Preparation and characterization of HA-PDA@IQ/DOX NPs
As shown in Scheme1, HA-PDA@IQ NPs were synthesized by the one-pot self-polymerization of dopamine initiated by KMnO4 in the presence of stabilizer (HA) and immune adjuvant (IQ). Then, the chemotherapy drug (DOX) was loaded via π-π stacking to form HA-PDA@IQ/DOX NPs. As shown in Figure 1a, the TEM result showed that the prepared HA-PDA@IQ/DOX NPs were spherical, uniformly distributed without obvious aggregation. The average size and zeta potentials of HA-PDA@IQ NPs and HA-PDA@IQ/DOX NPs were 136.4 nm and 150.9 nm, -23.1 mV and -18.8 mV, respectively (Figure 1b, c). The changes in size and zeta potential were attributed to DOX loading. In addition, elemental mapping images displayed that Mn element has been successfully integrated into HA-PDA@IQ/DOX NPs during oxidative polymerization (Figure 1d). XRD spectrum indicated that there was broad diffraction peak at 23 °, which represented non crystalline phase structure in HA-PDA@IQ/DOX NPs (Figure 1e). As shown in Figure 1f, the chemical structure and composition of HA-PDA@IQ/DOX NPs were analyzed by X-ray photoelectron spectroscopy (XPS). There were four peaks at 641.48 eV, 532.59 eV, 399.88 eV and 296.29 eV, corresponding to the four elements of manganese, oxygen, nitrogen and carbon, which proved the element composition of HA-PDA@IQ/DOX NPs. As shown in Figure 1g, there were two typical binding-energy peaks at 654.2 eV and 641.1 eV, corresponding to Mn (IV)2p1/2 and Mn (IV)2p3/2, respectively, which further confirmed the presence of Mn element in HA-PDA@IQ/DOX NPs. The C1s spectrum showed three main peaks, 287.7 eV, 286.2 eV, and 284.4 eV (Figure 1h), corresponding to C=0, C-N and C-C, respectively. The O1 spectrum had two obvious peaks at 532.98 eV and 531.1 eV, which were attributed to C=0 and C-O (Figure 1i). The N1s spectrum showed four peaks, 399.9eV, 399.5eV, 398.9ev and 400.5eV, which indicated the presence of N atoms (Figure 1j). These hydrophilic chemical groups endowed the HA-PDA@IQ/DOX NPs with superior dispersibility and stability in aqueous solution, which were essential for biomedical applications.
Photothermal properties of HA-PDA@IQ/DOX NPs
Photothermal conversion performance of HA-PDA@IQ/DOX NPs was investigated upon 808 nm laser irradiation (Figure 2a).Firstly, the UV-Vis spectra of HA-PDA@IQ/DOX NPs showed a wide absorption band extending to the infrared region, where the absorption intensity increased with the increase of concentration (0.2-1.0 mg/mL) in Figure 2b. Next, we studied the temperature changes of HA-PDA@IQ/DOX NPs solution with different concentrations under laser irradiation of 2 W/cm2, as shown in Figure 2c. Generally, the temperature of each group from 0.1 to 1 mg/mL quickly increased with the increase of HA-PDA@IQ/DOX NPs concentration. By contrast, there was no obvious change in temperature in de-ionized water group. Notably, the temperature of HA-PDA@IQ/DOX NPs group even reached 59.4 °C within 5 minutes under 808 nm laser irradiation when the concentration was 1 mg/mL. In addition, we tested the temperature changes of HA-PDA@IQ/DOX NPs (1mg/mL) group at different power densities from 1 to 2.5 W/cm2 (Figure 2d), and recorded the heat map using an infrared camera (Figure 2e). As we expected, the temperature exhibited power density-dependent manner, even exceeded 60°C under the power density of 2.5 W/cm2. Compared with commercial photosensitizer ICG, photothermal stability of HA-PDA@IQ/DOX NPs (200 μL, 1.0 mg/mL) was studied during the five cycles of on/off laser. As shown in Figure 2f, the temperature of the ICG solution gradually decreased after five rounds of repeated laser irradiation due to inherent photoquenching. However, the temperature of the HA-PDA@IQ/DOX NPs solution remained stable during five irradiation cycles, indicating that HA-PDA@IQ/DOX NPs have a highly stable photothermal effect. Besides, there was no significant change in the Vis-NIR absorbance of HA-PDA@IQ/DOX NPs before and after five rounds of repeated laser irradiation (Figure 2g). According to the calculation formula in Figure 2h and 2i, the photothermal conversion efficiency (η) of HA-PDA@IQ NPs was 41.2%, which was obviously much higher than those of previously reported PTT agents [39]. The high η of HA-PDA@IQ/DOX NPs could be attributed to its strong absorption at the near-infrared (NIR) wavelength of 808 nm and the increase in electron transport efficiency induced by manganese element.
