EGFR-mediated Intracellular Delivery of Pc 4 Nanoformulation for Targeted Photodynamic Therapy of Cancer: In Vitro Studies

Abstract

In photodynamic therapy (PDT), the light-activation of a photosensitizer leads to the generation of reactive oxygen species that can trigger various mechanisms of cell death. Harnessing this process within cancer cells enables minimally invasive yet targeted cancer treatment. With this rationale, here we demonstrate tumor-targeted delivery of a highly hydrophobic photosensitizer Pc 4 loaded within biocompatible PEG-PCL block-copolymer micelles. The micelles were surface-modified with EGFR-targeting GE11-peptides for active targeting of EGFR-overexpressing cancer cells, in vitro. Pc 4-loaded EGFR-targeted micelles were incubated with EGFR-overexpressing A431 epidermoid carcinoma cells for various time periods, to determine Pc 4 uptake by epifluorescence microscopy. The cells were subsequently photoirradiated and PDT-induced cell death for various incubation periods was determined by MTT assay and fluorescence Live/Dead assay. Our results indicate that active EGFR-targeting of the Pc 4-loaded micelles accelerates intracellular uptake of the drug. Consequently this enhances the PDT-induced cytotoxicity within shorter time periods.

Background

Photodynamic therapy (PDT) is a novel, minimally invasive cancer treatment modality, where photoactivation of photosensitizer (PS) molecules accumulated within a tumor leads to energy transfer cascades, ultimately resulting in the conversion of molecular oxygen to cytotoxic reactive oxygen species.¹ In recent years several PS formulations have been approved by health organizations in the US, Europe, and elsewhere for PDT of various pre-cancerous and cancerous lesions.² The efficacy of PDT is dependent upon the optimization of the ‘drug-light-oxygen triad’ and harnessing the cytotoxic effects selectively within the cancer tissue, while avoiding phototoxicity and photosensitivity side effects in healthy tissues. Optimization of the ‘drug’ component of the triad involves enhancing the selective accumulation of the drug in the tumor compared to the healthy tissues. Furthermore, it is beneficial if the drug itself has high molar absorptivity and photoactivation at tissue-penetrating deep red to near infrared wavelengths, so that light delivery to the drug becomes effective in the context of treating non-superficial tumors. Based on this rationale, we are investigating the cancer cell-selective actively targeted delivery of a novel photosensitizer, the silicon phthalocyanine Pc 4. Pc 4 has several superior properties compared to the current clinically approved PS drugs.³ For example, compared to FDA-approved Photofrin® (porfimer sodium), Pc 4 can be synthesized as a single high purity compound. Also, compared to the photoactivation wavelength of PhotofrinR (λ_max = 630 nm) and the EU-approved FoscanR (λ_max = 652 nm), Pc 4 has high molar absorptivity at longer wavelengths (λ_max = 675 nm), which allows greater tissue penetration of light. Pc 4 also shows much reduced cutaneous photosensitization and inflammatory effects.^3–7 Hence, cancer-selective delivery of Pc 4 can result in enhanced PDT efficacy compared to current clinical photosensitizers.

Delivery of photosensitizers suffers from the same limitations as that of cancer chemotherapeutic agents, i.e., the direct parenteral administration via intravenous injection results in a variable biodistribution. Such unpredictable, non-specific biodistribution of the PS results in significant drug loss, sub-optimal drug concentration at the target tumor and risks of lingering photosensitivity in healthy tissues (e.g., eyes and skin). Many PS molecules, including Pc 4, are highly hydrophobic and hence in current pre-clinical studies they are formulated using surfactants like Cremophor and Tween-80 as delivery excipients. However, such excipients can have significant hypersensitivity and toxicity issues, especially if multiple doses become necessary.^8–10 Similar issues with formulation and delivery of cancer chemotherapy drugs like doxorubicin and paclitaxel have been significantly resolved by the use of nanoformulation strategies, where the drug is packaged within biocompatible nanoparticle constructs (e.g., liposomal Doxil formulation for doxorubicin). These strategies prevent the drug from rapid renal clearance or non-specific accumulation in uninvolved tissues, protect the drug in plasma, and promote a high degree of drug accumulation within the target tumor via passive mechanisms of enhanced permeation and retention (EPR).^11–15 Following this rationale, we have previously demonstrated packaging of Pc 4 in biocompatible block-copolymer micelles for passive uptake and subsequent PDT of cancer cells in vitro.¹⁶

