SYNTHESIS AND OPTIMIZATION OF GEMCITABINE-LOADED MIL-101NH2 (Fe) NANOCARRIERS: RESPONSE SURFACE METHODOLOGY APPROACH

Objective: The objective of the present study is to synthesize and optimize gemcitabine (GEM)-loaded MIL-101NH2 (Fe) nanocarriers. The design of experiments is used to optimize the formulation for higher encapsulation efficiency (EE) for effective drug delivery. Materials and Methods: MIL-101NH2 (Fe) was synthesized by microwave-assisted hydrothermal method. Central composite design (CCD) under response surface methodology was used for the optimization of GEM encapsulation into the MIL-101NH2 (Fe). The most influential variable that affects the EE was investigated by Perturbation plot. Validation of the design was carried out by performing the experiments under optimal conditions. Further optimized formulation was physicochemically characterized for particle size, surface charge, and surface morphology using zetasizer and scanning electron microscopy (SEM), respectively. Structural integrity of the optimized formulation was carried out by Powder X ray crystallography (PXRD). Fourier-transform infrared (FTIR) spectroscopy was used for the confirmation of GEM loading. Accelerated storage stability analysis was also performed to find out the related parameters. Results: Here in this work, crystalline MIL-101NH2 (Fe) has been successfully synthesized by microwave radiation method. The optimization result revealed that process variables such as GEM concentration, pH, and time significantly affect the desired constraint, EE. Perturbation plot evidenced that among all the variables, pH is the most significant factor followed by drug concentration and time. The optimized formulation exhibited 76.4 ± 7% EE and average particle size of 252.9 ± 9.23 nm. PXRD and SEM results demonstrated that the optimized formulation was crystalline in nature. FTIR spectroscopy confirms the presence of drug inside the MIL-101NH2 (Fe). In vitro release profile revealed that MIL-101NH2 (Fe)-GEM exhibited the sustained release up to 72 h in comparison to the native GEM. Storage-stability studies also indicate that MIL-101NH2 (Fe)-GEM has a shelf life of 6 months. Conclusion: The EE of GEM in MIL-101NH2 (Fe) can be altered by varying the drug concentration and pH during the impregnation.


INTRODUCTION
Drug gemcitabine (GEM) is one among the Food and Drug Administrative authority-approved drugs, which is mainly used for the first-line therapy for advanced and metastatic pancreatic cancer [1][2][3]. It is a difluoro analog of deoxycytidine which is transported to cells by nucleoside transporters, where it is phosphorylated to difluorodeoxycytidine diphosphate (dFdCDP, ribonucleotide reductase inhibitor) and triphosphate (dFdCTP, compete with cytidine phosphate). It is found that the deficiency of these transporters causes resistance to GEM therapy [4][5][6]. GEM is used for the treatment of various cancers such as nonsmall cell lung, bladder, pancreatic, breast, colon, cervical, ovarian, and hepatocellular cancer [7]. Unfortunately, GEM has short plasma half-life of 8-17 min in human and 9 min in murine [8][9][10] because of rapidly and extensively deamination by cytidine deaminase in the blood, liver, kidney, and other tissues [11,12] to the inactive metabolite difluorodeoxyuridine (dfdU) which is excreted in the urine. However, the short half-life and low permeability are the major setback to the current clinical treatment with the drug. To achieve the therapeutic level of drug, frequent intravenous infusion is required at high dose which causes several side effects [12,13]. Thus, various attempts have been made to deliver GEM with the aim to improve pharmacokinetics and tumor delivery of this compound for more effective chemotherapy. GEM has been studied in different polymeric systems such as magnetic poly ε-caprolactone nanoparticles [14], micelles [15], and gold nanoparticles for pancreatic cancer treatment [16]. Several liposomal formulations have also been prepared and evaluated, e.g., pH-sensitive stearoyl-PEG-poly(methacryloyl sulfadimethoxine)-coated liposomes [17] and hyaluronic acid-coated liposomes for pancreatic adenocarcinoma cells [18]. Among these approaches, liposomes were the most effective carrier for delivery, but there are certain limitations faced by liposomes such as poor stability during storage [19]. Being a hydrophilic molecule, GEM is located in the aqueous compartment of liposomes and leads to diffusion of the drug during storage [19,20]. Therefore, encapsulation and drug release profile may change during the storage of formulation. 5-fluorouracil, ibuprofen, topotecan, and busulfan) [26]. MOFs can be synthesized in nanoregimen by well-established techniques and methods in nanotechnology [27]. Materials of Institut Lavoisier (MIL) family is the first group of MOF discovered by Férey et al. for the delivery of ibuprofen with chromium-based MIL-101 [28]. MIL family is engineered by trivalent metal centers connected with carboxylic acid. It was found that drugs with polar complexing groups bind eventually to the coordinative unsaturated iron Lewis acid sites (CUS), leading to high encapsulation efficiencies (EEs), high payloads, and controlled release [29,30]. MIL-101NH2 (Fe) is the most stable iron (Fe)-based MOF , already been used for bioactive [31] and magnetic compounds [32]. MIL-101NH 2 (Fe) is a rigid zeotype Mobil thirty nine (MTN) crystal structure that possesses two types of windows, i.e., large hexagonal windows with a pore diameter of 34 A˚ and small pentagonal window with a pore diameter of 29 A˚ [21].
Here, this is the first report on the optimization of drug-loading methods by central composite design (CCD) along with response surface methodology (RSM), an ideal tool for process optimization. MIL-101NH 2 (Fe) was synthesized by microwave-assisted method and characterized by different instrumental techniques, such as scanning electron microscopy (SEM), Powder X-ray diffraction (PXRD), and Fourier-transform infrared (FTIR) spectroscopy. Further encapsulation of GEM was carried out by impregnation method. The effect of GEM concentration, pH, and impregnation time on the EE was assessed CCD along with RSM [33,34]. The optimal condition for the achievement of higher EE was validated by performing the same experiments. The optimized formulation was further characterized for confirmation of drug loading and stability analysis.

