|Home | About | Journals | Submit | Contact Us | Français|
Multi-drug resistant (MDR) cancer is a significant clinical obstacle and is often implicated in cases of recurrent, non-responsive disease. Targeted nanoparticles were made by synthesizing a poly(D,L-lactide-co-glycolide)/poly(ethylene glycol)/epidermal growth factor receptor targeting peptide (PLGA/PEG/EGFR-peptide) construct for incorporation in poly(epsilon-caprolactone) (PCL) nanoparticles. MDR was induced in a panel of nine human breast and ovarian cancer cell lines using hypoxia. EGFR-targeted polymer blend nanoparticles were shown to actively target EGFR over-expressing cell lines, especially upon induction of hypoxia. The nanoparticles were capable of sustained drug release. Combination therapy with lonidamine and paclitaxel significantly improved the therapeutic index of both drugs. Treatment with a nanoparticle dose of 1 μM paclitaxel/10 μM lonidamine resulted in less than 10% cell viability for all hypoxic/MDR cell lines and less than 5% cell viability for all normoxic cell lines. Comparatively, treatment with 1 μM paclitaxel alone was the approximate IC50 value of the MDR cells while treatment with lonidamine alone had very little effect. PLGA/PEG/EGFR-peptide delivery system actively targets a MDR cell by exploiting the expression of EGFR. This system treats MDR by inhibiting the Warburg effect and promoting mitochondrial binding of pro-apoptotic Bcl-2 proteins (lonidamine), while hyperstabilizing microtubules (paclitaxel). This nanocarrier system actively targets a MDR associated phenotype (EGFR receptor over-expression), further enhancing the therapeutic index of both drugs and potentiating the use of lonidamine/paclitaxel combination therapy in the treatment of MDR cancer.
The development of multidrug resistant (MDR) cancer is a significant clinical obstacle that often results in non-responsive, recurrent disease and eventual metastasis 1-6. MDR refers to a state of resilience against structurally and/or functionally unrelated drugs 1. Multidrug resistant cancer is the biological result of microenvironmental selection pressures that contribute to tumor progression. These selection pressures include hypoxia and changes in the regulation/expression of oncogenes, tumor suppressors, and apoptotic factors. Cellular responses to concurrent selection pressures and chemotherapeutic agents determine if the cell will become quiescent (G0 phase), engage in apoptosis, sustain its current phase, or develop MDR. MDR can be intrinsic (innate) or acquired through exposure to chemotherapeutic agents 1.
Due to the haphazard and ever-evolving vasculature of a tumor, fluctuating states of chronic and transient hypoxia can occur within the same tumor mass 7. Under conditions of hypoxia and cell stress, cancer cells undergo complex phenotype changes that can result in MDR and resistance to radiation therapy 7-13. The hypoxic transformation begins when Hypoxia Inducible Factor alpha (HIF-1α) translocates from the cytoplasm to the nucleus where it complexes with HIF-1β, forming an active transcription factor that binds to Hypoxia Responsive Elements (HRE's) on target genes, inducing transcription 12, 14. This is demonstrated in Figure 1.A..
These target genes include P-glycoprotein (P-gp) and multi-drug resistance protein (MRP-1); two drug efflux pumps that are central to the MDR phenotype. One of the most predominate mechanisms of MDR is increased drug efflux through transmembrane pumps known as ATP-Binding Cassette (ABC) transporters 15-17. P-glycoprotein (Pgp) is an ABC transporter associated with malignant transformation, poor prognosis, and the development of MDR cancer 6,15-16, 18. Multi-drug resistance protein 1 (MRP-1) and breast cancer resistance protein (BCRP) are also ABC transporters that contribute to MDR cancer 16, 18-20.
Another target of HIF is epidermal growth factor receptor (EGFR) which is correlated with aggressive tumor behavior 21. Many disease cells over-express certain growth factors and receptors to increase survival and maintain homeostasis 22. Over-expression of these receptors leads to membrane clustering. Over-expression of EGFR leads to a higher number EGFR receptors and increased receptor density (clustering) in the cell membrane which potentiates the effects of EGF 22-23. Up-regulated EGFR confers a survival advantage for many types of MDR cancer 21,22-23. As demonstrated in Figure 1.A., this targeted nanocarrier system exploits the upregulation of EGFR through the use of a targeting peptide on the surface of the nanocarriers that binds to the EGFR receptor.
Cancer cells often resort to aerobic glycolysis for energy acquisition, this phenomenon was discovered by Otto Warburg in the 1930's and has since been coined the Warburg Effect 24-28. HIF has been demonstrated to target many glycolytic proteins such as the glucose transporters GLUT-1 and GLUT-3, hexokinase 1 and 2, phosphofructokinase, aldolase, glyceraldehyde-3-phosphate dehydrogenase, phosphoglycerate kinase, enolase, pyruvate kinase, and lactate dehydrogenase 7-8, 11-14, 29-33. Reverse activation; glycolytic proteins activating HIF-1α, has also been demonstrated which implies a possible feedback loop between the glycolytic and HIF-1α pathways that may be critical in the transformation of a normal cell into a cancer cell 30.
Hexokinase 2, perhaps the most fundamental enzyme involved in the Warburg effect, is transcriptionally activated by HIF-1α and is overexpressed in many forms of cancer 12, 34-37. Hexokinase catalyzes the first step of glycolysis, conversion of glucose and ATP to glucose-6-phosphate and ADP. The hexokinase 2 isoform is directly associated with the VDAC (voltage dependent anion channel) on the mitochondrial outer membrane and the ANT (adenine nucleotide translocator) on the inner mitochondrial membrane – this association allows hexokinase 2 to directly couple the ATP produced by ATP synthase (transported out of mitochondria by ANT) to aerobic glycolysis (Figure 1.B.) 35-37. The Warburg effect is dependent on both the over-expression of hexokinase 2 and the direct coupling of hexokinase 2 to the VDAC/ANT/ATPsynthase complex 35-37. Also of biological significance, by virtue of its association with mtPTP components, hexokinase 2 has spatial anti-apoptotic activity 35. Hexokinase 2 is associated with the VDAC and the ANT; two proteins that compose the mtPTP complex. Hexokinase 2 association prevents association of pro-apoptotic BcL-2 family member proteins (Bad, Bak, Bax) with the mtPTP complex; pro-apoptotic factor association is necessary for mitochondrial permeability transition and cytochrome c release (the apoptotic cascade) (Figure 1.B.) 35.
