Core-Shell PLGA Nanoparticles: In Vitro Evaluation of System Integrity

1. Introduction

Drug delivery by nanoparticles may considerably improve the safety and selectivity of many drugs by altering their pharmacokinetics and tissue distribution using the endogenous mechanisms. The first success of this strategy was achieved in antibiotic therapy by exploiting the known ability of colloidal particles, especially hydrophobic, to accumulate in the resident macrophages of the macrophage-rich organs, such as the liver, spleen, lungs, etc. (organs of the mononuclear phagocytic system, MPS) [1]. Loading of antibiotics into the nanoparticles improved their penetration into macrophages, which serve as a niche for pathogens, and thus considerably enhanced their efficacy [2]. Later, the discovery of the enhanced permeability and retention (EPR) effect, which enables the extravasation and retention of colloids in the tumor due to the differences between the vascular systems of solid tumors and normal tissues, inspired enormous interest in nanomedicines for cancer chemotherapy. Indeed, the preclinical studies provided numerous instances of the improved antitumor effect of the nanoparticle-bound drugs due to their enhanced trafficking to tumors. However, it was soon realized that the accumulation of nanoparticles in the MPS organs, although useful for drug delivery to these organs, considerably interferes with their ability to target other sites in the body, decreasing blood circulation half-life and targeting capability [3]. Moreover, the large heterogeneity of the EPR effect observed on both intra- and interindividual levels leads to heterogeneous therapeutic outcomes [4,5,6]. Indeed, the advantage of the approved nanoformulations of classical anticancer drugs designed to exploit the EPR effect, such as Caelyx®, Abraxane®, Genexol-PM® or Onivyde®, was achieved more due to their enhanced safety than to any improvement in antitumor effect or patient survival [6].

These disappointing results gave rise to numerous new trends in pharmaceutical nanotechnology, starting from the use of theranostics for more insightful pre-selection of patients who could benefit from nanomedicines [7,8] to the development of delivery systems for the next generations of drugs (immune check-point inhibitors [9], nanoadjuvants [10], RNA-based therapeutics [11], drug combinations, etc.) and new types of nanocarriers [12,13].

Among the latter, core-shell nanoparticles draw considerable attention; over the last few decades, they have been extensively studied for many applications, including in the biomedical field (reviewed in [14,15,16,17,18]). This technology indeed enables the optimization of plain “first-generation” nanoparticles by improving their in vitro and in vivo stability, as well as offering the possibility to load them with both hydrophilic and hydrophobic agents and tune their release profiles. Moreover, hydrophilic shells may shield nanoparticles from macrophages (the stealth effect), thus optimizing their biodistribution profile and enhancing their capability to reach other organs besides the MPS. Shells rich in functional groups could also enhance the feasibility of nanoparticle surface modification and conjugation with drugs or ligands for targeted delivery.

The biodegradable and biocompatible poly(lactide-co-glycolic) acid (PLGA) nanoparticles, one of the most popular nanocarriers in drug delivery [19], appear to be good candidates for this approach. These nanoparticles are relatively hydrophobic, and therefore their inherent disadvantage is the massive accumulation in macrophage-rich organs. Moreover, the attachment of bioligands that could enable active tumor targeting is difficult because the reactive groups on the surface of the PLGA nanoparticles are few (limited to the end carboxylic groups of the polymer) and sterically hindered. The advantages of core-shell PLGA nanocarriers, such as steric stabilization, modulation of drug release profiles, and facilitated surface functionalization, have been demonstrated in numerous studies using various shell-forming agents such as PEG or PEG derivatives, chitosan, alginate, albumin, collagen, silk fibroin, etc. were demonstrated in a number of studies [14,20,21,22,23,24]. The methods used to create the shells primarily involve the adsorption of shell-forming polymers onto nanoparticle surfaces, which can be performed either during their formation or by incubating preformed nanoparticles in a solution of the shell-forming agent [22]. Sometimes, cross-linking of the shell [25] or its conjugation with the PLGA core [26,27] is used to improve system stability.

Clearly, the integrity of the core-shell structure is essential for the successful performance of the system in the biological environment. At the same time, this integrity is not always well elucidated.

The objective of the present study was to evaluate the integrity of nanoparticles with a PLGA core and shells composed of different types of polymers in cell-free media and upon their interaction with cells, using Gl261 murine glioma cells as the model. The shell-forming polymers included the non-ionic polymeric surfactant poloxamer 188 (PLGA/P188 NP), DIVEMA (a copolymer of divinyl ether with maleic anhydride, PLGA/DIVEMA NP), and human serum albumin (PLGA/HSA NP). The methods of preparing the core-shell nanoparticles were developed in our previous studies [28,29,30]. Due to their structure, the selected polymers create a hydrophilic shell on the nanoparticle surface and improve their colloidal stability. A considerable number of functional groups exhibited by HSA (NH₂, COOH, and SH) and DIVEMA (COOH) shall also facilitate the binding of ligands to the core-shell nanocarrier.

These polymers are well characterized and appear to be suitable functional excipients. Both poloxamer 188 and HSA are effective surfactants and stabilizers with a long history of application in pharmaceutical products [31,32]. Moreover, coating PLGA nanoparticles with poloxamer 188 altered their biodistribution and enabled brain delivery of encapsulated doxorubicin; importantly, this formulation also improved the doxorubicin safety profile compared to the free drug [33,34,35]. The relatively low toxicity of low-molecular weight DIVEMA used in this study has been demonstrated in clinical studies [36]. A number of studies have revealed the broad spectrum of DIVEMA’s biological activities, such as antiviral and antibacterial activity and induction of immune responses to tumors [37,38].

Apart from the physicochemical evaluation, the core-shell system integrity was monitored by fluorescent spectroscopy using PLGA nanoparticles, the core and shell of which were labeled with pairs of fluorescent dyes with donor-acceptor properties (Cyanine5/Cyanine3 and Cyanine5/rhodamine B). These dye pairs are capable of exhibiting the Förster resonance energy transfer (FRET) phenomenon. This distance-dependent phenomenon enables the evaluation of core and shell integrity based on changes in donor and acceptor fluorescence intensity, which are observed only when the molecules are localized in close proximity [39,40]. Internalization of the core-shell nanoparticles in Gl261 cells was investigated using confocal laser fluorescent microscopy.

2. Materials and Methods

2.1. Materials

PLGA (Purasorb® PDLG 5004A, 50/50, η=0.4 dL/g, acid terminated) was purchased from Corbion Biomaterials (the Netherlands). N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDC), N,N′-dicyclohexylcarbodiimide (DCC), 4-(dimethylamino)pyridine (DMAP), N–hydroxysuccinimide (NHS), diisopropylethylamine (DIPEA), human serum albumin (HSA), polyvinyl alcohol (PVA, 9-10 kDa, 80% hydrolyzed), poloxamer 188 (P188), 2,2’-azobisisobutyronitrile (AIBN), rhodamine B (RhB) and rhodamine B isothiocyanate (RhBITC) were purchased from Sigma-Aldrich (USA). The reactive derivatives of the fluorescent dyes Cyanine5 (Cy5) amine and Cyanine3 (Cy3) amine were purchased from Lumiprobe (Russia). D-mannitol was purchased from Dia-M (Russia). All other chemicals were of analytical grade.

