P. hadiensis was provided by the Kirstenbosch National Botanical Gardens (Kirstenbosch, South Africa) and grown in Parque Botânico da Tapada da Ajuda (Lisbon, Portugal). The plant was air-dried at room temperature and stored in cardboard boxes protected from light and humidity to maintain its stability as previously described by our team (Domínguez-Martín et al., 2022; Magalhães et al., 2024).

The abietane diterpene, Roy (Figure 1), was isolated from the acetonic extract of P. hadiensis, as previously described by our team (Domínguez-Martín et al., 2022; Magalhães et al., 2024). Roy was purified by dry-column flash-chromatography, using silica as a stationary phase and a gradient of eluents of increasing polarity (n-Hexane:Ethyl acetate) as a mobile phase. The structure of both Roy was elucidated by nuclear magnetic resonance (NMR) spectra assignments and previously published (Ntungwe et al., 2021).

Roy-Bodipy was synthesized following the protocol previously described by our team (Domínguez-Martín et al., 2022). Briefly, succinic anhydride was added to a mixture of BODIPY (fluorescent probe) and triethylamine in anhydrous dichloromethane. The reaction mixture was stirred at room temperature. The crude material of succinimidyl-modified BODIPY was employed in the subsequent stage without undergoing additional purification. A mixture of carboxylic acid (succinimidyl-modified BODIPY) and abietane in anhydrous dichloromethane was treated with 1-[3-(dimethylamino)propyl]-3-ethylcarbodiimide methiodide (EDC) and a catalytic amount of 4-(Dimethylamino)pyridine (DMAP). The solution was diluted and washed. The crude material was purified through silica column chromatography to yield Roy-BODIPY.

U87 (HTB-14; brain-likely GB) and HMC3 (CRL-3304; microglia (MG)) cell lines were acquired at American Type Culture Collection (ATCC) and provided by Prof. Carla Vitorino (Faculty of Pharmacy, University of Coimbra) and Prof. Célia Gomes (Coimbra Institute for Clinical and Biomedical Research (iCBR), Faculty of Medicine, University of Coimbra), respectively. Cells were cultivated in Dulbecco’s Modified Eagle’s Medium-high glucose (DMEM-HG) (Biowest, Nuaillé, France) supplemented with 10% (v/v) of heat-inactivated fetal bovine serum (FBS) (Biowest, Nuaillé, France) and 1% (v/v) of penicillin-streptomycin (Sigma, St. Louis, MO, United States). Human brain capillary endothelial cells (HBMEC) were purchased from Innoprot (Bizkaia, Spain) and kindly provided by Prof. Carla Vitorino (Faculty of Pharmacy, University of Coimbra). HBMEC were grown in EBM™-2 Basal Medium (CC-3156) and EGM™-2 SingleQuots™ Supplements (CC-4176) (Lonza Group AG, Basel, Switzerland) in T75 flasks coated with 0.3 mg/mL rat tail collagen type I (BD Biosciences). According to the supplier recommendations, HBMEC cells were used until a maximum of 15 passages, counting from the initial vial. All cells were maintained at 37 °C and 5% CO₂ and subcultured by trypsinization when reaching 80%–90% confluence. All experiments were performed in cultures in log phase growth.

For all experiments, cells were treated with a single administration of 16 μM of Roy, a concentration previously shown in our work to exhibit cytotoxic effects in GB cells but not in non-tumoral cells (HMC3 cells and Primary cultures of astrocyte-enriched cells), indicating a tumor-specific cytotoxicity (Magalhães et al., 2024).

HBMEC cells (20 × 10³ cells/well) were seeded in a 96-well plate, 24 h before treatment. Following, cells were treated with 16 μM of Roy and further incubated for 4 h. Metabolic activity was assessed through a modified Alamar blue® assay, as previously described (Magalhães et al., 2024). Accordingly, 100 μL of a solution of EBM™-2 Basal Medium with 10% (v/v) of a stock solution of resazurin was added to the cells and incubated for 2 h, at 37 °C. The absorbance of the plate was read at 570 and 600 nm in a BioTek (BioTek Instruments, Inc., Winooski, VT, United States). The absorbance results were obtained by the Gen5 program. Metabolic activity was then calculated by Equation 1: M e t a b o l i c   a c t i v i t y   % = A 570 − A 600   o f   t r e a t e d   c e l l s A 570 − A 600   o f   c o n t r o l   c e l l s   × 100 % (1)

