Paclitaxel-loaded Folate-targeted Albumin-Alginate Nanoparticles Crosslinked with Ethylenediamine. Synthesis and In Vitro Characterization.

Among the different ways to reduce the secondary effects of antineoplastic drugs in cancer treatment, the use of nanoparticles has demonstrated good results due to the protection of the drug and the possibility of releasing compounds to a specific therapeutic target. The α-isoform of folate receptor (FR) is overexpressed on a significant number of human cancers; therefore, folate-targeted crosslinked nanoparticles based on BSA and alginate mixtures and loaded with paclitaxel (PTX) have been prepared to maximizing the proven antineoplastic activity of the drug against solid tumors. Nanometric-range sized particles (169 ± 28nm - 296 ± 57nm), with negative Z-potential values (between -0.12 ± 0.04 and -94.1± 0.4), were synthesized. The loaded PTX (2.63±0.19 - 3.56 ±0.13 µg PTX/mg Np) was sustainably released along 23 and 27h. Three cell lines (MCF-7, MDA-MB-231 and HeLa) were selected to test the efficacy of the folate-targeted PTX-loaded BSA/ALG nanocarriers. The presence of FR on cell membrane led to a significant larger uptake of BSA/ALG-Fol nanoparticles regarding to the equivalent nanoparticles without folic acid on its surface. The cell viability results demonstrated a cytocompatibility of unloaded nanoparticle-Fol and a gradual decrease in cell viability after treatment with PTX-loaded nanoparticles-Fol due to the sustainable PTX release.


Introduction
Targeted treatments and personalized medicine are two of the axes of current oncology. They refer to the design of drugs that respond specifically to some genetic characteristics of each patient's cancer. More efficiency is pursued with fewer adverse effects, adapting the procedures to the characteristics of each patient. Numerous studies have focused on the develop of targeted drug delivery systems [1] for cancer treatment in order to reduce side effects due to the unspecific effect of the anticancer drug on healthy cells. Nanocarriers based on natural polymers have the advantage of being a priori biocompatible and also biodegradable. Among biological polymers, polysaccharides have been used for therapeutic delivery systems [2] due to their non-toxic and non-reactogenic properties, as well as their physicochemical properties that allow chemical modifications and, therefore, an easy preparation of nanocarriers. Their reactive groups, such as hydroxyl, carboxyl or amine, can attach different molecules and so introducing new physicochemical characteristics [3]. Anionic polysaccharides have an important role in the preparation of drug delivery systems [4], they have extensively contributed to the development of several types of nanocarriers for cancer treatment and diagnosis. Among anionic polysaccharides, alginate, whose structure is based on a backbone of [1][2][3][4] linked β-D-mannuronic acid (M units) and α-L-guluronic acid (G units), have been used for preparing different nanocarriers for biomedical applications in the last years [5]; alginate has been declared safe by FDA [6] for application in humans as dental impression materials or wound dressing, and their hydrogels are the most assayed material for bone tissue engineering and bioprinting since they provide an appropriate niche for cell loading [7]. On the other hand, there is a significant interest in developing anticancer drug carriers based on serum albumin [8] due to the therapeutic efficacy of Abraxane®, an albumin-bound form of paclitaxel, that rapidly dissociate in serum losing some benefits of nanoformulations; some studies have indicated that Abraxane® induced the overexpression of P-glycoprotein and did not allow overcome the common small molecule drug resistance problem mediated by P-gp [9]; in addition, when PTX-loaded albumin nanoparticles crosslinked and non-crosslinked were compared, differences in pharmacokinetics were observed, due to the different physiological way they have of delivering the drug to the tumor [10]. In an attempt to improve the characteristics of nanocarriers, numerous studies have focused on the synthesis of nanosystems formed by proteins and polysaccharides; in this way, core-shell microcapsules based on BSA gel with polyelectrolyte complex multilayer shell of hyaluronic acid and chitosan, encapsulating Sorafenib, have been designed for hepatocellular cancer therapy [11], and also ionic crosslinked nanoparticles based on chitosan and BSA were developed as carriers for doxorubicin, showing their biocompatibility after intravenous administration in a Wistar rat model [12]. Nanocarriers can reach the internal part of solid tumors taking advantage of their angiogenesis, but the uptake of these nanosistems improves significantly if they are targeted to overexpressed receptors on the surface of cancer cells. Among those receptors [13], the axis folic acid/folate receptor is one of the most important ligand-receptor interactions used to target cancer. Folic acid reaches normal tissues through two carriers, the reduced folate carrier (RFC) and the proton-coupled folate transporter, but any of them is able to bind folate conjugates (such as folate drugs or folate nanocarriers). Folate-bounded nanocarriers are internalized by the membrane-bound folate receptor (FR), which has a high affinity by folate (Kd ∼ 0.1-1 nmol/L) although is expressed on few cell types. FR-α isoform is overexpressed on 40% of human cancers (such as ovary, kidney, breast, and liver tumors) where it is completely accessible to FAbounded systems, unlike what happens in healthy cells, where FR-α isoform is detected in the apical membrane of some epithelial cells, where folate conjugates do not have access. This fact makes folate-targeted nanosystems [14,15] very selective for cancer cells.
In this study folate-targeted crosslinked nanoparticles based on BSA and alginate mixtures have been prepared and loaded with paclitaxel (PTX) in order to maximizing the proven antineoplastic activity of the drug against many solid tumors [16]. These nanoparticles were also in vitro characterized to evaluate their application in targeted tumor therapy.