2.2. Preparation of Core-Shell PLGA Nanoparticle

2.2.1. Preparation of Dual-Labeled Nanoparticles with a PLGA Core and HSA Shell (PLGA-Cy5/HSA-RhBITC NP)

For fluorescence labeling, conjugates of PLGA with Cyanine5 amine (PLGA-Cy5) and HSA with rhodamine B isothiocyanate (HSA-RhBITC) were synthesized. PLGA-Cy5 was synthesized by conjugating the PLGA acid terminal group with the water-soluble amine derivative of the Cyanine5 dye (Cy5, λ_ex 651 nm, λ_em 670 nm) using NHS/EDC coupling reaction in the presence of DIPEA, as previously described in [28,41,42]. Briefly, solutions of Cy5 amine, DIPEA, EDC, NHS, and PLGA in dichloromethane (DCM) were combined, and the reaction mixture was stirred in the dark (48 h, ambient temperature). The resulting solution was washed three times with water and a water/methanol (1:1) mixture. The organic phase was then dried over anhydrous sodium sulfate, and the solvent was removed by evaporation. The residue was dissolved in ethyl acetate, and the polymer was precipitated by adding hexane (10-fold volume). The Cy5 content in the dried conjugate was measured spectrophotometrically. The PLGA-Cy5 conjugate with a dye-to-polymer ratio of 1:600 (w/w) was used in the experiments.

The PLGA-Cy5 conjugate was analyzed by gel permeation chromatography as described in [28]. Briefly, the analysis was performed using a Waters HPLC system equipped with a set of Styrogel HR5E and HR4E columns, a refractive index detector (Waters 2414 RI Detector) and a UV detector (Milton Roy UV-detector 3100 model, λ 264 nm). Tetrahydrofuran was used as the solvent and eluent at a flow rate of 1.00 mL/min. Data were analyzed using Z-lab software.

HSA was conjugated with rhodamine B isothiocyanate following the method described by Yang et al. [43]. Briefly, rhodamine B isothiocyanate was dissolved in DMSO, and this solution was added dropwise to a 1% HSA solution (0.15 mM) in a 9:1 v/v mixture of saline (0.15 М NaCl) and buffer (0.15 М NaHCO₃, pH 9.0). The RhBITC/HSA molar ratio was 5:1. The obtained solution was stirred overnight under cooling. A NH₄Cl solution (50 mM) was then added to stop the reaction, followed by an additional hour of stirring. The resulting fluorescently labeled HSA was purified by dialysis against phosphate-buffered saline (PBS, 0.01 М PBS, pH 7.4) at 4 °C in the dark. The concentration of RhBITC was determined by calculating the fluorophore/protein molar ratio (F/P):

$$\mathit{C}\left(\mathit{H}\mathit{S}\mathit{A}-\mathit{R}\mathit{h}\mathit{B}\mathit{I}\mathit{T}\mathit{C}\right)={\displaystyle \frac{{\mathit{A}}_{280}-\mathit{C}\mathit{F}\times {\mathit{A}}_{558}}{{\mathit{\epsilon }}_{\mathit{H}\mathit{S}\mathit{A}}}}\times \mathit{d}\mathit{i}\mathit{l}\mathit{u}\mathit{t}\mathit{i}\mathit{o}\mathit{n}\mathit{f}\mathit{a}\mathit{c}\mathit{t}\mathit{o}\mathit{r},$$

(1)

$${\displaystyle \frac{F}{P}}={\displaystyle \frac{A_{558}}{{\epsilon }_{RhBITC}\times C\left(HSA-RhBITC\right)}}\times dilutionfactor$$

(2)

where A₅₅₈ represents the absorbance (A) of RhBITC at 558 nm; A₂₈₀ is the absorbance (A) of HSA-RhBITC at 280 nm; ε_HSA is the molar extinction coefficient of HSA; ε_RhBITC is the molar extinction coefficient of RhBITC; CF is the correction factor to account for dye absorbance at 280 nm; dilution factor accounts for any dilution of the protein:dye sample for absorbance measurement. The HSA-RhBITC conjugate used in further experiments had an F/P ratio of 1.25.

Core-shell nanoparticles were prepared by the high-pressure homogenization — solvent evaporation technique (o/w). The HSA shell was formed using three different techniques: by using HSA as a stabilizer during nanoparticle formation (interfacial embedding of HSA) or by adsorbing or conjugating HSA onto previously prepared PLGA nanoparticles without a shell (Supplementary Figure S1) [29].

Interfacial embedding of HSA. A solution of a 1:1 mixture of PLGA and PLGA-Cy5 (150 mg + 150 mg) in dichloromethane was emulsified with an aqueous 0.5% HSA-RhBITC solution (30 mL) using a high-shear homogenizer (Ultra-Turrax T18 Basic, IKA-Werke, GmbH, Staufen, Germany). This coarse emulsion was further homogenized at high pressure (15,000 psi, Microfluidizer L10, Microfluidics, Newton, MA, USA), and the solvent was removed by vacuum evaporation. The resulting suspension was then filtered through a sintered glass filter and freeze-dried using 2.5% (w/v) D-mannitol as a cryoprotectant. The freeze-dried nanoparticles were stored at 4 °C. Unbound HSA was removed by repeated washing and centrifugation of the nanosuspension (48,060⨯g, 20 °C, 20 min). A portion of the sample was subjected to shell cross-linking with glutaraldehyde to improve shell stability [29]. Based on the findings of Langer et al. [44], 117 µL of an 8% glutaraldehyde solution (~0.1 mmol) per 1 mg of HSA (15 µmol) was added to the reaction mixture to ensure reliable HSA cross-linking.

Adsorption of HSA. To prepare PLGA/HSA nanoparticles by adsorption, PLGA-Cy5 nanoparticles were first prepared using 1% polyvinyl alcohol (PVA, 9–10 kDa, 80% hydrolyzed, Merck, Germany) as a stabilizer, as described in [41]. Then the nanoparticles were washed with water and resuspended in a 0.5% aqueous HSA-RhBITC solution, followed by 30 min of incubation under continuous stirring (37 °C, 130 rpm). Unbound HSA-RhBITC was removed by repeated centrifugation (48,060⨯g, 18 °C, 10 min). The supernatants were successively analyzed by spectrophotometry for the presence of unbound HSA. A portion of the sample was subjected to HSA cross-linking [29] followed by freeze-drying, as described above.

Conjugation of HSA. HSA amino groups were conjugated with the carboxyl end groups of PLGA through a coupling reaction using 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC) in phosphate buffer (1 h, 0.05 M, pH 6.8) [45]. An indirect acid-base titration in MES buffer (0.05 M, рН 5) revealed that the PLGA-Cy5 nanoparticles contained 2.4 µmol-eq/mL of functional carboxyl groups. A 10-fold molar excess of EDC (24 µmol-eq/mL, 0.46 mg/mL NP) and NHS (10-fold excess: 24 µmol-eq/mL, 0.46 mg/mL NP) was then added to the nanoparticle suspension. The mixture was incubated at 37 °С for 2 hours under stirring (90 rpm), washed by centrifugation (48,060×g, 30 min, 25 °С), and resuspended in MES buffer (0.05M, рН 5). Activated nanoparticles were then incubated in a 1% HSA-RhBITC solution under continuous stirring (90 rpm, 1 h, 37°С). Unconjugated HSA was removed by two-step washing (centrifugation at 48,060×g, 10 min, 18 °С). The precipitated nanoparticles were resuspended in HEPES buffer (0.1 M, pH 7.4). The HSA content in the supernatants was assessed spectrophotometrically based on the fluorescent properties of the labeled polymer. The nanosuspensions were then freeze-dried using 2.5% mannitol as a cryoprotectant.

