of NO with controlled release for the direct observation of vasodilation.
Nitrosothiols including S-nitrosoglutathione (GSNO), S-nitrosoalbumin (AlbSNO) and S-nitrosocysteine (CysNO) are found in both human plasma and tissue uids, and have been suggested to act as a therapeutic NO donor for the treatment of vascular disorders through the anti-platelet, anti-thrombotic and provasodilatory activities of NO. 9,10 In addition to heat and UV/visible light, which can cause the decomposition of the S-NO bond of nitrosothiols and the subsequent release of NO, metal ions such as Cu + , Fe 2+ , Hg 2+ and Ag + can also react with nitrosothiols leading to the generation of NO. 8,11 Accordingly, we propose that the djurleite (Cu 2Àx S) nanoparticles (NPs) have the potential to take part in the reaction of GSNO for the release of NO. On the other hand, the degradation of polymersomes made of poly-(lactic-co-glycolic acid) (termed PLGA polymersomes hereaer) is known to be sensitive to heating. Based on these notions, we present a strategy for the photo-triggered generation of NO from a polymersome cargo. Cu 1.6 S NPs have played dual roles in generating heat by laser irradiation and releasing NO by the reaction of GSNO. Our formulation comprises a PLGA polymersome with GSNO encapsulated in the core, and oil-phased djurleite (Cu 2Àx S) nanoparticles (NPs) packed inside the membrane, of the polymersome, respectively. Upon irradiation with photons of energy that matches the absorption band of Cu 2Àx S NPs, the embedded NPs become heated, thereby causing the increased permeability of the PLGA membrane and the subsequent escape of GSNO molecules from the core of the polymersome. When GSNO molecules inltrate through the PLGA membrane, they would react with the embedded Cu 2Àx S NPs leading to the generation of NO molecules (Fig. 1a). Our design notably improves the stability of NO donors (i.e. GSNO) and allows the generation of NO on demand, induced by light. These features are expected to increase the bioavailability of NO and open the possibility to spatiotemporally control the release of NO in vivo, based on which novel NO therapies could be envisioned.

Results and discussion
Characterization of GSNO/Cu 1.6 S-PLGA polymersomes Cu 1.6 S NPs were synthesized through a low-temperature (70 C) redox approach. 12 TEM images revealed that the Cu 1.6 S NPs were approximately 4.9 nm in size with a d spacing of 0.19 nm, which corresponds to the (1 1 3) lattice plane of the crystal (Fig. S1a †). XRD analysis identied particles belonging to the Cu 8 S 5 phase (JCPDS card no. 33-0491) (Fig. S1b †). Subsequently, the double emulsion method was employed to encapsulate the GSNO molecules and Cu 1.6 S NPs into the hydrophilic core and hydrophobic shells of the PLGA polymersomes, respectively (Fig. 1b). First of all, we prepared Cu 1.6 S-PLGA polymersomes without the inclusion of GSNO. Fig. 1c shows that the corresponding polymersomes can be successfully formed in a spherical morphology with a hydrodynamic diameter of 431 nm. Because GSNO was labile to light and heat, releasing NO, 8,11 GSNO was added into the hydrophilic core (W/O) under dark and low temperature (4 C) conditions during the rst emulsication process for the preparation of the GSNO/Cu 1.6 S-PLGA polymersomes. Next, the oilphase Cu 1.6 S NPs were embedded in the membrane of the polymersomes during the second emulsication process. Both TEM and SEM images show that the resultant GSNO/Cu 1.6 S-PLGA polymersomes retained the same morphology as the Cu 1.6 S-PLGA polymersomes but with a slightly increased diameter (441 nm; Fig. 1d and e). As illustrated in a cross-sectional image acquired with cryo-TEM, Cu 1.6 S NPs were distributed mainly in the membrane of the polymersomes (Fig. 1f). However, we can also nd that a small portion of Cu 1.6 S NPs were clustered in the cores in some polymersomes (Fig. S2 †). The elemental mapping analysis veried the presence of Cu and S in the polymersomes. The line-scan analysis showed a concave prole conrming a hollow structure with Cu and S mainly located around the polymersome membrane (Fig. S3 †). An NO assay kit following the Griess test 13 was employed to determine that 4.12 Â 10 À6 M (4.12 mM) of GSNO was encapsulated from the 200 ppm (copper ion concentration) of GSNO/Cu 1.6 S-PLGA polymersomes. To prevent the particles from being engulfed by macrophages in the blood, the positively charged bovine serum albumin (BSA) engaged in a favorable electrostatic interaction with negatively charged PLGA, thus modifying the surface of the PLGA polymersomes. The amount of BSA adsorbed on the GSNO/Cu 1.6 S-PLGA polymersomes was 13.3 mg BSA in 200 ppm GSNO/Cu 1.6 S-PLGA.

