Cell lines and culture conditions

We selected different HNSCC cell lines based on their sensitivity to cisplatin as described before [20]. The HNSCC cell lines VU-SCC-OE and VU-SCC-1131 were previously established at Amsterdam UMC location VUmc [21, 22]. The two cell lines are HPV-negative and near triploid [23]. VU-SCC-OE has recently been characterized as a typical HPV-negative HNSCC cell line with a classical copy number variation pattern and associated mutations in the known cancer driver genes (e.g. TP53, PIK3CA). The VU-SCC-1131 cell line is established from an FA patient with HNSCC. This cell line harbors a biallelic mutation in the FANCC gene [20, 21], the other morphological and genetic characteristics are very comparable to non-FA HNSCC cell lines with somatic mutations in the typical HNSCC driver genes and frequent copy number alterations [23]. However, the defect in the FA/BRCA DNA-crosslink repair pathway in VU-SCC-1131, causes a high sensitivity for cisplatin, while VU-SCC-OE has an intact FA pathway and is relatively cisplatin-resistant. Both cell lines were cultured in Dulbecco’s modified Eagle’s medium (DMEM), containing 5% fetal calf serum (FCS), 2mM l-glutamine. Cultures were maintained at 37 °C in a humidified atmosphere with 5% CO2.

Determination of cisplatin sensitivity and recovery after exposure

Sensitivity of cell lines to cisplatin was determined by a serial dilution assay as described by Nagel et al. [20]. In short, cells were seeded in 96-wells plates at densities that allowed exponential growth during the time of the experiment. During the experiments, cells were maintained at 37 °C in a humidified atmosphere with 5% CO2. Cisplatin was added 24 h after plating in concentrations ranging from 666 µM–0.635 nM. Cell viability was assessed using a CellTiter-Blue assay (Promega, Leiden, The Netherlands) after 72 h continuous exposure (sensitivity assay) and after 4 h exposure combined with 72 h cisplatin-free medium (recovery assay). Dose-response curves were calculated using Graphpad Prism (version 9.0.0, GraphPad Software, Boston, Massachusetts USA).

Cell cycle analysis

Cells were seeded in T25 flasks (Greiner) at 5 × 105 cells per flask with conditions mentioned before. In total 24 hours after plating, cells were treated for 4 hours with 0.04 µM and 0.2 µM cisplatin. Subsequently, cells were rinsed and cultured with cisplatin-free DMEM, and harvested for further analysis at 0, 72 and 96 hours after medium replacement. Harvested cells were labeled with 10µM 5-ethynyl-2’-deoxyuridine (EdU) for ten minutes. After cell dissociation with trypsin and rinsing,,m,, with phosphate buffered saline (PBS), cells were fixed in 2% para-formaldehyde and resuspended in ethanol 70% end left overnight at -20 °C. Cells were permeabilized by incubation with the Triton X-100-based buffer (Thermo scientific, Waltham, United states) for 30 min at room temperature, cells were stained with mitosis marker Alexa Fluor® 647 anti-Histone H3 Phospho (Ser10) (Biolegend 650806, clone 11D8) and DAPI. BD LSR II Fortessa™ (BD Biosciences, Vianen, The Netherlands) and BD FACSDiva™ software (V8.0.1.1, BD Biosciences) were used for flow cytometry and data analysis.

Production of Pt-195m radiolabeled cisplatin

Radioactive Pt-195m cisplatin was provided by NRG Advancing Nuclear Medicine (Petten, The Netherlands). Production of Pt-195m is described in the Additional file 1: Supplementary materials and methods, and synthesized according to a previously described protocol [24] In short, platinum-195m was produced by irradiation in the High Flux Reactor in separate batches (Pt-0009, Pt-0011 and Pt-0016). After irradiation, [195mPt]Cisplatin (CISSPECT®), or cis-[195mPt ][Pt(NH3)2Cl2], was synthesized according to published procedures [24] as a 1 mg/ml solution in 0.9% NaCl with a pH of 5 to 5.5. Part of the [195mPt]Cisplatin solution, containing 0.065–0.067 mg Pt, was analyzed for radioactivity and radionuclide purity (197Pt, 191Pt 192Ir, 194Ir, 198Au, 199Au) using a high purity Germanium detector coupled to a multi-channel analyzer system. The specific activity per mg Platinum at End of Irradiation (EoI) was 81MBq, 86 MBq and 131 MBq 195mPt/mg Platinum for Pt-0009, Pt-0011 and Pt-0016 batches respectively. At the end of synthesis, 48 h after EoI, the Activity Reference Time (ART) was set, at ART the radionuclide purity increased to 93.9%, 93.6% and 99.1% for Pt-0009, Pt-0011 and Pt-00016 respectively due to decay of mainly 197Pt. [195mPt]cisplatin showed characteristic chemical properties as known for cisplatin, in line with the European Pharmacopoeia regulations for cisplatin (Additional file 1: Supplementary materials and methods).

