Development of a novel formulation method to prepare liposomal Epacadostat
Abstract
Background: One of the important metabolic pathways in cancer progression is tryptophan catabolism by the indoleamin-2,3-dioxygenase (IDO) enzyme, which suppresses the immune system and induces tolerance. Inhi- bition of IDO1 is an important therapeutic goal for immunotherapy in many cancers such as metastatic mela- noma. Epacadostat (EPA) is a very strong inhibitor of IDO1, and its clinical studies are being performed in a higher clinical phase than other inhibitors. In this study, we have developed a new liposomal EPA formulation to reduce the dose, side effects, and treatment costs.
Methods: Liposomes containing EPA were formulated using a novel remote loading method. Their morphology, particle size, surface charge, total phospholipid content, and drug loading were evaluated. Validation method studies to assay of EPA were carried out according to ICHQ2B guidelines. For in-vivo study, B16F10 melanoma bearing C57BL/6 mice were treated with the free or liposomal forms of EPA, and then monitored for tumor size and survival rate.
Results: A validated method for EPA determination in liposomal form using UV-visible spectrophotometry was developed which was a precise, accurate and robust method. The particle size, zeta potential, and encapsulation efficacy of liposomes was 128.1 ± 1.1 nm, -16.5 ± 1 mV, and 64.9 ± 3.5, respectively. The half maximal inhibitory concentration (IC50) of liposomal EPA was 64 ng/ml that was lower than free EPA (128 ng/ml). In-vivo results also showed that tumor growth was slower in mice receiving liposomal EPA than in the group receiving free EPA.
Conclusion: A new method was developed to load EPA into liposomes. Moreover, the use of the nanoliposomal EPA showed more efficacy than EPA in inhibiting the tumor growth in melanoma model. Therefore, it might be used in further clinical studies as a good candidate for immunotherapy alone or in combination with other treatments.
1. Introduction
Cancer is one of the major threats to human health (Cheng et al., 2014). Common treatments for cancer, including surgery, chemo- therapy, and radiotherapy, are ineffective in treating cancer alone because of complications related to clinical application and treatment (Rebecca et al., 2003). About 90% of patients treated with chemo- therapy die from multidrug resistance (MDR) and cancer recurrence (Wang et al., 2017). Cancer recurrence is often related to the drug’s inability to appropriately target cancer cells, its low penetration, and unwanted micro-metastases into tumor tissue (Wolinsky et al., 2012). Moreover, one of the important reasons for the failure of chemotherapy in cancer is related to suppressing the immune system by the tumor microenvironment. Therefore, one of the newest and most effective approaches in cancer treatment is to improve the patient’s immune system (Sideras et al., 2014; Sun et al., 2021). Compared with conventional cancer therapies that damage healthy cells in addition to cancer cells, immunotherapy works more effectively on tumor cells, leading to fewer side effects (Tahaghoghi-Hajghorbani et al., 2017).
For successful immunotherapy in cancer, it is important to pay attention to successive biological pathways and gain sufficient knowl- edge for them. One of these critical metabolic pathways in cancer pro- gression is tryptophan catabolism by the indoleamin-2, 3-dioxygenase (IDO) enzyme, which suppresses the immune system and induces im- mune tolerance (Amobi et al., 2017; Liu et al., 2021). IDO1 has the highest activity in catalysis and degeneration of tryptophan (Km ~ 20 μM) (Zhai et al., 2018). Tryptophan is an essential amino acid that im- mune cells, especially T and NK cells, need for activation and function (Sza´nto´ et al., 2007). Many studies have shown that reduced tryptophan and increased kynurenine suppress the immune system and cause em- powers the growth of cancer cells, and eventually immune escape in cancer by preventing the activation and proliferation of T and NK cells and sensitizing them to apoptotic stimuli, reducing CD8 effector T cells, increasing CD4 regulatory T cells (Treg) and myeloid-derived suppressor cells (MDSC) (Yan et al., 2010; Yu et al., 2013; Greene et al., 2019). An increase in inflammatory factors, including interferon-gamma (IFN-γ) in the tumor environment, has led to a greater expression of the IDO1 enzyme, which is considered one factor in MRD resistance to chemo- therapy(Campia et al., 2015). Therefore, inhibition of IDO1 is a funda- mental therapeutic goal for immunotherapy in many cancers, including malignant and metastatic types.
Currently, IDO1 enzyme inhibitors considered in clinical trials include Epacadostat(INCB024360) (Kristeleit et al., 2017), Indoximod (Soliman et al., 2016), Navoximod (NLG-919) (Nayak-Kapoor et al., 2018), BMS986205(Siu et al., 2017), and PF-06840003 (Reardon et al., 2020). Comparison of EC50 and dosage of these inhibitors shows that Epacadostat is a potent inhibitor of IDO1, highly selective, and its clinical studies are being performed in a higher clinical phase than other inhibitors. Clinical trials of EPA for treating melanoma, lung, kidney, head, and neck cancers are being conducted in phase 3 trials (Liu et al., 2018). Given that previous studies have shown high levels of IDO1 enzyme in metastatic melanoma and promising immunotherapy in its treatment, metastatic melanoma is a good candidate for this study (Chen et al., 2007; Jia et al., 2018).
