A Hydrogel-Based Tumor Model for the Evaluation of Nanoparticle-Based Cancer Therapeutics

2.1. Chemicals and reagents

HA (500 kDa, sodium salt) was generously donated by Genzyme Corporation (Cambridge, MA). All chemicals necessary for the synthesis of HA derivatives and the amphiphilic block copolymers were purchased from Sigma-Aldrich (St. Louis, MO). Doxorubicin hydrochloride (Dox-HCl) was obtained from BIOTANG Inc. (Waltham, MA). Nile red (NR), bovine testicular HAase (30,000 U/mg), bovine serum albumin (BSA), sucrose, chlorpromazine hydrochloride, methyl-β-cyclodextrin (M-β-CD), genistein, colchicine and quercetin were purchased from Sigma-Aldrich (St. Louis, MO). Cell Titer-Blue® cell viability assay was obtained from Promega (Madison, WI). PEGylated (methoxy-PEG5000-SH) gold nanoparticles (PEG-AuNPs) were purchased from Cytodiagnostics (Burlington, ON). Cell Death Detection ELISA^PLUS was obtained from Roche Applied Science (Mannheim, Germany). Mouse anti-multidrug resistance protein 1 (MRP1) mAb (MRPm6) was purchased from Kamiya Biomedical Company (Seattle, WA). Mouse anti-lung resistance-related protein (LRP) mAb (LRP 1014) and goat anti-mouse IgG-FITC were obtained from Santa Cruz Biotechnology (Dallas, TX). All cell culture reagents were purchased from Life Technologies (Grand Island, NY). All other reagents were used as received unless otherwise noted.

2.2. Particle preparation and drug/dye encapsulation

Amphiphilic block copolymers consisting of hydrophilic PEG and hydrophobic polyester bearing pendent cyclic ketals [PEG-b-P(CL-ran-TSU)] were synthesized by ring-opening co-polymerization of ε-caprolactone (CL) and 1,4,8-trioxaspiro-(4,6)-9-undecanone (TSU), using α-hydroxyl, ω-methoxy PEG as the initiator and stannous octoate as the catalyst [31]. At a 20 wt% TSU in feed, the copolymer (abbreviated as ECT2) was found to contain 14 mol% TSU in the hydrophobic segment. ECT2 with a number average molecular weight (M_n) of 40.6 KDa was used to formulate Dox-loaded NPs. Prior to drug encapsulation, Dox-HCl was neutralized to generate Dox, following reported procedures [32]. NPs were prepared using a nanoprecipitation method. Briefly, an acetone/DMSO (1:1, v/v) solution of ECT2 (10 mg/ml, 1ml) was slowly added to a stirred (900 rpm) aqueous phase (5 ml DI water). The mixture was allowed to stabilize for 2 h under constant agitation at room temperature to obtain blank-NPs. Dox- or NR-loaded NPs were prepared using an acetone/DMSO (1:1, v/v) solution of ECT2 (10 mg/ml, 1ml) containing 2 mg/ml Dox or 0.1 mg/ml NR. Centrifugation (4,000 rpm for 10 min) was applied to all NP suspensions to remove large polymer aggregates. The supernatant containing NPs was collected, and the free drug or dye and the solvent were removed by filtration using Amicon regenerated cellulose centrifugal filters (MWCO=30kDa, EMD Millipore). The collected NPs were thoroughly washed with DI water or phosphate buffered saline (PBS, pH 7.4) three times by centrifugation using the centrifugal filters. The resultant formulations were immediately subjected to physical and biological analyses.

2.3. Characterization of NPs

The hydrodynamic diameters of NPs were measured using a Zetasizer nanoZS (Malvern Instruments, Westborough, MA) by dynamic light scattering (DLS) at a concentration of 0.5 mg/ml in PBS. Transmission electron microscopy (TEM) was used to examine the particle size and morphology. TEM samples were prepared by applying a drop of NP suspension (3 μl) directly onto a carbon-coated copper TEM grid. Samples were allowed to dry under ambient condition prior to imaging using a Tecnai G2 12 Twin TEM (FEI Company, Hillsboro, OR). The mean diameter of the NPs was estimated using ImageJ based on 50 counts of the particles from the TEM image.

