Potassium Iodide Nanoparticles Enhance Radiotherapy against Breast Cancer by Exploiting the Sodium-Iodide Symporter

Abstract

Iodine has shown promise in enhancing radiotherapy. However, conventional iodine compounds show fast clearance and low retention inside cancer cells, limiting their application as a radiosensitizer. Herein, we synthesize poly(maleic anhydride-alt-1-octadecene) coated KI nanoparticles (PMAO-KI NPs) and evaluate their potential for enhancing radiotherapy. Owing to the polymer coating, the KI core of PMAO-KI NPs is not instantly dissolved in aqueous solutions but slowly degraded, allowing for controlled release of iodide (I⁻). I⁻ is transported into cells via the sodium iodide symporter (NIS), which is upregulated in breast cancer cells. Our results show that PMAO-KI NPs can enhance radiation-induced production of reactive oxygen species such as hydroxyl radicals. When tested in vitro with MCF-7 cells, PMAO-KI NPs promote radiation-induced DNA double-strand breaks and lipid peroxidation, causing a drop in cancer cell viability and reproductivity. When tested in MCF-7 bearing mice, PMAO-KI NPs show significant radiosensitizing effects, leading to complete tumor eradication in 80% of the treated animals without inducing additional toxicity. Overall, our strategy exploits electrolyte nanoparticles to deliver iodide into cancer cells through NIS, thus promoting radiotherapy against breast cancer.

Graphical Abstract

Breast cancer is the most commonly diagnosed cancer among women and the second most common cause of cancer mortality among women.¹ Radiotherapy (RT) plays an integral role in breast cancer therapy. RT can be applied after lumpectomy or mastectomy to reduce tumor recurrence,^2–3 or given in a neoadjuvant setting for managing locally advanced breast cancer.⁴ However, the dose and efficacy of RT is limited by normal tissue toxicity. To improve treatment outcomes, radiosensitizers, agents that can enhance RT efficacy at the same irradiation dose, may be applied during or after radiation. Conventional radiosensitizers are chemotherapeutics such as paclitaxel, cyclophosphamide, and fluorouracil.⁵ More recently, high atomic number or high-Z nanoparticles (NPs) made of gold,⁶ hafnium oxide,⁷ and metal-organic frameworks (MOFs)⁸ are tested as radiosensitizers and show promising results. High-Z NPs afford large cross-section for high-energy photons thus increasing energy deposition in tumors. This results in elevated production of toxic reactive oxygen species (ROS) that kill cancer cells.^9–11 However, most high-Z NPs have relatively large sizes, restraining their ability to diffuse within the dense tumor tissues and their uptake by cancer cells.¹² This restriction may negatively affect radiosensitizing effects as ROS have a short effective range (10–400 nm) and limited ability to penetrate biological barriers such as the plasma membrane.¹³

In addition to high-Z NPs made of metals or metal oxides, iodine compounds have shown radiosensitizing benefits. As early as in the 50s and 60s, it was found that iodine compounds such as iodoacetate, iodoacetamide, and potassium iodide could enhance radiative lethality against bacteria and mammalian cells.^14–16 Recent studies suggest that iodinated computed tomography (CT) contrast agents, such as Iomeprol, Iopromide and Iomeron,^17–19 can increase radiation-induced cancer cell death. This discovery has led to clinical studies examining iodine-enhanced RT for brain tumor.⁸ However, iodinated contrast agents have very short circulation half-lives and low tumor retention.⁸ While measures such as infusion can be taken to improve tumor concentration of iodine at the time of irradiation, the efficacy is still limited by low accumulation of iodine inside cancer cells.^20–21

Herein, we report a potassium iodide (KI) NP-based radiosensitizer, which employs the Na⁺/I⁻ symporter (NIS) for iodine delivery and radiosensitization. NIS is a transmembrane protein that co-transports two Na⁺ and an I⁻ across the cellular membrane. NIS is mainly expressed on thyroid cells for uptake and incorporation of iodine for hormone synthesis.^22–24 Recent analysis showed that NIS is also positive in 76–87% of human invasive breast cancers, including triple negative cases.^25–26 We postulate that I⁻ ions can enter cancer cells via NIS and effectively sensitize them to RT. One challenge, however, is that iodide salts have a high water mobility and would be quickly cleared from the injection site. To address the issue, we synthesized KI NPs and coated them with poly(maleic anhydride-alt-1-octadecene) (PMAO). The polymer coating extends the half-lives of the electrolyte NPs to ~24 hrs in aqueous solutions. We hypothesize that these NPs can be injected into a breast tumor to enable sustained iodide release and cellular uptake, thereby improving RT efficacy. We tested this hypothesis first in vitro with MCF-7 cells,^{24 27} and then in vivo in a MCF-7 xenograft model. To the best of our knowledge, bio-applications of KI NPs have rarely been investigated, and their potential as a radiosensitizer has never been explored.