2.2.2. Preparation of Dual-Labeled Nanoparticles with a PLGA Core and DIVEMA Shell (PLGA-Cy5/DIVEMA-Cy3 NP)

The alternating copolymer of divinyl ether and maleic anhydride (DIVEMA) was synthesized by radical cyclopolymerization of the corresponding monomers in dry acetone, using 2,2’-azobisisobutyronitrile (AIBN) as the initiator and tetrahyfrofuran as the chain transfer agent [28,46]. The polymer structure was confirmed by IR and NMR spectroscopy, as previously described in [28]. A sample of DIVEMA copolymer with a molecular weight (MW) of 23 kDa and a characteristic viscosity of 0.16 dL/g (measured in borate buffer containing 0.2 M NaCl) was used in the following experiments.

Fluorescent labeling of DIVEMA was achieved by conjugating the carboxyl groups with the water-soluble amine derivative of Cyanine3 dye (Cy3, λ_ex 555 nm, λ_em 570 nm) via a carbodiimide coupling reaction, as described in [28]. First, the DIVEMA carboxyl groups (2.26 mmol DIVEMA solution in acetone [600 mg in 30 mL of acetone]) were activated by NHS (23.4 mg, 0.20 mmol in 1 mL of acetone) for 30 min under continuous stirring in the dark. A dicyclohexylcarbodiimide (DCC) solution (14.1 mg [68.3 μmol] in 0.5 mL of acetone) and a Cy3 amine solution in a mixed organic solvent (18 mg [28.7 μmol] in 1 mL acetone and 0.5 ml ethanol) were then added. The mixture was incubated for 3 hours under constant stirring, after which the conjugate was purified by three-fold precipitation in diethyl ether (ten times the volume). The precipitate was separated and washed with chloroform. The absence of unbound Cy3 in the wash solvent was monitored by the UV spectroscopy at 543 nm. The product was dried under vacuum to a constant weight. The resulting purified DIVEMA-Cy3 conjugate was stored in an evacuated flask or under an inert gas atmosphere. The conjugate formation was confirmed by gel permeation chromatography using a refractive index detector and a UV detector, as previously described [28]. The Cy3 content was quantified by UV spectroscopy at λ_max 552 nm. The conjugate containing 2.7% w/w of the dye (dye-to-polymer ratio of 1:40, w/w) was used in the experiments.

Preparation of PLGA-Cy5/DIVEMA-Cy3 nanoparticles by high pressure homogenization—solvent evaporation technique (PLGA-DIVEMA-H). A 1:1 mixture of PLGA and PLGA-Cy5 polymers (2 x 300 mg [2 x 2.3 mmol]) was dissolved in 7.2 mL of DCM. The resulting solution was mixed with a solution of DIVEMA-Cy3 in acetone (71.4 mg [0.268 mmol] in 4.8 mL), and the organic phase was added to a 0.5% aqueous PVA solution (60 mL). The mixture was first emulsified using a high-shear homogenizer (Ultra-Turrax T18, 23,600 rpm) and then subjected to high-pressure homogenization (Microfluidics M-110P, 15000 psi). Afterwards, the organic solvent was removed under vacuum, and the resulting suspension was filtered through a glass sintered filter (pore size 90-150 μm) and freeze-dried using 2.5% (w/v) of D-mannitol as a cryoprotectant. After lyophilization, the nanoparticles were resuspended in purified water and subjected to centrifugation: first at 10,000×g (10 min, 5 °C) to remove agglomerates, and then at 48,060×g (two times, 30 min, 5 °C). The resulting nanosuspension was freeze-dried with 2.5%mannitol as a cryoprotectant and stored at 4 °С. Plain PLGA nanoparticles (PLGA without a shell) were obtained in a similar way, with the addition of 4.8 mL of acetone without DIVEMA to the polymer solution.

Preparation of PLGA-Cy5/DIVEMA-Cy3 nanoparticles by nanoprecipitation (PLGA-DIVEMA-N). The PLGA/PLGA-Cy5 (1:1) mixture (60 mg; 0.46 mmol) was dissolved in a mixture of acetonitrile and acetone (1:2 v/v). Then, 24.7 mg (93 μmol) of DIVEMA-Cy3 was added, and the resulting solution was added dropwise to 60 mL of dH₂O under stirring (1,000 rpm) and incubated for 4 hours. The organic solvents were evaporated under vacuum. The resulting suspension was filtered through a sintered glass filter (pore diameter 90-150 μm) and freeze-dried with the addition of 2.5% (w/v) mannitol as a cryoprotectant. The resulting suspension was passed through a glass sintered filter with a pore diameter of 90-150 μm (Por 1); 2.5% mannitol was added as cryoprotectant. Plain PLGA nanoparticles (PLGA without a shell) were obtained similarly, but without DIVEMA and using 0.5% aqueous PVA solution instead of purified water.

2.2.3. Preparation of PLGA Nanoparticles Coated with Fluorescently Labeled Poloxamer 188

Synthesis of poloxamer 188 conjugate with RhB (P188-RhB). The conjugate of poloxamer 188 with rhodamine B (P188-RhB) was synthesized and characterized as reported previously [30]. Briefly, a solution of P188 (1,000 mg, 0.12 mmol), RhB (126 mg, 0.26 mmol), EDC (55 mg, 0.29 mmol), and 4-(dimethylamino)pyridine (DMAP, 59 mg, 0.48 mmol) in dimethylformamide (DMF, 15 mL) was stirred at ambient temperature for 3 days in the dark. Afterwards, the mixture was diluted with Et₂O, and the crude product was precipitated overnight in a freezer at −20 °C. The precipitate was washed three times with a mixture of DMF/Et₂O (1:1) while cooling. The crude product was then purified by gel filtration using a Sephadex G25 column. The combined fractions containing the product were then lyophilized. The absence of unbound dye was confirmed by thin-layer chromatography using a mixture of i-PrOH/H₂O (3:5) as the eluent. The stability of the conjugate in PBS (pH 7.4), DMEM, or human plasma was also evaluated by TLC (incubation for 2 h at 37 °C). The conjugate containing 4.49% w/w of the dye was used in the experiments.

Evaluation of poloxamer 188 adsorption on the PLGA nanoparticles’ surface. The PLGA-Cy5-P188-RhB nanoparticles were prepared by incubating the PLGA-Cy5 NP (prepared by the high-pressure homogenization—solvent evaporation technique as described above) in a 1% aqueous solution of P188-RhB for 30 minutes.

The evaluation of P188-RhB adsorption was carried out as reported previously [30]. To analyze the P188-RhB content in one vial, the nanoparticle lyophilizate was resuspended in 1.5 mL of the 1% P188-RhB solution. The suspension was incubated at room temperature while stirring at 130 rpm. The aliquots (400 μL) collected after 30, 90 and 180 minutes were centrifuged to separate the PLGA NP (48,068×g, 30 min, 5 °C). The precipitated nanoparticles were washed with water (1 mL), dried, and dissolved in DMSO (2 mL). The P188-RhB content in the resulting solution was determined spectrophotometrically at λ_max 555 nm (UV-1900i spectrophotometer, Shimadzu Corp, Japan) using the calibration curve. The amount of adsorbed P188-RhB in mg/m² of the PLGA NP surface was calculated as:

$$S={\displaystyle \frac{c{V}_s{\rho }_{PLGA}d}{6m}},$$

(3)

where m is the mass of PLGA nanoparticles in a vial, mg; с is the concentration of the P188 RhB conjugate in the analyzed sample, determined from the calibration curve, mg/mL; V_s is the sample volume, 2 mL; ρ_PLGA is the PLGA density, 1.2 g/cm³; d is the nanoparticle diameter, nm.