Lonidamine (1-[(2,4-Dichlorophenyl)methyl]-1H-indazole-3-carboxylic acid) is an experimental drug originally designed as an antispermatogenic agent that has been discovered to have pro-apoptotic and anti-glycolytic activity 38-40. In the 1980's lonidamine was demonstrated to inhibit aerobic glycolysis via direct inhibition of hexokinase 2 38. Lonidamine has also been shown to induce apoptosis via induction of the mitochondrial permeability transition pore complex 40. This is consistent with the role of hexokinase 2 as a spatial apoptotic inhibitor. In cell based studies, lonidamine has been successful in treating adriamycin resistant breast cancer cells (MCF-7), nitrosourea resistant glioblastoma cells (LB9), and doxorubicin resistant hepatocarcinoma (HepG2) cells 41-43. An anti-angiogenic effect of lonidamine has also been suggested 44. Lonidamine's potential has been explored as a treatment for benign prostatic hyperplasia; clinical trials of lonidamine progressed to phase II in the USA but were halted in 2006 due to the occurrence of elevated liver enzymes in several patients 45-46. This residual toxicity could be circumvented by improving and refining the drug delivery of lonidamine.
Active Targeting is achieved by this system by using nanocarriers that have been engineered to bind to the EGFR receptor by incorporation of a PLGA-PEG-EGFR peptide construct. Active targeting can help a system to overcome biological barriers (active transport verses non-specific endocytosis), decrease the residual toxicity of a system, and increase the therapeutic effect 47.
This study describes the design, development, and characterization of EGFR-targeted nanocarriers for combination paclitaxel/lonidamine delivery to treat MDR cancer. EGFR targeting is achieved by first grafting a PEG-maleimide residue onto poly(DL-lactide-co-glycolide) (PLGA), and subsequently grafting an EGFR-specific peptide to the PEG to create a PLGA-PEG-Peptide construct. This construct is then incorporated into the nanoparticle formulation along with PLGA-PEG and poly(epsilon-caprolactone (PCL) so that the PLGA and PCL form the core of the particles while the PEG and peptide residues modify the surface of the nanoparticles. SEM, dynamic light scattering, and ESCA were used to characterize the nanocarriers. Drug loading and release were also measured. Fluorescent microscopy and fluorescent quantification were used to assess the targeting ability of these nanocarriers in a panel of nine cell lines with various levels of EGFR expression.
The two therapeutic agents selected for combination therapy are paclitaxel and lonidamine. Paclitaxel is a common chemotherapeutic agent that hyper-stabilizes microtubules, preventing cell division. Although the clinical use of paclitaxel is common, its application is often limited by the residual toxicity associated with the drug. Lonidamine is a therapeutic agent with great clinical potential in the treatment of MDR cancer as it acts to inhibit aerobic glycolysis and induce apoptosis. Yet, the clinical application of lonidamine is impeded by the occurrence of residual liver toxicity. The low bioavailability and liver toxicity demonstrated by lonidamine in clinical trials make it an ideal candidate for nanoparticle delivery. The MTS assay was used to assess the efficacy of combination therapy.
An established EGFR specific peptide with the following sequence was used to achieve active targeting with the nanoparticle formulation: YHWYGYTPQNVI-GGGGC 48-49. This peptide, termed GE11, was originally synthesized and screened as an EGFR specific peptide by Zonghai Li and colleagues (Shanghai, PR China) 48-49. The “GGGG” sequence functions as an adequate spacer while the carboxyl terminal cysteine of the peptide reacts with the maleimide of the PLGA-PEG construct. The peptide was synthesized by Tufts University Core Facility, Boston, MA.
The peptide-PEG-PLGA construct was synthesized, with slight modifications, according to established methods 50-52. One gram 50:50 poly(DL-lactide-co-glycolide) (PLGA) with an inherent viscosity of 0.15-0.25 and molecular weight of 10,000 (Durect Lactel® Adsorbable Polymers; Pelham, AL) was dissolved in 2 mL chloroform (Fisher Scientific;Pittsburgh, PA) while stirring in a tightly sealed vial. Once dissolved, 45 mg EDC (Fisher Scientific) and 27 mg NHS (Fisher Scientific) was added and the mixture stirred overnight in a tightly sealed vial. The product was precipitated with 8 mL diethyl ether (Fisher Scientific) and centrifuged at 4,000g for 10 minutes. Following centrifugation, the supernatant was discarded and the polymer re-dissolved in 2 mL chloroform. This washing cycle of precipitation/dissolution was repeated three times before drying the activated PLGA-NHS ester under vacuum. Then 0.5 g PLGA-NHS was dissolved in 2 mL chloroform and once fully dissolved, 0.125 g amine- poly(ethylene glycol) PEG-maleimide (MW 2000; JenKem Technology; Allen, TX) and 20ul N,N-diisopropylethylamine (Fisher Scientific) was added, the vial tightly sealed, and the mixture stirred overnight. The product was precipitated with 8 mL of an ice-cold 80/20 mix of diethyl ether/methanol and centrifuged at 4,000g for 10 minutes. After discarding the supernatant the product was re-dissolved in 2 mL chloroform and the washing cycle repeated two more times before drying the PLGA-PEG under vacuum. To prepare a PLGA-PEG conjugate for non-targeted particles, m-PEG-amine (MW 2,000; LaysanBio; Arab, AL) was used in replacement of the amine-PEG-maleimide.