To establish an in vitro model of the BBB, HBMEC cells (5 × 10⁴ cells/cm²) were seeded on the apical side of transwell inserts (in 6-multiwell collagen pre-coated Transwell®, 0.4 μm pore PTFE membrane insert (Corning, Glendale, AZ, United States) and grown for 7 days at 37 °C and 5% CO₂. Cell medium was changed every 2 days. The formation of tight junctions (TJ) was monitored by measuring the transendothelial electrical resistance (TEER) of the cell monolayer by using an EVOM resistance meter (World Precision Instruments, Hertfordshire, UK). On day 7 of HBMEC culture in the transwell system, 16 µM of Roy-Bodipy was added to the apical (upper) compartment of the transwell. After 0.5, 1, 2, and 4 h, the medium in the basolateral (lower) compartment was collected and analyzed in a BioTek device (BioTek Instruments, Inc., Winooski, VT, United States) by measuring fluorescence at excitation and emission wavelengths 485 and 528 nm, respectively, for Roy-Bodipy determination. The initial concentration of Roy-Bodipy (16 µM) added to the upper compartment was considered as total of Roy-Bodipy, while the final concentration of Roy-Bodipy measured in the lower compartment was expressed as the percentage of Roy-Bodipy in the apical compartment at time 0. Permeability under steady-state conditions can be evaluated mathematically by the apparent permeability coefficient (Papp), according to Equation 2: P a p p = d Q d t 1 A C 0 (2) where dQ/dt is the solute flux (μg/s) across the barrier, A is the surface area of the transwell (cm²), and C₀ is the initial donor concentration (μg/mL). Whereas transport efficiency (TE) of Roy-Bodipy across the membrane was measured using Equation 3, as described by Pogačnik et al. (2016): T E = C l o w C u p × 100 % (3) where C_low is the concentration of Roy-Bodipy in the basolateral compartment and C_up is the concentration of Roy-Bodipy in the upper compartment, at the end of the experiment.

The permeability of sodium fluorescein (Na-F) (376.27 Da) was measured to confirm the integrity of the membrane as previously described (Pogačnik et al., 2016). 10 μg/mL of a Na-F solution, pH 7.4, was added to the upper compartment for 1 h. After 30 and 60 min, the medium in the lower chamber was collected and the concentration of Na-F was determined by fluorescence (excitation, 440 nm; emission, 525 nm) using a BioTek device (BioTek Instruments, Inc., Winooski, VT, United States).

Spheroids were grown in mono-, dual-, and triple-culture using U87 cells, U87 and HMC3 cells, and U87, HMC3, and HBMEC cells, respectively. For mono-culture, U87 cells were seeded at a density of 8 × 10³ cells per well, while for dual-culture and triple-culture, cells were seeded at a 1:1 ratio (U87:HMC3) and 2:1:0.2 ratio (U87:HMC3:HBMEC) per well, respectively, in ultra-low adherence round-bottomed 96-well plates (Corning, Glendale, AZ, United States) and further incubated for 72 h (Pires et al., 2012; Martins et al., 2023). At least 10 spheroids were generated per condition. After aggregation, spheroids were treated with a single administration of 16 µM of Roy, and media was replaced every 2 days for a total duration of 14-day treatment.

Spheroids were imaged, at days 0, 7, and 14, using the Motic Images Plus 3.0 (Barcelona, Spain), and spheroid size and circularity were determined using the ImageJ/Fiji software (NIH, United States) (Pires et al., 2012; Martins et al., 2023). For metabolic activity assessment, at day 14, spheroids were incubated with a solution of DMEM-HG with 10% (v/v) of a resazurin stock solution (Sigma. St. Louis, MO, United States) (0.1 mg/mL) for 4 h at 37 °C. After incubation, the absorbance of the plate was read at 570 and 600 nm in a BioTek (BioTek Instruments, Inc., Winooski, VT, United States). Metabolic activity was calculated using Equation 1.

U87 and HMC3 cells (75 × 10³ cells/well) were seeded in 12-well plates, while co-culture of U87 and HMC3 cells was seeded at a final density of 15 × 10⁴ cells per well in 12-well plates. After 24 h, the medium was replaced by fresh medium, and cells were further incubated for 48 h. Secretome from the different cultures was collected and centrifuged at 1,500 rpm for 5 min, to remove potential cell debris. The conditioned medium was used in the metabolic activity assay.