Preparation of BSA/ALG nanoparticles
Folate-targeted nanoparticles (Nps-Fol) based on bovine serum albumin (BSA) and alginate (ALG) mixtures have been prepared and stabilized by amide bonds using ethylenediamine. BSA 5% (w/v) and ALG 1% (w/v) aqueous solutions were prepared in the presence of EDAC 50 mM (Sigma-Aldrich). Then, several blends of BSA/ALG were obtained by mixing different volumes of the solutions prepared previously (Supplementary data  Table S1). Ethylenediamine (ED) was added to each blend according to 2:1 ratio (polymer: ED, w/w) under intense stirring conditions (magnetic stirrer), and pH was then adjusted to 3 with HCl to get amide bond formation. Stirring was maintained during 2h and blends were finally centrifuged (10000 rpm, 5min; Sigma SM202 centrifuge). Pellets were freeze dried (Heto PowerDry LL1500 Freeze Dryer, Thermo Electro Corporation) for 24 h at -110°C.
The conjugation of folate to nanoparticles was carried out by a carbodimiide reaction with NHS-folate. The N-Hydroxysuccinimide ester of folic acid (NHS-folate) was prepared according to the method developed by Lee and co-workers [17]. Folic acid (2.5 g) was dissolved in 50 ml of dry dimethyl sulfoxide (DMSO) plus 1.25 ml of triethylamine and reacted with N-hydroxysuccinimide (NHS) (1.3 g) in the presence of dicyclohexylcarbodiimide (2.35 g) overnight at room temperature. The by-product, dicyclohexylurea, was removed by filtration. The DMSO solution was then freeze-dried, and NHS-folate was precipitated in diethylether. The final product, NHS-folate, was washed several times with anhydrous ether, dried under vacuum, and yielded a yellow powder.
The conjugation of folate to nanoparticles was carried out following the method developed by Zhang and co-workers [18]: NHS-folate (50 mg/ml) was dissolved in 3 ml DMSO and added slowly to 6 ml of the stirred nanoparticle suspension (10 mg NP/ml, pH adjusted to 10 using 1 M carbonate/bicarbonate buffer). After stirring for 3 h at room temperature, the reaction mixture was centrifuged (10,000 rpm, 10 min) in order to separate the folate conjugated nanoparticles from unreacted folic acid and other by-products and then washed with DMSO (10,000 rpm, 10 min). Finally, folate-conjugated nanoparticles (30BSA/70ALG-Fol, 40BSA/60ALG-Fol; 50BSA/50ALG-Fol) were freeze dried for 24 h at −110°C.

Characterization of nanoparticles
Composition of BSA/AlG nanoparticles Composition of nanoparticles was determined according to the method developed by Martinez and co-workers [19]. The composition of BSA/ALG nanoparticles was studied measuring the concentration of BSA and ALG in the supernatant obtained after the centrifugation of the samples in the synthesis process. The difference between the initial concentration of each polymer and their concentration in the supernatant was determined, to reveal the proportion of each polymer in nanoparticle formation. Concentration of algi-nate in the supernatant was determined by measuring the absorbance of the colored products obtained after a hydrolysis reaction of alginate. The hydrolysis of alginate was carried out in extreme conditions (10N HCl, 5 h, 100ºC) and, after this reaction, samples were neutralized with 10N NaOH and cooled. Absorbance of the samples was measured at 277.6 nm. A standard curve of alginate (0.5-4 mg/ml) was prepared. Concentration of BSA in the supernatant was determined using Bradford's method [20]. The absorbance of blue colored products of this reaction can be measured at 595 nm. To avoid interactions between alginate and Bradford reagent, BSA in the supernatant was precipitated with 2M trichloroacetic acid (TCA). After the precipitation process, samples were centrifuged and the supernatant was removed. The pellet was suspended in 1ml 6M urea (urea solution in 2N HCl). A standard curve of BSA (0.125-0.5 mg/ml) was prepared.

