Nanoengineered photosensitizers for photodynamic priming to overcome P‐glycoprotein‐mediated multidrug resistance

Abstract

P‐glycoprotein (P‐gp, ABCB1)‐mediated multidrug resistance (MDR) remains a significant barrier to successful chemotherapy outcomes for cancer patients. While photoactivation of verteporfin (VP), a photosensitizer, has demonstrated success for overcoming MDR through direct protein aggregation upon photoactivation and through adenosine triphosphate (ATP) depletion, the impact of VP's formulation on P‐gp function and cellular energetics has not been fully characterized in this context. In this study, we screened four well‐established VP formulations—liposomal VP (L‐VP), lysophosphatidylcholine‐conjugated VP (lysoPC VP), liposomal formulation of lysoPC VP (L‐lysoPC VP), and a self‐assembled VP nanoaggregate (NanoVP), with a free form of VP as a control—for their ability to inhibit P‐gp. Using a combination of in vitro intracellular VP accumulation assays, P‐gp substrate retention experiments, and Seahorse‐based metabolic profiling, we identified NanoVP as the lead formulation for P‐gp modulation in cancer cells. NanoVP effectively depleted ATP in drug‐resistant cancer cells, while being recognized as a P‐gp substrate. Photodynamic priming with NanoVP at sub‐cytotoxic light doses enhanced P‐gp substrate retention within the cells without damaging P‐gp protein, indicating ATP depletion as the primary mode of functional inhibition. These findings highlighted NanoVP's clinical potential to enhance chemotherapeutic efficacy via photoactivation‐based modulation of P‐gp's function in multidrug‐resistant cancers.

Verteporfin's subcellular localization governs its ability to deplete ATP. LysoPC‐conjugated VP accumulates in lysosomes and spares mitochondria, while liposomal and NanoVP formulations target mitochondria, where light activation triggers membrane depolarization and ATP depletion, overcoming resistance seen with free‐form VP.

Abbreviations

ATP

adenosine triphosphate

BPD

benzoporphyrin derivative

BSA

bovine serum albumin

DMEM

Dulbecco's Modified Eagle Medium

FBS

fetal bovine serum

IMDM

Iscove's Modified Dulbecco's Medium

L‐lysoPC VP

liposomal formulation of lysoPC VP

L‐VP

liposomal‐VP

lysoPC

lysophosphatidylcholine

MDR

multidrug resistance

OXPHOS

oxidative phosphorylation

PDP

photodynamic priming

PDT

photodynamic therapy

P‐gp

P‐glycoprotein

TMRE

tetramethylrhodamine, ethyl ester

INTRODUCTION

Chemotherapy remains the cornerstone of cancer treatment; however, chemoresistance is responsible for nearly 90% of treatment failures in the clinic. ¹ A major contributor to drug resistance is the overexpression of ATP‐binding cassette (ABC) transporters, particularly P‐glycoprotein (P‐gp, ABCB1), which functions as an efflux pump that actively removes chemotherapeutic agents from cancer cells, reducing intracellular drug concentration and efficacy. As a result, patients with P‐gp‐overexpressing tumors often exhibit poor responses to chemotherapy and an increased likelihood of relapse. ² , ³ , ⁴ , ⁵

Approaches to overcome P‐gp‐mediated MDR have been characterized into three modes: engage, evade, and/or exploit. ⁶ , ⁷ Engaging P‐gp causes inhibition of drug efflux at the plasma membrane level, where the inhibitor interacts and binds to P‐gp. To evade P‐gp, agents that are not substrates of P‐gp would be employed, or they would be packaged in delivery systems that are not recognized by P‐gp. Exploitation takes advantage of the collateral sensitivity, where increased resistance to P‐gp leads to increased susceptibility to another drug or class of therapeutics. ⁶ Traditional methods for engaging P‐gp to overcome MDR include the co‐administration of small‐molecule inhibitors with chemotherapeutic agents. ⁸ , ⁹ , ¹⁰ , ¹¹ These inhibitors work to competitively bind to prevent substrate efflux, or cause inhibition of ATPase activity by blocking P‐gp conformation in either inward or outward facing, such as tariquidar. ¹² , ¹³ , ¹⁴ The major limitations of these inhibitors include low selectivity for diseased tissues, resulting in high toxicity to healthy tissues at the required dose. ¹⁵

Nanotechnology has shown promise in overcoming detection by P‐gp so that the drugs can evade recognition and exhibit their therapeutic potential. ¹⁶ In one such example, paclitaxel, a potent P‐gp substrate, was conjugated to gold nanoparticles, which resulted in a 17‐fold lower IC₅₀ compared to unconjugated paclitaxel when treating paclitaxel‐resistant non‐small cell lung cancer cells. ¹⁷ P‐gp expression was also observed to be decreased in these cells after the long‐term treatment with the nanoconjugate. ¹⁷ Further, the physical or chemical properties of nanoparticle formulations are being optimized and fine‐tuned such that they are not recognized by P‐gp. Polymer‐based nanomaterials and metal oxides for nanoparticles have also been shown to evade recognition by P‐gp, thus improving chemotherapy accumulation and increasing cytotoxicity in drug‐resistant cancer cells. ¹⁶ , ¹⁸ , ¹⁹ While the results from these approaches are promising, some of the major limitations for the implementation of nanotechnology‐based approaches for overcoming MDR include testing for safety, biocompatibility, and stability of the nanocarriers. ¹⁶

Exploitation of P‐gp includes targeting the upregulated pathways and processes that evolve due to P‐gp overexpression. One main approach for exploitation leverages the increased energy demand since P‐gp is an ATP‐dependent efflux transporter. Many studies have been aimed at targeting glucose uptake, glycolysis, and/or oxidative phosphorylation to indirectly decrease P‐gp function and enhance substrate accumulation. One study explored the use of 2‐deoxy‐D‐glucose (2‐DG), a glycolysis inhibitor, to enhance chemotherapy efficacy in idarubicin‐resistant, P‐gp‐overexpressing leukemia cells. The results showed that 2‐DG co‐treatment increased idarubicin‐induced cell death in vitro. ²⁰ The major limitation in this case was selective targeting toward cancer tissue, both chemosensitive and chemoresistant, which is a challenge for systemic administration of metabolic inhibitors. ²¹

To overcome these limitations, photochemical strategies have emerged as an alternative approach to inhibit drug efflux by P‐gp with a potential to engage, evade and/or exploit P‐gp. Photodynamic therapy (PDT) is a photochemistry‐based treatment modality that uses light activation of a photosensitizer at a specific wavelength to generate cytotoxic reactive oxygen species (ROS), inducing cell death. ²² , ²³ This approach offers dual selectivity, as treatment occurs only at the irradiated site and time of activation. ²² In the context of photoactivation, it has been previously shown that verteporfin (VP) is a substrate of P‐gp and, upon illumination, can cause aggregation of the protein at the plasma membrane. ²⁴ , ²⁵ To this effect, VP is a photosensitizer that can be used for photoactivation to engage with P‐gp. We previously demonstrated that photoactivation of VP with a photodynamic priming (PDP) light dose (subtherapeutic dose of PDT) effectively induces ATP depletion, leading to reduced P‐gp function and increased intracellular retention of P‐gp substrates without affecting surface expression levels. ²⁶ This highlights the potential for photoactivation to exploit the upregulated metabolic pathways resulting from P‐gp overexpression.

Given VP's dual ability to engage and exploit P‐gp through crosslinking of the protein itself and ATP depletion, an important question arises: can its nanoformulation be optimized to enhance intracellular retention while maintaining its ability to inhibit P‐gp function? Investigating different molecular and nanoengineered VP formulations may provide insights into how delivery strategies influence its therapeutic potential against MDR tumors. Clinically, VP is delivered with phospholipids, however, this formulation has not been investigated in the context of MDR. ²⁷ In this work, we aimed to screen clinically relevant nanoformulations of VP to determine the optimal formulation to engage and exploit P‐gp, and identify a formulation that, if able to evade P‐gp, can enhance inhibition after photoactivation. Here we screened four previously reported, well‐established molecular and nanoengineered formulations of VP: liposomal VP (L‐VP), ²⁸ lysoPC VP, ²⁴ , ²⁹ liposomal formulation of lysoPC VP (L‐lysoPC VP), ³⁰ , and carrier‐free nanoformulation of VP (NanoVP) ³¹ , ³² to evaluate their P‐gp substrate status, intracellular uptake levels, and their potential as ATP‐depletion agents, using free‐form VP as a control. Our results show that NanoVP is the most promising formulation, achieving both mitochondrial ATP depletion and P‐gp inhibition, making it the optimal candidate for modulation of P‐gp's function using PDP. By identifying the most effective VP formulation for targeting both drug efflux and cellular metabolism, this work aims to provide a clinically translatable approach for overcoming MDR and enhancing the efficacy of chemotherapy in drug‐resistant cancers.

MATERIALS AND METHODS

Chemicals

The fluorescent compounds calcein‐AM and rhodamine 123 were purchased from Invitrogen (Carlsbad, CA) and Sigma Aldrich (St. Louis, MO), respectively. The photosensitizer verteporfin (VP) was purchased from U.S. Pharmacopeia (Rockville, MD). All compounds were dissolved in dimethyl sulfoxide (DMSO), purchased from Sigma Aldrich (St. Louis, MO). All lipids used: 1‐palmitoyl‐2‐hydroxy‐sn‐glycero‐3‐phosphocoline (16:0 lyso PC), 1,2‐Dipalmitoyl‐sn‐glycero‐3‐phosphocholine (DPPC), 1,2‐dioleoyl‐3‐trimethylammonium‐propane (chloride salt) (DOTAP), 1,2‐distearoyl‐sn‐glycero‐3‐phosphoethanolamine‐N‐[methoxy(polyethyleneglycol)‐2000] (DSPE‐mPEG₂₀₀₀) were obtained from Avanti Polar Lipids. All other chemicals and reagents were purchased from Thermo Fisher Scientific (Waltham, MA) or Sigma‐Aldrich (St. Louis, MO).