Synthesis of MIL-101NH 2 (Fe)
MIL-101NH 2 (Fe) was synthesized by microwave-assisted hydrothermal method with minor modification in the previous method [29,35,36]. Briefly, a mixture of iron chloride (9 mmol) and amino-terephthalic acid (5 mmol) was dissolved in 400 ml of deionized water for 6 min at 170°C at 400 Watt (Mars-5, CEM, US). The synthesized MOF was collected by centrifugation for 10 min at 10,000 g. Product was washed with 50 ml of absolute ethanol to remove the residual ligand and collected by centrifugation. Encapsulation of GEM was performed by impregnation of the aqueous solution of drug in MOF. For encapsulation, 25 mg of lyophilized MIL-101NH 2 (Fe) was suspended in 5 ml of freshly prepared drug solutions (1 mmol) at 80% of the maximum solubility in water. The suspensions were stirred for a maximum of 24 h at room temperature, and 100 µl of the sample was collected at different time intervals at room temperature. These samples were further centrifuged and lyophilized [37].  (Table 1) for the synthesis of MIL-101NH 2 (Fe)-GEM. The variables were selected on the basis of preliminary experiments. Table 2 displays the range of selected variables.
The mathematical relationship between independent variables and their response can be modeled by polynomial model Equation 1: Where Y is the measured response associated with each factor level combinations; β 0 is the Intercept; βi (for i = 1, 2, and 3) are the linear effects, the βii are the quadratic effects, the βij's (for i, j = 1, 2, and 3, i < j) are the interaction between the i th and j th variables. Xi and Xj are the coded value for the processing variables.

Characterization
Particle size, polydispersity index, and zeta potential Zetasizer (Malvern Nano-ZS, Zetasizer Nano series, UK) was used for the determination of particle size. All the measurements were carried out at 90° angle. Each sample was diluted with Milli Q water. For zeta potential measurement, a 150 mV electric field was applied and electrophoretic velocity of samples was measured. All the experiments were performed in triplicate (n = 3).

Field emission scanning electron microscopy
The structure and morphology of the particles were analyzed by field emission SEM (FESEM; 4300S, Hitachi). For SEM analysis, dried samples were mounted on metal stubs with the help of double-sided carbon tape. The samples were sputter coated with gold under vacuum and then examined.

Fourier-transform infrared spectroscopy
FTIR spectra of samples were obtained using an FTIR spectrometer (Nicolet Continuum XL, Thermo Fisher Scientific). Briefly, the sample and potassium bromide were mixed well with a ratio of 1:10, followed by being compressed into a disk. Scans were carried out in wave number 400-4000 cm −1 at a resolution of 4 cm −1 . Infrared spectroscopy in ATR mode (Alpha Bruker) using Opus software was performed to confirm the presence of the drug in the formulation.