In vivo, such EPR-mediated passively accumulated nanoformulations within the tumor stroma can re-enter the blood stream via diffusion mechanisms over time resulting in decreased drug at the target tissue. Also, to be taken up within the cancer cells from the stromal space, such nanoformulations depend on time-resolved cell membrane-mediated processes.¹⁷ In order to avoid ‘diffusing out’ and to promote rapid intracellular uptake of the EPR-accumulated nanoformulations, one strategy is to utilize active targeting and binding of the nanoparticles to cancer cell-specific highly upregulated internalizing receptors.^{17, 18} In this mechanism, drug-loaded nanoparticles surface-modified by receptor-specific ligands or antibodies can bind the receptors, undergo cellular internalization via a receptor-mediated endosomal/lysosomal process, and subsequently the nanoparticle can undergo degradation/destabilization in the lysosomal compartment leading to intracellular release of the drug. The released drug can then bind to its target intracellular organelles and produce the desired therapeutic effects.

In order to investigate the utilization of this mechanism in the rapid intracellular delivery of Pc 4 and to analyze whether such delivery enhances the subsequent PDT effect, here we report on modifying our micelle-based Pc 4 nanoformulation with peptide ligands having specificity and affinity to epidermal growth factor receptor (EGFR). EGFR, a 170 kDa glycoprotein, is significantly upregulated on the surface of cancer cells, and the native ligand (e.g., EGF) binding to this receptor has been implicated in activation of cell signal pathways that inhibit apoptosis, promote cell proliferation, and increase the survival of the cancer.^19–21 Hence, EGFR has become a very important target for cancer immunotherapy and actively targeted cancer drug delivery. ^{19, 22}

To this end, we have surface-modified our micelle nanoformulations with a 12 aminoacid EGFR-targeting peptide, GE11.²³ The peptide has been reported to facilitate active EGFR targeting, receptor-mediated internalization and distribution of peptide-decorated liposomes in EGFR-overexpressing mouse xenografts.²² We have investigated modification of our Pc 4-loaded PEG-PCL micelles with multiple copies of the GE11 peptide and have studied their active targeting and uptake in vitro, on EGFR-overexpressing A431 human epidermoid carcinoma cell line. We have analyzed whether active targeting results in faster intracellular uptake of the micelles (hence Pc 4) compared to untargeted micelles and whether such intracellular uptake corresponds to enhanced PDT effects upon photoirradiation. Optimization of these strategies can lead to shorter drug-time intervals and site-selective PDT in EGFR-overexpressing cancers.

Materials and Methods

Materials

The EGFR-targeting GE11 peptide, YHWYGYTPQNVI, was custom synthesized by Abgent Inc. (San Diego, CA), with the addition of a cysteine residue on the N-terminus to facilitate thioether-mediated conjugation to the PEG block of the PEG-PCL copolymer that forms the micellar nanoformulation. Hence the final sequence was CYHWYGYTPQNVI. Hydroxyl-poly(ethylene glycol)-maleimide (Mal-PEG-OH, MW=3400 Da) was purchased from Laysan Bio Inc. (Arab, AL), and the chemical structure was confirmed by NMR. ε-Caprolactone, stannous octanoate (Sn(Oct)₂) and 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) were purchased from Sigma Aldrich (St. Louis, MO). A Live/Dead 3224 viability assay was obtained from Invitrogen (Carlsbad, CA). All solvents were obtained at HPLC grade from Sigma-Aldrich and used as purchased. Pc 4 was synthesized in the laboratory of Dr. Malcolm Kenney of the CWRU Department of Chemistry.

Synthesis of Mal-PEG-PCL

Synthesis of maleimide-functionalized poly(ethylene glycol-block-caprolactone) was carried out by reacting Mal-PEG-OH and ε-caprolactone in dry toluene at 130°C for three days with Sn(Oct)₂ catalyst. The resulting product was dissolved in dry tetrahydrofuran (THF) and precipitated into n-hexane followed by isolation of the polymer by vacuum filtration and thorough drying under vacuum. The polymer structure was confirmed by ¹H NMR in CDCl₃ (Figure 1).