For peptide conjugation to PLGA-PEG, 20 mg of the PLGA-PEG and 2 mg of peptide was dissolved in 800 uL of 50/50 acetonitrile/DMF, covered tightly, and stirred overnight. The product was precipitated with 3 mL of an ice-cold 80/20 mix of diethyl ether/methanol, centrifuged at 4,000g for 10 minutes, the supernatant discarded, and the product re-dissolved in 800 uL 50/50 acetonitrile/DMF. This cycle was repeated two more times and then the PLGA-PEG-peptide was dried under vacuum. NMR was used to assess grafting of PEG to PLGA and conjugation of the peptide to PEG.
EGFR targeted and non-targeted polymer blend nanoparticles were synthesized using a solvent displacement method 53-54. Briefly, the PLGA-PEG-peptide conjugate (or the PLGA-PEG conjugate for non-targeted particles), PCL (MW 10,000), and therapeutic agents were dissolved in 2 mL 50/50 acetonitrile/DMF, and placed in a 37°C water bath for 10 minutes to facilitate dissolution. This polymer/drug solution was added drop-wise to 20 mL distilled, deionized water while stirring. The preparation was covered with aerated parafilm, allowed to stir overnight, centrifuged at 10,000g for 30 minutes, and then resuspened in di water. To synthesize targeted nanoparticles the PLGA-PEG-peptide conjugate was added to the nanoparticle formulation at 20% w/w total polymer, with an additional 10% w/w of PLGA-PEG conjugate. Likewise, the PLGA-PEG conjugate was added at a concentration of 20% w/w total polymer. Control nanoparticles synthesized to determine the optimal loading efficiency were composed of either PCL or PLGA, surface modified with PEO (Pluronic® F-108 NF). To assess peptide modification of the nanoparticles samples were sent to the National ESCA and Surface Analysis Center for Biomedical Problems (Seattle, WA). XPS spectra were taken on a Surface Science Instruments S-probe spectrometer. Three spot analysis was done on each batch of nanoparticles.
After determining the optimal dose combination, dual agent loaded nanoparticles were synthesized with a 10:1 molar ratio of lonidamine to paclitaxel. Loading efficiency of the nanoparticles was determined by lyophilizing the nanoparticles to obtain a dry powder, dissolving the particles in acetone, and measuring the absorbance using a Bio-Tek Synergy HT plate reader at 200 nm for paclitaxel and 300 nm for lonidamine (Winooski, RI). Nanoparticle size and zeta potential were characterized using a Brookhaven ZetaPlus Particle Analyzer (Brookhaven Instruments; Holtsville, NY). SEM images of the nanoparticles were obtained using a Hitachi S-4800 microscope.
To measure drug release from nanoparticles, nanoparticles were lyophilized, weighed, and resuspended in two different PBS/0.1% Tween-80 buffers; one at pH 7.4 and one at pH 6.5 – then incubated in a 37°C water bath. At various time points between 15 minutes and ten days an aliquot of eluted drug medium was removed for quantification; this volume was replaced with fresh buffer. Drug release was quantified by measuring the absorbance of the release media using a Bio-Tek Synergy HT plate reader (Winooski, RI).
For initial dose response studies, single agent loaded nanoparticles were synthesized; paclitaxel was loaded at 10% w/w and after determining maximal loading, lonidamine was also loaded at 10% w/w. After determining the optimal dose combination, dual agent loaded nanoparticles were synthesized with a 10:1 molar ratio of lonidamine to paclitaxel.
SKOV3 cells, MDA-MB-231 cells, and OVCAR5 cells were obtained from ATCC (Manassas, VA). The SKOV3-TR cells and the MDA-MB-435 cells were a kind gift from Dr. Duan (Massachusetts General Hospital, Sarcoma Molecular Biology Laboratory). Cells were incubated at 37°C and maintained in RPMI-1640 media (Mediatech, Inc; Manassas, VA) supplemented with 10% fetal bovine serum (Gemini Bio-products; West Sacramento, CA) and 1% penicillin/streptomycin/amphotericin B mixture (Lonza; Walkersville, MD). Hypoxia was used to induce MDR. To create hypoxic conditions using low-oxygen gas; cell culture flasks were placed in a modular incubation chamber (Billups-Rothenberg, Inc.; Del Mar, CA), flushed with a 0.5% O2, 5% CO2, nitrogen balanced gas for five minutes, and incubated at 37°C for various time points.
For cell viability, cells were plated in 96-well plates at 2,000 cells per well and treated with various concentrations of nanoparticle formulations and drug solutions suspended in FBS supplemented media. Control treatments with supplemented media, blank nanoparticles and poly(ethyleneimine) were also conducted. Treatment continued undisturbed for 5 days under normoxic and hypoxic conditions. Cell viability was then measured using the MTS assay according to the manufacturer's protocol (Promega; Madison, WI).
Basal protein was extracted from cells grown to 90% confluency in 75 cm2 tissue culture flasks under normoxic and hypoxic conditions. Basal protein was extracted using a high salt lysis buffer at 4°C. Protein concentrations were quantified using the BCA Protein Assay (Pierce Biotechnology). Protein was separated on 4-20% gradient SDS-PAGE gels (PAGEgel, Inc.; San Diego, CA) and transferred onto PVDF membranes (0.45 μm pore; Millipore, Billerica, MA). Membranes were blocked for 30 minutes with StartingBlock™ buffer (Pierce Biotechnology) before a 1 hour incubation with the primary antibody. Membranes were then washed with TBST for 10 minutes (three times) and subsequently incubated with a horseradish peroxidase conjugated secondary antibody for 1 hour. Membranes were again washed with TBST, flash rinsed with deionized distilled water, incubated for 2-10 minutes in an enhanced chemiluminescence substrate (Pierce Biotechnology), and imaged using a Kodak FX Imaging Station (Rochester, NY). All blocking, probing, and washing steps were conducted at room temperature. P-glycoprotein antibody was purchased from Calbiochem while the EGFR and β-actin antibodies were purchased from Cell Signaling Technology (Danvers, MA). The secondary antibodies were purchased from Abcam (Cambridge, MA).
