Development of ²²⁵Ac-doped biocompatible nanoparticles for targeted alpha therapy

Abstract

Targeted alpha therapy (TAT) relies on chemical affinity or active targeting using radioimmunoconjugates as strategies to deliver α-emitting radionuclides to cancerous tissue. These strategies can be affected by transmetalation of the parent radionuclide by competing ions in vivo and the bond-breaking recoil energy of decay daughters. The retention of α-emitting radionuclides and the dose delivered to cancer cells are influenced by these processes. Encapsulating α-emitting radionuclides within nanoparticles can help overcome many of these challenges. Poly(lactic-co-glycolic acid) (PLGA) nanoparticles are a biodegradable and biocompatible delivery platform that has been used for drug delivery. In this study, PLGA nanoparticles are utilized for encapsulation and retention of actinium-225 ([²²⁵Ac]Ac³⁺). Encapsulation of [²²⁵Ac]Ac³⁺ within PLGA nanoparticles (Z_ave = 155.3 nm) was achieved by adapting a double-emulsion solvent evaporation method. The encapsulation efficiency was affected by both the solvent conditions and the chelation of [²²⁵Ac]Ac³⁺. Chelation of [²²⁵Ac]Ac³⁺ to a lipophilic 2,9-bis-lactam-1,10-phenanthroline ligand ([²²⁵Ac]AcBLPhen) significantly decreased its release (< 2%) and that of its decay daughters (< 50%) from PLGA nanoparticles. PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen significantly increased the delivery of [²²⁵Ac]Ac³⁺ to murine (E0771) and human (MCF-7 and MDA-MB-231) breast cancer cells with a concomitant increase in cell death over free [²²⁵Ac]Ac³⁺ in solution. These results demonstrate that PLGA nanoparticles have potential as radionuclide delivery platforms for TAT to advance precision radiotherapy for cancer. In addition, this technology offers an alternative use for ligands with poor aqueous solubility, low stability, or low affinity, allowing them to be repurposed for TAT by encapsulation within PLGA nanoparticles.

Graphical Abstract

Introduction

Targeted alpha therapy (TAT) is a radiotherapy that uses α-emitting radionuclides for the precise treatment of cancer [1,2,3]. This precise treatment is attained by virtue of the high energy deposition of α-particles within a short range in tissue (< 100 µm) [2, 4]. These characteristics and the cytotoxicity of α-particles can lead to unwanted side effects if they relocate to healthy tissue [5,6,7,8]. To mitigate this issue, α-emitting radionuclides can be targeted to a specific tissue based on either chemical affinity or highly expressed tumor antigens [6, 9, 10]. Targeting by chemical affinity is the mechanism behind Xofigo^®; a radiopharmaceutical composed of [²²³Ra]RaCl₂ that targets hydroxyapatite in bone undergoing remodeling [10]. Because [²²³Ra]Ra²⁺ does not directly target cancer cells, it exerts its effects through indirect irradiation and bystander effects [11]. Direct targeting of cancer cells is facilitated using radioimmunoconjugates composed of an α-emitting radionuclide attached to a targeting vector usually using a bifunctional chelator [12]. Bifunctional chelators with high radiolabeling efficiency and high stability have been developed to minimize the accumulation of α-emitting radionuclides in healthy tissue [13,14,15,16]. Despite recent developments, preventing the relocation of decay daughters from target sites remains a challenge due to the bond-breaking recoil energy of α-particles [17]. To address this issue, different approaches have been proposed: local administration of α-emitting radionuclides, rapid uptake of the radioimmunoconjugate by the cell, or the encapsulation of α-emitting radionuclides within nanoparticles [18]. The latter is an attractive approach for delivering α-emitting radionuclides at therapeutic doses as long as the nanoparticles do not accumulate in healthy tissue [18].

Nanoparticles with various morphologies, structures, and characteristics have been developed for TAT using organic and inorganic compounds [19,20,21]. Organic nanoparticles have been radiolabeled with unchelated and chelated α-emitting radionuclides using ionophore-mediated mechanisms and passive entrapment, respectively. [22,23,24,25,26,27] Retention of the parent radionuclide has been demonstrated to be greater than 90%, whereas the retention of decay daughters is dependent on the nanoparticle size [22,23,24,25,26, 28]. Inorganic nanoparticles are radiolabeled with α-emitting radionuclides by either incorporation within the nanoparticle structure, adsorption, or surface attachment with a ligand [29,30,31,32,33,34,35]. Enhanced retention of decay daughters has been achieved by leveraging the high electron density, core–shell structure, and ion-exchange capacity of inorganic nanoparticles [31, 34]. These results support the potential of inorganic and organic nanoparticles as novel radiopharmaceuticals in preclinical and clinical studies. Translation of nanoparticles into the clinic has been hampered by technical challenges in large-scale production, a lack of understanding of their biological fate, and unknown long-term effects [36,37,38]. Some of these challenges can be overcome by using nanoparticles composed of biocompatible and biodegradable lipids and block-copolymers.

Poly(lactic-co-glycolic acid) (PLGA) is a diblock copolymer—approved by the U.S. Food and Drug Administration—extensively studied for drug delivery and tissue engineering applications [39, 40]. In radiotherapy, PLGA nanoparticles have been used for delivery of radionuclides that emit β-particles ([¹⁷⁷Lu]Lu³⁺) and Auger electrons ([¹¹¹In]In³⁺) [41,42,43,44]. Recently, Sporer et al. radiolabeled PLGA micelles with [²¹¹At]At^– and [¹²⁵I]I^– using either a synthon-based approach or direct radiolabeling of the amphiphilic block copolymer [45]. In this work, PLGA nanoparticles were explored as a biodegradable and biocompatible delivery platform of [²²⁵Ac]Ac³⁺ for TAT. It is hypothesized that PLGA nanoparticles could increase the therapeutic potential of [²²⁵Ac]Ac³⁺ by increasing the retention of the parent and partially containing its decay daughters at the target site. It is also anticipated that PLGA’s status as an FDA-approved diblock copolymer will facilitate preclinical and clinical translation of PLGA nanoparticles encapsulating [²²⁵Ac]Ac³⁺ for therapeutic applications. Encapsulation of [²²⁵Ac]Ac³⁺ within PLGA nanoparticles was performed via active entrapment of the radionuclide during synthesis with a double-emulsion solvent evaporation method. This method was adapted to radiological applications by decreasing reagent volumes and using a cup horn sonicator. These changes were made to: (1) decrease the amount of radioactive waste generated during the optimization of synthesis parameters; and (2) contain the radioactive solutions within closed Eppendorf tubes during sonication. Different synthesis parameters were varied to enhance the encapsulation efficiency of [²²⁵Ac]Ac³⁺ within PLGA nanoparticles. A lipophilic 2,9-bis-lactam-1,10-phenanthroline ligand (BLPhen)—which has shown promise for lanthanide and actinide separation [46, 47]—was used to improve the encapsulation efficiency and retention of [²²⁵Ac]Ac³⁺. Encapsulation of [²²⁵Ac]Ac-macropa complexes was also evaluated due to the potential of this ligand to bind [²²⁵Ac]Ac³⁺ [48]. Retention of [²²⁵Ac]Ac³⁺ and its decay daughters, [²²¹Fr]Fr⁺ and [²¹³Bi]Bi³⁺, was assessed using dialysis against phosphate-buffered saline. The cytotoxicity of free [²²⁵Ac]Ac³⁺ and PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen was evaluated in vitro using murine and human breast cancer cells. PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen were significantly more cytotoxic to breast cancer cells than exposure to free ²²⁵Ac in solution. The enhanced cytotoxicity could be a combination of nanoparticle accumulation at the cell surface, nanoparticle uptake within cells, and the resultant retention of radionuclides near the cells. This study shows that PLGA nanoparticles offer an effective biocompatible delivery mechanism for α-emitting radionuclides that is facilitated by pre-chelation with a lipophilic ligand. PLGA nanoparticles could serve as a platform to repurpose ligands with poor aqueous solubility or low stability in TAT. These nanoparticles have the potential to also advance precision combination therapies by encapsulating a variety of therapeutic payloads, such as radionuclides and chemotherapeutic drugs [41, 42].

