Production of Curcumin-Loaded Silk Fibroin Nanoparticles for Cancer Therapy

Curcumin, extracted from the rhizome of Curcuma longa, has been widely used in medicine for centuries due to its anti-inflammatory, anti-cancer, anti-oxidant and anti-microbial effects. However, its bioavailability during treatments is poor because of its low solubility in water, slow dissolution rate and rapid intestinal metabolism. For these reasons, improving the therapeutic efficiency of curcumin using nanocarriers (e.g., biopolymer nanoparticles) has been a research focus, to foster delivery of the curcumin inside cells due to their small size and large surface area. Silk fibroin from the Bombyx mori silkworm is a biopolymer characterized by its biocompatibility, biodegradability, amphiphilic chemistry, and excellent mechanical properties in various material formats. These features make silk fibroin nanoparticles useful vehicles for delivering therapeutic drugs, such as curcumin. Curcumin-loaded silk fibroin nanoparticles were synthesized using two procedures (physical adsorption and coprecipitation) more scalable than methods previously described using ionic liquids. The results showed that nanoparticle formulations were 155 to 170 nm in diameter with a zeta potential of approximately −45 mV. The curcumin-loaded silk fibroin nanoparticles obtained by both processing methods were cytotoxic to carcinogenic cells, while not decreasing viability of healthy cells. In the case of tumor cells, curcumin-loaded silk fibroin nanoparticles presented higher efficacy in cytotoxicity against neuroblastoma cells than hepatocarcinoma cells. In conclusion, curcumin-loaded silk fibroin nanoparticles constitute a biodegradable and biocompatible delivery system with the potential to treat tumors by local, long-term sustained drug delivery.


Introduction
The yellow-orange compound 1,7-bis(4-hydroxy-3-methoxyphenyl)-1,6-heptadiene-3,5-dione ( Figure S1), popularly known as curcumin, is the main phenolic pigment extracted from turmeric, the powdered rhizome of Curcuma longa, which comprises, approximately, 2-5% of turmeric [1]. Commercial curcumin is a mixture of curcuminoids, containing approximately 77% curcumin, 18% demethoxycurcumin and 5% bisdemethoxycurcumin [2]. This perennial rhizome is commonly cultivated in the Asian continent, especially in India and China. Apart from being used as a spice and flavoring and coloring agent in cooking, curcumin has also been widely employed in Ayurvedic medicine for centuries [3]. The most relevant pharmacological effects of curcumin are its anti-inflammatory [4][5][6][7][8], anti-cancer [2,[7][8][9][10][11][12][13], anti-oxidant [8,12,[14][15][16] and anti-microbial [7,12,17,18] activities. Furthermore, the SF solutions obtained in these processes tend to be unstable and turn into a gel in a relatively short period of time (approximately one week). Apart from using ionic liquids as solvents, we also improved SF dissolution using high-power ultrasound, which requires lower temperature and represents a more efficient energy source compared with thermal energy. While the classical heating method needs hours to complete the dissolution of the SF fiber, the application of ultrasounds to the mixture achieves a significant reduction in the time necessary to complete the SF dissolution process in ionic liquids at less than 100 • C. Therefore, in this work, the solvent effect of the cations and anions of the ionic liquids that break the β-sheet hydrogen bonds network was enhanced by ultrasounds. In addition, ionic liquids can be recycled and used without the loss of their solvent properties in at least four successive cycles. Therefore, using ionic liquids and high-power ultrasounds to dissolve SF constitutes a more environmentally friendly option because the process needs lower temperature, is faster, the SF solutions obtained are very stable and the ionic liquid can be reused in subsequent dissolution processes [66]. All of these reasons, make the synthesis of Curc-SFNs developed in this paper a more scalable and continuous processing option than those already published in the literature.
The main aim of this work was to synthesize Curc-SFNs using a simpler and greener process of SF dissolution and loading the drug through two different procedures (physical adsorption and coprecipitation). Physical-chemical characterization and in vitro cytotoxicity studies were assessed.

Materials
B. mori silk cocoons were reared in the sericulture facilities of the IMIDA (Murcia, Spain) and raised on a diet of natural Morus alba L. fresh leaves. To obtain SF, raw silk cocoons were boiled twice in a 0.05 M Na 2 CO 3 aqueous solution for 45 min. The remaining SF was rinsed thoroughly with ultrapure water and dried prior to use. SF was dissolved in the ionic liquid 1-ethyl-3-methylimidazolium acetate, [emim + ][CH 3 COO − ], by high-power ultrasound, as previously reported [66]. The ionic liquid (95% purity) was purchased from IoliTec GmbH (Frankfurt, Germany) and was used without further purification. Curcumin (99% purity) was purchased from ChromaDex (Irvine, CA, USA). Purified water (18.2 MΩ·cm at 25 • C; from a Millipore Direct-Q1 ultrapure water system, Billerica, MA, USA) was used throughout. All other chemicals and solvents were of analytical grade and were used without further purification.