2.4. Drug/dye loading and release

Aliquots (1 ml each) of the Dox-NPs and NR-NPs suspension in DI water were collected and lyophilized. The dried powder was weighed accurately before being dissolved in DMSO (1 ml). The fluorescence intensity of the drug or dye was determined using a plate reader (DTX880 Multimode Detector, Beckman Coulter, Fullerton, CA) at 485 nm (excitation) / 595 nm (emission) for Dox and at 535 nm (excitation) / 595 nm (emission) for NR. The concentration of the drug or dye was then derived from a series of standards in DMSO at Dox concentrations of 3.125-25 μg/ml, and NR concentrations of 6.25-100 ng/ml. The drug-loading content (LC) was defined as the amount of drug (μg) loaded per milligram of Dox-NPs. The drug encapsulation efficiency (EE, as percentage of the total) was calculated by dividing the amount of Dox loaded into the NPs with the amount of Dox initially added during the nanoprecipitation process. All measurements were carried out in triplicate, and the results were indicated as the mean ± SD.

The in vitro release of Dox and NR was analyzed under sink conditions. Briefly, freshly formulated NPs dispersed in serum containing cell culture medium were incubated at 37 °C. At predetermined time points, the release medium was separated from the NPs using centrifugal filters as described above. NPs were re-dispersed in fresh cell culture medium to continue the release study. The collected filtrates containing the released Dox or NR at each time point were lyophilized, and the resultant solid was re-dissolved in DMSO (undissolved salt was removed by centrifugation). Subsequently, the fluorescence intensity of the Dox and NR was measured using the plate reader, and the concentration was derived based on the standard curve as described above. Three repeats were performed for each time point, and the cumulative release profile was calculated by dividing the amount of drug or dye released in one specific measurement time by the total mass initially loaded.

2.5. Hydrogel preparation

Acrylated HA (HA-AC) and thiolated HA (HA-SH) were synthesized following previously reported procedures [33, 34]. Briefly, HA-AC was synthesized by reacting adipic acid dihydrazide-modified HA (HA-ADH, 35% ADH incorporation determined by ¹H NMR [35]) with N-acryloxysuccinimide (NHS-AC) in DI water. The product was obtained as a lyophilized solid (yield: 70%). Sulfhydryl groups were incorporated in HA via the reaction with a disulfide-containing dihydrazide compound, followed by reduction with dithiothreitol (DTT). The dry solid HA-SH was obtained after lyophilization (yield: 80%). The HA derivatives were stored at −20 °C prior to use. ¹H NMR analyses revealed a degree of modification of 9% and 45% for HA-AC and HA-SH, respectively [17]. The respective weight average molecular weights, as determined by GPC, were 345 kDa and 314 kDa. HA hydrogels were prepared by mixing equal volume of HA-AC and HA-SH, dissolved in PBS at a concentration of 20 mg/ml. The gelation process started 17 min after mixing and was completed after 6 h incubation at 37 °C [17].

2.6. Analysis of hydrogel pore size

The pore size of the hydrogel was determined by a probe retention method [36] using a series of PEG-AuNPs with mean hydrodynamic diameters of 35, 50, 70 and 100 nm, with the respective Au core of 5, 20, 40 and 70 nm. The concentration of PEG-AuNPs in the stock solution, C₀ (mg/ml), was determined by UV-Vis spectrophotometer (Agilent Technologies, Santa Clara, CA) at a wavelength of 518, 524, 530 and 548 nm for 35, 50, 70 and 100 nm PEG-AuNPs, respectively. An aliquot (V₀ ml) of PEG-AuNPs stock solution was spun down by centrifugation and the pellet was re-dispersed in HA-AC solution in PBS (20 mg/ml, 75 μl). The HA-AC solution containing the disperse PEG-AuNPs was subsequently mixed with the HA-SH solution in PBS (20 mg/ml, 75 μl). The mixture was incubated at 37 °C for 6 h. The hydrogels were then immersed in fresh PBS solution at room temperature for two days under constant shaking. The supernatant was collected and the hydrogel was washed with PBS. The concentration of released PEG-AuNPs (C₁, mg/ml) in the combined supernatant and the washing solution (total volume V₁) was determined by UV-Vis spectrophotometer. The retention of PEG-AuNPs in the hydrogel was quantified as follows: ((C₀×V₀-C₁×V₁)/C₀×V₀) × 100.

