A bifunctional carborane-quinazoline combines tumor growth inhibition and boron delivery for boron neutron capture therapy in melanoma and non-small cell lung cancer models

1 Introduction

Boron neutron capture therapy (BNCT) is a binary radiotherapeutic strategy based on the preferential accumulation of boron-containing compounds in tumor cells followed by irradiation with low-energy neutrons. Neutron capture by ¹⁰B generates an excited ¹¹B nucleus that rapidly decays into an alpha particle (⁴He) and a high-linear-energy-transfer lithium nucleus (⁷Li), whose path length is comparable to cellular dimensions (Barth et al., 1996). This reaction enables highly localized radiation damage in boron-loaded cells, making BNCT particularly attractive for tumors in which selective irradiation is difficult to achieve using conventional radiotherapy. However, the therapeutic efficacy of BNCT depends critically on the selective delivery, intracellular accumulation, and retention of sufficient boron within tumor tissue (Wang et al., 2022). The clinically used boron delivery agents, mainly ¹⁰B-boronophenylalanine (BPA, Figure 1) and sodium ¹⁰B-borocaptate (BSH, Figure 1), have enabled important advances in BNCT, but their performance remains limited by heterogeneous tumor uptake, variable retention, and dependence on biological transport mechanisms (Xu et al., 2024; Selg and Hey-Hawkins, 2026). Although BPA remains the most clinically relevant boron delivery agent in BNCT, its efficacy can be affected by heterogeneous uptake and limited intracellular retention associated with transporter-mediated exchange mechanisms (Nomoto et al., 2020). Recent strategies, including polymer-BPA complexes, have sought to enhance tumor accumulation and reduce boron efflux, underscoring the relevance of boron retention as a key determinant of BNCT efficacy (Nomoto et al., 2020). Beyond optimization of BPA itself, increasing attention has been directed toward exploiting tumor-associated nutrient transporters and cell-surface receptors to achieve more selective boron delivery. Representative approaches include glucose-derived boron carriers designed to exploit GLUT1, glutamate-mimetic small molecules targeting ASCT2, folate receptor-α (FRα)-targeted boron conjugates, and boron-rich tyrosine kinase inhibitor platforms directed toward receptors such as EGFR, FLT3, KIT, and PDGFR-β (Imlimthan et al., 2025; Miura et al., 2026; Nishimura et al., 2023; Kaniowski et al., 2022; Couto et al., 2026a; Alamón et al., 2020). These strategies illustrate the broader shift toward biologically guided boron delivery based on molecular features preferentially expressed or exploited by tumor cells. In this context, multifunctional boron-containing compounds that combine boron delivery with intrinsic antitumor activity represent an emerging strategy for BNCT.

Cutaneous melanoma (CM) and non-small cell lung cancer (NSCLC) represent two biologically and clinically relevant models for evaluating this type of strategy. Melanoma has a long-standing connection with BNCT, particularly through BPA-based approaches, and remains a tumor type in which localized and selective therapeutic modalities may provide clinical benefit in selected settings (Mishima et al., 1989; Fukuda et al., 1999). NSCLC, in turn, is one of the most relevant solid tumors for targeted therapy, with EGFR signaling playing a key role in tumor proliferation, survival, and therapeutic response in molecularly defined subsets (Bethune et al., 2010). In addition, radiotherapy for lung cancer is frequently constrained by normal tissue toxicity and respiratory motion, making tumor-selective approaches such as BNCT conceptually attractive. Importantly, CM and NSCLC may differ in the expression of EGFR and LAT1, providing complementary biological contexts to assess both targeted kinase inhibition and boron transport mechanisms (Bethune et al., 2010; Lee et al., 2021; Shi et al., 2023; Watabe et al., 2023).

In this framework, EGFR-targeted boronated quinazolines constitute an attractive chemical biology platform. We previously reported a carborane-containing erlotinib-derived hybrid, herein referred to as Cmpd. 1 (Figure 1), designed to combine an EGFR tyrosine kinase inhibitor scaffold with a boron-rich cluster (Couto et al., 2017). Cmpd. 1 showed potent inhibition of wild-type EGFR, measurable activity against the T790 M mutant variant (Couto et al., 2018), favorable drug-like properties (Couto et al., 2019), tumor-cell selectivity (Alamón et al., 2023a; Alamón et al., 2023b), absence of mutagenicity, and no relevant acute oral toxicity in preclinical studies (Alamón et al., 2023a). Moreover, in glioma models, Cmpd. 1 promoted selective boron accumulation, showed improved in vitro BNCT activity compared with BPA-fructose under comparable conditions, and increased survival in glioma-bearing mice when administered either as a free compound or within a targeted liposomal formulation (Alamón et al., 2023b). These findings support Cmpd. 1 as a bifunctional molecule capable of integrating EGFR-oriented pharmacology with boron delivery for BNCT.

Nevertheless, whether this platform can be extended beyond glioma to other tumor types with distinct biological profiles remains insufficiently explored. Here, we evaluated Cmpd. 1 as a bifunctional agent for targeted antitumor activity and boron delivery in BNCT using CM and NSCLC models. The study integrates EGFR and LAT1 expression analysis, cellular boron uptake, neutron autoradiography, in vitro BNCT-mediated cytotoxicity, cell death assessment, in vivo tumor growth inhibition, and biodistribution studies. This approach was designed to determine whether a previously validated erlotinib-carborane platform can be applied to tumor models with different dependencies on EGFR signaling and boron transport pathway.