Dox loading and release
It is well known that aromatic molecules can be efficiently loaded on nanoparticles with delocalized π electrons through π-π stacking and hydrophobic interaction. Since PDA contains a delocalized π electronic structure, we envisioned that HA-PDA@IQ can also be used as a nano-drug delivery carrier to load the doxorubicin (common chemotherapy drug, DOX). In our experiment, HA-PDA@IQ and DOX were mixed in different ratios and stirred overnight at room temperature. After washing, the obtained HA-PDA@IQ/DOX NPs were detect by UV-Vis-NIR spectroscopy (Figure 3a). Compared with pure HA-PDA@IQ NPs, the HA-PDA@IQ/DOX NPs has a characteristic absorption peak at 490 nm, indicating that DOX was successfully loaded. As the weight ratio of DOX: PDA increased, the drug loading on the nanoparticles also increased (Figure 3b). The maximum loading reached about 130% (DOX: PDA, w/w), which seemed to be much higher than traditional polymer-based NDDS, and the loading capacity at this time reached 37.2%.
Next, we studied the drug release behavior of DOX in HA-PDA@IQ/DOX NPs.The supernatant was collected by dissolving HA-PDA@IQ/DOX in PBS with different pH values (5.0 and 7.4) with gentle agitation (Figure 3c). The amount of released DOX was analyzed by UV-Vis-NIR spectroscopy. After incubation of 24 hours, the released DOX at pH 5.0 was about 45%, while the released DOX at pH 7.4 was only 15.2%. Therefore, DOX release seems to be more pronounced at a lower pH (5.0) compared with physiological pH (7.4). This acid-triggered release was attributed to the protonation of the amino group in the DOX molecule under low pH. Then, we wondered whether the NIR-induced heating could promote the release of DOX. The HA-PDA@IQ/DOX NPs in pH 5.0 and 7.4 PBS were irradiated by 808 nm laser (2 W/cm2, 5 minutes) at different time points. As shown in Figure 3d, the cumulative release of DOX in HA-PDA@IQ/DOX NPs was measured by UV-Vis-NIR spectroscopy. Compared with no laser irradiation treatment, released DOX exhibited periodical fluctuation along laser on/off irradiation, demonstrating that irradiation could significantly enhanced the release of DOX from HA-PDA@IQ/DOX NPs. In view of these features, acidic environment of tumor tissue and HA-PDA@IQ/DOX NPs-mediated photothermal effect were conducive to the controllable release of DOX.
Biocompatibility and PTA effect of HA-PDA@IQ NPs in vitro
In order to eliminate the interference of DOX, HA-PDA@IQ NPs without DOX was used in the following experiments. To evaluate the biocompatibility, the CCK-8 experiment was performed using two types of cells, including normal mouse embryonic fibroblast (MEF) cells and mouse breast cancer cell 4T1 cells. Both MEF cells and 4T1 cells were cultured with different concentrations of HA-PDA@IQ NPs (0, 200, 400, 600, 800, 1000, 1200 μg/mL) for 48 h, respectively. As shown in Figure 5a, HA-PDA@IQ NPs showed negligible cytotoxicity to both MEF cells and 4T1 cells. Even at a high concentration of 1200 μg/mL, both these cell survival rate exceeded 80%. The hemolysis test was also used to evaluate the blood compatibility of HA-PDA@IQ NPs. As shown in Figure 5d, compared with the deionized water group, the absorbance at 541 nm did not exceed 0.04 even at 1200 μg/mL. These above results indicated that the prepared HA-PDA@IQ NPs has good biocompatibility.