2.7. Analysis of NP diffusion in HA hydrogels

HA hydrogels were prepared as described above. The precursor mixture was added to the channel (17×3.8×0.4 mm³) connected by two cylindrical reservoirs at each end on a channel slide (ibidi, Verona, WI). After 6 h incubation at 37°C, one of the reservoirs was filled with PBS containing NR-NPs (1 mg/ml, 70 μl) and the other with PBS. The device was equilibrated at 37 °C for 18 h. Control experiments using free NR (1 μg/ml in PBS) were performed as described above. Fluorescence images (16-bit) were taken with a PCO-Tech Pixelfly CCD camera after being magnified 5 times using a lab-built fluorescence microscope. A 1W 470 nm LED was used to excite the NR using an epifluorescence design, and the fluorescent light was filtered through a 570 nm long-pass filter. ImageJ was used to construct a 0.8 mm × 2.5 mm image from a row scan of four images, and to generate intensity line profiles. The background intensity was subtracted from all line profiles prior to curve fitting (Igor Pro, Wavemetrics). The concentration profile for NR-NPs was fitted to the following one-dimensional model to obtain the diffusion coefficient in the HA hydrogel [37]:

Where erfc is the complementary error function, x is the diffusion distance (μm), t is time (s), and C(x,t) is the measured concentration (intensity) as a function of time (t) at a known distance (x). The measurement was carried out with three independent repeats.

2.8. Cell culture

LNCaP prostate cancer cells were maintained in Corning (Lowell, MA) tissue culture flasks (75 cm²) at 37°C in 5% (v/v) CO₂ in a RPMI-1640 medium supplemented with 5% (v/v) fetal bovine serum (FBS), 100 U/mL penicillin G sodium and 100 μg/mL streptomycin sulfate in 0.085% (v/v) saline (P/S). The medium was changed every other day, and cells were routinely passaged using 0.25% (w/v) trypsin containing ethylenediaminetetraacetic acid (EDTA*4Na). Hydrogel culture of LNCaP prostate cancer cells were performed following previously reported procedures [17]. Briefly, cell culture inserts were pre-wet with PBS before being placed into a 24-well culture plate. HA-SH (100 μl, 20 mg/ml) was added to the HA-AC solution (100 μl, 20 mg/ml) containing dispersed LNCaP cells (0.2 million). After thorough mixing, the cell suspension was transferred into the cell culture insert and was incubated for 45 min at 37 °C before the medium was added around the insert. The resultant cell-gel construct had a cylindrical geometry of 1.8 mm in height and 12 mm in diameter. The cell culture medium was changed every day for each culture.

2.9. Expression of multidrug resistance proteins

LNCaP cells were cultured on 2D or in 3D as described above. At day 7, cells on 2D plastic surfaces were trypsinized and collected. Cells cultured in HA hydrogels were harvested by degrading the HA matrices using HAase (5KU/ml) at 37 °C for 4 h. RNA extraction, reverse transcription and quantitative polymerase chain reaction (qPCR) analysis using ABI Prism 7300 Sequence Detection System (AB Applied Biosystems, Foster City, CA) were performed as described previously [36]. Primer sequences for specific genes were as follows [17, 38]: MRP1 (XM_003281512.2): forward (5′-GTC GGA ACA AGT CGT GCC TG-3′), reverse (5′-CAA AGC CTC CAC CTC CTC A-3′); LRP (XM_004615945.1): forward (5′-TGG CTG AGG TGG AGG TGA AG-3′), reverse (5′-ATC GGT GAT GAG GGT TGA TTT C-3′); β-actin (NM_001101.3, the housekeeping gene): forward (5′-GGA CTT CGA GCA AGA GAT GG-3′), reverse (5′-AGC ACT GTG TTG GCG TAC AG-3′). The relative gene expression was calculated using the equation below [36]:

where E_target and E_ref are the real time PCR efficiencies of the target and the reference genes, respectively. ΔCt_target is the difference in the threshold cycle (Ct) of the target genes between the control and the test samples. ΔCt_ref is the difference in the Ct value of the control versus the test samples with reference to β-actin.