2 Materials and methods

2.1 Solutions preparation of boron-containing compounds

Cmpd. 1 was synthesized in our laboratory as previously reported (Couto et al., 2017) and used with the natural isotopic abundance of boron, corresponding to approximately 20% ¹⁰B and 80% ¹¹B. Boron concentrations determined by inductively coupled plasma optical emission spectroscopy (ICP-OES) refer to total boron content, as this technique does not discriminate between boron isotopes. For neutron autoradiography and BNCT experiments, the ¹⁰B-equivalent concentration of Cmpd. 1 was calculated from its known molecular boron content and the natural isotopic abundance of ¹⁰B. The amount of compound used in these experiments was adjusted accordingly to achieve the indicated ¹⁰B concentrations. Thus, the ¹⁰B concentration reported for neutron irradiation experiments refers to the calculated ¹⁰B fraction rather than to total boron. Cmpd. 1 was prepared as a 10 mM stock solution in dimethyl sulfoxide (DMSO) by dissolving 6.2 mg of Cmpd. 1 in 1 mL of DMSO. The final concentration of DMSO in cell culture assays was kept constant across the corresponding experimental groups and did not exceed 0.5%. ¹⁰B-enriched boronophenylalanine (BPA; 99% ¹⁰B enrichment) was purchased from Interpharma praha, a.s. (Prague, Czech Republic) and used as the reference boron delivery agent. BPA was prepared as a fructose complex at 30 mg/mL (0.14 M), adjusting the pH to 9.5–10.0 and subsequently to 7.5 using NaOH, as previously described (Couto et al., 2017).

2.2 Cell lines

The human cutaneous melanoma cell line Mel J and the human non-small cell lung cancer cell line NCI-H125 were used as the experimental tumor cell models. In addition, U87 MG human glioblastoma cells and 8505C human anaplastic thyroid carcinoma cells were included exclusively as positive controls for EGFR and LAT1 expression, respectively. Mel J cells were kindly provided by Dr. José Mordoh (Instituto Leloir, Argentina) (Guerra et al., 1989) while NCI-H125 cells were obtained from the American Type Culture Collection (ATCC). Mel J cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM), and NCI-H125 cells were maintained in RPMI-1640 medium. The human anaplastic thyroid carcinoma cell line 8505C, kindly provided by Dr. Cintia Rosemblit (BIOMED-UCA, Argentina), and the human glioblastoma cell line U87 MG, obtained from ATCC. 8505C was cultured in RPMI-1640 supplemented with 154 mg/L sodium pyruvate, 1.5 g/L sodium bicarbonate, 100 μg/mL gentamicin, and 10% FBS, whereas U87 MG was maintained in DMEM supplemented with 10% FBS. All cell lines were maintained at 37 °C in a humidified atmosphere containing 5% CO₂ and were passaged at approximately 80% confluence using 0.25% trypsin.

2.3 EGFR and LAT1 expression analysis by quantitative real-time PCR

EGFR and LAT1 expression levels were evaluated by quantitative real-time PCR (RT-qPCR). Total RNA was extracted from cultured cells using TRIzol™ Reagent (Invitrogen, Thermo Fisher Scientific) according to the manufacturer’s instructions. RNA concentration and purity were determined using a NanoDrop™ spectrophotometer (Thermo Fisher Scientific), and 1 μg of total RNA was reverse-transcribed into cDNA using M-MLV Transcriptase 2.0 (PBL Productos Bio-Lógicos). Quantitative PCR reactions were performed in a final volume of 10 μL containing 2 μL of diluted cDNA (1:10), 0.3 μL of gene-specific primers (10 µM), 2.4 μL of nuclease-free water, and 5 μL of 2× qPCR Master Mix (PBL Productos Bio-Lógicos). Amplification was carried out on a Rotor-Gene Q real-time PCR system (Qiagen) using 40 cycles of 95 °C for 15 s, 59 °C for 30 s, and 72 °C for 30 s. The following primers were used: LAT1 forward 5′-CGT​GAA​CTG​CTA​CAG​CGT​GA-3′ and reverse 5′- TTG​GAC​ACA​TCA​CCC​TTC​CC-3′; EGFR forward 5′-GCC​GCA​AAG​TGT​GTA​ACG​G-3′ and reverse 5′-GAT​CGC​CAC​TGA​TGG​AGG​TG-3′. Relative gene expression was normalized to the housekeeping gene GAPDH and quantified using the 2^−ΔΔCt method. U-87 MG human glioblastoma cells were used as a positive control for EGFR expression (Chen et al., 2017), whereas 8505C human anaplastic thyroid carcinoma cells served as a positive control for LAT1 expression (Rossich et al., 2023).

