Introduction

Alongside many other commonly used anti-cancer strategies and despite recently developed modalities that include targeted biological compounds and immunotherapeutic drugs, most cancer patients worldwide are still treated with some form of radiation therapy (RT), either as a main treatment or as an adjuvant approach [1,2,3]. The major challenge in RT lies in the delicate balance between delivering a destructive, lethal dose into the cancerous tissue while limiting the damage to the surrounding healthy perimeters and the radiotoxicity affecting vital organs. To that end, the use of low versus high linear energy transfer (low-LET and high-LET) differs significantly [4,5,6]. While 1 Gray (Gy) dose of X-rays (low-LET) induces thousands of single-strand breaks (SSBs) to the DNA of a mammalian cell and approximately 40 double strand breaks (DSBs), as LET increases, so does cytotoxicity [7, 8]. Notably, only a few hits of alpha particles (high-LET) to the DNA can produce clustered DSBs that will drive the cell towards an apoptotic course, mainly due to their high ionization density. Clustered DSBs are repaired less efficiently than isolated damage, suggesting that high-LET radiation therapy should be more robustly used for the eradication of tumors [9, 10].

Nevertheless, the vast majority of cancer patients are subjected to low-LET (most often X-rays) RT due to the heavy mass of alpha particles leading to strong interactions with tissue and their slow movement [11, 12]. Hence, the limited range of alpha particles in tissues (less than 100μm) has traditionally been considered a limiting factor, restricting the development of alpha-particle-based therapies [13].

Radiosensitizing agents (radiosensitizers) are substances and compounds which have the capacity to augment the effect produced by RT alone [14,15,RNA and real-time quantitative PCR

RNA was isolated with a mini-RNeasy kit (Qiagen, Germany) and reverse transcribed using Moloney murine leukemia virus reverse transcriptase and random hexamer primers (Promega, USA). Real-time qPCR was performed using SYBR Green Master Mix (Thermo Fisher) in a StepOnePlus instrument (Applied Biosystems). All primers were purchased from Sigma-Aldrich. Each sample was analyzed in triplicate and data were analyzed based on the comparative Ct (2 − ΔΔCt) method. The expression of target genes was normalized to GAPDH expression. The primers used in this study are detailed here:

NOXA: F: 5′-GAAGGGAGATGACCTGTGATTAG-3′/R:5′-TGCTGAGTTGGCACTGAAA-3′.

PUMA: F: 5′-GGA GCA GCA CCT GGA GTC / R: 5′-TA CTG TGC GTT GAG GTC GTC-3′.

GAPDH: F:5′-GGTGTGAACCATGAGAAGTATGA-3′/R:5′-GAGTCCTTCCACGATACCAAAG-3′.

Tumor inoculation

The study was approved by the Ben-Gurion University Institutional Animal Care and Use Committee and was conducted according to the Israeli Animal Welfare Act following the guidelines of the Guide for Care and Use of Laboratory Animals (National Research Council, 1996) [permit no. IL-47-07- 2019(E)]. Male nude mice (6–12 weeks old) were obtained from Envigo, Israel. Mice were inoculated subcutaneously with 5·106 cells (for all cell lines) in 100 μl Dulbecco’s phosphate-buffered saline (DPBS and DMEM) (Gibco, 14190144, Thermo Fisher Scientific, MA, USA) into the low lateral side of the back. At the day of source insertion and before the insertion, the mice were divided into treatment groups to create a similar as possible mean tumor volume for each group. Blinding was not conducted.

Ethics approval and consent to participate

No human subjects were used in this study, therefore ethics and consent are not applicable.

Frozen section preparation

Four to five days post the 224Ra treatment, the tumors were excised (as a whole). Each tumor was cut to two halves, in the estimated location of the seed center, perpendicular to the seed’s insertion axis. The seed was then pulled out using surgical tweezers and was placed in a 1.5 ml microcentrifuge tube filled with 1 ml of water. The tumor was placed for 1 h in −80 °C. The tumors were put in 20 ml santilation bottles on dry-ice and taken for a measurement with HIDEX gamma. Subsequently, both halves of the tumor were subjected to histological sectioning by LEICA CM 1520 cryostat (Buffalo Grove, IL, USA). The 10 µm-thick sections were then placed on positively charged glass slides (76 mm by 26 mm by 1–1.2 mm), with 250–300 µm intervals between each section, creating a series of sequential sections (between 5 and 15 per tumor). Following the sectioning, slides were fixed with 4% paraformaldehyde (sc-281692, Santa Cruz Biotechnology Inc., Dallas, TX, USA) for 10 min and rinsed twice with PBS for 10 min each time. Immediately after the fixation step, slides were taken to the autoradiography system. The same histological sections measured on the imaging plate, were later stained with hematoxylin-eosin (Surgipath, Richmond, IL). The pictures were taken using a Panoramic scanner (3D HISTECH Ltd, Budapest, Hungary).