2.3. Characterization of Nanoparticles

The mean hydrodynamic size and polydispersity index (PDI) of the nanoparticles were measured by dynamic light scattering (DLS) using the Zetasizer NanoZS (5 mW He-Ne laser, operating wavelength 633 nm, Malvern Instruments, Malvern, UK). The zeta-potential was determined by electrophoretic light scattering (ELS) in a U-shaped disposable folded capillary cell using the same instrument. Measurements of all samples were performed in triplicate. The samples were diluted to a concentration of 0.2 mg/mL polymer in Milli-Q water.

The total PLGA concentration in all samples was quantified using capillary electrophoresis (CAPEL 105M, Russia) as described previously [47]. The amount of PLGA was assessed by the amount of lactate formed after the nanoparticle hydrolysis in 1 M NaOH (37 °C, 24 h, continuous shaking at 200 rpm). The content of the –COOH groups in PLGA nanoparticles was determined by potentiometric titration [47].

The dye content in the nanoparticles was quantified by UV spectroscopy (Shimadzu UV-1800, Japan). The Cy3 and Cy5 content in the nanoparticles was measured at 553 nm and 649 nm for Cy3 and Cy5, respectively, after dissolving the core-shell nanoparticles in DMSO. The HSA-RhBITC shell content in the nanoparticles was calculated indirectly by the difference between the content of HSA-RhBITC (determined by UV spectroscopy at λ_max 558 nm in water) added and washed in the process of nanoparticle preparation as:

$${m\left(HSA\right)}_{NP}={m\left(HSA\right)}_{added}-{m\left(HSA\right)}_{washed},$$

(4)

where m(HSA)_NP is the content of HSA-RhBITC on the nanoparticle surface; m(HSA)_added is the content of HSA-RhBITC added during the nanoparticle synthesis; m(HSA)_washed is the content of HSA-RhBITC washed during the nanoparticle synthesis.

The morphology of the core-shell nanoparticles was studied by transmission electron microscopy (TEM) using a JEOL JEM-1400 electron microscope (JEOL, Japan) at an accelerating voltage of 120 kV. A diluted aqueous nanosuspension was applied to Cu grids coated with a Formvar polymer film, and the solvent was allowed to dry completely. To improve image quality, additional negative contrasting of the nanoparticles was performed using the UranyLess EM Stain (Electron Microscopy Sciences, USA).

The TEM images of the PLGA/HSA-IE cross-linked nanoparticles were also obtained using the Tecnai™ 12 G2 BioTwin Spirit microscope (ThermoFisher Scientific, USA) equipped with Eagle 4K detector (ThermoFisher Scientific, USA) at an accelerating voltage of 120 kV. The samples for the TEM study were prepared using negative staining. The 300 mesh Cu grids coated with a carbon support film were subjected to glow discharge in PELCO easiGlow™ for 30 sec with a 25-mA plasma current. The sample solution (3 μL) was applied to the grid for 30 seconds, after which the grid was washed with a droplet of distilled water and stained with a 1% aqueous solution of uranyl acetate for 60 seconds.

2.4. Evaluation of Fluorescent Properties

The quantum yield and brightness of the fluorescently labeled nanoparticles were assessed as reported previously [42]. The light absorption values and the integrated fluorescence intensities were determined for each nanoparticle type.

The quantum yields of the nanoparticles were calculated as:

$$\phi ={\phi }_{st}{\displaystyle \frac{tg\alpha }{tg{\alpha }_{st}}}{\left({\displaystyle \frac{n}{n_{st}}}\right)}^2,$$

(5)

where φ and φ_st are the quantum yield of the dye-labeled nanoparticles and standard, respectively; tg α and tg α_st are tangents of the slope of the dependences of the integrated fluorescence intensity on light absorption for the test compound and standard, respectively; n and n_st are refractive indices of the media in which the measurements were made.

The brightness of the nanoparticles was calculated as [48]:

$$B=\phi ·\epsilon ·N,$$

(6)

where ε is the extinction coefficient of the dye determined via Beer-Lambert law, l∙mol⁻¹∙cm⁻¹; φ is the quantum yield; N is thenumber of fluorescent dye molecules encapsulated inside the nanoparticles.

RhB (φ=0.7, $\epsilon$=106,000 M^-1 cm^-1 [49]) was used as a reference to determine the quantum yield and brightness of Cy3-, RhB- or RhBITC-labeled core-shell nanoparticles. Cy5 (φ=0.2, $\epsilon$=250,000 M^-1 cm^-1 [50]) was used as a reference to determine the quantum yield and brightness of Cy5-labeled nanoparticles.

Fluorescence spectroscopy FRET studies. The fluorescence spectra were measured for nanoparticles with either one or two fluorescent labels. The following types of nanoparticles were used for the study: PLGA-Cy5/HSA NP, PLGA/HSA-RhBITC NP, PLGA-Cy5/HSA-RhBITC NP, PLGA/DIVEMA-Cy3 NP, PLGA-Cy5/DIVEMA-Cy3 NP, PLGA-Cy5/P188 NP, PLGA/P188-RhB NP, and PLGA-Cy5/P188-RhB. Nanoparticle suspensions in water were prepared so that the fluorescent label content in nanoparticles with one label was equal to the content of the same dye in nanoparticles with two fluorescent labels. The fluorescence spectra of the PLGA/HSA NP and PLGA/P188 NP were recorded at λ_ex = 550 nm over the emission range of λ_em 565-800 nm, and at λ_ex = 644 nm over of λ_em 655-800. The fluorescence spectra for the PLGA/DIVEMA NP were taken at λ_ex = 530 nm over the emission range of λ_em 545-740 nm, and at λ_ex = 630 nm over of λ_em 640-800 nm.

2.5. Evaluation of Core-Shell Nanoparticle Stability by Physicochemical Methods

The colloidal stability of the nanoparticles was assessed at different time points by measuring their hydrodynamic diameters (DLS) at 37 °C in simulated physiological media: PBS (pH 7.4; 0.15 M) and 4.5% HSA solution in PBS.

The stability of the core–shell structures was also evaluated by assessing the amount of fluorescently labeled shell remaining on the nanoparticle surface through fluorescence measurements. The nanosuspensions (100 μg/mL as PLGA) were incubated in PBS or PBS containing 4.5% HSA. At selected time points (0, 0.5, 1, 2, 4, and 6 h), the 400-μL aliquots were taken, and then the fluorescence spectra of the suspension were recorded. Then the aliquots were centrifuged (48,060×g, 18 °C, 20 min) to precipitate the nanoparticles and the fluorescence intensity of each supernatant was measured. The fluorescence spectra were recorded for HSA-RhBITC and P188-RhB (λ_ex 560 nm, λ_em 570-800 nm), PLGA-Cy5 (λ_ex 649 nm, λ_em 660-800 nm), and DIVEMA-Cy3 (λ_ex 530 nm, λ_em 540-800 nm). The percentage of the detached shell at each time point was calculated as the ratio of the fluorescence intensities in the supernatant to those in the nanosuspension.

