Crystallization of Supersaturated PEG-b-PLA for the Production of Drug-Loaded Polymeric Micelles

Abstract

In this study, we propose the “crystallization from supersaturated solution” method for producing drug-loaded polymeric micelles. This method involves the formation of solid drug-encapsulating crystals of a diblock copolymer through isothermal crystallization from a supersaturated solution of the copolymer in low molecular weight PEGs containing the drug, followed by dissolution of the crystals to obtain drug-loaded micelles. We fabricated and characterized micelles loaded with several model drugs (paclitaxel, rapamycin, and docetaxel) and their oligo(lactic acid)₈-prodrugs using PEG_4kDa-b-PLA_2.2kDa as the micelle-forming copolymer and PEGs of varying molecular weights (200, 400, and 600 Da) as solvents.

Our findings indicate that the molecular weight of the solvent PEG and the target drug loading significantly influence the physicochemical properties of the resulting micelles, including loading efficiency and particle size distribution. Micelles produced with PEG200 as the solvent exhibited the highest loading efficiency, followed by those made with PEG600 and PEG400 for all the drugs and prodrugs tested. Increasing the target drug loading enhanced both the loading efficiency and average particle size across all formulations. Furthermore, prodrug-loaded micelles showed higher loading efficiency and improved stability in aqueous solutions compared to their parent drug counterparts. Crystals encapsulating both parent drugs and prodrugs could be stored at room temperature for extended periods, producing micelles with no significant differences in loading efficiency and particle size distribution compared to freshly prepared micelles. Additionally, the crystals demonstrated a rapid dissolution rate, forming uniform micelles after just 5 seconds of hydration and agitation. Cytotoxicity studies against 4T1 and MDA-MB-231 breast cancer cell lines revealed that the molecular weight of the PEG used as the solvent impacts the cytotoxicity of the resulting micelles, with those produced using PEG200 displaying the highest cytotoxicity, followed by PEG400 and PEG600.

Overall, the crystallization from supersaturated solution method proves to be an effective platform for prolonged storage and rapid formation of stable, drug-loaded polymeric micelles. It has the potential to eliminate the need for freeze-drying in the formulation and storage of drug-loaded polymeric micelles. These findings highlight the method’s potential for advancing drug delivery systems, particularly for the solubilization of hydrophobic drugs using micellar formulations.

1. Introduction

Polymeric micelles remain a widely utilized and highly promising subset of nanocarriers for drug delivery applications [1–7]. Among their diverse applications [8–12], the solubilization of water-insoluble or poorly soluble drugs with high potency and significant toxicity for injection stands out as a key function of this category of nanocarriers. Polymeric micelles enhance the safety profile of drugs, particularly for injectable formulations, by eliminating the need for toxic solvents such as ethanol or low molecular weight surfactants like Cremophor EL—a castor oil derivative used in the formulation of Taxol^® [13, 14].

The current literature suggests that physically entrapping water-insoluble small molecules into the micelle core represents the most viable route for clinical translation of polymeric micelles for injection [15]. Among such formulations, several have advanced to the clinical stage, either undergoing clinical trials or already approved for human use [15–18]. Examples include Genexol^® PM and Nanoxel^® M, approved in South Korea, as well as NK105 and NC-6004, both completed phase 3 clinical trials.

In the development of polymeric micelle formulations for injection, careful consideration of the manufacturing process is essential to ensure consistent physicochemical properties and scalability. The choice of manufacturing method can significantly influence key characteristics of the final polymeric micelle formulation, including drug loading, size distribution, and stability in aqueous media, which are crucial for translation [19].

Various manufacturing processes are available for producing polymeric micelles. These methods typically involve replacing a non-selective organic solvent in a drug-copolymer solution with a selective solvent for the shell-forming block to form drug-loaded micelles. In the thin film hydration technique [20–25], micellization is induced by the addition of water after complete removal of the organic solvent. In other approaches, micellization occurs upon the addition of water in the presence of the organic solvent, with micelles forming once a critical water concentration is achieved [26, 27]. The final product is obtained by thoroughly removing the organic solvent. Dialysis involves gradually replacing the organic solvent with water to induce micellization [28–30]. Cosolvent azeotrope evaporation entails the gradual addition of water to an organic solvent, followed by the removal of the formed azeotrope under reduced pressure to completely eliminate the organic solvent [31]. In the emulsification-solvent evaporation method, a water-immiscible organic solvent, to which water is added, forms an emulsion and induces micellization [32–34]. The organic solvent is then removed by evaporation. We recently introduced a novel fabrication method where low molecular weight polyethylene glycols (PEGs) are utilized as the non-selective solvent at elevated temperatures, and micellization is induced by the addition of water [35].

For long-term storage, the aqueous suspension of drug-loaded polymeric micelles needs to be transformed into a dehydrated solid form, typically achieved by lyophilization [36, 37]. Nanoparticles suspended in an aqueous medium are physically unstable due to particle aggregation and fusion. They are susceptible to drug leakage due to the hydrolytic action of water on the biodegradable polymer matrix and the formation of undesirable degradation products. Water also facilitates the growth of microorganisms, rendering the formulation unacceptable for administration [38–40]. Therefore, the current approach for formulating most drug-loaded polymeric micelles is a two-step procedure that involves (i) the production of an aqueous suspension of drug-loaded micelles and (ii) transforming it into a dehydrated solid form through lyophilization. Residual moisture limits are optimized for each lyophilized product based on its specific stability requirements. These limits are supported by stability data demonstrating that safety, purity, and potency are maintained at the recommended moisture levels throughout the product’s shelf life. Typically, the residual moisture content of lyophilized pharmaceutical products is in the range of 1–3% (w/w) [41]. The lyophilized drug-loaded polymeric micelles can be reconstituted through hydration before use. Only in a few cases, where the copolymer is water-soluble, a one-step freeze-drying method, which involves the use of a mixture of tert-butanol/water, can be employed successfully to obtain a dehydrated solid form that produces micelles upon hydration [42, 43].

While the additional lyophilization step is essential for long-term preservation, it may introduce scalability complications and lead to issues in formulation stability, such as leakage of the loaded drug during the process. Minimizing the steps in the manufacturing process to achieve and preserve high loading efficiency, while mitigating the detrimental side effects of manufacturing processes such as formulation instability, is highly desirable for translation [15].