To assess and visualize particle uptake kinetics and quantity, nanoparticles were prepared as previously mentioned and loaded with 1% w/w rhodamine 123 (Invitrogen; Carlsbad, CA). Cells were seeded at 10,000 cells per well and treated with rhodamine 123 loaded nanoparticles suspended in supplemented media (5 μM with respect to rhodamine). For competitive binding studies 100-fold molar excess of EGFR antibody (Cell Signaling Technology, Danvers, MA) was added to the media (relative to the amount of peptide on the nanoparticles). At various time points (from 15 minutes to 6 hours), the treatment solution was removed and cells were washed three times with supplemented media. For quantification of uptake, plates were read on a Bio-Tek Synergy HT plate reader (Winooski, RI) at 485ex/528em. For microscopy, wells were also treated with Hoechst 33342 nucleic acid stain (Invitrogen) 15 minutes before washing. Cells were visualized using an Olympus IX51 Microscope.
All statistical analysis was done using GraphPad Prism® software. Statistical significance of uptake was determined using one-way ANOVA and Bonferroni's Multiple Comparison Test. For cell viability studies, statistical significance of treatments were determined using a two-tailed, Student's t test at a 95% confidence interval (P< 0.05). For cell viability studies n = 8 for each treatment group. IC50 values for paclitaxel treatment were calculated using variable slope curve fitting of transformed data.
The primary objective in designing this drug delivery system was to design a system capable of delivering a therapeutic dose of lonidamine and paclitaxel to the site of a tumor. To accomplish this, the system must: (1) achieve high encapsulation of both drugs and (2) be surface modified with PEG to avoid immediate RES clearance, increasing resident circulation and ability to reach the tumor. A secondary design objective of this formulation was to synthesize a polymer-PEG conjugate and a polymer-PEG-EGFR peptide conjugate and incorporate these constructs to achieve PEG surface modification of the nanoparticles and to achieve active targeting of MDR cells (via EGFR binding). To accomplish this objective, m-PEG-amine (MW 2000) was conjugated to NHS-activated PLGA (to use for PEG modification of the formulation). For the active targeting construct, amine-PEG-maleimide was conjugated to NHS-activated PLGA and an EGFR specific peptide (YHWYGYTPQNVI-GGGGC) was conjugated to the PEG via cysteine/maleimide linkage. NMR spectra of the conjugate and raw materials were acquired to evaluate this reaction (data not shown).
To select the optimal polymer for this drug delivery system, a variety of polymeric nanoparticles were synthesized and drug loading efficiency was evaluated (Figure 2.A. and 2.B.). The loading efficiency of PEO (Pluronic® F-108 NF) modified PCL particles, PEO (Pluronic® F-108 NF) modified PLGA particles, and PLGA particles with the PLGA-PEG construct were compared. All formulations were evaluated for 10% w/w loading efficiency of paclitaxel and of lonidamine; 20% w/w loading of lonidamine was also examined for the PEO-PCL formulation. The highest loading efficiency of paclitaxel was achieved with the PEO-PLGA formulation (88%), yet the PEO-PCL and the PLGA/PEG-PLGA conjugate nanoparticles also achieved high loading efficiencies; 80% and 83% respectively. The loading efficiency of PEO-PCL nanoparticles loaded with 10% w/w lonidamine was 70%; this decreased to 30% loading efficiency when the dose was increased to 20% w/w. For both the PEO-PLGA nanoparticles and for the PLGA/PEG-PLGA construct nanoparticles, lonidamine loading was only 5%. The decrease in lonidamine loading with a higher w/w loading in PCL and the substantially lower loading in PLGA based nanoparticles is most likely due to the hydrophobicity of lonidamine. As PCL is more hydrophobic than PLGA, higher encapsulation is possible with PCL. Yet there seems to be a maximal threshold above which lonidamine encapsulation decreases, most likely due to excessive hydrophobicity of the polymer/drug solution which is not conducive to nano-precipitation via the solvent displacement method.
Based on these findings, it was determined that the predominate polymer constituent of the nanoparticle formulation needed to be PCL in order to achieve any appreciable encapsulation of lonidamine. To achieve the design goals of the system (high encapsulation and surface modification with a malleable construct), a polymer blend nanoparticle formulation of PCL, PLGA-PEG construct, and the PLGA-PEG-peptide construct was developed. The PCL and the PLGA of the conjugate should form the core of the nanoparticle with the PEG and peptide residues extending out from this core. To assess the capability of these polymer blend nanoparticles, nanoparticles were synthesized with single agent loading (paclitaxel alone and lonidamine alone) and with dual agent loading and the loading efficiency was determined (Figure 2.C.). Maintaining a predominately PCL based system allowed for encapsulation comparable to PEO-PCL nanoparticles (70% for lonidamine, 80% for paclitaxel). Combining the two drugs into one formulation did not have a great impact on the loading efficiency relative to formulations loaded with single agents.
To further characterize the polymer blend nanoparticles, the size and zeta potential of the combination drug loaded and unloaded polymer blend nanoparticles and targeted polymer blend nanoparticles were measured (Table 1). PEO-PCL nanoparticles were the largest particles which also had the most negative zeta potential (174.3 nm; -33.1 mV) while the PLGA/PEG-PLGA conjugate nanoparticles were the smallest, also bearing the most positive zeta potential (70.1 nm; -26.13 mV). The unloaded polymer blend nanoparticles were smaller than the PEO-PCL nanoparticles with a slight reduction in zeta potential (129.2 nm; -30.12 mV). This size reduction relative to the PEO-PCL nanoparticles may be attributed to the structural difference between the PEO and the PEG-PLGA graft. The PEO (Pluronic® F108 NF) is of an ABA structure where the more hydrophobic (B) block would interact with the core of the nanoparticle while the A-blocks branch out, forming a V-shape with the central V-apex interacting with the PCL core. The PEG-PLGA conjugate on the other hand, does not have this branching structure; the PLGA of the construct interacts with the PCL core while the PEG end protrudes on the surface of the particle. This more linear structure may be conducive to smaller particle sizes as the core becomes more compact (no branching V-type components). This would also explain the smaller size of the PLGA/PEG-PLGA conjugate nanoparticles compared to PEO-PLGA nanoparticles (70.1 nm verses 122.2 nm). Drug loading did not drastically alter the size or zeta potential of the polymer blend nanoparticles (123.4 nm; -31.4 mV for drug loaded and 129.2 nm; -30.12 mV for unloaded). Adding the targeting construct increased the particle size; from 123.4 nm to 139.6 nm for drug loaded nanoparticles and from 129.2 nm to 151.5 nm for unloaded nanoparticles. There was only a slight reduction in zeta potential; this is most likely due to the neutrality of the peptide (isoelectric point near 7.4).