Epacadostat (EPA) is an oral drug that can inhibit the IDO1 enzyme in advanced melanoma by 90% at doses of 300 mg twice daily. In phase 2 clinical trials, this dose has been suggested as the appropriate dose. Side effects of high doses include fatigue, loss of appetite, nausea and vomiting, rash, liver infection, pneumonia and dyspnea, and hypoxia (Naing et al., 2018).
As an IDO1 inhibitor, the EPA is an effective and practical remedy for immunotherapy versus various cancers. IDO1 overexpression is associ- ated with several diseases, including cancers. Since IDO1 is usually expressed throughout the body, particularly in inflammation and infection made by cytokines, a liposomal form of EPA can target the drug to the inflamed or infected site and reduced its off-target effects. Moreover, the required dose of EPA can be decreased when we use targeted delivery systems such as liposomes. This will reduce the product’s final price, particularly for expensive drugs like EPA, and make it affordable (Bondhopadhyay et al., 2020).
Nanoparticles increase efficacy and improve the pharmacokinetics of the drug, including distribution, circulation time, increased solubility, half-life, stability, and intracellular absorption and concentration, and reduced dosage and drug toxicity (Mishra et al., 2013). Nanoliposomes are most commonly used for various drug deliveries among the nano- particles, including injectable, oral, and topical delivery of anti-cancer drugs. The amphiphilic nature of liposomes allows the loading of a va- riety of hydrophilic and hydrophobic drugs. Also, liposomes have very high biocompatibility (Panahi et al., 2017). However, direct injection of cancer drugs has many side effects due to adverse effects on other cells in the body. PEGylation of liposomes prevents them from being detected by macrophages, thus increasing their circulation time in the blood and the drug’s half-life (Milla et al., 2012). Therefore, PEGylated nanoliposomes are more suitable for intravenous drug delivery and are more effective in clinical trials for anti-cancer drug delivery (Nikpoor et al., 2015). Also, since it is easy to produce them industrially, the final product can be controlled in quality and pharmaceutical specifications (Marcato and Dura´n, 2008).
To the best of our knowledge, there is only one recently published article which has formulated a pH-sensitive liposomal dual-delivery system for doxorubicin (DOX) and EPA (Chen et al., 2020). However, new, stable, and high drug-loaded nanoliposomal forms of EPA were developed in the present study. Moreover, we have developed a vali- dated UV-Vis spectroscopic method in the current study to quantify the amount of EPA in bulk and liposomal forms. Also, we could successfully quantify the amount of EPA in a liposomal formulation in a precise manner. At the same time, the lipid composition had no interaction with EPA, and no extraction procedure was needed, as well. Drug release profile, cellular uptake, cytotoxicity, and potency of the prepared for- mulations were studied to select the optimum formulation. Finally, a formula with the desirable in vitro results was nominated for in vivo experiments.
2. Materials and methods
2.1. Materials
EPA was purchased from MetonChem (China), Mouse IFN-γ was purchased from BioLegend (San Diego), Hydrogenated soy phosphati- dylcholine (HSPC), cholesterol (Chol), mPEG2000- distearoylphosphatidyl-ethanolamine (mPEG2000-DSPE), dipalmitoyl- sn-glycero-3-phosphoglycerol (DPPG) were purchased from Avanti Polar Lipid (Alabaster, USA). All solvents and materials were purchased from Merck (Germany) and were in the analytical grade.
2.2. Method validation studies of EPA
Using an isolator and in a sterile condition, 100 mg of the drug powder was dissolved in 20 ml of sterile ethanol and kept at -70◦C until use. Different concentrations of the drug in triplicate were prepared in methanol using the serial dilution method and scanned with a spectro- photometer (200-800 nm wavelengths). Then, the most optimal wave- length, which showed a single specific peak, was selected as a particular wavelength to prepare a calibration curve. The same was done for empty liposomes to ensure that no detectable absorbance for empty liposomes.
Validation studies were carried out according to ICH guidelines and are mentioned as follows;To evaluate the linearity of the assay, the calibration curve, which is the plot of absorbance against concentration (90 µg/ml), was plotted, and the regression line was determined. For determining accuracy, 80, 100, and 120% test concentrations were prepared and analyzed. Percent recovery, mean and RSD% were also calculated. Precision was deter- mined by assessing repeatability, intraday precision, and interday pre- cision. Six replicates of 100% test concentration were evaluated. Mean, SD, RSD% were determined to confirm the repeatability of the method. For determining intraday and interday precision, 100% test concentra- tion was analyzed at different time points in one day and three times for three consecutive days. Then, RSD% was calculated. The detection limit was calculated using the following equation, where σ is the standard deviation of response and S is the slope of the calibration curve.
