Curcumin encapsulated zeolitic imidazolate frameworks as stimuli responsive drug delivery system and their interaction with biomimetic environment

Metal organic frameworks (MOFs) exhibit unique features of finely tunable pore structures, excellent chemical stability and flexible surface structural functionality, making them advantageous for a wide range of applications including energy storage, compound separation, catalysis, and drug delivery. The present work enlightens a novel approach of single step fabrication of CCM-ZIF-8 as a drug carrier and its application as stimuli responsive drug delivery systems via external stimuli involving change in pH and in presence of biomimetic cell membrane like environment using liposomes and SDS micelles. The methodology is devoid of any post synthesis drug loading steps. The synthesized curcumin encapsulated ZIF-8 frameworks demonstrate ultrahigh drug encapsulation efficiency (ca. 83.33%) and good chemical stability. In vitro drug release of curcumin was three times higher in acidic medium than in physiological pH. Cytotoxicity results demonstrated enhanced therapeutic effect of CCM-ZIF-8 than free curcumin. Confocal microscopy results confirmed the easy cellular internalization of CCM-ZIF-8 in HeLa cells. Intracellular distribution studies at various incubation times confirmed the clathrin-mediated endocytosis to lysosomal pathway of CCM-ZIF-8, but without mitochondria being an intracellular fate. The results signify that CCM-ZIF-8 is an efficient drug carrier for passive tumor therapy in future for cancer treatments.

Nanomedicine is expressively stepping forward in improving the efficiency of therapeutic drugs. The conventional approach of the drug delivery systems (DDS) was to encapsulate therapeutic drugs, protect them from degradation and control the drug release kinetics 1 . However, with the progress in understanding of human physiology and disease pathology, robust DDS are being developed that are capable of effective delivery of drugs at the site of disease while minimizing drug doses and reducing side effects 2 . Accordingly, a lot of research is now being focused towards achieving the overall control in drug bio-distribution, cell targeting, in vivo drug stability, controlling the drug release kinetics and understanding the drug release mechanism 3 . A number of drug carriers such as liposomes 4 , polymer nanoparticles 5 , micelles 6 , dendrimers 7 , and inorganic NPs 8 specially iron oxide 9 , quantum dots 10 and metal organic frameworks 11 with diverse sizes, architectures and surface properties have been constructed and studied extensively. Nevertheless, all these carriers should have the preferred particle size of few hundreds of nanometers to allow systemic administration and ensure diffusion within the cell 12 .
Cancer is one of the world's leading causes of death and its diversity and complexity involving variety of cell types, physiological distinctiveness, and extracellular matrices limits the therapeutics efficiency 13 . Cancer tissues are marked with more acidic environment with respect to body physiological tissues 14 . Although, various anti-cancer therapeutic agents are strategically designed, these molecules are either too large or highly charged, making the metabolically unstable, and/or too insoluble to target cancer cells 15 . Moreover, most of these chemotherapeutics are highly toxic and their systemic administration leads to serious side effects leading to significant destruction of normal cells along with cancerous cells. Hence, an efficient DDS must be designed to achieve

Results and Discussion
Synthesis and Characterization of CCM-ZIF-8. CCM-ZIF-8 were synthesized via a single step procedure as reported earlier 31 . The synthesis parameters were optimized as shown in Table S1 (supplementary information). Maximum encapsulation efficiency was achieved within 15 min of the reaction onset. Subsequent increase in reaction time did not show a significant increase in encapsulation efficiency indicating the occurrence of saturation beyond 15 min. Hence, we considered 15 min as optimum reaction duration for further studies. The hydrodynamic diameter of ZIF-8 and CCM-ZIF-8 was found to be 140 ± 5 and 145 ± 5 nm respectively and the corresponding particle size distribution is presented in Fig. S1 (supplementary information). Zeta potential of the ZIF-8 and CCM-ZIF-8 was +7.40 and +7.22 mV respectively. No significant change in surface charge of ZIF-8 on encapsulation of curcumin infers minimum surface adsorption of curcumin on ZIF-8. TEM images demonstrated the formation of rhombic dodecahedron particles having an average particle size of 80 ± 5 nm (Fig. 1a). AFM image in Fig. 1b and S9 (supplementary information) displayed smooth surface morphology of CCM-ZIF-8. The CCM-ZIF-8 showed no framework disintegration even upto 30 days indicating high stability of the synthesized frameworks as shown in Fig. S2 (supplementary information). X-ray diffraction patterns of pure curcumin, ZIF-8 and CCM-ZIF-8 are shown in Fig. 1c. The XRD pattern of ZIF-8 matched well with reported literature 32 .The characteristic peaks of ZIF-8 at 2θ values of 7.29, 10.44, 12.83 and 18.10 correspond to (110), (200), (211) and (222) index planes respectively. There was no change in the XRD pattern of CCM-ZIF-8 as compared to ZIF-8 indicating the framework stability and minimum impact on its crystallinity. Thermal characteristics of curcumin and ZIF-8 (Fig. 1d) matched well with the reported literature 33 . Curcumin starts to decompose at 200 °C and gets completely decomposed between 200 and ~550 °C whereas ZIF-8 is stable up to 500 °C. Encapsulation of curcumin in the ZIF-8 is clearly evident from the thermal curve of CCM-ZIF-8. An early mass loss ~300 °C can be attributed to the decomposition of curcumin that is encapsulated in ZIF-8 frameworks.