Results and Discussion

For nanoparticle synthesis, potassium oleate reacts with I₂ in octadecene in the presence of oleylamine at room temperature. It is postulated that oleate and oleylamine function in collaboration to provide reverse micelle structures, within which KI nucleates and forms particles.²⁸ Oleylamine also functions as a mild reducing agent to convert I₂ to I⁻.²⁹ Typical reactions yielded monodisperse KI NPs with a diameter of 79.8 ± 1.7 nm based on transmission electron microscopy (TEM, Figure 1a). Dynamic light scattering (DLS) found a similar hydrodynamic size (Figure 1b). By adjusting precursor amounts of potassium oleate and I₂, KI NPs in the range of 20–500 nm can be produced (Figure S1). X-Ray diffraction (XRD), as well as scanning electron microscopy (SEM) with energy dispersive X-ray analysis (EDXA), confirmed that the particles were KI in composition (Figure 1c, d).

Figure 1. Physicochemical characterizations of as-synthesized KI NPs.

a) TEM images of as-synthesized KI NPs. Scale bar, 500 nm. Insert scale bar, 100 nm. b) DLS result of KI NPs in hexane. c) XRD spectra of KI NPs and KI (PDF #04-005-6718). d) EDS spectrum and element weight percentages of KI NPs.

KI NPs were then coated with a layer of PMAO to be rendered with water solubility (Scheme 1a, Figure 2a). PMAO was selected as the coating for a variety of reasons, including its biodegradability, biocompatibility, and readiness to be coupled with other molecules.³⁰ Briefly, KI NPs were mixed with PMAO and PEG-NH₂ in chloroform and, after removing the solvent by rotovap, bis-hexamethyl triamine in borate buffer (pH 8.2) was added. The hydrophobic alkyl chains of PMAO interacted with the oleate/oleylamine coating.³¹ The anhydride rings opened in the basic solution and reacted with primary amine, resulting in polymer crosslinking and PEGylation (Scheme 1a). The amount of PMAO can be tuned to adjust the final coating thickness. The resulting NPs are readily dispersed in water. Fourier-transform infrared spectroscopy (FT-IR) found OH and NH peaks, confirming successful PEGylation (Figure 2b). The zeta potential of the PMAO-KI NPs was −23.4 mV, which is attributed to multiple surface carboxyl groups generated from anhydride ring opening (Figure 2c).

Scheme 1. Schematic representation of PMAO-KI NP synthesis, particle delivery, and radiosensitizing.

a) Synthesis and surface coating of KI particles. KI NPs were synthesized in octadecene and then coated with a layer of PMAO to prevent fast degradation and I⁻ release. b) Mechanisms for PMAO-KI NPs based radiosensitization. I⁻ ions are slowly released from PMAO-KI NPs, and internalized by breast cancer cells through the NIS symporter. I⁻ promotes ROS generation under radiation, causing extensive DNA and lipid damage as well as reduced cell viability and clonogenicity.

Figure 2. Physical characterizations of PMAO-KI NPs.

a) TEM analysis of PMAO-KI NPs. Scale bars, Scale bar, 500 nm. Insert scale bar, 20 nm. b) FT-IR spectra of PMAO-KI NPs and PMAO polymer. c) Zeta potential of PMAO-KI NPs. Due to surface carboxyl groups, PMAO-KI NPs carry a negative surface charge. d) DLS of PMAO-KI NPs (in PBS) and uncoated KI NPs (in hexane). e) Iodide release, measured in PBS by iodide-selective electrode. f) Hydroxyl radical production, measured in PBS by APF assay. PMAO-KI NPs promoted hydroxyl radical generation at different radiation doses. *, p < 0.05.