Thermogravimetric Analysis (TGA)
TGA curves of BSA/ALG nanoparticles, as well as raw materials, were obtained using a Mettler Toledo thermal analyzer (TGA-SDTA 851®, Switzerland). Samples (3 mg) were placed onto the balance, the temperature increased from 25 to 600 •C at 10•C min−1 in a nitrogen atmosphere (nitrogen flow rate of 60 cm 3 min −1 ) and the mass continuously recorded as a function of temperature.

Determination of Folate-conjugated to BSA/ALG nanoparticles
The amount of folate conjugated to the amine groups of albumin was determined by spectrophotometric analysis [18]. 2 mg folate-conjugated nanoparticles (Nps-Fol) were hydrolysed by trypsin (0.05 mg/mg NP) with stirring at 37°C for 2 h. After the digestion process, the quantification of folate-conjugation was performed by spectrophotometric measurement of its absorbance at 358 nm (folic acid ε=8643.5 M−1 cm−1).

Morphology, size and Z-potential of nanoparticles
The morphology and size of Nps and Nps-Fol were studied by Transmission Electron Microscopy (TEM) (JOEL JEM 1010 microscope from ICTS Centro Nacional de Microscopía Electrónica, UCM). Size and zeta potential measurements of Nps and Nps-Fol were performed in deionized distilled water using a backscattered quasi-elastic light scattering device (Zetatrac NPA. Data were analysed with Microtrac Flex Software.

Preparation of PTX-Folate-conjugated nanoparticles
The paclitaxel (PTX) load into Nps-Fol was carried out as follow: a suspension of 30 mg of NPs-Fol in 1 ml of the solution 1 mg PTX/ml ethanol was prepared and incubated for 22 h using a circular rotary stirrer in darkness at room temperature. Finally, the suspension was centrifuged (12,000 rpm, 15 min), the pellet was washed with water and nanoparticles were freeze-dried.

Estimation of PTX content
Two different methods were carried out in order to determine the amount of PTX loaded into Nps-Fol. One of the methods consisted of a direct extraction of the drug from PTXloaded Nps-Fol using ethanol: 3 mg of PTX-loaded Nps-Fol were suspended in 1 mL of ethanol under stirring conditions during 5 h. After this time, the suspension was centrifuged (13,000 rpm, 5 min) and the supernatant was collected. The samples were analysed by HPLC. The other method consisted of an enzymatic digestion of PTX-loaded Nps-Fol with trypsin and a subsequent extraction of the drug with an organic solvent: 2 mg of PTX-loaded nanoparticles was suspended in 2 mL of phosphate-buffered saline (PBS, pH 7.4), and 50 µg of trypsin was added to the solution. The digestion process was performed under orbital stirring (100 rpm) at 37 ºC for 24 h. After this time, PTX was extracted from PBS solution by the addition of 4 mL of dichloromethane. The tubes were mixed in a vortex mixer for 10 min and then centrifuged for 10 min at 10000 rpm. The organic phase was transferred to flask vials and evaporated to dryness. The resulting residue was reconstituted in ethanol and then quantified by HPLC technique.
The concentration of PTX was determined by high-performance liquid chromatography (HPLC) using a Spectra Physics HPLC system, Spectra 100 UV-Vis detector and a SP8800 pump. The mobile phase consisted of acetonitrile/water 60:40 (v/v) and was pumped through 25×4.6mm RP-Spherisorb ODS2 C18 column (5 mm particle size, Waters) at a flow rate of 1 mL/min. The output at 227 nm was monitored and the peak area of each sample generated by ChromQuestTM software. The calibration curve was generated using PTX solutions between 0.1 and 25 µg/mL in acetonitrile/water (60:40, v/v), prepared from a PTX solution in ethanol, and a good linear correlation (r 2 = 0.99) was obtained. All quantifications were performed in triplicate.