2.4 In vitro cytotoxicity assay

Mel J and NCI-H125 cells were seeded into 96-well plates at a density of 2 × 10⁴ and 1 × 10⁴ cells per well respectively and incubated for 24 h at 37 °C in a humidified atmosphere containing 5% CO₂. Cells were then treated with increasing concentrations of Cmpd. 1 for 24 h, maintaining a constant final DMSO concentration (≤ 0.5%) in all experimental groups. Cell viability was determined using the CellTiter 96® AQueous One Solution Cell Proliferation Assay (MTS; Promega, Madison, WI, United States of America) according to the manufacturer’s instructions. Briefly, the MTS reagent was added to each well, incubated for the recommended time, and absorbance was measured at 490 nm using a microplate reader. Cell viability was expressed as the percentage of vehicle-treated control cells, and IC₅₀ values were calculated by nonlinear regression using GraphPad Prism.

2.5 Cellular uptake of boron compounds

For boron uptake studies, cells were seeded in 60 mm culture dishes and maintained under exponential growth conditions until reaching 60–70% confluence. Cells were then incubated with Cmpd. 1 (25 μM), a concentration selected below its IC₅₀ value determined in the in vitro cytotoxicity assay, or with BPA (2.33 mM). These conditions corresponded to approximately 2.7 ppm total boron for Cmpd. 1 and 25 ppm total boron for BPA. Boron uptake was evaluated after 2, 4, and 24 h of incubation. Following treatment, cells were washed twice with ice-cold phosphate-buffered saline (PBS) to remove extracellular compound and digested with 600 μL of nitric acid. A 300 μL aliquot was processed for boron quantification by inductively coupled plasma optical emission spectrometry (ICP-OES) following heating and sonication. After cooling, 100 μL of H₂O₂ was added, and the final volume was adjusted to 2 mL with Milli-Q water, resulting in final proportions of 15% nitric acid, 5% H₂O₂, and 80% Milli-Q water. The remaining aliquot was evaporated to dryness at 35 °C under a fume hood, resuspended in 0.3 N NaOH, and used for total protein determination by the Lowry method. Boron uptake was expressed as μg of boron per mg of total protein (Dagrosa et al., 2002).

2.6 Boron quantification by ICP-OES

Boron content was quantified by ICP-OES. The analytical emission line used for boron quantification was 249.677 nm. Daily calibration curves were prepared using matrix-matched boron standard solutions ranging from 0.03 to 0.20 mg/L for cells and 0.05–0.80 mg/L for tissues samples. Samples were analyzed under the same matrix conditions used for calibration. Results were expressed as total boron content.

2.7 Neutron autoradiography

Neutron autoradiography with UV-C sensitization was employed to qualitatively assess the intracellular localization of boron delivered by Cmpd. 1 and BPA in Mel J cells (Gadan et al., 2019). Neutron autoradiography allows qualitative assessment of boron distribution by detecting particle tracks produced by the ¹⁰B (n, α) ⁷Li nuclear reaction. Mel J cells were grown on Lexan polycarbonate substrates and incubated for 2 h with Cmpd. 1 and BPA at concentrations equivalent to 0.5 and 10 ppm of ¹⁰B, respectively. After incubation, cells were thoroughly washed with ice-cold PBS to eliminate extracellular compound prior to irradiation.

Samples were exposed to a thermal neutron beam at the RA-3 nuclear reactor facility (Centro Atómico Ezeiza, Argentina), under experimental conditions comparable to those used for BNCT assays. Following irradiation, the cell cultures on Lexan were stained with haematoxylin and exposed to UV-C radiation for 2 min. Then, the polycarbonate films were chemically etched to allow the visualization of latent tracks produced by the high-LET particles resulting from the neutron capture reaction and the imprint of the cell cultures generated by the UV-C exposure.

Autoradiographic and stained cell culture images were analyzed by optical microscopy, maintaining consistent magnification and field orientation. By adjusting the focal plane, the distribution of particle tracks was correlated with cellular structures, allowing a qualitative assessment of boron localization. This approach provided insight into the intracellular distribution pattern of BPA and Cmpd. 1, contributing to the evaluation of its suitability as a boron carrier for BNCT applications.

2.8 In vitro BNCT experimental design

Seven experimental groups were established to evaluate the effects of boron compounds and neutron irradiation, and their combination, as summarized in (Table 1).

TABLE 1. Experimental groups and treatment conditions used for the in vitro BNCT studies.

Cells were seeded in 96-well plates at a density of approximately 3,000 cells per well and allowed to adhere prior to treatment. Incubation with boron compounds was performed under standard culture conditions to achieve the desired intracellular boron concentration before irradiation. Neutron exposure was conducted to deliver a total physical dose of 2 Gy (± 10%), as described in the dosimetric characterization (Table 2). Following irradiation, cells were maintained in culture for subsequent biological analyses, including cell viability and cell death assessment, enabling comparison across all experimental conditions.

TABLE 2. Dosimetry for thermal neutron irradiation without and with boron (1 ppm¹⁰B).

BNCT, dose calculations were performed considering a¹⁰B concentration of 1 ppm, irrespective of the boron carrier (BPA, or Cmpd. 1).