224Radium‑loaded source preparation and insertion

Stainless steel seeds (0.1 mm in diameter, cut to a length of 6.5 mm) were loaded with 224Ra atoms (half-life of 3.7 days). To prevent Radium dissolution in the tissue fluids, the atoms were embedded a few atomic layers into the seed surface through thermal treatment. Seeds, either loaded with 224Ra or inert, were placed near the tip of an 18-gauge needle attached to a 2.5 ml syringe (Picindolor, Rome, Italy) and inserted into the tumor by a plunger placed internally along the syringe axis. The radioactive and inert seeds were inserted into the primary tumor under anesthesia with Isoflurane. Seed location was verified using a Geiger counter (RAM GENE-1, Rotem industries, Israel) after insertion process was completed and before tumor removal.

Tumor volume measurements

Local tumor growth was determined by measuring three mutually orthogonal tumor dimensions three times per week, according to the following formula: Tumor volume = length·weight·height·π/6. Mice were pre-excluded from the study based on tumor non-uniformity criteria (too big/small tumors before source insertion, double focal tumors, internal tumors) and if the source fell in first 5 days. Tumor volume over time was assessed and compared between the groups using repeated measures ANOVA analysis. The cubic root transformed volume was modeled as a function of group, day (categorical) and the day × group interaction with baseline volume entered as a covariate. Mean (least squares means) and confidence intervals were estimated from the interaction term for each day per group and were back transformed to the volume.

Survival data was depicted by a Kaplan–Meier plot; two curves were compared with a Log-rank test with p-values adjusted for multiple comparisons using the FDR method.

Drug preparation, storage and administration

APR-246 was purchased from Cayman Chemicals was dissolved in DPBS on the same day of the experiment. The stock was stored at −20 °C in powder condition. For the in vivo studies, APR-246 was dissolved DPBS, and 100 µl of 50 mg/kg APR-246 was injected i.p. twice a day; DPBS was used as a sham control. Treatment started 1 day after 224Ra source insertion (day 0) for a total of fourteen doses (days 1–7).

Spatial apoptosis assay

The DeadEnd™ Fluorometric TUNEL System (Promega) measures the fragmented DNA of apoptotic cells by catalytically incorporating fluorescein-12-dUTP at 3′-OH DNA ends using Terminal Deoxynucleotidyl Transferase (TdT), which forms a polymeric tail using the principle of the TUNEL (TdT-mediated dUTP Nick-End Labeling) assay. The fluorescein-12-dUTP-labeled DNA can then be visualized directly by fluorescence microscopy or quantitated by flow cytometry. Imaging was done on Zeiss Cell Discovery 7 system, equipped with a Plan-Apochromat 20×/0.95 objective lens and appropriate LED and filter configuration. Subsequent image analysis was conducted in QuPath software version 0.4.4. Initially, manual annotations, delineating the cancerous regions and the position of the capsules, were created for each slide. The cancer regions (ROIs) were subdivided into 100 μm2 tiles utilizing the built-in SLIC superpixel segmentation algorithm. The mean number of superpixels within each ROI was estimated at 500 ± 75, with a mean analyzed area of 0.5 ± 0.1 mm2. Quantitative descriptors capturing the intensity of the green channel were calculated for every superpixel, followed by the visualization of the average intensities of these superpixels as a heat map superimposed onto the images. Statistical analysis was performed using GraphPad Prism 10. Preliminary assessment of normal distribution was carried out using the Shapiro-Wilk test. The Wilcoxon nonparametric test was employed to evaluate the statistical variance between the control and AFP groups.

Autoradiography of 224Ra-treated tumors and 212Pb leakage probability measurements

A single 224Ra seed (6.5 mm length, 0.7 mm outer diameter), carrying 3 μCi 224Ra, was inserted to the center of a mice-borne HCT116 or PANC-1 tumors. Four to five days later, the tumor was excised (as a whole) and cut in two halves, at the estimated location of the seed center, perpendicular to the seed axis. The seed was then pulled out using surgical tweezers and placed in a water-filled tube for subsequent measurement by a well-type NaI(Tl) detector (Hidex Automatic Gamma Counter). The tumor was kept for 1 h at −80 °C. It was then taken, in dry ice, for measurement in the same gamma counter to determine the 212Pb activity it contains, by focusing on the 212Pb 239 keV gamma line. The measurements of the seed and tumor activity were used to determine the 212Pb leakage probability from the tumor (i.e., the probability that a 212Pb atom released from the seeds leaks out from the tumor through the blood before its decay).