Intra-cellular retention of [195mPt]cisplatin and DNA repair analysis

Both cell lines were seeded at a density of 1 × 106 cells per T25 flask (Greiner), and 2 days after seeding [195mPt]cisplatin (Pt-0011, Pt-0016) was added in three concentrations (5 µM, 20 µM and 75 µM). To assess uptake and retention of [195mPt]cisplatin we performed, the experiment under three conditions: (A) 4 h exposure to [195m]cisplatin and after which cells were harvested, (B) 4 h exposure to [195mPt]cisplatin after which the cisplatin-containing medium was removed, cells were rinsed with PBS and cisplatin-free DMEM was added and left for 24 h and (C) 24 h incubation with [195mPt]cisplatin after which cells were harvested. Cells were analyzed under the microscope for viability. For further analysis of all different fractions the cell suspension was centrifuged for 5 min at 500xg, the cell fraction was rinsed with PBS and collected. Cells were lysed and homogenized, and both RNA and DNA were isolated using the AllPrep DNA/RNA mini kit (Qiagen). The radioactivity of all fractions was determined using a gammacounter (LKB-Wallac, 1282 CompuGamma; Pharmacia, Woerden), and corrected for decay.

In vivo experiments on cisplatin and radioactive cisplatin

Athymic nude-Foxn1nu (Envigo) mice were subcutaneously injected with either the VU-SCC-OE or the VU-SCC-1131 HNSCC cell line on both flanks, with 2 × 106 cells per site and randomized to either cisplatin or vehicle-control group. Tumor volume was measured with electronic calipers (V= (L x W x H) x 0.5 where V = volume, L = length, W = width and H = height). When tumors reached an average size of 100 mm3 (range 80-200mm3) drugs were administered. In vivo sensitivity was determined after intra-peritoneal injection of 5 mg/kg cisplatin (CDDP, Accord) at day 0 and 7 after randomization. Both conditions were tested in 3–5 mice per group with 1–2 tumors per mouse, injected sites were discarded from further analysis when there was no tumor growth at baseline. Tumor volume was measured every 2–3 days for 21 days, mice were euthanized when tumor volume exceeded 1000mm3 or invaded the skin.

For the analysis of the biodistribution of radioactive cisplatin, mice were randomized to the [195mPt]cisplatin (Pt-0009 and Pt-0011) group (intravenously or intra-peritoneal) or the control group (n = 3 per group). We injected 2.5 mg/kg of Pt-195 m solution, 8 days after production and delivery. Standards of radioactivity were taken from the injection solution. Mice were euthanized at 1, 2, 6 and 24 h after injection, organs were dissected and subsequently weighed and radioactivity was determined using a gammacounter (LKB-Wallac, 1282 CompuGamma; Pharmacia, Woerden). The rate of counts per minute (CPM) was normalized as the percentage of CPM of the injected fraction, and subsequently standardized to the organ weight in grams. All animal experiments were performed according to Dutch and EU legislations, and the protocol was approved by the Institutional Review Board on animal experimentation.

Statistics

Baseline statistics were performed in Rstudio (R version 4.0.3 (2020-10-10)). Statistical analysis of comparative tumor growth repeated measures of xenografts were performed by fitting a linear mixed effects regression model, with time in days and each single mouse as random effects. We used the control mice to compute a growth speed using a linear regression model, with the regression equation per cell line we calculated an expected tumor volume per time point. The expected tumor volume was introduced in the linear mixed affects model as a random effect. Wilcoxon signed-rank test was used to compare tumor sizes at a predetermined timeframe.