Experimental methods

Materials

The chemicals Resomer RG 503 H poly(D,L-lactide-co-glycolide) (lactide:glycolide 50:50, acid terminated, M_w 24,000–38,000), D-α-tocopherol polyethylene glycol 1000 succinate (vitamin E-TPGS, CAS: 9002–96-4), lanthanum chloride heptahydrate (LaCl₃⋅7H₂O, ACS reagent, CAS: 10,025–84-0), ethyl acetate (EtOAc, ≥ 99.5%, ACS reagent, CAS: 141–78-6), phosphate-buffered saline tablets (PBS, P4417), methanol (MeOH, ≥ 99.8%, ACS reagent, CAS: 67–56-1), ammonium acetate (NH₄OAc, ≥ 98%, for molecular biology, CAS: 631–61-8), nitric acid (HNO₃, 70%, ACS reagent, CAS: 7697–37-2), dimethyl sulfoxide (DMSO, ≥ 99.9%, ACS reagent, CAS: 67–68-5), dichloromethane (DCM, ≥ 99.9%, HPLC Plus, CAS: 75–09-2), and 3-(N-morpholino)propanesulfonic acid (MOPS, ≥ 99.5%, CAS: 1132–61-2) were obtained from Sigma Aldrich and used without further purification. Deionized water (DI H₂O, 18.2 MΩ⋅cm) was obtained from a MilliporeSigma Milli-Q water purification system. Slide-A-Lyzer G2 dialysis cassettes (0.5 mL, 10,000 molecular weight cutoff) and the alamarBlue HS cell viability reagent were purchased from Thermo Scientific. The chemicals 1,2-Distearoyl-sn-Glycero-3-Phosphoethanolamine-N-azide (DSPE-N3, Cat# BP-26155) and cyanine5 red-emitting fluorophore with a dibenzocyclooctyne (Cy5-DBCO, Cat# D30F0) group were obtained from BROADPHARM and Lumiprobe, respectively.

Poly(lactic-co-glycolic acid) nanoparticle synthesis

A double-emulsion solvent evaporation method was adapted for the synthesis of PLGA nanoparticles [49, 50]. The main changes include: (1) using a cup-horn sonicator (Q125-110, Qsonica) to generate emulsions; and (2) decreasing the reagent volumes to minimize radiological waste generation. For a typical synthesis, 100 µL of PLGA (40 mg/mL in EtOAc) was mixed with 10 µL of a polar solution (i.e., DI H₂O, MeOH, or a mixture of both) in an Eppendorf tube. This mixture was sonicated for 30 s in a cup-horn sonicator (100% amplitude) to generate a water/oil emulsion. After sonication, 200 µL of vitamin E-TPGS (1 wt % in DI H₂O) was added immediately to the Eppendorf tube and mixed using a vortex mixer. Three sonication cycles of 30 s (100% amplitude) with a 30 s rest time were used to generate a water/oil/water emulsion. The water/oil/water emulsion was added to a beaker containing 2.5 mL of vitamin E-TPGS (0.5 wt % in DI H₂O). These suspensions were stirred at 500–700 rpm for at least 4 h to harden the nanoparticles. After hardening, PLGA nanoparticle suspensions were washed at least three times by centrifugation (15,000 rpm, 20 min, 8 ºC) followed by resuspension and sonication in 1 mL DI H₂O. The parameters and steps just mentioned were the standard conditions employed for the synthesis of PLGA nanoparticles. For radionuclide retention and cell viability studies, water/oil/water emulsions were transferred into a dialysis cassette and dialyzed against PBS (90 mL) stirring constantly (120 rpm). Emulsions were dialyzed for multiple days to assess the radionuclide retention. For cell-based assays, emulsions were dialyzed for at least 4 h in PBS stirring constantly (120 rpm) before diluting them in cell media. The water in the cup-horn sonicator was kept cold by adding ice during the emulsification steps described above.

A schematic representation of the PLGA nanoparticle synthesis method using double-emulsion solvent evaporation and a cup-horn sonicator is shown in Fig. 1. The influence of synthesis parameters on the size distribution of PLGA nanoparticles was assessed by varying PLGA and vitamin E-TPGS concentrations, PLGA payload, PLGA:TPGS volume ratios, sonication time and number of cycles, among other conditions.

Fig. 1

Controlled synthesis of PLGA nanoparticles with a cup-horn sonicator. Schematic representation of PLGA nanoparticles synthesized with a double-emulsion solvent evaporation method. Chemical structure of PLGA and TPGS are shown as a reference. The presence of methyl side groups in poly(lactic acid) [gray] makes it more hydrophobic than poly(glycolic acid) [orange]

Surrogate studies used LaCl₃ in dilute HNO₃ or in MOPS buffer as a payload during synthesis of PLGA nanoparticles. In these experiments, 10 µL of LaCl₃ solution was mixed with 200 µL of PLGA solution (40 mg/mL in EtOAc) in an Eppendorf tube. The mixture was sonicated for 30 s and then 400 µL of vitamin E-TPGS (1 wt % in DI H₂O) was added. Three sonication cycles of 30 s with a 30 s rest time were used to generate the water/oil/water emulsion. This emulsion was transferred into a beaker with 4 mL of vitamin E-TPGS (1 wt % in DI H₂O) and stirred overnight at 700 rpm. PLGA nanoparticle suspensions were collected by centrifugation and subsequently washed with DI H₂O. Both the PLGA nanoparticle suspension and the supernatant were assayed to determine the concentration of La and calculate its encapsulation efficiency within the nanoparticles. Copper-free click chemistry was used to prepare fluorescent PLGA nanoparticles by combining an azido phospholipid molecule and a cyanine5 red-emitting fluorophore with a dibenzocyclooctyne group [51, 52]. Briefly, 100 µL of PLGA (40 mg/mL), 12 µL DSPE-N3 (2.5 mg/mL), and 4 µL Cy5-DBCO (1 mg/mL) were mixed in an Eppendorf tube and sonicated for 30 s. The following steps for PLGA nanoparticle synthesis were performed as described above as the standard conditions.

Characterization

Morphological characterization was performed by scanning electron microscopy (SEM) using a Carl Zeiss MERLIN VP-SEM with a secondary electron detector operating at 1 kV accelerating voltage. A PLGA nanoparticle suspension was diluted tenfold in DI H₂O and then drop-casted onto a silicon wafer for SEM characterization. The size distribution of PLGA nanoparticle suspensions—diluted tenfold in DI H₂O—was characterized using dynamic light scattering (DLS) in a Malvern Zetasizer Nano ZS. PLGA nanoparticle suspensions prepared with different conditions were characterized in duplicate to determine the mean hydrodynamic size and polydispersity index (PDI). The stability of PLGA nanoparticles was assessed over time after dilution in DI H₂O, PBS, and cell media, respectively. Characterizations correspond to three runs with at least ten measurements each as defined in the DLS software. The DLS instrument was also used to measure the zeta potential of PLGA nanoparticles. Changes in the zeta potential after exposure to cell media were used to determine protein adsorption. PLGA nanoparticles were exposed to cell media, precipitated by centrifugation, and dispersed in DI H₂O for zeta potential characterization. Inductively coupled plasma mass spectrometry (ICP-MS) was performed using a ThermoFisher Scientific iCAP-RQ quadrupole ICP-MS instrument to assess the concentration of La within PLGA nanoparticles. Sample preparation, quantification, and analysis were performed as described previously [49].

Encapsulation and retention of radionuclides

Actinium-225 ([²²⁵Ac]Ac³⁺, t_1/2 = 9.9 d) was obtained from a thorium generator ([²²⁸Th]Th⁴⁺, [²²⁹Th]Th⁴⁺, and [²³²Th]Th⁴⁺) at Oak Ridge National Laboratory after various chemical separation processes involving anion and cation exchange chromatography [53]. For encapsulation and retention experiments, an [²²⁵Ac]Ac³⁺ stock solution in 0.1 M HNO₃ was mixed with NH₄OAc (0.1 M, pH 6.0), MeOH, or a 5 mM lipophilic BLPhen ligand solution in MeOH. Details on the synthesis of the lipophilic BLPhen ligand can be found elsewhere [47]. Encapsulation of [²²⁵Ac]Ac-macropa was also explored because of the high stability of this complex [48]. Chelation of [²²⁵Ac]Ac³ with macropa [2.7 mM] was performed as reported by Thiele et al. [48] Chelation of [²²⁵Ac]Ac³⁺ by BLPhen and macropa was assessed using thin-layer chromatography with alumina-backed silica plates and sodium citrate (0.4 M in 10% MeOH) as stationary and mobile phases, respectively. Autoradiography of the plates was performed using a Cyclone Plus phosphor imager (PerkinElmer) and super-resolution phosphor screens.