Figure 1.

Chemical characterization of Mal-PEG–PCL (protons labeled in A) showing (B) ¹H NMR in CDCl₃

Synthesis of GE11-PEG-PCL Conjugate and Formation of Peptide-decorated Micelles

To synthesize GE11-PEG-PCL, a 2:1 molar ratio of Cys-GE11 peptide to Mal-PEG-PCL was reacted in aqueous buffer for 4 hr at room temperature, to allow reaction of the Mal group with the sulfhydryl group of the cysteine residue to form a thioether linkage (Figure 2). Unreacted reagents and peptide were removed by dialysis against Millipore water using 2000 MWCO dialysis tubing (Spectra/Pore, CA). Conjugation was confirmed by ¹H NMR and MALDI-TOF mass spectroscopy. For micelle formation, GE11-PEG-PCL and unmodified PEG-PCL were dissolved in acetone at a 1:5 weight ratio and added dropwise to a 10X volume of water while removing acetone from the solution under reduced pressure. Following this, the resulting water/polymer suspension was dialyzed against Millipore water using 2000 MWCO dialysis tubing. The exposure of the amphiphilic polymer systems to the aqueous environment with the gradual diffusion and removal of acetone resulted in thermodynamically driven self-assembly of the polymer molecules into micelles decorated with the GE11 peptide. This micelle suspension was then lyophilized, reconstituted in Millipore water and sonicated for 1 hr to produce a narrow size distribution of the micelles. The micelle size range was characterized by dynamic light scattering (DLS) and confirmed by electron microscopy. The size ranges were found to be statistically similar to those reported in our previous publication (~80–100 nm) and hence will not be included in the results section of the current report. Pc 4-loaded EGFR-targeted (GE11-decorated) and non-targeted (no peptide) micelles were made similarly, with the additional step of dissolving a fixed concentration of Pc 4 in acetone along with the polymers, followed by dropwise addition into a 10X volume of water, removal of acetone under reduced pressure, dialysis against Millipore water, lyophilization, aqueous reconstitution, and sonication for 1 hr. The 70% loading capacity of Pc 4 in such micelle nanoformulations has been reported previously.¹⁶

Figure 2.

Schematic for synthesizing Mal-PEG-PCL, Cys-modified GE11-peptide conjugation to Mal termini via thioether linkage and Pc4-loaded GE-11-decorated micellar nanoformulation developed therefrom.

Micelle Stability Studies

Three batches of micelle formulations loaded with 2 μM Pc 4 were lyophilized overnight. One lyophilized batch was exposed to ammonium acetate, pH 4.2 (to simulate the lysosomal acidic environment), one to ammonium acetate, pH 6.5 (to simulate tumor-associated extracellular environment) and the other to 50% serum (by volume) in PBS (to simulate the blood circulation environment), in vials. It is important to determine the stability of the micelles in these environments, because ideally the micelles should not undergo disassembly/destabilization in the circulation (serum environment) in order to avoid leakage and loss of drug prematurely before reaching the target tumor. On the other hand, the micelles should degrade and release Pc 4 from the lysosomal compartment, which can occur by acid-catalyzed degradation of the PEG-PCL ester linkages. During exposure to the serum and acid environments, the micelles were maintained in a sterile incubator with humidified atmosphere of 5% CO₂ and 95% air at 37°C. At various timepoints, aliquots were removed from the vials and absorbance spectra at 669 nm ( ε_{Pc 4} = 2.4 × 10⁵ M⁻¹cm⁻¹) were taken using a microplate reader to determine the concentration of Pc 4 released from the micelles into the serum or ammonium acetate.

Cell Culture

Human epidermoid carcinoma A431 cells (American Type Culture Collection) were seeded at 2 × 10⁵ cm⁻² in 35-mm diameter culture dishes and grown in Dulbecco's modified Eagle's medium (Fisher Scientific, Inc.) supplemented with 10% fetal bovine serum, penicillin (50 units/mL), and streptomycin (50 μg/mL). MCF-7c3 human breast cancer cells were seeded similarly and grown in RPMI 1640 medium (Fisher Scientific, Inc.) supplemented with 10% fetal bovine serum, 2 mM L-glutamine and 1% penicillin/streptomycin. MCF-7c3 cells were used as a control in our studies, since they do not overexpress EGFR.²⁴ All cultures were maintained in a humidified atmosphere of 5% CO₂ and 95% air in a 37°C incubator. In all experiments, 70–85% confluent cultures were used.