Encapsulation efficiency
For the quantification of GEM loaded in MIL-101NH 2 (Fe), 10 mg of dried samples was treated with ultrasonic waves for 60 min using ethanol as extracting medium. The samples were centrifuged at 10,000 rpm for 10 min, and the supernatant was used to analyze nonadsorbed drug. The concentration of adsorbed and nonadsorbed drug was determined by ultraviolet (UV)-visible spectrophotometer at 268 nm based on the standard calibration curve of GEM in the range of 2-100 µg/ml. Equations 2 and 3 were used to calculate encapsulated drug concentration and EE, respectively.

In vitro drug release
For in vitro release, dialysis bag containing 2.5 mg of MIL-101NH 2 (Fe)-GEM was suspended in each of the 10 ml of phosphate buffer solution (PBS, pH = 7.4, 9.5 mM, Lonza). These suspensions were kept under rotary agitation up to 72 h. 0.5 ml sample was taken from the release medium and replaced by fresh buffer to maintain the sink condition and centrifuged to obtain supernatant. Supernatant was used to study drug release in different media by high-performance liquid chromatography (HPLC) [22]. HPLC analysis was performed using a water pump (600E) connected to a C18 column (4.6 mm × 250 mm) (Agilent technology 1220 Infinity LC, Germany) coupled with a UV detector (λ = 268 nm) and EZ Chrome Elite Software. The mobile phase consisted of a mixture of acetonitrile:H 2 O (10:90), NH 4 H 2 PO 4 (50 mM), and TEAA (5 mM). 50 µL of the sample was injected at a flow rate of 1 ml/min. Semiquantitative analysis was performed using standard calibration of different compounds in the range of 2-100 µg/ml [22].

Storage stability analysis
Stability analysis was carried out at 25 ± 2°C/60 ± 5% relative humidity (RH). Freshly prepared freeze-dried samples of each formulation were sealed in vials and placed in a stability chamber at 25 ± 2° C/60 ± 5% RH for 6 months. The samples were analyzed for color, aggregation, crystallinity, and particle size and entrapment efficiency with a sampling frequency of 1 month for 6 months. Experiments were performed in triplicate.

Statistical analysis
Design-Expert software was used for statistical analysis (Version 10.0.8.0), where analysis of relationship between the response variable "Y" and the entire set of "X" variables at 95% level of significance variance was significant, when p<0.05. An F-test was used to determine the overall regression. Selection of best fitting model was based on comparative study of different statistical parameters: coefficient of variation, correlation coefficients (R 2 ), adjusted correlation (adjusted R 2 ) coefficient.

Optimized synthesis of MIL-101NH 2 (Fe)-GEM by CCD
CCD methodology is the best method for optimization of the formulation using minimum number of experiments. Table 1 (4) where A, B, and C are GEM concentration, pH, and time, respectively. Synergistic and antagonistic effects were presented by + and -symbol, respectively [38]. Analysis of variance (ANOVA) was used to find out the effect of processing variables, interaction between the variables, and statistical significance. F value, P value, and sum of squares (SS) are the important parameters for the interpretation of ANOVA table. Higher F and SS value the imply significance of model and vice versa. p value parameter is contrary to F value parameter. Small probability p<0.05 indicates the significance of model and used to predict the response function precisely. Small probability value (p<0.0001) revealed that the selected quadratic model was highly significant and could be used for accurate prediction of responses as shown in Table 3. Goodness of fit was supported by large value of R 2 and adjusted R 2 . Table 3 reveals that A, B, A 2 , and B 2 were the significant factors (p<0.05) and C and C 2 were non-significant variables (p>0.05). The above statistical results revealed that CCD was adequate to optimize the GEM encapsulation within the range of determined variables. The