U87 cells (1 × 10⁴ cells/well) were seeded in 96-well plates, 24 h before treatment. Cells were pre-incubated with non-diluted conditioned medium from U87 and/or HMC3 for 1 h. Subsequently, the conditioned medium was removed, and cells were treated with 16 µM of Roy and further incubated for 48 h. After treatment, the medium containing Roy was completely removed to minimize potential direct chemical reduction of resazurin by Roy. Metabolic activity of cells was assessed through a modified Alamar blue® assay as previously described. Briefly, a solution was prepared of DMEM-HG medium with 10% (v/v) of a resazurin salt dye stock solution at a 0.1 mg/mL concentration, which was further added to each well after 48 h treatment. After 4 h of incubation at 37 °C and 5% CO₂, the absorbance of the plate was read at 570 and 600 nm in a BioTek (BioTek Instruments, Inc., Winooski, VT, United States). The absorbance results were obtained by the Gen5 program. Metabolic activity was determined using Equation 1.

Cell morphology was assessed using an inverted microscope Motic AE2000 (Barcelona, Spain) and images were obtained using the Motic Images Plus 3.0 (Barcelona, Spain).

Cytokines were quantified using a customized detection panel purchased from Bio-techne (R&D Systems, Inc., Minneapolis, MN, United States), as previously described (Valerii et al., 2021). Briefly, U87 cells (40 × 10³ cells/well) were seeded in 24-well plates. After 24 h, cells were treated with 16 µM of Roy and, further, incubated for 48 h. The supernatant was collected and centrifuged at 4000 RCF for 10 min. After centrifugation, the supernatant was collected into new tubes and stored at −80 °C until use.

The inflammatory cytokines evaluated were IL-1β, IL-6, TNFα, IL-4, IL-8, and IL-10. The assay was performed in 96-well filter plates by multiplexed Luminex® (Luminex Corp., Austin, TX, United States) immunoassay according to the manufacturer’s instructions. Microsphere magnetic beads coated with monoclonal antibodies against the different target analytes were added to the wells and incubated for 30 min. After, the wells were washed and biotinylated secondary antibodies were added. After 30 min, beads were washed and then incubated for another 10 min with streptavidin-PE conjugated to the fluorescent protein, phycoerythrin (streptavidin/phycoerythrin). After washing, the beads (a minimum of 100 per analyte) were analyzed in the BioPlex 200 instrument (BioRad®, Hercules, CA, United States). Sample concentrations were estimated from the standard curve using a fifth-order polynomial equation and expressed as pg/mL (Supplementary Figure S1) after adjusting for the dilution factor (Bio-Plex Manager software 5.0). Samples below the detection limit of the assay were recorded as zero, while samples above the upper limit of quantification of the standard curves were assigned the highest value of the curve.

Expression levels of Vascular Endothelial Growth Factor A (VEGFA), Signal Transducer and Activator Of Transcription 3 (STAT3), STAT5A, STAT5B, Janus kinase 2 (JAK2), IL6, IL8 and Cyclin-dependent kinase 4 (CDK4) mRNA were assessed by quantitative real-time PCR (qRT-PCR). U87 cells were seeded with a density of 1 × 10⁶ cells per well 24 h before treatment. After, cells were treated with 16 µM of Roy, and further incubated for 48 h. Total RNA was extracted using TripleXtractor solution (GRISP, Lisbon, Portugal) according to the manufacturer’s protocol. RNA was converted into cDNA through the Xpert cDNA Synthesis Supermix (GRISP, Lisbon, Portugal), following the manufacturer’s protocol and 100 ng of cDNA were amplified by qRT-PCR using the primer sequences described in Supplementary Table S1. Each optimized reaction was performed using Xpert Fast SYBR Green Mastermix 2X with ROX (GRISP, Lisbon, Portugal) and samples were subjected to the amplification protocol described by the manufacturer, using a melting temperature of 60 °C. Relative gene expression was determined by the 2^−ΔΔCT method and normalized to the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene, the endogenous reference, and relative to the untreated control cells. Optimization conditions are described in Supplementary Table S2.

U87 cells (0.8 × 10³ cells/well) were seeded in 6-well plates. After 24 h, cells were treated with 16 µM of Roy and further incubated for 14 days with the media being changed every 2 days. Non-treated cells were considered the control group. After, cells were washed with PBS, fixed with 4% PFA in PBS for 15 min, and then stained with 0.1% crystal violet solution for 30 min. The surviving fraction (SF) was calculated as mean colonies/number of cells seeded. A minimum of 50 cells was required to define a colony.

