Characterization of physicochemical properties of DIMP NPs
DIMP NPs were obtained through a series of reactions and the particle size was determined with particle size analyzer. From Fig. 1A, the size was 131.9 nm, and the polymer dispersity index (PDI) was about 0.25, indicating that DIMP NPs had suitable particle size and uniform dispersion in aqueous solution. The scanning electron microscope (SEM) showed that the morphology of DIMP NPs was irregular, nearly round, with the particle size of about 100 nm. The larger hydrated particle size of DIMP obtained with dynamic light scattering (DLS) may be caused by swelling in the aqueous solution.
As show in Fig. 1B, the potential of ZIF-8 was 0.65 mV, and the potential after loading DOX and ICG was 0.86 mV, which was conducive to the drug circulation in the blood. The stability of the DIMP was monitored (Fig. 1C), and the particle size and PDI were measured and recorded for 7 days, suggesting good stability. UV–Vis spectrophotometer (Fig. 1D) and fluorescence spectrophotometer (Fig. 1E) were applied to measure the optical properties of DIMP. The UV results showed that the DOX and ICG peaks in DIMP were at 488 nm and 780 nm, respectively, which were corresponding to the free drugs. The fluorescence results showed that the peaks of DOX and ICG in DIMP were at 550 nm and 820 nm, which were also consistent with the peak positions of free drugs. These results collectively indicated that two drugs were successfully loaded on ZIF-8.
Furthermore, the lysosome in cells showed an acidic condition. In order to study the release behavior of DIMP in tumor cells, the in vitro release ability of DOX in DIMP was explored under different pH values (Fig. 1F). The release kinetics of DOX was investigated by measuring the fluorescence at 535 nm. The DOX release rate of the DIMP system was 27.3% at pH = 7.4, which was not obvious. At pH = 6.8, the content of DOX was increased in the buffer, reaching a release rate of 57.7%. When pH = 5.0, the DOX release rate was 90.7%. The acid-responsive DIMP was beneficial to chemotherapy, because imidazole can be easily decomposed when exposed to acid. It indicated that DIMP could efficiently decompose and release chemotherapeutic drugs in the TME, thus improving the efficacy of tumor treatment.
The crystal structure of DIMP was verified by XRD (XRD-700). As shown in Additional file 1: Fig. S2, the characteristic peaks of ZIF-8 were at 7.6°, 10.5°, 13°, and 18.5°, respectively. And DIMP also had the same characteristic peak in the corresponding position, which showed that DIMP had the crystalline form of ZIF-8. The elements in DIMP was analyzed with scanning results of XPS. As shown in Additional file 1: Fig. S2, the photon lines at the binding energy of about 270, 400, 520, and 1020 were attributed to N1s, C1s, O1s, and Zn2p, respectively. Among them, the Zn2p spectrum was shown in Fig. 1G, and the peaks of Zn–O were at 1021.6 and 1044.7 eV. The DOX and ICG (wt%) of DIMP grafted on ZIF-8 surface was evaluated with thermogravimetric analysis (TGA) (Fig. 1H). ZIF-8 NP was thermally stable at temperature as high as 300 °C, while DIMP tended to degrade at around 200 °C with a weight loss of about 16.84%.
Characterization of photothermal properties of DIMP
In order to observe the photothermal effects of the DIMP, the temperature changes were recorded with thermal imager. After irradiation, the temperature changes of DIMP with different concentrations (0, 5, 10, 20 and 40 μg/mL) were recorded every 1.5 min within 6 min. From Fig. 1I, the temperature of DIMP with a concentration of 40 μg/mL can eventually reach about 50 °C.
The irradiation time was extended for 15 min to verify the temperature change curve of DIMP NPs in different concentrations. In Fig. 1J, the temperature of DIMP NPs at various concentrations was gradually increased under the irradiation of 1.5 W/cm2 808 nm laser, in which it can exceed 60 °C at the maximum concentration of 40 µg/mL. The temperature of the PBS group did not change significantly. Further, DIMP remained a stable temperature without light. As the power gradually increased, DIMP NPs show different tendency (Fig. 1K). At 1.5 W/cm2, the temperature was around 60 °C. When it was increased to 2 W/cm2, it only took 8 min, and the temperature was already close to 90 °C. Then the On–Off effect of the DIMP NPs was verified (Fig. 1L), where it could remain unchanged after three consecutive laser On/Off cycles, showing an excellent thermal stability.