To analyze the expression MDR proteins at the protein level, tumoroids in 3D hydrogel cultures were collected by the enzymatic degradation using HAase as described above. Single cell suspension was obtained by treating the tumoroids with trypsin. Cells on the monolayer after 7 days of culture were treated with the same concentration of HAase as used in 3D, and then were collected by trypsinization. Single cell suspensions from both 2D and 3D cultures were plated onto the 8-well chamber slides at the same density (5 × 10⁴ cells per well). Immunocytochemistry staining for MRP1, LRP and cell nucleus were performed 24h after seeding, following previously reported procedures [17]. Mouse anti-MRP1 mAb (1:4 in 3% BSA) and mouse anti-LRP mAb (1:20 in 3% BSA) were used as the primary antibodies for the MRP1 and LRP, respectively. Goat anti-mouse IgG-FITC (1:400 in 3% BSA) was used as the second antibody. Cell nucleus was stained with Draq5 (1:1000 in PBS). Samples were imaged using the Zeiss510 NLO confocal microscope.

2.10. Cellular uptake of Dox and Dox-NPs

LNCaP cells were grown on 8 well chamber slides as 2D monolayer or in HA hydrogels as 3D tumoroids, following the procedures described above. Cells on 2D and in 3D were incubated with free Dox (50 μg/ml) or Dox-NPs (containing 50 μg/ml equivalent Dox) for 2 h at 37 °C. After extensive wash with PBS, cells or cell-gel constructs were inspected using Zeiss510 NLO confocal microscope to detect the internalized Dox or Dox-NPs (excitation wavelength at 488 nm with a band pass filter of 565-615 nm). Because of the short duration of the uptake study, tumoroids close to cell/gel constructs were imaged. Z-stack images were collected at a 3-μm interval to observe the localization of Dox-NPs in the tumoroid.

2.11. Endocytic mechanism

LNCaP cells seeded in a 96 well plate at a density of 6.4 × 10³ per well were incubated for 24 h. After the medium was removed, cells were incubated at 37 °C for 30 min with one of the following compounds dissolved in cell culture medium: 0.45 M hypertonic sucrose and 10 μg/ml chlorpromazine hydrochloride (both are inhibitors of clathrin-mediated endocytosis); 5 mM m-β-CD and 50 μg/ml genistein (both are inhibitors of caveolae-mediated endocytosis); 8 μg/ml colchicine (an inhibitor of macropinocytosis) and 6 μg/ml quercetin (an inhibitor of caveolae- and clathrin-independent endocytosis), followed by a 2h incubation with NR-NPs (0.5 mg/ml). Subsequently, 2D cultures were thoroughly washed with PBS, and the internalized NR was recovered by DMSO extraction and quantified using the plate reader. For 3D samples, LNCaP cells were cultured in HA-based hydrogel for 7 days before the treatment. After the inhibitor treatment, cell-gel constructs were exposed to NR-NPs for 24 h at 37 °C. Constructs were then degraded by using HAase and the collected cells were subjected to DMSO extraction. The mean fluorescence intensity of cells cultured on 2D and in 3D without any inhibitor treatment served as the control and was considered to be 100%. The fluorescence intensity of cells treated with various inhibitors was compared with the control group to obtain the percentage of cellular uptake.

2.12. Drug treatment and apoptosis assay

Cells cultured on 2D in a 12-well plate and in 3D in HA hydrogels as stated above were allowed to proliferate for 24 h and 7 days respectively, to reach a similar cell number of 2.5 × 10⁵ [17]. Fresh medium containing free Dox or Dox-NPs at various concentrations were then added to the respective cultures and the drug treatment was allowed to proceed for 24 h. Subsequently, cells were collected by trypsinization or HAase degradation for 2D and 3D respectively. Cell apoptosis analysis was performed using Cell Death Detection ELISA^PLUS according to manufacturer’s instructions. Colorimetric detection by absorbance was carried out using the plate reader at 405 nm.