The quantification limit was calculated using the following equation: LOQ=10σ/S And as the final step, robustness was determined by carrying out the procedure at different wavelengths. Recovery%, mean, and RSD% were calculated. A liposomal formulation encapsulating a certain amount of EPA was synthesized, and the amount of the drug was analyzed using the proposed method.
2.3. Preparation of liposomal EPA by remote loading method
As shown in Table 3, specific amounts of lipids were taken from a stock solution (40mg/ml lipid) in chloroform and poured into round- bottomed glass flasks. The chloroform solvent was removed using a vacuum rotary device at 45◦C for 2 h followed by freeze-drying for 3 h to
remove trace amounts of solvent. The lipid film was hydrated by different concentrations of buffer (250 mM and 300 mM ammonium sulfate buffer) to reach different final concentrations of total lipid (43 mM, 50 mM, and 75 mM). The milky colloid containing multilamellar vesicles (MLVs) was first shaken vigorously and placed in a bath sonicator at 65◦C. Then, empty liposomes were extruded using an extruder
(Avestin, Canada) through different polycarbonate filters (400nm, 200nm, 100nm, and 50nm filters) to achieve small unilamellar vesicles (SUVs). A dialysis bag (~12 kDa cut off, Spectrum, Houston, USA) was used to remove ammonium sulfate buffer from liposomes and substituted with a 10 mM histidine buffer containing 10% w/v of su- crose (pH 6.8).
To encapsulate EPA using the remote loading method, various con- centrations of drug were added to 1000 µl of empty liposome according to Table 3 (F1- F11) and placed in a water bath in different temperatures and incubation time (65 ◦C, 68 ◦C, 70◦C for 10,15, and 20 mins). Then,they were transferred immediately to the cold-water bath to reach a temperature below Tm of the liposomes. A dialysis bag was used to separate the encapsulated drug from the non-encapsulated one in the presence of 10 mM histidine buffer containing 10% w/v of sucrose (pH 6.8). The final liposomes were stored at 4 ◦C until being used.
2.4. Physicochemical characterization of liposomes
Physicochemical properties of liposomes, including morphology, size, surface charge, total phospholipid content, and drug loading, were measured. Liposome particle size, polydispersity index (PDI), and zeta potential were measured by dynamic light scattering (DLS) using a particle size analyzer (Nano-ZS; Malvern, UK). 50 µl of samples were mixed with 950 µl of used buffer and added to size measurement cuvette. For zeta potential measurement, 20 µl of samples were mixed with 980 µl of MOPS (morpholinopropane-1-sulfonic acid) buffer (pH 6.5). The morphology of nanoliposomal EPA was investigated using transmission electron microscopy (TEM). 50 µl of liposome formulation was diluted with hydration buffer (1: 50 ratio) and dropped onto the carbon-coated copper grid. Fresh and filtered uranyl acetate solution (2%w/v, pH4.5) was added to the dried film on the grid to negative stain liposomes. A dried grid was then photographed using a Leo 912 AB transmission electron microscope (Zeiss, Jena, Germany).
The total phospholipids content was determined by the Bartlett phosphate assay method (Bartlett, 1959). The principle of the Bartlett assay is based on the colorimetric determination of inorganic phosphate.Briefly, a 400µl of H2SO4 (10mM) was added to each liposomal sample and phosphorous standards and then placed at 200◦C for one hour., After cooling to room temperature, 100 µl of H2O2 (10%) was added to each sample and placed at 200◦C for 10 min. Next, a certain amount of ascorbic acid and ammonium molybdate was added to each sample and incubated at 100◦C for 10 min. Finally, the samples were cooled rapidly to room temperature, and a spectrophotometer measured their absorp-
tion at a wavelength of 800 nm.
The encapsulation efficacy of EPA was calculated by dissolving liposome bilayer using pure methanol. The 100 µl of liposome sample was added to 900 µl methanol and stirred in a water bath at 60◦C for a few minutes. Then, the absorbance of empty liposomes and sample were
read using a UV-visible spectrophotometer (SPEKOL 1300, Analytik Jena, Germany) at 287 nm based on the validation method mentioned earlier. The encapsulation efficacy was calculated using the following formula:To evaluate liposomal EPA stability, liposome particle size, PDI, zeta potential, and encapsulation efficiency were measured at 0, 7, 14, and 21 days after preparation and storage at 4◦C.