Synthesis of PLGA nanoparticles encapsulating [²²⁵Ac]Ac³⁺ was performed using standard conditions. A 10 µL solution of ²²⁵Ac in NH₄OAc, a MeOH/DI H₂O mixture, or [²²⁵Ac]AcBLPhen in a MeOH/DI H₂O mixture was used as the payload. The activity of [²²⁵Ac]Ac³⁺ in these solutions ranged between 3.7 kBq and 55.5 kBq (0.1 µCi and 1.5 µCi, respectively). Adjusting the [²²⁵Ac]Ac³⁺ activity resulted in various MeOH:DI H₂O volume ratios (e.g., 95:5, 90:10, and 75:25) during PLGA nanoparticle synthesis. PLGA nanoparticle suspensions were washed three times with DI H₂O to determine the encapsulation efficiency. After each centrifugation, the supernatant was carefully removed with a transfer pipette and stored in a 15 mL Falcon tube. To account for total radioactivity distribution, the PLGA nanoparticle suspension, the supernatant, and the Eppendorf tube used for sonication were assayed using γ-ray spectroscopy on a high purity germanium (HPGe) detector. The HPGe detector had a crystal active volume of ~ 100 cm³ and a beryllium window; and was coupled to a PC-based multichannel analyzer (Canberra Industries, Meriden, CT). Energy and efficiency calibrations were determined by γ-ray sources traceable to the National Institute of Standards and Technology. Radionuclide retention was assessed by quantifying the fraction of [²²⁵Ac]Ac³⁺, [²²¹Fr]Fr⁺ (t_1/2 = 4.5 min), and [²¹³Bi]Bi³⁺ (t_1/2 = 45.6 min) activity in the dialysate over time. The water/oil/water emulsion (~ 0.5 mL) was transferred into a dialysis cassette and dialyzed against PBS (90 mL) stirring constantly (120 rpm). Aliquots (5 mL) were taken from the dialysate at different time points (e.g., 0.5, 4, 24, 120, and 168 h) and assayed immediately on an HPGe detector to quantify the activity of [²²¹Fr]Fr⁺ and [²¹³Bi]Bi³⁺ [54, 55]. The same aliquot was counted the next day—decay daughters are now in equilibrium with [²²⁵Ac]Ac³⁺—to assess the fraction of [²²⁵Ac]Ac³⁺ in the dialysate [54, 55]. The equations used to determine the encapsulation efficiency and retention of radionuclides are presented in the supporting information. A 5 mL solution of fresh PBS was added to the beaker to maintain a constant dialysate volume.

Quantum chemical calculations of [ ²²⁵ Ac]AcBLPhen

Electronic structure calculations of the [²²⁵Ac]AcBLPhen complex were performed employing density functional theory (DFT) with the Gaussian 16, Revision A.03 software package [56]. We utilized the hybrid B3LYP functional [57, 58] and the DFT-D3 approach of Grimme with zero damping [59] to account for van der Waals interactions. A standard 6–31 + G(d) basis set was used for main group elements and hydrogen for geometry optimizations. The f-element Ac was modeled using large-core (LC) relativistic effective core potential (RECP) and the associated (7s6p5d)/[4s3p3d] basis set [60]. Since LC RECP calculations include the 4f electrons in the core, they were performed on a pseudo singlet state configuration. Frequency calculations at the B3LYP-D3/LC/6–31 + G(d) level were performed to ensure real vibrational modes for the minimum ground state structures and to provide zero-point energies (ZPEs). ZPEs and thermal corrections (T = 298.15 K) were added to the total energy to obtain the Gibbs free energy. Thermal contributions to the gas-phase Gibbs energies were calculated using standard molecular thermodynamic approximations [61], except that vibrational frequencies lower than 60 cm^–1 were raised to 60 cm^–1, adopting a methodology introduced by Truhlar et al. [62] based on the so-called quasiharmonic approximation. Using the gas-phase geometries, implicit solvent corrections were obtained at 298.15 K with the SMD [63] solvation model as implemented in Gaussian 16. Complexation free energies in aqueous solution, ΔG_aq, were calculated using the thermodynamic cycle approach described in our previous works on vanadium- and uranyl-containing complexes [64, 65]. Chemical bonding analysis was performed for the DFT-optimized structures using the natural bond orbital (NBO) methodology [66].

Tissue culture

Cell lines E0771 murine breast carcinoma (CRL-3461™), MCF-7 human breast adenocarcinoma (HTB-22), and MDA-MB-231 human breast adenocarcinoma (HTB-26) were obtained from the American Type Culture Collection. Cells were seeded in a 75 cm² tissue culture flask and cultured at 37 °C in humidified air containing 5% CO₂. A 1:1 mixture of Dulbecco’s Modified Eagle’s Medium and Ham’s F12 medium, supplemented with 10% fetal calf serum, 2 mM L-glutamine, 100 I.U./mL penicillin, and 100 µg/µL streptomycin, was used to culture the three cell lines (DMEM/F12 complete media). Cells were passaged by trypsinization and seeded at ~ 3 × 10³ cells per well in a flat-bottom tissue culture–treated 96-well plate (Corning Costar). Cells were incubated for 48 h at 37 °C in humidified air containing 5% CO₂, reaching 70% to 80% confluency before treatment.

Breast cancer cell lines were exposed for 24 h to different [²²⁵Ac]Ac³⁺ activities delivered as free [²²⁵Ac]Ac³⁺, [²²⁵Ac]AcBLPhen, and PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen. Radioactive solutions were initially prepared in PBS and then serially diluted into DMEM/F12 complete media. The activities used in this work ranged between 0.14 ± 0.01 kBq/mL and 7.2 ± 1.6 kBq/mL (3.7 ± 0.2 nCi/mL and 194 ± 42 nCi/mL, respectively). This activity range was defined based on in vitro experiments with targeted [²²⁵Ac]Ac³⁺ and [²²⁵Ac]Ac³⁺ encapsulated within nanoparticles [30, 32, 67,68,69,70]. After 24 h of exposure to [²²⁵Ac]Ac³⁺, the radioactive media (200 µL/well) was removed and cells were washed three times with PBS. Metabolic activity of cells was used as a measure of viability after exposure to [²²⁵Ac]Ac³⁺. Cell viability was assessed 1 h and 72 h after exposure to [²²⁵Ac]Ac³⁺ using the alamarBlue HS cell viability reagent. Cell viability was calculated as percent viability relative to untreated cells, a negative control. AlamarBlue reagent was mixed with DMEM/F12 complete media (10 µL alamarBlue per 90 µL media) and then added to the 96-well plate (100 µL per well). Cells were incubated at 37 °C for 1 h before measuring the alamarBlue fluorescence (λ_exc = 560 nm and λ_emi = 590 nm) with a BioTek Cytation 1 cell imaging multimode reader (red filter block, λ_exc = 530 nm and λ_emi = 590 nm). After the fluorescence was recorded, the alamarBlue media was removed and replaced with fresh DMEM/F12 complete media, and cells were incubated at 37 °C, 5% CO₂ in the humidified incubator. The viability assay was repeated 72 h after exposure to ²²⁵Ac following the steps described previously. To assess the influence of nanoparticle concentration on cell viability, cells were exposed for 24 h to similar [²²⁵Ac]Ac³⁺ activities delivered by PLGA nanoparticles prepared with different polymer concentrations (e.g., 10 mg/mL, 20 mg/mL, and 40 mg/mL). The PLGA nanoparticle concentration ranged between 16.7 µg/mL and 266.7 µg/mL. The viability assay for the different nanoparticle concentrations was performed as described previously 1 h post-exposure. To determine the effect of exposure time on cell viability, E0771 and MCF-7 cells were exposed to free [²²⁵Ac]Ac³⁺ in PBS and PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen for different time points (e.g., 1 h, 3 h, 6 h, and 24 h). Cell viability was evaluated 24 h and 48 h after exposure to ²²⁵Ac.