UV-Vis Spectrophotometric Estimation of Curcumin
Spectroscopic analysis was carried out using a UV-Vis HELIOSα spectrophotometer (Thermo Scientific, Waltham, MA, USA) and good linear correlations were obtained between absorbance and concentration in the range 0.5-3.5 µg/mL with a correlation coefficient of 0.9974 in water, and in the range 1.0-7.0 µg/mL with a correlation coefficient of 0.9995 in ethanol. The spectrophotometric detection was determined at an absorption maximum of 421 nm using ethanol or water as solvent.

Preparation of SFNs
The preparation of SFNs was based on the method described previously by Lozano-Pérez et al. [66], with modifications. Briefly, an SF-ionic liquid (SIL) solution (10 wt %) was prepared by adding 0.5 g of SF to 4.5 g of [emim + ][CH 3 COO − ]. The mixture was treated with a 3/8 tapered horn of a Sonifier Branson 450D (Emmerson Ultrasonic Corporation, Danbury, CT, USA), with pulsating ultrasonication steps at 30% amplitude at a controlled temperature below 90 • C until complete dissolution. To this solution freshly prepared, 3 mL of ultrapure water was slowly added to reduce viscosity. The final concentration of the SIL solution after diluting with 3 mL of ultrapure water was 6.66 wt %. After heating to 60 • C, the SIL solution was propelled using a peristaltic pump and then sprayed onto 100 mL of gently stirred methanol at −20 • C by a thermostatically controlled 0.7 mm two-fluid nozzle (from a Mini Spray Dryer B-290, BÜCHI Labortechnik, Flawil, Switzerland, Part No. 044698) which uses compressed N 2 to disperse the  A milky white suspension appeared and the suspension was allowed to reach  room temperature while stirring for 2 h. Then, the nanoparticle suspension was transferred to centrifuge  vials and centrifuged at 13,400 rpm for 15 min, at 4 • C (Sigma 3-18K Centrifuge with a 19,776 H angle rotor, Osterode, Germany). The supernatant, which is free of nanoparticles, was removed and reserved for subsequent recycling of the ionic liquid. An equal volume of fresh methanol was added to the vial, and the white precipitate was suspended by vigorous stirring in a vortex mixer for 2 min and 5 min of ultrasonication with a Branson 450D sonicator (Emmerson Ultrasonic Corporation, Danbury, CT, USA). The centrifugation step was repeated under the same conditions. The white precipitate was subjected to successive rinses with ultrapure water to remove the remaining methanol and ionic liquid. The particles were lyophilized in an Edwards Modulyo 4K Freeze Dryer (Thermo Scientific, Waltham, MA, USA) for 72 h, at −55 • C and 0.5 mbar to obtain dry particles. The methanolic fractions were mixed before recovery of the ionic liquid by removing of the methanol/water on a BÜCHI RE-111 rotary evaporator (Flawil, Switzerland) at 80 • C and 80 mbar. The ionic liquids were kept in a desiccator until reuse.

Synthesis of Curc-SFNs
The drug was loaded into the SFNs by two different experimental procedures (physical adsorption, Curc-SFNs 1, and coprecipitation, Curc-SFNs 2) to obtain Curc-SFNs. The average yield of the process after lyophilization was of 95.8% and 90.3% for Curc-SFNs 2 and Curc-SFNs 1, respectively. All the experiments have been carried out at least in triplicate.

Drug Loading by Physical Adsorption
For loading of curcumin by physical adsorption, forty milliliters of a 1 mg/mL solution of curcumin in ethanol was used to resuspend 325 mg of SFNs obtained as described in Section 2.3. The suspension was ultrasonicated for 5 min using 30% amplitude with 15 s ON and 15 s OFF pulses and gently stirred at 30 rpm in a MX-RD-ProAnalog Tube Rotator (Scilogex, Rocky Hill, CT, USA) for 24 h. Then, Curc-SFNs 1 were collected by centrifugation at 13,400 rpm for 15 min. Next, the nanoparticles were washed with water to remove the remaining ethanol. The amount of the loaded drug in the SFNs was determined by an indirect method measuring UV-absorbance of curcumin (421 nm) in the supernatants of the centrifugation (in ethanol and in water) and the initial 1 mg/mL solution.
Drug loading content (DLC) and entrapment efficiency (EE) of the Curc-SFNs 1 obtained were calculated according to the following expressions: Weight of the drug in nanoparticles Weight of the nanoparticles × 100 (1) EE = Weight of the drug in nanoparticles Weight of the feeding drugs × 100

Drug Loading by Coprecipitation
In the case of coprecipitation, the curcumin was loaded during the nanoparticles synthesis process. Briefly, 25 mg of curcumin was dissolved in 3 mL of 0.1 M NaOH solution and the resulting solution was immediately dissolved in 5 g of SIL solution (10 wt %). The curcumin-SIL solution was heated to 60 • C and sprayed onto 100 mL of ethanol using the same equipment as described in Section 2.3. The orange suspension was stirred for 2 h before being centrifuged at 13,400 rpm for 15 min, at 4 • C. In this case, three washes with water were carried out to remove the ionic liquid. Lyophilization was carried out under the experimental conditions described in Section 2.3.
The characteristic parameters (DLC and EE) of the Curc-SFNs 2 obtained were also indirectly calculated by the above expressions from the measurements of UV-Vis absorbance of curcumin.