The encapsulation of curcumin into ZIF-8 frameworks was indicated by the change in yellow color of curcumin to orange (see Fig. 2a inset). UV-Vis and PL spectroscopic studies were also performed to confirm the curcumin encapsulation into ZIF-8. UV-Vis absorption spectra of pure ZIF-8, curcumin and CCM-ZIF-8 are shown in Fig. 2a. The characteristic absorption peaks of curcumin and ZIF-8 appeared at 425 and 213 nm respectively. The absorption peak at 425 nm in curcumin spectra is assigned to π-π* transitions which corresponds to enol form of curcumin. CCM-ZIF-8 spectra showed a red shift in curcumin peak to 485 nm which demonstrated the encapsulation of curcumin in ZIF-8 framework 34 . UV spectroscopy results correlated with the photoluminescence (PL) spectroscopy studies. The emission peak of pure curcumin was observed at 537 nm and it red shifts towards 610 nm in CCM-ZIF-8 as shown in Fig. 2b. FTIR spectra of pure curcumin, ZIF-8 and CCM-ZIF-8 are shown in Fig. 2c. The characteristic peaks of ZIF-8 were observed at 3135, 2928, 1606 and 1580 cm −1 corresponding to aromatic C-H stretching, aliphatic C-H stretching, C-C stretching and C-N stretching of imidazole respectively 35 . In curcumin spectra, the band at 3490 cm −1 corresponds to vibrations of free hydroxyl group of phenols. Other observed peaks at 1511, 1279 and 1152 cm −1 are attributed to C = C stretching of benzene ring, aromatic C-O stretching, and C-O-C stretching respectively 36 . In CCM-ZIF-8 spectrum, peaks at 3426, 3055, 2920, 2359, 1605, 1184, 1033 and 894 cm −1 correspond to OH stretching of phenol group, aromatic C-H stretching, aliphatic C-H stretching, C-C stretching, C-N stretching, C-O stretching, C-O-C stretching and C-C stretching respectively. However, stretching peak of phenolic group of curcumin was blue shifted from 3490 to 3426 cm −1 in CCM-ZIF-8 and confirm the encapsulation of curcumin in ZIF-8 34 . BET adsorption studies were performed to determine the surface area and porosity of ZIF-8 and CCM-ZIF-8. The BET isotherms as shown in Fig. 2d and e underline linear adsorption isotherm at low relative pressures. Significant increase in nitrogen uptake during adsorption experiments at low relative pressures confirmed microporosity of ZIF-8 and CCM-ZIF-8. Additionally, the decrease in BET surface area and pore volume of CCM-ZIF-8 (ca. 1922.6 m 2 g −1 /0.762 cm 3 g −1 ) as compared to ZIF-8 (ca. 2229.3 m 2 g −1 /0.827cm 3 g −1 ) also confirmed the effective encapsulation of curcumin in ZIF-8 frameworks. Raman spectroscopy as depicted in Fig. S3 (supplementary information) showed intense bands of ZIF-8 corresponding to methyl group and imidazole ring vibrations. The vibrational bands spotted at 168, 686, 1146 and 1458 cm −1 correspond to Zn−N stretching, imidazole ring (C−N) stretching and methyl bending respectively 37 . Raman spectrum of pure curcumin confirmed the enol form due to the absence of (C = O) peaks in the1650-1800 cm −1 region. This observation matched well with 1 H NMR spectral studies of curcumin The 1 H NMR spectra of ZIF-8 showed two peaks at δ 7.37 and δ 2.61 ppm corresponding to aromatic C-H protons and CH 3 protons of imidazole respectively. Both these peaks were present in 1 HNMR spectrum of CCM-ZIF-8 at δ 7.34 and δ 2.60 ppm respectively, albeit with slight chemical shift. 1 H NMR spectra of pure curcumin displayed peaks at δ 5.95 and 5.93 ppm (in CD 3 OD and CD 3 OD + CF 3 COOD solutions respectively, Fig. S4, supplementary information) which correspond to methine protons (H c , scheme 1, supplementary information) and confirmed the enol form of curcumin. However, the same peak was absent in CCM-ZIF-8 spectra. It is due to the interaction between Zn and di-keto form of curcumin. This also affects the trans alkene protons (H a and H b , scheme 1, supplementary information) adjacent to di-keto group which deshields and shows downfield in spectra with slight change in chemical shift. The detailed NMR spectra and respective chemical shift values of CCM and CCM-ZIF-8 are presented in Fig. S4 and Table S2 (supplementary information) respectively. 1 H NMR spectra of curcumin and ZIF-8 were similarly reported in literature 39,40 . The presence of 1 H NMR peaks of curcumin in CCM-ZIF-8 spectra successfully implied that curcumin was present in CCM-ZIF-8. From 1 H NMR spectra, the relative intensities of methoxy peak in pure curcumin and CCM-ZIF-8 revealed that almost 3.2% of curcumin was present in the carrier system. This was in conformity with the drug loading capacity (3.42%) evaluated by UV-Vis spectra.