In vitro drug release studies
The release of PTX from Nps-Fol was carried out in centrifuge tubes. Drug-loaded Nps-Fol (3 mg) were suspended in PBS (4 mL pH 7.4, 37 •C) containing 0.1% (w/v) Tween 80® [21,22]. Release studies were carried out in dark at orbital shaking (100 rpm; Ecotron IN-FORS HT, Switzerland) for 96 h. Sink conditions were maintained during drug release experiments and all experimental conditions were performed in triplicate. At predetermined time intervals, the solutions were centrifuged at 10,000 rpm for 1 min and the supernatants collected for PTX concentration analysis. The Nps-Fol samples were dispersed in fresh PBS and placed back on the shaker. The PTX was extracted from the supernatants by DCM (1 mL) by stirring using a vortex mixer for 10 min at room temperature. The DCM layer collected was evaporated at room temperature for 24 h, and the dried PTX was dissolved in acetonitrile/water (1 mL; 60:40 (v/v)). The percentage of PTX released into the supernatant was determined by HPLC.
In order to analyze the drug delivery mechanism used by the NPs-Fol, the in vitro drug release data were fitted to various kinetic models. Higuchi, Korsemeyer-Peppas, zeroorder and first-order models were selected for this study, applying the following set of equations [23]:

Cellular uptake of nanoparticles
Nps and Nps-Fol (10 mg) were loaded with 6-coumarin (98%, Aldrich, Spain) by immersion in 1 mL of a 1 mg/mL 6-coumarin solution (dissolved in ethanol) for 24 h at room temperature in dark and under orbital shaking. Then, the suspension was centrifuged (7,000 rpm, 5 min), and the 6-coumarin-loaded NPs were washed twice with ethanol and freeze-dried for 24 h at −110 °C Cellular uptake of Nps and Nps-Fol was investigated in a monolayer of MCF7, MDA-MB-231 and HeLa cells. Cells were seeded in 96 well flat-bottom plates at 10,000 cells/well for MCF7 and MDA-MB-231, and 5,000 cells/well for HeLa. After 24 h, the medium was replaced with 100 µL of 1% FBS medium containing 6-coumarin-loaded Nps or Np-Fol and the plates were incubated in the 5% CO2 incubator at 37•C for 2, 4 and 24 h. The concentration of nanoparticles was the equivalent to obtain a concentration of 1 µM of drug, considering the PTX load of each nanoparticle composition. Thus, the concentration of 6-coumarin-loaded Nps and Nps-Fol was 0.24 mg/ml (composition 50BSA /50ALG) and 0.325 mg/ml (composition 30BSA/70ALG) respectively. As a positive control, cells were grown with 0.2 µg/mL of 6-coumarin in FBS medium, in the absence of nanoparticles. At each time interval, cells were washed twice with PBS to remove any uninternalized nanoparticles and lysed with 100 µL lysis reagent (PBS with 2% SDS v/v and 50 mM EDTA).
Internalization of nanoparticles was evaluated by fluorimetry at 488 nm, using a spectrophotometer (Varioskan, Thermo FisherScientific, Barcelona, Spain) [27]. Calibration curves (fluorescence versus concentration) were prepared with 6-coumarin (1ng/ml-1 µg/mL) dissolved in the lysis medium in order to calculate the amount of 6-coumarin incorporated into nanoparticles. All experimental conditions were performed in quintuplicate.

Cell viability
Cell viability was evaluated by using the MTT method. All experimental conditions were performed in quintuplicate. Each experiment was carried out in triplicate.
In preliminary experiments, increasing concentrations of PTX from 0.1 to 500 nM were tested in culture. The IC50 of PTX was determined in the three cell lines tested. According to that, the selected concentrations of PTX in cell viability studies were 2.5 and 7.5 nM for HeLa and MCF7 cell lines, and 7.5 and 30 nM for MDA-MB-231 cells.
Cells were seeded in 96-well flat-bottom plates at 5,000 cells/well in the case of MCF7 and MDA-MB-231, and at 2,500 cells/well in the case of HeLa. After 24 h, the medium was replaced with 100 µL medium with 1% FBS containing Nps-Fol, PTX-loaded Nps-Fol or the drug in solution. The amount of Nps-Fol was the needed to obtain a final concentration of 30, 7.5 and 2.5 nM of PTX, considering the PTX load of each nanoparticle composition; for this purpose, a nanoparticle suspension of 0.24 mg/ml of 50BSA/50ALG and 0.325 mg/ml of 30BSA/70ALG Nps-Fol, (equivalent to 1,000 nM PTX) was used. After 1, 3 and 6 days, 10 µL of MTT solution (5 mg/mL) were added to each well. After 2 h of incubation at 37ºC, 5% CO2, each well was replaced with 100 µL DMSO [28]. The cell viability was determined by measuring the absorbance at 570 nm using a spectrophotometer (Varioskan, Thermo Fisher Scientific, Barcelona, Spain). Results are expressed as the percentage survival in relation to untreated control cells.

