Novel Worm-like Micelles for Hydrochloride Doxorubicin Delivery: Preparation, Characterization, and In Vitro Evaluation

Doxorubicin hydrochloride (DOX) is one of the widely used antineoplastic agents in treating various cancers, yet it is always associated with the occurrence of adverse reactions that limit its clinical use. Currently, encapsulating DOX in micelles may represent a promising strategy to reduce toxicity and side effects of the drug. This study aimed to explore a novel acitretin-based surfactant (ACMeNa) with good solid stability to encapsulate DOX to form micelles (ACM-DOX). In this work, ACM-DOX micelles were prepared by a microfluidic method free of organic solvents. The characteristics of ACM-DOX micelles were assessed, including morphology, particle size, stability, entrapment efficiency, and drug loading. An in vitro cytotoxicity experiment of the micelles on MDA-MB-231 (a human breast cancer cell line) was also performed. The micelle formation mechanism suggested that the insoluble ACMeNa/DOX complex was formed by electrostatic interaction, and subsequently encapsulated by self-assembly into micelles. The designed ACM-DOX micelles had an average particle size of 19.4 ± 0.2 nm and a zeta potential of −43.7 ± 2.4 mV, with entrapment efficiency and drug loading efficiency of 92.4 ± 0.5% and 33.4 ± 0.3%, respectively. The ACM-DOX micelles had worm-like structures under a Cryo-transmission electron microscope and exhibited good stability within 8 hours after reconstitution and 4- to 32-fold dilution of its reconstituted solution. ACM-DOX micelles released 80% of DOX within 24 hours in a medium of pH = 5.0, and its drug profile can be described by a first-order model. Moreover, ACM-DOX micelles showed cytotoxicity against MDA-MB-231 in a dose-dependent manner, and displayed a higher antitumor activity as compared with free DOX, with IC50 values of DOX and ACM-DOX micelles being 6.80 ± 0.50 and 4.64 ± 0.32 μg/mL, respectively. Given above, ACMeNa has great application potential as a DOX carrier for the treatment of cancers.