Curcumin Release from Curc-SFNs
Curcumin release studies were carried out with Curc-SFNs 1 and Curc-SFNs 2. Twenty milligrams of Curc-SFNs were dispersed in 1 mL of PBS 1× (0.5% Tween 80) by ultrasonication. The samples were incubated at 37 • C in a tube rotator for 3 days to study curcumin release. At predetermined time points (0.5, 1, 1.5, 2, 3, 3.5, 4, 5, 25, 50 and 73 h), the samples were centrifuged at 13,400 rpm and the curcumin in the supernatant was monitored using UV-Vis spectrophotometry at 421 nm. The PBS solution was replaced with the same volume (1 mL) of fresh solution. The release experiments were carried out in triplicate for every sample. Experimental data were fitted using four release kinetic models found in the literature (zero order, first order, Higuchi and Ritger-Peppas) [67] to know the release mechanism of curcumin from the Curc-SFNs in PBS. The mean hydrodynamic diameter (Z-average), the Polidispersity Index (PdI), the Zeta Potential and the Electrophoretic Mobility were measured using a Zetasizer Nano ZSP instrument (Malvern Instruments Ltd., Worcestershire, UK) by DLS. All measurements were performed in purified water at 25 • C, a 173 • angle relative to the source and with a nanoparticle concentration of 0.66 mg/mL. The influence of the SFNs concentration on size distribution and Zeta Potential was evaluated and it was checked that both parameters were practically constant in the range from 0.30 to 40 mg/mL and they remained constant up to 3 months. The mean values of the measurements performed in triplicate are reported.

Field Emission Scanning Electron Microscopy (FESEM)
The morphology of the nanoparticles was examined by FESEM using an FEI Scios TM microscope (Thermo Scientific, Waltham, MA, USA). The sample was placed on a pedestal as powder and then coated with gold.

Transmission Electron Microscopy (TEM)
Sample preparation for TEM was carried out using 0.06 mg/mL nanoparticle suspensions in purified water and ultrasonification for 3 min with 30% amplitude. A drop of this suspension was placed on a 200 mesh copper grid coated with carbon. Once the drop was dried at room temperature, a drop of uranyl acetate was added and the copper grid was imaged in an FEI Tecnai TM 12 microscope (Thermo Scientific, Waltham, MA, USA) operated at an acceleration voltage of 120 kV.

Attenuated Total Reflectance Fourier Transformed Infrared Spectroscopy (ATR-FTIR)
ATR-FTIR analysis was performed to detect the possible structural changes of SF after loading with curcumin. Each spectrum was acquired on a Nicolet TM iS5 spectrometer (Thermo Scientific, Waltham, MA, USA), equipped with an iD3 ATR accessory (Thermo Scientific, Waltham, MA, USA) controlled by OMNIC Software Ver. 6.1.0.0038 (Waltham, MA, USA), measuring in absorbance mode with a resolution of 4 cm −1 , a spectral range of 4000-550 cm −1 , and 64 scans, using N-B strong apodization and mertz phase correction. Fluorescence was studied on a Synergy MX UV-Vis spectrometer (Bio Tek Instruments Inc., Winooski, VT, USA). The fluorescence excitation spectra of curcumin samples or Curc-SFNs were obtained using 96 well non-sterile Costar black stripwell plates (N • 3914) (Corning, NY, USA) by varying excitation from 300 to 510 nm (5 nm step) and recording emission intensity at 530 nm. Fluorescence emission spectra of curcumin samples or Curc-SFNs were obtained by exciting at 420 nm and monitoring emission from 450 to 700 nm (5 nm steps).

Free Radical Scavenging Activity of Curc-SFNs (DPPH Assay)
It is well known that substances which have radical scavenging activity such as curcumin induce color bleaching of a 2,2-diphenyl-1-picrylhydrazyl radical (DPPH·) solution, which can be measured spectrophotometrically [14]. To assess the DPPH· free radical scavenging capacity of the samples, the method of Blois [68] was used with slight modifications. Briefly, aliquots of 100 µL of the 1 mM (100 nmol) DPPH· stock solution in methanol (prepared daily and protected from light) were added to the Eppendorf ® vials containing a mixture of 800 µL of methanol and 100 µL of a water suspension of the samples at 1 mg/mL of curcumin or Curc-SFNs.
Vials were thoroughly mixed and kept in darkness for 30 min. In the same conditions, ascorbic acid solutions (0-300 nmol in methanol) were tested as positive controls. After 30 min of reaction at 25 • C in darkness, the suspensions were centrifuged in an Eppendorf Centrifuge 5415D (20 min, 16,100 g at room temperature) to remove particles in the supernatants that could affect the spectrophotometric measurements. The radical scavenging was evaluated by measuring the absorbance at 515 nm, by using a Synergy MX UV-Vis spectrometer (BioTek Instruments Inc., Winooski, VT, USA) adding 200 µL/well. The radical scavenging activity was presented as Ascorbic Acid Equivalents per milligram of Curc-SFNs (nmol AAE/mg Curc-SFNs). All tests were done in quintuplicate and results were expressed as mean ± standard deviation (SD).