Curcumin changes its absorbance when bonded with metals and such change in absorbance can be used to detect curcumin in host molecules 16 . The interaction between curcumin and ZIF-8 was studied using fluorescence spectroscopy. Solutions of different concentrations of curcumin (1-10 µM) were added to the ZIF-8 and decrease in the emission peak at 310 nm was observed. Fluorescence quenching was calculated from Stern-Volmer equation (1)  information). The drug release studies were carried out for 72 hrs. Gradual increase in the absorbance at 425 nm with time was observed as shown in Fig. S7 (supplementary information). The cumulative drug release of 88% under acidic medium and 28% in physiological medium was achieved in 72 hrs (Fig. 2f). This shows three-fold increase in drug release under acidic medium as compared to physiological pH. This is due to protonation of imidazolate ions under acidic condition whereby, the co-ordination linkage between Zn and imidazolate ions breaks and leads to the effective increase in curcumin release.
pH stimuli responsive curcumin release from CCM-ZIF-8 was further probed under biomimetic cell membrane like environment of liposomes and SDS micelles 43 . The pK a value of curcumin lies in between 7.9-10.5, and this indicates that curcumin is almost neutral at physiological conditions. SDS solutions of different concentrations, viz. above critical micelle concentration (CMC) (30 mM), at CMC (8 mM) and below CMC (3 mM) were prepared to observe curcumin release form CCM-ZIF-8 through fluorescence spectroscopy. The decrease in emission peaks was noticed with subsequent time. The percentage fluorescence quenching of emission peak intensity indicated the drug release and was calculated by using equation ( The percentage fluorescence quenching was observed for 3 mM (ca. 72.9%), 8 mM (ca. 70.9%) and 30 mM (ca. 60.6%) shown in Fig. 3. Additionally, emission peak of CCM-ZIF-8 at 580 nm showed blue shift with subsequent time in 3 and 8 mM SDS solutions but no such shift was observed in 30 mM. The fluorescence quenching behavior shows more hydrophobic interaction of CCM-ZIF-8 with SDS micelles above CMC (i.e. 30 mM). The bioadhesion of CCM-ZIF-8 with cell membrane like environment offers a strong stand towards its bioavailability. Thus, two commonly used lipids such as 1, 2-dimyristoyl-sn-glycero-3-phosphatidylcholine (DMPC) and 1, 2-dimyristoyl-sn-glycero-3-phospglycerol (DMPG) and both together DMPC: DMPG (9:1) were used for further studies. Figure 4 shows strong emission band of CCM-ZIF-8 in all the three liposomes. The emission peak shifts towards lower wavelength at 560, 580 and 550 nm in DMPC, DMPG and DMPC: DMPG (9:1) respectively from 625 nm. The maximum flouroscence quenching of CCM-ZIF-8 was observed in DMPC (ca. 90.76%) followed by DMPC:DMPG (9:1) (ca. 88.81%) and DMPG (ca. 74.4%). This specifies an electrostatic interaction between CCM-ZIF-8 and liposomes; as polarity of the medium influences the blue shift. The shift could be related to curcumin release and explains a strong synergy between CCM-ZIF-8 and liposomes. This clearly indicates that CCM-ZIF-8 is suitable for cellular internalisation under invitro cell uptake studies and could pave the way for therapeutic treatments. Cytotoxicity and cellular internalization studies. The observed photoluminescence spectra of CCM-ZIF-8 suggest that optical imaging and cellular internalization can be successfully attained through confocal laser scanning microscopy (CLSM) without any imaging probe. We evaluated cellular internalization of CCM-ZIF-8 in HeLa cell lines at three different incubation times viz. 0.5, 2 and 12 hrs. CLSM imaging in Fig. 5 clearly shows that green fluorescence in HeLa cells incubated with CCM-ZIF-8, implying cellular internalization. This highlights ZIF-8 as an efficient carrier protecting curcumin against degradation and enhancing the bioavailability by effective cellular internalization.