Introduction
Doxorubicin hydrochloride (DOX) is one of the widely used powerful antineoplastic agents in the clinical practice for the treatment of a wide range of cancers, including hematological malignancies, soft tissue sarcomas, and solid tumors. 1 It is generally used in its hydrochloride form to increase drug solubility in water. Unfortunately, when it was intravenously administered, due to its low binding rate to plasma proteins, rapid distribution occurs in the heart, kidney, liver, spleen, and lung, with bone marrow suppression and significant dose-dependent cardiotoxicity. 2,3 Severe adverse reactions, such as congestive heart failure, have limited its clinical application. 4 Taking these into account, encapsulating drugs into nanoparticles appears to have distinct advantages, including controlling drug delivery, improving drug pharmacokinetics, enhancing efficacy, and reducing side effects. 5 The developed drug delivery strategies provide an alternative to simple intravenous infusion of aqueous solution of DOX to reduce toxicity and side effects of the drug. 5,6 The most successful one is Doxil, an unique form of liposomal DOX, that was approved by Food and Drug Administration in 1995. Doxil is undoubtedly successful in reducing DOX cardiotoxicity, 7,8 but the process for liposome preparation is complex, leading to a great increase in the cost of drug manufacturing and bringing some side effects caused by lipid components. 9 Micelles are self-assembled core-shell nanocarriers formed by surfactants or polymers. Micellar drug formulation, due to its smaller particle size (10-100 nm), could passively target solid tumors (even poorly permeable tumors) and internalize cells more effectively. 10 The formulation has high vascular permeability, which is not easy to be opsonized by plasma proteins, but is also efficient to avoid reticuloendothelial system phagocytosis and reduce renal excretion, thus, the drug could circulate in the blood for a long time. 11 DOX has also been encapsulated in micelles to achieve passive targeting of the drug, and thereby, reducing adverse effects of the drug. 10 Micelles can be formed by simple self-assembling in solution. Considering that, micelles may be characterized by an easier preparation and a higher scale-up feasibility in comparison to other nanocarriers, such as polymeric nanoparticles and liposomes that require more complex, long-lasting, and costly manufacturing procedures. 12 In contrast to micelles formed by surfactants of low molecular weight, polymeric micelles are associated with main advantages like lower critical micelle concentration (CMC) and higher kinetic stability. 13 One of the representatives is Genexol PM, which is a polymeric-micelle formulation loading 16 wt% of paclitaxel (PTX) and approved in South Korea. 14 However, due to the complex structure of nanoparticles and potential immunogenicity, novel amphiphilic materials with superior physical and chemical properties have been extensively explored. 15 Surfactant XR17 has been employed in Apealea, 16 a marketed PTX micelle approved by the European Medicines Agency in 2018. XR17 is a mixture of two iso-forms of N- Abstract Doxorubicin hydrochloride (DOX) is one of the widely used antineoplastic agents in treating various cancers, yet it is always associated with the occurrence of adverse reactions that limit its clinical use. Currently, encapsulating DOX in micelles may represent a promising strategy to reduce toxicity and side effects of the drug. This study aimed to explore a novel acitretin-based surfactant (ACMeNa) with good solid stability to encapsulate DOX to form micelles (ACM-DOX). In this work, ACM-DOX micelles were prepared by a microfluidic method free of organic solvents. The characteristics of ACM-DOX micelles were assessed, including morphology, particle size, stability, entrapment efficiency, and drug loading. An in vitro cytotoxicity experiment of the micelles on MDA-MB-231 (a human breast cancer cell line) was also performed. The micelle formation mechanism suggested that the insoluble ACMeNa/DOX complex was formed by electrostatic interaction, and subsequently encapsulated by self-assembly into micelles. The designed ACM-DOX micelles had an average particle size of 19.4 AE 0.2 nm and a zeta potential of À43.7 AE 2.4 mV, with entrapment efficiency and drug loading efficiency of 92.4 AE 0.5% and 33.4 AE 0.3%, respectively. The ACM-DOX micelles had worm-like structures under a Cryo-transmission electron microscope and exhibited good stability within 8 hours after reconstitution and 4-to 32-fold dilution of its reconstituted solution. ACM-DOX micelles released 80% of DOX within 24 hours in a medium of pH ¼ 5.0, and its drug profile can be described by a first-order model. Moreover, ACM-DOX micelles showed cytotoxicity against MDA-MB-231 in a dose-dependent manner, and displayed a higher antitumor activity as compared with free DOX, with IC 50 values of DOX and ACM-DOX micelles being 6.80 AE 0.50 and 4.64 AE 0.32 μg/mL, respectively. Given above, ACMeNa has great application potential as a DOX carrier for the treatment of cancers. retinoyl-L-cysteic acid methyl ester sodium salt, which are both amphiphilic compounds with well-defined structures and low molecular weights. However, XR17 is unstable in the context of light, heat, and oxygen because its structure contains retinol, which is responsible for its rapid degradation. 17 However, acitretin could avoid the reaction of conjugated rearrangement and oxidation of the retinol ring, and thus it has a stable structure. 18 In this study, we tried to replace the retinol part of XR17 with an acitretin to overcome the poor stability of XR17. The new amphiphilic surfactant, namely ACMeNa, was synthesized from acitretin and Lcysteic acid methyl ester (LAME) through an amide bond, and then self-assemble in water to form micelles at concentrations above CMC. ACMeNa can not only stably encapsulate lipophilic drugs including docetaxel, 19 but also stably encapsulate water-soluble drugs including DOX. In this work, DOXloaded polymeric micelles (ACM-DOX) were prepared via a microfluidic method. Our data suggested that ACM-DOX micelles were good in morphology, particle size, and stability, and associated with high entrapment efficiency (EE) and drug loading. ACM-DOX may represent a valuable strategy for the delivery of DOX in clinical treatment.
High-performance liquid chromatography (HPLC) was used to assess the stability of ACMeNa in solid under different temperatures and light conditions for successive 120 days. Analysis was performed using a reverse-phase HPLC with C18 column (250 mm Â 4.6 mm, 5 μm). ACN (containing 20 mmol/L potassium dihydrogen phosphate) in water (55/45, v/v) was used as an eluent. UV detection was performed at 350 nm and the column temperature was 40°C. The flow rate was set at 1.2 mL/min. Each sample was dissolved in methanol (0.5 mg/mL), and 10 μL of which was injected for HPLC analysis.