2.5. Evaluating EPA release profile from the liposomes
The release profile of EPA from liposomes was also assessed. Under the sterile condition, 0.5 ml of the liposome formulation (F10 or F12) was transferred to a dialysis bag (cut off ~14 kDa) and then dripped into a phosphate-buffered saline (PBS) buffer, then incubated at 37◦C while gently stirred for 24 hours. The experiment was done at three different pH of 5.5, 6.5, and 7.4 to mimic the pH of endosome, tumor environ- ment, and plasma, respectively. 2 ml (3 duplicates) of the medium was removed and replaced with an equal volume of fresh medium at different time points (0, 2, 4, 6, 10, 12, 24 and 48 hours). The released drug amount was detected at each time point via UV-Vis spectropho- tometry. The cumulative percentage of drug release was calculated and plotted versus time.
2.6. Evaluating cellular binding & uptake of the nanoliposomal product by B16F10 cancer cells
To confirm the uptake of EPA in free or liposomal form by cells, B16F10 cells (5 × 105 cells) were cultured on 12-well plates and stored for 24 hours at 37◦C. The following day, tumor cells were washed with PBS (pH 7.4), and then a particular concentration of EPA in free or F10 and F12 were added to each well and placed at 4 ◦ C and 37◦C for 4 h. The cells were then removed from the bottom of the wells, and their absorption was read at a wavelength of 287 nm.
2.7. Cytotoxicity assay
MTT assay was performed on the B16F10 cancer cell and the normal 3T3 (NIH) cell line to assess the cellular toxicity of EPA in liposomal and free form. In summary, B16F10 or NIH cells (5,000 cells/100 μl) in DMEM culture medium were cultured in 96-cell plates. After 24 hours of incubation at 37◦C, various concentrations of EPA, F1, F11, F10, and F12 were added in triplicate to each well. MTT reagent and then DMSO were added to the plates after incubation for hours and finally read with ELISA Reader (Stat Fax 2100). 20 mg/ml of EPA and DMEM culture medium was used as a positive control and negative control, respectively.
2.8. Evaluating the inhibitory potency of EPA and Liposomal EPA
B16F10 cells (3,000 cells) were cultured in 96-cell plates and DMEM culture medium containing L-tryptophan. After 24 hours of incubation at 37◦C, the cells were induced by INF-γ (100 g/ml) to stimulate IDO enzyme production. 24 hours later, different concentrations of EPA and
liposomal EPA (1000, 256, 128, 64, 32, 32, 16, 8, 4, and 2 ng/ml) were added to each well in triplicate in an environment containing L-gluta- mine and L-tryptophan (50µg/ml). A certain amount of trichloroacetic acid (for the hydrolysis of N-formyl-kynurenine to kynurenine) was added to each well after 24 hours. The plate was placed at 50◦C for half an hour. The plate was centrifuged at 2500 rpm to remove the sediment produced by hydrolysis. After adding the Erlich reagent to each well, the amount of produced kynurenine was measured by the ELISA Reader. No INF-γ was added to three wells which were negative controls. For pos- itive control, no IDO inhibitor drug (EPA) was added to three wells.
2.9. In vivo anti-tumor study
Induction of melanoma tumors in C57BL6 mice was performed using the B16F10 cell line. Female six-to-eight-week-old C57BL6 mice were obtained from Pasteur Institute (Iran, Tehran) and maintained in the animal house (BuAli Research Institute, Mashhad, Iran) under pathogen- free conditions and approval of the Institutional Ethical Committee of Mashhad University of Medical Sciences (dated November 21, 2018;
proposal code 970822). For the in-vivo study, 5 × 105 B16F10 tumor cells were injected subcutaneously at the flank of the mice (total 21 mice). On day 10, after tumor cells challenges, the mice with palpable tumors were divided into three different groups of seven mice, including a) PBS, b) EPA, and c) liposomal EPA (F10). Mice received 60µg of free or liposomal EPA intravenously, three times at three-day intervals. The control group received PBS. The mice were monitored for tumor size and survival rate throughout the study. Tumor masses were measured with a digital caliper in three orthogonal diameters, and volumes were deter- mined. The mice were monitored for up to 22 days unless one of the following conditions for euthanization was met: (a) body weight decrease greater than 15% of initial weight, (b) tumor growth greater than 2000 mm3, and (c) finding them dead. All surviving mice were euthanized on day 22.