In Vitro Cytotoxicity Studies
Three cell lines were studied: human hepatocellular carcinoma (Hep3B), human neuroblastoma (Kelly Cells) and human bone marrow-derived mesenchymal stem cells (hBMSCs). Hep3B cells were obtained from the American Tissue Culture Collection (ATCC, Manassas, VA, USA) and Kelly Cells were purchased from Sigma Aldrich (St. Louis, MO, USA). The hBMSCs were obtained as previously described in a paper with silk materials and curcumin of the Kaplan research group [44]. The culture media used were Roswell Park Memorial Institute (RPMI) for Hep3B and Kelly Cells and Dulbecco's modified Eagle's medium (DMEM) for hBMSCs. Both media were supplemented with 10% fetal bovine serum, 1% penicillin and streptomycin, 1% non-essential amino acids and 1 ng/mL basic fibroblast growth factor (human recombinant, Invitrogen), at 37 • C in a 5%-CO 2 incubator. Cells were culturally expanded using trypsin passaging at a 1:3 split ratio. The medium was changed twice a week. The size distribution was stable in cell culture media DMEM with 10% fetal bovine serum up to 72 h.
For cytotoxicity studies, the three cell lines were seeded onto 96-well tissue culture plates at a density of 1.0 × 10 4 cells/well (for Hep3B and hBMSCs) and 2.0 × 10 4 cells/well (for Kelly Cells) and left for 24 h, after which the growth medium was removed and replaced with the medium containing SFNs, Curc-SFNs 1 or Curc-SFNs 2. In each experiment, growth medium without nanoparticles was used as a control. After 24 h, the culture medium of each well was replaced with fresh medium containing SFNs or Curc-SFNs in a wide concentration range: from 1.62 to 6650 µg/mL. Initial suspensions of nanoparticles in ultrapure water were previously sonicated 3 min with 30% amplitude and then diluted to 10% (v/v) with the medium. After 48 h, the medium was removed and AlamarBlue ® assay (ThermoFisher Scientific, Waltham, MA, USA) was performed following the manufacturer's protocol. Then, post reaction medium aliquots (90 µL) were transferred to black 96-well plates and quantified for fluorescence intensity within a SpectraMax M2 plate reader (Molecular Devices, Sunnyvale, CA, USA) using an excitation wavelength of 560 nm and an emission wavelength of 590 nm. The data are presented as mean values ± SD calculated from at least three samples per condition.

Characterization of SFNs and Curc-SFNs
The SFNs, Curc-SFNs 1 and Curc-SFNs 2 were characterized by DLS to ascertain their hydrodynamic diameter (expressed as Z-average), PdI, Zeta Potential and Electrophoretic Mobility (Table 1). Figure 1 shows the size and the Zeta Potential distributions of the SFNs and the Curc-SFNs obtained by both methods. Table 1. Physical characterization of the silk fibroin nanoparticles (SFNs), curcumin-loaded silk fibroin nanoparticles synthesized by physical adsorption (Curc-SFNs 1) and curcumin-loaded silk fibroin nanoparticles synthesized by coprecipitation (Curc-SFNs 2). of 590 nm. The data are presented as mean values ± SD calculated from at least three samples per condition.

Dynamic Light Scattering (DLS)
The SFNs, Curc-SFNs 1 and Curc-SFNs 2 were characterized by DLS to ascertain their hydrodynamic diameter (expressed as Z-average), PdI, Zeta Potential and Electrophoretic Mobility (Table 1). Figure 1 shows the size and the Zeta Potential distributions of the SFNs and the Curc-SFNs obtained by both methods. Table 1. Physical characterization of the silk fibroin nanoparticles (SFNs), curcumin-loaded silk fibroin nanoparticles synthesized by physical adsorption (Curc-SFNs 1) and curcumin-loaded silk fibroin nanoparticles synthesized by coprecipitation (Curc-SFNs 2).  The results showed that SFNs had a lower diameter while the PdI was similar (and lower than 0.15) for all of the samples, resulting in near monodisperse size distributions. The Zeta Potential of the SFNs and Curc-SFNs was in the same range and sufficiently negative, which is especially relevant if nanoparticles are to have high colloidal stability. The absolute Zeta Potential values of the SFNs was higher than the values found in the literature for SFNs obtained by other classical methods [50,54,69,70], which reflects the improvement in stability of the SFNs and hence of the Curc-SFNs obtained with this new procedure.