Here, we propose a new formulation strategy for one-step production of drug-loaded polymeric micelles directly from a solid form, eliminating the need to first prepare an aqueous suspension. We previously showed that drug-loaded polymeric micelles can be fabricated using low molecular weight PEGs as the solvent [35]. PEG is an FDA-approved polymer widely utilized in pharmaceutical formulations, valued for its customizable physicochemical properties and established safety profile, both of which are essential considerations in excipient selection during formulation development. PEG is commonly utilized as a vehicle in oral and parenteral dosage forms due to its low toxicity, excellent miscibility with aqueous fluids, and ability to dissolve many poorly water-soluble compounds. For drugs with low aqueous solubility and bioavailability challenges, PEG-based formulations have demonstrated significantly enhanced bioavailability and reduced inter-subject variability in plasma concentrations, especially when administered as solutions or suspensions [44, 45]. In this study, we hypothesize that PEG-b-PLA crystals, formed by crystallization from a supersaturated solution of the copolymer and drug in low molecular weight PEG, have the capacity to encapsulate the drug during their formation and subsequently transform into drug-loaded polymeric micelles upon hydration. Based on our previous work modeling mixtures of low molecular weight PEG and PEG-b-PLA diblock copolymers, such systems may undergo liquid-liquid phase separation (LLPS) under specific conditions [46], such as reduced temperature. This LLPS could concentrate the copolymer within phase-separated droplets, into which the drug may also partition and concentrate. Subsequent crystallization of the crystallizable PEG block within the copolymer may then result in the formation of drug-loaded solid crystals, providing the foundation for the hypothesis explored in the present study. We also investigated the impact of different parameters including the molecular weight of PEG and crystallization conditions on the properties of the obtained micelles. PEG_4kDa-b-PLA_2.2kDa and paclitaxel were used as the micelle building block and the model drug, respectively, but the applicability of the method to other payloads was also investigated. Additionally, we studied the impact of the formulation parameters on cytotoxicity.

2. Experimental details

2.1. Materials

Polyethylene glycol-b-poly(DL-lactide) (PEG-b-PLA, PEG block Mw: 4 kDa, PDLLA block Mw: 2.2 kDa) diblock copolymer was purchased from JenKem Technology (Plano, TX). Different molecular weight PEGs (200, 400, 600 Da) were obtained from Sigma-Aldrich (St. Louis, MO). Paclitaxel, rapamycin and docetaxel were acquired from LC Laboratories (Woburn, MA). oligo(lactic acid)₈-paclitaxel, oligo(lactic acid)₈- rapamycin and oligo(lactic acid)₈-docetaxel were synthesized by Genesis Drug Discovery and Development (Hamilton, NJ) according to our previously published protocols [25, 47, 48].

2.2. Isothermal crystallization kinetics and crystalline structure of PEG-b-PLA

The isothermal crystallization kinetics of PEG_4kDa-b-PLA_2.2kDa from supersaturated solutions of the copolymer in low molecular weight PEG at room temperature was studied using differential scanning calorimetry (DSC 404 F1 Pegasus, NETZSCH, Germany). Briefly, diblock copolymer granules (5.0 wt.%) were added to liquid PEG (Mw 200–600 Da), heated to 60 °C and vortex-shaken to obtain a transparent mixture. The mixture was incubated at 60 °C for 1 h before being transferred to aluminum crucibles. The thermal history of the samples was erased by 15 min isothermal hold at 80 °C. The isothermal crystallization was then performed by cooling the samples to 25 °C and isothermal hold for various periods ranging from 0.5 to 48 h. The samples were finally heated to 80 °C in the second scan to measure the melting onset temperature and enthalpy. A heating/cooling rate of 5 °C/min was used for all the scans.

To study the crystalline structure, PEG_4kDa-b-PLA_2.2kDa crystals were prepared by isothermal crystallization from solution and analyzed using powder X-ray diffraction (PXRD). Briefly, a 5.0 or 1.0 wt.% PEG_4kDa-b-PLA_2.2kDa solution in PEG (200, 400 or 600 Da) was prepared by heating the mixture to 60 °C, vortexing and incubation for 1 h. Isothermal crystallization was conducted by transferring the samples to 25 °C and incubating for 72 h. The crystals were collected through centrifugation at 10000 g for 20 min and analyzed using PXRD (Bruker D8 Advance diffractometer with a Cu Kα source, λ = 1.54178 Å). The diffraction patterns were collected at 2θ between 10–40° at a scanning rate of 5°/min.

Laser microscopy (LEXT OLS4000, Olympus, Japan) was used to image the morphology of the crystals in the solution.

2.3. Preparation and characterization of drug-loaded crystals and micelles

To prepare paclitaxel-loaded PEG-b-PLA crystals, paclitaxel was added to PEG-b-PLA solution in liquid PEG prior to crystallization. PEG_4kDa-b-PLA_2.2kDa/paclitaxel/PEG at 1.0 or 0.5 wt.% copolymer concentration and 33.0 wt.% paclitaxel target loading (1.00/0.50/98.50 or 0.50/0.25/99.25 copolymer/paclitaxel/PEG w/w/w) was heated to 60 °C and vortexed to obtain a transparent mixture. After 1 h of incubation at 60 °C, the samples were transferred to 25 °C in glass shell vials or syndiotactic polypropylene microtubes and allowed to crystallize isothermally for up to 72 h. The crystals were separated via centrifugation at 10000 g for 20 min after isothermal crystallization at 25 °C and analyzed.

The paclitaxel content in the crystals was measured through a direct or indirect method utilizing high-performance liquid chromatography (HPLC, Prominence, Shimadzu, Japan) equipped with an autosampler, a C18 column and a UV-Vis detector. In the direct method, crystals were dissolved in acetonitrile (ACN) and analyzed. Paclitaxel was detected at 228 nm absorption wavelength following injection of 10 μL sample using a 70/30 v/v ACN/water mixture as the mobile phase at a flow rate of 1 mL/min. The same procedure was repeated for the supernatant to obtain the ratio of paclitaxel loaded into the crystals and of that remaining in the solution. In the indirect method, crystals were first hydrated using DI water to obtain drug-loaded particles at a target paclitaxel concentration of 1 mg/mL considering the initial amount of paclitaxel in the PEG solution. The samples were then analyzed for paclitaxel content and particle size distribution. To measure paclitaxel content, the unencapsulated drug was first removed by centrifugation at 10000 g for 10 min. Following centrifugation, the supernatant of each sample was analyzed using HPLC. Dynamic light scattering (DLS, nano ZS, Malvern) was used to measure particle size distribution. The measurements were carried out at 25 °C using water as the dispersant.

The following parameters were frequently calculated/measured for different samples when a drug was present in the mixture:

$$\mathit{\text{Target}}\mathit{\text{drug}}\mathit{\text{loading}}=\frac{W_{\mathit{\text{initial}}\mathit{\text{drug}}}}{W_{\mathit{\text{initial}}\mathit{\text{drug}}}+W_{\mathit{\text{initial}}\mathit{\text{copolymer}}}}$$

$$\mathit{\text{Loading}}\mathit{\text{or}}\mathit{\text{Encapsulation}}\mathit{\text{efficiency}}=\frac{W_{\mathit{\text{encapsulated}}\mathit{\text{drug}}}}{W_{\mathit{\text{initial}}\mathit{\text{drug}}}}$$

The impact of isothermal crystallization period on the paclitaxel loading efficiency and particle size distribution of micelles was investigated by collecting the PEG_4kDa-b-PLA_2.2kDa crystals at different time points ranging from 1 to 72 h, followed by hydration using DI water and analysis. The feasibility of prolonged storage of crystals and its impact on the resulting micelles were evaluated by producing paclitaxel-loaded PEG_4kDa-b-PLA_2.2kDa crystals from solution, followed by storing the crystals for 0, 15, or 30 days at 25 °C and 50% RH, then hydrating them using DI water and analyzing the paclitaxel loading efficiency and particle size distribution of the resulting micelles. The dissolution rate of paclitaxel-loaded PEG_4kDa-b-PLA_2.2kDa crystals or the rate of their transformation to drug-loaded polymeric micelles was investigated by adding DI water followed by vortexing for 1, 5, 10, or 15 seconds and measuring both particle size distribution and paclitaxel concentration in the solution.