2.9 Neutron irradiation conditions

Neutron irradiation was carried out at the RA-3 nuclear reactor, using a thermal neutron beam facility suitable for BNCT experimental studies. The irradiation setup was designed to ensure homogeneous neutron exposure across the samples while maintaining controlled environmental conditions. The neutron beam was characterized by a predominant thermal neutron component, with a flux in the order of 5 × 10⁹ n·cm^-2·s^-1, accompanied by epithermal and gamma contamination within acceptable limits for biological experiments. Dosimetric characterization was performed according to established protocols, considering the physical dose contributions from gamma radiation, the ¹⁴N(n,p)¹⁴C reaction, and the ¹⁰B(n,α)⁷Li reaction.

Samples were positioned at a fixed distance from the beam exit, and irradiation time was adjusted based on previously determined dose rates to achieve the desired exposure. During irradiation, cells were maintained in appropriate culture conditions to preserve viability and minimize environmental stress.

For neutron irradiation experiments, the concentrations of Cmpd. 1 and BPA were adjusted to provide a final ¹⁰B concentration of 1 ppm. For Cmpd. 1, which contains boron at natural isotopic abundance (approximately 20% ¹⁰B), the required concentration was calculated from its molecular boron content and the natural abundance of ¹⁰B.

2.10 Cell viability after neutron irradiation

Cell viability was evaluated 7 days after irradiation using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. Briefly, cells were incubated with MTT solution (5 mg/mL in PBS) under standard culture conditions to allow the reduction of MTT to insoluble formazan crystals by metabolically active cells. Formazan crystals were subsequently dissolved, and absorbance was measured at 595 nm using a microplate reader. Cell survival was expressed as the percentage of absorbance relative to the corresponding untreated control group. This assay was used to evaluate the long-term effects of neutron irradiation, boron compounds, and BNCT on cell viability.

2.11 Cell death analysis

Cell death was evaluated 48 h after irradiation by fluorescence staining with Hoechst 33,258, fluorescein diacetate (FDA), and propidium iodide (PI). Viable cells were identified by positive FDA staining, intact plasma membrane (PI-negative), and normal nuclear morphology. Apoptotic cells were identified primarily by characteristic nuclear changes, including chromatin condensation and/or nuclear fragmentation visualized by Hoechst staining. Necrotic cells were identified by loss of plasma membrane integrity, evidenced by PI uptake, in the absence of the characteristic fragmented or condensed nuclear morphology used to classify apoptotic cells. At least 200 cells per sample were scored in randomly selected microscopic fields, and the percentages of apoptotic and necrotic cells were calculated separately for each experimental condition.

2.12 In vivo antitumor efficacy study

Athymic NIH nude mice (6–8 weeks old, 20–25 g, both sexes) were used for the in vivo studies. Subcutaneous xenografts were established by injecting 2.5 × 10⁶ Mel J cells or 1 × 10⁶ NCI-H125 cells into the right flank in a maximum volume of 100 μL PBS. Approximately 2 weeks after inoculation, when tumors reached 40–70 mm³, animals were randomly assigned to the experimental groups (n = 5 per group).

For animal administration, Cmpd. 1 was dissolved in DMSO at a concentration of 1 mg/mL. The treatment group received Cmpd. 1 by intraperitoneal injection at 1 mg/kg body weight three times per week over an 18-day treatment period, for a total of eight administrations, whereas control animals received an equivalent volume of DMSO. The administered volume was adjusted according to individual body weight.

Body weight and tumor size were monitored throughout the study. Tumor dimensions were measured with a digital caliper, and tumor volume was calculated using the formula:where A represents the longest tumor diameter and B the shortest diameter. Tumor growth was expressed as relative tumor volume (Vf/Vi) over time.

V = A × B 2 / 2

All animal procedures were performed in accordance with institutional guidelines for the care and use of laboratory animals and were approved by the Institutional Committee for the Care and Use of Laboratory Animals (CICUAL, CNEA; Protocol No. 05_2020_V3).

2.13 End-of-treatment boron tissue biodistribution

The tissue biodistribution of boron was evaluated at the end of the in vivo antitumor efficacy study, using the same tumor-bearing nude mice described in Section 2.12. This experiment was designed to determine tissue boron concentrations under the conditions present at the conclusion of the repeated-dose treatment protocol rather than to characterize the biodistribution of Cmpd. 1 following a single administration previously untreated animals. No washout period was introduced before the final administration. Following completion of the 18-day treatment schedule (total 8 injections), animals from each tumor model were evaluated in two groups (n = 5 per group): BPA alone or Cmpd. 1 plus BPA. Mice received a final intraperitoneal administration of the corresponding boron compounds (Cmpd. 1, 1 mg/kg b.w; BPA-fructose, 350 mg/kg b.w., 0.14 M).

One hour after the single final administration of the boron compounds, animals were anesthetized with ketamine (120 mg/kg) and xylazine (21 mg/kg) and euthanized by cervical dislocation. Tumor, blood, surrounding skin, distal skin, lung, kidney, liver, and spleen were collected, weighed, and processed for boron analysis. Tissue samples (10–50 mg) were digested in 750 μL of nitric acid at 70 °C for 1 h, then H₂O₂ was added (Olivera et al., 2025) and boron concentrations were determined by ICP-OES as described above for cell samples. Results were expressed as μg boron per gram of tissue (ppm).