The KI cores remain largely intact throughout the surface modification (Figure 2a), but slowly degraded in aqueous solutions. To investigate this, we loaded PMAO-KI NPs onto a dialysis cassette and evaluated iodide released from the NPs using an iodide-selective electrode. The time it took for half of the iodine to be released, or t_1/2, is negatively correlated with the coating thickness. We have settled on a formulation with a coating thickness of ~50 nm (Figure 2a, d) and a t_1/2 of ~24 hours (Figure 2e) for follow-up studies.

We then tested whether KI NPs can enhance radical production under beam irradiation. This was analyzed by subjecting PMAO-KI NP solutions to X-rays (5 Gy) and measuring radicals produced using chemical sensors including dihydroethidium (DHE), singlet oxygen sensor green (SOSG), and aminophenyl fluorescein (APF). These fluorogenic probes are selectively responsive to superoxide, singlet oxygen, and hydroxyl radicals, respectively. Hydroxyl radical, which is the most relevant in RT, showed a significant radiation dose dependent increase (Figure 2f). Superoxide and singlet oxygen levels, on the other hand, were not significantly elevated with PMAO-KI NPs (Figure S2).

We then tested PMAO-KI NPs in vitro with MCF-7 cells, which are NIS positive (Figure S3). There was no significant cytotoxicity when the concentration of PMAO-KI NPs was below 100 μg iodine/mL (Figure 3a). Next, we incubated MCF-7 cells with PMAO-KI NPs at 25 μg/mL for 24 h, and measured iodine contents in cells by inductively coupled plasma mass spectrometry (ICP-MS). Trans retinoic acid (tRA, 1 μM)^{32 33} was added into the incubation medium to promotes NIS expression (Figure 3b).^{26, 33} For comparison, cells treated with KClO₄, an NIS inhibitor, were also tested. Relative to untreated MCF-7 cells, incubation with PMAO-KI NPs doubled the intracellular iodine levels. The iodine uptake was further elevated when tRA was applied and reduced in the presence of KClO₄. Meanwhile, fluorescence microscopy (Figure 3c) flow cytometry (Figure 3d) and found insignificant difference in cellular uptake of Cy5-labeled PMAO-KI NPs between cells treated or not treated with tRA. These results indicate that a significant portion of iodine enters cells in the form of free iodide rather than intact NPs.

Figure 3. Cytotoxicity and cellular uptake.

, assessed with MCF-7 cells. a) MTT assay evaluating cytotoxicity of PMAO-KI NPs. Low cytotoxicity was observed when nanoparticle concentration was below 100 μg iodine/mL. b) Iodine uptake, measured by ICP-MS. Increased cellular iodine content was observed with PMAO-KI NPs; the uptake was further increased when cells were treated with tRA (1 μM) that promotes NIS expression. *, p < 0.05; ***, p < 0.001. c) Cell uptake of PMAO-KI NPs, examined by fluorescence microscopy. Cy5 labeled PMAO-KI NPs were incubated with MCF-7 cells. No significant difference in uptake was observed between cells treated with or without tRA. d) Cell uptake of PMAO-KI NPs, measured by flow cytometry. For comparison, tRA (1 μM) or KClO4 (1 μM) was co-incubated.

We next examined RT-induced DNA damage by anti-γH₂AX staining. Relative to radiation alone (RT, 5 Gy), radiation in the presence of PMAO-KI NPs increased γH₂AX foci number from 17.5 to 28.7 foci/cell. This number was increased further to 32.8 foci/cell when tRA was added (Figure 4a). Meanwhile, PMAO-KI NPs plus RT in the presence of tRA elevated lipid peroxidation level by 20.72%, compared to 5.68% for RT alone (Figure 4b). The increased DNA damage and lipoid peroxidation is attributed to an enhanced production of hydroxyl radical as a result of iodide accumulation in cells. Notably, previous studies show that iodide can accumulate in the lipid membrane through physical interaction,²⁴ which may contribute to enhanced lipid peroxidation.

Figure 4.