Data were analyzed using GraphPad Prism v.10.1.0. All experiments are representative of at least three independent experiments and acquired results were expressed as mean ± standard deviation (SD). Statistical analysis was performed by t-student test, one-way, and two-way ANOVA, using the unpaired comparison and the multiple comparisons tests Tukey and Dunnett, respectively. A value of p < 0.05 was considered significant.

The BBB presents a significant challenge for treating GB, as it protects the brain from both exogenous and endogenous molecules, thereby hindering the entry and accumulation of most therapeutic agents in brain tissue, which results in lower treatment efficacy. In this study, we evaluated the transport efficiency of Roy across the BBB using the HBMEC model, after 4 h, and assessed its cytotoxicity, for the same incubation time, to ensure it does not damage the BBB cells (Figure 2). We used a concentration of 16 µM of Roy for the treatment, as, in our previous work, this concentration had previously demonstrated an antitumor effect on GB cells while being safe for non-tumor cells (MG and astrocytes) (Magalhães et al., 2024). Our findings indicated that this concentration of Roy did not cause damage to HBMEC cells, being considered safe (Figure 2A). Subsequently, we investigated Roy’s ability to cross the BBB in vitro. For this purpose, Roy was conjugated to the fluorescent probe Bodipy (Roy-Bodipy), enabling its quantification by fluorescence. Notably, conjugation with Bodipy did not alter Roy’s bioactivity (Data not shown). We used TEER and Na-F as markers to confirm BBB integrity. The permeability value of Na-F below 1.2 × 10⁻³ cm/min indicated no adverse effect on the cell monolayer integrity, which was confirmed by TEER values averaging 140–150 Ω.cm². The apparent permeability of Roy (Papp = 8.32 × 10⁻⁶ cm/s) and the effective concentration of this natural compound that permeated the membrane (expressed as TE = 70.55%) (Figure 2B) demonstrated Roy’s ability to cross the BBB.

To assess the effectiveness of Roy treatment in cell models that better mimic the TME, 3D multicellular spheroid models were employed. These models are widely recognized in vitro tools for investigating innovative therapeutic strategies against cancer due to their ability to simulate the TME in vitro (Pires et al., 2012; Zanoni et al., 2016). In this study, we generated a heterogeneous 3D spheroid model of GB microenvironment, incorporating its three main cellular components: tumor cells, immune cells, and endothelial cells. In primary GBs, immune cells can represent 30%–50% of the tumor mass and play a crucial role in promoting tumor proliferation, angiogenesis, and immune suppression (Larionova et al., 2021; Li et al., 2022; Chen et al., 2022). Endothelial cells, which constitute approximately 5%–30% of the tumor mass, are key mediators of tumor angiogenesis, being involved in the modulation of the pro-tumor immune response and therapeutic resistance (Martins et al., 2023; Leone et al., 2024). This 3D model therefore allowed us to assess the efficacy of Roy treatment within a more realistic tumor structure. To represent tumor cells, the well-characterized human GB cell line U87 was chosen, as recommended by ATCC and supported by our team’s prior experience using this model to develop GB spheroid systems (Pires et al., 2012). Additionally, HMC3 cells, a human MG cell line, and HBMEC, human brain microvascular endothelial cells, were selected to represent the immune and endothelial components, respectively, in this model.

Monocellular and multicellular spheroids were formed and treated with 16 µM of Roy for 14 days. We monitored spheroid size and circularity during this period (Figure 3B). Notably, treatment with Roy significantly reduced spheroid size after 7 and 14 days of treatment in multicellular spheroid cultures, whereas this significative effect was only observed after 14 days in mono-culture spheroids (Figure 3A). However, we found no significant differences in spheroid circularity after treatment (Figure 3B). Moreover, at day 14, treated spheroids exhibited a substantial decrease in metabolic activity and a marked reduction in size, accompanied by apparent structural changes, when compared to untreated spheroids, despite overall spheroid circularity being preserved (Figure 3). This underscores Roy’s strong cytotoxic and potential cytostatic effects, as well as, its potential to mitigate the inflammatory state of the GB microenvironment, demonstrating greater efficacy in complex tumor structures, reinforcing its promise as a natural lead compound.