2.13. Statistical analysis

All quantitative measurements were performed on three replicate samples. The results were reported as mean ± standard deviation. Statistical significance was evaluated by analysis of variance (one-way ANOVA), followed by Tukey’s post-hoc test. A P-value of <0.05 was considered to be statistically different.

3.1. Formulation and characterization of drug-loaded NPs

The amphiphilic block copolymer, ECT2 (M_n: 40.6 kg/mol, PDI: 1.57) containing 14 mol% TSU in the hydrophobic block, was synthesized following a previously reported procedure [31], and used to formulate the NPs. Dox or a fluorescent hydrophobic dye, NR, was loaded into the NPs by a nanoprecipitation method without using of any surfactants. Particle size analysis by DLS (Figure 1A) showed that Dox-NPs and NR-NPs were similar in size (Dox-NPs: 54 ± 1 nm; NR-NPs: 50 ± 7 nm) and exhibited an overlapping size distribution. Inspection of Dox-NPs by TEM revealed the presence of spherical objects with an estimated diameter of 40 ± 5 nm (Figure 1B). The TEM results are in good agreement with DLS analyses, considering the fact that the TEM image was acquired on dry samples under vacuum whereas the DLS profile was obtained in an aqueous solution. The formulated NPs did not dissociate upon dilution, nor did they aggregate during prolonged incubation in both PBS and serum containing cell culture medium (Figure S1). Cytotoxicity evaluations on blank NPs confirmed that the NPs were not toxic to LNCaP cells under experimental conditions employed in the study (Figure S2).

Dox was effectively entrapped in the NPs with a loading content and encapsulation efficiency of 75.6 ± 3.5 μg/mg and 39.3 ± 0.9 %, respectively. Comparable values were obtained from our previous studies using this type of amphiphilic copolymer as the drug carrier [39]. In vitro release was evaluated by incubating Dox-NPs in serum-containing cell culture medium under sink conditions at 37 °C for up to 7 days (Figure 1C). From day 0 to day 3, Dox was released at a cumulative rate of 15.9 ± 0.3 wt% per day. Thereafter, Dox was released at a slower rate (8.4 ± 0.2 wt% per day) until day 7 when the experiments were terminated. By day 7, a total of 81.2 ± 0.8 wt% of Dox initially loaded was released. Separately, the loading content and encapsulation efficiency for NR were quantified as 3.2 ± 0.1 μg/mg and 32.3 ± 0.4 %, respectively. Less than 0.4 wt% of NR was released at day 1 and 12.1 ± 2.4 wt% NR was detected in serum containing media by day 7 (Figure 1C). The limited release of NR is attributed to the hydrophobicity of the dye [40]. The high retention of the hydrophobic fluorescent probe by NPs demonstrates the feasibility to use NR to track the location of NPs without undesirable dye leaking concerns.

3.2. NP diffusion in HA hydrogels

In our study, the HA hydrogel was developed to mimic the tumor-associated ECM, which is the first barrier to the NP-based cancer therapeutics [37]. We speculate that the diffusion of the ECT2-derived NPs through the hydrogels would be strongly dependent on the average pore size of the hydrogels, not the NP-matrix interaction, considering the fact that the NPs are PEGylated. To analyze the hydrogel microstructure, PEG-modified gold nanoparticles (PEG-AuNPs) ranging from 35 to 100 nm were employed as probes. PEG-AuNPs were encapsulated in the HA hydrogels in situ and the relative retention of PEG-AuNPs reflects the average pore size of the hydrogel matrix. For 35 nm PEG-AuNPs, approximately 27 ± 4% of initially loaded particles was retained by the hydrogel after two days of incubation (Figure 2A). As the size of PEG-AuNPs increased, the average retention capacity increased accordingly, reaching a plateau value of 77 ± 2% for 70 nm PEG-AuNPs. Further increase in particle size didn’t result in any significant increase in the average retention for 100 nm PEG-AuNPs (79 ± 7%). When the size of the PEG-AuNPs is smaller than the average pore size of the hydrogel, the probe can diffuse out of the hydrogel without significant restriction, consequently a relatively low overall retention was observed. When the average pore size of the hydrogel became comparable to the size of PEG-AuNPs, high retention was observed. Large PEG-AuNPs entrapped in the hydrogel could not diffuse out due to the steric hindrance imposed by the hydrogel matrix. Our probe retention experiments suggest that the HA hydrogel contains water-filled pores of approximately 70-100 nm. The estimated pore size is in agreement with the results observed by cryogenic scanning electron microscopy (CryoSEM) as reported in our previous study [17].