Verteporfin (VP) formulations

Four formulations of VP were investigated: nanoliposomal VP (L‐VP), lysoPC VP, nanoliposomal lysoPC VP (L‐lysoPC VP), and NanoVP, and free‐form VP in DMSO was used as a control.

A previously used protocol was used to synthesize L‐VP. ²⁸ Dipalmitoyl‐phosphatidylcholine (DPPC), cholesterol, distearoyl‐phosphatidylethanolamine‐methoxy polyethylene glycol (DSPE‐mPEG2000), and dioleoyltrimethyl‐ammoniumpropane (DOTAP) were combined in chloroform at a 20:10:1:2.5 molar ratio. 200 nanomoles of VP were co‐dissolved with the lipid mixture at a 2.4% VP‐to‐lipid mole ratio. The chloroform was removed by rotary evaporation to form a thin lipid film, which was subsequently rehydrated with 1 mL of ultrapure water and subjected to freeze–thaw cycles between 4°C and 45°C. The lipid suspension was extruded through a polycarbonate membrane (0.1 μm pore size) at 45°C. The extruded liposomes were dialyzed against 1× phosphate‐buffered saline (PBS) at 4°C to remove unencapsulated material. Liposome size, polydispersity index (PDI) and zeta potential were measured on NanoBrook Omni (Brookhaven). UV–Vis spectroscopy was used to measure VP concentration as previously shown. ²⁸ , ³³

The (16:0) lysoPC‐VP conjugate was synthesized by esterifying the carboxylic acid group of VP with the hydroxyl functional group of (16:0) lysoPC, similar to previously published protocols. ²⁴ , ²⁹ Briefly, 1‐palmitoyl‐2‐hydroxy‐sn‐glycero‐3‐phosphocholine ((16:0)lysoPC), VP, 1‐ethyl‐3‐(3‐dimethylaminopropyl) carbodiimide (EDC), 4‐(dimethylamino) pyridine (DMAP), and N,N‐diisopropylethylamine (DIPEA) were combined in dichloromethane (DCM) at a fixed molar ratio of 1:5:50:25:60 and allowed to react at room temperature for 24 h. The crude product was purified using a preparatory thin‐layer chromatography silica plate and was run with a 10% methanol in DCM mobile phase. The purified conjugated product was extracted in a 2:1 DCM:methanol solution and was sonicated (60 s, 30 kHz, Branson bath sonicator). The sample was centrifuged, and the supernatant was removed by rotary evaporation. This process was repeated twice. The product was redissolved in pure DCM and the precipitated silica was filtered using a 0.22‐micron polytetrafluoroethane (PTFE) filter. The supernatant DCM was removed by rotary evaporation at 40°C. The final conjugated product was redissolved in chloroform and purity was tested by mass spectrometry (SCIEX x500B QTOF), and concentration was measured by UV–Vis spectroscopy. ²⁸ , ³³ LysoPC VP nanoliposomes (L‐lysoPC VP) were synthesized similarly to L‐VP, however, 200 nmol of pre‐conjugated lysoPC VP was synthesized as described above and was used in place of 200 nmol free‐form VP in DMSO. Liposome size, polydispersity index (PDI) and zeta potential were measured on NanoBrook Omni (Brookhaven). UV–Vis spectroscopy was used to measure VP concentration as previously shown. ²⁸ , ³³

The protocol from Quinlan et al. ³² was used to synthesize NanoVP (condition 4). VP was mixed with DMSO and was then added from 15 cm above the water level dropwise (10 μL) into rapidly stirred ultrapure water (1500 RPM) to form self‐assembled J‐type nanoaggregates. The nanoparticles were dialyzed against 1X PBS. Nanoparticle size, PDI and zeta potential were measured on NanoBrook Omni (Brookhaven). UV–Vis spectroscopy was used to determine VP concentration. ³¹ , ³²

Cell culture

The human triple‐negative breast cancer (TNBC) cell line MDA‐MB‐231 was obtained from the National Cancer Institute. ³⁴ The vinblastine‐resistant subline (VBL‐MDA‐MB‐231) was generated through continuous culture of MDA‐MB‐231 cells in 100 ng/mL vinblastine. ³⁵ MDA‐MB‐231 cells were maintained in Dulbecco's Modified Eagle Medium (DMEM) (Gibco), while VBL‐MDA‐MB‐231 cells were cultured in Roswell Park Memorial Institute (RPMI) 1640 Medium (Gibco). Both media were supplemented with 10% (v/v) fetal bovine serum (FBS), 100 U/mL penicillin, and 100 μg/mL streptomycin.

The human ovarian cancer cell line OVCAR8 was obtained from the Division of Cancer Treatment and Diagnosis Tumor Repository, National Cancer Institute (Frederick, MD). The doxorubicin‐resistant subline (NCI‐ADR/RES) was generated through continuous culture of OVCAR8 cells in 100 ng/mL doxorubicin (determined by us). Both cell lines were cultured in RPMI 1640 Medium, supplemented with 10% (v/v) FBS, 100 U/mL penicillin, and 100 μg/mL streptomycin. All cell lines were routinely tested and confirmed to be mycoplasma‐free and were maintained at 37°C in a humidified incubator with 5% CO₂.

Intracellular VP fluorescence using flow cytometry

Intracellular VP fluorescence was quantified by adapting previously published methods. ²⁴ , ³⁶ Briefly, 500,000 cells were seeded into 35 mm dishes and were allowed to incubate overnight for attachment at 5% CO₂ in 37°C. The next day, cells were incubated with 2 μM of each VP formulation for 1 h at 5% CO₂ at 37°C. The cells were washed twice with PBS containing calcium and magnesium and fresh medium was added to each dish. Cells were incubated for an additional 1 h at 5% CO₂ at 37°C to allow for VP efflux. Cells were washed once with PBS, then trypsinized and collected in Iscove's Modified Dulbecco's Medium (IMDM) supplemented with 5% (v/v) FBS. Cells were counted and 300,000 cells were aliquoted into 5 mL round‐bottom polystyrene tubes. Sample measurements were performed using a fluorescence‐activated cell sorting (FACS) flow cytometer (BD FACSCelesta, BD Biosciences). The fluorescence emission of VP (Ex/Em: 435/690 nm) was detected using a 405 nm laser and a PerCP‐Cy5‐5 filter (BV670 nm). Parental cell lines were used to set voltages to establish a control mean fluorescence intensity (MFI) between 2000 and 2500. Voltages between each formulation remained the same for measurements within the same cell pair but changed between formulations (e.g., a voltage was set for MDA‐MB‐231 cells treated with free‐form VP and the same voltage was used to measure the fluorescence of free‐form VP in VBL‐MDA‐MB‐231 cells, but this voltage was not the same for all formulations). All experimental conditions were conducted at least in triplicate (N ≥ 3). For each flow cytometry analysis, 10,000 events were recorded. Data analysis was performed using FlowJo V10 software (Tree Star, Inc., Ashland, OR), focusing on gated single‐cell populations.

VP extraction

Intracellular VP accumulation was quantified by adapting a previously published method from Vig et al. ³⁶ Briefly, 500,000 cells were seeded into 35 mm dishes and were allowed to incubate overnight for attachment at 5% CO₂ in 37°C. The next day, cells were incubated with 2 μM of each VP formulation for 90 min at 5% CO₂ at 37°C. Cells were washed twice with PBS without ions, then lysed using RIPA buffer for 1 h (Thermo Fisher Scientific, USA). Fluorescence signals of the VP formulations were measured using a BioTek Synergy Neo2 plate reader at VP's excitation/emission wavelengths of 435/689 nm. Protein concentration (mg/mL) was determined using the Bicinchoninic Acid (BCA) Protein Assay (Thermo Fisher Scientific). Intracellular VP concentrations were quantified using standard curves and normalized to total protein content as determined by the BCA assay. All experiments were performed in at least duplicate with duplicate measurements on three separate days (N = 3).

Photodynamic priming (PDP), TMRE accumulation, and cell viability

Forty thousand cells were seeded into a monolayer and allowed to incubate overnight for attachment in 5% CO₂ at 37°C. The next day, PDP was performed using each VP formulation as a photosensitizer. Cells were incubated with VP formulations in media at 37°C for 90 min. The cells were then washed twice with PBS containing calcium and magnesium, and fresh medium was added to each well. PDP was performed at 690 nm and an irradiance of 10 mW/cm² and a radiant exposure of 1 J/cm² with a 3‐well by 3‐well illumination area using an ML8500 laser system (Modulight Corp., Tampere, Finland). Cells were then incubated in 5% CO₂ at 37°C for 2 h post‐illumination, followed by measuring mitochondrial membrane potential and cell viability.

Mitochondrial membrane potential depolarization was evaluated using the tetramethylrhodamine, ethyl ester (TMRE) assay (Abcam). ²⁶ Cells were incubated with TMRE (300 nM) and the P‐gp inhibitor tariquidar (2.5 μM) for 2 h. Following incubation, cells were washed twice with PBS containing 0.2% BSA, and fresh PBS + 0.2% BSA was added to each well. TMRE fluorescence was quantified using a BioTek Synergy Neo2 plate reader and visualized using the BioTek Lionheart Imager.

CellTiter Fluor assay, which measures live cell proteases as a measure of cell viability, was used to determine the photodynamic priming range of each VP formulation after illumination with 1 J/cm². The CellTiter Fluor assay buffer was prepared by adding 1 μL of glycyl‐phenylalanyl‐aminofluorocoumarin (GF‐AFC) substrate to 1 mL of assay buffer (example: 10 μL CF‐AFC substrate into 10 mL assay buffer). DMEM and the pre‐made assay buffer were added to each well at a 1:1 ratio and were incubated at 37°C for 2 h under constant shaking. The fluorescent product that is generated when cleaved by live cell protease was measured using a BioTek Synergy Neo2 plate reader with an excitation wavelength of 390 nm and an emission wavelength of 505 nm.