2.10. Statistical analysis
Data were analyzed using GraphPad Prism version 6 (GraphPad Software, San Diego, CA, USA). Quantitative variables were expressed as the mean ± the standard deviation (SD). Statistical analysis of the difference between groups was determined by one way of analysis of variance (ANOVA) and Tukey’s post hoc test for one-way ANOVA. Survival was analyzed by Mantel–Cox test. P values < 0.05 were considered statistically significant. 3. Results 3.1. Evaluation of EPA specifications Different concentrations of the EPA were prepared in methanol sol- vent using serial dilution and triplicate and scanned with a spectro- photometer. UV/visible spectra scan of EPA showed an absorption peak at 287 nm without overlapping with liposomes (Figs. 1 and S1). Therefore, there is a linear relationship between the concentration and the absorbance of EPA; the correlation coefficient was 0.9999 (Fig. 2, Table 1). As shown in Table 2, UV/visible validation data for EPA exhibited a Limit of detection (LOD) and limit of quantitation (LOQ) 2.37 and 7.19 µg/ml, respectively. Therefore, the samples were prepared in higher concentrations based on the results of LOD and LOQ. The three test concentrations (80%, 100%, and 120%) and accepted reference values were in close agreement, and the relative standard deviation (RSD) % were admitted according to AOAC guidelines. Therefore, the accuracy of the method was confirmed (Table S1). The results showed that RSD% values for robustness (Table S2),repeatability (Table S3), intraday (Table S4), and interday precision values (Table S5) were at the acceptable range. Thus, the method can be considered precise. Fig. 1. UV/visible spectra of EPA in Methanol. The absorption spectrum shows an absorption peak at 287 nm for EPA without overlapping with liposomes ingredients. EPA: Epacadostat.
Fig. 2. Calibration curve for UV/visible spectrophotometric determination of EPA in methanol under various concentrations (λ max = 287 nm). Data are presented as Mean ± SD (n=3). ABS: Absorbance.
3.2. Drug encapsulation and characteristics of liposomes
Drug encapsulation and physicochemical characteristics of lipo- somes were given in Table 3. In this study, we prepared two different empty liposome formulations. The total lipid concentration of empty liposomes was adjusted to 43, 50, or 75 mM. Then, EPA was loaded into these liposomes using the remote loading method (F2, F3, F4, F5, F6, F7, F8, F9, F10, F12, F13 and F14). Almost all formulations showed mon- omodal and similar particle size distribution given the Z-average and PDI, which were no significant differences between F1, F10, F11, F12, and F13. But F2, F3, and F14, with a total lipid concentration of 43mM or 75mM (described in detail in Table 3), showed a larger particle size than other formulations. There were no significant differences between the F1, F10, F11, F12, and F13 formulations in terms of Z-potential, but a significant (*p<0.05) difference in surface charge was seen between F2, F3, and F14 with other formulations. PDI of all formulations showed a monomodal system with values less than 0.1 in all formulations. The encapsulation efficacy of F2, F3, F4, F5, F6, F7, F8, F9, F10, F12, F13, and F14 was obtained 76.4%, 70.7%, 71.8, 44%, 47.1%, 24.4%, 51.8%, 49.6%, 64.9%, 74.7%, 67.7% and 65.3% respectively. Increasing the total lipid concentration from 50 mM to 75 mM did not significantly affect the encapsulation efficacy of F14. The use of 5% and 10% DPPG in F12 and F13 formulations increased the encapsulation efficacy 2.78% and 9.86% compared to groups without DPPG, respectively. Moreover, F2, F3, F4, and F12 showed encapsulation efficacy about 10% higher than others. Concerning the acceptable size, higher zeta potential, and higher encapsulation efficacy of formulas F10 and F12, they were considered optimal formulas, and subsequent in vitro and in vivo tests were performed. The phosphorus content of optimal formulations (F10, F12) agreed with the expected lipids molar ratio, as shown in Table 4. The drug release rate from the liposomes was investigated using PBS buffer at pH 5.5, 6.5, and 7.4. The initial concentration of the drug in F10 and F12 was 130 and 150 µg/ml, respectively. The maximum per- centages of EPA released from F10 and F12 were 9.9% and 19.3% at pH 7.4, respectively. Cumulative EPA release at pH 6.5 was obtained at 63.8% and 14.9% for F10 and F12, respectively, after 24 hours. More- over, the EPA release at pH 5.5 was obtained 91.3% and 11.1% for F10 and F12 after 24 hours, respectively (Fig. 3). The release study showed that the EPA release rate of F12 was significantly (****p<0.0001) higher than F10 at pH 7.4. But the EPA release from F10 was significantly (****p<0.0001) higher than from F12 at pH 6.5 and 5.5. The results of the physical stability of liposomes showed that10.23% and 25.36% of the EPA has leaked from F10 and F12 within 14 days, respectively. There was a significant difference between the leakage of EPA from F10 and F12 on days 2, 4, 8, 12, and 14 (***p<0.001). In visual observation, there was a little aggregation in F12after 14 days at 4◦C (Fig. 4). Particle size average, PDI, and zeta potential of F10 and F12 through 3 weeks stored at 4◦C were also