Evidence of controllable NO release with the reaction of Cu 1.6 S NPs
Prior to performing the laser irradiation of the GSNO/Cu 1.6 -S-PLGA polymersomes, the reaction of Cu 1.6 S NPs with GSNO   was examined to verify the production of NO. To determine that NO can be generated from the reaction between Cu 1.6 S NPs and GSNO, the oil-phase Cu 1.6 S NPs were transferred to aqueous solution by the modication of hexadecyltrimethylammonium bromide (CTAB) surfactant yielding CTAB-Cu 1.6 S NPs. A solution of CTAB-Cu 1.6 S NPs with a concentration xed at 100 ppm (copper ion concentration) reacted with a solution of GSNO of varied concentration, and the prole of NO release was determined. The generation of NO increased with the concentration of GSNO, with the NO becoming detectable when the concentration of GSNO exceeded 10 À5 M (Fig. S4a †). A GSNO solution of xed concentration (10 À3 M) was treated with CTAB-Cu 1.6 S NPs solution of varied concentration, and the results show that the release of NO increased with the concentration of CTAB-Cu 1.6 S NPs and reached a plateau at a concentration of 100 ppm (Fig. S4b †); the strong dependence of NO release on the concentration of CTAB-Cu 1.6 S NPs indicates that NO was generated mainly through the reaction between CTAB-Cu 1.6 S NPs and GSNO but not the direct decomposition of GSNO.
We proceeded to investigate whether NO can be triggered to be released from GSNO/Cu 1.6 S-PLGA polymersomes by light irradiation. The UV-visible absorption spectra of PLGA, Cu 1.6 S NPs and Cu 1.6 S-PLGA polymersomes were analyzed (Fig. 2a). PLGA alone did not have discernible absorption with red light, whereas the absorbance of the Cu 1.6 S NPs began to rise from approximately 600 nm. When the Cu 1.6 S NPs were encapsulated in polymersomes (Cu 1.6 S-PLGA), the absorption increased signicantly, which was possibly caused by the aggregation of Cu 1.6 S NPs in the polymersome membrane. The appreciable absorption of Cu 1.6 S-PLGA polymersomes in the red spectral range enables the induction of a photothermal effect with red light, which presumably exerts less phototoxicity to biological cells relative to blue and UV light, and is preferable considering the prospective biomedical applications. Fig. 2b displays the temperature elevation curves obtained from the solutions of Cu 1.6 S-PLGA polymersomes of varied concentration aer illumination with a diode laser (l ¼ 633 nm) at 1 W cm À2 for 15 min. The temperature of the solutions of all concentrations increased effectively, with the solutions of greater concentration exhibiting more rapid heating. A TEM image conrmed that light irradiation caused destruction of the Cu 1.6 S-PLGA polymersome and a leak-out of Cu 1.6 S NPs from the polymersome (Fig. 2c). Having demonstrated photothermally induced rupturing of the Cu 1.6 -S-PLGA polymersomes through the absorption of Cu 1.6 S NPs, we proceeded to examine whether NO was produced from GSNO/ Cu 1.6 S-PLGA polymersomes aer light irradiation under the same conditions. The release of NO was evidenced and increased with the concentration of polymersomes (represented with the concentration of copper ions) (Fig. 2d). These collective results indicate that NO was generated through the reaction of GSNO with Cu 1.6 S NPs as a result of the disintegration of PLGA polymersomes induced by a photothermal effect.
It is known that the generation of NO through the reaction of nitrosothiols with Cu + is accompanied with the formation of Cu 2+ . 8,11 To further verify our preceding notion, X-ray photoelectron spectroscopy (XPS) was employed to determine the oxidation states of the copper ions of Cu 1.6 S NPs before and aer exposing GSNO/Cu 1.6 S-PLGA polymersomes to light from a 633 nm laser. The binding energies of Cu + are at 932.6 and 952.5 eV, whereas those of Cu 2+ are at 933.6 and 954.1 eV. 12 As shown in the XPS spectra, the Cu 2+ /Cu + ratio, which was 0.31 (2P 3/2 ) and 0.29 (2P 1/2 ) prior to laser irradiation (Fig. S5a †), increased to 0.76 (2P 3/2 ) and 0.43 (2P 1/2 ) aer 15 min of laser irradiation (Fig. S5b †), and hence supports our preceding deduction.