Seahorse assay

An Agilent Seahorse Mito Stress Test Assay kit was used to measure the changes in oxygen consumption rate (OCR) and ATP‐production coupled respiration after PDP via stimulation from electron transport chain (ETC) inhibitors oligomycin, carbonyl cyanide p‐(trifluoromethoxy)phenylhydrazone (FCCP) and rotenone/antimycin A. 40,000 cells were seeded into the Seahorse 96‐well microplate and were allowed to attach overnight in 5% CO₂ at 37°C. The Seahorse utility plate was hydrated overnight with Ultrapure water (Invitrogen) at 37°C in a 0% CO₂ environment, while the XF calibrant was also incubated under the same conditions. The following day, the XF calibrant was added to the utility plate and incubated at 37°C in 0% CO₂ until ready for inhibitor loading. Seahorse DMEM was supplemented with glutamate (2 mM), pyruvate (1 mM), and glucose (10 mM). VP (0–2 μM) was incubated with cells in Seahorse DMEM for 90 min at 37°C in 5% CO₂. After incubation, cells were washed twice with PBS containing calcium and magnesium, and fresh Seahorse DMEM was added to each well. The cells were then incubated for an additional 2 h at 37°C. Electron transport chain inhibitors, oligomycin (15 μM), carbonyl cyanide p‐(trifluoromethoxy) phenylhydrazone (FCCP) (20 μM), and rotenone/antimycin A (5 μM), were prepared and loaded onto the utility plate. After the two‐hour incubation, the incubation medium was replaced with fresh Seahorse DMEM in each well. The Seahorse analyzer was calibrated using the utility plate, after which the cell microplate was loaded into the instrument for measurement. To normalize the data to cell number, cells were fixed with 4% neutral buffered formalin, followed by nuclei staining with NucBlue (Thermo Fisher). Total well imaging was performed at 4X magnification using the BioTek Lionheart Imager, and cell counts were obtained using Gen5 software.

P‐gp Substrate Retention

Flow cytometry was used to measure the intracellular fluorescence of rhodamine 123 and calcein (the fluorescent product after the acetoxymethyl ester (AM) group is cleaved from calcein‐AM by intracellular esterases). ²⁶ Briefly, 500,000 cells were seeded into 35 mm dishes and were allowed to incubate overnight for attachment at 5% CO₂ in 37°C. The next day, cells were incubated with 2 μM of NanoVP for 90 min at 5% CO₂ at 37°C. Cells were washed twice with PBS, and fresh medium was added to the dish. Cells were illuminated at 690 nm, at 1 J/cm², and were then allowed to incubate for 2 h post‐illumination. Cells were washed once with PBS, then trypsinized and collected in IMDM supplemented with 5% (v/v) FBS. Cells were counted and 300,000 cells were aliquoted into 5 mL round‐bottom polystyrene tubes. To investigate P‐gp function, cells were incubated with P‐gp substrates: rhodamine 123 (1.3 μM) for 45 min or calcein‐AM (0.5 μM) for 10 min at 37°C. ³⁷ After being washed with cold IMDM, cells were resuspended in PBS containing 1% BSA. 1 μL of 0.2 mg/mL of DAPI (Sigma Aldrich) in DI water was added directly to each tube as a dead cell exclusion dye (final concentration 0.001 μg/μL). Flow cytometry analysis was conducted using a FACSCelesta (BD Biosciences) to assess substrate retention. ³⁸ Flow cytometry data were analyzed using FlowJo software and dead cells were gated out based on DAPI staining (Tree Star, Inc., Ashland, OR).

UIC2 binding

To determine the cell surface expression of P‐gp, flow cytometry was used to measure fluorescence from FITC‐labeled secondary antibody conjugated to UIC2. The cells were treated with NanoVP‐PDP and harvested and counted using the same steps as in the substrate accumulation assay. The monoclonal UIC2 antibody is conformation‐sensitive and to enhance its binding to P‐gp, cells were pretreated with 20 μM cyclosporine A (CsA) for 5 min at 37°C. UIC2 was then added directly into the tube for 30 min at 37°C. Cells were incubated with IgG2a isotype as a control for 1 h at 37°C. Cells were washed with cold IMDM by centrifuging at 1500 RPM for 5 min at 4°C. The cells were incubated with anti‐mouse FITC‐conjugated secondary antibody for 30 min at 37°C, washed again in IMDM, and resuspended in PBS containing 1% BSA. 1 μL of 0.2 mg/mL of DAPI in DI water was added directly to each tube. Flow cytometry was performed using a FACSCelesta (BD Biosciences) to measure P‐gp surface expression. ²⁶ Flow cytometry data were analyzed using FlowJo software and dead cells were gated out based on DAPI staining.

RESULTS

Nanoengineering alters the uptake of VP

Our previous work showed that using PDP with free‐form VP (2 μM, 1 J/cm²) could deplete ATP in drug‐resistant cells, without directly damaging P‐gp and increasing substrate accumulation. ²⁶ We aimed to evaluate the effect of nanoengineering on VP to determine an optimal formulation that enhances intracellular retention, while maintaining its ability to inhibit P‐gp function. We explored four VP formulations with free‐form VP as a control (Figure 1) and characterized each formulation (Table 1). Once prepared, a cell lysate extraction method was used to measure the levels of intracellular VP accumulation from each formulation in P‐gp‐overexpressing cells (VBL‐MDA‐MB‐231, NCI‐ADR/RES) and control cells (MDA‐MB‐231, OVCAR8). ³⁶

FIGURE 1.

Verteporfin (VP) formulations after molecular and nanoengineering. Free‐form VP (monomer) is shown solubilized in DMSO (A). L‐VP was synthesized by combining VP DPPC, cholesterol, DSPE‐PEG2000 and DOTAP in chloroform and then evaporating the chloroform via rotary evaporation. The film was then rehydrated in Ultrapure water, nanoliposomes were formed using a freeze–thaw method and nanoparticles were extruded through a 200‐micron membrane and were dialyzed against 1× PBS (B). LysoPC VP was synthesized by combining 16:0 lyso PC, VP, EDC, DMAP and DIPEA in dichloromethane, then evaporating the dichloromethane via rotary evaporation. The dried product was purified using a TLC silica plate and redissolved in chloroform (C). L‐lysoPC VP was synthesized the same way as L‐VP was synthesized, using lysoPC VP as previously described instead of free‐form VP (D). NanoVP was synthesized by adding VP in DMSO dropwise to stirred water to form a self‐assembled J‐type nanoaggregate and was dialyzed against 1× PBS (E).

TABLE 1.

Physical characterization of VP formulations. Size, polydispersity index and zeta potential were all measured using a particle sizer and zeta potential analyzer (NanoBrook Omni, Brookhaven Instruments).

Formulation	Molecular weight (g/mol)	LogP values	Size (nm) mean ± SD	Polydispersity index mean ± SD	Zeta potential (mV) mean ± SD
Free‐form VP	718.79	3.74 [40, 63]	–	–	–
L‐VP	–	–	223.40 ± 4.9	0.110 ± 0.02	15.14 ± 2.06
LysoPC VP	1194.8	–	–	–	–
L‐LysoPC VP	–	–	226.97 ± 1.0	0.123 ± 0.04	17.76 ± 0.79
NanoVP 32	–	–	119.41 ± 0.45	0.212 ± 0.03	−18 ± 8.0

In MDA‐MB‐231 cells, intracellular VP accumulation was measured to be 231.7 ± 72.7 pmol/mg protein compared to 50.4 ± 27.2 pmol/mg protein in VBL‐MDA‐MB‐231 cells (Figure 2A). In OVCAR8 cells, intracellular VP accumulation was measured to be 277.9 ± 117 pmol/mg protein and 50.1 ± 26.2 pmol/mg protein in NCI‐ADR/RES cells (Figure 2A). In MDA‐MB‐231 cells, intracellular VP accumulation from L‐VP was measured to be 175.1 ± 44.2 pmol/mg protein and in VBL‐MDA‐MB‐231 cells, intracellular VP accumulation was measured to be 69.3 ± 21.8 pmol/mg protein from L‐VP (Figure 2B). In OVCAR8 cells, intracellular VP accumulation from L‐VP was measured to be 212.1 ± 73 pmol/mg protein and in NCI‐ADR/RES cells, intracellular VP accumulation was measured to be 53.9 ± 20.5 pmol/mg protein from L‐VP (Figure 2B). In MDA‐MB‐231 cells, intracellular VP accumulation from lysoPC VP was measured to be 8.8 ± 4.1 pmol/mg protein and in VBL‐MDA‐MB‐231 cells, intracellular VP accumulation was measured to be 8.7 ± 4.7 pmol/mg protein from lysoPC VP (Figure 2C). In OVCAR8 cells, intracellular VP accumulation from lysoPC VP was measured to be 13.9 ± 8.9 pmol/mg protein and in NCI‐ADR/RES cells, intracellular VP accumulation was measured to be 8.1 ± 5.7 pmol/mg protein from lysoPC VP (Figure 2C). In MDA‐MB‐231 cells, intracellular VP accumulation from L‐lysoPC VP was measured to be 12.3 ± 4.3 pmol/mg protein and in VBL‐MDA‐MB‐231 cells, intracellular VP accumulation was measured to be 8.9 ± 5.4 pmol/mg protein from L‐lysoPC VP (Figure 2D). In OVCAR8 cells, intracellular VP accumulation from L‐lysoPC VP was measured to be 12.9 ± 4.9 pmol/mg protein and in NCI‐ADR/RES cells, intracellular VP accumulation was measured to be 16.8 ± 11.6 pmol/mg protein from L‐lysoPC VP (Figure 2D). In MDA‐MB‐231 cells, intracellular VP accumulation from NanoVP was measured to be 204.1 ± 74.3 pmol/mg protein and in VBL‐MDA‐MB‐231 cells, intracellular VP accumulation was measured to be 89.8 ± 31.6 pmol/mg protein from NanoVP (Figure 2E). In OVCAR8 cells, intracellular VP accumulation from NanoVP was measured to be 262.6 ± 101.7 pmol/mg protein and in NCI‐ADR/RES cells, intracellular VP accumulation was measured to be 74.4 ± 28.8 pmol/mg protein from NanoVP (Figure 2E). Thus, these results indicate that there are higher levels of intracellular VP accumulation in formulations that do not contain lysoPC.