Immunostaining for γ-H2AX was performed following the protocol established by Noubissi et al. [71] After exposure to [²²⁵Ac]Ac³⁺, cells were washed with PBS (× 3) and fixed using a 4% (v/v) paraformaldehyde solution in PBS for 10 min. After washing the cells with PBS (× 3), they were permeabilized with a 0.3% (v/v) Triton X-100 solution in PBS for 10 min. Following permeabilization, cells were washed with PBS (× 3) and blocked with a MAXblock blocking solution overnight at 4 ºC. After washing the cells with PBS (× 3), cells were incubated with the primary antibody for 1 h at 37 ºC followed by 30 min at room temperature with rocking. The primary antibody—Anti-phospho-Histone H2A.X (serine 139) clone JBW301 (MilliporeSigma) —was diluted to 2 µg/mL in MAX bind staining medium. Following this incubation, cells were washed with PBS (× 3) and incubated with the fluorescent-tagged secondary antibody. The secondary antibody—Anti-Mouse IgG (H + L), F(ab’)2 fragment CF 488A (Millipore Sigma)—was diluted to 1 µg/mL in MAX bind staining medium. Incubation with the secondary antibody was performed for 1 h at 37ºC. After incubation, cells were washed with PBS (× 3) and stained with Hoechst. Hoechst (10 mg/mL) was diluted in PBS (1:2000) and incubated for 5 min at room temperature. Cells were washed with PBS (× 3) and maintained in 100 µL of PBS for imaging and image acquisition. Images were acquired using a Biotek Cytation 1 cell imaging multi-mode reader (Biotek Instruments Inc.). The filter cubes used to image the cells were DAPI (Cat #1,225,100) and GFP (Cat #1,225,101). A spot counting macro in ImageJ was used for direct foci counting.

Interaction of fluorescent PLGA nanoparticles with E0771 and MDA-MB-231 cells was evaluated using confocal microscopy. Cells were seeded at ~ 3 × 10³ cells per well in a 6-well plate, incubated for 48 h at 37 °C in humidified air containing 5% CO₂, and treated with 16.5 µg/mL and 66.4 µg /mL of fluorescent PLGA nanoparticles per well. Images were recorded after 2 h and 24 h of incubation with PLGA nanoparticles.

Results

A double-emulsion solvent evaporation method was used to synthesize PLGA nanoparticles using vitamin E-TPGS as emulsifier and stabilizer [49, 50]. Spherical PLGA nanoparticles with different size distributions were obtained after varying synthesis parameters (Fig. 2a and Fig. S1). The size distribution was influenced by the method used for solvent removal (Fig. 2b) and the parameters used during synthesis (Fig. S2). PLGA nanoparticles had a smaller mean hydrodynamic size when using stirring (Z_ave = 151.8 nm) for solvent removal relative to those obtained with dialysis (Z_ave = 255.4 nm). Solvent removal by dialysis also resulted in a broader size distribution relative to that obtained with stirring (Fig. 2b). The concentration of PLGA and vitamin E-TPGS, number of sonication cycles, and payload solvent significantly influenced the mean hydrodynamic size and polydispersity of PLGA nanoparticles (Fig. S2). A lower concentration of vitamin E-TPGS (0.1 wt %) yielded the largest mean hydrodynamic size, Z_ave > 1,000 nm (Fig. S2b). As shown in Fig. S2 h, a single sonication cycle increased both the hydrodynamic size (Z_ave = 627.8 nm) and polydispersity (PDI = 0.428) of PLGA nanoparticles relative to those synthesized using standard conditions (Z_ave = 155.3 nm and PDI = 0.172). The intensity size distribution was used to assess the stability of the PLGA nanoparticles in DI H₂O, PBS, and DMEM/F12 complete media over time (Fig. S3). The mean hydrodynamic size was similar for PLGA nanoparticles dispersed in DI H₂O and PBS up to 7 days (Table S1). After 12 days, PLGA nanoparticles in PBS had a larger mean hydrodynamic size (Z_ave = 250.4 nm) compared to those in DI H₂O (Z_ave = 158.5 nm). The size distribution of PLGA nanoparticles in DMEM/F12 complete media was characterized by two peaks (Fig. S3C): a small one below 100 nm (< 3.0%) and a large one around 200 nm (> 15–20%). The presence of the smaller peak decreased the mean hydrodynamic size of PLGA nanoparticles to ~ 130 nm (Table S1). The zeta potential of PLGA nanoparticles remained unchanged after exposure to DMEM/F12 complete media for 24 h and 72 h. The average zeta potential of PLGA nanoparticles diluted in DI H₂O was − 24.4 ± 1.9 mV compared to those diluted in DMEM/F12 complete media − 29.3 ± 4.3 mV and − 22.1 ± 0.1 mV, respectively.

Fig. 2

Spherical PLGA nanoparticles with on average a uniform size distribution were obtained by a double-emulsion solvent evaporation method. a SEM image of PLGA nanoparticles synthesized using standard conditions and evaporated through stirring. Inset corresponds to higher magnification micrograph of PLGA nanoparticles synthesized as described previously. b Intensity size distribution of PLGA nanoparticles synthesized with stirring or dialysis for solvent evaporation

Chelation of [²²⁵Ac]Ac³⁺ with a lipophilic BLPhen ligand was demonstrated using thin-layer chromatography and autoradiography. Free [²²⁵Ac]Ac³⁺ moves with the solvent front to the top of the plate, whereas [²²⁵Ac]AcBLPhen remains at the baseline (Fig. 3a and b). Lipophilic BLPhen ligand chelates [²²⁵Ac]Ac³⁺ rapidly after co-mixing both solutions at room temperature. The chelation efficiency is 92.0 ± 3.7% at 5 mM BLPhen ligand (n = 3); however, it decreases below 70% for concentrations < 200 µM (Fig. 3c). To gain structural insights into the coordination chemistry of [²²⁵Ac]Ac³⁺ with the BLPhen ligand, quantum chemical calculations were performed using DFT for the truncated version of BLPhen, where the long hexyl substituents were replaced by methyl groups. Based on a recent x-ray absorption fine structure spectroscopy (XAFS) study [72], pointing to the presence of a 1:2 local coordination in the Ln–BLPhen species formed in the organic phase, the calculations were performed for the cationic [²²⁵Ac][Ac(BLPhen)₂]³⁺ complex. The electrostatic potential map of the BLPhen molecule clearly shows the highly electronegative regions formed by the amide oxygen (O_amide) and phenanthroline nitrogen (N_phen) groups constituting the ideal chelating sites suitable for [²²⁵Ac]Ac³⁺ binding and retention (Fig. 3d). These favorable binding interactions were confirmed by the calculated negative Gibbs free energy value for the displacement of nine water molecules from the Ac³⁺ aqua ion by two BLPhen ligands, according to reaction 1:

Fig. 3

Understanding the chelation and coordination of [²²⁵Ac]Ac³⁺ by the lipophilic BLPhen ligand. a Autoradiographic images of [²²⁵Ac]Ac³⁺ and [²²⁵Ac]AcBLPhen at different times after co-mixing. Color scale shows different activities/intensities of [²²⁵Ac]Ac³⁺ and decay daughters with blue and red as the lowest and highest, respectively. b Histograms of autoradiographic images of [²²⁵Ac]Ac³⁺ and [²²⁵Ac]AcBLPhen at different times after co-mixing. c Chelation of [²²⁵Ac]Ac³⁺ as a function of BLPhen ligand concentration, 5 min after co-mixing. Insert corresponds to the chemical structure of the lipophilic BLPhen ligand.⁴⁷. d Chemical structure of BLPhen with plotted electrostatic potential map (0.01 e/a₀³ isovalue) and charges on O and N atoms obtained from natural population analysis. e Structure of the [²²⁵Ac][Ac(BLPhen)₂]³⁺ complex optimized at the B3LYP-D3/LC/6–31 + G(d) level of theory. Color scheme: Ac, sky blue; C. grey; H, white; N, blue; O, red

$$Ac{(H}_{2}O{)}_{9}^{3+}+2BLPhen\rightleftharpoons Ac(BLPhen{)}_{2}^{3+}+9{H}_{2}O, \Delta {G}_{aq}= -28.12\frac{kcal}{mol}$$