Preparation of the ACM-DOX Micelles
Micelles were prepared using the microfluidics system with carrier and DOX at a ratio of 1.8:1 (w/w). Then 5.8 mg/mL of DOX was prepared by dissolving a certain amount of DOX in water. An ACMeNa solution (5.1 mg/mL of) was prepared by dissolving ACMeNa in water followed by adjusting the pH to approximately 6 using dilute hydrochloric acid. Both the solutions were prepared independently for separate inlets and mixed within the chip. Herein, an in-plane micromixing structure called Tesla mixer was used. Flow rates of the solutions of DOX and ACMeNa were 3 and 6 mL/min, respectively, and controlled with syringe pumps. The micellar solution was filtered through a 0.22 μm microporous membrane filter to obtain micelles, which were lyophilized with a freeze dryer (VirTis Advantage, SP Scientific) through a designed procedure. The obtained micelle formulation was named ACM-DOX.

Preparation of ACMeNa/DOX Complex and ACMeNa/DOX Physical Mixtures
Our preliminary experiment of preparing DOX-loaded micelles suggested that when DOX solution was added dropwise to the ACMeNa solution, turbidity appears, and after shaking, the precipitate disappeared and the solution became clear. When the ACMeNa solution was added dropwise to the DOX solution, obvious turbidity appeared, but after shaking, the precipitation remained undissolved. It may be speculated that an insoluble ACMeNa/DOX complex was formed.
ACMeNa/DOX complexes were prepared to reveal the assembly mechanism of the ACM-DOX micelle. Briefly, ACMeNa solution (1 mL) was added dropwise into the DOX solution (23.2 mg/mL in water, 4 mL) under magnetic stirring. Stirring was continued for a further 10 minutes after the formation of a suspension. Then, it was centrifuged at 7,500 g for 10 minutes. The supernatant was discarded. The precipitate was resuspended in water, vortexed briefly, and centrifuged. The operation was repeated until a clear upper liquid was obtained. The precipitate was filtered and dried under vacuum at 40°C for 24 hours to obtain the ACMeNa/DOX complex.
Physical mixtures of ACMeNa/DOX were prepared by mixing ACMeNa (20.4 mg) and DOX powders (23.2 mg) using a vortex mixer for 1 minute.
The thermal behavior of the samples was analyzed on a DSC (Q2000, TA Instruments) equipped with a refrigerated cooling system. Experiments were performed in a nitrogen atmosphere using approximately 2 mg samples crimpled in aluminum pans. Samples were heated from 40 to 280°C at a heating rate of 10°C/min. PXRD data were collected on a Bruker D8 X-ray diffractometer using Cu Kα radiation at 40 kV and 40 mA passing through a nickel filter. Samples were scanned in the 2u range of 10 to 30°. Measurement was done by placing a powder sample onto a glass plate.
FTIR spectra were measured on a Shimadzu IRTracer-100 spectrometer (Shimadzu, Japan) using KCl pellets. The scanning ranged from 4,000 to 400 cm À1 and the resolution was 1 cm À1 .

Characterization of ACM-DOX Micelles
Particle Size, Zeta Potential, Morphological, and Stability Analyses Dynamic light scattering (DLS; Nicomp 380 N3000, PSS Company, United States) was used to assess particle size, polydispersity index, and zeta potential of the ACM-DOX micelles at 25°C. Cryo-transmission electron microscopy (Cryo-TEM; Talos F200C, Thermo Scientific, United States) was used to investigate the morphological features. Particle stability was investigated using the same method after reconstitution over 8 hours with 4-to 32-fold dilutions of their reconstituted solutions.

Determination of Entrapment Efficiency and Drug Loading Efficiency
First, 16 mg of the ACM-DOX micelles was reconstituted in 3 mL water. Unentrapped drug was removed by ultrafiltration centrifugation (7,500 g, 20 minutes). HPLC (Shimadzu, Japan) was used to determine the amount of DOX encapsulated. The system was equipped with a C18 column (250 mm Â 4.6 mm, 5 μm) with a detection wavelength at 254 nm. The mobile phase consists of 0.01 mol/L sodium lauryl sulfate solution, ACN, and methanol at a ratio of 50:50:6 (v/v). The flow rate was 1 mL/min. The injected sample volume was 10 μL. The calibration curve for DOX was obtained from standard solutions of DOX with the concentration ranging from the analyte limit of quantification (0.001) up to 1.0 mg/mL, with the regression coefficient greater than 0.999. EE and drug loading efficiency (DLE%) were calculated using the following Eqs. (1) and (2):