Transmission Electron Microscopy (TEM)
To detect any differences in the size and morphology of SFNs, Curc-SFNs 1 and Curc-SFNs 2, The results showed that SFNs had a lower diameter while the PdI was similar (and lower than 0.15) for all of the samples, resulting in near monodisperse size distributions. The Zeta Potential of the SFNs and Curc-SFNs was in the same range and sufficiently negative, which is especially relevant if nanoparticles are to have high colloidal stability. The absolute Zeta Potential values of the SFNs was higher than the values found in the literature for SFNs obtained by other classical methods [50,54,69,70], which reflects the improvement in stability of the SFNs and hence of the Curc-SFNs obtained with this new procedure. To detect any differences in the size and morphology of SFNs, Curc-SFNs 1 and Curc-SFNs 2, TEM was performed. The TEM observations (Figures 2 and S2) showed globular granules for SFNs and Curc-SFNs 1 with some aggregation and a slightly elongated shape for Curc-SFNs 2. The granules were about 60-95 nm in diameter. This diameter range was obtained from analysis of the TEM images using GIMP software (ver. 2.8, Mountain View, CA, USA). Unlike the more defined shape of the Curc-SFNs 1, the SFNs and Curc-SFNs 2 were uneven in appearance. and Curc-SFNs 1 with some aggregation and a slightly elongated shape for Curc-SFNs 2. The granules were about 60-95 nm in diameter. This diameter range was obtained from analysis of the TEM images using GIMP software (ver. 2.8, Mountain View, CA, USA). Unlike the more defined shape of the Curc-SFNs 1, the SFNs and Curc-SFNs 2 were uneven in appearance.

Field Emission Scanning Electron Microscopy (FESEM)
FESEM is a commonly used technique to study the morphology and size of a variety of nanoparticles. FESEM observation (Figure 3) of these nanoparticles provided similar results to the TEM analysis. Figure 3 shows nanospherical morphology for SFNs and Curc-SFNs 1 and again a more elongated shape for Curc-SFNs 2. The TEM and FESEM images revealed smaller diameters compared with the DLS results, probably due to swelling of the particles in the water solution, as the DLS measurements were performed with suspensions of the particles in water, while the TEM and FESEM samples were dried. This difference has also been observed by other authors [53,69].

Field Emission Scanning Electron Microscopy (FESEM)
FESEM is a commonly used technique to study the morphology and size of a variety of nanoparticles. FESEM observation (Figure 3) of these nanoparticles provided similar results to the TEM analysis. Figure 3 shows nanospherical morphology for SFNs and Curc-SFNs 1 and again a more elongated shape for Curc-SFNs 2. The TEM and FESEM images revealed smaller diameters compared with the DLS results, probably due to swelling of the particles in the water solution, as the DLS measurements were performed with suspensions of the particles in water, while the TEM and FESEM samples were dried. This difference has also been observed by other authors [53,69].

Attenuated Total Reflectance Fourier Transformed Infrared Spectroscopy
Information about the secondary structure of the proteins can be obtained from the position of the infrared signals and their relative intensity, as well as identifying specific molecules by comparison of their spectra with the databases [69]. In the present study, we used spectroscopy to verify that nanoparticles retained the β-sheet structure after the curcumin loading process. The Curc-SFNs spectra were coincident with the spectrum of nanoparticles prepared by the method described by Zhang et al. [71] by dissolving SF in the Ajisawa solvent system [65] ( Figure S3). The spectrum of the Curc-SFNs showed the characteristic β-sheet signals, such as 1624-1626 cm −1 (Amide I, C=O Stretching), 1517-1521 cm −1 (Amide II, N-H Bending) and 1232 cm −1 (Amide III, C-H Stretching), as in the literature [69,71,72], confirming that β-sheet in the SFNs was not modified after the curcumin loading processes, either by coprecipitation or by adsorption. The main signals in the spectrum of curcumin were 1627, 1603, 1507, 1428, 1283, 1233, 1205, 1154, and 1114 cm −1 , which match those previously described in the literature [73] (Figure S3). In the Curc-SFNs spectra, the intense absorbance of the SF signals, in addition to a much larger mass ratio, masked the curcumin signals.

UV-Vis Absorbance Spectroscopy
Curcumin is a tautomeric compound that absorbs light in the visible range and displays a yellow color when dissolved in organic solvents such as chloroform, dimethylsulfoxide, ethanol, methanol or acetone. Curcumin is practically insoluble in aqueous media at neutral pH [74]. The UV-Vis spectrum of Curc-SFNs dispersed in ultrapure water had a peak with a λmax = 435 nm, similar to free curcumin in low polar solvents such as octanol (λmax = 430 nm). Curcumin dissolved in water had a poorly resolved wide band with λmax~450 nm [37,74,75] (Figure S4). As the photochemical properties of curcumin depend strongly on the microenvironment [76], this hypsochromic shift of the maximum absorption peak confirmed that the curcumin was in a more hydrophobic environment, or less exposed to the solvent, than in the free state.

Attenuated Total Reflectance Fourier Transformed Infrared Spectroscopy
Information about the secondary structure of the proteins can be obtained from the position of the infrared signals and their relative intensity, as well as identifying specific molecules by comparison of their spectra with the databases [69]. In the present study, we used spectroscopy to verify that nanoparticles retained the β-sheet structure after the curcumin loading process. The Curc-SFNs spectra were coincident with the spectrum of nanoparticles prepared by the method described by Zhang et al. [71] by dissolving SF in the Ajisawa solvent system [65] ( Figure S3). The spectrum of the Curc-SFNs showed the characteristic β-sheet signals, such as 1624-1626 cm −1 (Amide I, C=O Stretching), 1517-1521 cm −1 (Amide II, N-H Bending) and 1232 cm −1 (Amide III, C-H Stretching), as in the literature [69,71,72], confirming that β-sheet in the SFNs was not modified after the curcumin loading processes, either by coprecipitation or by adsorption. The main signals in the spectrum of curcumin were 1627, 1603, 1507, 1428, 1283, 1233, 1205, 1154, and 1114 cm −1 , which match those previously described in the literature [73] (Figure S3). In the Curc-SFNs spectra, the intense absorbance of the SF signals, in addition to a much larger mass ratio, masked the curcumin signals.