(1)

The DFT-optimized geometry of the resulting [²²⁵Ac][Ac(BLPhen)₂]³⁺ complex shows that the Ac–O_amide bond distances are 0.25 Å shorter than the Ac–N_phen distances. This indicates that the amide oxygen atoms possess relatively stronger electron donating ability toward [²²⁵Ac]Ac³⁺ in comparison with the phenanthroline nitrogen atoms (Fig. 3e). To further probe these interactions, we performed NBO analysis of chemical bonding in the actinide complex. As expected, the calculated Wiberg bond indices, which serve as a measure of bond order, were higher for Ac–O_amide bonds (0.152) than for Ac–N_phen bonds (0.111). Additionally, the NBO analysis indicated that the formal polarized Ac–O_amide and Ac–N_phen dative bonds originated from a σ-type donation of electron density from the donor O and N lone pairs to the vacant acceptor orbitals of primarily 6d character on [²²⁵Ac]Ac³⁺.

The encapsulation efficiency of La within PLGA nanoparticles increased with decreasing payload volume (1.27%, 0.67%, and 0.45% for 5 µL, 10 µL, and 20 µL, respectively). Co-mixing LaCl₃ with a MOPS buffer (0.1 M, pH 7.2, 0.15 M NaCl) increased the La encapsulation efficiency to 53.2%. The encapsulation efficiency of [²²⁵Ac]Ac³⁺ followed a similar trend to that of La: It was less than 5% for [²²⁵Ac]Ac³⁺ dispersed in DI H₂O; and it increased to 35.2 ± 8.1% after reconstituting [²²⁵Ac]Ac³⁺ in NH₄OAc solution (0.1 M, pH 6.0). Encapsulation of [²²⁵Ac]Ac-macropa complex was less than 2% within PLGA nanoparticles. Chelation of [²²⁵Ac]Ac³⁺ to BLPhen ligand increased the encapsulation efficiency (> 20%) relative to free [²²⁵Ac]Ac³⁺ in DI H₂O and [²²⁵Ac]Ac-macropa complex. The encapsulation of [²²⁵Ac]Ac³⁺ in a MeOH/DI H₂O mixture was evaluated as a control for the encapsulation of [²²⁵Ac]AcBLPhen complex. The encapsulation efficiency of [²²⁵Ac]Ac³⁺ in a MeOH/DI H₂O mixture was higher than that of [²²⁵Ac]AcBLPhen for the same water content (Fig. 4). Increasing the water content in the MeOH/DI H₂O mixture decreased the encapsulation efficiency within PLGA nanoparticles for both [²²⁵Ac]Ac³⁺ and [²²⁵Ac]AcBLPhen (Fig. 4).

Fig. 4

Encapsulation of [²²⁵Ac]Ac³⁺ within PLGA nanoparticles favors a hydrophobic environment. The encapsulation efficiency of [²²⁵Ac]Ac³⁺ within PLGA nanoparticles increased by decreasing the water content of the [²²⁵Ac]Ac³⁺ solution. The plot represents the encapsulation efficiency of [²²⁵Ac]Ac³⁺ within PLGA nanoparticles as free cations and when chelated to a BLPhen ligand, using a MeOH/DI H₂O mixture with different water contents. Values and error bars correspond to the mean and standard error of three replicates, respectively. *P < 0.05 one-way ANOVA followed by Tukey multiple comparisons post-test

The retention of [²²⁵Ac]Ac³⁺ and its decay daughters, [²²¹Fr]Fr⁺ and [²¹³Bi]Bi³⁺, within PLGA nanoparticles was assessed by measuring the fraction of radioactivity in the dialysate. Water/oil/water emulsions with [²²⁵Ac]Ac³⁺ in a MeOH/DI mixture, [²²⁵Ac]AcBLPhen, and [²²⁵Ac]Ac³⁺ in NH₄OAc as payload were transferred into dialysis cassettes and dialyzed against PBS. The water content in the MeOH/DI mixture used for [²²⁵Ac]Ac³⁺ and [²²⁵Ac]AcBLPhen was 5% (Fig. 4). Retention of [²²⁵Ac]Ac³⁺ within PLGA nanoparticles was influenced by the payload solution used during synthesis (Fig. 5a). PLGA nanoparticles encapsulating [²²⁵Ac]Ac³⁺ in NH₄OAc displayed a substantial release of [²²⁵Ac]Ac³⁺ (45.3 ± 2.5%) after 24 h in dialysis (Fig. 5a). A controlled release of [²²⁵Ac]Ac³⁺ was observed when using the [²²⁵Ac]Ac³⁺ in a MeOH/DI H₂O mixture as payload (Fig. 5a). This mixture resulted in a maximum of 53.7 ± 1.7% of [²²⁵Ac]Ac³⁺ found in the dialysate after 168 h (Fig. 5a). The release of [²²⁵Ac]Ac³⁺ was less than 2% when encapsulating [²²⁵Ac]AcBLPhen within PLGA nanoparticles (Fig. 5a). Release of ²²¹Fr was similar for PLGA nanoparticles encapsulating [²²⁵Ac]Ac³⁺ in NH₄OAc and [²²⁵Ac]Ac³⁺ in a MeOH/DI H₂O mixture, > 60% after 24 h in dialysis (Fig. 5b). The fraction of [²²¹Fr]Fr⁺ released from PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen varied between 32 and 35% over time (Fig. 5b). Similar results were obtained for [²¹³Bi]Bi³⁺: PLGA nanoparticles with [²²⁵Ac]Ac³⁺ in NH₄OAc and [²²⁵Ac]Ac³⁺ in a MeOH/DI H₂O mixture exhibited a greater release of [²¹³Bi]Bi³⁺ than did PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen (Fig. 5c).

Fig. 5

Enhanced retention of [²²⁵Ac]Ac³⁺ and its decay daughters, [²²¹Fr]Fr⁺ and [²¹³Bi]Bi⁺, was achieved by encapsulating [²²⁵Ac]AcBLPhen within PLGA nanoparticles. The fraction of a [²²⁵Ac]Ac³⁺, b [²²¹Fr]Fr⁺, and c [²¹³Bi]Bi³⁺ detected in the dialysate for PLGA nanoparticles encapsulating [²²⁵Ac]Ac³⁺ in a MeOH/DI mixture, [²²⁵Ac]Ac³⁺ in NH₄OAc solution, and [²²⁵Ac]AcBLPhen. Values and error bars correspond to the mean and standard error of three replicates, respectively. The release of [²²¹Fr]Fr⁺ and [²¹³Bi]Bi³⁺ considers the fraction of activity originating from [²²⁵Ac]Ac³⁺ in the dialysate as defined by the equations in the supporting information

The cytotoxic effect of PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen was evaluated against free [²²⁵Ac]Ac³⁺ and [²²⁵Ac]AcBLPhen, hereafter referred to as controls. E0771, MCF-7, and MDA-MB-231 breast cancer cells were exposed to all [²²⁵Ac]Ac³⁺ solutions for 24 h. Cell viability was assessed 1 h and 72 h after exposure to different ²²⁵Ac activities. At 1 h post-exposure, the viability of E0771 (Fig. 6a) and MDA-MB-231 (Fig. 6e) cells decreased significantly when using activities greater than 1.3 kBq/mL delivered within PLGA nanoparticles. For MCF-7 cells, no significant difference in cell viability was observed between controls and PLGA nanoparticles when using activities below 4.1 kBq/mL (Fig. 6c). At 72 h post-exposure, the viability of E0771 cells was not significantly different between controls and PLGA nanoparticles for activities below 2.3 kBq/mL (Fig. 6b). For MDA-MB-231 cells, exposure to activities greater than 2.3 kBq/mL was required to observe a significant difference in viability between PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen to controls (Fig. 6f). MCF-7 cell viability was not significantly different between PLGA nanoparticles and controls for the [²²⁵Ac]Ac³⁺ activities used in this work at 72 h post-exposure (Fig. 6d). Overall, it appears that the cell killing effect of [²²⁵Ac]Ac³⁺ is more pronounced when delivered to cells using the PLGA nanoparticles. The MDA-MB-231 and the E0771 cells showing significant cell death at 1.3 kBq/mL and 2.3 kBq/mL at 1 h. Persistent loss of viability was observed for these cell lines at doses above 2.3 kBq/mL at 72 h, while some recovery was noted at 1.3 kBq/mL. Interestingly, the MCF-7 cells showed significant cell death to PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen at 4.1 kBq/mL and 7.2 kBq/mL when exposed for 1 h and these cells appear to show some recovery at 72 h.