Radiosensitizing effects of PMAO-KI NPs, tested in vitro with MCF-7 cells. a) Anit-γH2AX staining results. Cells were incubated with PMAO-KI NPs (25 μg/mL), and irradiated by X-ray (5 Gy). Images were acquired at 24 h. Average foci number per cell was compared. *, p < 0.05. b) Lipid peroxidation, evaluated by BODIPY 581/591 assay. *, p < 0.05. c) Dose modifying effects, evaluated on the basis of MTT results. MCF-7 cells were incubated with PMAO-KI NPs (25 μg/mL, in the presence of tRA), and irradiated (5 Gy). Viability drop relative to cells receiving radiation only (without incubation with PMAO-KI NPs) at 24 or 72 hrs was computed and presented. *, p < 0.05. A dose dependent response is observed at both time points with 72 hrs showing a greater dose enhancement. d) Clonogenic assay results, performed in the presence or absence of 25 μg/mL PMAO-KI NPs. The results were fitted to the liner-quadratic formula, S(D)/S(0)=exp-(aD+bD²), where S is the survival fraction, D is the radiation dose in Gy, and a and b are fitting coefficients. ***, p < 0.001. e) Dose modifying factors, based on results from d. D₁₀, dose required for 10% survival. DMR_, dose modification ratio based on D₁₀.

The impact of PMAO-KI NPs on cell viability under radiation was assessed by MTT assays. In the presence of tRA, PMAO-KI NPs plus RT (5 Gy) reduced the 24-h cell viability by 34.66% relative to treatment without tRA (Figure S4). The benefits of PMAO-KI NPs were also confirmed by a dose-dependent decrease of cell viability at both 24 hrs and 72 hrs (Figure 4c). The radiosensitizing effects of PMAO-KI NPs were further analyzed by clonogenic assays. Compared to RT alone, PMAO-KI NPs plus RT significantly reduced colony formation at all tested doses (Figure S5). The survival fraction was 0.68 ± 0.10 for RT alone at 2 Gy, which was reduced to 0.22 ± 0.01 for PMAO-KI NPs plus RT. At 6 Gy, the survival fraction was only 0.0053 ± 0.0006 for PMAO-KI NPs plus RT, compared to 0.040 ± 0.015 for RT alone (Figure S5). When fitting the results into the linear-quadratic model (S(D)/S(0)=exp-(aD+bD²)), it was determined that the dose required for 10% survival (D₁₀) was 4.6126 for RT, 3.9399 for PMAO-KI NPs plus RT, and 2.6006 for PMAO-KI NPs plus RT in the presence of tRA (Figure 4d). This represents a dose modifying radio (DMR) of 1.7737 for the PMAO-KI NPs plus tRA combination.

Tumor retention of iodine was evaluated in vivo in MCF-7 tumor bearing mice. To this end, we synthesized ¹³¹I-doped PMAO-KI NPs (¹³¹I-PMAO-KI NPs, Supporting Information), and intratumorally (i.t.) injected them (~4.6 MBq) into animals (n=3). tRA was pre-administered into tumors to promote NIS expression.³³ For comparison, ¹³¹I-KI solution at the same dosimetry plus tRA (¹³¹I-KI solution+tRA) or ¹³¹I-PMAO-KI NPs alone (¹³¹I-PMAO-KI NPs) were i.t. injected. After 72 hrs, we euthanized the animals, harvested tumors and major organs, and subjected them to gamma counting to evaluate iodine bio-distribution. Relative to ¹³¹I-PMAO-KI NPs only, ¹³¹I-PMAO-KI NPs plus tRA (¹³¹I-PMAO-KI NPs+tRA) showed a significant increase in iodine tumor retention (Figure 5a,b). The tumor uptake was high compared to the ¹³¹I-KI solution+tRA group (Figure 5a,b), suggesting that controlled iodine release also contributed to improved tumor retention. In addition to tumors, significant radioactivities were found in the thyroid and stomach (Figure 5c), where NIS expression is also high.³⁴ Of note, stomach radioactivity was high in the ¹³¹I-PMAO-KI NPs+tRA group relative to the ¹³¹I-PMAO-KI NPs group, though difference was insignificant (p = 0.1179). This result suggests that some tRA may have entered blood circulation and promoted NIS expression in extratumoral tissues.

Figure 5. Biodistribution of iodine.