Effect of process parameters and their interaction
After analyzing the most significant variables which affect the main constraint (EE), RSM methodology was used for optimization. Fig. 1 presents the three-dimensional (3D) response plots of EE versus significant variables.
Design-Expert software was used to plot the response in 3D form to show the interaction of three independent variables and dependent variable. Fig. 1 shows that pH and drug concentration significantly affect the EE. Equation 4 revealed that GEM showed positive effect on EE while pH showed its negative effect also supported by Fig. 1. At lower pH, EE was less; however, when pH was increased up to 4.5, then, there was a significant increase in EE; however, at higher pH, i.e., 6, the EE was reduced (Fig. 1a). On the other hand, similar trends were also shown by increasing drug concentration regardless of the pH. As per the literature, pH is an important process parameter which affects the encapsulation of drug, and it also influences the surface charge and chemical structure of the molecules [39,40]. At lower pH, GEM and MIL-101NH 2 (Fe) possess positive charge which leads to less encapsulation of GEM [41]. When pH was changed by NaOH, MOF becomes negative by giving the proton to the solvent [42]. The optimum pH for higher EE was found 4.5. At higher pH, MOF became more negative and GEM was neutral at 6-7 pH [43]. Hence, there is less interaction between GEM and MOF. Fig. 1b presents that when drug concentration was increased, EE was also increased to a certain extent; however, at higher concentration, there was no impact of drug concentration. At low drug concentration, enough site was available for encapsulation; however, when concentration was increased, all sites became saturated and no impact of increased drug concentration [40]. Fig. 1c demonstrates that time has not significant (p>0.05) on EE. In Design-Expert, there are three ways for optimization: graphical optimization, numerical optimization, and point prediction [44]. In this study, we have selected numerical optimization, in which we have selected the best target for each factor, i.e., higher EE. To confirm the model adequacy for predicting the response function, we again performed the experiment using optimal condition given in Table 4. Table 4 presents that the experimental results and predicted results are very close, suggesting the reliability of optimized formulation. Fig. 2 presents the perturbation plots, used for better understanding the optimization procedure. The steepest curve presents the response  sensitiveness to specific variable. Fig. 2 reveals that variable B (pH) is more influential on EE followed by A and C.

Physicochemical characterization of optimized formulation
Particle size analysis by dynamic light scattering (DLS) of freeze-dried MIL-101NH 2 (Fe) revealed an average particle diameter of 158.1 ± 10 nm. However, there was a slight increase in particle size of MIL-101NH 2 (Fe)-GEM, i.e., 252.9 ± 9.23 nm. The increase in particle size was attributed to the encapsulation of drug into the MIL-101NH 2 (Fe). Zeta potential of MIL-101NH 2 (Fe) and MIL-101NH 2 (Fe)-GEM was determined according to the Helmholtz-Smoluchowski equation from their electrophoresis mobility and was found to be 30.75 ± 3.8 mV which shows good stability. MIL-101NH 2 (Fe) has shown a facetted-type architecture as observed by FESEM ( Fig. 3a and b) and mean diameters of 158.1 ± 10 nm correlated with DLS. SEM of MIL-101NH 2 (Fe) and MIL-101NH 2 (Fe)-GEM was well crystalline structures. This was further confirmed by PXRD data in Fig. 3c which shows crystalline nature of MOF, good agreement with the literature [45]. No peak shift was observed in GEM-loaded MOF (MIL-101NH 2 [Fe]-GEM) which confirms that drug was comfortably accommodated in this porous MOF. GEMloaded MOF showed partial amorphization demonstrated by PXRD patterns (Fig. 3c).

Fourier-transform infrared spectroscopy
FTIR spectroscopy was performed for the confirmation of drug in these nanocarriers. The FTIR spectra (Fig. 4a) of GEM showed characteristics bending vibrations of amines at 1418 cm −1 and 1636 cm −1 and stretching vibration of amine at 3359 cm −1 . Fig. 4b shows the characteristic peak of MIL-101NH 2 (Fe) at 1382, 1522, 3614, and 3742 cm −1 , corresponding to C-N stretch C-O stretching and N-H vibration (asymmetric and symmetric), respectively [45,46]. Drug-loaded MOF showed the presence of N-H bands around 1642 cm −1 , corresponding to the N-H groups from the amine moieties and shift of band 3359 cm −1 to 3349 cm−1, showing the presence of GEM NH2 groups (Fig. 4c) In vitro drug release of optimized formulation Fig. 5 presents the in vitro drug release profile of optimized formulation. GEM-loaded optimized formulation was found quite stable during the study time, i.e., 72 h. GEM release profile from MIL-101NH 2 (Fe)-GEM was also significant (one-way ANOVA, p<0.05) in comparison to native GEM. MIL-101NH 2 (Fe)-GEM showed burst effect up to 2 h of the study; however, later on, up to 72 h sustained release was observed. The reason of burst effect might be loosely attached molecule which was present on the surface of MOF. On the contrary, GEM was progressively released under physiological simulated conditions (PBS, 37°C) with 25, 60, and 99.9% of drug release after 0.5, 2, and 6 h, respectively. Thus, MIL-101NH 2 (Fe)-GEM nanoparticles clearly appeared as promising candidates for the delivery of GEM, showing both very high payloads and progressive drug release in physiological-simulated conditions.

Storage stability analysis
Stability studies were carried out at 25 ± 2°C/60 ± 5% RH to find out the effect of storage temperature. The results of stability testing in Table 5 revealed no significant change in color, aggregation, and crystallinity, but there was a slight change in size and EE. Therefore, optimized formulation was found to be stable for 6 months.