2.6. Cells

Gl261 murine glioblastoma cells were purchased from the American Type Culture Collection (ATCC). The cells were cultivated in DMEM supplemented with GlutaMax™ (2 mM, Gibco), 10 % fetal bovine serum (Biowest, US) and penicillin (100 U/mL), and streptomycin 100 g/mL (Gibco). The cells were cultured under standard conditions (37° C, 5% CO₂) and used between passages 5-9.

2.7. Investigation of Nanoparticle Internalization by Confocal Microscopy

Images were obtained using a Nikon A1R MP+ multiphoton confocal microscope (Nikon Instr., INC, USA) equipped with four lasers: 405 nm (λ_em 425-475 nm), 488 nm (λ_em 500-550 nm), 561 nm (λ_em 570-620 nm), and 638 nm (λ_em 663-738 nm), and Plan Apo 20x/0,75 Dic N, Apo IR 60x/1.27 WI, and Apo TIRF 60x/1,49 oil Dic objectives. Data were analyzed using NIS-Elements AR Software. Gl261 cells were seeded onto confocal 35-mm high coverslip dishes (Ibidi, USA) (70x10³ cells) 48 hours prior to the study. The cells were incubated with the samples for 15, 30 and 45 min (final concentration 100 μg/mL in serum-containing medium). The cells were then washed three times with DPBS to remove nanoparticles from the cell surface. Lysosomes were stained using LysoTracker Green DND26 (50 nМ, Molecular Probes, USA). Plates were transferred to the microscope stage for live-cell imaging.

To assess the core-shell structure integrity of the nanoparticles as well as their colocalization with lysosomes (for the evaluation of their major NP internalization pathway and intracellular localization), Manders’ overlap coefficient (MOC) values were calculated. MOC values were calculated within the selected ROIs — three random ROIs per image. Colocalization plots were built to compare the core-shell integrity of the nanoparticles (HSA, DIVEMA or P188 shells with PLGA cores) obtained by different techniques.

2.8. Investigation of Nanoparticle Uptake by Flow Cytometry

The Gl261 cells were seeded into 24-well plates (Corning) (100x10³ cells /well) and incubated for 24 hours to allow attachment and reach 70-90% confluence. Then the samples were added to a final concentration of 100 μg/mL in cell culture medium (DMEM supplemented with 10% FBS) and incubated for 15, 30 and 45 minutes.

At the specified time points, the cells were washed twice with HBSS buffer, detached using TrypLE solution (ThermoFisher, USA), and centrifuged at 1,000×g (5 min, 18 °C). The cell pellets were resuspended in FACS buffer and analyzed on the FACS MoFlo XDP (Beckman Coulter, USA). The nanoparticle uptake by the cells was determined as the percentage of NP-positive cells: PLGA-Cy5-positive or double-positive (PLGA-Cy5- and HSA-RhBITC//DIVEMA-Cy3/RhB-positive cells).

2.9. Statistics

The data were analyzed using GraphPad Prism 9 software (San Diego, CA, USA) with one- or two-way ANOVA followed by Tukey’s and Dunnett’s multiple comparisons tests. Characterization of particle size, size distribution, zeta potential, and all in vitro measurements were conducted in triplicate.

3. Results and Discussion

3.1. Core-Shell Nanoparticle Preparation and Characterization

The main objective of this study was to develop an integrated approach for the evaluation of core-shell nanoparticle stability using PLGA nanoparticles with different polymeric shells as model structures. As mentioned above, the following shell-forming polymers were used: 1) poloxamer 188 (P188), a nonionic triblock copolymer of polyethylene oxide and polypropylene oxide; 2) human serum albumin (HSA), one of the most abundant plasma protein with high ligand binding capacity and a natural carrier of many exogenous and endogenous substances [51]; 3) a copolymer of divinyl ether and maleic anhydride (DIVEMA), a polyanion with versatile biologic activity [52]. DIVEMA is a polyanhydride that, upon hydrolysis, turns into a polycarboxylate containing four carboxyl groups per monomeric unit.

To enable visualization of core-shell nanoparticles’ interaction with cells, both the core- and the shell-forming polymers were labeled with covalently bound bright fluorescent dyes. As shown in our previous studies, only covalent linking ensures stable retention of fluorescent labels by nanoparticles, which helps to avoid artifacts in imaging the nanoparticle interactions with biological objects [42]. Therefore, the following dye-polymer conjugates were synthesized using carbodiimide chemistry: PLGA–Cyanine5 (PLGA-Cy5), HSA–rhodamine B (using RhBITC derivative, HSA-RhBITC), poloxamer 188–rhodamine B (P188-RhB), and DIVEMA–Cyanine3 (DIVEMA-Cy3). The covalent binding of the dyes to the polymers was confirmed by GPC or TLC analysis, as described in [28,29,30]. The chemical structures of the fluorescently labeled polymers are schematically presented in Figure 1.

The core PLGA nanoparticles were prepared by the commonly used method of high-pressure homogenization followed by solvent removal. This method involves the emulsification of a PLGA solution in DCM in an aqueous solution of polyvinyl alcohol (PVA), one of the most popular stabilizers used for the preparation of PLGA nanoparticles [31]. The core-shell nanoparticles were prepared using previously established techniques, including adsorption of the shell-forming polymer onto the previously prepared core PLGA nanoparticles (P188, HSA), interfacial embedding achieved by adding the shell-forming polymer in the course of nanoparticle preparation (DIVEMA, HSA), or by covalent linking of the shell to the PLGA carboxyl end groups (HSA) [28,29,30].

The types of dual-labeled core-shell PLGA nanoparticles prepared and investigated in this study are shown schematically in Figure 2 and Supplementary Figure S1. The physicochemical parameters of these nanoparticles are shown in Table 1. The nanoparticles had hydrodynamic diameters of 100-250 nm (volume distribution) depending on the preparation method and shell type, and they were negatively charged –parameters considered optimal for drug delivery [53].

The methods of shell formation were chosen based on the polymer properties. In the case of HSA, three techniques were used for shell formation to reveal the optimal conditions for the use of this versatile and complex molecule. HSA is known to adsorb to various surfaces depending on their charge and hydrophobicity [54]; it can also provide steric stabilization of the colloidal carriers, protecting them from recognition by macrophages [55]. In the case of the adsorption technique, shell formation was achieved by incubating the “bare nanoparticles” in the HSA solution. For the interfacial embedding of HSA, the PLGA NP were also prepared using the high-pressure homogenization—solvent evaporation technique; however, in this case, HSA (0.5% solution) was used as a stabilizer in the outer aqueous phase instead of PVA. Indeed, due to its amphiphilic properties, HSA could be located in the interfacial area between the organic and aqueous phases, with its hydrophilic fragments protruding into the water, while the more hydrophobic fragments are “immersed” into the PLGA core. As shown in Table 1, the method of interfacial embedding was the most effective, producing a considerably more abundant shell on the nanoparticle surface compared to HSA adsorption or conjugation (0.52, 0.11, and 0.48 mg HSA per mg PLGA, respectively). Similar results were obtained in the study by Hyun at el., where the interfacial embedding of HSA yielded nanoparticles with a higher amount of HSA compared to adsorption [56]. For better stability, the HSA shells in the PLGA/HSA NP prepared by adsorption and interfacial embedding were additionally cross-linked by glutaric aldehyde. This approach was shown to improve the stability of the HSA-coated nanoparticles upon dilution and in the presence of other proteins (endogenous albumin) [57].