After that, we evaluated the effect of HA-PDA@IQ NPs mediated PTA on tumor growth in vitro. The tumor cells 4T1 and normal cell MEF were co-cultured with different concentrations of HA-PDA@IQ NPs and irradiated with 808 nm laser for 5 minutes (2 W/cm2). It can be seen from Figures 4b and 4c that the cell viability of 4T1 cells was lower than 50 % at 800 μg/mL, while the cell viability of MEF cells was higher than 80 % even at same concentration. The findings proved that PTA significantly inhibited the tumor cell viability, but has little effect on normal cells. The possible reason was that normal cells are more resistant to temperature than cancer cells. Furthermore, live/dead cell assay was conducted to verify the effect of PTA on tumor cells. It can be seen from Figure 4e that there was almost no cell death in control, NIR and HA-PDA@IQ NPs treatment. In contrast, HA-PDA@IQ NPs+NIR caused mass cell death, indicating that HA-PDA@IQ NPs had promising potential in PTA of tumor.
The effect of HA-PDA@IQ NPs on DC maturation.
Firstly, the intracellular localization and real-time cell uptake of HA-PDA@IQ/DOX NPs were studied via fluorescence microscopy after incubation 4T1 cells with HA-PDA@IQ/DOX NPs for 24 hours. As shown in Figure 5a, red fluorescence could be observed in the cytoplasm, but no signal in nucleus. These findings showed that HA-PDA@IQ/DOX NPs could be phagocytosed into the cytoplasm and distributed around nucleus.
During the antitumor immune response, DC as a type of antigen presenting cell (APC) plays an essential role in capturing the neoantigen and activating cytotoxicity T lymphocyte. Therefore, immature bone marrow-derived dendritic cells (BMDC) were used to evaluate the effect of HA-PDA@IQ NPs on DC maturation. Flow cytometry analysis was used to study the expression of CD80 and CD86 on the surface of BMDCs, which represented the level of DC maturation. Figure5b showed the expression of CD80 and CD86 on the surface of BMDCs increased with the increase of HA-PDA@IQ NPs concentration. Compared with PBS treatment (only 2.33%), CD80 and CD86 expression in HA-PDA@IQ NPs treatment was up to 10.31%. Based on these findings, we have reason to believe that resultant HA-PDA@IQ NPs have a powerful effect on triggering DC maturation, which may be beneficial to anti-tumor immune activation.
MRI performance of HA-PDA@IQ/DOX NPs in Vitro and in Vivo.
Due to the Mn element doping, we speculated that HA-PDA@IQ/DOX NPs might possess the capability to be utilized as a potential contrast agent (CAs) for MR imaging. Herein, the 3T MR scanner was used to study the T1-weighted MR of HA-PDA@IQ/DOX NPs in vitro. As shown in Figure 7a, there was a positive correlation between HA-PDA@IQ/DOX NPs solutions with different Mn concentrations (0-0.2 nM) and T1-weighted MR signal intensity. Furthermore, HA-PDA@IQ/ DOX NPs encapsulated with F127 hydrogel was in situ injected into tumor site to access their imaging capability. As shown in Figure 7b, strong signal could be observed in tumor region after HA-PDA@IQ/DOX NPs injection. Then, we further studied the retention and distribution of HA-PDA@IQ/DOX NPs at tumor site for 48 hours. As shown in Figure 7c, single HA-PDA@IQ/ DOX NPs disappeared quickly within 6 hours, while HA-PDA@IQ/DOX HG still well remained in tumor region under 48 hours. It could be seen that HA-PDA@IQ/DOX HG had a strong retention effect at tumor site, which was conducive to the recruitment of APCs as well as recognition of neoantigen derived from necrotic tumor cells.