In Vitro Cell Binding Studies

After the cells reached 80% confluence, Pc 4-loaded GE11-decorated (targeted) micelles or unmodified (non-targeted) micelles were diluted in DMEM or RPMI 1640 and added 6-well culture dishes to bring the Pc 4 concentration of each dish to 400 nM. As an additional control Pc 4-loaded targeted micelles were also incubated with MCF-7c3 human breast cancer cell lines which do not overexpress EGFRs and hence are amenable to only membrane-mediated passive uptake mechanisms. After incubation for various times (10 min to 24 hr), the micelle-containing medium was removed, and the cells were washed carefully with PBS to remove free or loosely bound micelles. Fluorescence images of Pc 4 in the cells were captured using a Carl Zeiss Axio Observer D1 Inverted Epifluorescence Microscope fitted with a dichromatic filter cube (bandpass λ_ex= 630–640 nm; bandpass λ_em = 650–690 nm).

In Vitro PDT Studies on A431 Cells

In vitro efficacy of PDT with targeted vs. non-targeted Pc 4nanoformulation was evaluated using the MTT assay and further verified using a Live/Dead fluorescence assay. Cells were plated in 96-well flat-bottomed plates at 2 × 10⁴ cells per well, for testing PDT or for testing ‘dark toxicity’ (toxicity without photoirradiation) upon incubation with EGFR-targeted or non-targeted Pc 4 nanoformulations. Pc 4 was loaded in the micelles as described previously. Cells were incubated with either EGFR-targeted or nontargeted micelle-based Pc 4 nanoformulations at final Pc 4 concentrations of 25-400 nM. After various incubation periods (1–24 hr), some plates were exposed to Pc 4 photoactivating red light using a diode array (EFOS, Mississauga, ONT, Canada) at a fluence of 200 mJ/cm² (λ_max = 675 nm; width of output peak at half maximum, 24 nm) for 200 seconds at room temperature. Other plates were not photoirradiated and were shielded from ambient light. Plates were incubated at 37°C for 24 hr, and then with MTT for 4 hr at 37°C to allow conversion of MTT to purple formazan by active mitochondria of live cells. Plates were subsequently centrifuged, medium was removed, and 150 μL DMSO was added to lyse the cells and dissolve the formazan. The formazan was quantified by UV/Vis absorption at 540 nm on a microplate reader (Molecular Devices). The absorbance measurement for treated cells was normalized to that of the controls to report cell viability.

As a complimentary method of analysis, cells were stained with a Live/Dead cell viability assay. This assay uses two fluorescent probes, calcein AM and ethidium homodimer-1. Non-fluorescent calcein-AM is hydrolyzed intracellularly by viable cells into the green fluorescent calcein, while ethidium homodimer-1 is a red fluorescent nuclear stain that can only pass through the compromised membrane of dead cells. Fluorescence images were then taken using a Carl Zeiss Axio Observer D1 Inverted Microscope, and the relative green and red fluorescence of cells were used to assess the PDT-induced cell death qualitatively.

Data Analysis

Fluorescence quantification from the cell targeting/binding studies was carried out using raw image analysis in AxioVision software (Carl Zeiss). Where applicable, statistical analyses were performed using Minitab (Minitab Inc., State College, PA). ANOVA tests were used to analyze the data from MTT assays. They were used to compare four mean values at a time for different time points for the same type of micelle nanoformulation, as well as, to compare the EGFR-targeted vs. nontargeted formulations to determine statistical significance. ANOVA tests were also used to compare the means from different incubation times. Significance were reported with p<0.05.