The effects of payload, target drug loading, and target drug concentration were also examined. Paclitaxel, rapamycin, docetaxel, or their oligolactic acid-conjugated o(LA)₈-prodrugs were incorporated into a PEG_4kDa-b-PLA_2.2kDa/PEG mixture at a copolymer concentration of 1.0 wt.% and drug/copolymer ratios of 1:2 or 1:1 w/w, aiming for 33.0 or 50.0 wt.% target drug loading, respectively. The samples were heated to 60 °C and vortexed to obtain a transparent mixture. Following 1 hour of incubation at 60 °C, the samples were cooled to 25 °C and allowed to crystallize isothermally for up to 72 hours. The resulting crystals were separated via centrifugation at 10000 g for 20 minutes following isothermal crystallization at 25 °C. Drug-loaded polymeric micelles were then obtained by hydrating the crystals using DI water at target drug concentrations of 1 or 5 mg/mL and subsequently analyzed for loading efficiency, particle size distribution, and stability over time. The feasibility of prolonged storage of the crystals and their dissolution rate were measured for selected samples according to the protocol mentioned above. Stability was assessed by monitoring the particle size distribution and drug leakage over a 24-hour period.

To evaluate the capacity of crystals to encapsulate a combination of two drugs while preserving the initial feed ratio in the resulting micelles, paclitaxel and rapamycin were co-loaded into PEG_4kDa-b-PLA_2.2kDa at paclitaxel/rapamycin w/w ratios of 2:1, 1:1, or 1:2, targeting a total drug loading of 33.0 wt.%, following the same protocol described earlier. The crystals were subsequently hydrated using DI water at target drug concentrations of 1 or 5 mg/mL and analyzed for their properties.

2.4. Cell Culture and Cytotoxicity Assay

4T1 and MDA-MB-231 cells (3×10⁴ cells/mL, 100 μL) were obtained from American Type Culture Collection (ATCC) and seeded in a 96-well plate and incubated for 24 hours at 37 °C. The cells were then treated with paclitaxel-loaded PEG_4kDa-b-PLA_2.2kDa micelles prepared by crystallization from supersaturated solutions of PEG200, PEG400, or PEG600, followed by a further 72-hour incubation at 37 °C. After incubation, cell viability was evaluated using the CellTiter-Blue assay (Promega, Madison, WI).

2.5. Statistical analysis

All measurements are reported as mean ± SD. Statistical differences between groups were analyzed using one-way ANOVA, followed by Tukey’s Honest Significant Difference (HSD) test for multiple comparisons. A p-value of less than 0.05 was considered statistically significant.

3. Results and Discussion

3.1. PEG-b-PLA can encapsulate drugs in solid form when crystallized from a supersaturated solution and subsequently form drug-loaded polymeric micelles upon hydration

Cooling a 5.0 wt.% PEG_4kDa-b-PLA_2.2kDa mixture in liquid PEG (200–600 Da) from 60 °C to room temperature causes the sample to become turbid, eventually forming a waxy solid due to the crystallization of the PEG block of the copolymer at lower temperatures. In other words, a 5.0 wt.% PEG_4kDa-b-PLA_2.2kDa solution in liquid PEG is supersaturated at 25 °C. We propose that, under these conditions, the crystals can encapsulate a significant amount of the drug as they separate from the solution at temperatures below the crystallization temperature of PEG-b-PLA.

The kinetics of crystallization of PEG_4kDa-b-PLA_2.2kDa are highly dependent on the molecular weight of the PEG used as the solvent. Figure 1a shows the enthalpy of fusion as a function of isothermal crystallization time for copolymer crystals formed at 25 °C, either from the melt or from a 5.0 wt.% solution in PEG200, PEG400, or PEG600. The data indicate that the crystallization rate is inversely proportional to the molecular weight of the solvent PEG. Given sufficient crystallization time, the majority of copolymer chains crystallize out of the solution, as evidenced by the comparable enthalpy of fusion between the copolymer crystallized from the solution and that crystallized from the melt.

Figure 1.

(a) Enthalpy of fusion as a function of isothermal crystallization time and (b) melting onset for PEG_4kDa-b-PLA_2.2kDa crystals formed at 25 °C from the melt or from a 5.0 wt.% solution in PEG200–600. (c) XRD pattern of crystals collected after isothermal crystallization from a 5.0 or 1.0 wt.% PEG_4kDa-b-PLA_2.2kDa solution in liquid PEG in polypropylene tubes. (d) Laser microscope images of crystals obtained from a 1.0 wt.% PEG_4kDa-b-PLA_2.2kDa solution in PEG200 or PEG600 through isothermal crystallization at 25 °C. (e) Paclitaxel (PTX) content in crystals/paclitaxel remaining in the PEG solution, and (e’) particle size distribution for crystals obtained from a PEG_4kDa-b-PLA_2.2kDa/paclitaxel/PEG solution with 1.0 or 0.5 wt.% copolymer and 33.0 wt.% target drug loading in polypropylene tubes. (f) Paclitaxel content and (f’) particle size distribution of micelles produced by hydration of crystals obtained from a 1.0/0.5/98.5 PEG_4kDa-b-PLA_2.2kDa/paclitaxel/PEG w/w/w solution as a function of isothermal crystallization time. The effect of prolonged storage of crystals at room temperature and 50% RH on (g) paclitaxel encapsulation and (g’) particle size distribution of the micelles obtained by hydrating them. The dissolution rate of the crystals represented by (h) paclitaxel concentration in the solution and (h’) particle size distribution as a function of vortex time. (i) Schematic illustration of the “crystallization from supersaturated solution” method for the fabrication of drug-loaded polymeric micelles, showing crystallization and separation of drug-loaded solids in polypropylene tubes and their transformation into drug-loaded polymeric micelles through hydration. Statistical analysis: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

Figure 1b illustrates the effect of the molecular weight of the solvent PEG on the melting onset of crystals formed isothermally from a 5.0 wt.% PEG_4kDa-b-PLA_2.2kDa solution in liquid PEG. As shown, the lower the molecular weight of the solvent PEG, the lower the melting onset of the resulting crystals. The melting onset of the copolymer crystallized from the melt was recorded at 48.3 ± 0.1 °C, slightly lower than that of PEG_4kDa homopolymer, which measured 51.6 ± 1.4 °C (see supporting information, Figure S1). The presence of amorphous PLA blocks may restrict the diffusion of copolymer chains and inhibit crystallization, as previously reported [49]. Crystallization from PEG600, PEG400, and PEG200 solutions yielded crystals with melting onsets of 42.7 ± 0.6, 36.5 ± 0.8, and 33.0 ± 0.4 °C, respectively. The depression of the melting point in crystals obtained from the solution suggests the possible incorporation of low molecular weight PEG into the solid structures formed and/or changes in crystal size due to differences in crystallization kinetics [50–52]. A similar trend, albeit at higher temperatures, is observed for different concentrations of PEG_4kDa homopolymer in liquid PEG (see supporting information, Figure S1), indicating the significant impact of the solvent PEG’s molecular weight and the presence of the PLA block on the structure and properties of the copolymer crystals obtained from solution.