Determination of the Critical Micelle Concentration
Pyrene was used as a fluorescent probe to estimate CMC, according to a reported study. 20 Briefly, pyrene (1 mg/mL) was dissolved in acetone to achieve a final concentration of 6 Â 10 À6 mol/L. To a series of vials was added the pyrene solution and evaporated under nitrogen, followed by the addition of different concentrations of ACMeNa aqueous solution (0.00001-1 mg/mL). The mixtures were vortexed for 2 minutes. For each vial, fluorescence intensity was recorded using a spectrofluorometer at an excitation wavelength of 334 nm and emission wavelengths of 375 nm for I 1 and 384 nm for I 3 . The CMC was given as the concentration for which an intersection of two straight lines of the I 3 /I 1 ratio was observed.  A parallel dialysis, in which a dialysis bag containing the same amount of free DOX, was conducted for comparison. The concentration of DOX in the dissolution medium was determined using HPLC as described in the section "Deter-mination of Entrapment Efficiency and Drug Loading Efficiency."

Cytotoxicity Assay
Cell Counting Kit-8 (CCK-8) was used to assess cell viability. Briefly, aliquots (100 μL) of MDA-MB-231 cells were seeded in a 96-well plate (5 Â 10 3 cells/well) and incubated for 24 hours before the experiment. MDA-MB-231 cells were treated with 100 μL of DMEM containing different concentrations of DOX, ACMeNa, and ACM-DOX micelles, and incubated for 24 hours. The micelle concentration was calculated depending on the DOX content. At the same time, cells treated with DMEM were used as a blank control. The culture medium was removed. Cells were washed twice with PBS, mixed with 100 μL of culture medium containing 10% CCK-8, and incubated for another 2 hours. Cell viability (%) was calculated based on absorbance at 450 nm, measured by a microplate reader (Thermo Scientific Multiskan FC). Each group has six wells.

Results and Discussion
Successful Synthesis of ACMeNa The chemical stability of solid ACMeNa is shown in ►Table 1. Our data showed that ACMeNa had a good stability when protected from light at À18°C and 4°C. However, at 120 days, approximately 2% ACMeNa was degraded at ambient temperature (25°C), in the presence of light or protected from light. Our data recommended ambient storage of ACMeNa in the dark.

Formation Mechanism of ACM-DOX Micelles
DSC is a powerful tool to reflect the purity, crystalline form, and stability of substances by measuring the thermal effects of substances in the process of phase transition so as to distinguish and identify different substances. 21 As shown in ►Fig. 3, there is a single endothermic peak in DSC plots of both ACMeNa and DOX, indicating a high purity of the two substances. However, in comparison to ACMeNa or DOX, there are two endothermic peaks of physical mixtures of ACMeNa/DOX that shifted to the left, and this was consistent with the melting-point depression rule for mixtures. Interestingly, only one exothermic peak was found in the ACMeNa/DOX complex, which was distinct from the other three samples, indicating that the ACMeNa/DOX complex existed in a unique form. PXRD can qualitatively and quantitatively analyze the crystalline morphology of solid substances and reflect small changes in the crystalline composition of solid drug components in the PXRD patterns. 22 ►Fig. 4 demonstrates a crystalline state of ACMeNa and DOX. The peaks of PXRD patterns of ACMeNa/DOX physical mixtures are nearly a superposition of spectra of ACMeNa and DOX. However, almost no diffraction peak of the ACMeNa/DOX complex is observed, implying an amorphous state of the complex. Combined with the results of DSC patterns (►Fig. 3), the ACMeNa/DOX complex was in an amorphous state, and the exothermic peak in the DSC may be generated when the complex transformed from an amorphous state to a crystalline state.
Through detecting the subtle changes in functional groups of the chemical structure in FTIR spectroscopy, 23 interaction between the sulfonic acid group of ACMeNa and the amino group of DOX was assessed. In ►Fig. 5, the infrared spectrum of the ACMeNa/DOX complex was a superposition of ACMeNa and DOX spectra, and the characteristic absorption peaks corresponding to ACMeNa and DOX were shown in the infrared spectrum of the ACMeNa/DOX complex, indicating no new chemical bond is formed between ACMeNa and DOX. However, subtle changes in characteristic peaks in the fingerprint region of the ACMeNa/DOX complex demonstrated hydrogen bonds or van der Waals forces between ACMeNa and DOX. At the same time, the amino stretching vibration peak disappeared at 3,500 cm À1 , indicating that the amino group contributed to the formation of the ACMeNa/DOX complex.
Above all, an amorphous form of the ACMeNa/DOX complex may be a result of electrostatic interaction between the sulfonic acid group of ACMeNa carriers and the amino group of DOX. Therefore, the ACMeNa/DOX complex may be involved in the process of DOX loading. It can be speculated that an insoluble ACMeNa/DOX complex (7 µg/mL in water; DOX concentration of the supernatant in the section "Preparation of ACMeNa/DOX Complex and ACMeNa/DOX Physical Mixtures" was determined using HPLC as described in the section "Determination of Entrapment Efficiency and Drug Loading Efficiency") was first formed, and subsequently encapsulated by the excessive carriers during the self-assembly process to form ACM-DOX micelles.