Statistical analysis
Statistical comparisons were performed with unpaired Student's t-test. A value of p < 0.05 was considered significant.

Composition of Nanoparticles
Final nanoparticle composition was determined by the concentration of BSA and ALG in the supernatant collected after centrifugation, in the synthesis process. Thus, the contribution of each polymer within the final formulation of nanoparticles was calculated by the difference between the initial concentration and the final concentration after the synthesis protocol ( Table 1). The BSA/ALG ratio was calculated considering the concentration of each polymer incorporated into nanoparticles. These results demonstrated that not all the BSA nor ALG added in the initial solution of synthesis was incorporated in the final nanoparticles. Surprisingly, an increased proportion of BSA (w/w) in the initial composition was not translated into a higher incorporation of BSA into Nps. The thermal stability of raw materials and BSA/ALG nanoparticles was studied. The TGA first derivative of the materials is plotted in Figure S1 (Supplementary material). Nps showed different thermal profile from raw materials, which could indicate chemical modification. Furthermore, results revealed that the different ratio of polymers incorporated in each Np composition determine their differential thermal profiles. Thus, the thermal profile of 30BSA/70ALG Nps, whose BSA/ALG ratio is 1:1, showed two peaks at 217 and 317 ºC, whereas 40BSA/60ALG and 50BSA/50ALG Nps showed a single peak at 212 and 217ºC, respectively. Thus, in the case of the nanoparticles whose BSA/ALG ratio is 0.25:1 (40BSA/60ALG) and 0.5:1 (50BSA/50ALG) a pronounced peak was not observed at higher temperatures. In addition, results clearly showed that all the compositions synthesized showed a mass loss above 200ºC.
According to these results, the compositions 30BSA/70ALG-Fol and 50BSA/50ALG-Fol were finally chosen for further studies due to their greater content of folate incorporated.

Morphology, size and Z-potential of nanoparticles
The morphology of 30BSA/70ALG and 50BSA/50ALG Nps was analysed by TEM. Micrographs showed nanometric-range sized particles (<100 nm), with a spherical appearance. This morphology was maintained after folate conjugation (Figure 1). These nanoparticles were also analyzed using quasi-elastic light scattering to characterize them in terms of mean size and surface charge ( Table 2). The results showed that all synthesized systems had a nanometric size in the range of 169 ± 28nm and 296 ± 57nm and a polydispersity index between 1.2 ± 0.4 and 1.8 ± 0.4. Moreover, an increase in size was observed after the incorporation of folate and the loading of paclitaxel in both nanoparticle compositions. However, these differences in size were not statistically significant. Differences in nanoparticle size among TEM photographs and DLS data can be attributed to the different characteristics of these techniques. In this way, DLS measures are realized in water solution without stirring, and nanoparticles tends to stablish interactions and aggregate, increasing the observed size. Results of Z-potential analysis ( Table 2)  The negative value of this parameter was increased after the conjugation of folic acid and the incorporation of PTX in both formulations, those differences were statistically significant.

Estimation of PTX content in folate-conjugate nanoparticles
The estimation of PTX content in each formulation was determined by two methods: extraction with ethanol and enzymatic degradation using trypsin ( Table 3). Independently of the method, results showed statistically significant differences (p<0.05) in drug content comparing both compositions, showing a higher PTX load in 50BSA/50ALG-Fol nanoparticles. The cumulative release curve of PTX is represented in Figure 2. The maximum PTX release (82% and 88% of drug load) was achieved at 23 and 27 h from 30BSA/70ALG-Fol and 50BSA/50ALG-Fol, respectively. Drug release in the presence of Tween 80® occurred fast in the first 4 h (Table 4), followed by a slower release rate in both types of nanosystems.
In both stages of drug delivery, the release of PTX was faster from the 50BSA/50ALG-Fol composition.  Additionally, the PTX release was also evaluated in Tween 80® free medium at 74 h. In this experimental condition, only 16% and 23% of the PTX load was released from 50BSA/50ALG-Fol and 30BSA/70ALG-Fol, respectively.