FIGURE 2.

Molecular and nanoengineering of VP alter intracellular accumulation in both chemosensitive and chemoresistant TNBC and ovarian cancer cells. A photosensitizer extraction method was used to quantify the amount of VP (pmol) that was delivered intracellularly after a 90 minute incubation with the different formulations per mg of protein from cell lysates. Free‐form VP (A), L‐VP (B) and NanoVP (E), all showed similar levels of intracellular VP fluorescence. Intracellular fluorescence of VP from LysoPC VP (C) and L‐lysoPC VP (D) was approximately 20× lower in parental cells and 5× lower in drug‐resistant cells. Data presented as mean ± SD (n = 3), *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001, unpaired two‐tailed t‐test.

These results were validated using flow cytometry to measure intracellular fluorescence of VP after delivery from each formulation. VP efflux when delivered from L‐VP and NanoVP was similar to VP efflux from free‐form VP in both drug‐resistant TNBC (Figure S1) and drug‐resistant ovarian cancer (Figure S2). The VP retention values compared to the parental cell lines are recorded in Table S1. The average change in MFI (ΔMFI) between control cells and P‐gp‐overexpressing cells (Tables S2, S3), was calculated to determine the accumulation of each formulation, using free‐form VP as a control.

Interestingly, VP efflux from formulations containing lysoPC did not follow a trend between both cell pairs. In TNBC cells, VP from lysoPC VP (Figure S1E,F) and L‐lysoPC VP (Figure S1G,H) appeared to be significantly effluxed from VBL‐MDA‐MB‐231 cells. In ovarian cancer cells, the VP from lysoPC VP (Figure S2E,F) and L‐lysoPC VP (Figure S2G,H) was not significantly effluxed from NCI‐ADR/RES cells. The contradictory results prompted us to investigate this further. In the VP extraction experiments, all cells were treated with the same concentration of each formulation (2 μM), and since the levels of accumulation were approximately 20 times lower in the lysoPC VP and liposomal lysoPC VP treatments compared to others in the parental cell lines (Figure S3A), we measured the fluorescence of each VP formulation in RIPA buffer, the same buffer that the cell lysates were made in (Figure S3B) and found that all of the levels of fluorescence were approximately on the same order of magnitude. This proves that any differences in accumulation in the lysoPC‐conjugated formulations were due to decreased diffusion across the plasma membrane to deliver VP intracellularly, not due to any quenching effects. Because of this, in the flow cytometry experiments, the voltage for the lysoPC‐containing formulations had to be increased to approximately twice that of the free‐formVP control to establish a similar baseline fluorescence in all parental cells, while the voltages used to measure VP fluorescence from L‐VP and NanoVP formulations were approximately the same as the free‐formVP control (Figure S3C,D).

Accumulation of L‐VP and NanoVP was compared to free‐form VP since their levels of accumulation were the most similar to the free‐form VP control in TNBC cells (Figure S4A) and in ovarian cancer cells (Figure S4B). The results showed that VP accumulation from L‐VP is significantly lower than free‐form VP by 24.2% (p < 0.05), but NanoVP was similar in parental TNBC cells. VP from all formulations had no statistically significant differences in accumulation levels in parental ovarian cancer cells. However, VP accumulation from NanoVP was significantly higher than free‐form VP by 77.9% (p > 0.05), but L‐VP was similar in drug‐resistant TNBC cells. The same result was observed in ovarian cancer cells.

Free‐form VP decreases ATP‐production coupled respiration and mitochondrial membrane potential in a dose‐dependent manner in parental and drug‐resistant cells

We moved forward in investigating how L‐VP, lysoPC VP, and NanoVP interacted with key parameters of mitochondrial function after red light illumination at a photodynamic priming (PDP) dose using free‐form VP as a control. We used the Agilent Seahorse MitoStress Test assay to measure oxygen consumption rate (OCR), and ATP‐production coupled respiration. We used the Abcam TMRE assay to measure changes in mitochondrial membrane potential. Finally, we used the Cell Titer Fluor cell viability assay to confirm that any changes to mitochondrial function occurred within a PDP range (≤30% decrease in cell viability). All assays were performed 2 h after illumination at 690 nm, 10 mW/cm², 1 J/cm².

MDA‐MB‐231 cells treated with free‐form VP showed a 73% decrease in ATP‐production coupled respiration starting at 0.125 μM VP and decreased in a dose‐dependent manner (Figure 3A). Significant decreases in TMRE accumulation were observed starting at 0.5 μM VP with a 24.3% decrease in accumulation and continued to a 70.8% decrease at 2 μM VP after illumination (Figure 3B). Based on the Cell Titer Fluor assay, the concentration of VP where there is ≤30% decrease in cell viability at 1 J/cm² is 0.77 ± 0.07 μM VP (Figure 3C). This indicates that the results of decreases in ATP‐production coupled respiration and mitochondrial membrane potential depolarization are within the priming range in the parental cell line. In the drug‐resistant cell line, a 77% decrease in ATP‐production coupled respiration was observed at the lowest tested concentration of 0.125 μM VP and a decrease in a dose‐dependent manner, similar to the effect in the parental cell line (Figure 3D). A 46.5% decrease in TMRE accumulation was observed at 2 μM VP after illumination (Figure 3E). The PDP range in the drug‐resistant cells is below 1.26 ± 0.12 μM VP (Figure 3F). The decrease in ATP‐production coupled respiration was within the priming range; however, the decrease in mitochondrial membrane potential depolarization was only observed at 2 μM which was outside of the priming range.

FIGURE 3.

Free‐form VP‐PDP decreases ATP‐production coupled respiration and mitochondrial membrane potential in a dose‐dependent manner in parental and drug‐resistant TNBC cells. Cells were treated with free‐form VP (0–2 μM in DMSO) for 90 minute prior to the PDP treatment. Cells were irradiated from the bottom up at 1 J/cm² and then allowed to incubate for 2 h. The Agilent Seahorse Mito Stress Test Assay kit was used to measure changes in mitochondrial oxygen consumption rate and ATP‐production coupled respiration after irradiation of MDA‐MB‐231 cells (A) and VBL‐MDA‐MB‐231 cells (D). A TMRE assay was used to measure the changes to mitochondrial membrane potential after irradiation of MDA‐MB‐231 cells (B) and VBL‐MDA‐MB‐231 cells (E). CellTiter Fluor was used to measure the cell viability (after irradiation of MDA‐MB‐231 cells (C) and VBL‐MDA‐MB‐231 cells (F). Data presented as mean ± SD values (n = 3), *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001, two‐tailed t‐test.

Similar results were observed in the ovarian cancer cell pair. OVCAR8 cells treated with free‐form VP‐PDP showed a 79.7% decrease in ATP‐production coupled respiration at 0.125 μM VP (Figure 4A). A 9.6% decrease in TMRE accumulation was observed in OVCAR8 cells at 0.125 μM VP treatment after illumination and continued to decrease in a dose‐dependent manner to a 60.9% decrease in accumulation after illumination at 2 μM VP (Figure 4B). The PDP range was measured to be below 0.62 ± 0.07 μM VP where a 30% decrease in cell viability was observed after illumination (Figure 4C). The decreases in ATP‐production coupled respiration and mitochondrial membrane potential depolarization occur within the priming range for the parental cell line. NCI‐ADR/RES cells treated with VP‐PDP showed a 69.9% decrease in ATP‐production coupled respiration at 0.125 μM VP and continued to decrease in a dose‐dependent manner (Figure 4D). A 9.8% decrease in TMRE accumulation was observed in NCI‐ADR/RES cells at 0.25 μM VP post‐illumination and continued to decrease in a dose‐dependent manner to a 67.9% decrease in accumulation after illumination at 2 μM VP (Figure 4E). The PDP range was measured to be below 2.04 ± 0.10 μM VP (Figure 4F). The decreases in ATP‐production coupled respiration and mitochondrial membrane potential depolarization occur within the priming range for the drug‐resistant ovarian cancer cell line. Table 2 summarizes the IC₃₀ and IC₅₀ values for the mitochondrial membrane potential depolarization and cell viability. ATP‐production coupled respiration IC₃₀ and IC₅₀ values could not be calculated due to non‐normalization of the data and unstable values.

FIGURE 4.

Free‐form VP‐PDP decreases ATP‐production coupled respiration and mitochondrial membrane potential in a dose‐dependent manner in parental and drug‐resistant ovarian cancer cells. Cells were treated with free‐form VP (0–2 μM in DMSO) for 90 min prior to the PDP treatment. Cells were irradiated from the bottom up at 1 J/cm² and then allowed to incubate for 2 h. The Agilent Seahorse Mito Stress Test Assay kit was used to measure changes in mitochondrial oxygen consumption rate and ATP‐production coupled respiration after irradiation of OVCAR8 cells (A) and NCI‐ADR/RES cells (D). A TMRE assay was used to measure the changes to mitochondrial membrane potential after irradiation of OVCAR8 cells (B) and NCI‐ADR/RES cells (E). CellTiter Fluor was used to measure the cell viability (after irradiation of OVCAR8 cells (C) and NCI‐ADR/RES cells (F). Data presented as mean ± SD values (n = 3), *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001, two‐tailed t‐test.

TABLE 2.

Summarized IC₃₀ and IC₅₀ values for TMRE accumulation and cell viability after free‐form VP‐PDP. Summarized IC₃₀ and IC₅₀ values for TMRE accumulation (mitochondrial membrane potential depolarization) and cell viability after a 90 min incubation with 0–2 μM free‐form VP followed by PDP at 1 J/cm² and measured 2 h post‐illumination.