SEM images of the non-targeted and targeted nanoparticles confirmed the nanometer scale of the particles (Figure 3.A. and Figure 3.B.). The targeting construct is depicted in Figure 3.C. To confirm that the EGFR peptide was on the surface of the nanoparticles electron spectroscopy for chemical analysis (ESCA) was used to examine the surface of the particles at an approximate depth of 50 Å (~ 5 nm) (Figure 3.D). The percent composition of carbon and oxygen was similar for the targeted and non-targeted nanoparticles, suggesting similar PEG modifications and polymer compositions. However, there was no detectable level of nitrogen on the surface of the non-targeted nanoparticles whereas the targeted nanoparticles had approximately 2% nitrogen composition. This correlates to the total w/w ratio of the peptide in the targeted nanoparticles formulation.
An important characteristic of a nanoparticle formulation is its drug release profile. Just as drug release from common dosage forms can determine clinical application, the release of drugs from a nanocarrier system must be sustained to justify the system and ensure the drug is released at the target site, not during systemic circulation. The release of lonidamine and paclitaxel from combination polymer blend nanoparticles (dual loaded), singly loaded polymer blend nanoparticles (either paclitaxel alone or lonidamine alone), targeted combination polymer blend nanoparticles (dual loaded), targeted singly loaded polymer blend nanoparticles (either paclitaxel alone or lonidamine alone), and from PEO-PCL nanoparticles was measured over the course of ten days (Figure 4). Paclitaxel release is presented in the top panel of Figure 4 (Figure 4.A. and Figure 4.B.) while lonidamine release is depicted in the bottom panel (Figure 4.C. and Figure 4.D.) Release was measured both at physiological pH 7.4 (Figure 4.A. and Figure 4.C.) and at a pH common to the microenvironment of a tumor (6.5) (Figure 4.B. and Figure 4.D.). The figure insets represent the early time points (from 0 to 6 hours).
As demonstrated by the early time points (Figure 4, insets), none of the formulations exhibited burst release of lonidamine at pH 7.4 (Figure 4.C.) and at pH 6.5 (Figure 4.D.). The release kinetics of lonidamine from all three formulations did not appear to be greatly affected by pH as the profiles were similar for both pH 7.4 and pH 6.5. By 72 hours, 100% of the dose is released from both dual loaded polymer blend nanoparticles and the singly loaded polymer blend nanoparticles at pH 7.4 (the percentage of release after 72 hours at pH 6.5 is slightly reduced to 93%). After 72 hours only 58% of the lonidamine dose has been released from the PEO-PCL nanoparticles at pH 7.4 (57% at pH 6.5); complete release from the PEO-PCL occurs after 144 hours (both pH values). After 72 hours only 39% of the total lonidamine dose was released from the targeted, dual loaded polymer blend particles and 41% of total lonidamine was released by the targeted, singly loaded polymer blend nanoparticles (neither of these values changed in response to pH). The targeted nanoparticles exhibited the most sustained release for lonidamine; 100% of the total loaded lonidamine was released after 168 hours for the dual loaded formulation and after 192 hours for the singly loaded formulation (there was no change with a change in pH). This prolonged release may be attributable to the spatial interactions of the polymer/drug mixture in the nanoparticle core, the density of the particle, and the surface protection of the formulation. The PEO-PCL particles may be more rigid and stable (owing to the B-block interaction of PEO with PCL), this rigidity may enhance the association of lonidamine with the polymers and as PCL degradation is very slow, this association may extend the release of the drug. The targeted nanoparticles, on the other hand, have extensive surface modification relative to the other formulations; this may prolong their degradation. The targeted nanoparticles may also have a very dense core of drug/PCL/PLGA tightly packed by the peptide and PEG protrusions.
Paclitaxel release kinetics were also sustained over the course of several days for the five formulations (Figure 4.A. and Figure 4.B.). None of the formulations exhibited burst release at early time points at both pH 7.4 (Figure 4.A., inset) and pH 6.5 (Figure 4.B., inset). The dual loaded polymer blend nanoparticles have a much greater retention of paclitaxel than lonidamine; after 72 hours at both pH 7.4 and pH 6.5, 100% of lonidamine is released where as at 72 hours only 52% of total loaded paclitaxel is released (57% at pH 6.5). Singly loaded lonidamine and paclitaxel polymer blend nanoparticles also exhibit this retention difference. This could be due to the bulkier structure of paclitaxel; it may be that paclitaxel is more tightly associated with the core polymers. One hundred percent of the loaded paclitaxel dose is not released until 168 hours for both the dual loaded and singly loaded polymer blend nanoparticles at pH 7.4 (this value is the same for the dual loaded polymer blend nanoparticles at pH 6.5 but is reduced to 144 hours for the singly loaded polymer blend nanoparticles at pH 6.5). As with lonidamine, the PEO-PCL nanoparticles demonstrate more prolonged release of paclitaxel; it is not until 216 hours (at both pH values) that this formulation releases 100% of the paclitaxel dose. Similar to lonidamine release kinetics, the targeted formulations exhibit sustained release for paclitaxel; it is not until 216 hours that 100% of the paclitaxel dose is released by the targeted formulations (this is released to 192 hours for the dual loaded targeted particles at pH 6.5). The sustained release kinetics of lonidamine and paclitaxel from the targeted combination polymer blend nanoparticles ensures that the therapeutics will not completely leach out of the particles during circulation, but will be released once the particles are internalized and digested by the target cells.