Using the melting point data obtained from DSC and applying the Nishi-Wang equation [53], which correlates the depression of the melting point of a crystalline polymer in a blend to the molar fractions, degree of polymerization, and interaction parameter, the amount of residual low molecular weight PEG was estimated. Assuming an interaction parameter of zero (${\chi }_{12}=0$) between the low molecular weight PEG and the PEG block of the copolymer—due to their identical chemical structure—the residual low molecular weight PEG in the final product was calculated to be in the range of 20–30% molar fraction.

XRD analysis was employed to investigate the structure of these crystals. The crystals were collected via centrifugation following isothermal crystallization from a 5.0 or 1.0 wt.% PEG_4kDa-b-PLA_2.2kDa solution in liquid PEG within syndiotactic polypropylene tubes for 72 hours. Pure liquid PEG served as the control for each measurement (see supporting information, Figure S2). The XRD patterns of all crystals obtained from the 5.0 wt.% solution (Figure 1c) exhibited two distinct diffraction peaks at 19.6° and 23.4°, corresponding to interplanar spacings of 4.65 Å and 3.82 Å, respectively [54, 55]. These two peaks represent the (120) and (032) crystal plane diffractions of PEG, respectively [56, 57], and are also distinctly observed in the spectrum of PEG-b-PLA crystallized from the melt (see supporting information, Figure S2). Additionally, the XRD patterns revealed two minor peaks at 32.6° and 36.3°, particularly when higher molecular weight solvent PEGs were used, which have been previously reported for PEG crystals [58–60]. These minor peaks also appear in the spectrum of PEG-b-PLA crystallized from the melt (see supporting information, Figure S2). The major diffraction peaks remained consistent in the XRD patterns of the crystals obtained from the 1.0 wt.% solution when incubated in syndiotactic polypropylene containers (Figure 1c).

We also examined the macroscopic morphology of the obtained crystals. Polymer crystals exhibit ordering across various dimensional levels, from interatomic spacings to macroscopic structures. Despite similar crystal structures confirmed by XRD patterns, crystals obtained from copolymer solutions in different PEG solvents displayed some differences. Figure 1d presents laser microscope images of crystals obtained from 1.0 wt.% PEG_4kDa-b-PLA_2.2kDa in PEG200 or PEG600 through isothermal crystallization at 25 °C. The formation of banded spherulites is evident in both types of crystals. However, the slower crystallization kinetics in the PEG600 solution allows for uniform spherulite growth in all directions, while the crystals formed in PEG200 exhibit more arbitrary growth patterns. The formation of banded spherulites has been previously reported for various homopolymers and block copolymers [61–66]. These banded spherulites are believed to arise from a periodic change in radial lamellar orientation or discrete radial lamellar packing [65].

Overall, DSC, XRD, and microscopy confirmed the crystallization of the PEG block in the PEG-b-PLA copolymer from a blend of PEG-b-PLA and low molecular weight PEGs. The higher-order structure of crystalline block copolymers or polymer blends is thought to form through the interplay between two types of phase transitions: crystallization and liquid-liquid phase separation (LLPS). In polymer blends, crystallization-induced LLPS and LLPS-induced crystallization may occur kinetically during structure formation. In this context, the amorphous component, such as solvent PEG, can either be excluded from or incorporated into the inter-crystalline lamellar region, depending on the competition between the rates of crystallization and molecular diffusion [67]. The restricted chain mobility due to tethered PEG chains bonded by amorphous PDLLA blocks is believed to play a critical role in the crystallization and thermal behavior of PEG in PEG-b-PLA [49]. The connectivity of amorphous PDLLA chains can significantly reduce the crystallization rate of the PEG block.

To investigate whether the obtained crystals could encapsulate the drug during separation from the solution, we conducted a 72-hour isothermal crystallization at 25 °C with paclitaxel present in the solution. Crystallization was performed using PEG_4kDa-b-PLA_2.2kDa/paclitaxel/PEG at copolymer concentrations of 0.5 or 1.0 wt.% and a target paclitaxel loading of 33.0 wt.% (0.50/0.25/99.25 or 1.00/0.50/98.50 copolymer/paclitaxel/PEG w/w/w) in syndiotactic polypropylene tubes. The crystals were separated via centrifugation at 10000 g for 20 minutes. The rationale for using low copolymer concentrations was to facilitate the formation of discrete crystals that could be easily separated by centrifugation. At higher copolymer concentrations (> 5.0 wt.%), crystallization produces a waxy solid or gel-like mass, making it difficult to distinguish and separate the two phases. Incomplete separation of the phases could introduce substantial errors in measuring the amount of encapsulated paclitaxel. However, the slower crystallization kinetics at lower concentrations extended the process to achieve adequate crystal content. We measured the paclitaxel content using both direct and indirect methods.

As shown in Figure 1e, PEG_4kDa-b-PLA_2.2kDa crystals obtained from both 0.5 and 1.0 wt.% solutions in low molecular weight PEG successfully encapsulated paclitaxel, as verified through direct measurement of drug content (by dissolving the crystals) and indirect measurement (by hydrating the crystals and subsequently measuring the drug content). Upon hydration of the crystals produced in polypropylene tubes from the PEG_4kDa-b-PLA_2.2kDa/paclitaxel/PEG solutions (0.50/0.25/99.25 or 1.00/0.50/98.5 w/w/w), uniform drug-loaded polymeric micelles were formed, with average sizes ranging from approximately 30 to 40 nm across nearly all molecular weights of the solvent PEG (Figure 1e’).

For the 0.5 wt.% PEG_4kDa-b-PLA_2.2kDa solution in PEG, the paclitaxel content in the crystals, measured by the direct versus the indirect method, was 16.8 ± 0.1% versus 18.9 ± 0.5%, 4.7 ± 0.0% versus 5.5 ± 0.2%, and 9.4 ± 0.0% versus 9.3 ± 1.4% w/w when PEG200, PEG400, or PEG600 were used as solvents, respectively (Figure 1e). No significant difference (p > 0.05) was observed between the paclitaxel content measured using the direct method (dissolving the crystals) and the indirect method (hydrating the crystals and measuring the drug content) for any of the examined solvent PEGs. This finding indicates that drug loss occurs exclusively during the preparation of crystals from the low molecular weight PEG solution and that paclitaxel-loaded crystals can be efficiently transformed into paclitaxel-loaded polymeric micelles without incurring additional drug loss. After confirming that the drug content remains consistent between the crystals and the drug-loaded micelles formed through their hydration, all subsequent loading efficiencies were measured post-hydration, using the indirect method.