	Free‐form VP (1 J/cm2)
	Mitochondrial membrane potential depolarization		Cell viability
	IC30 (μM)	IC50 (μM)	IC30 (μM)	IC50 (μM)
MDA‐MB‐231	0.59 ± 0.01	0.93 ± 0.06	0.77 ± 0.07	1.54 ± 0.07
VBL‐MDA‐MB‐231	1.59 ± 0.38	2.07 ± 0.20	1.26 ± 0.12	2.23 ± 0.09
OVCAR8	0.17 ± 0.08	0.49 ± 0.07	0.27 ± 0.07	1.84 ± 0.18
NCI‐ADR/RES	0.78 ± 0.07	1.29 ± 0.05	2.04 ± 0.10	3.00 ± 0.07

L‐VP decreases ATP‐production coupled respiration and mitochondrial membrane potential in a dose‐dependent manner in parental and drug‐resistant cells

MDA‐MB‐231 cells treated with L‐VP exhibited a 55.4% decrease in ATP‐production coupled respiration starting at 0.125 μM L‐VP and decreased in a dose‐dependent manner up to an 82.6% decrease at 2 μM L‐VP post‐illumination (Figure 5A). TMRE accumulation decreased significantly starting at 1 μM L‐VP to a 56.7% decrease in accumulation and decreased further to 74.6% at 2 μM L‐VP post‐illumination (Figure 5B). The PDP range is below 0.70 ± 0.08 μM L‐VP in the control cells (Figure 5C). Drug‐resistant cells also showed a 46.8% decrease in ATP‐production coupled respiration starting at 0.125 μM L‐VP and decreased in a dose‐dependent manner up to a 91.9% decrease at 2 μM VP post‐illumination (Figure 5D). Decreases in TMRE accumulation were not observed in drug‐resistant cells after illumination at 1 J/cm² at any tested concentration of L‐VP (Figure 5E). The PDP range in drug‐resistant cells is below 3.92 ± 0.44 L‐VP (Figure 5F). Only decreases in ATP‐production coupled respiration occur within the priming range in both cell lines, suggesting that VP levels from liposomal delivery were only high enough to disrupt the electron transport chain after light activation.

FIGURE 5.

L‐VP‐PDP decreases ATP‐production coupled respiration and mitochondrial membrane potential in a dose‐dependent manner in parental TNBC cells, but it only decreases ATP‐production coupled respiration in drug‐resistant TNBC cells. Cells were treated with L‐VP (0–2 μM in PBS) for 90 min prior to the PDP treatment. Cells were irradiated from the bottom up at 1 J/cm² and then allowed to incubate for 2 h. The Agilent Seahorse Mito Stress Test Assay kit was used to measure changes in mitochondrial oxygen consumption rate and ATP‐production coupled respiration after irradiation in MDA‐MB‐231 cells (A) and VBL‐MDA‐MB‐231 cells (D). A TMRE assay was used to measure the changes to mitochondrial membrane potential after irradiation in MDA‐MB‐231 cells (B) and VBL‐MDA‐MB‐231 cells (E). CellTiter Fluor was used to measure the cell viability (after irradiation in MDA‐MB‐231 cells (C) and VBL‐MDA‐MB‐231 cells (F). Data presented as mean ± SD values (n = 3), *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001, two‐tailed t‐test.

OVCAR8 cells treated with L‐VP were found to have a 50.8% decrease in ATP‐production coupled respiration starting at 0.125 μM VP and do not further decrease as concentration increases (Figure 6A). TMRE accumulation decreased significantly starting at 0.5 μM L‐VP to a 35.5% decrease in accumulation and decreased further to 57.9% 2 μM L‐VP post‐illumination (Figure 6B). The PDP range is below 0.92 ± 0.13 μM μM L‐VP in the control cells (Figure 6C). NCI‐ADR/RES cells treated with L‐VP were found to have a 60.8% decrease in ATP‐production coupled respiration starting at 0.125 μM VP and decreased in a dose‐dependent manner up to 78.3% post illumination (Figure 6D). TMRE accumulation decreased significantly at 2 μM L‐VP to 39.4% post‐illumination (Figure 6E). The PDP range is below 7.31 ± 0.96 μM L‐VP in the control cells (Figure 6F). In both cell lines, decreases in ATP‐production coupled respiration and TMRE accumulation occur within the priming range. Table 3 summarizes the IC₃₀ and IC₅₀ values for the mitochondrial membrane potential depolarization and cell viability. ATP‐production coupled respiration IC₃₀ and IC₅₀ values could not be calculated due to non‐normalization of the data and unstable values.

FIGURE 6.

L‐VP decreases ATP‐production coupled respiration and mitochondrial membrane potential in a dose‐dependent manner in parental and drug‐resistant ovarian cancer cells. Cells were treated with L‐VP (0–2 μM in PBS) for 90 min prior to the PDP treatment. Cells were irradiated from the bottom up at 1 J/cm² and then allowed to incubate for 2 h. The Agilent Seahorse Mito Stress Test Assay kit was used to measure changes in mitochondrial oxygen consumption rate and ATP‐production coupled respiration after irradiation in OVCAR8 cells (A) and NCI‐ADR/RES cells (D). A TMRE assay was used to measure the changes to mitochondrial membrane potential after irradiation in OVCAR8 cells (B) and NCI‐ADR/RES cells (E). CellTiter Fluor was used to measure the cell viability (after irradiation in OVCAR8 cells (C) and NCI‐ADR/RES cells (F). Data presented as mean ± SD values (n = 3), *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001, two‐tailed t‐test.

TABLE 3.

Summarized IC₃₀ and IC₅₀ values for TMRE accumulation and cell viability after L‐VP‐PDP. Summarized IC₃₀ and IC₅₀ values for TMRE accumulation (mitochondrial membrane potential depolarization) and cell viability after a 90 min incubation with 0–2 μM L‐VP followed by PDP at 1 J/cm² and measured 2 h post‐illumination.

	L‐VP (1 J/cm2)
	Mitochondrial membrane potential depolarization		Cell viability
	IC30 (μM)	IC50 (μM)	IC30 (μM)	IC50 (μM)
MDA‐MB‐231	0.88 ± 0.14	1.16 ± 0.07	0.70 ± 0.08	1.71 ± 0.14
VBL‐MDA‐MB‐231	–	–	3.92 ± 0.44	7.03 ± 0.80
OVCAR8	0.66 ± 0.21	1.10 ± 0.14	0.92 ± 0.13	3.65 ± 0.21
NCI‐ADR/RES	1.76 ± 0.11	2.22 ± 0.07	7.31 ± 0.96	14.8 ± 1.38

LysoPC VP does not decrease ATP‐production coupled respiration and/or mitochondrial membrane potential in parental and drug‐resistant cells at a given PDP dose

In the MDA‐MB‐231 cells, there were no significant changes observed in the ATP‐production coupled respiration, TMRE accumulation or cell viability after illumination at 1 J/cm² (Figure 7A–C). No changes were observed in VBL‐MDA‐MB‐231 cells for ATP‐production coupled respiration, TMRE accumulation or cell viability, when treated with lysoPC VP after illumination at 1 J/cm² (Figure 7D–F).

FIGURE 7.

LysoPC VP‐PDP does not decrease ATP‐production coupled respiration or and mitochondrial membrane potential in parental and drug‐resistant TNBC cells. Cells were treated with lysoPC VP (0–2 μM in DMSO) for 90 min prior to the PDP treatment. Cells were irradiated from the bottom up at 1 J/cm² and then allowed to incubate for 2 h. The Agilent Seahorse Mito Stress Test Assay kit was used to measure changes in mitochondrial oxygen consumption rate and ATP‐production coupled respiration after irradiation in MDA‐MB‐231 cells (A) and VBL‐MDA‐MB‐231 cells (D). A TMRE assay was used to measure the changes to mitochondrial membrane potential after irradiation in MDA‐MB‐231 cells (B) and VBL‐MDA‐MB‐231 cells (E). CellTiter Fluor was used to measure the cell viability (after irradiation in MDA‐MB‐231 cells (C) and VBL‐MDA‐MB‐231 cells (F). Data presented as mean ± SD values (n = 3), two‐tailed t‐test.

There were also no observed changes to ATP production coupled respiration, TMRE accumulation or cell viability in OVCAR8 cells (Figure 8A–C) and in NCI‐ADR/RES cells treated with lysoPC VP after illumination at 1 J/cm² (Figure 8D–F). These results are likely due to the decreased uptake observed in the lysoPC VP formulation.

FIGURE 8.

LysoPC VP‐PDP does not decrease ATP‐production coupled respiration or and mitochondrial membrane potential in parental and drug‐resistant ovarian cancer cells. Cells were treated with lysoPC VP (0–2 μM in DMSO) for 90 min prior to the PDP treatment. Cells were irradiated from the bottom up at 1 J/cm² and then allowed to incubate for 2 h. The Agilent Seahorse Mito Stress Test Assay kit was used to measure changes in mitochondrial oxygen consumption rate and ATP‐production coupled respiration after irradiation in OVCAR8 cells (A) and NCI‐ADR/RES cells (D). A TMRE assay was used to measure the changes to mitochondrial membrane potential after irradiation in OVCAR8 cells (B) and NCI‐ADR/RES cells (E). CellTiter Fluor was used to measure the cell viability (after irradiation in OVCAR8 cells (C) and NCI‐ADR/RES cells (F). Data presented as mean ± SD values (n = 3), two‐tailed t‐test.