In certain cancer cells, hypoxia has been shown to contribute to MDR 7-13. We have conducted extensive analysis of the relationship between hypoxia and MDR in this panel of cell lines (work submitted for publication). In the current study, we used hypoxia to induce MDR in a panel of cell lines. Pgp expression was used as a positive marker for MDR (Figure 5). EGFR expression was also analyzed. As demonstrated in Figure 5, hypoxia induced the expression of Pgp in SKOV3 cells after 5 days of hypoxia and in MDA-MB-231 cells after 3 days of hypoxic exposure. The OVCAR5 cells appeared resistant to hypoxic induction of Pgp. The established MDR cell line, SKOV3-TR cells were a positive control for Pgp expression. Hypoxia also increased the expression of EGFR in the SKOV3 cell line and in the MDA-MB-231 cell line. The OVCAR5 cells have a high basal level of EGFR. The MDA-MB-435 cells were used as a negative control for EGFR expression.
To confirm that the nanoparticles were being taken up by the cells, nanoparticles were loaded with rhodamine 123 and incubated with the panel of cells (SKOV3-TR, SKOV3, MDA-MB-231, OVCAR5, and MDA-MB-435 cells) under normoxic and hypoxic conditions. At time intervals spanning from 15 minutes to 6 hours images of the cells were taken (Figures 6--10,10, Panel A; only the 15 and 30 minute time points are shown) and the fluorescent intensity was quantified as a percentage of the total administered dose (Figures 6--10,10, Panel B). The panel of cell lines were treated with targeted nanoparticles (T), targeted nanoparticles in the presence of competitive EGFR antibody (TC), non-targeted nanoparticles (N), and non-targeted nanoparticles in the presence of competitive EGFR antibody (NC). Cells were treated with a 5 μM dose of the nanoparticles (relative to rhodamine loading) and examined after 15 minutes of treatment, 30 minutes, 1 hour, 2 hours, 3 hours, and 6 hours. Nuclei were stained with Hoechst 33342 (blue). For competitive studies, a 100-fold molar excess of the EGFR antibody (relative to the peptide content of the nanoparticle dose) was added to the cells 10 minutes before treatment with the particles.
Figure 6.A. demonstrates the uptake in the MDR cells after 15 minutes and 30 minutes of treatment. The cellular uptake after 15 minutes and 30 minutes of treatment was also monitored in hypoxic and normoxic derivatives of SKOV3 cells (Figure 7.A.), MDA-MB-231 cells (Figure 8.A.), OVCAR5 cells (Figure 9.A.), and MDA-MB-435 cells (Figure 10.A.). Although the microscopy demonstrates the differences in uptake kinetics between the different formulations and the between the time points; the data is qualitative and does not demonstrate clear differences between the hypoxic and normoxic derivatives. As such, quantitative nanoparticle uptake was measured on a Bio-Tek Synergy HT plate reader and the results are presented in Panel B of Figures 6--1010.
The degree of targeted nanoparticle uptake at the early time points (15 minutes and 30 minutes) paralleled the expression level of EGFR in the panel of cell lines. The highest expression of EGFR was in the OVCAR5 normoxic and hypoxic derivatives, followed by the SKOV3TR cells and the hypoxic MDA-MB- 231 cells, then the SKOV3 hypoxic cells, the SKOV3 normoxic and MDA-MB231 normoxic cells, and lastly the MDA-MB-435 cells which had no EGFR expression. Inhibition of targeted nanoparticle uptake in the presence of excess EGFR antibody is indicative of competitive binding between the nanoparticles and the free antibody. This inhibition was significant in all EGFR-expressing cell lines after 15 and 30 minutes of treatment and also after 1 hour of treatment in the OVCAR5 derivatives (Figures 6--10).10). No competition was observed in the EGFR-negative MDA-MB-435 cell line (Figure 10). This competitive binding verifies that the targeted nanoparticles are engaging the EGFR receptor.
The degree of targeted nanoparticle uptake relative to non-targeted nanoparticle uptake was also significant at the early time points for all EGFR expressing cell lines. There was no appreciable uptake of the non-targeted nanoparticles in any cell line after 15 minutes of treatment. After 30 minutes of treatment, non-targeted nanoparticle uptake ranged from 1.92% of the total administered dose (in the SKOV3 derivatives) to 2.20% (in the normoxic MDA-MB-435 cells). There was no significance between the non-targeted nanoparticles in the presence and absence of free EGFR antibody (Figures 6--1010).
As indicated by the quantitative data, there was significance between targeted nanoparticle uptake in normoxic and hypoxic derivatives of the SKOV3 and MDA-MD-231 cells after 30 minutes of treatment. The lack of significance at the 15 minute time point may be because 15 minutes is not adequate time to demonstrate a difference between the hypoxic and normoxic uptake by saturating the receptors (although it is adequate time to demonstrate a difference between the non-targeted and targeted nanoparticles). At the 30 minute time point in the the SKOV3 and MDA-MD-231 cell lines there was no difference between the accumulation of targeted nanoparticles in normoxic cells and between targeted nanoparticles in the presence of competitive EGFR antibody under normoxic conditions, and between non-targeted nanoparticles under normoxic and hypoxic conditions. For the MDA-MB-231 cell line there was also no difference between the uptake of targeted nanoparticles under normoxic conditions and targeted nanoparticles in the presence of competitive EGFR antibody under hypoxic conditions. At this time point in these cell lines, targeting enhances uptake proportionate to EGFR expression.
The lack of significance between the uptake of targeted nanoparticles in normoxic and hypoxic derivative of the OVCAR5 cells may be due to the excessively high basal expression of EGFR in the normoxic derivatives. Targeting does show a clear engagement of the EGFR receptor at 15 minutes (demonstrated by competitive binding) and enhanced uptake relative to non-targeted nanoparticles at 15 minutes for the SKOV3 cells and the MDA-MD-231 cells and upto the 1 hour time point for the OVCAR5 cells.