2.14 Statistical analysis

Data are presented as mean ± standard deviation (SD) or mean ± standard error of the mean (SEM), as indicated in the corresponding figure and table legends. Statistical analyses were performed using GraphPad Prism version 10.0 (GraphPad Software, Boston, MA, United States of America).

For in vitro studies, comparisons among multiple groups were performed using one-way ANOVA followed by Dunnett’s or Tukey’s multiple comparisons test, as appropriate. Dunnett’s test was used when multiple groups were compared with a single reference group, whereas Tukey’s test was used when pairwise comparisons among experimental groups were required, including the BNCT cell survival experiments.

For in vivo studies, tumor growth curves were analyzed using two-way ANOVA followed by Bonferroni’s multiple comparisons test. Biodistribution data involving comparisons of multiple tissues against a single reference tissue were analyzed using one-way ANOVA followed by Dunnett’s multiple comparisons test. Comparisons between two independent groups were performed using an unpaired two-tailed Student’s t-test.

A p value < 0.05 was considered statistically significant. The number of biological and technical replicates is indicated in the corresponding figure and table legends.

3 Results

3.1 EGFR and LAT1 expression in Mel J and NCI-H125 cells

Figure 2 shows the relative expression levels of EGFR and LAT1 in NCI-H125 and Mel J cells. EGFR, the molecular target of erlotinib, is directly involved in the mechanism of action of Cmpd. 1, whereas LAT1 is the major amino acid transporter mediating the cellular uptake of BPA.

Quantitative PCR analysis revealed distinct expression patterns for both genes in the two tumor cell lines. NCI-H125 cells exhibited significantly higher EGFR expression than Mel J cells (Figure 2A). Likewise, LAT1 transcript levels were higher in NCI-H125 than in Mel J cells (Figure 2B). These findings indicate that the two tumor models differ in the transcript expression involved in targeted boron delivery, supporting their use to compare the performance of the EGFR-targeted boron carrier Cmpd. 1 with the LAT1-mediated boron uptake mechanism of BPA.

3.2 In vitro cytotoxicity and cellular boron uptake of cmpd. 1

The in vitro cytotoxic activity of Cmpd. 1 was first evaluated in Mel J and NCI-H125 bnccells. Dose-response analysis showed that NCI-H125 cells were more sensitive to Cmpd. 1 than Mel J cells, with IC₅₀ values, after 24 h of incubation, of 39.73 μM and 60.79 μM, respectively (Figures 3A,B). Based on these results, a concentration of 25 μM, below the IC₅₀ in both cell lines, was selected for subsequent boron uptake studies.

Cellular boron uptake was quantified by ICP-OES after incubation with BPA or Cmpd. 1 for 2, 4, and 24 h. In both Mel J cells y NCI-H125 cells, Cmpd. 1 produced a marked increase in intracellular boron accumulation after 24 h compared with BPA, despite being added at a substantially lower total boron concentration (2.7 ppm for Cmpd. 1 versus 25 ppm for BPA). When uptake was normalized to the amount of boron added to the culture medium, Cmpd. 1 exhibited markedly greater boron delivery efficiency than BPA, with normalized Cmpd. 1/BPA uptake ratios of 56.65 in Mel J cells and 24.06 in NCI-H125 cells at 24 h (Figures 3C,D). These findings demonstrate that Cmpd. 1 is an efficient boron delivery agent in both tumor models.

3.3 Neutron autoradiography

Neutron autoradiography was used to qualitatively compare the cellular microdistribution of boron delivered by BPA and Cmpd. 1. Mel J cells incubated with BPA and the corresponding neutron autoradiograph are shown in Figures 4A,B, respectively, whereas cells treated with Cmpd. 1 and the corresponding autoradiograph are shown in Figures 4C,D. Detectable particle tracks were observed after incubation with both compounds. Although BPA was used at 10 ppm ¹⁰B and Cmpd. 1at 0.5 ppm ¹⁰B, Cmpd. 1-Treated cells displayed a clear autoradiographic signal, supporting efficient cellular boron delivery under these conditions. Track distribution was heterogeneous among cells and, under the imaging conditions used, qualitative inspection did not reveal an obvious preferential localization in either nuclear or cytoplasmic regions.

3.4 Cell survival following BNCT

Cell survival was evaluated 7 days after neutron irradiation by the MTT assay to assess the long-term effects of BNCT on tumor cell viability (Figure 5). BNCT significantly reduced cell viability in both CM and NSCLC cells compared with their respective controls (p < 0.05). In melanoma cells, BPA- and Cmpd. 1-Mediated BNCT produced comparable reductions in cell survival. In contrast, NSCLC cells exhibited a more pronounced decrease in viability, particularly following Cmpd. 1-Mediated BNCT (p < 0.001), indicating a stronger potential antitumor response. Additionally, neutron irradiation alone (NCT) significantly reduced NSCLC cell viability (p < 0.01), suggesting a higher intrinsic radiosensitivity of these cells. Notably, Cmpd. 1 alone induced a modest but significant reduction in NSCLC cell viability (p < 0.05), supporting its role not only as a boron delivery agent but also as a compound with intrinsic cytotoxic activity.