Results

Block Copolymer Characterization

Figure 1A-B depicts the structure of the maleimide-functionalized PEG-PCL block copolymer (Mal-PEG-PCL) along with its ¹H NMR (in CDCl₃) spectrum with corresponding labeled peaks. The chemical shift representing the protons on the maleimide group is seen at δ=6.7 ppm and indicates that the maleimide remained intact during polymerization of Mal-PEG-PCL from Mal-PEG-OH. The theoretical number-averaged molecular weight, M_n, of the Mal-PEG-PCL was calculated from end group analysis using the integration values of the PCL protons at δ=4.1 ppm and the end group maleimide peak at δ=6.7 ppm. The M_n of the PCL block was calculated to be 4 kDa and the total M_n was 7.4 kDa.

Micelle Stability Studies

Over a period of seven days, the micelles remained stable when exposed to 50% serum (by volume) with negligible Pc 4 release into the serum as seen in Figure 3A. When exposed to a slightly acidic pH of 6.5, the micelles remained stable for approximately one day (Figure 3B). In contrast, when exposed to a highly acidic pH characteristic of lysosomes, Pc 4 was released within the first day (Figure 3C).

Figure 3.

Pc 4 release kinetics from micelles at (A) pH=7.4 with serum exposure, (B) pH=6.5 and (C) pH=4.2.

In Vitro Cell Binding Studies

Pc 4 fluorescence was used to compare the relative cellular uptake of Pc 4 when delivered by the EGFR-targeted (GE11-modified) micelles or the nontargeted unmodified micelles. Figure 4A-F shows representative fluorescence images of EGFR-overexpressing A-431 cells incubated with EGFR-targeted or non-targeted Pc 4 nanoformulations for 1, 5 and 24 hr. Figure 4G-L show corresponding representative images of EGFR-deficient MCF-7 cells. At the shortest incubation time studied (1 hr), it is apparent that the actively targeted micelles delivered more Pc 4 to the EGFR-expressing A431 cells (panel B) than did untargeted micelles (panel A), whereas neither formulation delivered a detectable amount of Pc 4 to MCF-7 cells (4G and 4H). These data suggest a role for receptor-mediated internalization of the actively targeted Pc 4 nanoformulation for early delivery of Pc 4 by micelles. The qualitative evaluation of the images is supported by the quantitative estimation of Pc 4 fluorescence in the cells per cell area (Fig. 4M), which demonstrates significantly enhanced uptake of Pc 4 at early times (10 min, 30 min and 1 hr) when EGFR-targeted micelles are presented to EGFR-expressing cells.

Figure 4.

Representative fluorescence images comparing the Pc 4 uptake at various time points when delivered via the EGFR-targeted versus nontargeted nanoformulation on A431 cells (A-F) and MCF-7c3 cells (G-L); the plot (M) shows quantitative Pc 4 fluorescence intensity per cell area for various incubation time periods in these cells for the targeted and non-targeted nanoformulations; all data are for incubation with nanoformulation containing 400 nM Pc 4.

After 5 hours of incubation, the representative fluorescence images (4C, 4D, 4I, 4J) indicate that all cells have taken up a certain extent of the Pc 4 nanoformulation. The quantitative data (4M) indicate that the intracellular uptake of the EGFR-targeted formulation on A-431 cells was statistically higher than untargeted formulation on A-431 cells or both formulations on MCF-7 cells. Such statistical differences between the test and control samples (and conditions) seem to go away at long incubation period (24 hr), as shown in the quantitative data in 4M. This data suggest that without active EGFR targeting, the intracellular uptake of the Pc 4-loaded micellar nanoformulation possibly occurs via membrane-based processes (e.g. fusion, pinocytosis etc) which are well known to be time-dependent delayed processes.^{25, 26} This is the case for unmodified micelles on EGFR-expressing A-431 cells, and both GE11-modified and unmodified micelles on EGFR-deficient MCF-7 cells. In contrast, with active EGFR-targeting, Pc 4-loaded GE11-modified micellar nanoformulations are rapidly internalized possibly via receptor-mediated endocytosis. Similar mechanisms have been recently reported for a number of ligand-decorated drug-loaded block copolymer micelle formulations targeted to relevant cancer biomarkers like EGFR, folate receptor and integrin α_Vβ₃.^27–29 This is the case for enhanced Pc 4 fluorescence in A-431 cells incubated with these targeted formulations at shorter incubation periods (10 min-5 hr). At 24 hr, due to the long incubation period, both the membrane-mediated passive uptake (unmodified micelles on A-431 cells and both types of micelles on MCF-7 cells) and the receptor-mediated active uptake (GE11-modified micelles on A-431 cells) end up becoming similar in extent and hence the statistical difference in intracellular Pc 4 fluorescence between the two processes become negligible. ANOVA analysis confirms the difference in the mean fluorescence intensity values for the different incubation times.