As previously noted, Roy demonstrated the ability to inhibit the growth of tumor spheroids over time with a nearly complete spheroid disintegration by day 14 (Figure 3). Considering these findings, we intended to determine whether this lead compound could maintain its cytotoxic/antiproliferative activity in the presence of conditioned medium. To investigate this, we exposed GB cells to conditioned medium derived from cultures of GB cells, MG cells, or co-culture of GB and MG cells, and subsequently treated them with 16 µM of Roy (the same concentration used in the previous assays) (Figure 4). The results showed that Roy’s effectiveness increased, leading to a significant reduction in the metabolic activity of GB cells when compared to its activity in the presence of fresh medium (Figure 4A). Moreover, in the presence of conditioned medium from MG cells, Roy treatment caused a 40% reduction in metabolic activity compared to the control group (non-treated cells) (Figure 4A). In contrast, under normal conditions (using fresh media), Roy induced only a 20%–30% decrease in metabolic activity relative to the control (Figure 4A). This data was supported by changes in the morphology of treated cells that match with cells undergoing apoptosis (Figure 4B). Interestingly, Roy’s antitumor effect on spheroids (Figure 3) appears to be linked to its improved efficacy in the presence of a complex TME.

Various cytokines within the TME of GB play a crucial role in shaping immune and inflammatory responses, thus influencing tumor development, angiogenesis, and progression. Targeting signaling molecules associated with neuroinflammation processes may be a promising strategy to combat this invasive brain tumor. Building on our previous findings, where Roy demonstrated enhanced efficacy in models that better mimic the TME (Figures 3, 4), and considering the critical role of cytokines in shaping the GB microenvironment, we evaluated the impact of Roy treatment on the expression of cytokines secreted by GB cells that are involved in the tumorigenesis process.

Our findings suggested that Roy exhibited an immunomodulatory effect on GB cells, significantly reducing the levels of several cytokines involved in the progression of neuroinflammation, angiogenesis, and tumor growth, namely IL-1β, IL-6, IL-8, and TNF-α (Figure 5). Additionally, Roy treatment also led to a decrease in IL-4 and IL-10 levels, cytokines with a dual role, typically anti-inflammatory but seemingly pro-inflammatory in the context of GB (Figure 5). Clearly, Roy also displayed an immunomodulatory impact, reinforcing its therapeutic potential as observed in our previous experiments.

Based on the previous findings, we observed that Roy exhibited a strong antitumor activity, inhibiting tumor growth and possibly modulating immune-related targets. With this in mind, we assessed the mRNA levels of genes linked to the IL-6/JAK2/STAT3 signaling pathway, known to be overexpressed in GB and associated with uncontrolled proliferation, invasion, and survival. Following a 48-h treatment period with Roy, we analyzed the transcriptional profiles of IL6, JAK2, STAT3, STAT5A, and STAT5B in GB cells (Figure 6). Notably, we observed a significant downregulation of IL6, JAK2, STAT3, STAT5A, and STAT5B mRNA levels, all of which are found to be upregulated in GB cells, compared to the control group (untreated cells) (Figure 6). These findings align with the observed inhibition of GB immune-endothelial spheroids size (Figure 3) and the reduction in IL-6 levels secreted by GB cells (Figure 5), further supporting that Roy potentially suppresses IL6/JAK2/STAT3 pathway activity, consequently inhibiting GB proliferation and survival.

Considering the reduction in spheroid size following Roy treatment (Figure 3) and its effects on immune-related targets implicated in GB proliferation and invasion (Figures 5, 6), we next explored the antiproliferative activity of this lead compound on GB clonal growth (Figure 7A). By the results obtained, we observed that treatment with Roy led to almost complete impairment of colony formation as sustained by the reduced surviving fraction (SF) factor (Figure 7A). Based on this effect, significant inhibition of CDK4 mRNA levels, a gene encoding Cdk4 an important regulator of cell cycle progression, following Roy treatment was also observed (Figure 7B), supporting the antiproliferative and antitumor effect of this lead compound. Moreover, besides uncontrolled proliferation, GB is also characterized by extensive vascularization, leading to robust angiogenesis. Building upon our previous findings, it was also investigated how treatment with Roy affects the expression of VEGFA mRNA levels. As anticipated, Roy induced downregulation of VEGFA mRNA levels in GB cells (Figure 7C). These results align with the antitumor potential of Roy previously observed, suggesting an antiproliferative effect and ability to impair tumor growth and promote cell death.