Cell up-taking and cell viability
The toxicity of DIMP and free drugs was determined with MTT. As shown in Fig. 2A, the toxicity of DOX was dominant at low concentrations. However, when the concentration was increased to 5 µg/mL, the toxicity of the DIMP group was much higher than that of the free drug group, with a significant difference. The cell viability of the DIMP + L group was less than 10%.
The cytotoxicity was further verified with live cell staining. The cells were photographed with a fluorescence microscope and quantified with Image J (Fig. 2B). The number of cells in the DIMP group was smaller, and the image obtained in DIMP + L group was significantly different from other control groups (Fig. 2C). Above results indicated that the DIMP group exhibited high cytotoxic effects under 808 nm laser.
To further study the ability of the drug to enter cells, the fluorescence intensity of 4T1 cells after treating with DIMP for 2 h and 6 h was observed by CLSM (Fig. 2D). The fluorescence of DOX and ICG was increased significantly with the lasted time (Fig. 2E, F). In order to accurately control the fluorescence intensity, the measurement tool of CLSM was applied for quantification. As shown in Fig. 2G, there was a significant difference in mean fluorescence intensity (MFI) between the treatment of 2 h and 6 h. Flow cytometry was applied in exploring the drug internalization. The drug intake was significantly increased over time, with cell phagocytosis exceeding 50% at 1 h and over 90% at 6 h, indicating that DIMP could be quickly ingested by cells (Fig. 2H).
In vitro photodynamic effects
1, 3-Diphenylisobenzofuran (DPBF) was used as a singlet oxygen capture agent, and the relative consumption of DPBF under light condition was measured with ultraviolet–visible spectrophotometer. In the DPBF solution containing DIMP, under 808 nm laser irradiation, the UV absorption intensity of DPBF continued to decrease with the increase of the irradiation time (Additional file 1: Fig. S3), indicating that the consumption of DPBF increased with the increase of the irradiation time, to manifest the existence of singlet oxygen. The experimental results proved that DIMP had excellent active oxygen generation ability. The ROS production of DIMP was demonstrated with the fluorescent probe SOSG. As shown in Fig. 2I, the green fluorescence of the control groups was negligible without laser irradiation, and the green fluorescence of 4T1 cells incubated with DMIP was not obvious in the dark. However, when the 4T1 cells were irradiated with 808 nm laser (1 W/cm2) for 5 min, green fluorescence was obvious. The average fluorescence intensity of SOSG in the cells was shown in Fig. 2J. All these results indicated that DMIP had a strong ability to produce ROS, which had a promising application in PDT.
Lysosomes and mitochondria co-localization assay
As an important processing plant for cells, lysosome was important organelles for the study of drug internalization pathways. As shown in Additional file 1: Fig. S4, the fluorescence intensity of DOX and ICG at 6 h was significantly stronger than 2 h. The Mander overlap coefficient (R) shows that DOX increased from 0.55 to 0.84, and ICG increased from 0.33 to 0.89. The above results indicate that DIMP has a good co-localization effect.
As a cellular energy factory, mitochondria played vital roles in cell stability. The co-localization of lysosome and mitochondria was observed by CLSM. As shown in Fig. 3A, the Mander coefficient was 0.56 for DOX and 0.74 for ICG at 2 h. At 6 h, the Mander coefficient was significantly increased, which was 0.77 for DOX and 0.91 for ICG. The cellular fluorescence intensity was quantified (Fig. 3B) and the fluorescence intensity of DOX and ICG had a significant difference over time.
Cell damage test
The cell damage was tested with JC-1 stained experiment for determining the change of mitochondrial transmembrane potential induced by DMIP. When the membrane potential of cells was normal, JC-1 entered into the mitochondrial matrix through the mitochondrial membrane and formed a polymer emitting red fluorescence due to the increased concentration. However, for the apoptotic cells, the membrane potential was low. JC-1 was released from the mitochondria, which could not be gathered in the mitochondrial matrix to form a monomer emitting green fluorescence. The results could be qualitative and quantitative by detecting green and red fluorescence. As shown in Fig. 3C, comparing with the control group, the green fluorescence of DIMP + L group was the strongest, and the red fluorescence was the weakest. Meanwhile, the green fluorescence of DIMP group was only weaker than that of DIMP + L group, and the red fluorescence was only stronger than that of DIMP + L Group. As shown in Fig. 3D, the average fluorescence of DIMP + L group was the strongest. These results indicated that our DIMP could decline mitochondrial membrane potential and promote apoptosis.