UV-Vis Absorbance Spectroscopy
Curcumin is a tautomeric compound that absorbs light in the visible range and displays a yellow color when dissolved in organic solvents such as chloroform, dimethylsulfoxide, ethanol, methanol or acetone. Curcumin is practically insoluble in aqueous media at neutral pH [74]. The UV-Vis spectrum of Curc-SFNs dispersed in ultrapure water had a peak with a λ max = 435 nm, similar to free curcumin in low polar solvents such as octanol (λ max = 430 nm). Curcumin dissolved in water had a poorly resolved wide band with λ max~4 50 nm [37,74,75] (Figure S4). As the photochemical properties of curcumin depend strongly on the microenvironment [76], this hypsochromic shift of the maximum absorption peak confirmed that the curcumin was in a more hydrophobic environment, or less exposed to the solvent, than in the free state.

Fluorescence Spectroscopy
While significant curcumin fluorescence was not recorded in water, the nanoparticles loaded both by adsorption or by coprecipitation had similar fluorescence spectra, with a single peak for both excitation and emission Figure S5). This peak was similar to those found in the literature [33]. The maximum of the excitation spectrum appeared at 450 nm in both cases, although Curc-SFNs 1 had higher fluorescence intensity at the same particle concentration. The maximum emission spectrum of the nanoparticles was measured at 535 nm in both cases, the spectrum of Curc-SFNs 1 showed greater intensity. Both facts agree with the higher DLC of the nanoparticles. Figure S6b shows, from left to right, the fluorescence at 365 nm of free curcumin, Curc-SFNs 1 and Curc-SFNs 2 and SFNs at three different concentrations of nanoparticles: 10, 1 and 0.1 mg/mL. The only fluorescent samples were the Curc-SFNs obtained in this work. The different intensity of the fluorescence between the three cuvettes of each sample was due to the different concentration of the nanoparticle suspensions in water. Figure S6a shows the same samples but imaged with white light.

Drug Loading
Drug loading in three independent coprecipitation and physical adsorption experiments gave the results shown in Table 2. The DLC in the physical adsorption experiments was higher than in the coprecipitation assays, mainly because the initial curcumin/SF ratio (weight) was much lower in the latter. However, the EE had approximately the same value of around 50%. Both parameters (DLC and EE) were generally similar or higher than those obtained in similar studies using SFNs as drug vehicles [53,54]. To the best of our knowledge, only a recent study of Perteghella et al. [77] reached DLC values higher than 30% for Curc-SFNs. In this case, they were synthesized by acetone desolvation method. Differences in DLC values could be due to the synthesis technique. Table 2. Drug loading and encapsulation efficiency of the curcumin-loaded silk fibroin nanoparticles synthesized by physical adsorption (Curc-SFNs 1) and curcumin-loaded silk fibroin nanoparticles synthesized by coprecipitation (Curc-SFNs 2).

Curcumin Release from Curc-SFNs
The release behavior of the two types of Curc-SFNs (Curc-SFNs 1 and Curc-SFNs 2) was evaluated. The studies were carried out at 37 • C in PBS 1× (with 0.5% Tween 80 due to the poor solubility of curcumin in PBS). A notable difference in the release rate was seen between the two Curc-SFNs. For both types of nanoparticles (Figure 4), burst release was observed in the first 5 h, becoming plateau after 24 h. Curcumin release was higher for the Curc-SFNs 1, reaching a maximum value of approximately 35% of the loaded drug, while in the case of the Curc-SFNs 2, the maximum value was around 10% (Figure 4). This difference in the release of the two Curc-SFNs was probably due to the lower drug loading of the Curc-SFNs 2. In addition, in the authors' opinion, it might also be because Curc-SFNs 2, which are synthesized by coprecipitation, probably have the drug encapsulated inside the nanoparticle, making more difficult its release. On the contrary, in the case of Curc-SFNs 1, the drug is mainly adsorbed onto the surface so the release could be easier. Curcumin release profiles (from 0.5 to 5 h) have been fitted to several release models: zero order, first order, Higuchi and Ritger-Peppas. Table 3 shows the fitting equations obtained and the goodness of the fittings. We found that curcumin release profiles of Curc-SFNs 1 and Curc-SFNs 2 can be better agreed with Higuchi equation, although all fittings are quite good for both types of Curc-SFNs. Higuchi equation is based on Fick's Law where the release occurs by the diffusion of drugs within the delivery system [78]. This equation is a conceptually simple model but it is valid only under specific assumptions. It is known that, under some experimental conditions (swelling, glassy/rubbery transitions, dissolution or concentration-dependent diffusion), the release mechanism can deviate from Fickian diffusion following an Anomalous transport (non-Fickian) [79]. Therefore, when the drug release mechanism is not well known or more than one type of release phenomena is involved, a more generic equation is necessary [78]. In our case, to know the release mechanisms of curcumin from Curc-SFNs, Ritger-Peppas equation [80] has been used because of swellable silk fibroin matrix in PBS. Ritger and Peppas claimed that the mechanistic limits of the diffusional exponent, n, are dependent on the geometry (film, cylinder, and sphere) of the associated release device (Table 4) [80]. For Curc-SFNs 1 and Curc-SFNs 2, n, which is indicative of the transport mechanism, has a value of 0.80 ± 0.05 and 0.66 ± 0.04, respectively, indicating that the curcumin release from Curc-SFNs (assuming spheres) is controlled by Anomalous (non-Fickian) transport in both cases. This means that the curcumin release from Curc-SFNs is not a pure diffusional (Fickian) mechanism but involves a relaxational or convection mechanism which is usually associated with a major state or phase change [79]. The same drug release mechanism was found in the literature for polymeric microspheres loaded with simvastatin with DLC values similar to those presented in this work [67] and for polymeric nanoparticles containing diazepam [81]. Table 3. Release models of curcumin-loaded silk fibroin nanoparticles synthesized by physical adsorption (Curc-SFNs 1) and curcumin-loaded silk fibroin nanoparticles synthesized by coprecipitation (Curc-SFNs 2).