Fig. 6

PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen are cytotoxic to breast cancer cells. The viability of E0771, MCF-7, and MDA-MB-231 breast cancer cells after exposure for 24 h to free [²²⁵Ac]Ac³⁺, [²²⁵Ac]AcBLPhen, and PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen. Cell viability relative to untreated cells was assessed (a, c, e) 1 h and (b, d, f) 72 h after exposure to [²²⁵Ac]Ac³⁺ using alamarBlue. Data is presented as the mean and standard deviation for at least three biological replicates. **P < 0.01 (relative to both free [²²⁵Ac]Ac³⁺ and [²²⁵Ac]AcBLPhen) one-way ANOVA followed by Tukey multiple comparisons post-test

Delivering similar [²²⁵Ac]Ac³⁺ activities with different concentrations of PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen (e.g., 10 mg/mL, 20 mg/mL, and 40 mg/mL) did not significantly affect cell viability (Fig. 7). The viability of all cell lines exposed to PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen was different from that of free [²²⁵Ac]Ac³⁺ when using activities greater than 0.8 kBq/mL (Fig. 7). E0771 were more sensitive to free [²²⁵Ac]Ac³⁺ and PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen relative to MCF-7 and MDA-MB-231 cells. E0771 cell viability decreased below 80% and 40% after exposure to activities greater than 0.4 kBq/mL delivered as free [²²⁵Ac]Ac³⁺ and [²²⁵Ac]AcBLPhen encapsulated within PLGA nanoparticles, respectively (Fig. 7a). As shown in Fig. S4, E0771 cells had lower internalization and uptake of fluorescent PLGA-Cy5 nanoparticles relative to that of MDA-MB-231 cells (Fig. S5). The damage induced by α-particles was evaluated by measuring γ-H2AX foci, a marker of DNA double-strand breaks (Fig. S6). Cells were exposed to free [²²⁵Ac]Ac³⁺ and PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen for 24 h. When using free [²²⁵Ac]Ac³⁺, E0771 cells exhibited more γ-H2AX foci indicating that they are more susceptible to DNA damage than MCF-7 and MDA-MB-231 cells (Figs. S7a, S8a, and S9a). For E0771 cells, the number of foci per nucleus increased with greater activities of free [²²⁵Ac]Ac³⁺ (Figs. S6 and S7a). For PLGA nanoparticles, this trend was not as significant compared to that of free [²²⁵Ac]Ac³⁺ because the number of viable cells decreased drastically at the highest [²²⁵Ac]Ac³⁺ activities (Fig. 6). The effect of exposure time on E0771 and MCF-7 cell viability was evaluated with free [²²⁵Ac]Ac³⁺ and PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen. Viability analysis 24 h post-exposure showed similar viability of E0771 cells for the different activities and incubation times when using free [²²⁵Ac]Ac³⁺ (Fig. S10a). E0771 cells exposed to free [²²⁵Ac]Ac³⁺ (> 2 kBq/mL) for 24 h showed a slight decrease in viability when evaluated after 48 h (Fig. S10b). No appreciable effects were noted for shorter exposures to free [²²⁵Ac]Ac³⁺ under these conditions (Fig. S10b). By comparison the incubation of E0771 cells with PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen for 24 h shows a clear impact to cell viability: a reduction of viability to 43.7% and 34.7% when exposed to 1.6 kBq/mL at 24 h and 48 h post-exposure, respectively (Fig. S10C and S10D). Exposure to 3.1 kBq/mL for 24 h decreased E0771 cell viability to 8.9% and 4.3% at each time post-exposure (Figs. S10c and S10d).

Fig. 7

A similar cytotoxic effect was observed when [²²⁵Ac]Ac³⁺ was delivered with different concentrations of PLGA. Delivering [²²⁵Ac]Ac³⁺ with different concentrations of PLGA nanoparticles did not influence its cytotoxic effect on a E0771, b MCF-7, and c MDA-MB-231 cells. Cell viability after exposure to [²²⁵Ac]Ac³⁺ in PBS and PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen at 10 mg/mL, 20 mg/mL, and 40 mg/mL for 24 h. Cell viability was assessed 1 h post-exposure to [²²⁵Ac]Ac³⁺ using alamarBlue assay relative to untreated cells. Cells were exposed to free [²²⁵Ac]Ac³⁺ (0.6 kBq/mL, 1.1 kBq/mL, and 2.3 kBq/mL), PLGA([²²⁵Ac]AcBLPhen) [10 mg/mL] (0.4 kBq/mL, 0.8 kBq/mL, and 1.7 kBq/mL), PLGA([²²⁵Ac]AcBLPhen) [20 mg/mL] (0.4 kBq/mL, 0.8 kBq/mL, and 1.7 kBq/mL), and PLGA([²²⁵Ac]AcBLPhen) [40 mg/mL] (0.5 kBq/mL, 0.9 kBq/mL, and 1.8 kBq/mL). Reported values correspond to the mean of 12 technical replicates and n = 1 experiment. Error bars show the relative error

For MCF-7, little change in viability was observed when assessing it 24 h post-exposure to free [²²⁵Ac]Ac³⁺ for all doses and times (Fig. S11a). At 48 h, free [²²⁵Ac]Ac³⁺ resulted in cytotoxic effects when using 4.0 kBq/mL with viability dropping to 67.2% (Fig. S11b). In contrast, the viability of MCF-7 cells decreased significantly after incubation with PLGA nanoparticles at 3.1 kBq/mL (Fig. S11c and S11d). At 24 h and 48 h, a decrease in viability greater than 75% was observed in cells exposed to highest activity (3.1 kBq/mL) for 24 h. A lower activity of 1.6 kBq/mL showed a further decrease in MCF-7 cell viability when tested at 48 h (54.4%) from that observed 24 h post-treatment (78.8%) (Fig. S11c and S11d).

Discussion

A double-emulsion solvent evaporation method was adapted for the synthesis of PLGA nanoparticles in a radiological laboratory. The main adaptations include using small reagent volumes and a cup-horn sonicator to generate emulsions. Spherical PLGA nanoparticles are obtained with this method and their mean size and size distribution can be tailored by adjusting various parameters (Fig. S2). Of these parameters, the solvent removal method and the emulsifier concentration during sonication had a significant impact on the distribution and mean size of PLGA nanoparticles, respectively (Fig. 2b and S2b). As a solvent removal method, dialysis was also used for removal of free radionuclides and facilitate buffer exchange into PBS. It is expected that the sample will achieve a purification factor equal to the volume ratio between sample and dialysate (approximately 1:180–200). Additionally, the diffusion rate through the membrane is affected by dialysis parameters including temperature and stirring in addition to molecule/contaminant characteristics such as concentration and molecular weight. Because dialysis was conducted at room temperature with constant stirring (120 rpm), it is anticipated that free radionuclides—having a small molecular weight and a relatively low concentration—will diffuse readily through the dialysis membrane. Additionally, the type of dialysis device used—Slide-A-Lyzer cassettes—will also improve the diffusion rate of radionuclides and organic solvents. The enhancement of diffusion rate is related with the flat configuration of the dialysis cassettes, which facilitates that the molecules are close to the membrane. Based on these factors, 4-h dialysis with constant stirring at room temperature is sufficient time for the diffusion of radionuclides and organic solvents from the sample to the dialysate. Low concentration of emulsifier is not desirable due to a large mean hydrodynamic size (Z_ave > 1000 nm), broad size distribution, and incomplete particle formation (Fig. S1d and S1f). The increase in size and partial formation of PLGA nanoparticles was due to a lower fraction of emulsifier molecules than that required for controlled particle growth. Increasing the concentration of emulsifier contributes to a smaller size and more uniform size distribution (Fig. S2b). These characteristics are required for translation of these PLGA nanoparticles into future in vivo studies [36, 37]. Adaptation of a double-emulsion solvent evaporation method for radiological synthesis of PLGA nanoparticles resulted in particles with a similar size distribution to those obtained previously [49, 50]. PLGA nanoparticles were found to be stable in DI H₂O, PBS, and DMEM/F12 complete media over multiple ²²⁵Ac half-lives (Fig. S3). Only incubation in PBS resulted in appreciable broadening of the size distribution after 14 days (Fig. S3b). A slight broadening of the main hydrodynamic size peak was observed for the nanoparticles incubated with complete media containing 10% serum relative to those in DI H₂O (Fig. S3c). No change in the peak shape was observed for PLGA nanoparticles diluted in DI H₂O or complete cell culture media over the time course of the experiment. The presence of vitamin E TPGS may contribute to the stability of PLGA nanoparticles in the different solvents as well as prevents the surface adsorption of proteins from DMEM/F12 complete media.