¹³¹I-PMAO-KI NPs were i.t. injected into MCF-7 tumor bearing mice that had been pre-administered with tRA (¹³¹I-PMAO-KI NPs+tRA; n=3). For comparison, animals were treated with ¹³¹I-KI solution and tRA (¹³¹I-KI solution+tRA; n=3), or ¹³¹I-PMAO-KI NPs only (¹³¹I-PMAO-KI NPs; n=3). Animals were euthanized after 72 hrs. Gamma counting was performed with dissected tumors and normal tissues to examine iodine distribution. a) Iodine remained in tumors. *, p < 0.05. b) Autoradiographs of tumor that had been treated with ¹³¹I-PMAO-KI NPs+tRA, ¹³¹I-KI solution+tRA, and ¹³¹I-PMAO-KI NPs. Images were displayed in pseudocolors based on the scale appearing to the right. c) Distribution of iodine in normal tissues. Due to NIS expression, radioactivities were found in the thyroid and stomach.

Lastly, we tested treatment efficacy in MCF-7 tumor bearing mice. Briefly, PMAO-KI NPs (50 μg iodine/mL, 50 μL) were i.t. injected, followed by radiation (5 Gy) applied to tumors (n=5). The rest of the animal body was protected by lead. For comparison, PBS, RT alone and PMAO-KI NPs alone were investigated (n=5). tRA was pre-administered into all animals to promote NIS expression. Animals in the PBS and PMAO-KI NPs groups showed rapid tumor progression (Figure 6a), resulting in a relatively short animal survival (32 and 50 days on average, respectively). RT only caused a moderate tumor suppression and a slightly extended animal survival (57 days, Figure 6b). However, 60% of the animals in the RT group had either died or met a humane endpoint by day 75. As a comparison, all animals in the PMAO-KI NPs plus RT (PMAO-KI NPs+RT) group experienced significant tumor regression. Eighty percent of the tumors became impalpable after two weeks, and 60% of the animals remained tumor-free after 60 days. All animals in the PMAO-KI NPs+RT group remained alive at the end of the study on Day 75 (Figure 6b). No mice in the group showed acute toxicity or significant body weight drop throughout the experiment (Figure S6).

Figure 6.

In vivo therapeutic studies, tested in a MCF7 xenograft model (n = 5). PMAO-KI NPs (50 μg iodine/mL in 50 μL PBS) were i.t. injected, followed by 5-Gy beam radiation applied to tumors (PMAO-KI NPs+RT). For comparison, PBS, radiation alone (RT), and PMAO-KI NPs alone (PMAO-KI NPs) were tested. tRA was pre-injected to promote NIS expression. a) Tumor growth curves. Relative to RT alone, PMAO-KI NPs+RT significantly improved tumor suppression. 80% of the treated mice remained tumor-free at the end of the experiment. *, p < 0.05. b) Animal survival. All animals remained alive in the PMAO-KI NPs+RT group. c) Ki-67 staining results.

Post-mortem histology was performed to validate treatment efficacy. H&E staining found large areas of nuclear shrinkage and cell death in the PMAO-KI NPs+RT group (Figure S7). Ki-67 staining also confirmed that the combination led to a significant reduction in cancer cell proliferation compared to PBS, RT alone and PMAO-KI NPs alone (Figure 6c). Meanwhile, H&E staining found no signs of adverse effects in major organ tissues, including the brain, heart, kidney, liver, lung and spleen (Figure S8).

Conclusions

In summary, we have successfully synthesized PMAO-KI NPs and investigated their role as a radiosensitizer for RT against breast cancer. We observed significant improvement in tumor suppression among animals receiving i.t. injection of PMAO-KI NPs before irradiation. As discussed above, iodine-containing molecules such as Iomeprol have short circulation half-lives and low tumor uptake. In our approach, NPs-enabled controlled release of iodide and NIS-mediated cell uptake of the ion is exploited, providing a solution to these issues. While iodine molecule (I₂) shows anti-proliferative effects,^35–36 we found little toxicity with iodide (Figure 3a), which agrees with previous reports.²⁴ The tumor suppression was attributed to iodide-based enhancement in RT-induced production of radicals and the resulting damage to the cellular components, not direct cancer cell killing by iodide.

In the clinic, breast cancer patients after mastectomy or lumpectomy often receive multi-session fractionated RT to prevent recurrence. It is envisioned that KI NPs can be injected into the tumor bed for sustained iodide release and radiosensitization. The current PMAO-KI formulation has t_1/2 of ~24 h. It is possible to increase the coating thickness and/or degree of crosslinking to extend iodide release so that one injection can benefit multiple RT sessions. Last but not least, radioisotope such as ¹³¹I and ¹²⁵I can be incorporated into PMAO-KI NPs, producing “hot” NPs that function as a radiopharmaceutical or a brachytherapy agent for cancer treatment. For these applications, PMAO-KI NPs will be systemically administered and rely on nanoparticles’ passive or active targeting to home to tumors.³⁸ More studies however are needed to evaluate whether the approach can mitigate the accumulation of radioactivities in the thyroid and other normal tissues.