Furthermore, we have examined the intracellular distribution of CCM-ZIF-8 at three different incubation times viz. 0.5, 2 and 12 hrs using two different cell organelle tracker dyes LysoTracker Red (for Lysosomes) and MitoTracker Red (for Mitochondria). Figure 6 shows the time dependent colocalization of LysoTracker Red and CCM-ZIF-8, wherein the green fluorescence of CCM-ZIF-8 is majorly overlapping with red fluorescence of LysoTracker Red after 2 hrs of incubation (Pearson's coefficient = 0.87). With minimal Pearson's coefficient values (Fig. 7), it is clear that CCM-ZIF-8 is not localized into mitochondria of HeLa cells, even after 12hrs of incubation. These results confirmed the clathrin-mediated endocytosis to lysosomal pathway of CCM-ZIF-8, with maximum colocalization with lysosome at 2 hrs, but without mitochondria being an intracellular fate. Our results are in accordance to the previous studies on the lysosomal fate of curcumin 44 .
The cytotoxicity study using MTT assays was performed to evaluate the viability of free curcumin, ZIF-8 and CCM-ZIF-8 against HeLa cell lines. The cell viability and toxicity experiments were performed with 0.4% DMSO and 0.025% Tween 80 as positive controls. Prior to experiments, curcumin was dispersed in 0.4% DMSO whereas ZIF-8 and CCM-ZIF-8 were dispersed in PBS solutions containing Tween 80 (0.025%). The experiments were performed at different concentrations of curcumin, ZIF-8 and CCM-ZIF-8 in the range of 15.6 to 125 μg/ mL for 24, 48 and 72 hrs (Fig. 8). It is clearly evident from the data shown in Fig. 8 that ZIF-8 has no obvious cytotoxicity upto 72 hrs when its concentration was below 50 μg/mL. However, a slight increase in cytotoxicity was observed at higher concentrations of pure ZIF-8 in all time points. On the other hand, CCM-ZIF-8 exhibited a strong cell growth inhibition effect on HeLa cells compared to free curcumin. High cytotoxicity of CCM-ZIF-8 than free curcumin was evident in all the concentrations at various time points. Only 62.5 μg/mL of CCM-ZIF-8 induced almost 57% of cell death compared to a mere 25% cell death caused by free curcumin (effective concentration = 2.39 μg/ml). The cytotoxicity was more pronounced at higher concentration of CCM-ZIF-8 as 70% of cell death was observed at 125 μg/mL. This must be due to the strong cellular internalization of CCM-ZIF-8 by HeLa cells. Additionally, the results also show a significant improvement in the bioavailability after the encapsulation of curcumin in ZIF-8. Moreover, using ZIF-8 as a drug carrier not only enhances cellular internalization but also ensures safe delivery of curcumin resulting in better cytotoxicity and bioavailability. This could be attributed to the small particle size and positive zeta potential of the present CCM-ZIF-8 frameworks. The small particle size favors the passive tumor targeting approach by enhanced permeability and retention (EPR) effect 45 . Additionally, positive zeta potential could lead to enhanced electrostatic interaction with the cell membranes, thereby improving cellular internalization of drug carriers.