Evaluation of GSNO stability in polymersomes
As GSNO is unstable under physiological conditions, it would be most desirable if encapsulation of GSNO in PLGA polymer-somes could improve the chemical stability of GSNO. Free GSNO decomposed spontaneously at 37 C, resulting in a signicant release of NO within one hour, and the generation of NO reached a maximum at the third hour in both H 2 O and PBS (Fig. S6a †). Although the decomposition of GSNO in polymersomes was still evident, the encapsulation signicantly improved the stability of GSNO in both H 2 O and PBS; specically, the release of NO from GSNO/Cu 1.6 S-PLGA polymersomes in H 2 O was reduced in the time course between 3 and 4 h relative to that from free GSNO (Fig. S6b-e †). Overall, these results show that the encapsulation of GSNO in PLGA polymersomes signicantly improved the stability of GSNO relative to free GSNO especially in short runs (e.g. <3 h), a favorable feature considering its prospective application as an NO donor under physiological conditions. Another test shows that GSNO encapsulated polymersomes can be effectively stored without disintegration at 4 C in H 2 O and PBS (Fig. S7 †). The low temperature signicantly extended the stability of GSNO and prevented the spontaneous release of NO from GSNO for up to 3 days. The SEM images show that most of the GSNO/Cu 1.6 S-PLGA polymersomes retained their general spherical morphology in both H 2 O and PBS at 4 C aer 7 days (Fig. S8 †). Careful examination of polymersomes stored in PBS revealed that approximately 13% of the polymersomes showed features of minor to moderate destruction in their structures.

Photo-triggered intracellular release of NO from polymersomes
We next demonstrated photo-triggered intracellular release of NO on demand in vitro. MRC-5 cells were incubated in a medium with or without GSNO/Cu 1.6 S-PLGA polymersomes, and confocal uorescence images were acquired. The nuclei were highlighted with Hoechst (blue uorescence), whereas the generation of NO was probed with a uorescent sensor of NO, DAF-2 DA (green uorescence) (Fig. 3); when the cell-permeable NO sensor enters cells, it is hydrolyzed to DAF-2 by intracellular esterase, and becomes uorescent aer reaction with NO. 14,15 For cells alone, no discernible NO-based green uorescence was seen neither before nor aer light irradiation (633 nm, 1 W cm À2 , 15 min). When cells were cultivated in a medium containing free GSNO (4.12 mM) for 1 h, the green uorescence became evident, indicating the generation of NO, presumably from the spontaneous decomposition of GSNO at 37 C; nevertheless, the uorescence intensity did not vary signicantly with or without light irradiation (633 nm, 1 W cm À2 , 15 min). We proceeded to examine two experimental groups, which corresponded to cells incubated for 1 h in a medium of GSNO/Cu 1.6 S-PLGA polymersomes (200 ppm, corresponding to encapsulated GSNO of 4.12 mM) with or without exposure to light irradiation (633 nm, 1 W cm À2 , 15 min). No green uorescence was detected for the group not subjected to light irradiation, a result consistent with our preceding observation that encapsulated GSNO was stable for a few hours. In contrast, strong green uorescence was observed for the group of cells subjected to light irradiation, and the uorescence intensity was signicantly greater than that observed for cells incubated with free GSNO. Similar results were obtained for the cells cultured in a medium containing GSNO/Cu 1.6 S-PLGA polymersomes (200 ppm) for a longer duration (3 h vs. 1 h) (Fig. S9 †). Nevertheless, green uorescence was discernible for the group of cells not subjected to light irradiation; the result indicates that NO was spontaneously generated, which is consistent with the observation of the stability of GSNO-PLGA polymersomes and GSNO/Cu 1.6 S-PLGA polymersomes at 37 C ( Fig. S6 †). Comparison of the results obtained for cells incubated with GSNO/Cu 1.6 S-PLGA polymersomes for varied durations (3 h vs. 1 h) shows that the longer duration of incubation resulted in a greater uorescence intensity, which indicates that the uptake of polymersomes by cells increased with the duration of incubation. It is noted that Cu 1.6 S-PLGA polymersomes without encapsulation of GSNO that received a 15 min irradiation by a 633 nm diode laser at 1 W cm À2 exhibited at least 90% cell viability, indicating no apparent overheating effect to damage cells.