Methods

Potassium iodide NP synthesis.

6.68 g of potassium oleate (40 wt. % paste in H₂O Sigma-Aldrich, 291242) was added to a round bottom flask with 200 mL 1-octadecene (Sigma-Aldrich, O806, 90%), and mixed at 290 °C until oleate was fully solubilized. Once the reaction is cool 10 mL of Oleylamine (Sigma-Aldrich, O7805, 70%) is added and mixed. After, 5 g of Iodine (Sigma-Aldrich, 207772) is added and the reaction is sealed and mixed overnight. Then 8 aliquots of NPs are centrifuged at 12096 × g for 10 mins. The particles are then washed three times in EtOH and placed in an oven until dry.

PMAO coating.

Previously synthesized KI NPs in 3 mL of chloroform was added to the mixture and sonicated into suspension. After which 120 mg Poly(maleic anhydride alt-1-octadecene) (Sigma-Aldrich, 776866) in 3 mL chloroform (Sigma-Aldrich, C2432) and 1 mL PEG-bis-amine (43 mg in 10 mL chloroform) (Sigma-Aldrich, P9906) were added. The reaction was then mixed for 2 hrs. The solution was then poured into a round bottom flask. The solvent was removed under a rotary evaporator and gentle heat (<40 °C) until dry. 1.5 mL bis-hexamethyl triamine (43 mg in 10 mL chloroform) (Fisher Scientific, B181425G) was added to the flask and sonicated until the film was resuspended. The solution was then placed under Rotovap again until dry at room temperature. 50 mM Borate buffer is then added to cover film and sonicated until film is fully in solution. After which the solution is filtered to remove excessive aggregates then centrifuged at 10,000 rpm for 10 minutes. After which and the supernatant is discarded and the precipitate dried below 120 °C to avoid polymer softening. The dried and sealed PMAO-KI NPs can be stored for at least 6 months with no discernable loss in quality.

Characterization of KI NPs.

KI in ethanol were dried on formvar TEM grids at 50 °C overnight for TEM and STEM imaging. TEM images were acquired on a 120 kV HT7830 and a 300 kV high resolution H9500 TEM (Clemson Electron Microscopy Facility). SEM images and elemental mapping were taken on a FEI Teneo operating at 15 kV for images and 30 kV for elemental mapping (Georgia Electron Microscopy Facility). For release profile determination, NPs both coated and uncoated were suspended in PBS and transferred to a molecular weight cutoff 10,000 kDa (Slide-A-Lyzer Thermo Scientific). The filtrate was sampled at multiple timepoints over 72 hrs and the residual collected as well. The iodine content was then calculated based on an iodine electrode (perfectION™ Mettler Toledo). IR spectra was taken from the dried powder of NPs on a Nicolet iS10 FT-IR spectrometer. Dynamic light scattering of both coated and uncoated particles was performed using a Malvern Zetasizer Nano S90 and zeta potential measured with the same.

ROS production.

Initial ROS testing was done using methylene blue trihydrate (MP Biomedicals) at a concentration of 5 mM with 25 μg/mL NPs. Irradiation was delivered via a Mini-X X-Ray Tube (Ampek) 50 keV at a dose rate of approximately 20 cGy/minute and would remain consistent throughout experiments. Absorbance change was measured via SynergyMX. A similar approach was used for APF (3’-(p-aminophenylfluorescein), Hydroxyl Radical, Hypochlorite or Peroxynitrite Sensor) (Invitrogen, 36003), DHE (Dihydroethidium (Hydroethidine), a superoxide indicator) (Invitrogen, D11347) and SOSG (Singlet Oxygen Sensor Green, Invitrogen, S36002) following the kit manufacturing instructions with 50 μg/mL NPs.

MTT assays.