For a 1.0 wt.% PEG_4kDa-b-PLA_2.2kDa solution in PEG, the paclitaxel content in the crystals obtained by the indirect method was 29.2 ± 1.4%, 11.0 ± 1.8%, and 12.8 ± 0.6% w/w when PEG200, PEG400, or PEG600 were used as solvents, respectively. This increased encapsulation compared to crystals obtained from a 0.5 wt.% PEG_4kDa-b-PLA_2.2kDa solution highlights the critical role of the initial copolymer concentration in enhancing drug encapsulation through an improved yield of crystal production. Given the 1.0 wt.% copolymer content relative to the 98.5 wt.% liquid PEG, it can be concluded that paclitaxel has a strong tendency to partition into the formed crystals rather than remain in the solution. It should be noted that while the crystallizable PEG block in PEG-b-PLA might not be highly receptive to impurities, the PLA block remains amorphous and is capable of entrapping paclitaxel.

One-way ANOVA indicates that the effect of solvent PEG molecular weight on the amount of paclitaxel entrapped in the crystals is highly significant (p < 0.0001). The paclitaxel content entrapped in the crystals was highest for PEG200, followed by PEG600 and PEG400. This outcome may be attributed to the interplay of two factors: the molecular weight-dependent crystallization rate, which determines the total crystal content at a given time, and the molecular weight-dependent solubility of paclitaxel in the solvent PEG.

Paclitaxel is highly soluble in low molecular weight PEGs. PEG400, for example, has been reported to dissolve paclitaxel at concentrations exceeding 125 mg/mL [68]. Our previous study revealed that the addition of water to a paclitaxel/PEG200–600 mixture at a water content of 50 wt.% or higher significantly decreases paclitaxel solubility, reducing it to less than 1 mg/mL [35]. This decrease in solubility was inversely proportional to the molecular weight of the PEG solvent, with the highest solubility observed in PEG200. These findings not only underscore the molecular weight-dependent solubility of paclitaxel but also suggest that the presence of PEG alone is inadequate to sustain paclitaxel solubility in water at the high water content achieved during crystal hydration. Instead, polymeric micelles play a pivotal role in the solubilization of paclitaxel under these conditions.

While higher paclitaxel solubility in the solvent PEG leads to lower paclitaxel entrapment, a higher crystallization rate results in a greater amount of paclitaxel being encapsulated. To further investigate the impact of copolymer crystallization kinetics on the entrapped paclitaxel content, we conducted a time-dependent paclitaxel encapsulation experiment. A PEG_4kDa-b-PLA_2.2kDa/paclitaxel/PEG solution (1.0/0.5/98.5 copolymer/paclitaxel/PEG w/w/w) was subjected to isothermal crystallization at 25 °C. At various time points, the crystals were separated, hydrated at a target concentration of 1.0 mg/mL paclitaxel, and analyzed for paclitaxel content and particle size distribution. The kinetics of paclitaxel encapsulation (Figure 1f) closely mirrored the kinetics of copolymer crystallization from solution (Figure 1a). Polymeric micelles were the dominant phase in all samples at every time point (Figure 1f’). The percentage of encapsulated paclitaxel reached a plateau for all solvent PEGs after approximately 24 hours (Figure 1f), corresponding with the near-zero crystallization rate observed after this time (Figure 1a). The percentage of encapsulated paclitaxel stabilized at around 25% w/w for micelles formed from PEG200 solution crystals and around 10% w/w for micelles formed from PEG400 or PEG600 solution crystals at extended times, consistent with previously measured paclitaxel content in the crystals. The faster rate of paclitaxel encapsulation in micelles derived from PEG200 solution crystals can be attributed to the quicker crystallization kinetics, which results in a higher crystal content compared to PEG400 and PEG600 at a given point. While PEG400-derived crystals and the resulting micelles were expected to exhibit a higher rate of paclitaxel encapsulation compared to those from PEG600 due to faster crystallization kinetics, both the rate of encapsulation and the final paclitaxel content were higher for micelles obtained by PEG600-derived crystals (Figure 1f). This could be due to the higher solubility of paclitaxel in PEG400 compared to PEG600, which likely reduces its partitioning into the crystals and compensates for the difference in crystallization rate.

It should be noted that the crystallization conditions can significantly impact the structure of the resulting solid, thereby altering the properties of the hydrated formulation. For instance, the use of a foreign heterogeneous surface during crystallization can influence the kinetics and the resulting crystal structure and morphology, primarily by affecting the nucleation rate—this is particularly relevant for polymers that have difficulty with homogeneous nucleation [69, 70]. When borosilicate glass was used during crystallization, the resulting solid exhibited a slightly different XRD pattern and a bimodal particle size distribution upon hydration (see Figure S3 in the supporting information). The cooling rate is another well-known factor that impacts the crystallization process, primarily by altering the degree of crystallinity, crystal size, and melting point [71, 72]. Varying the cooling rates from 1 to 60 °C/h altered the bimodal particle size distribution in some samples but never produced a uniform micellar solution when crystallization was conducted in borosilicate glass. In these cases, particles larger than 100 nm appeared in the hydrated sample, regardless of the solvent PEG’s molecular weight and the cooling rate. In contrast, when polypropylene was used for crystallization, polymeric micelles formed as the dominant species, even with significant changes to the copolymer concentration or cooling rate (see Figure S3 in the supporting information). These observations underscore the importance of crystallization conditions, particularly the crucial role of the surface material in determining the outcome of crystallization and the characteristics of the particles formed upon hydration.

The crystals obtained from supersaturated solutions of copolymer in liquid PEG can be considered an alternative solid form to lyophilization for storage. We next investigated the impact of prolonged storage of these crystals on the properties of the resulting micelles. Crystals obtained from a PEG_4kDa-b-PLA_2.2kDa/paclitaxel/PEG solution (1.0/0.5/98.5 copolymer/paclitaxel/PEG w/w/w) were stored at 25 °C and 50 %RH for 0, 15, or 30 days and then hydrated to a target paclitaxel concentration of 1 mg/mL, based on the initial amount of paclitaxel in the solution. Storing the crystals for 15 or 30 days led to no significant difference (p > 0.05) in encapsulated paclitaxel content (Figure 1g) or particle size distribution (Figure 1g’) compared to day 0.

We also examined the dissolution rate of these crystals by vortex-agitating them in water at 5-second intervals, measuring the paclitaxel concentration in the solution, and monitoring the formation of polymeric micelles. As shown in Figure 1h, the paclitaxel concentration reached a plateau after 5 seconds of agitation, demonstrating the rapid dissolution rate of these crystals in water. The final paclitaxel concentration—approximately 0.2 mg/mL for PEG200 and ~0.1 mg/mL for PEG400 and PEG600—matches the expected values considering the target paclitaxel concentration of 1 mg/mL and the paclitaxel encapsulation capacity of the crystals (Figure 1e). Particle size distribution analysis indicates that 5 seconds of agitation results in the formation of uniform polymeric micelles (Figure 1h’).

Based on the presented evidence, the described method can be utilized for the production of drug-containing crystals, which can be stored for prolonged periods and subsequently converted into drug-loaded polymeric micelles upon hydration and slight agitation. This method, hereafter referred to as “crystallization from supersaturated solution,” is schematically illustrated in Figure 1i.

3.2. The “Crystallization from Supersaturated Solution” method enables the loading of various drugs into polymeric micelles at different target loadings and drug concentrations.