NanoVP decreases ATP‐production coupled respiration and mitochondrial membrane potential in a dose‐dependent manner in parental and drug‐resistant cells

MDA‐MB‐231 cells treated with NanoVP showed a 68.9% decrease in ATP‐production coupled respiration starting at 0.125 μM L‐VP and decreased in a dose‐dependent manner to 82.2% at 2 μM (Figure 9A). TMRE accumulation decreased significantly starting at 0.5 μM NanoVP to 19.3% and decreased further to 75.4% at 2 μM NanoVP post‐illumination (Figure 9B). The PDP range is below 0.68 ± 0.09 μM NanoVP in the control cells (Figure 9C). VBL‐MDA‐MB‐231 cells showed a 74.3% decrease in ATP‐production coupled respiration starting at 0.125 μM NanoVP and decreased in a dose‐dependent manner to 95.2% at 2 μM post‐illumination (Figure 9D). A 25.2% decrease in TMRE accumulation was observed starting at 1 μM NanoVP in drug‐resistant cells after illumination at 1 J/cm² and continued to decrease to 48.5% at 2 μM post‐illumination (Figure 9E). Finally, the PDP range in drug‐resistant cells is below approximately 1.26 ± 0.11 μM NanoVP (Figure 9F). In both cell lines, decreases in ATP‐production coupled respiration and TMRE accumulation occurred within the priming range.

FIGURE 9.

NanoVP decreases ATP‐production coupled respiration and mitochondrial membrane potential in a dose‐ dependent manner in parental and drug‐resistant TNBC cells. Cells were treated with NanoVP (0–2 μM in PBS) for 90 min prior to the PDP treatment. Cells were irradiated from the bottom up at 1 J/cm² and then allowed to incubate for 2 h. The Agilent Seahorse Mito Stress Test Assay kit was used to measure changes in mitochondrial oxygen consumption rate and ATP‐production coupled respiration after irradiation in MDA‐MB‐231 cells (A) and VBL‐MDA‐MB‐231 cells (D). A TMRE assay was used to measure the changes to mitochondrial membrane potential after irradiation in MDA‐MB‐231 cells (B) and VBL‐MDA‐MB‐231 cells (E). CellTiter Fluor was used to measure the cell viability (after irradiation in MDA‐MB‐231 cells (C) and VBL‐MDA‐MB‐231 cells (F). Data presented as mean ± SD values (n = 3), *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001, two‐tailed t‐test.

OVCAR8 cells treated with NanoVP were found to have a 62.4% decrease in ATP‐production coupled respiration starting at 0.125 μM NanoVP and further decreases to 86.02% as concentration increases to 2 μM NanoVP (Figure 10A). TMRE accumulation decreased significantly starting at 0.25 μM NanoVP to a 31.4% decrease in accumulation and decreased further to 64.10% at 2 μM NanoVP post‐illumination (Figure 10B). The PDP range is below 1.09 ± 0.17 μM NanoVP in the control cells (Figure 10C). NCI‐ADR/RES cells treated with NanoVP were found to have a 73.9% decrease in ATP‐production coupled respiration starting at 0.125 μM NanoVP and a decrease in a dose‐dependent manner up to 97.9% at 2 μM NanoVP post illumination (Figure 10D). A 7.9% decrease in TMRE accumulation was observed starting at 0.5 μM NanoVP in drug‐resistant cells after illumination at 1 J/cm² and continued to decrease to 66.4% at 2 μM post‐illumination (Figure 10E). The PDP range is below 2.21 ± 0.16 μM NanoVP in the control cells (Figure 10F). In both cell lines, decreases in ATP‐production coupled respiration and TMRE accumulation occur within the priming range. Table 4 summarizes the IC₃₀ and IC₅₀ values for the mitochondrial membrane potential depolarization and cell viability. ATP‐production coupled respiration IC₃₀ and IC₅₀ values could not be calculated due to non‐normalization of the data and unstable values.

FIGURE 10.

NanoVP decreases ATP‐production coupled respiration and mitochondrial membrane potential in a dose‐ dependent manner in parental and drug‐resistant ovarian cancer cells. Cells were treated with NanoVP (0–2 μM in PBS) for 90 min prior to the PDP treatment. Cells were irradiated from the bottom up at 1 J/cm² and then allowed to incubate for 2 h. The Agilent Seahorse Mito Stress Test Assay kit was used to measure changes in mitochondrial oxygen consumption rate and ATP‐production coupled respiration after irradiation in OVCAR8 cells (A) and NCI‐ADR/RES cells (D). A TMRE assay was used to measure the changes to mitochondrial membrane potential after irradiation in OVCAR8 cells (B) and NCI‐ADR/RES cells (E). CellTiter Fluor was used to measure the cell viability (after irradiation in OVCAR8 cells (C) and NCI‐ADR/RES cells (F). Data presented as mean ± SD values (n = 3), *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001, two‐tailed t‐test.

TABLE 4.

Summarized IC₃₀ and IC₅₀ values for TMRE accumulation and cell viability after NanoVP‐PDP. Summarized IC₃₀ and IC₅₀ values for TMRE accumulation (mitochondrial membrane potential depolarization) and cell viability after a 90 min incubation with 0–2 μM NanoVP followed by PDP at 1 J/cm² and measured 2 h post‐illumination.

	NanoVP (1 J/cm2)
	Mitochondrial membrane potential depolarization		Cell Viability
	IC30 (μM)	IC50 (μM)	IC30 (μM)	IC50 (μM)
MDA‐MB‐231	0.67 ± 0.10	0.99 ± 0.06	0.68 ± 0.09	1.49 ± 0.12
VBL‐MDA‐MB‐231	1.37 ± 0.23	0.20 ± 0.17	1.26 ± 0.11	2.19 ± 0.08
OVCAR8	0.26 ± 0.08	0.56 ± 0.06	1.09 ± 0.17	2.36 ± 0.29
NCI‐ADR/RES	0.99 ± 0.08	1.44 ± 0.05	2.21 ± 0.16	3.55 ± 0.14

NanoVP‐PDP increases P‐gp substrate accumulation in TNBC and ovarian cancer cells

NanoVP was the formulation where we found a decrease in ATP production coupled with respiration and a depolarization of the mitochondrial membrane potential within the PDP range at 1 J/cm² that outperformed free‐form VP in DMSO. Based on this criterion, NanoVP was used to evaluate the effect of substrate accumulation using fluorescent P‐gp substrates rhodamine 123 and calcein‐AM to measure their intracellular fluorescence after NanoVP‐PDP (2 μM, 1 J/cm²). The results showed that there were no significant changes in the rhodamine 123 accumulation after NanoVP‐PDP in the MDA‐MB‐231 cells or OVCAR8 cells (Figure 11A). There were no observed changes in rhodamine 123 accumulation in VBL‐MDA‐MB‐231 cells; however, a significant increase in rhodamine 123 was observed in NCI‐ADR/RES cells (Figure 11A). Similarly, there were no changes in calcein accumulation after NanoVP‐PDP in MDA‐MB‐231 cells or OVCAR8 cells (Figure 11B). There were observed increases in calcein accumulation in both drug‐resistant cell lines; however, the increase was only significantly different in the VBL‐MDA‐MB‐231 cell line after NanoVP‐PDP (Figure 11B).

FIGURE 11.

NanoVP‐PDP calcein accumulation in drug‐resistant cells. Cells were treated with 2 μM NanoVP for 90 mins followed by photodynamic priming at 1 J/cm². Flow cytometry was used to measure the intracellular fluorescence of P‐gp substrates rhodamine 123 (A) and calcein (B) 2 h post‐illumination. Data presented as mean ± SD values (n = 3), one‐tailed t‐test, Mann–Whitney test, p values are presented on plots.

NanoVP‐PDP does not affect surface expression of P‐gp in TNBC and ovarian cancer cells

Finally, to ensure that any increases in P‐gp substrate accumulation were a result of ATP depletion, we measured changes to P‐gp surface expression after NanoVP‐PDP. CsA was used to enhance UIC2 binding to P‐gp. In MDA‐MB‐231 cells, the fold change remained approximately 1.0 both without and with CsA pre‐treatment (Figure 12A), which was expected because of a lack of P‐gp expression on the parental cell line. The same result was observed in VBL‐MDA‐MB‐231 cells both without and with CsA pre‐treatment (Figure 12B), indicating no changes to P‐gp surface expression. In OVCAR8 cells, the fold change also remained approximately 1.0 both without and with CsA pre‐treatment (Figure 12C), which was expected because of a lack of P‐gp expression on this cell line. The same result was observed in NCI‐ADR/RES cells both without and with CsA pre‐treatment (Figure 12D), indicating no changes to P‐gp surface expression.

FIGURE 12.

NanoVP‐PDP does not change P‐gp surface expression levels in drug‐resistant TNBC cells. Cells were treated with 2 μM NanoVP for 90 mins followed by photodynamic priming at 1 J/cm². Flow cytometry was used to FITC‐UIC2 fluorescence bound to surface P‐gp in MDA‐MB‐231 cells (A), VBL‐MDA‐MB‐231 (B), OVCAR8 (C), and NCI‐ADR/RES cells (D) with and without CsA 2 h post‐illumination. Data presented as mean ± SD values (n = 3), one‐tailed t‐test, Mann–Whitney test, p values are presented on plots.

DISCUSSION

Photoactivatable approaches have emerged as a promising strategy for decreasing P‐gp‐mediated drug efflux and overcoming MDR. ⁷ Verteporfin (VP) is a photosensitizer that is currently being investigated for photochemical treatment of cancer. VP is highly hydrophobic and has historically been delivered using lipid‐based carriers to improve solubility in aqueous environments. ³⁹ In fact, VP was previously FDA approved to treat age‐related macular degeneration in a lipid‐based formulation called Visudyne® ²⁷ , ⁴⁰ but is being repurposed and investigated clinically for phototherapy for cancers such as pancreatic cancer (NCT06381154). However, since VP is also a substrate for P‐gp, key questions remain regarding whether VP formulations can be optimized for clinical translation to enhance ATP depletion and sensitize tumors to chemotherapy.