In Vitro PDT Studies on A431 Cells

The difference in cellular internalization extent of Pc 4-loaded EGFR-targeted vs untargeted micelle nanoformulations at shorter time periods (up to 5 hr) and similarity of such internalization at longer time periods (24 hr), also corresponded to their respective in vitro PDT effects (extent of cancer cell death upon photoirradiation). The MCF-7 cells internalized the Pc 4 formulations in a time-dependent passive way irrespective of presence or absence of GE11 modification, since these cells do not normally overexpress EGFR. Hence their Pc 4 and subsequent PDT response upon photoirradiation were very similar to that reported in our previous publication¹⁶ and is not included in the current article. Here we show PDT results of A431 cells incubated with Pc 4-loaded EGFR-targeted (GE11-modified) versus untargeted micellar nanoformulation, for various Pc 4 concentrations (Figure 5), as well as, various incubation periods (Figure 6A). Figure 5 shows the cell viability data (analyzed by MTT assay) where the A-431 cells were incubated with Pc 4-loaded targeted versus untargeted micelles for 24 hrs at various Pc 4 concentrations, followed by photoirradiation at the activating wavelength of Pc 4 and subsequent cell death analysis. As expected from the equivalence of ‘intracellular uptake’ data between targeted vs untargeted systems at long (24 hr) incubation period, there was no statistical difference in cell viability between the targeted vs untargeted population at specific Pc 4 dose concentrations. Also, there was a decreasing trend in cell viability (increased cell death) upon PDT with increasing concentration of Pc 4 dose in the micellar nanoformulations.

Figure 5.

Dose response of A-431 cells to various concentration of Pc 4 when delivered via targeted vs. nontargeted nanoformulation and incubated for 24 hrs; irrespective of Pc 4 concentrations there was no statistical difference between cell killing with the two formulations at long (24 hr) incubation periods.

Figure 6.

A-431 cell response to PDT with 400 nM Pc 4 delivered by EGFR-targeted or nontargeted nanoformulations for various incubation periods; (A) shows quantitative cell viability data demonstrating that incubation with the targeted formulation results in statistically significant cell death within a short period of time (1 hr) possibly due to rapid receptor-mediated active intracellular uptake; at longer time periods (beyond 5 hr) the PDT effect for targeted and nontargeted Pc 4 formulations become comparable; (B) shows representative post-PDT Live/Dead images of A-431 cells incubated with EGFR- targeted and nontargeted Pc 4 nanoformulations for 1 hr and 24 hrs.

As shown in Figure 6, if the Pc 4 concentration was maintained constant, for example at 400 nM, there was a significant statistical difference in PDT response (cell viability) between EGFR-targeted vs nontargeted nanoformulations at shorter time period (1 hr) compared to longer time period (24 hr). This is in accordance with the fact that at shorter time periods, for the EGFR-targeted Pc 4 nanoformulation, a higher amount of intracellular uptake of Pc 4 occurs via receptor-mediated mechanism, compared to the time-dependent passive uptake of the nontargeted formulation. Consequently photoactivation of the cell-internalized Pc 4 causes higher cell death (PDT effect) in the population incubated with the GE11-modified formulation. At longer time periods the uptake of targeted and nontargeted formulations become similar and therefore, though the total amount of cell death is more at 24 hr compared to 1 hr incubation, there is no statistical difference between the targeted and nontargeted formulations at 24 hr.

The quantitative MTT-based cell viability data is complemented by the representative Live/Dead fluorescence images shown in Figure 6B1-B4, where photoirradiation of cells at 1 hr incubation points followed by Live/Dead staining showed higher extent of red fluorescence (from dead cells) for the cells exposed to EGFR-targeted Pc 4 nanoformulation compared to the cells exposed to the nontargeted formulation. At 24 hr incubation and photoirradaiation, cells exposed to either targeted or nontargeted Pc 4 nanoformulations showed comparable red fluorescence, indicating similar extent of cell death.