Experiments using La³⁺ as a surrogate for [²²⁵Ac]Ac³⁺ showed that encapsulation of free La³⁺ is not suitable for evaluation in cell culture studies (< 2%). Reconstituting La cations with MOPS buffer significantly increased the encapsulation of La³⁺ within PLGA nanoparticles, as evidenced in our previous work [49]. Building from these results, the encapsulation of [²²⁵Ac]Ac³⁺ in PLGA nanoparticles was explored as (1) a free cation; (2) reconstituted in NH₄OAc, and (3) chelated to a ligand. Encapsulation of free [²²⁵Ac]Ac³⁺ in DI H₂O was < 5% because [²²⁵Ac]Ac³⁺ remained in the aqueous phase and was thus excluded from the hydrophobic environment of PLGA nanoparticles. Reconstituting [²²⁵Ac]Ac³⁺ with NH₄OAc increased its encapsulation efficiency to 59.4 ± 3.1% owing to the formation of [²²⁵Ac]Ac-acetate complexes. Chelation of [²²⁵Ac]Ac³⁺ to the hydrophilic macropa chelator (LogP = 0.222 ± 0.7 calculated using Advanced Chemistry Development (ACD/Labs) Software V11.02) [73] did not increase the encapsulation efficiency compared to that of free [²²⁵Ac]Ac³⁺ in DI H₂O. Like free [²²⁵Ac]Ac³⁺, [²²⁵Ac]Ac-macropa complexes had poor encapsulation efficiency because of the hydrophilic nature of the complex. Chelation of [²²⁵Ac]Ac³⁺ by a lipophilic BLPhen ligand (LogP = 10.3 ± 1.3 calculated using Advanced Chemistry Development (ACD/Labs) Software V11.02) [73] was therefore investigated as an alternative to improve the encapsulation efficiency within PLGA nanoparticles. Successful chelation of [²²⁵Ac]Ac³⁺ by lipophilic BLPhen ligand (> 90%) was achieved at room temperature for ligand concentrations greater than 5 mM (Fig. 3c). This is the first instance of a BLPhen ligand binding [²²⁵Ac]Ac³⁺ and opens the potential of this ligand for TAT. Lower BLPhen ligand concentrations caused a drop in chelation efficiency. This decrease in efficiency may be associated with the rigidity of the BLPhen ligand, which has been shown to impact the affinity of the BLPhen ligand for lanthanide cations [46, 47]. Ligand-lanthanide complexes with both 1:1 and 2:1 stoichiometries have been observed in the organic phase for lipophilic BLPhen ligand [46]. Computational analyses also showed that BLPhen coordinates to [²²⁵Ac]Ac³⁺ in a tetradentate fashion to form a 2:1 complex by exploiting a rigid and preorganized binding pocket. It is anticipated that modifying the BLPhen ligand structure will alter its rigidity which may increase its chelation efficiency for [²²⁵Ac]Ac³⁺ at lower ligand concentrations.

The lipophilic nature of the [²²⁵Ac]AcBLPhen complex was expected to: (i) decrease its water partitioning during nanoparticle synthesis, (ii) improve the encapsulation of [²²⁵Ac]Ac³⁺ within PLGA nanoparticles, and (iii) minimize [²²⁵Ac]Ac³⁺ release from PLGA nanoparticles. An encapsulation efficiency of 49.3 ± 12.4% was obtained with [²²⁵Ac]AcBLPhen when the water content in the mixture was 5% (Fig. 4). The encapsulation efficiency of [²²⁵Ac]Ac³⁺ in a DI H₂O/MeOH mixture was studied as a control for the lipophilic BLPhen ligand. Actinium-225 in this mixture had a higher encapsulation efficiency within PLGA nanoparticles relative to that of the lipophilic BLPhen ligand (Fig. 4). Increasing the water content in the [²²⁵Ac]Ac³⁺ DI H₂O/MeOH mixture to 25% resulted in a decrease in encapsulation efficiency to 48.8 ± 11.6% (Fig. 4). The lower [²²⁵Ac]Ac³⁺ encapsulation efficiency paralleled an increase in water content in the payload mixture, which is associated with a decrease in miscibility with the organic phase [74]. It is hypothesized that transmetalation of ²²⁵Ac during emulsification causes the lower encapsulation efficiency for [²²⁵Ac]AcBLPhen with respect to [²²⁵Ac]Ac³⁺ in DI H₂O/MeOH. The transmetalation of ²²⁵Ac during emulsification could be caused by a lack of stability of the [²²⁵Ac]AcBLPhen and unfavorable reaction conditions. The reaction conditions could also promote the precipitation of the BLPhen complexes and, thus, the transmetalation of ²²⁵Ac. The encapsulation efficiency of [²²⁵Ac]Ac³⁺ within PLGA nanoparticles was higher than that reported for liposomes encapsulating [²²⁵Ac]Ac-DOTA (< 10%) [22, 23]. Polymersomes passively loaded with ²²⁵Ac using ionophore- or tropolone-mediated mechanisms reported a higher encapsulation efficiency (> 60%) than those obtained in this work [26, 28]. Additionally, similar encapsulation efficiency was obtained by loading [¹⁷⁷Lu]Lu-DOTA-TATE into preformed PLGA nanoparticles (> 60%) [43, 44]. Overall, active encapsulation of [²²⁵Ac]Ac³⁺ within PLGA nanoparticles yielded similar encapsulation efficiencies to those obtained with passive encapsulation into preformed polymeric nanoparticles.