Release Model
Curc-SFNs 1 Curc-SFNs 2 Zero order y = 6.36x + 2.92 y = 1.89x + 1.  Curcumin release profiles (from 0.5 to 5 h) have been fitted to several release models: zero order, first order, Higuchi and Ritger-Peppas. Table 3 shows the fitting equations obtained and the goodness of the fittings. We found that curcumin release profiles of Curc-SFNs 1 and Curc-SFNs 2 can be better agreed with Higuchi equation, although all fittings are quite good for both types of Curc-SFNs. Higuchi equation is based on Fick's Law where the release occurs by the diffusion of drugs within the delivery system [78]. This equation is a conceptually simple model but it is valid only under specific assumptions. It is known that, under some experimental conditions (swelling, glassy/rubbery transitions, dissolution or concentration-dependent diffusion), the release mechanism can deviate from Fickian diffusion following an Anomalous transport (non-Fickian) [79]. Therefore, when the drug release mechanism is not well known or more than one type of release phenomena is involved, a more generic equation is necessary [78]. In our case, to know the release mechanisms of curcumin from Curc-SFNs, Ritger-Peppas equation [80] has been used because of swellable silk fibroin matrix in PBS. Ritger and Peppas claimed that the mechanistic limits of the diffusional exponent, n, are dependent on the geometry (film, cylinder, and sphere) of the associated release device (Table 4) [80]. For Curc-SFNs 1 and Curc-SFNs 2, n, which is indicative of the transport mechanism, has a value of 0.80 ± 0.05 and 0.66 ± 0.04, respectively, indicating that the curcumin release from Curc-SFNs (assuming spheres) is controlled by Anomalous (non-Fickian) transport in both cases. This means that the curcumin release from Curc-SFNs is not a pure diffusional (Fickian) mechanism but involves a relaxational or convection mechanism which is usually associated with a major state or phase change [79]. The same drug release mechanism was found in the literature for polymeric microspheres loaded with simvastatin with DLC values similar to those presented in this work [67] and for polymeric nanoparticles containing diazepam [81]. Table 3. Release models of curcumin-loaded silk fibroin nanoparticles synthesized by physical adsorption (Curc-SFNs 1) and curcumin-loaded silk fibroin nanoparticles synthesized by coprecipitation (Curc-SFNs 2).

Release Model
Curc  Table 4. Diffusional exponent n of the Ritger-Peppas equation and drug release mechanism from polymeric controlled delivery system for different geometries [82]. The DPPH· assay has been previously used to test the free radical scavenging capacity of curcumin and curcumin-loaded nanocarriers [14,82]. In its radical form, DPPH· strongly absorbs at 515 nm, but its absorption decreases after reaction with an antioxidant. The radical scavenging activity of both free curcumin and curcumin-loaded nanoparticles Curc-SFNs (Curc-SFNs 1 and Curc-SFNs 2) have been evaluated and compared to the equivalents of an antioxidant standard needed to obtain the same absorption reduction in the same reaction conditions. It should be noted that two equivalents of ascorbic acid react with one equivalent of DPPH· but Curcumin interacts in the ratio 1:1 [70]. Thus, the Ascorbic Acid Equivalents per milligram of loaded nanoparticles in aqueous solutions obtained for the curcumin loaded nanoparticles varied from 8.15 ± 1.17 nmol AAE/mg Curc-SFNs 1 to 4.93 ± 0.92 nmol AAE/mg Curc-SFNs 2, while the empty nanoparticles also exert a small effect (1.1 ± 0.1 nmol AA/mg SFNs). Silk fibroin contain about a 4% of aromatic residues that present radical scavenging activity as previously reported [59,83].

