De Novo Design of Polymeric Carrier to Photothermally Release Singlet Oxygen for Hypoxic Tumor Treatment

Intratumoral hypoxia extremely limits the clinic applications of photodynamic therapy (PDT). Endoperoxides allow thermally releasing singlet oxygen (1O2) in a defined quantity and offer promising opportunities for oxygen-independent PDT treatment of hypoxic tumors. However, previous composite systems by combining endoperoxides with photothermal reagents may result in unpredicted side effects and potential harmful impacts during therapy in vivo. Herein, we de novo design an all-in-one polymer carrier, which can photothermally release 1O2. The strategy has been demonstrated to effectively enhance the production of 1O2 and realize the photodamage in vitro, especially in hypoxic environment. Additionally, the polymer carrier accumulates into tumor after intravenous injection via the enhanced permeation and retention effects and accelerates the oxygen-independent generation of 1O2 in tumors. The oxidative damage results in good inhibitory effect on tumor growth. Realization of the strategy in vivo paves a new way to construct photothermal-triggered oxygen-independent therapeutic platform for clinical applications.

Polycyclic aromatic hydrocarbons are the most reliable organic compounds to cleanly supply 1 O 2 in a defined quantity without side reaction [20][21][22][23][24]. They can trap 1 O 2 yielding endoperoxides (EPOs) and release 1 O 2 upon elevating the temperature, which provides a new and powerful concept in 1 O 2 delivery for the treatment of hypoxic tumors. Till now, only few reports have developed the composite systems by combining EPOs with photothermal reagents (such as gold nanorods) to release 1 O 2 in cancer cells [25,26]. Although they realized photothermal-triggered oxidative damage of cancer cells in in vitro experiments, the composite structures may result in unpredicted side effects and potential harmful impacts on biological environments during therapy in vivo. Moreover, the utilization of gold nanorods substantially increases the difficulty of clearance from the body and leads to long-term toxicity towards healthy tissues and organs [27]. In this regard, it is highly desirable to de novo design an allin-one strategy without the above-mentioned concerns.

P2 P3
Methylene blue, ／ 2 , Ｂ] Ｂ] only an excellent organic photothermal agent, but also potential imaging reagents in vivo because of good physiological stability and near-infrared absorption and luminescence [28]. The model polymers P2 and P3 were also obtained ( Figure 1(b)). The detailed synthesis procedures of the polymer dots with DMN and aza-BODIPY pendants were illustrated in supporting information. The monomers and polymers were characterized by NMR spectra, gel permeation chromatography (GPC), transmission electron microscope (TEM) images, dynamic light scattering (DLS), and matrixassisted laser desorption ionization time of light mass spectrometry (MALDI-TOF MS).
The number-average molecular weights (Mn) of P1-P3 were 19674, 21024, and 16024, respectively. The polydispersity index (Mw/Mn) of P1-P3 was 1.20, 1.31, and 1.24, respectively. A uniform spherical morphology of P1 dots was clearly revealed by TEM images (Figure 2(a)). The diameters of P1 dots were estimated to be around 20 nm. DLS indicated that the P1 dots were well-dispersed in water and had an average hydrodynamic size of 23 nm (Figure 2(c)), which contributed to be enriched in tumor tissues via EPR effect [29]. After loading 1 O 2 , the results of TEM and DLS of P1-SO dots remain almost the same as those of P1 dots (Figures 2(b) and 2(d)), demonstrating that capture of 1 O 2 did not change the morphology, particle size, or dispersity of the polymer dots.

