ML19H02 is the Ru PS with the chemical formula rac-[Ru(phen)₂(IP-4T)](Cl)₂, where phen=1,10-phenanthroline and IP-4T is imidazo[4,5-f] phenanthroline tethered to α-quaterthiophene. ML19H02 was prepared as previously described (24) and characterized by 1D and 2D ¹H NMR spectroscopy and Electrospray Ionization (ESI⁺) mass spectrometry. Its purity was estimated at >95% by HPLC.

The synthesis technique follows our previously reported radioPDT production in Dinakaran et al. (2020) with some modifications (9). Briefly, 45 mmol of lanthanum chloride (LaCl₃, 99.9%) and 5 mmol of cerium (III) chloride (CeCl₃.6H₂O, 99.9%) were dissolved in 24 mL anhydrous methanol by stirring and vortexing at room temperature for 30 minutes. Similarly, 100 mmol of ammonium fluoride (NH₄F) was dissolved in 180 mL of anhydrous methanol. Next, the NH₄F solution was maintained at 70 ⁰C with constant stirring while lanthanum (III) chloride and cerium (III) chloride solution (90% LaCl₃, 10% CeCl₃) was added to the NH₄F dropwise and kept stirring for 2 hrs. The reaction occurred in a nitrogen protected environment with all oxygen and water purged. The resultant LaF₃:Ce³⁺ (10% doping ratio) NSC was pelleted by ultracentrifugation, and the final product was resuspended in water. The quality of LaF₃:Ce³⁺ NSC was characterized by transmission electron microscopy (TEM), UV-Vis, and dynamic light scattering (DLS).

Next, as described earlier, LaF₃:Ce³⁺ NSC and PPIX or Ru were encapsulated into PEG-PLGA by the nanoprecipitation method into PPIX/radioPDT NP or Ru/radioPDT, respectively (9). For PPIX/radioPDT NP, 100 mg of PEG-PLGA, 15 mg of LaF₃:Ce³⁺ NSC, and 2 mg of PPIX were dissolved in 10 mL of acetonitrile by sonicating for 15 mins and rotating at room temperature for 30 mins. For Ru/radioPDT, 2 mg Ru, 100 mg PEG-PLGA, and 15 mg/mL LaF₃:Ce³⁺ NSC were quickly dissolved in 10 mL of acetonitrile without sonication. Of note, the PS and NSC loading amounts were chosen based on the favorable physical properties of the resultant nanoparticle. The organic phase reactants were added dropwise to 200 mL of MilliQ water using a variable flow peristaltic pump. The organic solvent was removed by vacuum evaporation using a rotavapor system, and the left-over solution was stirred at room temperature overnight. Next, the NPs were purified and concentrated using a tangential flow filtration (TFF) to generate PPIX/radioPDT NP or Ru/radioPDT NP. A hollow fiber cartridge (GE Healthcare, MA, USA) with 500,000 NMWC pore size was used for TFF. Via TFF, the radioPDT NP was washed in 20X excess milliQ H₂O with a transmembrane pressure maintained at 1 bar. A final volume of 3 mL of NPs was collected, and the particle recovery and encapsulation were measured by DLS and TEM.

UV-Vis absorbance spectra were measured to ensure the successful loading of PS in PEG-PLGA. 1.5 µl of diluted radioPDT NP was placed in a nanodrop (DS-11 FX⁺, DeNovix Inc., Wilmington, DE, USA), and the absorbance spectra were collected over a range of 190-800 nm. Dynamic Light Scattering (DLS) was used to determine the size distribution, particle concentration, and polydispersity index (PDI) of the NPs. The sample was diluted 1000X in milliQ H₂O, and 1 mL of solution was taken in a cuvette for measurement using Brookhaven Zetasizer (Brookhaven, Holtsville, NY, USA). The zeta potential was also measured with the particles suspended in milliQ H₂O.

A 10x dilution in milliQ H₂O (3 µL) was placed on a 400-mesh copper grid coated with carbon (Electron Microscopy Sciences, PA, USA), allowing 3 minutes to stabilize before excess water was removed with a filter paper. 3 µl of 2% uranyl acetate (UA) was added to counterstain the grid for 2 minutes before the excess was similarly removed. The NPs were imaged on a JEM 2100 TEM (JEOL, Tokyo, Japan) using a beam energy of 200 kV under a Gatan Quantum GIF energy imaging filter. Images were captured with a Orius SC200D camera under brightfield conditions.