Characterization of ACM-DOX Micelles
Drugs can be encapsulated in the micelles during their formation, or in a second step, depending on the method used for preparation. Micelles can be prepared by different methods including direct dissolution, dialysis, emulsion with solvent evaporation, and solution casting followed by film hydration. 24 When an insoluble drug was encapsulated, environmentally unfriendly organic solvents were inevitably used in those methods. Encouragingly, DOX and ACMeNa are soluble in water, and the limitation of the organic solvents can be avoided. Considering easy manipulation conditions and high scale-up feasibility, microfluidic devices were used to prepare DOX-loaded ACM-DOX micelles (►Fig. 6A). The microchannel schematic is shown in ►Fig. 6B: DOX solution was mixed with the carrier at a volume ratio of 1:2. At the junction of the streams, enhanced mixing occurs through the Tesla structures as the focused streams flow along the channel. Given that properties such as polarity and hydration degree of micelles are not uniform within the carrier, drugs can be hosted in different sites, close to the surface or in the inner core, depending on their properties. Most of the time, hydrophobic drugs are loaded and hosted in the inner core. Due to the lipophilic nature of the ACMeNa/DOX complex, DOX could be incorporated into the micelle core at a carrier/drug ratio of 1.8:1 (w/w). Thus, in the present study, the ACM-DOX micelles were prepared using a microfluidic method that does not require organic solvents.
The smaller size of the micelles is suitable for passive tumor targeting, 25 and the smaller negative zeta potential of the surface was associated with a lower protein adsorption, which avoids mononuclear phagocyte system clearance and prolongs the circulation lifetime of the drug. 26,27 Consistent with this, in this work, low particle size (19.4 AE 0.2 nm) (►Fig. 6C) and zeta potential (À43.7 AE 2.4 mV) were also detected for our ACM-DOX micelles. At mean time, ACM-DOX micelles also exhibited high EE (92.4 AE 0.5%) and DLE (33.4 AE 0.3%). DLS results showed a size of 10.76 and 17.28 nm of blank ACM and ACM-DOX micelles, respectively, suggesting the comparable size of the two micelles (►Fig. 6C). Cryo-TEM results showed a spherical system of blank ACM micelles, and a worm-like morphology of ACM-DOX micelles (►Fig. 6D), which was also found in squalenoylated doxorubicin micelles reported by Maksimenko et al. 28 A micellar shape can be influenced by different factors including the structure of carriers used, as well as the surrounding environment in terms of temperature, pH, and composition. As for squalene-acylated doxorubicin micelles, the concentration of loading drugs, ionic strength, and anionic properties have been reported to affect the morphology of micelles. 29 However, these factors may not affect the worm-like appearance of ACM-DOX micelles, but may be closely associated with the unique interaction between carriers and DOX. This can be supported by a previous reported study, 30 in which cetyltrimethylammonium bromide (CTAB) formed worm-like CTAB/NaSal micelles at lower concentrations, in the presence of sodium salicylate (NaSal). Micelle morphology was fundamental in predicting micelles' behavior in vivo, as the shape of nanocarriers plays a key role in influencing the circulation time, biodistribution, and cellular uptake. For instance, the soft and flexible nature of worm-like micelles could minimize phagocytosis, 31 prolong their circulation in the blood, 32 and increase maximum tolerated dose. 33 Thus, worm-like ACM-DOX micelles were more likely to be favored in delivering and releasing DOX at tumor sites. Notably, in this study, a stable average particle size of approximately 20 nm was obtained within 8 hours after the reconstitution of the lyophilized powders with water (►Fig. 6E). The particle size and size distribution did not change upon 32-fold dilution of the initial dispersion (►Fig. 6F), indicating their superior colloidal stability.
The pyrene method was used for CMC determination, due to its high sensitivity and easy application. 34 The CMC of ACMeNa was determined by plotting fluorescence intensity (using pyrene as a fluorescent probe) against ACMeNa concentration. Our data confirmed a low CMC of ACMeNa, approximately 6.50 μg/mL (►Fig. 7), which is much lower than those of conventional surfactants, such as sodium dodecyl sulfate (2.39 mg/mL) and reported polypeptide, and similar to those of PEG-poly(amino acid) block copolymers used in clinical trials. 35,36 Due to low CMC, ACM-DTX micelles could remain stable at very low ACMeNa concentrations, which makes them relatively insensitive to dilution (as described above with 32-fold dilution), resulting in an enhanced circulation time of the drug.
Furthermore, a release study was performed at pH 5.0 to mimic a low pH value of tumor environments. 37 Our data suggested that free DOX diffused out of the dialysis bag was 80.0 AE 0.5% in the first 4 hours, and 99.1 AE 2.3% at 12 hours (►Fig. 8), suggesting a rapid release profile of free DOX. However, the released DOX from ACM-DOX micelles was only 17.3 AE 0.4% in first 4 hours, 49.4 AE 1.3% at 12 hours, and 81.7 AE 4.3% at 24 hours, suggesting a slower release profile of DOX from ACM-DOX micelles in comparison to free DOX. Then, various mathematical models were applied to evaluate in vitro drug release profile of ACM-DOX micelles, including zero-order kinetics, first-order kinetics, the Higuchi model, the Hixson-Crowell model, and the Ritger-Peppas formula model (►Table 2). Correlation coefficient (R 2 ) analysis showed that ACM-DOX micelles have a sustained release effect to some extent, and first-order kinetics, with R 2 ¼ 0.9991, turned out to be the most suitable model to describe the release profile of DOX from ACM-DOX micelles.