Next, we investigated the capability of the “crystallization from supersaturated solution” method to encapsulate different drugs in polymeric micelles under various conditions and formulation parameters. We examined the impact of increasing the target drug loading and target drug concentration in the aqueous medium after dissolution of the crystals on the properties of the micelles produced. Paclitaxel, rapamycin, and docetaxel were used as model drugs.

First, while maintaining the PEG_4kDa-b-PLA_2.2kDa concentration in PEG200–600 at 1.0 wt.%, we increased the target loading from 33.0 wt.% (1.0/0.5/98.5 Copolymer/Drug/PEG w/w/w) to 50.0 wt.% (1.0/1.0/98.0 Copolymer/Drug/PEG w/w/w) and hydrated the crystals, aiming for a 1 mg/mL drug concentration in the aqueous medium. As shown in Figure 2a, increasing the target loading in the initial supersaturated solution led to a proportional increase in the amount of drug encapsulated in the resulting polymeric micelles. For instance, using PEG200, when the target loading was increased from 33.0 wt.% to 50.0 wt.% (a 1.5-fold increase), the loading efficiency for paclitaxel, rapamycin, and docetaxel increased by approximately 1.38 times (from 19.7% to 27.1%), 1.35 times (from 20.9% to 28.3%), and 1.54 times (from 19.4% to 29.9%), respectively. Increasing the target loading for all the drugs across all PEG molecular weights resulted in higher loading efficiencies. Interestingly, the average particle size also increased for all drugs and solvent PEGs when the target loading was raised (Figure 2a’). A similar trend was observed when the final drug concentration in the aqueous solution was set at 5 mg/mL, and the target loading was increased from 33.0 wt.% to 50.0 wt.% (Figure S4).

Figure 2.

The “crystallization from supersaturated solution” method can be used for the fabrication of various drug-loaded polymeric micelles, including paclitaxel (PTX), rapamycin (RAP), and docetaxel (DTX). The effect of increasing the target loading from 33.0 wt.% (1.0/0.5/98.5 copolymer/drug/PEG w/w/w) to 50.0 wt.% (1.0/1.0/98.0 copolymer/drug/PEG w/w/w) on the (a) loading efficiency (left column (dark) 33.0 wt.%; right column (light) 50 wt.%) and (a’) particle size distribution (solid lines 33.0 wt.%; dashed lines 50 wt.%) of paclitaxel-, rapamycin-, and docetaxel-loaded PEG_4kDa-b-PLA_2.2kDa micelles made using PEG200–600 at a target drug concentration of 1 mg/mL. The effect of increasing the target drug concentration from 1 mg/mL to 5 mg/mL on the (b) loading efficiency (left column (dark) 1 mg/mL; right column (light) 5 mg/mL) and (b’) particle size distribution (solid lines 1 mg/mL; dashed lines 5 mg/mL) of paclitaxel-, rapamycin-, and docetaxel-loaded PEG_4kDa-b-PLA_2.2kDa micelles made using PEG200–600 at 33.0 wt.% target loading. The effect of prolonged storage of rapamycin-encapsulating crystals at room temperature and 50% RH on (c) encapsulation efficiency and (c’) particle size distribution of the micelles obtained by hydrating them. The effect of prolonged storage of docetaxel-encapsulating crystals at room temperature and 50% RH on (d) encapsulation efficiency and (d’) particle size distribution of the micelles obtained by hydrating them. (e) Ratiometric loading efficiency and (e’) unimodal particle size distribution of paclitaxel/rapamycin co-loaded micelles (33.0 wt.% total target loading, 1 mg/mL total drug concentration) demonstrate the capability of the “crystallization from supersaturated solution” method for co-formulating drug combinations. Statistical analysis: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

Next, we kept the target loading constant at 33.0 wt.% with a PEG_4kDa-b-PLA_2.2kDa concentration in PEG200–600 of 1.0 wt.% and increased the target drug concentration in the aqueous medium from 1 mg/mL to 5 mg/mL. This change resulted in no significant difference in either loading efficiency (Figure 2b) or particle size distribution (Figure 2b’) for any of the drugs or PEG molecular weights. A similar trend was observed when the target loading was fixed at 50.0 wt.% while the target concentration was increased from 1 mg/mL to 5 mg/mL (Figure S4).

Notably, regardless of the type of drug, target loading, or target drug concentration in the aqueous medium, micelles derived from crystals obtained from PEG200 supersaturated solutions consistently exhibited the highest loading efficiency, followed by those from PEG600 and PEG400. These observations suggest that the conditions of the supersaturated solution, such as the concentration of different components, including the drug and the molecular weight of the solvent PEG, are more critical in determining the final properties of drug-loaded micelles than the hydration conditions.

We next explored the potential for prolonged storage of rapamycin- and docetaxel-encapsulating crystals before hydrating them to produce drug-loaded micelles. Similar to the results for paclitaxel (Figure 1g and 1g’), crystals were stored at room temperature for 0, 15, or 30 days and then hydrated to a target drug concentration of 1 mg/mL, based on the initial amount of drug in the solution. Storing the rapamycin (Figure 2c and 2c’) and docetaxel (Figure 2d and 2d’) crystals for 15 or 30 days led to no significant differences (p > 0.05) in encapsulated drug content (Figures 2c and 2d) or particle size distribution (Figures 2c’ and 2d’) compared to day 0.

Finally, we examined the capability of the method for co-loading drug combinations at specific ratios consistent with the feed ratio in the supersaturated solution. Paclitaxel and rapamycin at different ratios (2:1, 1:1, and 1:2 w/w) were added to the solution while maintaining a copolymer concentration of 1.0 wt.% and a target total drug loading of 33.0 wt.%. As shown in Figure 2e, the ratio of the loaded drugs into the resulting micelles from crystals was consistent with the feed ratio in the initial supersaturated solution for all molecular weight solvent PEGs, and uniform micelles were produced (Figure 2e’, Table S1) using the crystallization from supersaturated solution method (total drug concentration = 1 mg/mL). The co-loading can also be performed at a higher total drug concentration (5 mg/mL) while still preserving the feed ratio and producing uniform micelles (Figure S5, Table S2).

For some drugs, such as those used here as model drugs (i.e., paclitaxel, rapamycin, and docetaxel), although the drug can be solubilized in aqueous media by polymeric micelles, incompatibility between these hydrophobic drugs and the micelle core [73] may result in instability, with the drug remaining stable in the micelle core for only a few hours. While the crystallization from supersaturated solution method addresses the stability challenge by producing a stable solid form that can be stored for months prior to hydration, the resulting drug-loaded micelles may remain stable for only a few hours following hydration (see Figure S6). To address this, we previously introduced a platform where an oligo(lactic acid) moiety is attached to the parent drug to form an o(LA)_n-prodrug, which compatibilizes the drug with the micelle core and improves the aqueous stability of the drugs [25, 47, 48, 74, 75]. Here, as the next step, we investigated the capability of the crystallization from supersaturated solution method for the production of polymeric micelles encapsulating o(LA)₈-prodrugs of different drugs, namely paclitaxel, rapamycin, and docetaxel. These formulations are expected not only to be stable for months in solid form but also to remain stable in aqueous media for days.