The stability of the NPs was determined by size, zeta potential and UV-Vis measurements in physiologic media. The NPs with different constituents (PEG-PLGA, NSC/PEG-PLGA, PPIX/radioPDT, and Ru/radioPDT) were added to 8% FBS containing phenol red-free DMEM medium. Samples were incubated at 37°C for 24 hrs and 48 hrs. The size distribution and zeta potential of the NPs were measured using Brookhaven Zetasizer (Brookhaven, Holtsville, NY, USA). UV-Vis was measured using a nanodrop (DS-11 FX⁺, DeNovix Inc., Wilmington, DE, USA).

Single oxygen yield was measured via the SOSG probe and its direct fluorescence similar to Dinakaran et al. (9). For the SOSG measurement, 1×10^12/mL of radioPDT NP and control NSC NPs, and UV-Vis standardized equivalent PPIX or Ru were added to a 96-white well plate. Subsequently, 10 μM of SOSG was added, and the volume was adjusted to 100 μL with 2X PBS at a pH of 7.4. One plate was kept in the dark while the other was irradiated at 5 J/cm² with a 402 nm monochromatic LED light source (HouLight, China) at a fluence of 12 mW/cm², which was calibrated for fluence and homogeneity (+/- 5%) using a Thorlabs PM100D photometer (Thorlabs Inc, Newton, NJ, USA). After light irradiation, the SOSG fluorescence was measured (excitation 485 nm, emission 520 nm) by a FLUOstar Omega plate reader (BMG Labtech, Ortenberg, Germany). Each light irradiated experimental condition was normalized to the respective dark condition for analysis.

For direct singlet oxygen measurement, singlet oxygen sensitization (Φ_Δ) was measured at room temperature as suspensions in D₂O, or a solution in 99:1 D₂O:DMSO for PPIX. The values were inferred from the intensity of the singlet oxygen emission band near 1276 nm, measured on a PTI Quantamaster spectrometer (Horiba Scientific, Kyoto, Japan) equipped with a near-infrared sensitive Hamamatsu R5509-42 photomultiplier tube cooled to −80 ⁰C. The most intense absorbance region in the excitation spectrum was chosen to do relative quantification. The quantum yield was calculated by relative actinometry, as shown in Equation 1, where I is the integrated intensity of the emission, A is the baseline-corrected absorbance of the solution or suspension at the excitation wavelength, and η is the refractive index of the solvent. The subscript S denotes the standard solution [Ru(bpy)₃](PF₆)₂ in acetonitrile, for which Φ _Δ,s =0.56 (26).

PC3 prostate cancer cell line was cultured in DMEM with 8% FBS supplemented with 100 U/mL penicillin and 100 μg/ml streptomycin (Life Technologies, USA). Cells were cultured under standard conditions in a humidified incubator at 37°C with 5% CO₂.

Cytotoxicity was measured as cell viability by colorimetric assay and cell survival by clonogenic assay. These assays were chosen to more closely examine PDT/radioPDT effect alone, as represented by the colorimetric assay, and radioPDT effect in combination with radiotherapy as represented by the clonogenic assay. For the colorimetric assay, PC3 cells were incubated overnight at 5000 cells/well density in black, clear bottom, 96 well plate (ThermoFisher, Waltham, MA, USA). 1×10¹² NPs/mL and UV-Vis standardized equivalent amount of free PPIX or Ru were added to the cells for 3 hrs. For light activated PDT, cells were irradiated at 2 J/cm² (12 mW/cm²) with a monochromatic LED. Emitting 405 nm light For radiation activated PDT (radioPDT), cells were irradiated at 8 Gy using a Cesium Irradiator (JL Shepherd and Associates, San Fernando, CA, USA). The cell media was changed 2 hrs post-irradiation to remove the nanoparticles and incubated for 72 hrs prior to viability measurement with Alamar blue dye (10% v/v) using fluorescence measurement (excitation 544 nm, emission 590 nm) with a FLUOstar Omega plate reader.