Conclusion
We demonstrated successful encapsulation of a hydrophobic drug curcumin inside ZIF-8 frameworks via a single step synthesis process avoiding any kind of post synthesis drug loading. The small size of CCM-ZIF-8 was found to be optimal for cellular uptake through EPR effect in cancer cells. A high drug encapsulation efficiency of 83.3% was achieved. The characteristic pH stimuli responsive behavior of ZIF-8 can be used to trigger drug release in external stimuli change in pH and biomimetic cell membrane like environment of micelles and liposomes. In vitro drug release results showed almost 88% drug release over a period of 72 hrs in acidic conditions. Cytotoxicity and   Characterization of CCM-ZIF-8. The crystalline phase of the prepared frameworks was analyzed using powder x-ray diffraction (Rigaku smart lab) with CuKα (λ = 1.5418 Å) radiation. The samples were scanned between 5-40° 2θ range. The morphology of the samples was characterized by transmission electron microscopy (TEM) using FEI Tecnai TEM equipped with a LaB 6 source operating at 200 kV. The AFM imaging was done using Dimension ICON (Bruker) AFM in tapping mode at scanning rate of 0.9 Hz. UV-Vis absorption spectra were measured with Shimadzu U-2450 UV-Vis spectrophotometer in the wavelength range of 200-800 nm. Fluorescence spectra were recorded by Cary fluorescence spectrophotometer (Agilent Technologies) in the wavelength range of 200 to 800 nm. FTIR spectra were recorded with Agilent Technologies Cary 6000 series FTIR spectrometer at wavenumber from 400 to 4000 cm −1 . The particle size distribution and zeta potential was measured by dynamic light scattering (Malvern Zetasizer). The Brunuer-Emmett-Teller (BET) surface area measurements were performed on Quanta chrome, automatic volumetric instrument with pressure ranging from 0 to 760 Torr with low-pressure volumetric nitrogen adsorption measurements. Thermal properties were measured by Perkin Elmer Pyris Thermogravimetric (TGA) analyzer under nitrogen atmosphere from room temperature to 700 °C with the heating rate of 5 °Cmin −1 . Raman studies were performed using a Horiba LABRAM high resolution UV-Vis-NIR Raman spectrophotometer equipped with a confocal microscope. The samples were analyzed using a 534 nm He-Ne laser employing a 10X objective lens to achieve an average spot size of 3 µm diameter. 1 H NMR spectra were collected on a Bruker Advance III 500 MHz NMR spectrometer in CD 3 OD and CD 3 OD and CF 3 COOD solvents. The curcumin content was calculated by using the following formula, Preparation of Liposomes. The liposomes and SDS micelles were prepared using a method reported elsewhere 43 . The aqueous PBS solution was taken in round bottom flask, and heated above the phase transition temperature of the used lipids. On the other hand, the required amount of lipids was dissolved in the ethanol solution (less than 1% v/v) and stirred till lipids were completely dissolved. Subsequently, the lipid ethanol solution was immediately injected into the prepared PBS solution above the phase transition temperature of lipids under continuous stirring.
Cell culture preparation. HeLa cell lines were procured from National Centre for Cell Sciences (NCCS), Pune, India and were cultured in DMEM (GibcoTM) supplemented with 10% of fetal bovine serum (FBS) and 1% penicillin-streptomycin (GibcoTM) in a humidified CO 2 (5%) incubator at 37 °C temperature and passaged using Trypsin (GibcoTM) when the cells become confluent.
Cytotoxicity Assay. A total of 150 µl of cells suspended in 1% FBS were seeded in 96 well plate at a density of 3000 cells/well and allowed to adhere for 24 hrs. The concentration of free CCM and pure ZIF-8 that were used was in relative to the loading percentage of CCM in CCM-ZIF-8. DMEM with 0.025% Tween-80 (for pure ZIF-8 and CCM-ZIF-8) and 0.4% Dimethyl sulfoxide (DMSO) (for free CCM) were used to disperse the samples. Treated cells were incubated for 24, 48 and 72 hrs at 37 °C. 0.025% Tween-80 and 0.4% DMSO were used as vehicle control along with positive (only cells) and negative (only media) controls. 20 µl of 3-(4, 5-dimethylthiazol-2-yl) -2, 5-diphenyl tetrazolium bromide (MTT) was added to each well and further incubated for 3hrs followed by the addition of 100 µl of DMSO to each well. Absorbance was read at 570 nm with 650 nm as reference.
Cellular uptake study. 300 µl of HeLa cell suspension was seeded in a 6 well plate on a sterile cover slip