We first encapsulated o(LA)₈-paclitaxel, o(LA)₈-rapamycin, and o(LA)₈-docetaxel in PEG_4kDa-b-PLA_2.2kDa micelles at a target loading of 33.0 wt.% (1.0/0.5/98.5 Copolymer/Prodrug/PEG w/w/w) and investigated the impact of increasing the target loading to 50.0 wt.% (1.0/1.0/98.0 Copolymer/Prodrug/PEG w/w/w) on loading efficiency and particle size distribution when the micelles were prepared at a 1 mg/mL target drug concentration in the aqueous medium (Figure 3a and 3a’).

Figure 3.

The “crystallization from supersaturated solution” method can be used for loading various prodrugs, including o(LA)₈-paclitaxel (oLA₈-PTX), o(LA)₈-rapamycin (oLA₈-RAP), and o(LA)₈-docetaxel (oLA₈-DTX), into polymeric micelles. The effect of increasing the target loading from 33.0 wt.% (1.0/0.5/98.5 copolymer/prodrug/PEG w/w/w) to 50.0 wt.% (1.0/1.0/98.0 copolymer/prodrug/PEG w/w/w) on the (a) loading efficiency (left column (dark) 33.0 wt.%; right column (light) 50.0 wt.%) and (a’) particle size distribution (solid lines 33.0 wt.%; dashed lines 50.0 wt.%) of o(LA)₈-paclitaxel-, o(LA)₈-rapamycin-, and o(LA)₈-docetaxel-loaded PEG_4kDa-b-PLA_2.2kDa micelles prepared using PEG200–600 at a target drug concentration of 1 mg/mL. The stability of paclitaxel versus o(LA)₈-paclitaxel in PEG_4kDa-b-PLA_2.2kDa micelles at room temperature at a 1 mg/mL target concentration, presented by (b) % encapsulated drug and (b’) particle size distribution over time. The effect of prolonged storage of (c, c’) o(LA)₈-paclitaxel-, (d, d’) o(LA)₈-rapamycin-, and (e, e’) o(LA)₈-docetaxel-encapsulating crystals at room temperature and 50% RH on (c, d, e) encapsulation efficiency and (c’, d’, e’) particle size distribution of the micelles obtained by hydrating them. The dissolution rate of the crystals, represented by (f) % initial o(LA)₈-paclitaxel detected in the solution, and (f’) particle size distribution as a function of vortex time. Statistical analysis: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

The first observation was that using prodrugs instead of the parent drugs for all three model compounds, regardless of the solvent PEG molecular weight, significantly increased the loading efficiency at a given target loading. For example, at a 33.0 wt.% target loading using PEG200, the loading efficiency increased from 19.7% to 38.5% for o(LA)₈-paclitaxel compared to paclitaxel, from 20.9% to 31.8% for o(LA)₈-rapamycin compared to rapamycin, and from 19.4% to 32.5% for o(LA)₈-docetaxel compared to docetaxel. A similar trend was observed across all molecular weight PEG solvents at both 33.0 and 50.0 wt.% target loadings (Figure 3a vs. Figure 2a), possibly due to better compatibility of the o(LA)₈-prodrugs with the PLA block.

Consistent with the parent drugs, prodrug-loaded micelles derived from crystals obtained from PEG200 supersaturated solutions consistently exhibited the highest loading efficiency, followed by those from PEG600 and PEG400. Furthermore, increasing the target loading in the initial supersaturated solution across all PEG molecular weights led to a higher loading efficiency in the resulting polymeric micelles. For instance, using PEG200, when the target loading was increased from 33.0 wt.% to 50.0 wt.% (a 1.5-fold increase), the loading efficiency for o(LA)₈-paclitaxel, o(LA)₈-rapamycin, and o(LA)₈-docetaxel increased by approximately 1.16 times (from 38.5% to 44.6%), 1.55 times (from 31.8% to 49.4%), and 1.41 times (from 32.5% to 45.8%), respectively. The increase in loading efficiency was more significant for o(LA)₈-rapamycin and o(LA)₈-docetaxel compared to o(LA)₈-paclitaxel. Further studies are required to investigate the origin of this observation, but one possible explanation could be that o(LA)₈-paclitaxel already exhibits a significantly higher loading efficiency at 33.0 wt.% target loading, nearing its plateau point for loading. This may limit the additional prodrug that can be encapsulated, even with an increased amount of prodrug in the solution—essentially, it reaches its loading saturation point faster.

Similar to the parent drugs, the average particle size also increased for all prodrugs and solvent PEGs when the target loading was raised (Figure 3a’). It is important to note that the larger average particle size of prodrug-loaded micelles compared to their corresponding drug-loaded ones has been observed in our previous studies using other fabrication methods [25, 47, 48, 74, 75], and is not specific to the crystallization from supersaturated solution method.

A similar trend was observed for all the analyses mentioned above when the final prodrug concentration in the aqueous solution was set at 5 mg/mL, and the target loading was increased from 33.0 wt.% to 50.0 wt.%, with no significant difference in loading efficiency and particle size distribution between the 1 mg/mL and 5 mg/mL target drug concentrations (Figure S7). In other words, similar to the parent drugs, the conditions of the supersaturated solution, such as the concentration of different components, including the drug and the molecular weight of the solvent PEG, are more critical in determining the final properties of prodrug-loaded micelles than the hydration conditions.

Next, we investigated the impact of using prodrugs versus parent drugs on the stability of the encapsulated drug in polymeric micelles in aqueous solution at a 33.0 wt.% target loading and 1 mg/mL target drug concentration. As shown in Figure 3b, using o(LA)₈-paclitaxel, the entire encapsulated drug remained stable for over 24 hours at room temperature, while for the parent drug, encapsulated paclitaxel began to decrease after 2 hours, with only a negligible amount remaining after 6 hours. This pattern was consistent across all molecular weights of solvent PEG. Particle size distribution measurements (Figure 3b’) further confirmed these observations, as o(LA)₈-paclitaxel-encapsulating micelles maintained a uniform particle size distribution throughout the 24-hour period, whereas paclitaxel-encapsulating formulations began to exhibit a non-uniform bimodal particle size distribution after 2 hours at room temperature. Similar to paclitaxel, rapamycin and docetaxel prodrugs demonstrated improved stability in PEG_4kDa-b-PLA_2.2kDa micelles at room temperature in aqueous solution compared to the parent drugs (Figure S8).

We then explored the potential for prolonged storage of o(LA)₈-paclitaxel-, o(LA)₈-rapamycin-, and o(LA)₈-docetaxel-encapsulating crystals before hydrating them to produce prodrug-loaded micelles. Similar to the parent drugs, crystals were stored at room temperature for 0, 15, or 30 days and then hydrated to a target prodrug concentration of 1 mg/mL, based on the initial amount of prodrug in the solution. Storing the o(LA)₈-paclitaxel (Figure 3c and 3c’), o(LA)₈-rapamycin (Figure 3d and 3d’), and o(LA)₈-docetaxel (Figure 3e and 3e’) crystals for 15 or 30 days led to no significant differences (p > 0.05) in encapsulated drug content (Figures 3c, 3d, and 3e) or particle size distribution (Figures 3c’, 3d’, and 3e’) compared to day 0.