Cellular uptake of nanoparticles
To evaluate the uptake of the synthesized nanosystems by the cells, three different human tumour cell lines, MCF-7, MDA-MB-231 and HeLa were selected. BSA/ALG-Fol nanoparticles and BSA/ALG nanoparticles were loaded with a fluorescent marker (6-coumarin) and added to cells in order to study their cellular uptake. Fluorescent signal into cells was quantified after 2, 4 and 24 h of incubation. Results are shown in Figure 3. As results showed, either targeted or not targeted systems were internalized in cells. However, folate-conjugated nanoparticles enhanced their uptake in the three cell lines. 50BSA/50ALG nanoparticles were less internalized than 30BSA/70ALG. Both unloaded nanoparticles (50BSA/50ALG-Fol and 30BSA/70ALG-Fol) were cytocompatible and no significant decrease in viability (cell survival between 91±6% and 89±3% for MCF7; 83±5% and 84±13% for MDA-MB-231; 100±3% and 100±12% for HeLa) was observed for the three tested cell lines at longer times of exposure.
PTX administered into 50BSA/50ALG-Fol and 30BSA/70ALG-Fol nanoparticles (Figure 4) caused more gradual decrease in cell viability than PTX in solution from the first day, at the highest drug concentration assayed for MCF7, MDA-MB-231 and HeLa cells.

Discussion
Paclitaxel (PTX) is one of the common drugs used for the treatment of different types of solid tumors [16]. PTX acts promoting microtubules polymerization by binding to β-subunit of tubulins, which causes mitotic arrest at G2/M phase and results in cell death by apoptosis pathway. Besides, PTX exhibits cytotoxic properties that lead to the inhibition of cellular growth. Both characteristics contribute to antitumor efficacy of PTX [29]. Since it is a hydrophobic drug, a vehicle is required to dissolve it, which causes numerous side effects, limiting its optimal clinical utility as an anticancer agent. For this reason, different nanocarriers have been prepared to improve the safety, efficacy and pharmacokinetic profile of PTX. Thus, in recent studies PTX has been encapsulated in nanoparticles [30,31], liposomes [32] and polymeric micelles [33]. On the other hand, folate receptor is overexpressed on many tumors [34]; the FRα isoform is the most widely expressed of all the FR isoforms [35] and is overexpressed (100-300 times more than in healthy cells) in a large number of cancers of epithelial origin. Besides, the high binding affinity (KD = 0.1-1 nM) [34] of FRα for folic acid and other oxidized folates has led to the development of folateconjugated nanosystems for targeting to cancer cells [37,38]. For these reasons, folic acid was selected to target tumor cells with folate-bounded drug-loaded BSA/ALG nanoparticles.