For the clonogenic cytotoxicity assay PC3 cells were seeded into 12-well plates at a density of 500 cells/well. The cells were treated as described for the colorimetric viability assay but with a lower radiation dose of 3 Gy in a single fraction to better represent the contemporary understanding of prostate radiobiology and radiotherapy dose single fraction as used clinically (27). In contrast, 8 Gy was necessary in the colorimetric assay as there was little contribution from radiotherapy’s DNA damage-induced lack of clonogenicity in the 72 hr assay period in comparison to PDT’s more immediate cytotoxic effects, necessitating a higher activating radiation dose for radioPDT to obtain a cell viability readout. Cell media was changed every 3 days until the plate was developed at day 14 by washing once with PBS, fixation with 1% formaldehyde and 1% methanol, and staining with 0.05% crystal violet. The colonies were counted by scanning the plates on a High Content MetaXpress SLS (Molecular Devices, San Jose, CA, USA) system. Plating efficiency was calculated to derive surviving fractions as previously established for clonogenic assays (28).

PC3 cells were grown to 80% confluency in Aclar film in 6-well plate and then treated with Ru/radioPDT for 4 hrs. Primary fixation of cells was done in 2% PFA + 2.5% GTA/0.1 M Cacodylate buffer with 2 mM CaCl_2. After a series of washing steps, the cells were dehydrated in increasing ethanol concentrations, followed by infiltration in 1:1 Durcupan resin and acetone. Cells were polymerized at 60°C over 48 hrs by placing the Aclar sheet on top of a beam capsule with the cells facing down. After polymerization, the Aclar sheet was peeled off from the block, cut with a Leica EM UC6 ultramicrotome, and transferred to a carbon grid for TEM imaging to detect the presence of NPs within the cells via detecting Lanthanum within the LaF₃:Ce³⁺ NSC using elemental mapping by energy-filtered TEM (EFTEM) on a JEM 2100 TEM machine (JEOL, Tokyo, Japan) with a Gatan Quantum GIF system (Gatan, CA, USA).

All experiments were done in triplicate and repeated at least 3 times. Data are reported as mean ± SEM. One-way ANOVA was used to evaluate for statistically significant difference with a p-value cutoff of <0.05 using Prism 8 (GraphPad 5, San Diego, USA).

LaF₃:Ce³⁺ NSC at a 10% dopant ratio ( Table 1 ) was synthesized in a single step wet chemistry process ( Supplemental Flowchart 1 ) as previously reported (9). TEM of LaF₃:Ce³⁺ NSC revealed hexagonal nanocrystal structures ( Figure 2A ), and the dynamic light scattering (DLS) measurement showed a homogenous population (PDI <0.2) of LaF₃:Ce³⁺ NSC with an average size ranging from 30 ± 7 nm. UV-Vis absorption spectroscopy is shown in Figure 2B .

LaF₃:Ce³⁺ NSC and PS (PPIX and Ru) were then encapsulated into polymeric PEG-PLGA NP ( Supplemental Flowchart 2 ). TEM showed the LaF₃:Ce³⁺ NSC encapsulated within the PEG-PLGA NP to produce the Ru/radioPDT and PPIX/radioPDT systems ( Figures 2C, E ). UV-Vis measurements showed the presence of PPIX or Ru in PPIX/radioPDT or Ru/radioPDT, respectively ( Figures 2D, F ). Together, these data showed the successful loading of LaF₃:Ce³⁺ NSC and photosensitizers into PEG-PLGA.

DLS measurement showed the average hydrodynamic diameter of the unloaded PEG-PLGA NP was 81 ± 5 nm, while the sizes of PEG-PLGA loaded with LaF₃:Ce³⁺ NSC was 88 ± 2 nm ( Figures 3A, B ). The full radioPDT system consisting of PEG-PLGA loaded with LaF₃:Ce³⁺ NSC and PPIX (PPIX/radioPDT) was 96 ± 6 nm and 118 ± 3 nm for the LaF₃:Ce³⁺ NSC, and Ru loaded PEG-PLGA (Ru/radioPDT) ( Figures 3C, D ). The PDI of all the NPs preparation was less than 0.25, indicating a relatively homogenous population of NPs. The measured average zeta potential was -17.4 ± 0.7 mV for Ru/radioPDT NP, as opposed to -27.4 ± 0.5 mV and −19 ± 1 mV for PPIX/radioPDT NP and NSC NP, respectively.