The difference between non-targeted nanoparticle uptake (no appreciable uptake) and targeted nanoparticle uptake at the early time points indicate that the two different formulations may engage in distinct mechanisms of uptake. It has been established that non-targeted nanoparticles are internalized via non-specific endocytosis 47, 55-59. Most likely, the EGFR targeted nanoparticles are internalized by mechanisms similar to endogenous EGFR substrates (EGF). Studies have shown that after endogenous substrate binding to EGFR and by environmental activation (radiation), EGFR and the bound substrate are internalized via a process characteristic to this protein tyrosine kinase 60-61. This characterized receptor/substrate internalization is a rapid process, occurring within 10 to 20 minutes of EGFR stimulation 61. This process is more similar to a flip-flop mechanism than to a membrane budding process 60-61. As the EGFR-targeted nanoparticles are an extension of an EGFR substrate (the peptide), the mechanism of targeted nanoparticle uptake is most likely via the receptor/substrate internalization process.
After one hour of treatment for every cell line (except the highly EGFR expressing OVCAR5 derivatives), there is no significance between the targeted and non-targeted nanoparticle uptake (Figure 6--10,10, Panel B). This suggests that the kinetic differences between non-specific endocytosis and receptor/substrate internalization are normalizing. Also at this time point, there is no observed competition between the targeted nanoparticles and the free antibody in all cell lines (again, except for the OVCAR5 derivatives). It is possible that after 1 hour, the free antibody has already engaged EGFR receptors and been internalized, restoring the receptors for targeted nanoparticle binding. It is also likely that after 1 hour of treatment uptake of all formulations by non-specific endocytosis masks any kinetic differences due to competitive receptor/substrate internalization.
The uptake studies verify that cellular uptake kinetics are undoubtedly cell-type specific. However, there is a general trend of increasing particle uptake in each cell line from 15 minutes until 6 hours at which point between 80% of the total administered dose (for the MDA-MB-231 normoxic cells treated with targeted nanoparticles in the presence of competitive antibody) and 100% of the administered dose (for the OVCAR5 hypoxic derivatives treated with non-targeted nanoparticles) had been taken up. Interestingly, there does not appear to be any appreciable nanoparticle uptake in the MDA-MB-231 cells between 2 and 3 hours (Figure 8). The MDA-MB-231 cells may reach a saturation threshold (after 2 hours) at which point the cell increases metabolism of the internalized particles and vesicles, restores phospholipids to the cell membrane, apparently ceasing additional uptake, before resuming internalization at a slower rate. This would explain the dramatic increase in particle uptake between 1 hour and 2 hours (an average of 39% of the total dose), the lack of change between 2 and 3 hours, and the subsequent slow uptake between 3 hours and 6 hours (an average rate of 10% total dose per hour). Also of note, the MDA-MB-435 cells appear to have a slower initiation of particle uptake. After two hours of treatment approximately 35% of the administered dose was taken up by the MDA-MB-435 cells which is 10% lower than uptake by the SKOV3 derivatives (with no significance between the nanoparticle formulations). For each cell line, the particle uptake kinetics visualized during microscopy (Figures 6--10,10, Panel A) were consistent with the kinetics observed in quantifying cumulative nanoparticle uptake (Figures 6--10,10, Panel B) and confirmed interaction of the targeted nanoparticles with the EGFR receptor.
It is not unusual that the benefit of active targeting was only demonstrated during the early time points. This analysis was done in a stationary cell culture system where saturation (even by non-targeted nanoparticles) is fairly rapid and inevitable as particle uptake becomes dominated by non-specific endocytosis. However, an in vivo study would be expected to demonstrate the true potential of a targeted nanocarrier system. We have also completed in vivo studies which show improved pharmacokinetic parameters and enhanced efficacy of this targeted system (work to be published).
To determine the optimal combination of lonidamine and paclitaxel, a dose response study was conducted in three cell lines (SKOV3 ovarian cancer cells, MDA-MB-231 breast cancer cells, and OVCAR5 ovarian cancer cells) under normoxic and hypoxic conditions as well as in MDR cells (SKOV3-TR ovarian cancer cells). Cells were treated with drug loaded nanoparticle formulations and with solution forms of the drugs. First, single agent treatment was examined (data not shown). Dose response to paclitaxel encapsulated in nanoparticles and paclitaxel solution was measured over the range of 0.001 μM to 10 μM. The cells that were the most resistant to paclitaxel were the MDR cells (SKOV3-TR), the hypoxic SKOV3 cells, and the hypoxic MDA-MB-231 cells. The OVCAR5 cells were the most sensitive to paclitaxel. As paclitaxel is a Pgp substrate, these responses can be correlated with the expression profile of Pgp in the cells (Figure 5). Hypoxia induced the expression of Pgp in SKOV3 and MDA-MB-231 cells (which are the most resistant, second to the MDR cells), yet the OVCAR5 cells did not express basal levels of Pgp and expression was not induced by hypoxia (most sensitive to paclitaxel).