Composition, characterization and PTX release from folate-conjugated nanoparticles
Nanoparticles of different BSA/ALG composition, cross-linked with ethylenediamine (ED), were synthesized. BSA was better incorporated in 30BSA/70ALG (BSA:ALG 1:1) nanoparticles; meanwhile, alginate was similarly incorporated in all compositions, which can be attributed, at least in part, to the formation of amide bonds between ED and carboxylic groups of both ALG and BSA (each BSA molecule has 58 residues of Glu and 41 residues of Asp).
Covalent attachment of folic acid to the three systems was achieved. There was not a correlation among the amount of BSA in the nanoparticles and the extent of folate on their surface. Whereas 30BSA/70ALG-Fol and 50BSA/50ALG-Fol nanoparticles bound a similar a: Significant differences (p<0.05) regards HeLa control at day 1 b: Significant differences (p<0.05) regards HeLa control at day 3 c: Significant differences (p<0.05) regards HeLa control at day 6 amount of folic acid (FA), 40BSA/60ALG-Fol nanoparticles showed a significant lower amount of FA-attached. This fact could be related with the disposition of the amine groups of the side chain of amino acids in the nanoparticles; one BSA molecule has 60 residues of Lys and 26 residues of Arg, which can participate in FA attaching. Thus, the low amount of FA-attached was the main reason to reject 40BSA/60ALG-Fol nanoparticles for later experiments.
TEM micrographs showed spherical, nano-sized particles without significant differences between nanoparticles of different BSA/ALG composition or those FA-bounded. The size of the unloaded nanosystems, determined by DLS (Table 2), was larger, but statistically no significant, for 50BSA/50ALG-Fol (268±102nm) than for 30BSA/70ALG-Fol (189±81nm), which can be due to the larger proportion of alginate in those nanoparticles, which would induce a higher swelling of the systems. However, the value of the polydispersity index was the same for both type of nanoparticles, and it is like the values obtained for other polymeric systems [12,39].
The Z-potential value of all studied nanosystems was negative ( Table 2). The presence of alginate in 30BSA/70AGL and 50BSA/50ALG nanoparticles showed lightly negative Zpotential values. The FA-bounded to nanosystems significantly increased Z-potential values (-69.3 mV and -66.2 mV for 30BSA/70ALG-Fol and 50BSA/50ALG-Fol, respectively), which were very much larger than -30 mV, the Z-potential value (positive or negative value) that has been established as the minimum to obtain a physically stable nanosuspension and avoid nanoparticle aggregation [38]. Besides, when these nanosystems were loaded with PTX, Z-potential was of the same magnitude for PTX-loaded 30BSA/70ALG-Fol, and significantly larger (-94.1 mV) for PTX-loaded 50BSA/50ALG-Fol; this increase could be explained by the possible superficial distribution of part of the drug on the nanosystem. Thus, the low tendency to establish interactions among them, indicated by the Z-potential value of these unloaded and PTX-loaded FA-attached nanosystems, make them very suitable for later in vivo intravenous administration studies.
PTX-load of nanosystems-Fol was determined by extraction with ethanol and also after enzymatic digestion by trypsin of the nanoparticles (Table 3). Differences were not observed between the results obtained by both methods. Regardless of the method used, PTX-load of 50BSA/50ALG-Fol was larger (3.56 -3.28 µg PTX/mg Np) than that of 30BSA/70ALG-Fol (2.63 -2.42 µg PTX/mg Np). Interactions between PTX and nanosystems seem to be very similar in both types of nanoparticles since the value of PTX-load obtained by the two methods is almost the same; according to the expected results, in the case of a high interaction between PTX and components of the nanoparticles, the drugload determined by ethanol extraction would have been lower than that by enzymatic digestion [24], but the observed results were different. PTX release studies showed larger and a faster drug release from 50BSA/50ALG-Fol nanoparticles; PTX release took place in two stages from both type of nanosystems ( Figure 3). The first stage of release was during the first four hours (Table 4), and the release was faster from 50BSA/50ALG-Fol, which could be related to a part of the drug being located more superficially as indicated by the Z-potential values. A second stage of drug release was determined up to 27 hours, and also PTX release was faster from 50BSA/50ALG-Fol. In no case the total amount of PTX loaded was release, 88% of PTX-load was released from 50BSA/50ALG-Fol, and 82% of it from 30BSA/70ALG-Fol. In all these experiments, the release conditions were favored by the presence of Tween 80® in the medium; however, in the absence of the surfactant, an experimental condition closer to biological systems, a slower release occurred, which would make possible a significant amount of PTX loaded into these nanoparticles-Fol to be delivered to cancer cells in vivo.
In order to determine the predominant drug release mechanism from BSA/ALG-Fol nanoparticles, the obtained data were fitted to different physical models with a significant correlation coefficient (r 2 = 0.791-0.979) (Supplementary material; Table S2). According to the results, PTX release kinetics from 50BSA/50ALG-Fol nanoparticles fitted in the order of Korsemeyer-Peppas = Higuchi = First Order > Zero Order; PTX release from 30BSA/70ALG-Fol nanoparticles fitted in the order of Korsemeyer-Peppas> Higuchi > First Order > Zero Order. The Higuchi model indicates drug diffusion from the matrix, with no matrix dissolution and constant drug diffusivity [41]. The release of PTX from 30BSA/70ALG-Fol nanoparticles showed the lowest value of KH, which implies a slower release in comparison with 50BSA/50ALG-Fol nanoparticles. An equivalent behavior has been observed for the release of naproxen from cyclodextrin-based hydrogel matrices [42]. The release of PTX was favored by the swelling capability of the nanoparticles, which made the diffusion of the entrapped drug easier. In the case of the Korsemeyer-Peppas model, the nanosystems showed values of the release exponent (n) different from the standard value for declaring Fickian release behavior (n = 0.43; considering spherical shape) [43,44], it was 0.36 and 0.70 for 30BSA/70ALG-Fol and 50BSA/50ALG-Fol nanoparticles, respectively. Values of n parameter lower than 0.43 can be related to a dispersion of nanoparticle sizes. For spherical particles, when n value is between 0.43 and 0.85 an anomalous transport (non-Fickian; intermediate between Fickian and Case II transport, zero order) is considered, and then a mixed diffusion and chain relaxation mechanisms takes place, which happens for PTX release from 50BSA/50ALG-Fol; this nanoparticle composition presents an acceptable fitting to the first order kinetic model.