After radioPDT synthesis, we sought to determine its stability in physiologic media for determining the effective treatment time window. DMEM (without phenol red) supplemented with 8% FBS at 37°C was used to incubate the NP and the relevant controls over 48 hrs. The size and PDI measured by DLS showed the NPs after 24 hrs and 48 hrs remained relatively unchanged over these time points ( Figure 3E ), suggesting the NPs maintain their stability without aggregation or breakdown in physiologic serum conditions. In addition, the zeta potential ( Supplemental Table 1 ) and UV-Vis ( Supplemental Figure S2 ) measurements showed stable negative charges and the presence of photosensitizer for a period of 48 hours under physiological conditions. Taken together, these data confirmed the stability of radioPDT NPs under these conditions.

In PDT, the cytotoxic effect relies on the ability of PS to generate ROS, importantly singlet oxygen (10, 29). To determine the efficiency of Ru/radioPDT, we used the singlet oxygen sensor green (SOSG) kit (ThermoFisher, MA, USA) as a highly specific probe for singlet oxygen. The generation of singlet oxygen, as measured by SOSG, was significantly higher in free Ru compared to PPIX, as would be expected by Ru’s significantly higher quantum yield (30) ( Figure 4 ). SOSG signal for free Ru was higher than free PPIX and it was also higher for Ru/radioPDT NP compared to PPIX/radioPDT NP. Both PPIX/radioPDT NP and Ru/radioPDT NP gave significantly higher SOSG signals compared to control NSC NPs without PS ( Figure 4 ). Of note, the average concentration of free PS and nanoparticulated PS was standardized, indicating any difference in efficacy may be due to absorption, triplet state formation, and/or sensitization pathways. These trends were corroborated by singlet oxygen quantum yield measurements performed in D₂O based on singlet oxygen emission, whereby Ru/radioPDT NP produced a singlet oxygen quantum yield of up to 5% compared to well under 1% for PPIX/radioPDT and free PPIX or NSC NP.

Next, we sought to determine the photodynamic efficacy of Ru/radioPDT compared to previously reported PPIX/radioPDT. For comparison, PC3 cells were treated with Ru/radioPDT NP or PPIX/radioPDT NP (and the analogous light-based PDT conditions) and assessed for cell viability using the colorimetric assay. Both PPIX and Ru and their NSC formulations exhibited similar cytotoxicity when irradiated with 402-nm light at a radiant exposure of 2 J/cm² (irradiance of 12 mW/cm²) ( Figure 5A ). The NSC NPs alone did not produce any photocytotoxicity under similar conditions. Following a single dose ionizing radiation of 8 Gy, free PPIX or Ru did not produce any measurable cytotoxic effects ( Figure 5B ). Ru/radioPDT NP was more potent than PPIX/radioPDT NP but not statistically different.

Further investigations examined the combined effect of radioPDT and radiotherapy on PC3 survival following delivery of a single 3 Gy ionizing radiation dose in combination with Ru/radioPDT NP or PPIX/radioPDT NP. PC3 survival was significantly decreased in both Ru/radioPDT and PPIX/radioPDT compared to radiation alone, with cell survival reduction of 49% and 57% of the radiation-only condition, respectively ( Figure 6 , Supplemental Figure S3 ). A statistically nonsignificant trend toward decreased survival in the Ru/radioPDT NP group was seen over PPIX/radioPDT NP at this dose of radiation.

Determining intracellular localization of Ru/radioPDT NP indicates the likely organelles affected by the short-range singlet oxygen species during PDT effect in the context of our experimental setup. PC3 cells treated with Ru/radioPDT NP were prepared for TEM imaging as described in the method section. Of note, acetone was used as the fixative during the sample preparation, which did not significantly interfere with the stability of Ru/radioPDT NP. TEM demonstrates the localization of the high contrast inorganic LaF₃:Ce³⁺ NSC in cytoplasmic organelles, including in phagolysosomes and endoplasmic reticulum ( Figure 7A ), but remaining outside the nucleus ( Figure 7B ). This is further confirmed by elemental mapping of Lanthanum EFTEM ( Figures 8A, B ). Together, Ru/radioPDT NP appears to be endocytosed into PC3 cells and localizes to organelles outside the nucleus, particularly critical structures such as the endoplasmic reticulum, and presumably damages these structures when the PDT effect is induced by radiation to achieve cell kill. This contrasts with the well-known DNA damage-mediated radiotherapy effect (31).