Photophysical, Photothermal, and Photodynamic Properties of P1-SO Dots. The photophysical properties of P1
and P1-SO dots were investigated by UV-Vis absorption and emission spectra. As illustrated in Figure 2(e), the absorption spectra of P1 dots displayed a strong absorption at 243 nm and a weak absorption at 689 nm, which were consistent with the absorption of DMN and B1 monomer, respectively. After capture of 1 O 2 , the destruction of conjugated structure of DMN leaded to a sharp decrease at 243 nm in the UV-Vis absorption spectra [26]. In addition, P1 dots and P1-SO dots showed almost the same emission maxima at 730 nm ( Figure 2(f)), which was attributed to the luminescence of B1 (see Figure S1 in the Supplementary Material) and negligibly affected by the 1 O 2 loading.
Thermal effects were the key factor to trigger the releasing of 1 O 2 . Therefore, the photothermal performance of monomer B1, P1, and P1-SO was carried out via thermal infrared imager (FLIR E40). The temperature change was recorded at different concentration in dimethylsulfoxide (DMSO) under continuous exposure to irradiation (690 nm, 400 mW/cm 2 ). As shown in Figure S2 in the Supplementary Material and Figures 3(a) and 3(b), the increasing concentration resulted in the elevation of temperature. After irradiating for 360 s, the temperatures increased by 21.5 ∘ C, 20.0 ∘ C and 19.8 ∘ C for B1 (40 M), P1 (300 g/mL), and P1-SO (300 g/mL), respectively, demonstrating that the polymer containing B1 units displayed good photothermal effects.
To investigate the 1 O 2 capture ability of P1, MB was added to the DMSO solution of P1. As shown in Figure 3(c), when the mixture solution was irradiated by a 660 nm laser (4 mW/cm 2 ), the absorption band at 243 nm decreased gradually with the extension of irradiation time, but the absorption of MB at 665 nm remained unchanged, indicating that 1 O 2 sensitized by MB was captured by P1 to form P1-SO. After removing MB, the abilities of 1 O 2 release of P1-SO were studied at different temperatures ( Figure 3(d)). When the temperature was kept at 37 ∘ C, rare 1 O 2 was released according to the negligible response of the absorption band of the 1 O 2 indicator, 1,3-diphenylisobenzofuran (DPBF). In contrast, a sharp decrease of absorption band was observed when the temperature was increased to 50 ∘ C, indicating of a large amount of 1 O 2 generated. These results demonstrated that elevated temperature facilitated the release of 1 O 2 . Moreover, to investigate the release of 1 O 2 under photothermal stimulation, the generation of 1 O 2 of P1-SO was measured under irradiation of laser (690 nm, 400 mW/cm 2 ) in air and hypoxia environment that was produced by bubbling with nitrogen gas (Figure 3(e)). In the DMSO solution of P1-SO and DPBF, obvious decrease of the absorption band ascribed to DPBF was observed in both air and hypoxia environments, indicating that the oxygen levels had negligible influence for the 1 O 2 release. On the other hand, in the control group, when P1 was used instead of P1-SO, it was hardly able to generate 1 O 2 even in air since 1 O 2 was not trapped in the DMN units and heavy atom-free B1 had no ability to generate 1 O 2 . Furthermore, to demonstrate the reversibility of the capture and release of 1 O 2 for P1-SO, we added MB to the DMSO solution containing P1 and exposed the mixture under light irradiation (660 nm, 4 mW/cm 2 ). The absorption band at 243 nm was decreased after the light irradiation (660 nm, 4 mW/cm 2 ) for 120 min owing to the capture of 1 O 2. When the light irradiation (690 nm, 400 mW/cm 2 ) was prolonged for 25 min, the absorption band at 243 nm was increased because of the 1 O 2 releasing. MB was not removed in this process. The reversible performance was displayed for 5 cycles (Figure 3 All the results indicated that the photothermal effects of B1 could provide enough heat to trigger 1 O 2 release and P1-SO had excellent phototoxicity effects in solution. More importantly, oxygen is unnecessary in 1 O 2 release process of P1-SO compared to conventional photosensitizers, revealing enormous potential for improving the therapeutic effects of hypoxia-associative PDT.