We next evaluated the diffusive properties of the polymeric NPs in HA hydrogels (Figure 2B). NR was entrapped in NPs as the fluorescent probe. The NR-NPs suspension was placed in contact with one end of the hydrogel-filled channel. Infiltration of NR-NPs into the HA matrices was visualized using a fluorescence microscope after NPs were equilibrated with the HA hydrogel for 18 h. NR-NPs were found to distribute across the hydrogel-filled channel in a gradient fashion, reaching the interior of the channel at least 1000 μm away from the source reservoir in 18 h (Figure 2B(a)). By fitting the intensity profile of the NR-NPs (Figure 2B(b)) to a one-dimensional diffusion model [37], the diffusion coefficient (D₁) of NR-NPs was calculated as 4.04 ± 0.03 μm²/s. As a control experiment, free NR solution (1 μg/ml) was prepared and the diffusion study was performed similarly as that for NR-NPs. As expected, no fluorescence signal was detected for the free NR within the hydrogel. This can be attributed to the fact that the fluorescence of the free dye was quenched in water [41]. Therefore, the fluorescent signal detected is from NR molecules entrapped in the NPs, not from free NR release in the medium. The diffusion coefficient of NR-NPs in water (D₀) was calculated as 9.21 μm²/s using the Stokes-Einstein equation: $D=\frac{k_{B}T}{6\pi r}$, where k_B is the Boltzmann’s constant, the viscosity (η) of water at 295.6 K is 9.4 ×10⁻⁴ Pa∙s [37], and the hydrodynamic radius (r) of NR-NPs is 25 nm as measured by DLS. The lower D₁ value, in comparison to D₀, indicates the ability of the HA matrix to hinder the NP diffusion [37, 42].

3.3. Characterization of engineered tumoroids

In our previous investigation [17], we characterized PCa tumoroids grown from LNCaP cells in HA hydrogels, in terms of cell morphology, cell-cell adhesion, and the expression of pro-angiogenic factors. Herein, the expression of two important multidrug resistance proteins, MRP1 and LRP were first evaluated at the mRNA level by qPCR analysis. Our results showed that the hydrogel culture resulted in a significant upregulation of MRP1 and LRP expression (Figure 3A). Specifically, cells in 3D expressed 3.2 ± 1.2 and 1.4 ± 0.1 times more MRP1 and LRP than those cultured on 2D at day 7 (p<0.05, Figure 3A). To further corroborate our finding, cellular production of multidrug resistance proteins was assessed by immunocytochemical staining (Figure 3B and 3C). LNCaP cells with a history of 7-day 2D or 3D culture were trypsinized and re-plated on 2D for direct comparison purposes. Twenty four hours post plating, cells with a 2D culture history appeared larger and their cytoplasm was more elongated, whereas those with a 3D culture history appeared smaller and were prone to aggregate. The difference in cell morphology implies that cells with a history of 3D culture may still maintain their 3D phenotype after being re-plated on 2D plastic surface. For both MRP1 and LRP, only a faint background staining was detected for cells cultured on 2D. Contrarily, cells with a history of 3D hydrogel culture exhibited robust MRP1 and LRP staining (Figure 3B and 3C), indicating a significant overexpression of these two multidrug resistance proteins in cells grown under 3D conditions. Overall, the trend obtained from immunostaining is consistent with the qPCR results.