A unique advantage of VP over other photosensitizers is its ability to crosslink P‐gp, resulting in aggregation upon illumination, rendering the transporter nonfunctional at the cell membrane level. ²⁵ Additionally, Kessel & Reiners found that VP interacts with mitochondria and when light‐activated, can depolarize the mitochondrial membrane potential. ⁴¹ Inspired by those findings, we demonstrated that PDP with verteporfin (VP‐PDP) effectively induces ATP depletion, leading to reduced P‐gp function and increased intracellular retention of P‐gp substrates. ²⁶ This dual effect of structural inactivation of P‐gp at the membrane and ATP depletion at the mitochondrial level, makes VP an ideal candidate for combating drug resistance. Thus, in this study, we sought to compare various molecular and nanoengineered VP formulations to identify those that optimize ATP depletion while maintaining sufficient intracellular retention. Specifically, we aimed to determine whether certain formulations could improve VP retention without altering its P‐gp substrate status, thereby ensuring sufficient VP accumulation in mitochondria for ATP depletion. By leveraging these multi‐level effects, our overall goal is to develop a clinically translatable approach for enhancing chemotherapy efficacy in drug‐resistant cancers.

Molecular engineering and nanotechnology have been widely investigated for VP delivery. In our study, we investigated four molecular and nanoengineered formulations of VP: liposomal VP (L‐VP), ²⁸ lysoPC VP, ²⁴ , ²⁹ liposomal lysoPC VP (L‐lysoPC VP), ³⁰ and NanoVP, ³¹ , ³² using free‐form VP as a control. We found that L‐VP and NanoVP are effluxed similarly from P‐gp compared to free‐form VP in drug‐resistant cells. The extraction assay also showed similar levels of intracellular uptake of L‐VP and NanoVP in parental and drug‐resistant cells compared to free‐form VP, however statistical analysis revealed that L‐VP was statistically significantly lower in MDA‐MB‐231 (parental) cells compared to free‐form VP and NanoVP. The lower accumulation of L‐VP compared to free‐form VP and NanoVP was also observed in OVCAR8 cells, however, it was not statistically significantly different. NanoVP was taken up at a similar level compared to free‐form VP in both parental cell lines. A similar observation was reported by Quinlan et al. ³¹ where liposomal VP accumulation was compared to free‐form VP and NanoVP accumulation in U87 glioblastoma cells. Previous studies have reported that liposomes hinder drug uptake by cancer cells compared to free drugs or other formulations, which has been seen with free paclitaxel and doxorubicin compared to their liposomal formulations. ⁴² , ⁴³ , ⁴⁴ NanoVP is speculated to readily disassociate into its monomeric molecular form in the presence of serum or cells, which allows for passive diffusion through the cell membrane to achieve intracellular uptake, whereas liposomes must be endocytosed. ³¹ , ⁴⁵

We also observed that intracellular VP accumulation from NanoVP was statistically significantly higher compared to free‐form VP and L‐VP in both drug‐resistant cell lines. Drug efflux from P‐gp has been theorized to follow either a “flippase model” (also known as the “hydrophobic vacuum cleaner model”) or a “pump model”. ⁴⁶ , ⁴⁷ , ⁴⁸ In the flippase model, the substrate first interacts with the lipid bilayer, then accesses P‐gp via the inner leaflet of the lipid bilayer and then interacts with the transmembrane domain, where the drug is “flipped” to the outer leaflet of the plasma membrane for efflux extracellularly. ⁴⁸ The pump model assumes that the drug binds to P‐gp at the substrate binding pocket in the transmembrane domain and is then effluxed with no interaction with the lipid bilayer. ⁴⁶ , ⁴⁸ We speculate that the free‐form VP drug is already in its monomeric form, which is highly hydrophobic, and could be interacting with the lipid bilayer to follow the “flippase” efflux model. VP accumulation from NanoVP is more complex because it was shown that when NanoVP is introduced into serum, there is a burst release effect where there is initial VP release from NanoVP into its unquenched monomeric form in biologically relevant media. The remaining nanoparticle decreases in size where it could be small enough to undergo passive diffusion across the plasma membrane for further intracellular delivery. ³¹ , ³² With this proposed transport mechanism, NanoVP efflux could follow the “pump” efflux model, where VP is directly effluxed out of the cell with little interaction with the lipid bilayer. If NanoVP follows the pump efflux model, this would cause a temporary increase in intracellular VP concentration because the drug must accumulate inside the cell before P‐gp can actively transport it outside of the cell. ⁴⁸ The loading capacity of NanoVP is approximately 706,000 ± 38,000 VP molecules per particle, so even after the burst release effect, VP influx is potentially higher than the rate of efflux, also resulting in temporarily higher intracellular concentrations after a 90‐min incubation. ³¹ This suggests that the nanoaggregate acts as a temporary shield for VP, to at least bypass initial recognition by P‐gp at the plasma membrane and allow for accumulation intracellularly before the efflux mechanism begins. This proposed “nanoparticle shielding” effect has been observed by Cavaco et al. ⁴⁹ where they encapsulated paclitaxel into solid lipid nanoparticles (SLN) and treated MDA‐MB‐436 cells. They observed a lower IC₅₀ in cells treated with the paclitaxel SLN (1.48 μg/mL) compared to free paclitaxel treatment (2.20 μg/mL), which they hypothesize is a result of the SLN's evasion of P‐gp efflux. They also observed a decrease in P‐gp expression in cells treated with the SLN's, which could result in overall better cell killing. ⁴⁹ Additionally, P‐gp has been observed to transport at least two molecules at a time, which further bolsters the notion that NanoVP must be degraded into its monomeric form to be effluxed by P‐gp, regardless of how many of those molecules are effluxed per cycle. ⁴⁶

LysoPC VP formulations showed different results compared to L‐VP, NanoVP and the free‐form control. The VP extraction assay revealed that the intracellular lysoPC VP levels were approximately 20‐times lower in parental cells and approximately 5‐times lower in drug‐resistant cells compared to free‐form VP. The flow cytometry assay showed that lysoPC‐containing VP formulations appeared to be less effluxed by P‐gp in VBL‐MDA‐MB‐231 cells and were not effluxed from NCI‐ARE/RES cells. We also measured the fluorescence of VP from lysoPC VP in RIPA buffer to determine if this effect was due to quenching of the fluorescence of VP. The results revealed that the fluorescence intensity of free‐form VP and VP from lysoPC VP were on the same order of magnitude (10⁵), indicating that quenching was not responsible for the significant reduction in fluorescence and was likely due to limited diffusion through the plasma membrane. We additionally investigated the liposomal formulation of lysoPC VP (L‐lysoPC VP), which has been investigated for photoactivation in previous studies. ²⁹ , ³⁰ The level of VP efflux when delivered via L‐lysoPC VP was significantly decreased in VBL‐MDA‐MB‐231 cells compared to the control, however the similar levels were observed in NCI‐ADR/RES cells compared to the control. The VP extraction assay revealed no difference between the lysoPC VP accumulation and L‐lysoPC VP accumulation, which were also not a result of quenching. These results are consistent with previous reports. LysoPC has been reported to have slow trans‐bilayer movement, with half times reported to exceed 40 h. ⁵⁰ Another study found that lysoPC has a strong preference toward the outer leaflet of the plasma membrane. ⁵¹ This could potentially be a reason why diffusion is limited through the plasma membrane when VP is conjugated to lysoPC, resulting in inefficient VP uptake. Previous studies also found that lysoPC VP was a weaker substrate for P‐gp and was thus not effluxed as well as free‐form VP. ²⁴ , ²⁵ Since the influx of VP from lysoPC‐containing formulations was so low, the efflux that we measured from the VBL‐MDA‐MB‐231 cells is likely unreliable. We had to increase the voltage on the flow cytometer to detect the signal in parental cells, also confirming the low uptake, showing that overall, that formulation is a poor one to use in cancer treatment in general at this time point. For these reasons, L‐lysoPC VP was not further investigated for PDP effects. A future direction could include extending the incubation time point to allow the formulation more time to diffuse through the plasma membrane.

These differences in intracellular accumulation appear to influence how each formulation behaves to induce mitochondrial dysfunction after illumination. Interestingly, L‐VP and NanoVP interacted similarly to decrease ATP‐production coupled respiration at 0.125 μM VP after photoactivation with 1 J/cm² compared to free‐form VP in both parental cells and drug‐resistant cells. This indicates that despite VP efflux from P‐gp, enough VP remained intracellularly to introduce minor perturbations to the electron transport chain enough to decrease ATP production in mitochondria. Changes to mitochondrial membrane potential were different between parental and drug‐resistant cells. Across both cell pairs, MMP depolarization occurred at lower concentrations of VP in the parental cell lines compared to the drug‐resistant cell lines, further supporting the role of P‐gp efflux in reducing effective intracellular VP levels, thus, higher VP concentrations were required to depolarize MMP with PDP in drug‐resistant cells. In both MDA‐MB‐231 cells, the lowest to highest VP delivery for MMP depolarization was VP from L‐VP, NanoVP, then free‐form VP. Interestingly, in OVCAR8 cells, the best VP delivery for MMP depolarization was VP from NanoVP and L‐VP and NanoVP both depolarized the MMP at the same concentration. Free‐form VP has the best level of MMP depolarization because VP that passively diffuses into the cell is readily shuttled to the mitochondria, where it preferentially localizes. ³¹ VP release from the nanoformulations may require more time to be released due to endocytosis of L‐VP or further degradation into monomeric VP from NanoVP. ³² In VBL‐MDA‐MB‐231 cells, the lowest to highest VP delivery for MMP depolarization was VP from L‐VP, free‐form VP, then NanoVP. In NCI‐ADR/RES cells, the worst to best VP delivery for MMP depolarization effects was VP from L‐VP, NanoVP, then free‐form VP. Liposomal delivery being the least effective for both cell pairs indicates that the encapsulation of VP in liposomes may limit its immediate bioavailability for mitochondrial localization and function. This could be due to slower cellular uptake via endocytosis, requiring additional time for lysosomal processing and VP release into the cytoplasm before it can reach the mitochondria. ⁴⁵ The fact that NanoVP exhibited enhanced MMP depolarization in VBL‐MDA‐MB‐231 and NCI‐ADR/RES cells compared to free‐form VP suggests that excipient‐free nanoparticle formulations may help counteract P‐gp‐mediated efflux to some extent, potentially by shielding VP from direct recognition by P‐gp until its release occurs intracellularly. Fluorescence images of the effects of photoactivation of each nanoformulation on mitochondrial membrane potential are shown in Figures S5–S8 for each cell line.