As seen from Table 1, the parameters of the PLGA/HSA nanoparticles depended on the method of shell formation. Thus, adsorption and conjugation of HSA led to an increase in the hydrodynamic diameter (from ≈ 110 nm to ≈ 140 nm) and partial compensation of the negative surface zeta potential (from ≈ -20 to ≈ -7 mV) compared to the bare PLGA NP, which correlated with the results of other authors [45]. In contrast, the interfacial embedding of HSA (acting as a surfactant in the absence of PVA) led to the formation of 90-100-nm particles with a higher negative zeta potential (≈ -30 mV). These data correlate with the results of Hyun et al., who observed similar tendencies for the PLGA/HSA nanoparticles obtained by interfacial embedding and by physical adsorption of HSA [56] It is noteworthy that cross-linking of the HSA shell did not exert any influence on the nanoparticle parameters compared to the non-cross-linked nanoparticles.

Interfacial embedding was also applicable for the preparation of PLGA/DIVEMA NP. Indeed, the radical copolymerization of maleic anhydride and divinyl ether yields DIVEMA in its polyanhydride form (Figure 1). In this form, DIVEMA is readily soluble in acetone but not in water. When added into the aqueous-organic mixture during nanoparticle formation, the anhydride cycles are gradually hydrolyzed, converting the copolymer into an amphiphilic molecule. This process is rather slow, so the DIVEMA-Cy3 conjugate retains its amphiphilic nature and is capable of interacting with PLGA on one hand and creating a hydrophilic shell on the nanoparticle surface on the other [58]. The presence of DIVEMA in the PLGA/DIVEMA nanoparticles was previously confirmed by FTIR spectroscopy [28]. Moreover, the presence of the outer shell rich in carboxyl groups on the PLGA/DIVEMA NP surface was confirmed by the pH-dependent changes in their size, which decreased upon lowering the pH level or in the presence of salts due to the decreased dissociation of DIVEMA carboxyl groups. Both phenomena are typical for polyelectrolytes [28,58,59].

The fluorescently labeled PLGA-Cy5/DIVEMA-Cy3 NP were prepared by either the high-pressure homogenization—solvent evaporation technique (PLGA/DIVEMA-H NP) or by nanoprecipitation (PLGA/DIVEMA-H) (Figure 2). Dual labeling was achieved by using a (1:1) mixture of PLGA and PLGA-Cy5 conjugate as the PLGA constituent and DIVEMA-Cy3 conjugate. In both cases, the presence of DIVEMA led to a considerable increase in the mean hydrodynamic diameter (measured by DLS) (Тable 2), which suggests the presence of a more pronounced hydration shell and abundant polyanionic groups on the surface of the hybrid nanoparticles. The sizes of the PLGA/DIVEMA-N and PLGA/DIVEMA-H were about 180 nm and 264 nm compared to 130 nm and 178 nm for the bare particles, respectively (Table 1). The content of the –COOH groups in both types of the PLGA/DIVEMA NP was considerably higher compared to the plain PLGA NP: 0.72 ± 0.06 vs 0.09 ± 0.04 mmol/g PLGA and 0.46 ± 0.05 vs 0.04 ± 0.01 mmol/g PLGA for the PLGA/DIVEMA-H and PLGA/DIVEMA-N nanoparticles, respectively. Accordingly, these nanoparticles had higher negative zeta potentials compared to the plain PLGA NP (Table 2). The relative content of the shell (DIVEMA-Cy3) was higher in the case of the nanoprecipitation method, which is probably due to the particle formation in the absence of PVA in the aqueous phase: 0.47 vs 0.10 mg/mg PLGA, respectively (Table 1). Due to this more considerable shell, combined with their more favorable size and size distribution parameters, only the PLGA/DIVEMA-N NP were used in further experiments.

The PLGA-Cy5/P188-RhB nanoparticles were prepared by incubating the plain nanoparticles in a 1% P188-RhB solution [30]. Conjugation of P188 with rhodamine B did not affect its surfactant properties: the amount of P188-RhB adsorbed on the nanoparticle surface was even increased compared to P188 (0.38 vs 0.14 mg/m² of NP or 0.017 vs 0.006 mg/mg PLGA, respectively). Modification of the PLGA NP with poloxamer 188 and P188-RhB did not induce any changes in their hydrodynamic diameter and surface potential (Table 1), which correlates with our previous observations [61]. It is noteworthy that the P188 shell content is the lowest compared to HSA-RhBITC and DIVEMA-Cy3 (0.017 mg/mg PLGA).

As shown by TEM, all types of the core-shell nanoparticles had a spherical shape (Figure 3). Due to its electron density, the HSA shell was well visualized by TEM [62]. The presence of the hydrated HSA shell was confirmed by the smaller nanoparticle size observed by TEM, compared to the DLS measurements of the same sample: 50-100 nm vs 80-150 nm (Vol, PDI <0.20). This observation correlates with the data obtained by other authors [60,62].

The most prominent shell was observed on the PLGA/HSA NP prepared by interfacial embedding with cross-linking (PLGA/HSA IE cross-linked NP, Figure 3d). These nanoparticles had the highest HSA/PLGA ratio (0.52 mg/mg PLGA, Table 1), compared to the adsorption and conjugation methods (Figure 3a,b); a similar phenomenon was also observed by Hyun H. et al. [45,56]. These authors suggested that HSA introduced onto the nanoparticle surface during the emulsification stage could form a shell consisting of several HSA layers. The high density of HSA on the PLGA/HSA-IE NP surface is also consistent with the relatively high negative surface charge similar to the nanoparticles obtained in [45,56].

In the case of the PLGA/DIVEMA-N, the direct comparison of the sizes measured by TEM and DLS is not possible due to the high polydispersity of the PLGA/DIVEMA-N NP (PDI >0.2); however, it appears that in the TEM micrographs, the majority of the hybrid nanoparticles had smaller diameters than the diameter measured by DLS: ≈ 200 nm vs 280 nm (Volume, Table 1). Accordingly, this phenomenon is most likely explained by the hydration of the hydrophilic shell in the aqueous medium. The P188-RhB shell was not visible in TEM images, which has been noted in a number of other studies [63,64].

3.2. Evaluation of Integrity by Physicochemical Methods

The stability of the core-shell structures in biorelevant media was initially studied using DLS and spectrofluorimetry. As shown by the DLS measurements, all HSA- and DIVEMA-coated nanoparticles preserved their initial hydrodynamic diameters and polydispersity for at least 1 hour (Supplementary Table S1) in the HSA-containing medium (pH 7.4, PBS buffer with 4.5% of HSA) and for 6 hours (Supplementary Table S2) in the absence of HSA (PBS pH 7.4) (Figure 4). An increase in the average particle size (by volume) in the presence of HSA may indicate the adsorption of HSA on the surface of the particles [66], the replacement of the shell with HSA molecules, loss of stability, and aggregation. The formation of the protein corona on the PLGA/HSA NP (adsorption and interfacial embedding) during nanoparticle incubation in serum was confirmed in [56].

The next step involved evaluating the core-shell structure integrity of the nanoparticles by fluorescence spectroscopy. As mentioned above, the core and the shells of the nanoparticles were labeled with pairs of fluorescent dyes with known donor-acceptor properties, where Cy5 is the acceptor molecule, and Cy3 and rhodamine B are the donors [67]. As shown by the fluorescence measurements (Table 2), all dual-labeled PLGA nanoparticles had sufficient brightness for visualization based on a comparison of the parameters of free dyes and nanoparticles (Supplementary Table S3) [49,50,67].