A MTT assay (Thiazolyl Blue Tetrazolium Bromide, Sigma, M2128) was used to estimate the cytotoxicity of PMAO-KI NPs and phosopholipid (DSPE-PEG(2000) Amine) coated KI NPs on MCF-7 cells as well as radiosensitizer efficacy. Cells were seeded in 96 well-plates (Corning, 3599) with 10,000 cells/well. 24 hrs after cellular seeding, cells were treated with PMAO-KI NPs and KI solution of equivalent iodine content at various ‎concentrations (1 μg/mL to 400 μg/mL). The cellular media was replaced with media of the appropriate concentration of materials and incubated for a further 24 hours. MTT reagent was then added in the culture medium, followed by incubation at 37°C for 3h and the resulting purple crystal solubilized in DMSO. The absorbance was measured on a microplate reader (SynergyMx) at 590 nm, with the measurements performed in at least triplicate.

For X-ray therapy a similar protocol was followed. 24 hours after cellular seeding tRA in DMSO was added to the cell wells to a final tRA concentration of 1 μM, with 1% DMSO media serving as the control. After 24 hours for the NP positive groups PMAO-KI NPs were added at a final concentration of 25 μg/mL. After an additional 24 hrs incubation the NPs were irradiated using the previously described protocol at various doses. 24 hours and 72 hours later, depending on the trail, the MTT reagent was added and protocol proceeded as above.

Iodine accumulation and particle uptake in MCF-7 cells w/ or w/o tRA.

For iodine uptake study, MCF-7 cells were seeded in 6-well plate (Corning, 3516) at a density of 450,000 cells/mL. 24 hours later the media was replaced, and for tRA positive studies, cell culture medium with a concentration of 1 μM all trans-retinoic acid was added for 24 additional hours. The medium was replaced with a fresh cell medium containing either 0, 12, 25, or 50 μg/mL PMAO-KI NPs. X-ray therapy followed a similar protocol as the MTT experiment. Cells were trypsinized and lysed, and iodine was quantified using the Inductively Coupled Plasma Mass Spectroscopy (ICP-MS) at the Center for Applied Isotope Studies, University of Georgia. For particle uptake studies, MCF-7 cells were seeded in an 8-well Nunc™ Lab-Tek™ II Chamber Slide™ System (Thermo Fisher, 154534PK) at the same density as the iodine ICP study. The PMAO-KI NPs were functionalized with Cy5 dye and the cell nuclei stained with DAPI. Images were taken using a Brightfield Microscope (OLYMPUS TH4-100, Japan). NP uptake quantification was analyzed based on the Cy5 channel fluorescence intensities in the images.

Clonogenic assay.

MCF-7 cells were seeded into 6 well plates at a density of either 500,000 cells/mL depending on treatment. After seeding the cells were incubated for 24 hrs before NIS induction as in the MTT assay for the tRA positive groups. 24 hours after tRA addition NPs were added a concentration 50 μg/mL for the NP therapy groups and incubated for an additional 24 hrs. The medium was then replaced, and the groups were irradiated at 0, 2, 4, or 6 Gy before being plated at a cellular density of 500, 1000, 10,000, 50,000, or 500,000 cells/mL respectively. Colonies were stained with crystal violet and CFU/cells were then calculated after about 12 days. These groups were performed in triplicate. Efficacy was calculated through $PlatingEfficiency=\frac{\#ofcoloniesformed}{\#ofcellsseeded}\times 100\%$, $SurvivalFraction=\frac{\#coloniesformedafterirradiation}{\#ofcellsseeded\times PE}$. The results were fitted into a linear-quadratic equation (S(D)/S(0)=exp-(aD+bD²)), where SF(D)=Survival Fraction at dose D, $\left[1-{(1-e^{\frac{-D}{D_0}})}^n\right]$, n=1 for a single dose, D₀ = the dose to reduce cell survival to 37% of its value at any point on the final near exponential part of the curve. Dose modification radio at 10% survival ${DMR}_{10\%}=\frac{D_{10\%,Cont}}{D_{10\%,NPs}}$, where D_10%=Does required for 10% survival.

Lipid peroxidation.

Lipid peroxidation was measured using an Image-iT™ Lipid Peroxidation Kit (Invitrogen, C10445) for live cell analysis. MCF-7 cells were preincubated with 1 μM tRA for tRA positive cells and peroxidation was measured at 5 Gy with 25 μg/mL PMAO-KI NPs for NP positive cells. All the experiments were performed following manufacturer instructions. The ratio of green absorbance to red absorbance measured using SynergyMx 510nm/590nm and compared to pre X-ray controls to indicate induced lipid peroxidation.

γH2AX assay.