3.4. Evaluation of Dox-NPs using the engineered tumor models

Dox’s intrinsic fluorescence enables the direct monitoring of the intracellular transport of the NPs via confocal microscope. Free Dox and Dox-NPs were found to localize in different intracellular compartments after being exposed to cells cultured on 2D for 2 h. The short exposure time was chosen to avoid complications introduced by Dox released from NPs at a longer incubation time. While a stronger fluorescent signal was found in the cell nuclei in the case of free Dox (Figure 4A), Dox-NPs were excluded from the nuclei and distributed predominantly in the cytoplasm (Figure 4B). These results confirm that Dox-NPs can be readily internalized by LNCaP cells and that limited Dox release from NPs occurred within the time frame of the experiment, given the fact that Dox signal in the cell nuclei was undetectable (Figure 4B). Interestingly, under 3D culture conditions, free Dox localized predominantly in the cytoplasm of individual cells in the tumoroids (Figure 4C and Figure S3A), rather than in the cell nuclei as observed for free Dox in 2D cultures (Figure 4A). On the other hand, Dox-NPs were found to be distributed mainly in the cytoplasm of 3D tumoroids (Figure 4D and Figure S3B). The faint pinkish background staining surrounding the tumoroid in Figure 4D could be a result of the Dox-NPs being retained by the HA matrix. To ascertain that Dox-NPs were indeed internalized by cells in 3D and not simply adhered to the surface of the spheroids, Z-stack slides of the confocal image at a 3-μm interval were examined. Our results (Figure 5) confirm the homogeneous penetration of Dox-NPs through the entire tumoroid.

Our cellular uptake studies show that Dox-NPs can be readily internalized by LNCaP cells. Next, we investigated the endocytic pathways involved in the cellular uptake of NPs, employing specific endocytic inhibitors (Figure 6). NR-NPs were used in these studies in place of Dox-NPs to eliminate possible cytotoxic effects induced by the released Dox. Under the 2D conditions when treated with colchicine, a macropinocytosis inhibitor, cellular uptake of NR-NPs was reduced by 32 ± 4% (p<0.05) relative to the untreated controls. Treatment on 2D monolayer cultures with other inhibitors did not significantly alter the degree of NP uptake. As shown in Figure 6, the inhibitory effects of M-β-CD, genistein (both inhibit caveolae-mediated endocytosis) or colchicine on 3D cultures was evidenced by 34 ± 1%, 27 ± 2% and 23 ± 2% (p<0.05) reduction of NP uptake, respectively. These results suggest that cells cultured on 2D internalize NPs predominantly via the macropinocytosis mechanism, whereas those in 3D employ both caveolae-mediated endocytosis and macropinocytosis mechanisms in NP uptake.

Following the cellular uptake studies, we evaluated the toxicity of free Dox and Dox-NPs on 2D and in 3D using a standard cell apoptosis assay. As shown in Figure 7, the drug induced a dose-dependent toxicity to the cultured cells, both on 2D and in 3D. The apoptosis experiments conducted on 2D cultures revealed that as the Dox concentration increased, both for free Dox and Dox-NPs, the respective apoptosis levels increased accordingly. A similar plateau level was reached (50.3 ± 4.6 and 52.2 ± 4.2, fold change compared to untreated control) when the Dox concentration was higher than 10 and 20 μM for free Dox and Dox-NPs, respectively. The half maximal inhibitory concentration (IC₅₀) for free Dox and Dox-NPs on 2D was calculated as 5.0 ± 0.1 μM and 11.0 ± 0.8 μM, respectively. The apoptosis experiments conducted on 3D cultures showed a similar trend of dose dependence, although the plateau values of apoptosis were significantly lower as compared to the 2D cultures. Compared to cells cultured in 3D without any drug treatment, the apoptosis level for cells receiving the treatment of free Dox and Dox-NPs was 26.8 ± 1.9 and 25.0 ± 1.2 fold higher, respectively. The calculated IC₅₀ value for free Dox and Dox-NPs in 3D cultures is 15.0 ± 0.4 and 12.3 ± 0.8 μM, respectively.