Discussion

PDT efficacy can benefit significantly from ‘nanomedicine’ approaches to enhance tumor site-selective (and cell-specific) delivery and release of photosensitizers. This benefit can happen from a combination of both passive delivery (EPR-mediated) and active targeting (receptor-mediated) mechanisms to allow for favorable tumor bioavailability and therapeutic index of the drug, reduced drug dosage parameters, and potentially reduced ‘waiting period’ between photosensitizer administration and tumor photoirradiation (the drug-light interval). For hydrophobic photosensitizers administered intravenously using Cremophor/Ethanol or Propylene glycol based excipient formulations, there is significant partitioning of the PS into the various lipoproteins in blood and other uninvolved tissues. This is the reason behind their non-specific accumulation into different tissues (e.g. liver, skin etc), variability in biodistribution and tumor bioavailability of the PS, and potential risk of phototoxicity and photosensitivity of uninvolved tissues. ^{30, 31} Because of this, although PDT has significantly less systemic toxicity than chemotherapy or ionizing radiation therapy, its efficacy and clinical repertoire in the treatment of non-superficial cancers have been limited. It is believed that for such excipient-based formulations, the PS accumulates into the tumor possibly via partitioning into blood lipoproteins (e.g. LDL) and subsequent interaction with upregulated LDL receptors on the tumor cells.^{32, 33} With time, the PS is expected to gradually clear from the uninvolved tissues via the lymphatic drainage, while it stays retained within the tumor tissue due to compromised lymphatic drainage. Only then can the tumor tissue be irradiated and a PDT response selective to the tumor site can be expected. This mechanism has lead to the conventional regimen of so-called ‘drug-light interval’, which can vary unpredictably between 1–4 days for various current clinical formulations.^{34, 35} Also, in some aggressive type of tumors, for example head-and-neck cancers, repeated locoregional therapy becomes necessary after primary treatment.^36–38 In such cases, compared to chemotherapy, radiotherapy or surgery, PDT can become a much safer, functionally preserving and cosmetically forgiving treatment modality. But repeated use of Cremophor type of formulations may result in serious issues of toxicity and hypersensitivity, similar to that reported for Taxol (a Cremophor formulation of paclitaxel).^{39, 40} Formulating photosensitizers within nanoscale vehicles can resolve many of the above issues by protecting the PS within the vehicle (i.e. preventing leakage and partitioning into blood proteins), preventing non-specific uptake into healthy tissue while allowing enhanced passive uptake within the tumor tissue (i.e. EPR mechanism), and further providing a way to enhance cancer-specific intracellular uptake via ligand-based receptor-mediated uptake mechanisms. Furthermore, choosing a biocompatible material for the nanovehicle can allow safe multiple doses of the formulation if needed. Because of such potential advantages of nanomedicine approaches in PDT, in recent years there have been multiple research activities in formulating different PS drugs in a variety of nanoscale vehicles (e.g. liposomes, dendrimers, lipidic and polymeric micelles, nanospheres, gold nanoparticles, silica nanoparticles, quantum dots etc), and enhancing cell-selective delivery of the formulations by using tumor-targeted ligands and antibodies to modify the nanovehicle surface.^41–44