As mentioned above, donor-acceptor pairs are capable of exhibiting the distance-dependent Förster resonance energy transfer (FRET) phenomenon, observed only in the case of the close proximity (< 10 nm) of both dye molecules, enabling the energy transfer from the donor fluorophore to the acceptor fluorophore [66,68,69]. In the present study, FRET was detected for the Cy5 – Cy3 (PLGA-Cy5/DIVEMA-Cy3 NP) and Cy5 – RhBITC (PLGA-Cy5/HSA-RhBITC) pairs. As shown in Figure 5 (representative images), the fluorescence intensity of the acceptor in double-labeled nanoparticles is detected when the excitation occurs at the donor wavelength (550 nm for RhBITC and 530 nm for Cy3).

The FRET phenomenon was observed for all types of PLGA/HSA NP, indicating the closeness of the shell and PLGA core (interaction between RhBITC and Cy5). Moreover, the most prominent FRET phenomenon was observed for the HSA shell obtained by the interphase embedding method (Supplementary Figure S2). Similarly, the PLGA/DIVEMA NP demonstrated core-shell integrity through a strong FRET signal between Cy3 (DIVEMA) and Cy5 (PLGA), indicating the close proximity of the donor (in the PLGA core) and the acceptor (in the shell) (Figure 5b). No FRET phenomenon was detected for the PLGA/P188 NP (data not shown), which may be due to the large distance between the donor in the P188-PhB shell and the acceptor in the PLGA nanoparticle core. Poloxamer 188 reacts with rhodamine B through terminal hydroxyl groups of polyoxyethylene (Figure 1) [30]. Presumably, when P188-RhB is adsorbed on the surface of PLGA nanoparticles, the hydrophobic chains of polyoxypropylene most likely interact with the hydrophobic core of PLGA, while the hydrophilic chains of polyoxyethylene with the grafted dye are solubilized by water molecules and protrude into the surrounding medium [70,71], which prevents them from approaching the nanoparticle surface and hinders efficient energy transfer from the donor to the acceptor.

To evaluate how the modification method influenced core-shell structure stability, the amount of detached shell was assessed within 6 hours of nanoparticle incubation in PBS and 4.5% HSA solution (imitating HSA concentration in plasma). The amount of detached shell was evaluated based on the fluorescence intensity of the supernatant after sample centrifugation at high acceleration for complete nanoparticle sedimentation (48,060×g). Stability of the core-shell structure was evaluated as the percentage of the shell remaining on the PLGA nanoparticle surface (Figure 6).

As shown in the graphs, considerable fractions of the HSA/RhBITC shell detached immediately upon their addition into the media for all nanoparticle types (Figure 6a). This initial burst effect was more pronounced for the PLGA/HSA NP obtained by conjugation and adsorption without crosslinking (~30%). In contrast, in the case of nanoparticles with cross-linked shells (adsorption-linked and interfacial embedding-linked), above 80% of the shell remained attached to the surface within 1 hour of incubation (Figure 6a).

The HSA-RhBITC shells appeared to be less stable in the serum-containing medium (4.5% HSA in PBS), with more than 50% of the shell detaching from the nanoparticle surface within the first hour (Figure 6b, Supplementary Table S4). Interestingly, the PLGA/HSA-IE non-cross-linked nanoparticles, which were highly stable in albumin-free medium, demonstrated the lowest stability in the presence of 4.5% HSA in the medium. The initial amount of detached HSA-RhBITC in all cases, including the PLGA/HSA-C NP obtained by conjugation, could be explained by the presence of HSA, loosely associated with the surface. The nanoparticle samples prepared with additional cross-linking stage were more stable compared to the non-linked ones. Thus, the cross-linked nanoparticles (PLGA/HSA-IE cross-linked and PLGA/HSA-A cross-linked) were chosen for further in vitro experiments.

A similar trend was observed for PLGA/DIVEMA NP (Figure 6c, Supplementary Table S4). The amount of detached shell was evaluated based on the relative fluorescence intensity (DIVEMA-Cy3) of the supernatant after sample centrifugation. Apparently, the high initial amount of detached DIVEMA-Cy3 resulted from a shell loosely associated with the surface.

3.3. Investigation of the Integrity of the Core-Shell Nanoparticles Upon Internalization in Gl261 Cells by Confocal Microscopy

For the in vitro experiments, it is essential to track the integrity of the core-shell structures during their incubation with the cell culture media/serum. This study aimed to provide a more comprehensive overview of the nanoparticles’ structural changes (lifecycle) once they were added to the cells. Dual labeling of the nanoparticles allowed for evaluation of core-shell integrity over time by confocal laser scanning microscopy (CLSM). Life cell imaging was performed at different time intervals during nanoparticle incubation with Gl261 cells. The Manders’ Overlap Coefficient (MOC) was chosen to evaluate the core-shell colocalization, as it is widely used in fluorescence microscopy to quantify colocalization and is implemented in most biological image analysis software packages [73]. Unlike Pearson’s correlation, MOC is primarily sensitive to the co-occurrence of signals from two channels, almost independent of the signal level and proportionality.

To investigate how the modification method affected the integrity of the PLGA/HSA NP (i.e., the attachment of surface-bound albumin), colocalization coefficients between the PLGA-Cy5 (core) and HSA-RhBITC (shell) were calculated upon incubation with cells for different time periods (Supplementary Table S5). Only the cross-linked samples were used in these experiments due to their higher stability in model media. All three selected nanoparticle samples were efficiently internalized by the Gl261 cells (Figure 7). Colocalization coefficients above 0.5 indicated a correlation between the nanoparticle core and shell localization in the course of the experiment (Figure 7a). However, the MOC values were significantly lower at all time points for the PLGA/HSA NP obtained via interfacial embedding (0.63 compared to 0.83 for the nanoparticles produced by HSA conjugation or adsorption, Supplementary Table S5). Interestingly, unlike other PLGA/HSA nanoparticles (Figure 7b,c), the PLGA/HSA NP obtained by interfacial embedding (Figure 7d) tended to accumulate on the outer membrane after 15-30 min of incubation with cells, while the HSA-RhBITC signal was detected in the cell cytoplasm (Supplementary Figure S3). Subsequent incubation of this sample (30-45 min) led to increased colocalization between the core and the shell (0.73 after 45 min of incubation). These nanoparticles’ shells were most probably partly detached from the core upon introduction into the cell culture medium, due to dilution, and were internalized faster than the nanoparticles (PLGA-Cy5), suggesting that the PLGA/HSA nanoparticles obtained by interfacial embedding are less stable upon internalization into cells than those obtained by conjugation or adsorption (Supplementary Figure S4, S5).

Similar experiments were performed for the nanoparticles coated with DIVEMA, produced by the nanoprecipitation method (Figure 7e). The MOC values ranged from 0.63 to 0.79, suggesting that these nanoparticles preserved their core-shell structure (Supplementary Table S6). At 15 minutes, the MOC value was 0.42, indicating the presence of unbound or easily detached DIVEMA-Cy3. As a result, a significant amount of free DIVEMA-Cy3 accumulated in the cell cytoplasm during the first minutes of incubation (Supplementary Figure S6).

The P188 coating of the PLGA nanoparticles appeared to be relatively stable (Supplementary Figure S7). During the first 15 minutes of incubation of poloxamer-coated NP with cells, the MOC was 0.48 (Figure 7f, Supplementary Table S7), indicating that the uptake of free poloxamer by the cells occurred faster than the internalization of the PLGA nanoparticles. Further incubation (30-45 min) increased the average values of colocalization coefficients, indicating a high level of overlap between the PLGA-Cy5 and P188-RhB signals. The colocalization coefficients varying from 0.49 to 0.62 during incubation with the cells suggested that the core-shell structure partially preserved its integrity upon nanoparticle internalization and intracellular trafficking.