Thin Film
These results agree with the DLC of curcumin in nanoparticles, previously described in Drug Loading Section, and remarks that curcumin retain the full radical scavenging activity after being adsorbed onto SFNs in the assay conditions. This antioxidant protective effect of SFNs has been previously described with other sensitive phenolic compounds as Resveratrol [52] or Quercetin [59].

In Vitro Cytotoxicity Studies
The cytotoxicity of SFNs, Curc-SFNs 1 and Curc-SFNs 2 were evaluated after 48 h of exposure using AlamarBlue ® in three cell lines: Hep3B cells, Kelly cells and hBMSCs. We found similar trends for the tumor cell lines (Hep3B cells and Kelly cells) with regard to nanoparticle cytotoxicity (Figure 5a,b). After 48 h of incubation with several concentrations of nanoparticles, the results showed that cell viability slowly decreased with an increased concentration of nanoparticles. In addition, Curc-SFNs were more cytotoxic than SFNs, while the Curc-SFNs 1 were more cytotoxic than Curc-SFNs 2. These findings suggest that nanoparticle cytotoxicity is dependent on drug concentration because the Curc-SFNs 1 had a higher DLC than Curc-SFNs 2 ( Table 2). However, in the case of healthy cells (hBMSCs), no cytotoxicity was observed, even after exposure to the most heavily drug loaded-SFNs (Figure 5c). Similar results were obtained by Chang et al. [84], who used spherical amphiphilic nanoparticles as a curcumin carrier. These authors found that the curcumin loaded-amphiphilic nanoparticles had significant selective cytotoxicity against MG-63 osteosarcoma cells compared with normal osteoblasts. The fact that curcumin was non-toxic to healthy cells has been shown in several previous works [85,86], which established the potential of curcumin as a chemopreventive and chemotherapeutic agent in malignancy. The Curc-SFNs obtained by the two different procedures were cytotoxic for the three cell lines in the order: Kelly cells > Hep3B cells > hBMSCs.
In cancer cells, metabolic pathways are reprogrammed to satisfy tumor cell proliferation and survival requirements. In these cells, glycolysis and glutaminolysis are strongly increased. These metabolic processes and the role of mitochondria in supporting tumor cell metabolism are probably the reason of the different behavior between Hep3B, Kelly Cells and hBMSCs. Syng-ai et al. [86] showed that curcumin induce apoptosis in human breast carcinoma cell lines as well as in human hepatoma cells but failed to do so in normal rat hepatocyte primary cultures. Their results indicate that Glutathione (GSH, also known as γ-L-glutamyl-L-cysteineglycine) plays a vital role in the sensitivity of these cell lines to curcumin. Depletion of GSH further sensitized the cells to curcumin effects, and the cell death is caused by the generation of reactive oxygen species. Curcumin also down-regulated the expression of bcl-2 protein in tumor cells, which may be responsible for making them vulnerable to apoptotic death. down-regulated the expression of bcl-2 protein in tumor cells, which may be responsible for making them vulnerable to apoptotic death.

Conclusions
Curc-SFNs were successfully synthesized by two environmentally friendly procedures using ILs and high-power ultrasound to dissolve the SF. High DLC and EE values were obtained in both cases compared with those in the literature, and curcumin release was influenced by the synthesis method of the Curc-SFNs. The SFNs and the Curc-SFNs obtained showed a narrow size distribution, with a hydrodynamic diameter of <175 nm, and high Zeta Potential (in absolute terms), which make them excellent nanocarriers for use in therapeutic treatments. The nanospherical morphology was confirmed by FESEM and TEM, and the curcumin was fluorescent when encapsulated or adsorbed on the SFNs but not in its free state. The antioxidant activity against DPPH· showed that the curcumin loaded in Curc-SFNs retains full antioxidant activity. The Curc-SFNs obtained in this paper enhanced the antitumor activity of curcumin towards the two different tumor cell lines studied, while the viability of the healthy cells did not decrease. This work broadens the possibility of using these SFNs, which have been synthesized by an industrial process, as future systems for other drugs of hydrophilic or hydrophobic nature, such as curcumin.

Conclusions
Curc-SFNs were successfully synthesized by two environmentally friendly procedures using ILs and high-power ultrasound to dissolve the SF. High DLC and EE values were obtained in both cases compared with those in the literature, and curcumin release was influenced by the synthesis method of the Curc-SFNs. The SFNs and the Curc-SFNs obtained showed a narrow size distribution, with a hydrodynamic diameter of <175 nm, and high Zeta Potential (in absolute terms), which make them excellent nanocarriers for use in therapeutic treatments. The nanospherical morphology was confirmed by FESEM and TEM, and the curcumin was fluorescent when encapsulated or adsorbed on the SFNs but not in its free state. The antioxidant activity against DPPH· showed that the curcumin loaded in Curc-SFNs retains full antioxidant activity. The Curc-SFNs obtained in this paper enhanced the antitumor activity of curcumin towards the two different tumor cell lines studied, while the viability of the healthy cells did not decrease. This work broadens the possibility of using these SFNs, which have been synthesized by an industrial process, as future systems for other drugs of hydrophilic or hydrophobic nature, such as curcumin.