Anticancer Investigation In Vitro.
To demonstrate the feasibility of the polymer carrier to generate 1 O 2 in vitro, 2,7-dichlorifluoresceindiacetate (DCFH-DA), which can be oxidized to 2,7-dichlorofluorescein (DCF) by intracellular ROS in live cells, was utilized as a ROS tracer agent. The laserscanning confocal luminescence microscopy was employed to investigate the ROS generation in HeLa cells. Upon irradiation at 690 nm, weak luminescence in the cells was observed under 21% (see Figure S6 in the Supplementary Material) and 5% oxygen concentration (Figure 4(a)) when   the cells were incubated with P1. But bright green luminescence of DCF was exhibited in the cells incubated with P1-SO under 21% and 5% oxygen concentration. Without light irradiation, the generation of ROS hardly occurred in the cells incubated with P1-SO, since the physiological temperature (37 ∘ C) was relatively low so that it could not trigger the rapid release of 1 O 2 from endoperoxides (see Figure S7 in the Supplementary Material). The control group containing the cells only incubated with DCFH-DA displayed negligible luminescence in the absence and presence of irradiation at 690 nm (see Figure S8 in the Supplementary Material). These results demonstrated that the intracellular release of 1 O 2 from endoperoxides could be accelerated by photothermal effects of B1 in P1-SO. The methyl thiazolyl tetrazolium (MTT) assay was used to evaluate the cytotoxicity of the polymer carrier towards HeLa cells. HeLa cells were incubated with different concentrations of P1 or P1-SO at 37 ∘ C for 24 h in dark. Low dark cytotoxicity of P1 and P1-SO were shown under 5% and 21% oxygen. When the P1 incubated cells were treated by 690 nm laser, the cell viability was relatively high for cells in 21% or 5% oxygen concentrations, indicating low toxicity of P1 (Figures 4(c) and S9 in the Supplementary Material).
Additionally, the cells incubated with 1 O 2 loading P1-SO showed relatively lower cell viability in 21% (see Figure S10 in the Supplementary Material) or 5% (Figure 4(d)) oxygen compared to the group of P1-treated cells under irradiation, revealing that the 1 O 2 release induced the oxidative damage and was the dominant reason to kill cells.
To study the therapy performance via cell apoptosis assay, Calcein-AM and PI were used to label the living and dead cells as indicators by staining the cytoplasm with green fluorescent AM and the nucleus with red fluorescent PI, respectively. The cells remained alive in dark. Under irradiation, the cells incubated with P1 remained alive even under 21% oxygen, whereas the cells incubated with P1-SO were dead under either 21% or 5% oxygen concentration (Figures 4(b) and S11 in the Supplementary Material). Without irradiation, the cells treated with P1 or P1-SO mostly remained alive under 21% and 5% oxygen levels (see Figure S12 in the Supplementary Material). These results confirmed that the oxidative damage could be achieved by P1-SO in hypoxic cancer cells. To further determine the cell population at different stages of apoptosis, the flow cytometry experiments were performed (Figures 4(e) and 4(f) and S13 in the Supplementary Material). Under irradiation by laser at 690 nm, the cells incubated Research 7 with P1 were still alive but those incubated with P1-SO were dead after 6 hours under 5% oxygen condition (Figures 4(e) and 4(f)). All the results indicated that photothermal effect of P1-SO triggered the ROS generation and the oxidative damage dominated the therapeutic effects, encouraging us to investigate the potential application of P1-SO for cancer therapy in hypoxia environments.

Anticancer Investigation In Vivo.
To test the in vivo behaviors of P1-SO, the polymer carrier was intravenously injected into the HeLa tumor-bearing mice, and then their biodistributions were evaluated at 1,4,8,12,24, and 48 h after injection (Figure 5(a)). After injection of P1-SO for 1 h, the gradually enhanced fluorescence at the tumors was observed. A maximized distribution was exhibited in tumor at 8 h after injection compared with other major organs ( Figure 5(b)), demonstrating that the highest level of P1-SO was accumulated in tumor. After 48 h, no fluorescence was observed in tumor, revealing the metabolism of P1-SO with time prolonging (Figure 5(c)). Hence, the PDT treatments will be carried out at 8 h after injection of P1-SO. After the anticancer treatments, P1-SO can be eliminated from the body. These results indicated good capacity of the polymer dots to accumulate in tumor, owing to the EPR effect mediated by appropriate particle size.
To prove the photothermal effects of the polymer carrier in vivo, four groups of tumors bearing mice were investigated after 8 h after injection ( Figure 5(d)). Upon irradiation at 690 nm (400 mW/cm 2 ) by a FLIR camera, mice injected with P1-SO and P1 showed temperature increase to about 48 ∘ C in the tumor area. Similar temperature increase was also observed when the mouse was injected with P1-SO and Nacetyl-L-cysteine (NAC), which is a ROS-scavenger. In the control experiment where the mouse was injected with PBS, the temperature was almost kept at 37 ∘ C ( Figure 5(e)). To investigate the ability of ROS generation in tumor, the tumor biopsies of mice with different treatments were studied by confocal imaging (Figure 5(f)). The in vivo generation of 1 O 2 from P1-SO under irradiation was observed using DCFH-DA as the indicator. After intravenous injection of P1-SO for 8 h, NAC was injected into the tumor. With a 690 nm laser irradiation (400 mW/cm 2 ) for 6 min, no luminescence of DCF was found, and P1 treated tumor also displayed no 1 O 2 generation under irradiation. These results prove that P1-SO has strong ability to generate 1 O 2 in vivo under irradiation.
To investigate the in vivo PDT efficacy, P1-SO or P1 was intravenously injected into the mice bearing HeLa tumors of 100-300 mm 3 in the presence or absence of NAC, followed by irradiation at 690 nm laser for 6 min (400 mW/cm 2 ). The tumor size and weight of the mice were recorded every 2 days (Figures 6(a)-6(c)). As a control, the mice injected with PBS showed a 19-fold increase of tumor volumes regardless of irradiation, indicating that the irradiation of laser hardly had evident influence to the growth of tumors. Without irradiation, P1 and P1-SO treated mice also displayed similar tumor growth rate compared to those injected with PBS, demonstrating negligible anticancer efficacy of P1 and P1-SO due to their low dark cytotoxicity. When tumors were exposed to irradiation, P1 treated tumors showed a 10fold increase of tumor volumes, but clear shrink of tumors was found on the mice injected with P1-SO after 2 weeks, indicating that the photothermal effect of B1 was unable to eliminate tumor but can trigger the release of 1 O 2 in P1-SO. Importantly, 1 O 2 generation by P1-SO caused irreversible oxidative damage towards tumor and inhibited the growth of tumor. To further prove the influence of 1 O 2 generation for tumor therapy, the ROSscavenger NAC was intratumorally injected into the P1-SO treated mice before therapy. The tumors did not stop growing and displayed 11-fold increases of tumor volumes after 14 days under irradiation. All the photographs of the mice after treatments were shown in Figure S14 in the Supplementary Material. These results demonstrated that the released 1 O 2 played a dominant role in the therapeutic process.
To further evaluate the detail anticancer efficacy of P1 and P1-SO, the proliferation and morphology of the tumors and organs were investigated by hematoxylin & eosin (H&E) staining ( Figure 6(d)). Firstly, obvious tumor damage was not observed in group PBS with or without irradiation. Few tumor cells were found in the group of P1-SO with irradiation, indicating an ideal ability to release 1 O 2 of P1-SO under irradiation. A small amount of tumor necrosis displayed in group P1-SO + NAC and P1 with irradiation owing to the photothermal effects. No distinct tumor damage was found in P1-SO or P1 without irradiation, which indicated that irradiation played a key role in the therapeutic process. These results demonstrated that tumor damage caused by P1-SO under irradiation was mainly attributed to oxidative damage with few photothermal effects. In addition, P1-SO exhibited negligible influence to the normal tissues (such as heart, liver, spleen, lung, and kidney) during the therapeutic process (see Figure S15 in the Supplementary Material).