3.5 BNCT-induced cell death mechanisms

Cell death mechanisms were evaluated 48 h after neutron irradiation by fluorescence microscopy using nuclear morphology and membrane integrity criteria. Representative images of Mel J and NCI-H125 cells are shown in Figures 6A,B, respectively, and the corresponding quantification of apoptotic and necrotic cells is presented in Figures 6C,D. In both tumor models, BNCT increased cell death compared with the corresponding non-irradiated controls. In Mel J cells, BNCT mediated by either BPA or Cmpd. 1 predominantly induced necrotic cell death, with necrosis reaching approximately 50%, whereas apoptosis remained around 20–30% (Figure 6C). In contrast, NCI-H125 cells exhibited a mixed pattern of cell death following BNCT. Cmpd. 1-Mediated BNCT induced approximately 35% apoptosis and 35% necrosis, whereas BPA-mediated BNCT resulted in approximately 17% apoptosis and 33% necrosis (Figure 6D). These results show differences in the cell-death response between the 2 cell lines, with a larger apoptotic fraction observed in NCI-H125 cells following Cmpd. 1-Mediated BNCT.

3.6 In vivo antitumor activity of cmpd. 1

The intrinsic antitumor activity of Cmpd. 1 was evaluated in melanoma and NSCLC xenograft models following repeated intraperitoneal administration (1 mg/kg) in the absence of neutron irradiation (Figure 7). In both tumor models, treatment with Cmpd. 1 delayed tumor growth compared with the corresponding vehicle-treated controls, as reflected by lower relative tumor volumes throughout the treatment period (Figures 7A,C). These findings demonstrate that Cmpd. 1 retains intrinsic antitumor activity in vivo independently of BNCT. Body weight remained stable throughout the study, with no significant differences between treated and control animals in either tumor model (Figures 7B,D), indicating that the treatment was well tolerated and produced no evidence of overt systemic toxicity based on this parameter.

3.7 Boron biodistribution in CM and NSCLC xenograft models

End-of-treatment tissue boron concentrations were evaluated 1 h after a final intraperitoneal administration of BPA alone or Cmpd. 1 plus BPA in CM and NSCLC xenograft models. The biodistribution profile of the combined treatment is shown in Figure 8, whereas the complete quantitative boron concentrations for both the BPA-alone and Cmpd. 1 plus BPA groups are presented in Tables 3,4, allowing direct comparison between the two experimental conditions.

TABLE 3. Biodistribution studies of different treatments with Cmpd. 1 and BPA in CM model.

Table 3 Athymic NIH, nude mice bearing Mel J cutaneous melanoma xenografts (2.5 × 10⁶ cells) were treated with Cmpd. 1 administered intraperitoneally (i.p.) three times per week for 18 days (total 8 injections). Subsequently, animals received a final intraperitoneal dose of Cmpd. 1 (1 mg/kg body weight) in combination with BPA (350 mg/kg body weight) (Group 1), or BPA, alone (350 mg/kg body weight) (Group 2). All compound administrations were performed via the intraperitoneal route. Boron concentrations were determined 1 h after the final administration by ICP-OES, and are expressed as μg B/g tissue (mean ± SEM, n = 5 mice per group). Statistical comparisons between the Cmpd. 1 + BPA, and BPA-alone groups were performed for each tissue using an unpaired two-tailed Student’s t-test. No statistically significant differences were observed (p > 0.05).

TABLE 4. Biodistribution studies of different treatments with Cmpd. 1 and BPA in NSCLC model.

Table 4 Athymic NIH, nude mice bearing NCI-H125, non-small cell lung cancer (NSCLC) xenografts (1 × 10⁶ cells) were treated with Cmpd. 1 administered intraperitoneally (i.p.) three times per week for 18 days (total 8 injections). Subsequently, animals received a final intraperitoneal dose of Cmpd. 1 (1 mg/kg body weight) in combination with BPA (350 mg/kg body weight) (Group 1), or BPA, alone (350 mg/kg body weight) (Group 2). All compound administrations were performed via the intraperitoneal route. Boron concentrations were determined 1 h after the final administration by ICP-OES, and are expressed as μg B/g tissue (mean ± SEM, n = 5 mice per group). Statistical comparisons between the Cmpd. 1 + BPA, and BPA-alone groups were performed for each tissue using an unpaired two-tailed Student’s t-test. No statistically significant differences were observed (p > 0.05).