Immediately after the gamma measurement, both halves of the tumor were subjected to histological sectioning using a LEICA CM 1520 cryostat (Buffalo Grove, IL, USA). Sections were cut at 250–300 μm intervals with a thickness of 10 μm, and were then placed on positively charged glass slides, fixed with 4% paraformaldehyde (sc-281692, Santa Cruz Biotechnology Inc., Dallas, TX, USA) and rinsed twice with PBS. Typically, there were 5–15 sections per tumor, spanning a length of 1.5–5 mm. Shortly after their preparation, the glass slides were placed, faced down, for a duration of 1 h, on a phosphor imaging plate (Fujifilm TR2040S) protected by a 12-μm Mylar foil and enclosed in a light-tight casing. Alpha particles emitted from the sections in the decays of 212Pb progeny atoms, 212Bi and 212Po, penetrate through the foil and deposit energy in the active layer of the phosphor imaging plate. Immediately after exposure, the plate was read out by a phosphor-imaging scanner (Fujifilm FLA-9000). The intensity (in units of photo-stimulated luminescence) was converted to 212Pb activity using suitable 212Pb calibration samples. By calculating the total area corresponding, in a given tumor section, to an asymptotic 212Bi/212Po alpha dose larger than 10 Gy, the effective diameter is defined by: deff = 2[A(DBiPo > 10 Gy)/π]1/2. The 10-Gy dose is chosen as a convenient reference for actual therapeutic alpha-particle doses that are expected to be in the range ~10–20 Gy. The same histological sections measured on the imaging plate were later stained with hematoxylin–eosin (H&E) (G-biosciences, St Louis MO, USA) for tissue damage detection. H&E staining was correlated with the activity distribution measurements. The pictures were taken using a Panoramic scanner (3D HISTECH Ltd., Budapest, Hungary).

Western blotting

Cells were lysed with 30 µL of RIPA 1× buffer supplemented with Protease inhibitor cocktail (1:100, Thermo Fisher Scientific, Waltham, MA, USA). Protein concentration in the lysates were determined via the Pierce BCA protein assay kit (Thermo Fisher Scientific, Waltham, MA, USA). Lysates were mixed with sample buffer (5×) (Thermo Fisher Scientific, Waltham, MA, USA) and boiled at 95 °C for 10 min, loaded into gel (Invitrogen Novex WedgeWell 4 to 20%, Tris-Glycine, 1.0 mm, Mini Protein Gel, 10-well, Thermo Fisher Scientific, Waltham, MA, USA), and separated via PAGE. Proteins were then transferred into the nitrocellulose membrane (Greiner, 10-6000-02), blocked for 1 h at room temperature with 5% BSA in TBS and then followed by exposure to primary antibodies (dil 1:1000) overnight at 4 °C: anti-p53, anti-p21, and anti-GAPDH. The membranes were then washed and incubated with HRP-conjugated secondary antibody (dil 1:5000) for 1 h at room temperature, washed, and exposed to SuperSignal West Pico PLUS Chemiluminescent Substrate (Thermo Fisher Scientific, Waltham, MA, USA) as per manufacturer’s protocol before visualizing the membranes in iBrightCL1000 (Invitrogen, A32749, Carlsbad, CA, USA).

Antibodies

Mouse Monoclonal anti-p53 (DO-1) was purchased from Santa Cruz Biotechnology, Santa Cruz, CA, USA. Mouse anti-GAPDH was purchased from Sigma-Aldrich, St. Louis, MO, USA. Anti-p21 antibody [EPR362] was purchased from abcam, Cambridge Biomedical Campus, Hills Road, Cambridge. Anti-Cleaved Caspase-3 antibody (Ab-2302) was purchased from abcam, Cambridge Biomedical Campus, Hills Road, Cambridge.

Biostatistical analysis

All statistical tests were performed in GraphPad PRISM 8.0. and presented as the mean ± standard of measurement. Continuous variables were compared by using the Student’s t-test. Categorical variables were compared by using Chi-square (or the Fisher exact test when appropriate). No statistical methods were used to predetermine the sample size. The variance was similar between the groups that were being statistically compared. Tumor volume over time was assessed and compared between the groups using repeated measures Two-way ANOVA unless stated otherwise. Each experiment was analyzed until the time point at which the first animal died. Survival curves are depicted by a Kaplan–Meier plot and compared with a Log-rank test. A p-value < 0.05 was considered statistically significant.