Retention of [²²⁵Ac]Ac³⁺ within PLGA nanoparticles was enhanced by its chelation to a BLPhen lipophilic ligand and by the hydrophobic composition of the payload solution (Fig. 5A). The [²²⁵Ac]Ac³⁺ reconstituted with NH₄OAc experienced a burst release from PLGA nanoparticles due to the combined effects of PLGA nanoparticle hydrolysis, surface adsorption of [²²⁵Ac]Ac³⁺, and the rapid diffusion of encapsulated [²²⁵Ac]Ac³⁺ through the polymeric matrix into the dialysate [75]. A controlled release of [²²⁵Ac]Ac³⁺ from PLGA nanoparticles was obtained after encapsulation of free [²²⁵Ac]Ac³⁺ in a MeOH/DI H₂O mixture. The extent of hydrophobicity within the PLGA nanoparticles—determined by the amount of water present in the [²²⁵Ac]Ac³⁺ solvent—gives rise to differences in release rates. A lower water content may alter the rate of diffusion through water pores and thus the rate of PLGA hydrolysis [75]. PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen retained 98% of [²²⁵Ac]Ac³⁺ over time (Fig. 5a). The enhanced retention of [²²⁵Ac]Ac was due to the stability and lipophilic nature of the [²²⁵Ac]AcBLPhen complex and the low water content within these PLGA nanoparticles. The lack of a burst release from [²²⁵Ac]Ac³⁺ suggests that the [²²⁵Ac]AcBLPhen complex is not adsorbed onto the outer particle surface during synthesis [43]. It is expected that the release of [²²⁵Ac]Ac³⁺ may increase when challenged with biologically relevant media like serum or cerebrospinal fluid under dynamic in vivo conditions [45, 76]. The release of radionuclides in biologically relevant media may also depend on the labeling conditions and the transchelation of radionuclides by proteins and competing cations [45, 77].However, as observed in this study the cytotoxic effect of PLGA([²²⁵Ac]AcBLPhen) nanoparticles incubated with breast cancer cells in the presence of 10% serum was not adversely impacted when compared to [²²⁵Ac]Ac³⁺ alone. Overall, the retention of [²²⁵Ac]Ac³⁺ within PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen was similar to that obtained with zwitterionic liposomes (88% over 30 days) and polymersomes (> 92% after 48 h) [22, 23, 26, 28].A high release of decay daughters, [²²¹Fr]Fr⁺ and [²¹³Bi]Bi³⁺, from PLGA nanoparticles was obtained after encapsulation of [²²⁵Ac]Ac³⁺ in a MeOH/DI H₂O mixture and in an NH₄OAc solution (Fig. 5b and 5c). These results were expected owing to the relationship between particle size and the retention of decay daughters within polymeric nanoparticles [22, 23, 26, 28]. The adsorption of [²²⁵Ac]Ac³⁺ onto the nanoparticle surface combined with an increased rate of PLGA hydrolysis would in turn increase the release of decay daughters into the dialysate [26]. The adsorption and implantation of decay daughters into adjacent PLGA nanoparticles was not significant under the experimental conditions tested (i.e., the nanoparticle concentration was approximately 8 mg/mL within the dialysis cassette) [78]. It is hypothesized that the implantation into adjacent nanoparticles is negligible because of the low electron density and amorphous structure of PLGA nanoparticles [79]. Encapsulating [²²⁵Ac]AcBLPhen within PLGA nanoparticles decreased the release of [²²¹Fr]Fr⁺ and [²¹³Bi]Bi³⁺ into the dialysate to 33% and 44%, respectively. The similar encapsulation efficiency of [²²¹Fr]Fr⁺ and [²¹³Bi]Bi³⁺ could be attributed to their chemical properties and origin. The transport of [²¹³Bi]Bi³⁺ through the membrane is potentially slower than that of [²²¹Fr]Fr⁺, whereas the majority of [²¹³Bi]Bi³⁺ found in the dialysate may originate from [²²¹Fr]Fr⁺.³³ The enhanced retention of [²²¹Fr]Fr⁺ and [²¹³Bi]Bi³⁺ within PLGA nanoparticles may be explained by a random distribution of [²²⁵Ac]AcBLPhen within each nanoparticle [26], which was likely achieved as a result of the active loading during synthesis. Future research will evaluate the implantation and potential complexation of decay daughters by adjacent PLGA nanoparticles and BLPhen ligands following the experimental setup used by Kozempel et al. [78].

Exposure of MCF-7, MDA-MB-231, and E0771 cells to PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen significantly decreased their viability compared with controls (Fig. 6). The high cytotoxic effect of PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen is associated with their spatial distribution within the well and their uptake into cells. The accumulation of PLGA([²²⁵Ac]AcBLPhen) around cancer cells will increase the dose delivered and, thus, their cytotoxicity. The quantitative retention of [²²⁵Ac]Ac³⁺ and the partial retention of decay daughters within PLGA nanoparticles is enhanced by the cellular uptake of PLGA nanoparticles, increasing their cell killing potential [18]. Fluorescent PLGA-Cy5 nanoparticles accumulated around E0771 and MDA-MB-231 cells and were internalized into the cytoplasm after 2 h of incubation (Figs. S4 and S5). Noticeably, MDA-MB-231 cells appeared to internalize the PLGA nanoparticles to a greater extent than E0771 cells. Accumulation of nanoparticles around and within the cancer cells increases the resultant DNA damage. DNA damage was determined by measuring γ-H2AX foci a marker of double-strand breaks caused, in this study, by α-particles. Higher concentrations of [²²⁵Ac]Ac³⁺ increased the number of γ-H2AX foci per nucleus for both [²²⁵Ac]Ac³⁺ and PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen. This trend was clearly observed for free [²²⁵Ac]Ac³⁺ and to a lesser extend with PLGA nanoparticles. The differences in viability and DNA damage between cells may be associated with their sensitivity to α-particles [67] and the rate of PLGA nanoparticle uptake and internalization [80]. Additionally, counting of γ-H2AX foci was challenging with PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen because there was a lower fraction of viable cells available. These remaining cells likely had lower amounts of DNA damage which allowed them to remain adhered to the tissue culture surface.

The concentration of PLGA nanoparticles in solution did not induce cytotoxic effects in the human and murine breast cancer cells confirming that PLGA is a biocompatible delivery vehicle (Fig. 7) [39]. In this study, the relative fraction of [²²⁵Ac]AcBLPhen within PLGA nanoparticles determines the cytotoxic effect since the amount of fluorescent PLGA-Cy5 nanoparticles internalized at low and high concentrations was not significantly different for a given cell line (Fig. S4 and S5). It is anticipated that both free [²²⁵Ac]Ac³⁺ and [²²⁵Ac]AcBLPhen will remain in solution and uniformly distributed over the well volume, decreasing the dose delivered to the cells which are attached to the bottom of the plate. The exposure time to PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen correlated with increasing cell death because the cumulative dose delivered to the cells increased over time. Exposure to the highest activity [²²⁵Ac]AcBLPhen within PLGA nanoparticles, for the longest incubation time, resulted in the greatest decrease in cell viability for both E0771 and MCF-7 (Figs. S10 and S11). PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen have the potential to treat solid tumors and micrometastases by either intratumoral injection or active targeting after modifying PLGA nanoparticles with targeting vectors that recognize unique or overexpressed receptors on cancer cells.

Conclusions

PLGA nanoparticles were evaluated as a biocompatible delivery platform for [²²⁵Ac]Ac³⁺ in TAT. A double-emulsion solvent evaporation method was adapted for radiological synthesis, decreasing the generation of radioactive waste, and mitigating radioactive contamination because of the aerosolization of radionuclides. The adapted synthesis method resulted in PLGA nanoparticles with a mean particle size of ~ 150 nm when standard conditions were used. The size and size distribution are easily tailored by adjusting various synthesis parameters. Encapsulation of free [²²⁵Ac]Ac³⁺ and [²²⁵Ac]Ac-macropa was low because of their hydrophilic nature. A high encapsulation efficiency of free [²²⁵Ac]Ac³⁺ in DI/MeOH mixture and in NH₄OAc solution within PLGA nanoparticles was achieved by tailoring the hydrophobicity of the payload mixture and chelating [²²⁵Ac]Ac³⁺ with a lipophilic ligand. Encapsulating free [²²⁵Ac]Ac³⁺ within PLGA nanoparticles resulted in a burst release of [²²⁵Ac]Ac³⁺ and a low retention of its decay daughters. Chelation of [²²⁵Ac]Ac³⁺ to a lipophilic BLPhen ligand enhanced the retention of [²²⁵Ac]Ac³⁺ and its decay daughters within PLGA nanoparticles. The enhanced radionuclide retention combined with the targeted delivery of [²²⁵Ac]Ac³⁺ to cells—facilitated by nanoparticle deposition on the cell surface and subsequent internalization by breast cancer cells—resulted in an increased cytotoxic effect of PLGA nanoparticles encapsulating [²²⁵Ac]AcBLPhen relative to controls. These results demonstrate that PLGA nanoparticles are a promising delivery platform for [²²⁵Ac]Ac³⁺ in TAT. This study indicates that PLGA nanoparticles have the potential to be combined with other ligands that are not suitable for in vivo delivery of radionuclides owing to their low binding affinity, poor solubility in water, or low stability in biological environments. Therefore, PLGA nanoparticles offer a new opportunity to repurpose these ligands for TAT with the synergistic benefit of enhancing encapsulation and retention of radionuclides within a biocompatible delivery platform. Further optimization of size distribution, radionuclide encapsulation, and bioconjugation is required for successful translation of radioactive PLGA nanoparticles to pre-clinical experiments.