Previous experiments have proved that DIMP was well located in lysosomes. Due to the decomposition behavior of DIMP triggered in the acidic environment of lysosomes, acridine orange (AO) staining was performed to determine the damage of DIMP to lysosomes. As shown in Additional file 1: Fig. S5, bright red fluorescence was observed in intact lysosomes in untreated control cells, resulted from the accumulation of AO in acidic environment of lysosomes. After the treatment with DIMP + L, the red fluorescence disappeared and the cells shrank, indicating the destroyed lysosomes and apoptotic cancer cells.
The cellular DNA damage of drugs was studied with a step diagram (Fig. 3E). Though the CLSM image (Fig. 3F), it could be clearly seen that the green fluorescence intensity was stronger in the DIMP group and the DIMP + L group, indicating more accumulation of γ-H2AX. The fluorescence intensity was obtained quantitatively in image (Fig. 3G). It confirmed that the fluorescence intensity was strongest in DIMP + L group, with significant difference from other groups. A cell was randomly selected and quantified to obtain the fluorescence curve (Fig. 3H), and it further illustrated that the fluorescence intensity of the DIMP + L group was significantly higher than that of the other groups.
Immunologic cell death effect by DIMP
Conventional tumor therapies such as chemotherapy and phototherapy have been considered as effective tool for boosting immunogenic cell death (ICD) against a broad spectrum of solid tumors. Considering the powerful ability of combination therapy of DIMP, we further gained the results of biochemical correlates including calreticulin (CRT) and high-mobility group box 1 (HMGB1) in the extracellular milieu and in-vivo tumor sections (Fig. 4A). As showed in Fig. 4B–D, we could clearly witness the overexpression of CRT, which released a “eat me” signal and provided basics for the phagocytosis of DC cells. Compared with PBS group, all drug groups showed an obvious CRT expression, and the DIMP + L group exhibited greater effect than that of DIMP group, which was owing to the combined treatment of chemotherapy and PTT/PDT for the former. Meanwhile, the HMGB1 protein could be obviously spread to the extracellular matrix from cell nucleus, which could promote the maturation process of dendritic cells (DCs). Further, the ICD process was also carried out in tumor tissue (Fig. 4E–G), and the similar results to cell level were obtained. After irradiation with laser (808 nm), the DIMP + L group generated forceful ICD effect. All these excellent results collectively validated that DIMP could be served as outstanding ICD-inducer for rousing immune response.
The size and shape of nano-carriers made important effects on the tissue infiltration of anti-tumor drug [41, 42]. In order to simulate the tumor penetration of DIMP, MCSs of 4T1 cells were used as an in vitro model. The MCSs was established (Fig. 5A) and incubated with DIMP (containing 20 μg/mL DOX) for 8 h (Fig. 5B). The DOX with green fluorescence and ICG with red fluorescence were observed and quantified on MCSs (Fig. 5C, D). It could be found that the fluorescence intensity of DOX and ICG was the strongest when the scanning depth reached 40 μm, respectively. And the fluorescence signal intensity distribution of the MCS was analyzed in Additional file 1: Fig. S6. It proved that DIMP could be enriched in tumor site with good permeability.
In vivo bio-distribution analysis
As a clinical NIR dye, ICG can efficiently absorb NIR light and convert it into heat. Under the irradiation of 808 nm laser, the temperature change of the whole body of the mouse was recorded by a thermal imager. As shown in the Fig. 5E, with the increase of laser irradiation time, the temperature of the PBS group did not change significantly, while the temperature of the ICG group was gradually increased. It was obvious that the temperature of the tumor site in the DIMP group was increased significantly. The thermal imager recorded the temperature increase over time. The temperature of tumor was recorded (Fig. 5F), which could be as high as 50 °C within 6 min of irradiation. It proved that ICG could be enriched in the tumor thus making excellent photothermal effects.
As a new type of biomedical imaging method, PA imaging could obtain high-resolution and high-contrast tissue fluorescence signal images. As an effective contrast agent for PA, ICG was conductive to analyze the fluorescent signal distribution of DIMP in tumor models. A mouse tumor model was used to verify the performance of DIMP in vivo. The fluorescence imaging results suggested that the drug was enriched in the tumor site within 24 h after drug injection (Fig. 5G). The DIMP group had obvious PA signal, which was stronger than that of ICG group. Above conclusion was verified with quantitative analysis (Fig. 5H). In general, DIMP could be accumulated in tumor tissues and applied for imaging, which could provide accurate information on tumor micro-structure.