Immune responses assessment in vivo
In view of the excellent photothermal performance and biocompatibility, we further studied the immunoregulatory function of HA-PDA@IQ/DOX NPs in vivo using 4T1 subcutaneous tumor-bearing mice as animal model (Figure 7a). When the average tumor volume reached about 100 mm3, the 4T1 tumor-bearing mice were divided into five groups (PBS, HA-PDA NPs, HA-PDA@IQ NPs, HA-PDA@IQ/DOX NPs and HA-PDA@IQ/DOX HG) for in-situ injection under the 808 nm NIR irradiation (2 W/cm2, 5 min). The maturation of DC cells in the inguinal lymph nodes of mice with breast cancer was detected by flow cytometry (Figure 7b). The data showed that the ratio of CD11c+/CD80+ (1.27%) and CD11c+/CD86+ cells (1.52%) in HA-PDA treatment group was a little bit higher than that in control group PBS (1.03%, 1.46%), which may be attributed to photothermal treatment influence. Notably, the proportion of CD11c+/CD80+ cells and the proportion of CD11c+/CD86+ cells in HA-PDA@IQ NPs treatment group were higher than those in HA-PDA NPs treatment group, confirming that immune adjuvant IQ doping could stimulate DC maturation. Meanwhile, the ratio of CD11c+/CD80+ and CD11c+/CD86+ cells in HA-PDA@IQ/DOX HG group were higher than that in HA-PDA@IQ/DOX NPs group. In the HA-PDA@IQ/DOX HG group, the DC maturation rate was the highest (6.43% for CD11c+/CD80+ cells and 6.67 % for CD11c+/CD86+ cells). The reason may be that the addition of the gel prolonged the residence time of the nanoparticles in the tumor and enhanced the recruitment of APC cells. In order to study the potential utility of this treatment strategy for the long-term prevention of tumor recurrence, we used flow cytometry to detect the inducibility of effector T cells in inguinal lymph nodes (CD3+/CD8+/CD44high/CD62Llow). As shown in figure 7c, the ratio of memory T cells reached 22.64% under HA-PDA@IQ/DOX HG treatment, which was 6.2 times compared with control group (PBS). These results demonstrated that HA-PDA@IQ/DOX HG had the ability to promote memory T cells production.
Cytotoxic T lymphocytes (called CTL or CD8+ T cells) are essential for the anti-tumor immune response where they are activated by tumor-derived antigens and directly destroy target cells. In this experiment, each group of splenocytes was co-stained with anti-CD3 and anti-CD8a antibodies and was measured by flow cytometry. As shown in Figure 7d, the percentage of CD3+/CD8+ T cells in PBS, HA-PDA NPs, HA-PDA@IQ NPs, HA-PDA@IQ/DOX NPs and HA-PDA@IQ/DOX HG group were 2.30%, 2.68%, 5.05%, 5.43% and 5.68%, respectively. It was worth noting that the percentage of CD3+/CD8+ T cells in HA-PDA@IQ NPs (5.05%) group was about 1.88 times as that in HA-PDA NPs (2.68%). This enhancement of CTL might be ascribed to the introduction of IQ, which remarkably stimulated the DC cells maturation and enhanced the tumor-derived antigens presentation function.
The antitumor effect of HA-PDA@IQ/DOX HG in vivo
To verify the therapeutic effect of HA-PDA@IQ/DOX HG, subcutaneous 4T1-Luc breast tumor model was established in the inner mammary gland of lower left posterior breast (Figure 8a). After 7 days of treatment, the tumors were remarkably inhibited in HA-PDA@IQ/DOX HG group compared with other groups in figure 8b. It was noted that the tumors in HA-PDA@IQ/DOX HG group completely disappeared on the 20th day. Individual tumor growth curves of each group further confirmed this trend (Figure 8c). It can be seen that HA-PDA@IQ/DOX HG with irradiation possessed application potential in eliminating the tumor in situ.
The PTA performance of HA-PDA@IQ/DOX HG in vivo
In order to evaluate the photothermal performance in vivo, digital NIR thermal imager was chosen to real-time monitor temperature profile of tumor site after in situ injection of HA-PDA@ IQ/DOX HG. As shown in Figure 9a and 9b, the surface temperature of the tumor in each group increased sharply, except for the PBS group. Notably, the temperature of tumor in HA-PDA@IQ/DOX HG group reached its maximum and even up to 56.6 °C after 5 min of laser irradiation. After that, irradiation was performed every other day for a total of three times and corresponding tumor volume was recorded at certain time interval. As shown in Figure 9c, the growth curves of the tumor within 20 days further verified that HA-PDA@IQ/DOX HG group exhibited optimal anti-tumor effect compared to other control groups. However, there was no significant change in body weight among all groups in Figure 9d. It was worth noting that the survival period of the mice in HA-PDA@IQ/DOX HG group was as long as 60 days, while the average life span of mice in the control group was shorter than 30 days in Figure 9e. The life span of mice in other treatment groups was extended to varying degrees, which was attributed to the lack of DOX and hydrogel collaboration. These results proved that PTA, hydrogel and chemotherapy could reinforce each other, finally obtain an ideal therapeutic effect. In addition, in order to verify its biological safety in vivo, the heart, liver, spleen, lung and kidney of the mice were taken out for H&E staining analysis (Figure 9f). The images showed that there were no inflammatory lesions or obvious tissue damage.