Colocalization coefficients were also calculated between the PLGA-Cy5 nanoparticle core and lysosomes (Supplementary Tables S5-7). Late endosomes and lysosomes are the main compartments involved in clathrin-mediated endocytosis, which is known to be one of the most common pathways for polymeric nanoparticle internalization in the vast majority of cell cultures [74]. Interestingly, the modification of PLGA nanoparticles with different shells did not significantly alter the colocalization of the PLGA-Cy5 nanoparticle core with lysosomes in Gl261 cells: the MOC values were found to vary from 0.3 to 0.5 for almost all samples, suggesting only partial accumulation of nanoparticles in lysosomes. Thus, it can be hypothesized that clathrin-mediated endocytosis is not the main internalization route for PLGA nanoparticles. These data align with the results of Beigulenko et al., who demonstrated that only a small portion of PLGA nanoparticles coated with fluorescently labeled poloxamer (P188-RhB) was associated with lysosomes [30]. It is also worth mentioning that caveolin-mediated endocytosis plays the primary role in albumin uptake into cells [75]. Therefore, the albumin coating of nanoparticles could be expected to affect the intracellular trafficking of PLGA nanoparticles. However, this was not observed in the present study. One possible reason could be a change in the conformation of albumin molecules on the nanoparticle surface, leading to alterations in the interaction between nanoparticles and cells [56]. These findings, however, require further investigation of the endocytosis routes, as they may also be attributed to the specific characteristics of the cell culture.

3.4. Investigation of the Dynamics of Nanoparticle Uptake by Gl261 Glioma Cells

To gain a deeper understanding of nanoparticle interactions with cells, the dynamics of their accumulation in Gl261 cells was investigated using flow cytometry. The nanoparticle uptake by the cells was determined as the percentage of NP-positive cells: PLGA-Cy5-positive or double-positive (both PLGA-Cy5- and HSA-RhBITC-, DIVEMA-Cy3- or P188-RhB-positive) cells (Figure 8). First, the percentage of PLGA-Cy5 positive cells was calculated to evaluate the possible influence of nanoparticle surface modification on their interaction with cells. Then, the percentage of cells positive for both PLGA-Cy5 (the core) and HSA-RhBITC, RhB or Cy3 (the shell) was determined to assess the structural integrity of the nanoparticles upon interaction with the cells over time. The nanoparticle uptake by the cells (percentage of nanoparticle-positive cells) tended to increase with time for nanoparticles obtained by all shell-forming techniques.

The presence of the HSA shell increased the number of nanoparticle-positive cells (Figure 8a, Supplementary Table S8). The percentage of PLGA-Cy5 positive cells significantly increased compared to control PLGA-Cy5 nanoparticles without shells (p < 0.0001 for all samples), regardless of the preparation method used (only statistically insignificant data are shown in the graph). This phenomenon could be attributed to the interaction of nanoparticles with albumin-binding proteins on the cell membrane. Interestingly, based on the PLGA-Cy5 fluorescence (nanoparticle core), the maximal percentage of cellular uptake was observed for the PLGA-HSA NP obtained by interfacial embedding: around 65 % of PLGA-Cy5-positive cells were detected after only 15 minutes of incubation, compared to 27% for PLGA/HSA nanoparticles produced by conjugation or adsorption (p <0.0001). However, the percentage of double-positive cells (positive for both PLGA-Cy5 and HSA-RhBITC) was the highest for the PLGA/HSA nanoparticles obtained by conjugation (Figure 8b). The number of double-positive cells was twice as high for PLGA/HSA-C compared to other HSA-coated batches at all investigated time points. These results are consistent with the confocal experiments, which demonstrated the high stability of the core-shell structure for PLGA/HSA-C.

The nanoparticles modified with P188 have been successfully used in several in vivo studies [33,35], suggesting that the P188 surface modification is sufficient to significantly enhance nanoparticle interaction with cells. Indeed, the PLGA nanoparticle coating with poloxamer 188 increased the number of PLGA-Cy5-positive cells by about 45%, compared to bare PLGA-Cy5 nanoparticles after 30 minutes of incubation with the cells (Figure 8a, Supplementary Table S9).

As shown in Figure 8a, the presence of the DIVEMA shell also significantly increased the number of PLGA-Cy5 positive cells (Figure 8a, Supplementary Table S10); however, the percentage of double-positive cells remained below 10% throughout the entire incubation period. These results are consistent with the stability tests performed in a 4.5% HSA solution, where a significant portion of the shell already detached from the nanoparticle surface within the first minutes of incubation in the HSA-containing medium. It is worth mentioning that flow cytometry can measure the number of events but cannot distinguish whether the nanoparticle core or shell is internalized by the cells or merely accumulated on the cell membrane [76]. In contrast, confocal microscopy enables detailed investigation of nanoparticle uptake, intracellular localization, and trafficking. Therefore, it is essential to use several methods, such as flow cytometry in conjunction with confocal microscopy, to fully characterize the behavior of nanoparticle formulations.

Thus, despite the different nature and properties of the shell-forming polymers, all shells enhanced the nanoparticle uptake by Gl261 cells as compared to uncoated nanoparticles. In the case of the PLGA/HSA nanoparticles, the highest stability of the HSA shell upon incubation in the HSA-containing media and with Gl261 cells was achieved by conjugation (PLGA/HSA-C), compared to the adsorption (PLGA/HSA-A) and interfacial embedding methods (PLGA/HSA-IE). Interestingly, the interfacial embedding method provided the highest shell content (HSA, DIVEMA) per polymer weight (~0.5 mg/mg PLGA) and high FRET-effect; however, in both cases low percentage of the core-shell double-positive cells (flow cytometry) and low Manders’ overlap coefficient (CLSM) indicated the loss of the core-shell integrity in the presence of HSA in the incubation medium. The differences between the core-shell systems may become more pronounced in the in vivo conditions, when they will be introduced into a very dynamic multicomponent environment and various blood components will be competing for the space on the particle surface. At the same time, the suggested combination of the evaluation techniques enables thorough analysis of the core-shell structure stability and may be useful for selection of the optimal preparation method already at the early stages of nanoparticle development.

4. Conclusions

In this study, the properties of core-shell nanoparticles with a PLGA core and shells made of polymers with significantly different structures and physicochemical properties were compared, with a focus on their structural integrity. The basic parameters of the nanoparticles, such as their size and zeta potential, depended on the shell composition and preparation technique; however, all particles had hydrodynamic diameters of ≤ 250 nm and were negatively charged — parameters considered optimal for drug delivery. Notable, despite differences in their chemical structures, polarity, molecular mass and hydrophobicity, all shells were largely retained by the core nanoparticles in both biorelevant media and within cells. Moreover, all shells enhanced the cellular uptake of the nanoparticles. The structural integrity of the core-shell PLGA nanoparticles was confirmed using a variety of methods, including physicochemical evaluation, fluorescence spectroscopy, confocal microscopy, and flow cytometry. The latter methods were facilitated by dual labeling of the nanoparticles with dyes possessing donor-acceptor properties. This comprehensive approach provides valuable insights into the behavior of core-shell nanoparticles in different environments and their interactions with tumor cells.