It has been well established that free‐form VP, L‐VP and NanoVP preferentially localize to the mitochondria. ³¹ , ⁴¹ , ⁵² Our results also suggest that PDP induces mitochondrial dysfunction first at the electron transport chain level, where we hypothesize some interaction between VP and the ETC complexes, and then depolarizes the mitochondrial membrane potential, driven by ROS‐induced pore formation, and finally induces apoptosis. ⁵³ , ⁵⁴ , ⁵⁵ Despite this tiered response in the mitochondria, where ATP is depleted at low concentrations of VP, we only observe enhanced substrate accumulation at concentrations where mitochondrial membrane potential depolarization is observed (2 μM NanoVP), suggesting that physical mitochondrial damage, no matter how transient, is required to decrease ATP production to reduce P‐gp efflux function. This mirrors our observations from our previous work, where significant increases in P‐gp substrate accumulation were observed at 2 μM free‐form VP at 1 J/cm². ²⁶ In the present study, increases in P‐gp substrate accumulation were observed in the drug‐resistant cell lines after NanoVP‐PDP, but only for calcein in the drug‐resistant TNBC cells, and only rhodamine 123 in the drug‐resistant ovarian cancer cells, without damaging the surface level P‐gp as detected with the monoclonal UIC2 antibody. A potential improvement is to extend the incubation time with NanoVP before photoactivation to allow VP release after lysosomal degradation of NanoVP.

Additionally, the PDP ranges (IC₃₀ values) are summarized in Tables 2, 3, 4. In our studies, the PDP range is determined based on drug‐resistant cells, as the priming range in parental cells is too low to induce measurable changes in P‐gp function. Since cell death is a primary goal of photochemical treatment, achieving priming in drug‐resistant cells while also inducing cell death in parental cells at the same PDP range is considered a therapeutic success. ²⁶ From this, we observed that the priming ranges are similar when comparing free‐form VP and NanoVP. Since NanoVP is hypothesized to accumulate passively after the initial burst release,' where free monomer is released into biological media, this would mimic free‐form VP accumulation. This suggests that both free‐form VP and NanoVP are bioavailable in the mitochondria at comparable concentrations, enabling effective PDP‐induced apoptosis upon photoactivation. In contrast, the higher IC₃₀ PDP range observed for L‐VP in both TNBC and ovarian cancer cells suggests that the degradation kinetics of the liposomal formulation may delay VP bioavailability in the mitochondria. ⁴⁵ This slower release profile could impact the timing of mitochondrial VP accumulation, leading to a lag in ATP depletion and P‐gp inhibition, ultimately requiring higher VP concentrations to achieve effective priming. Given that mitochondrial localization is critical for PDP efficacy, these results further reinforce the importance of optimizing drug delivery systems to enhance bioavailability in resistant cancer cells. Future studies should explore whether adjusting the degradation kinetics of L‐VP formulations or modifying the lipid composition could improve VP release dynamics, allowing for enhanced mitochondrial targeting and more effective PDP‐mediated drug sensitization.

LysoPC VP was not observed to decrease ATP‐production coupled respiration or depolarize the MMP at any tested concentration in any of the parental or drug‐resistant cell lines. Previous work has characterized the subcellular localization of VP conjugated to 16:0lysoPC and found that the conjugation changes the subcellular localization from preferential mitochondrial localization to lysosomal localization. ⁵² This shift affects how VP exerts its therapeutic effects. Mitochondrial localization is crucial for ATP depletion and mitochondrial membrane depolarization, which are key mechanisms in photodynamic priming for MDR reversal. Given that lysoPC VP did not induce ATP depletion or MMP depolarization at any tested concentration, its altered localization likely reduces its ability to disrupt mitochondrial function and inhibit P‐gp activity through metabolic targeting. We were also not able to identify a photodynamic priming range at 1 J/cm² at any tested concentration of lysoPC VP (0–20 μM). A study from Rizvi et al. showed that a PDT dose (μM × J/cm²) that resulted in a 20% decrease in viability was 3.6 for liposomal delivery of VP but was 48.5 for liposomal delivery of lysoPC VP. ³⁰ Additionally, they showed that a PDT dose of 60 induces approximately a 25% decrease in tumor volume when VP is delivered with liposomal lysoPC VP. However, a PDT dose of 10 induced approximately a 75% decrease in tumor volume when VP is delivered with liposomal unconjugated VP at the same concentration (1 μM). ³⁰ study further mirrors the differences in uptake we observed when VP is conjugated to lysoPC vs. free form or the nanoformulations when delivered at the same concentration (2 μM). We would likely have to deliver significantly higher concentrations of lysoPC VP to observe photoactivation effects at a PDP dose. LysoPC VPs' decreased overall VP bioavailability, resulting in reduced photodynamic effects, in combination with the lack of bioavailable VP in the mitochondria making lysoPC conjugation of VP a less effective candidate for PDP‐mediated P‐gp inhibition.

These results also suggest a potential correlation between the affinity of VP for mitochondria and P‐gp affinity to overcome MDR. We found that L‐VP and NanoVP engage with P‐gp to a similar degree as free‐form VP. We also observed that L‐VP and NanoVP exploit the increased energy demand in drug‐resistant cells to deplete ATP to a similar level as free‐form VP. However, we observed that lysoPC maintains some level of P‐gp engagement but cannot exploit the metabolism like the other formulations can. A way to test this potential correlation would be to test a photosensitizer that is not a substrate for P‐gp but localizes to the mitochondria. This would effectively tell us whether mitochondrial localization alone is sufficient to induce ATP depletion and impact P‐gp function, or if P‐gp engagement is a prerequisite for metabolic exploitation in drug‐resistant cells. If a mitochondria‐targeted photosensitizer that is not a P‐gp substrate can still deplete ATP and sensitize MDR cells, this would suggest that mitochondrial disruption alone is a viable mechanism for overcoming resistance, independent of P‐gp interaction. Conversely, if only P‐gp substrates that also localize to mitochondria exhibit this dual effect, it would reinforce the hypothesis that P‐gp affinity and mitochondrial targeting work synergistically to deplete ATP and impair drug efflux.

A major limitation of this study is the reliance on in vitro models, which, while informative, may not fully capture the complexity of MDR in heterogeneous tumor microenvironments. In vivo studies are needed to determine whether NanoVP or L‐VP formulations can effectively reverse drug resistance in tumor‐bearing models and how formulation‐dependent VP retention impacts chemotherapy efficacy over time. Additionally, to establish the proposed mitochondrial‐P‐gp correlation, it will be necessary to test mitochondria‐targeted photosensitizers that are not P‐gp substrates, as well as engineered VP analogs with altered mitochondrial affinity or efflux susceptibility. Moreover, all the tested nanoformulations in this study are non‐targeted. Importantly, future studies should investigate if targeted formulations of VP such as photoimmunoconjugates (PICs), either alone or conjugated to nanoliposomes, might be able to modulate P‐gp specifically at cancer cells. ⁵⁶ , ⁵⁷

Molecular and nanoengineering of VP have been investigated for since the clinical formulation of VP, such as Visudyne, which is formulated with lipids. ²⁷ Despite the easy implementation of repurposing L‐VP, which is already FDA approved, NanoVP appears to outperform L‐VP and free‐form VP in the context of photoactivation to overcome MDR by enhancing PDP‐mediated ATP depletion and improving substrate accumulation. In the context of the tested cell pairs, TNBC and ovarian cancer, these results lay the framework for potential clinical translation where inhibition of P‐gp via photodynamic priming can be incorporated into a clinical workflow to complement the current standard of care treatments. TNBC is extremely aggressive with limited treatment options available, and P‐gp is often overexpressed. ⁵⁸ , ⁵⁹ Similarly, ovarian cancer is highly aggressive and can become highly metastatic when not diagnosed in early stages, with P‐gp expression reported in between 15% and 35% of cases. ⁶⁰ , ⁶¹ , ⁶² In both cases, we provide a mechanistic framework for how photoactivation works to sensitize cells to chemotherapy to improve treatment outcomes in the context of MDR.

CONCLUSION

Our findings highlight NanoVP as a promising nano‐formulation for overcoming MDR through a multi‐faceted mechanism of action. Unlike free‐form VP, which is rapidly recognized and effluxed by P‐gp, NanoVP exhibits a temporary shielding effect, allowing for increased intracellular accumulation before efflux mechanisms take effect. This suggests that NanoVP can somewhat evade P‐gp initially, facilitating mitochondrial localization where it can exploit the increased energy demands of drug‐resistant cells to induce ATP depletion. Additionally, NanoVP maintains its ability to engage P‐gp at the plasma membrane, making it a unique formulation that interacts with both P‐gp function and mitochondrial metabolism. While complete P‐gp evasion remains a challenge, the ability of NanoVP to partially bypass efflux while still triggering ATP depletion provides a strategic advantage for PDP in MDR cancers. These results underscore the potential of NanoVP as a clinically translatable nanotechnology‐driven approach to sensitizing resistant tumors to chemotherapy and overcoming P‐gp‐mediated drug efflux.

AUTHOR CONTRIBUTIONS

IR, SVA, and HCH conceived the project and supervised the study, with help from AS. IR, SVA, HCH, and AS designed the experiments. IR, AM, and KAM synthesized the nanoformulations, performed experiments and analyzed the data. IR, AS, SVA, and HCH drafted the manuscript. SVA, HCH, and AS provided a critical revision of the manuscript.

CONFLICT OF INTEREST STATEMENT

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Supporting information

Figure S1.

Rahman I, Meda A, Moore KA, et al. Nanoengineered photosensitizers for photodynamic priming to overcome P‐glycoprotein‐mediated multidrug resistance. Photochem Photobiol. 2026;102:328‐350. doi: 10.1111/php.70049

DATA AVAILABILITY STATEMENT

All the data supporting the findings of this study are available within the article and supplementary materials from the corresponding author.