Of the various nanovehicles mentioned above, we rationalize that block polymeric micelles provide some unique advantages regarding Pc 4 nanoformulation for cancer-targeted PDT. This micellar nano-assembly system enables enhanced control of the chemical tailoring, block length and stoichiometry of various functional components individually and independently, before assembling them at controlled compositions in the final drug-loaded nanoconstruct. Such an approach makes the constructs amenable for multifunctional modifications (e.g. with ligands, drugs, imaging agents etc) as and when desired. Also, the scale up of processes becomes potentially convenient for translation. In addition, for formulation of hydrophobic PS like Pc 4, micelles provide high loading capacity, guided by London dispersive forces between the hydrophobic drugs and the hydrophobic block of the polymer and also by the Flory-Huggins interaction parameter between the polymer and the drug.^{45, 46} Such thermodynamic ‘physical’ mechanisms of high drug loading enable avoiding of extra ‘chemical’ conjugation/modification steps that are otherwise needed in gold, quantum dot or silica based constructs in order to prevent drug dissociation and leakage. Compared to dendrimers, micelles are significantly easier to fabricate in the 20–100 nm diameter range which is an optimum size range to avoid rapid renal clearance or leakage through normal vasculature, yet allow EPR-mediated accumulation through tumor-associated leaky vasculature. For dendrimers, such size ranges are possible only in higher generations which essentially involve repeated chemical synthesis and amplification (convergent or divergent) steps, thereby increasing the process complexity and cost. Compared to polymer micelles, polymer nanospheres and nanoshells require enhnaced management of stabilizer and surfactant additives along with the polymer itself, in order to maintain stability in an aqueous environment. Compared to liposomes of the same diameter, block copolymer micelles allow higher hydrophobic volume for drug loading (i.e. liposome membrane vs micelle core) and hence have higher loading capacity. Also, polymer-based micellar assemblies have low critical micelle concentrations (CMC), which renders them resistant to dilution effects and destabilization in blood circulation, thereby ensuring prevention of encapsulant leakage. This is particularly important for preventing drug leakage and premature loss in circulation before reaching the tumor. Beyond targeted delivery, micelles also provide refined ways of site-selective ‘drug release’. For example, utilizing pH-sensitive components (e.g. the ester link in the PEG-PCL block copolymer in our research), the micelles can be induced to undergo destabilization/disassembly in a highly acidic environment, allowing stimuli (pH)-responsive drug release. Our results suggest that our Pc 4 nanoformulation can stay stable in circulation (dilution resistant and serum stable), can potentially undergo passive accumulation and receptor-mediated rapid active intracellular uptake selectively within EGFR-overexpressing tumors, and then can enable intracellular release of encapsulated Pc 4 for binding to target organelles (e.g. mitochondria) for PDT. Our research approach is also validated by several recent reports emphasizing the advantage and efficacy of micelles for in vivo tumor-targeted delivery of hydrophobic PS, for improved PDT.^{41, 47–49}

It is also important to note that following intracellular uptake, whether a PS needs to be released from its formulation vehicle to bind specific organelles prior to photoirradiation, or whether it can stay within the formulation as long as molecular oxygen can interact with it following photoirradiation and reactive oxygen species formed can diffuse within cells to cause oxidative damage, is an open question. Our micelle-based Pc 4 nanoformulation can allow destabilization/disassembly of the vehicles in the lysosomal compartment to further allow Pc 4 release and may enhance mitochondrial membrane binding of the Pc 4 for enhanced PDT. At the same time, Ohulchanskyy et al have recently reported on successful cancer PDT with Pc 4 covalently bound within silica nanoparticles and Cheng et al have recently reported similarly with Pc 4 bound on the surface of gold nanoparticles, both of which demonstrate tumor-specific ‘drug delivery’ but do not have any information on actual intracellular ‘drug release’.⁵⁰ Hence the need for ‘intracellular release’ of PS like Pc 4 to correlate to PDT efficacy, remains a topic of further mechanistic research.

Utilizing a ligand-decorated cancer-targeted nanoformulation strategy, we have demonstrated the feasibility of using EGFR specific peptides for surface modification of PEG-PCL micelles for encapsulating the photosensitizer Pc 4 for targeted delivery, rapid intracellular uptake and enhanced PDT of EGFR-overexpressing cancer cells. Such strategies can provide efficient ways of enhancing targeted PDT while reducing the drug-light interval, as well as, the drug dosage parameters. Compared to the nontargeted formulations, the EGFR-targeted Pc 4 formulations resulted in statistically higher intracellular uptake and corresponding PDT response (cell death) upon photoirradiation within shorter periods of time (10 minutes-5 hr). These results provide the rationale that in vivo these nanoformulations may allow targeted Pc 4 delivery, tumor-selective accumulation, rapid intracellular uptake and enhanced Pc 4-PDT in EGFR-overexpressing tumors and may shorten the drug-light interval. Investigation of such in vivo potential will be part of future research and reports.

Supplementary Material

01. Figure 1.

Mass spectroscopy (MALDI-TOF) data for (A) GE11 peptide and (B) GE11-PEG-PCL confirms peptide conjugation to polymer.