Biocompatibility of GSNO/Cu 1.6 S-PLGA polymersomes The toxicity of the GSNO/Cu 1.6 S-PLGA polymersomes was assessed on the cell line of MRC-5 using the MTT assay. The protocols of the preceding uorescence imaging experiments were followed, except cells were incubated with GSNO/Cu 1.6 -S-PLGA polymersomes for varied durations (1, 2, 3 and 4 h) at 37 C; the survival rate exceeded 95% for an incubation of 1 h but decreased to 80% when the incubation was prolonged to 4 h (Fig. 4a). Cellular survival was also examined for cells incubated with GSNO/Cu 1.6 S-PLGA polymersomes of varied doses but for the same duration (2 h); the viability of the cells decreased slightly with the dose of GSNO/Cu 1.6 S-PLGA polymersomes (Fig. 4b). Particularly, under 2 h incubation, the GSNO/Cu 1.6 -S-PLGA polymersomes exerted some toxicity but still had good biocompatibility with cells, with a survival rate of at least 85% up to a dosage of 200 ppm; conversely, Cu 1.6 S-PLGA polymersomes (200 ppm) with no inclusion of GSNO exerted negligible toxicity to cells incubated for 24 h at 37 C (Fig. 4c). These observations indicate that NO released from GSNO can cause a certain degree of cytotoxicity. In short, these results show that the biocompatibility of GSNO/Cu 1.6 S-PLGA polymersomes can be adjusted through the dose and the duration of incubation. Notably, the dosage used for the later vasodilation experiments in zebrash (0.04 ppm of GSNO/Cu 1.6 S-PLGA polymersomes with 0.82 nM encapsulation of GSNO) displayed no toxicity in cells aer 24 h of incubation at 37 C (Fig. 4c).

Cerebral vasodilation in zebrash model
Having demonstrated the photo-triggered generation of NO in vitro, the in vivo studies were performed to investigate the possibility of inducing the generation of NO by light and the bioactivity of the generated NO using zebrash (Danio rerio) as a model. Before the experiments, we rst validated the protocols for the detection of NO using a uorescent probe (DAF-FM DA) in living zebrash, and for the determination of the diameter of cerebral blood vessels of zebrash. For the latter, a transgenic line of zebrash, Tg(kdrl:mcherry), which expresses mCherry in the vascular endothelium, was employed. As described, GSNO decomposes spontaneously to produce NO, and hence was chosen to act as an NO donor for this test. As expected, the uorescence intensity was much stronger for the larva soaked sequentially with GSNO accompanied with the NO probe relative to that soaked with the NO donor alone (Fig. S10a †). On further examination using the transgenic zebrash, the larva soaked with GSNO exhibited signicant vasodilation with the diameter of the basilar artery increasing from 12.9 mm to 15.4 mm; in contrast, the untreated control showed a negligible change in the vascular diameter (Fig. S10b †).
We then proceeded to examine whether light irradiation can cause NO release in vivo. The green uorescence increased signicantly aer the irradiation of red light on larvae that were injected with GSNO/Cu 1.6 S-PLGA polymersomes; in contrast, irradiation on the control group, which was not injected with GSNO/Cu 1.6 S-PLGA polymersomes, caused only a minor increase in the uorescence (Fig. 5a). Statistical analysis shows further that the percentage increase of the uorescence for the experimental group is signicantly greater than that of the control group (39.4 AE 5.8% vs. 4.7 AE 0.6%; n ¼ 3, p < 0.01) (Fig. 5b).
Similar to the preceding test, larvae of transgenic zebrash, Tg(kdrl:mcherry), were injected with GSNO/Cu 1.6 S-PLGA polymersomes, irradiated by red light, and then imaged with confocal microscopy. Light irradiation caused a dramatic increase in the diameter of the basilar artery; distinctly, light irradiation caused less of an increase in the diameter of the basilar artery of a larva that was injected with a PBS solution containing no polymersomes ( Fig. 5c and d). In particular, the percentage increase in the vascular diameter was signicantly greater than that observed for the control (27.4 AE 4.7%, n ¼ 8 vs. 17.5 AE 1.1%, n ¼ 4; p < 0.001) (Fig. 5e). With these results taken together, we conclude that NO can be triggered by light to release from GSNO/Cu 1.6 S-PLGA polymersomes in vivo, and such photo-triggered generation of NO can exert vasodilatory activity by enlarging the diameter of cerebral blood vessels of larval zebrash.

Conclusions
In this study, we presented an exogenous stimulus strategy to achieve the stimuli-responsive release of NO on demand. Our study represents for the rst time NO produced by stimulus through the reaction between nanomaterials and GSNO and illustrates a prospective application of our design to act as a photo-controllable vasodilator. The release of NO can be controlled by light irradiation when GSNO and Cu 1.6 S NPs are packed in the core and membrane of PLGA polymersomes, respectively. The encapsulation of GSNO in polymersomes also improves the stability of GSNO, which would otherwise decompose spontaneously under physiological conditions. We anticipate that the increased stability of encapsulated GSNO and the unique ability of photo-triggered NO release will open up new possibilities of novel NO based therapies.