given in Table 5. There was a significant difference in the zeta potential of F12 between day 0 and day 14 (****p<0.0001), but it was not true for F10. Evaluating cellular binding and uptake of EPA in Liposomal or free form was shown in Fig. 5. The binding of F10 and F12 was not different from the free EPA. The cellular uptake of the liposomal drug was nearly decreased by half compared to the free EPA (**p<0.01).According to the relevant stability results of the F10 formulation, it was selected as an optimal formulation. The TEM image of the F10 formulation was shown in Fig. 6. The image shows a uniform spherical liposome with a diameter around 130 nm. Fig. 3. Cumulative percentage of EPA release profile of F10 and F12 formulations in PBS (pH 7.4) (A), (pH 6.5) (B) and (pH 5.5) (C) at 37◦C. The error bars were obtained from triplicate samples. The initial concentration of the drug in F10 and F12 was 130 and 150 µg/ml, respectively. **** p<0.0001, **** p<0.0001 relative to F12. Fig. 4. Cumulative percentage of EPA leakage from liposomes in histidine su- crose buffer (10 mM, sucrose 10%, pH 6.8) at 4◦C. Data represent mean and standard deviations of three replicates (Mean ± SD, n = 3). *** p<0.001 relative to F12. 3.3. In vitro cytotoxicity & potency of EPA and liposomal EPA The cytotoxic effect of various concentrations of EPA and liposomal EPA (6.25, 12.5, 25, 50, 100, 200, 400, 800, and 1600 µg/ml) were evaluated on two mice cell lines including (NIH (3T3)) as the normal cells and (B16F10) as a cancerous cell line using MTT assay. As shown in Fig. 7A, MTT results demonstrated that EPA, F10, and F12 had a cyto- toxic effect on NIH (3T3) cells at concentrations 350, 500, and 505 µg/ ml, respectively. Also, EPA, F10, and F12 showed a cytotoxic effect on B16F10 cells at concentrations 300, 432, and 435 µg/ml, respectively (Fig. 7B). However, the cytotoxic effect of empty liposomes (F1 and F11) was not found in any of the cell lines. There was a significant difference in cytotoxic potency of EPA and F11 at a concentration of 100 µg/ml and also EPA and F1 at a concentration of 200 µg/ml on the NIH (3T3) cells (*p<0.05). Furthermore, a significant difference has been found between cytotoxicity of F11 and F12 at a concentration of 200 µg/ml (*p<0.05) and F1 and F10 at a concentration of 400 µg/ml on the NIH (3T3) cells (**p<0.01) (Fig. 7A). Also, a significant difference was seen between EPA and F11 cytotoxicity at a concentration of 100 µg/ml (*p<0.05) and between EPA and F1 at a concentration of 400 µg/ml on the B16F10 cells (**p<0.01). Furthermore, a significant difference in enzyme and treated with different concentrations of EPA and liposomal EPA (256, 128, 64, 32, 32, 16, 8, 4, and 2 ng/ml). Eventually, the pro- duction of kynurenine was investigated. As shown in Fig. 8, the half maximal inhibitory concentration (IC50) of liposomal EPA was lower than free EPA. The IC50 values for EPA, F10, and F12 were 128, 64, and 65 ng/ml, respectively. A significant difference was found between the inhibitory potency of EPA, F1, and F10 (*p<0.05) and F12 (**p<0.01). Fig. 5. Evaluating cellular binding and uptake of EPA in liposomal or free form by B16F10 cells. A similar concentration of EPA in liposomal or free form was added to each well and placed at 4◦C to evaluate their binding capacity (A) and at 37◦C to evaluate their uptake assay (B) after 4 hours. **p<0.01 relative to EPA, F10, and F12. Error bars represent standard deviations of three replicates. Fig. 6. Transmission electron microscopy (TEM) image was used to explore the morphology of F10. The magnification was X10000 (left side) and X50000 (right side), respectively. The internal buffer was 250 mM ammonium sulfate, pH 3.6. The external buffer was histidine sucrose (10 mM, sucrose 10%, and pH 6.8). The image shows a uniform spherical liposome with a mean diameter of ~130 nm. 3.4. In vivo anti-tumor effects of EPA and liposomal EPA We examined the in vivo anti-tumor effects of free or liposomal EPA three times at days 10, 13, and 16 in B16F10 melanoma–bearing mice. Mice received intravenously 60µg of free or liposomal EPA at days 10, 13, and 16 after tumor inoculation. The tumor volume was measured with 3 day intervals. As shown in Fig. 9, the tumor growth in the mice that received liposomal EPA was slower than that of the other groups (****p<0.0001). The survival percent was higher in this group than in the PBS group but not statistically significant. 