The DNA damage was studied using anti-γH2AX (Alexa 647) antibody (Millipore Sigma, 07-164-AF647). MCF-7 cells were seeded into 6-well plates at a density of either 500,000 cells/mL and incubated overnight. Cells were treated with 1 μM tRA for tRA positive groups. PMAO-KI NPs were added a concentration 50 μg/mL for the NP therapy groups and incubated for an additional 24 hrs. X-ray radiation at 6 Gy was delivered. Cells were incubated for another 24 h at 37 °C. The cells were collected, fixed, permeabilized, and stained with anti-γH2AX antibody according to the protocol from the manufacturer. The presence of γH2AX protein was analyzed using a Millipore Sigma Image Stream Mark II, and both single cell image and statistical data was collected to evaluate DNA damage efficacy.

Animals and tumor inoculation.

All experiments were conducted in accordance with the guidelines from the University of Georgia institutional animal care and use committee (Animal Use Protocol Number: A2017 11-002-A10). Female nude mice (4–5 weeks old, 20) were obtained from (Charles River, USA). One week after mice arrival drinking water was supplemented with 8 mg/L estradiol for MCF-7 tumor model establishment. After one week of pre-supplementation estradiol, 5 million MCF-7 cells in the mixture of 150 μL Matrigel® Matrix (Corning, 354234) and PBS at 1:1 ratio were inoculated into the right flank of mice.

Therapy studies.

MCF-7 tumor bearing mice were randomly divided into 4 groups (n = 5 for each group), including 1) PBS, 2) PMAO-KI NPs, 3) Radiation (RT) at 5 Gy, 4) PMAO-KI NPs+RT (5 Gy). Therapy was started once tumor volume reaches 100 mm³. tRA was intratumarally injected at 1 μM in 50 μL PBS on Day 0, 3 and 5 to stimulate uniform NIS expression. PMAO-KI NPs were intratumarally injected at 50 μg/mL in 50 μL PBS on Day 4. The tumor size and body weight were inspected every 3 days. The tumor was measured in two dimensions with a caliper, and the tumor volume was estimated as (length)×(width)²/2. Estradiol supplementation was continued throughout the endpoints. At the end of the therapy experiment, autopsies were performed. The tumor and major organs were dissected for morphological and histological examination. In particular, tissues in all groups were fixed in 10% (v/v) paraformaldehyde. These tissues were then sectioned into 4 μm slices for H&E, Ki-67 staining to evaluate cell death and proliferation, respectively. Images were acquired using a light microscope (OLYMPUS TH4-100, Japan).

Synthesis of ¹³¹I-KI NPs.

3.9 mCi ¹³¹I-NaI aqueous solution (40 μL) was dried in an Eppendorf tube (1.5 mL) on a thermal mixer at 95 °C over 2 hours before it was cooled down and 0.2 mL hexane was added into the tube to suspend the ¹³¹I-NaI. At the same time, potassium acetate (42.45 mg) was added into a glass tube vial (10 mL) with an aluminum screw cap and a stir bar inside before EtOH (1.25 mL) was injected and stirred until clear solution formed. Then hexane (1.05 mL) was added into the tube vial and stirred until the solution became clear. Sonicate it if necessary. Oleyamine (0.125 mL, heated to be clear and then cooled down to room temperature) was then added into the tube vial and stirred. Acetyl iodide (32 μL) was slowly added into the 0.2 mL ¹³¹I solution in the Eppendorf tube before the mixed iodide was slowly transferred into the tube vial. It was stirred at room temperature for another 30 minutes. The final suspension was a yellow solution with white precipitate. It was separated into aliquots (800 μL × 3) in Eppendorf tubes and centrifuged for 10 minutes at 10,000 rpm. Finally, the brown supernatants were discarded and white precipitates were rinsed with hexanes (100 μL × 3) before the white particles were completely dried on a thermal mixer at 100 °C for 15 minutes. The nanoparticles were cool to room temperature and activities were determined with a dose calibrator.

Statistical analysis.

Comparison of multiple assays was performed using a one-way ANOVA followed by a Turkey test. Comparisons of only two groups was performed using a paired t-test Significance was set at p < 0.05 represented by * in graphs with ** and *** representing p < 0.01 and p < 0.005 respectively. All experiments were performed with at least three replicates unless specified. All the data is represented as mean ± SEM.