Based on the conclusions of above in vivo experiments, 24 h after drug injection was regarded as the best time point for laser irradiation. The fluorescence distribution at 12 h, 24 h, and 48 h after the injection of the drug (Fig. 5I). For the ICG group, the drug could reach the tumor site at 12 h, and then the fluorescence was decreased at 24 h, and almost no fluorescence could be observed at 48 h. For the DIMP group, the drug began to gather at the tumor site at 12 h, and fluorescence was observed throughout the body at 24 h, mainly at the tumor site. At 48 h, the drug only gathered at the tumor site. At 48 h, the main organs of the mice were dissociated for fluorescence scanning. The fluorescence was almost invisible in ICG group, which may be resulted from the metabolism of drugs. In the DIMP group, a part of the fluorescence could be still observed in the tumor site, and little fluorescence was found in other organs. This may be because that the drug was cleared through the blood circulation. The fluorescence quantification results of both tumors and major organs by IVIS imaging software also verified above conclusions (Fig. 5J, K). CLSM of the dissociated tumor tissue slices could also show that the DIMP group had higher DOX and ICG fluorescence signals (Additional file 1: Fig. S7). It suggested that DIMP had excellent permeability and enhanced permeability and retention (EPR) effects, which could be efficiently gathered at the tumor site, and be not easily cleared by the blood circulation. The above results brought confidence for the further in vivo experiments.
In vivo anti-cancer activity
Encouraged by above results, the 12-day in vivo treatment experiment was quickly launched (Fig. 6A). Three times of tail vein injection were required, and light was given 24 h after each injection. The DIMP group performed well in tumor treatment (Fig. 6B), and the tumor size in the DIMP + L group showed a steady trend of decline. It showed the excellent anti-tumor effects of DIMP.
During the treatment, the body weight of the mice was recorded daily (Fig. 6C), and the body weight of the mice in each group maintained stable without significant changes. The dissociated mice tumors were weighed (Fig. 6D) and photographed (Fig. 6E). The weight of the tumor in the DIMP + L group was the smallest with a significant difference from other groups, which further illustrated the excellent anti-tumor effects in DIMP + L group.
The main organs were sliced and stained with H&E, Ki67 and TUNEL. The H&E results (Fig. 6F and Additional file 1: Fig. S9) showed no obvious lesion. As shown in Fig. 6G, after calculation, the number of proliferation factors in the DIMP + L group was significantly different from that of other groups. In Fig. 6H, TUNEL results were quantified using image J software, indicating more apoptotic area in DIMP + L group. The tumor tissues of the treatment group were sliced to observe the drug fluorescence, and the DIMP + L group also showed stronger drug fluorescence than free drugs. It showed that DIMP had excellent anti-tumor effects in vivo.
In order to evaluate the in vivo biosafety of DIMP, healthy Balb/c mice were injected with Control (PBS), DOX, ICG and DIMP via tail vein. After 7 days, the heart, liver, spleen, lung and kidney of mice were taken and sliced for H&E staining. As showed in Fig. 7A, the DOX group showed a cardiotoxicity  and the DIMP could effectively ameliorate this phenomenon. Furthermore, as shown in Fig. 7B, the weight of the mice did not change significantly within 7 days. Further, whole blood was collected for hematological analysis (Fig. 7C–K), involving several parameters including WBC, RBC, MCHC, RDW, LYM, HCT, HGB, MPV and PTL etc. Comparing with the control group, there was no significant difference in all these parameters. The DIMP NPs (at concentration of 1, 5, 10, 25, 50 μg/mL) were mixed with the blood and the leaked hemoglobin was detected at 570 nm by UV–Vis absorption spectra (Additional file 1: Fig. S10). Both the DIMP NPs and PBS control showed a leakage of less than 5% (Fig. 7I), suggesting a good biological safety.
In summary, we have successfully developed an intracellular acidity-responsive MOF nanoreactor of DIMP for cancer therapy. The well-designed DIMP presented high drug loading capacity, controlled drug release, low side effects and good tumor suppression effect. After laser irradiation, ICG in DIMP could not only produce ROS, but also elevated temperature at the tumor site, exhibiting improved efficacy than that of free ICG. Further, DIMP was armed with a multi-mode FL/IR/PA imaging and a combined PTT/PDT/chemotherapy for effectively inhibiting the growth of 4T1 tumors in vivo. These results proved the potential therapeutic value of DIMP in cancer theranostics, and may provide a new strategy for the establishment of an integrated nanoplatform for tumor diagnosis and treatment.