4. Discussion A better understanding of the mechanisms that enable tumors to escape the immune system is reached during the past decade, and it was elucidated that most immunotherapies to treat solid tumors fail due to the immunosuppressive microenvironment of tumors (Prendergast and Jaffee, 2007). One of the mechanisms that tumors utilize to hinder im- mune responses is the secretion of IDO1, which is involved in immu- nometabolism and inflammatory programming via its role in tryptophan catabolism (Munn and Mellor, 2016). As it’s shown in pre-clinical models, targeting the IDO1 enzyme can enhance tumor therapy ap- proaches such as radiotherapy, chemotherapy, and immune checkpoint blockade without any increase in their side effects (Prendergast et al., 2017). EPA is a particular inhibitor of IDO1 and has shown some promising results when combined with anti-PD1 therapy in patients with mela- noma, head and neck carcinoma, renal and urothelial cancers, which is suggested to be through the increase in IL-2 secretion and CD8+ T cells’ proliferation (Spranger et al., 2014; Rose, 2017). However, EPA can cause some serious complications for a patient’s health, which could be due to inhibition of a crucial enzyme, IDO1, in various organs and tissues and the fact that it mimics tryptophan which can lead to induction of pro/anti-inflammatory pathways (Günther et al., 2019). To overcome these issues, we decided to develop nanoliposomes with different for- mulas for carrying EPA and delivering it to the tumor site. Formulated liposomes had to have high stability to avoid a rapid drug release and a prolonged circulation time. Our formulation is based on the PEGylated Liposome and better works out for intravenous administration than the other drug delivery systems, particularly micelles in terms of stability in the blood circulation. Although hydrophobe drugs may be encapsulated in micelles easier than liposomes, they are extensively dependent on the critical micelle concentration (CMC) despite the liposomes. Therefore, when a liposomal drug enters the bloodstream, it is not affected by dilution factor because of blood volume (5 L), and the cargo’s structure will still be stable with a lower drug release, but this is not true for micelles. Fig. 7. Cell survival analysis of free EPA or Liposomal EPA. The graphs (A and B) demonstrate NIH (3T3) & B16F10 cells viability and survival rate under treatment with different concentration of EPA in Liposomal or free form at 48 hours post-treatment. All experiments were performed in triplicate and data are presented as means ± SD. *p<0.05, **p<0.01 and ****p<0.0001 relative to EPA, F1, F10, F11 and F12. Fig. 8. Evaluation of the inhibitory potency of EPA in Liposomal or free form on the Kynurenine producing of B16F10 cells after induction of IDO1 enzyme by IFN-γ treatment. Cells were treated with different concentrations of EPA or liposomal EPA (256, 128, 64, 32, 32, 16, 8, 4, and 2 ng/ml). Kynurenine levels were measured in the media by the ELISA method. The IC50 values for EPA, F10, and F12 were 128, 64, and 65 ng/ml, respectively. The IC50 values for F1 and F11 were >256 ng/ml. *p<0.05, **p<0.01 relative to EPA, F1, F10, F11 and F12. Error bars represent standard deviations of three replicates. Fig. 9. In vivo anti-tumor effects of EPA in liposomal or free form. (A)The tumor volume (mm3) of each mouse in each group was measured three times a week until day 22. The graph shows average with SD (n=7). ****p<0.0001,**p<0.01 relative to PBS, EPA and F10 on days 16, 19 and 22 after tumor inoculation. (B) The survival percent of the mice were analyzed by log-rank (Mantel–Cox) test over 22 days. On the other hand, smaller nanoparticles such as micelles can deeply penetrate the tumor microenvironment, which is impossible for big particles (Wakaskar, 2018; Priev et al., 2002). At present, two Nano-drugs approved by FDA, Doxil and Abraxane , are available as cancer remedies (Barenholz, 2012). Although using the Doxil has detracted from the side effects of the drug, one of the upcoming chal- lenges is the precipitation of the drug and the absence of drug-releasing in the tumor microenvironment. The anti-tumor efficacy of Doxil is hindered by the poor release of the active drug from the liposome at the tumor sites. Hence, many studies are conducting to resolve the up- coming problem by different strategies (Zhao et al., 2013). Also, Abraxane has faint colloidal perseverance and decomposes in the bloodstream after intravenous injection owing to being diluted (Ruttala and Ko, 2015). The high chemical stability and suitable transition temperature of HSPC made it an excellent candidate to be used in our nanoliposomes. Also, PEGylation was used to enhance in-vivo liposomal stability by preventing plasma proteins from interacting with liposomes and pre- venting rapid clearance from circulation via macrophages or renal filtration. In addition, DPPG is a negatively charged lipid that can enhance encapsulation efficacy by increasing the size of the liposomes and the percent of an entrapped drug, which has a positive charge at the pH of our formulation. Therefore, based on what was mentioned, we developed several different liposome formulations and evaluated them regarding particle size, PDI, Z-potential, EE%, and appearance. Among the developed liposomes, two of them (F10 and F12) were superior in respect to the above criteria. Therefore, the rest of the tests were performed using these liposomes. Chen et al. have previously worked on nanoliposome-based delivery of EPA; although they have reached some decent results, their formulation is different, and they have used DOTAP, DOPE, Chol, and DSPE-PEG2000 in their construct, so it is the first time that a lipid formulation composed of HSPC/cholesterol/ MPEG-2000- DSPE/DPPG is used to carry EPA to the tumor site (Chen et al., 2020). Compared to Chen et al. study, our nanoliposomes demonstrated better stability at physiological pH, which could be due to the use of remote