In vitro evaluation of folate-conjugated nanoparticles in tumor cell lines
The presence of FR in several human solid tumors, such as breast cancer, has been clearly demonstrated. Thus, these receptors provide an interesting therapeutic target for solid tumor treatment. PTX is widely used as a first-line treatment for patients with breast cancer [45]. Cancer cell lines of human breast adenocarcinoma MCF-7 and MDA-MB-231 express a low and high level of folate receptor, respectively [46]. However, whereas MCF-7 cells express estrogen receptors (ER), MDA-MB-231 cells are triple negative breast cancer (TNBC) cells, they do not express ER, progesterone receptor (PR), and amplification of HER-2/Neu [47]. Besides, among drugs used to treat stage IVB cervical cancer is PTX [48]; and HeLa cells are human cervical carcinoma cells, positive for FR, adequate to study the efficacy of these folate-targeted PTX-loaded BSA/ALG nanocarriers.
Therefore, these three different cellular lines expressing folate receptors (FR) were chosen for evaluating the effectiveness of these BSA/ALG-Fol nanoparticles. Different studies carried out by our group [24] and for other researches [49] have shown that MCF7, MDA-MB-231 and HeLa cells express different levels of FR on their surface, allowing they can be used as in vitro test of folate-targeted nanosystems efficacy. In general, good cytocompatibility values were obtained with the two composition of BSA/ALG-Fol nanoparticles for the three cell lines, although a slightly lower value was observed for MDA-MB-231 cells, which can be attributed to the larger amount of nanoparticles in the experiments with this cell line. Therefore, the uptake of unloaded nanoparticle-Fol observed in these cell lines seemed not to be harmful even after six days of incubation. The presence of FR on cell surface led to a larger uptake in these cell lines for both nanoparticle-Fol compositions in comparison with the corresponding nanoparticle composition without folic acid bounded on their surface. The uptake of Nps without FA on their surface revealed there was a nonspecific uptake mechanism that would act in addition to FR-mediated internalization. In this way, different studies indicate that negative-charged nanoparticles showed high cellular uptake due to electrostatic interactions established with some parts of cellular membrane [24,50], which would be in accordance with the negative values of Z-potential of these BSA/ALG nanoparticles. This nonspecific uptake was lower in the case of 50BSA/50ALG nanoparticles in the three cell lines studied, and so they would be less internalized into cells without FR on their surface, that is the case of no cancer cells.
PTX-loaded nanoparticles-Fol caused a gradual decrease in cell viability values when PTX was released from the both nanosystems, mainly at higher PTX load (7.5 nM PTX for MCF7 and HeLa; 30 nM for MDA-MB-231) at longer time of incubation. Significant differences in decreasing viability were not observed between PTX-loaded 50BSA/50ALG-Fol and 30BSA/70ALG-Fol nanoparticles in all cell lines studied. Although viability values obtained in the presence of PTX-loaded nanoparticles-Fol were similar to those of PTX in solution, it must be taking into account that PTX amount within the cells treated with PTXloaded nanosystems-Fol will be lower than the amount of drug available in cells treated with PTX solutions, in accordance with in vitro drug release studies. Once PTX-loaded nanoparticles-Fol were uptake into cells via folate receptor mediated endocytosis pathway, the unligate folate receptor may then recycle to the cell surface at a rate in the range of 0.5-5 hours [51], and it is expected that more nanoparticles were uptake into the cells; PTX release would take place mainly after the enzymatic degradation of nanoparticles, which would cause a delay in the cytotoxic action of PTX-loaded nanoparticles-Fol regards to PTX in solution.

Conclusions
These folate receptor-targeted PTX-loaded BSA/ALG nanoparticles have demonstrated good effectiveness against cancer cells that overexpress folate receptors on their surface, reducing cell survival in a cervical carcinoma cell line, a breast adenocarcinoma ER+ cell line and a TNBC cell line, which causes a very aggressive cancer type with a difficult treatment. Therefore, these folate-targeted PTX-loaded BSA/ALG nanoparticles are very promising nanosystems for further in vivo studies of biocompatibility and tumor grow suppression.