The combined administration of Cmpd. 1 and BPA resulted in preferential boron accumulation in tumor tissue relative to blood in both tumor models. In melanoma-bearing mice, tumor boron concentration reached 16.38 ± 5.49 μg B/g tissue, corresponding to a tumor-to-blood ratio of approximately 4.7. Tumor boron levels also exceeded those measured in surrounding normal skin, yielding a tumor-to-normal ratio (T/N) of approximately 1.6. In NSCLC-bearing mice, tumor boron concentration reached 22.91 ± 6.07 μg B/g tissue, corresponding to a tumor-to-blood ratio of approximately 6.3, while the tumor-to-surrounding normal skin ratio was approximately 2.2. In addition, a tumor-to-lung ratio of approximately 3.7 was observed in NSCLC mode. Although the tumors were established as subcutaneous xenografts, the tumor-to-lung ratio was included as an additional disease-relevant parameter because normal lung represents the tissue of origin of NSCLC and would constitute a clinically relevant normal tissue in a potential lung-directed BNCT application. Comparison with the BPA-alone group, whose complete quantitative biodistribution data are provided in Tables 3,4, showed that, in CM xenografts, tumor boron concentrations were similar following administration of Cmpd. 1 plus BPA and BPA alone (15.53 ± 1.33 μg B/g tissue). In contrast, in the NSCLC model, combined administration of Cmpd. 1 plus BPA resulted in a higher mean tumor boron concentration than BPA alone (16.84 ± 1.54 μg B/g tissue). Overall, these findings demonstrate preferential boron accumulation in tumor tissue relative to blood and normal tissues in both xenograft models.

4 Discussion

The present study demonstrates the preclinical potential of Cmpd. 1 as a bifunctional boronated agent that combines intrinsic antitumor activity with efficient boron delivery for BNCT. By evaluating two tumor models with distinct biological characteristics, cutaneous melanoma (CM) and non-small cell lung cancer (NSCLC), our findings also provide new insights into the influence of tumor biology on the therapeutic performance of EGFR-targeted boron carriers. Overall, the results support further development of EGFR-targeted boronated compounds as multifunctional agents for high-LET radiation-based therapies.

The differential EGFR mRNA expression observed between NSCLC and CM cells provided a biological basis for interpreting the distinct responses to Cmpd. 1. NCI-H125 cells exhibited markedly higher EGFR expression than Mel J cells, suggesting a greater dependence on EGFR-mediated signaling pathways. EGFR activation regulates downstream cascades, including the PI3K/AKT and RAS/RAF/MEK/ERK pathways, which play key roles in cell proliferation, survival, and resistance to apoptosis (Wee and Wang, 2017). Accordingly, the higher EGFR transcript levels observed in NCI-H125 cells may have contributed, at least in part, to their greater sensitivity to Cmpd. 1 under the experimental conditions evaluated (Aran and Omerovic, 2019). Although the present study did not directly assess EGFR pathway inhibition in Mel J or NCI-H125 cells, the previously reported EGFR inhibitory profile of Cmpd. 1 supports the hypothesis that EGFR-related signaling may contribute to its anti-tumor activity. This interpretation is also consistent with the well-established pharmacological relevance of EGFR inhibition in NSCLC (Shepherd et al., 2005; Rosell et al., 2012). In contrast, although CM cells expressed lower EGFR levels, they still responded to treatment with Cmpd. 1, indicating that its biological activity is not exclusively dependent on high EGFR expression. Moreover, despite the generally lower and more heterogeneous EGFR expression reported in CM, EGFR signaling has been associated with tumor progression, invasiveness, and therapeutic resistance in advanced stages of the disease (Girotti et al., 2013). These biological differences may partly explain the distinct responses observed between NSCLC and CM following treatment with the boronated erlotinib derivative.

Regarding LAT1, the major amino acid transporter responsible for BPA uptake, NSCLC cells exhibited higher LAT1 mRNA expression than melanoma cells. This finding is consistent with the greater BPA-mediated boron uptake observed in both the in vitro and in vivo studies. Increasing evidence indicates that the efficacy of BPA-based BNCT is strongly influenced by LAT1 (SLC7A5) expression, as this transporter mediates BPA internalization into tumor cells. Reduced LAT1 expression has been associated with lower intracellular boron accumulation, metabolic reprogramming, enhanced autophagy, and the development of resistance to BNCT (Lin et al., 2025). Therefore, the higher LAT1 expression observed in NSCLC cells likely contributed to the greater boron accumulation achieved with BPA in this tumor model. In contrast to BPA, the accumulation of Cmpd. 1 is unlikely to be directly determined by EGFR expression. Cmpd. 1 is a small-molecule tyrosine kinase inhibitor that interacts with the intracellular catalytic domain of EGFR and therefore must first cross the plasma membrane. Its cellular entry is expected to depend primarily on its physicochemical properties and ability to cross the plasma membrane, rather than on EGFR abundance itself. This may explain why Mel J and NCI-H125 cells showed similar total boron accumulation after 24 h despite their markedly different EGFR transcript levels. Interestingly, NCI-H125 cells showed somewhat higher boron accumulation at earlier time points, suggesting that the kinetics of uptake and/or retention may differ between the 2 cell lines before reaching similar levels after prolonged exposure. Moreover, ICP-OES measures total cell-associated boron and does not distinguish between EGFR-bound and unbound intracellular Cmpd. 1. Thus, EGFR abundance may have a greater influence on target engagement and biological response than on the total intracellular accumulation of the compound. Consistent with this interpretation, NCI-H125 cells were more sensitive to Cmpd. 1 than Mel J cells despite showing comparable boron accumulation after 24 h. Consequently, EGFR-targeted boronated inhibitors may complement the LAT1-dependent uptake mechanism of BPA, expanding the range of biological strategies available for selective boron delivery in BNCT.