loading method that led to the entrapment of EPA mainly in the core of liposomes. In contrast, in their study, the EPA was loaded using the passive loading method. It was inserted into the lipid bilayer, which probably releases drug content faster in blood than ours (Chen et al., 2020). A validation method was developed for EPA assay in the lipo- somal form, which is a rapid tool using the spectrophotometric-based method. According to the ICH guidelines, multiple validation parame- ters such as precision, accuracy, and reproducibility were tested to confirm the method’s robustness. It was confirmed that at the selected wavelength, 287 nm, we can evaluate the EPA assay. Regarding drug encapsulation percent, passive loading methods cannot sufficiently encapsulate hydrophobic drugs, but active loading methods such as the remote loading method are more efficient in hy- drophobic drugs than passive methods, and as a result, it leads to high entrapment and less drug losing during the preparation method. In addition, we could prevent the rapid in vivo drug leakage by using a remote loading method that is usually seen in passively encapsulated drugs in the bilayer and benefits from a more controlled drug release (Vakili-Ghartavol et al., 2020). Here, we managed to remotely load EPA using an ammonium sulfate gradient, a hydrophobic drug with a neutral charge at pH 6.8 and positively charged in pH <4, into the core of our liposomes. EPA’s interaction with ammonia makes it a neutral drug that can diffuse into the liposome; in the hydrophilic liposomal core where the pH is acid, the drug binds to protons and becomes charged and entrapped within the core. Loading the EPA in the core of liposomes led to increased stability of the drug. In fact, it is one of the advantages of our loading method as hydrophobic drugs are usually loaded in the hydrophobic bilayer of liposomes. One of the indicators of stability is the size distribution of the lipo- somes. Our analysis showed that the size of F10 and F12 was between 127 and 138 nm, and they had excellent uniformity of size distribution (PDI ≈ 0.1), indicating the stability of nanoliposomes. Furthermore, the EPA release profile of F10 and F12 in various pHs exhibited that the EPA in the F10 formulation has an apparent pH-sensitive release. The EPA is released upon reaching acidic environments, as the tumor microenvi- ronment possess an acidic pH, this could lead to selective delivery of EPA to the tumor site and decrease the complications induced by its systemic delivery. Therefore, because of this characteristic of F10 and its lower drug leakage, we chose it as our optimal formulation for further evaluation on animal models. The F10 and F12 formulations were found to exert cytotoxicity against B16F0 melanoma cells (at 432 and 435 µg/ml, respectively), though this was weaker than that exerted by free EPA (at 300 µg/ml). Due to the EPA’s price and its cytotoxic effects on various organs at high concentrations, one of the main goals of this study was to decrease the required dose of the EPA to inhibit kynurenine production without affecting the EPA’s inhibitory potential effect. Interestingly, the results of the kynurenine assessment demonstrated that the IC50 of F10 and F12 formulations (64, and 65 ng/ml, respectively) are lower than free EPA. Therefore, they can inhibit kynurenine production of B16F0 melanoma cells more efficiently, which shows the great potential of the nano- liposomal form of EPA to replace the free form of the drug in the clinical context. Regarding cellular binding and uptake of EPA by cells, although the binding efficacy of F10 and F12 was not different from the free EPA, the cellular uptake of the liposomal drug was nearly decreased by half compared to the free EPA, which could be because of steric hindrance provided by PEG in liposome formulation. However, the lower uptake of nanoliposomes can be compensated by their higher accumulation at the tumor site (Yingchoncharoen et al., 2016). The in-vivo study demonstrated that F10 could exert better anti- tumor effects than free EPA. The tumor growth was significantly slower in F10-treated mice in comparison to the EPA-treated group. Regarding survival, although the difference was not significant, the F10- treated group’s survival was improved. The combination of free EPA and anti-PD1 antibody have already been tested in phase 3 clinical trial (NCT02752074), suggesting that nanoliposomal-delivery of EPA could lead to far better results if we make use of it as a combinatory treatment along with chemotherapies, vaccines, or immune checkpoint inhibitors such as anti-PD-1 antibodies. 5. Conclusion The EPA determination using the UV-visible spectrophotometric method was found to be precise, accurate, and robust. Also, this method was successfully applied to quantify the amount of EPA in liposome formulation and achieve the encapsulation efficacy. In addition, the use of the nanoliposomal EPA reduces the dose, side effects, and treatment costs, and it’s more effective than the free EPA in inhibiting tumor growth in melanoma cancer. Therefore, it might be used in further clinical studies as a good candidate for immunotherapy alone or in combination with other treatments.