ACM-DOX Micelles Inhibited Viability of MDA-MB-231 Cells
As shown in ►Fig. 9A, the carrier ACMeNa exhibited almost no cytotoxicity, and the cells maintained a survival rate of 86.20 AE 4.66% even at a high concentration (100 μg/mL), indicating a good cell compatibility of the carrier. DOX and ACM-DOX micelles inhibited cell viability of MDA-MB-231  cells in a dose-depended manner. At the same concentration, the cytotoxicity of ACM-DOX micelles was significantly higher than that of free DOX. Besides, IC 50 values of DOX and ACM-DOX were 6.80 AE 0.50 and 4.64 AE 0.32 μg/mL, respectively (►Fig. 9B). Given above, ACMeNa is noncytotoxic, and in comparison to free DOX, DOX encapsulated in ACM-DOX micelles was more cytotoxic.

Conclusion
In this work, a novel and stable amphiphilic surfactant, ACMeNa, was synthesized to encapsulate water-soluble DOX to control the release and enhance the efficiently of the drug. ACM-DOX micelles can be prepared by a microfluidic method, which avoided the use of organic solvents, and was highly feasible to scale up. The mechanism of micelle formation suggested that an insoluble ACMeNa/DOX complex was formed by electrostatic interaction, and then effectively encapsulated to form ACM-DOX micelles during self-assembly. ACM-DOX micelles had a worm-like morphology, a particle size distribution of approximately 20 to 100 nm, as well as high EE% and DLE%. Moreover, ACM-DOX micelles were associated with good dilution stability in vitro and a slow-release effect, and showed significant concentration-dependent toxicity. The strategy of ACMeNa as nanocarriers provides a means to encapsulate water-soluble DOX. The stability and size of worm-liked ACM-DOX micelles make them particularly suitable for cancer therapy with tumor-selectivity potential. Further work will focus on analyzing the exact localization ACM-DOX micelles and their efficacy in vivo.