The IC50 values for paclitaxel treatment are listed in Table 2. The IC50 values for MDR SKOV3 cells treated with nanoparticle formulations of paclitaxel are 92 times the IC50 values for the wild-type, normoxic SKOV3 cells with similar treatment; the MDR cells treated with paclitaxel solution had an IC50 value 79 times that of wild-type normoxic cells treated with solution. While the IC50 values for the hypoxic SKOV3 cells are 72 times that of the wild-type, normoxic SKOV3 cells treated with nanoparticle encapsulated paclitaxel and the hypoxic SKOV3 cells treated with solution have an IC50 value 64 times that of wild-type normoxic SKOV3 cells treated with paclitaxel solution. This increased IC50 value in hypoxic cells verifies the dynamic cellular transformation that occurs during hypoxia; shifting the cells toward a MDR character (portrayed in Figure 5). Likewise, the IC50 for hypoxic MDA-MB-231 cells treated with nanoparticle paclitaxel is 70 times that of the wild-type, normoxic cells treated with the formulation and hypoxic MDA-MB-231 cells treated with solution paclitaxel had an IC50 65 times that of wild-type, normoxic breast cancer cells treated with solution. Yet, the IC50 values of the hypoxic OVCAR5 cells treated with paclitaxel nanoparticles was only 13 times that of the wild-type, normoxic, OVCAR5 cells treated with the same formulation and the IC50 of the hypoxic OVCAR5 cells treated with paclitaxel solution was 20 times that of the wild-type, normoxic OVCAR5 cells treated with solution paclitaxel. This minimal increase in IC50 values for the OVCAR5 cells subjected to hypoxia coincides with the lack of hypoxic transformation in the OVCAR5 cells demonstrated by the protein analysis (Figure 5). The IC50 values again clarify that hypoxia seemed effective in inducing MDR character in the SKOV3 cells and in the MDA-MB-231 cells, yet this selection pressure did not reach the threshold for transforming OVCAR5 cells. Lonidamine treatment alone had very little effect at 1 μM and 10 μM doses and for most cells did not even reach an IC50 value with a 100 μM dose (data not shown). The response of the different cell types to lonidamine is also consistent with the protein expression profile and glycolytic character of the cells (previous work submitted for publication). The OVCAR5 cells had extremely low glycolytic character and as lonidamine is a hexokinase-2 inhibitor, lonidamine alone had very little effect even at a 100 μM dose.
Based on the dose response of single agents, doses of 0.1 μM paclitaxel (below the IC50 of MDR and hypoxic cells), 1 μM (the approximate IC50 of MDR and hypoxic cells), and 10 μM paclitaxel (above the IC50 of MDR and hypoxic cells) were selected for a combination study with 1, 10, and 100 μM lonidamine (data not shown). As a result of the dose responses of the cell lines to paclitaxel/lonidamine combination treatment, 1 μM paclitaxel and 10 μM lonidamine were selected as the optimal dose combination. This dose enhanced the efficacy of both drugs in the MDR cell line and in the hypoxic and normoxic SKOV3 and MDA-MB-231 cells.
After selection of the optimal dose combination, EGFR-targeted nanoparticles dual loaded with lonidamine and paclitaxel at a dose ratio of 10:1 were synthesized using PCL, the PLGA-PEG conjugate (10% w/w and the PLGA-PEG-EGFR peptide construct (20% w/w). Cells were treated with the nanoparticles at a dose of 1 μM paclitaxel/10 μM lonidamine and viability was assessed after five days of treatment (Figure 11). Treatment with the lonidamine/paclitaxel loaded EGFR targeted nanoparticles resulted in less than 10% cell viability for all cell lines and less than 5% cell viability for the wild-type, normoxic SKOV3 and MDA-MB-231 cells (Figure 11.A.). For all cell lines, there was significance between the EGFR-targeted combination blend nanoparticles and drug solution (Figure 11.A.). This enhanced efficacy is most likely due to the sustained drug release from the nanoparticles.
Treatment with the dual loaded lonidamine/paclitaxel EGFR-targeted nanoparticles was compared to treatment with nanoparticles only loaded with paclitaxel; also, treatment with a solution combination of lonidamine and paclitaxel was compared to treatment with paclitaxel solution alone (Figure 11.B.). Combination treatment was at a dose of 1 μM paclitaxel and 10 μM lonidamine; this was compared to treatment with 1 μM paclitaxel alone. Combination treatment with lonidamine/paclitaxel nanoparticles enhanced the cell kill efficacy relative to treatment with paclitaxel nanoparticles (extremely significant, p < 0.0001, for the MDR, all hypoxic cell derivatives, and the normoxic MDA-MB-231 cells; and significant, p < 0.05, for the SKOV3 wild-type normoxic cells and the OVCAR5 wild-type normoxic cells). Solution treatment with a combination of lonidamine and paclitaxel enhanced the cell kill efficacy relative to treatment with paclitaxel solution alone for all cell lines except for the wild-type, normoxic OVCAR5 cells (significant at p < 0.0001, for the SKOV3 hypoxic cells and for the MDA-MB-231 normoxic cells; very significant, p < 0.01, for the MDR and MDA-MB-231 hypoxic cells; significant, p < 0.05, for the SKOV3 normoxic cells and for the OVCAR5 hypoxic cells; and not significant for the OVCAR5 normoxic cells). As previously mentioned, the decreased response of the OVCAR5 normoxic cells to lonidamine is most likely due to the lower glycolytic character of these cells. For the hypoxic and normoxic SKOV3 and MDA-MB-231 cells as well as for the MDR cells, combination treatment with lonidamine and paclitaxel at the dose ratio of 10 μM lonidamine: 1 μM paclitaxel greatly enhances the cell kill efficacy relative to treatment with 1 μM paclitaxel alone (Figure 11.B.).
This novel drug delivery system achieves the functional goals of the system; improved efficacy with combination therapy and active EGFR targeting. The nanocarriers are capable of appreciable drug encapsulation and sustained drug release. ESCA analysis confirmed that the surface of the nanocarriers were modified with the targeting construct. This was further confirmed with the demonstrated targeting of the nanocarriers in the panel of cell lines; uptake kinetics emulated the expression of EGFR in the cell lines and competitive inhibition by an EGFR antibody was apparent. Combination therapy with 1 μM paclitaxel and 10 μM lonidamine resulted in 5-10% cell viability whereas treatment with 1 μM paclitaxel alone was the approximate IC50 of the MDR cells. This drug delivery system is a well characterized platform for the delivery of paclitaxel and lonidamine to treat MDR cancer.
This study was supported by the National Cancer Institute, National Institutes of Health through grants R01 CA-119617 and R01 CA-119617S1 (ARRA Supplement), and R21 CA-135594. Scanning electron microscopy was performed by William Fowle at the Electron Microscopy Center at Northeastern University.