Discussion
In this study, a novel all-in-one polymeric 1 O 2 carrier, which can rapidly release cytotoxic 1 O 2 in cancer cells under photothermal stimulation, was developed to overcome the restriction of hypoxic tumor during PDT process in vivo. The capture of 1 O 2 was attributed to the DMN units and the release of 1 O 2 was triggered by the photothermal effect of B1 under NIR light irradiation. The strategy has been demonstrated to effectively enhance the production of 1 O 2 in vitro and realize the photodamage to cancer cells, especially in hypoxic environments. Additionally, introduction of nearinfrared excitable B1 facilitates the potential imaging-guided therapy in vivo. The polymer dots accumulate into tumor after intravenous injection via EPR effect and accelerate the oxygen-independent generation of 1 O 2 . The oxidative damage towards tumor results in good inhibitory effect on tumor growth in vivo. The realization of this concept in vivo not only is a huge boost to the novel thermaltriggered PDT strategy, but also provides a valuable means to construct photothermal-triggered oxygen-independent therapeutic platform for clinical applications.

4.4.
In Vivo Therapy. 24 mice were divided into 8 groups averagely. P1-SO, P1 (300 g/L, 100 L), or PBS were injected. After 8 h, the mice were exposed to a 690 nm laser (400 mW/cm 2 ) for 6 min or not. The weight and tumor volumes were recorded every two days. Volume of tumors was calculated by equation: volume = length × width 2 /2. The relative tumor volume = v/v 0 , v was the tumor volume at different day, v 0 was the tumor volume at first day. All the mice were sacrificed after treatments and tumors and main organs (heart, liver, spleen, lung, and kidney) were fixed by using 4% formalin solution for further histomorphological analysis.

Data Availability
All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials. Additional data related to this paper may be requested from the authors.

Conflicts of Interest
The authors declare no competing financial interests. Figure S1: the UV-Vis absorption and emission spectra of B1. Figure S2: the photothermal properties of B1. Figure  S3: the singlet oxygen trapping properties of P2. Figure  S4: the singlet oxygen generation of P2 and P2-SO. Figure  S5: the singlet oxygen generation of P3. Figure S6: the ROS generation of P1 and P1-SO under irradiation in vitro. Figure S7: the ROS generation of P1 and P1-SO without irradiation in vitro. Figure S8: the ROS generation without photosensitizers in vitro. Figure S9: the cytotoxicity of P1. Figure S10: the cytotoxicity of P1-SO. Figure S11: the PDT effects of P1 and P1-SO under irradiation in vitro. Figure S12: the PDT effects of P1 and P1-SO without irradiation in vitro. Figure S13: flow cytometry of PDT effects under hypoxia. Figure S14: photographs of the mice. Figure S15: H&E stained of the main organs. (Supplementary Materials)