Finally, we investigated whether the prodrug-loaded crystals could be rapidly converted into prodrug-loaded micelles upon hydration, similar to the parent drugs. We examined the dissolution rate of o(LA)₈-paclitaxel-encapsulating crystals by vortex-agitating them in water at 5-second intervals, measuring the o(LA)₈-paclitaxel content in the solution (expressed as a percentage of the initial prodrug), and monitoring the formation of polymeric micelles. As shown in Figure 3f, the o(LA)₈-paclitaxel content in the solution reached a plateau after 5 seconds of agitation, demonstrating the rapid dissolution rate of these crystals in water. Particle size distribution analysis indicates that 5 seconds of agitation results in the formation of uniform polymeric micelles (Figure 3f’).

Based on the presented evidence, the crystallization from supersaturated solution method was concluded to be effective for producing prodrug-containing crystals, which can be stored for prolonged periods and subsequently converted into drug-loaded polymeric micelles upon hydration and slight agitation, yielding a stable aqueous formulation.

It is important to note that while the current study provides 30-day stability data, definitive conclusions about the long-term storage potential of the proposed solid form (>6 months) require further investigation and are a focus of future studies. However, we anticipate that long-term storage of the proposed system is achievable based on two primary factors: (1) the absence of water during the manufacturing process, which significantly minimizes hydrolytic degradation and restricts exposure to residual moisture—a primary source of instability; and (2) the crystalline nature of the solid form, which represents a thermodynamically stable state that is inherently resistant to environmental fluctuations. Supported by these considerations, theoretical principles, and the 30-day empirical data, we hypothesize that these drug-loaded crystals can maintain their structural and functional integrity over extended periods when stored under appropriate conditions.

3.3. The formulation process impacts the cytotoxicity of the final product

The method proposed in this study allows for the production of drug-loaded polymeric micelles from crystals obtained from supersaturated solutions using PEG as the solvent. The use of different PEG molecular weights or varying target loadings results in differences in the physicochemical properties of the formulations, such as loading efficiency and/or particle size distribution. We also investigated whether altering process parameters, such as the molecular weight of the solvent PEG or target loading, would affect the cytotoxicity of the formulations.

To explore this, paclitaxel-loaded PEG_4kDa-b-PLA_2.2kDa micelles were produced using crystals obtained from PEG200–600 supersaturated solutions (1 wt.% copolymer) at 33.0 or 50.0 wt.% target loading, and their cytotoxicity against 4T1 and MDA-MB-231 breast cancer cell lines was assessed at different paclitaxel concentrations. According to our results (Figure 4), the molecular weight of the solvent PEG has a significant impact on the cytotoxicity of the formulations against both cell lines, while the impact of target loading is less pronounced. Formulations produced using PEG200 showed the highest cytotoxicity against both cell lines across almost all concentrations tested. Paclitaxel-loaded micelles made with PEG200, regardless of the target loading, were able to kill more than 80% of 4T1 and MDA-MB-231 cells at paclitaxel concentrations as low as 0.16 and 0.08 mg/mL, respectively. In contrast, when PEG600 was used, a paclitaxel concentration greater than 1.25 mg/mL was required to achieve comparable cytotoxicity. Overall, the data suggest that the cytotoxicity of the samples prepared using PEG200 is the highest, followed by PEG400, and then PEG600.

Figure 4.

The impact of solvent PEG molecular weight and target loading on the cytotoxicity of paclitaxel (PTX)-loaded PEG_4kDa-b-PLA_2.2kDa micelles obtained by the crystallization from supersaturated solution method. The viability of 4T1 cells as a function of paclitaxel concentration at a target loading of (a) 33.0 wt.% and (a’) 50.0 wt.%. The viability of MDA-MB-231 cells as a function of paclitaxel concentration at a target loading of (b) 33.0 wt.% and (b’) 50.0 wt.%. Statistical analysis: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

This observed cytotoxicity cannot be solely attributed to the impact of PEG molecular weight on loading efficiency, as PEG600 samples consistently showed higher loading efficiency than PEG400 samples. Although PEG itself can have concentration-dependent cytotoxicity [76, 77], the observed effects are unlikely to result from the cytotoxic effect of the solvent PEG for several reasons:

1. While a substantial amount of PEG is used as the solvent in the proposed method, the majority is not incorporated into the final formulation, as the crystals are separated from the PEG solution and then hydrated to produce the drug-loaded PEG-b-PLA micelles. This process may leave behind residual amounts of low molecular weight PEG, either mixed with or incorporated into the crystals, but the PEG content in the final formulation is expected to be negligible. Considering the highest PTX concentration used to measure IC50 (2.5 mg/mL), the lowest observed loading efficiency of PTX (10% w/w, associated with crystallization from PEG400 at 33 wt.% target loading), and the 20–30% molar fraction of residual PEG400 in the final crystals, these values collectively translate to less than 1 mg/mL of PEG400 (or other low molecular weight PEGs) in the cell culture media, even at the highest PTX concentration tested.

2. The IC50 range reported for PEG200 (e.g., 300 mg/mL against Caco-2 cells [76]) is significantly higher than the concentrations of PEG that the cells would be exposed to in this context (e.g. 1 mg/mL calculated for PEG400 in the previous example). Such high PEG concentrations would only be encountered if the cells were exposed to the entire supersaturated solution.

3. Comparable cytotoxicity has been reported for low molecular weight PEGs [76], but the significantly different cytotoxicity pattern observed here suggests that the observed cytotoxicity may not be due to the direct effect of PEG on the cells.

We hypothesize that the impact of the solvent PEG molecular weight on the cytotoxicity of the formulations may be an indirect result of its influence on the structure of the crystals and the resulting micelles. For instance, we demonstrated how changing the molecular weight of solvent PEG can affect the melting onset of the resulting crystals (Figure 1b). Incorporation of short PEG chains into the crystals (20–30% molar fraction), and subsequently into the micelle structure, may influence molecular assembly and drug release. Considering the stability of paclitaxel-loaded PEG_4kDa-b-PLA_2.2kDa micelles (Figure 3b), the steepest drop in encapsulated paclitaxel over time was observed in micelles prepared using PEG200, followed by those made with PEG400 and PEG600. This suggests that micelles prepared with lower molecular weight PEG may provide higher levels of free unencapsulated paclitaxel at certain time points, potentially enhancing cytotoxicity. However, drawing definitive conclusions about the underlying mechanisms responsible for these observations requires further investigation, which is part of our planned future studies.

4. Conclusions

In conclusion, the “crystallization from supersaturated solution” method provides a robust and versatile platform for producing stable, drug-loaded polymeric micelles. This technique effectively encapsulates various parent drugs and their oligo(lactic acid)-prodrugs within PEG-b-PLA micelles, achieving high loading efficiency while maintaining drug stability over extended storage periods. The method’s flexibility in using different molecular weights of PEG as solvents and adjusting formulation parameters, such as target loading, allows for precise control over micelle properties, including particle size and loading efficiency. Furthermore, the rapid dissolution of crystals into uniform micelles upon hydration, along with the elimination of the need for freeze-drying, underscores the method’s practicality for large-scale production and prolonged storage. Overall, this approach holds significant promise for advancing the formulation of hydrophobic drugs in micellar systems, particularly by enhancing their stability and storage in solid form.