A central requirement for BNCT efficacy is the achievement of sufficient intracellular boron levels. In this regard, Cmpd. 1 showed high boron delivery efficiency compared with BPA under the in vitro conditions evaluated. The significantly higher cellular boron uptake observed in both tumor models, particularly after normalization (56.65-fold in melanoma and 24.06-fold in NSCLC), confirms its superior delivery efficiency and supports its use as an effective boron carrier. The biological impact of this enhanced uptake was reflected in cell survival assays. BNCT significantly reduced viability in both tumor types, with a more pronounced effect in NSCLC, especially with Cmpd. 1. Interestingly, neutron irradiation alone also decreased NSCLC viability, suggesting a greater intrinsic sensitivity of these cells to neutron exposure. In addition, Cmpd. 1 alone induced significant reduction in NSCLC viability, indicating an intrinsic antiproliferative effect under the experimental conditions used. Given the previously reported EGFR inhibitory activity of Cmpd. 1 (Couto et al., 2018), modulation of EGFR signaling may contribute to this response by attenuating downstream pathways involved in cell proliferation and survival. Cell death analysis showed that BNCT predominantly induced necrosis in both tumor models, consistent with the high-LET radiation generated during the boron neutron capture reaction. Nevertheless, NSCLC cells displayed a greater apoptotic component than melanoma cells, particularly after treatment with Cmpd. 1 and Cmpd. 1-Mediated BNCT. These findings indicate that the cellular response to Cmpd. 1 differs between both tumor models. The increased apoptotic response observed in NSCLC cells could be related to inhibition of EGFR signaling by Cmpd. 1, leading to suppression of downstream pathways such as PI3K/AKT and RAS/RAF/MEK/ERK, which are known to promote cell survival, inhibit apoptosis, and drive tumor progression.

The in vivo studies further supported the therapeutic potential of Cmpd. 1. Repeated administration of the compound significantly delayed tumor growth in both CM and NSCLC xenograft models in the absence of neutron irradiation, demonstrating that Cmpd. 1 retains intrinsic antitumor activity in vivo. Although the underlying mechanism was not investigated in the present study, this effect is consistent with the known antiproliferative activity of EGFR tyrosine kinase inhibitors. In addition, body weight remained stable throughout the treatment period, suggesting that the treatment was well tolerated and produced no evidence of overt systemic toxicity based on this parameter. Boron biodistribution was evaluated at the end of the same repeated-dose treatment protocol employed in the in vivo antitumor efficacy study. Therefore, residual Cmpd. 1 from previous administrations cannot be completely excluded and may have contributed to the total boron concentrations measured. This represents a limitation of the present experimental design, which was intended to assess tissue boron levels at the completion of treatment rather than the single-dose biodistribution of Cmpd. 1 in animals that had not previously received the compound. In the NSCLC model, the addition of Cmpd. 1 resulted in a higher mean tumor boron concentration compared with BPA alone. The decision to evaluate Cmpd. 1 in combination with BPA was based on its bifunctional nature, combining intrinsic antitumor activity with boron-delivery capacity, and on our previous experience with a structurally related boron-containing lapatinib derivative. In that study, intraperitoneal administration of the compound alone at 10 mg/kg resulted in low tumor boron levels (Couto et al., 2026b). Considering this previous observation and the considerably lower dose of Cmpd. 1 used here (1 mg/kg), evaluation of Cmpd. 1 alone using the same intraperitoneal administration protocol was considered likely to provide limited information. Furthermore, the in vitro uptake studies demonstrated efficient boron delivery by Cmpd. 1 alone in both tumor models despite the substantially lower boron concentration used compared with BPA. Therefore, biodistribution was assessed following co-administration with BPA to evaluate the potential contribution of Cmpd. 1 to tumor boron accumulation under the experimental conditions used. A higher mean tumor boron concentration was observed in the NSCLC model followingesto co-administration of Cmpd. 1 and BPA compared with BPA alone, whereas no appreciable difference was observed in CM. Nevertheless, the present experimental design does not establish the independent in vivo biodistribution profile of Cmpd. 1 or allow direct comparison of its tumor selectivity with that of BPA. Future studies should therefore optimize the administration protocol for Cmpd. 1, including alternative intravenous or intratumoral routes, before evaluating its independent biodistribution and tumor-to-normal tissue and tumor-to-blood ratios. Taken together, the present findings show that Cmpd. 1 retains intrinsic antitumor activity in melanoma and NSCLC xenograft models and can efficiently deliver boron to tumor cells in vitro. The biodistribution studies further show that Cmpd. 1 can be incorporated into a BPA-based treatment regimen without preventing preferential tumor boron accumulation. Together with its previously reported EGFR inhibitory activity, these results support the further investigation of boronated EGFR inhibitors as multifunctional agents combining pharmacological activity with boron delivery for BNCT.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. Generative AI (ChatGPT) was used to assist with language editing, rephrasing, and improving the clarity and readability of selected sections of the manuscript. It was not used to generate or analyze data, produce figures, or determine the scientific conclusions. All AI assisted text was critically reviewed, verified, and revised by the authors, who take full responsibility for the final content.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Antonello Merlino, University of Naples Federico II, Italy

Reviewed by: Komal Anjum, Ocean University of China, China

Kazuki Miura, Institute of Innovative Research, Japan