Abstract
Background/Aim: Human papillomavirus-positive (HPV+) oropharyngeal squamous cell carcinoma (OPSCC) has a better outcome than HPV-negative (HPV−) OPSCC, but not all are cured with chemoradiotherapy, thus new therapies are required. Recently, we showed that e.g., phosphoinositide 3-kinase (PI3K) and fibroblast growth factor receptor (FGFR) inhibitors reduced viability in both HPV+ and HPV− OPSCC cell lines. Here, to better mimic the in vivo environment, the effects of supernatants of a normal fibroblast cell line on the responses of HPV+/HPV− OPSCC cell lines to various drugs were examined.
Materials and Methods: HPV+ (CU-OP-2, CU-OP-20) and HPV− (CU-OP-17) OPSCC cell lines were treated for 72 h with PI3K (BYL719), FGFR (JNJ-42756493), cyclin-dependent kinase (CDK) 4/6, (PD-0332991) and AKT (AZD-5363) inhibitors as well as cisplatin and docetaxel, with/without supernatants from the normal BJ-hTERT fibroblast, or the cancer-associated fibroblast (CAF) KS35 cell lines. Effects on viability and cell confluence/proliferation were then analyzed.
Results: All drugs abrogated viability and confluence in all cell lines. Combining BJ-hTERT cell supernatants with BYL719, JNJ-42756493 and PD-0332991 ameliorated their potential to reduce viability in some cell lines, while this was not the case for AZD-5363 and cisplatin. Docetaxel-induced cell death was only slightly affected, and only in CU-OP-20 cells. BJ-hTERT cell supernatants mainly had analogous effects on confluence to those observed on viability. CAF supernatants had more variable results due to the influence of growth factors needed and added to their media and were not pursued further.
Conclusion: Co-administering fibroblast BJ-hTERT cell supernatants with BYL719, JNJ-42756493, and PD-0332991 had an inhibitory effect on their effects on viability and confluence in some cell lines. This was never the case with cisplatin, and very rarely the case with docetaxel or AZD-5363.
Introduction
Human papillomavirus-positive (HPV+) oropharyngeal squamous cell carcinoma (OPSCC) has increased in incidence and has a better prognosis than HPV-negative (HPV−) OPSCC (1-10). Chemoradiotherapy, given irrespective of tumor HPV status, fails dismally to cure all patients (9, 10). To improve survival, research has focused on identifying prognostic/targetable biomarkers in HPV+ OPSCC (11-15). Mutations were thereby found in e.g., the phosphatidylinositol-4,5-bisphosphate 3-kinase, catalytic subunit alpha (PIK3CA) and the fibroblast growth factor receptor (FGFR) 3 (11-15). Notably, in breast and urothelial cancer with PIK3CA or FGFR3 mutations/translocations, the Food and Drug Administration (FDA) has approved PI3K and FGFR inhibitors (16, 17). We therefore tested e.g., phosphoinositide 3-kinase (PI3K), FGFR and CDK4/6 inhibitors and found dose-dependent effects on OPSCC cell lines (18, 19). Moreover, when combining the drugs with each other or cisplatin and docetaxel used for OPSCC therapy, synergistic effects were revealed (18, 19). Later, to better mimic in vivo conditions, some inhibitors were tested in a spheroid culture (20). Here, to further mimic in vivo conditions, components of the microenvironment were studied. Normal fibroblast- and cancer-associated fibroblast (CAF)-secreted factors (21-24) were examined for their effect on treatment responses to PI3K (BYL719), FGFR (JNJ-42756493), AKT (AZD-5363) and CDK4/6 (PD-03329991) inhibitors, cisplatin and docetaxel in HPV+/HPV− OPSCC cell lines.
Materials and Methods
Cell lines and seeding. OPSCC cell lines, HPV+ CU-OP-2, CU-OP-20 and HPV− CU-OP-17, provided by Dr N. Powell, Cardiff University, UK, adapted to growth without feeders (20, 25, 26) were cultured in GMEM (Sigma Aldrich, Gillingham, UK) with 10% fetal bovine serum (FBS; Gibco: Invitrogen Corporation, Paisly, UK), 500 ng/ml Hydrocortisone, 0.1 nM Cholera toxin, and 1x Penicillin/ Streptomycin (Sigma-Aldrich). For monolayer (2D) analysis, cells were seeded at 5,000 cells/well with 90 μl of medium in 96-well plates and treated 24 h after seeding. CU-OP-20 spheroids (3D) were generated in Thermo Scientific™ Nunclon Sphera 96-Well, Nunclon Sphera-Treated, U-Shaped-Bottom Microplates (ThermoFischer, Stockholm, Sweden) as reported previously (20). Seeding 2,000 cells/well in 100 μl GMEM medium resulted in the formation of a spheroid with a diameter of 300-400 μm after 72 h, then treatment was administered for an additional 72 h. The normal fibroblast cell line BJ-hTERT kindly obtained from Prof. A. Östman Karolinska Institutet, Stockholm, Sweden, was cultured in (D-MEM, Gibco), and the primary CAFs, obtained from a head and neck cancer patient, KS35, was cultured in keratinocyte-SFM supplemented medium (ThermoFischer Scientific). Both media were supplemented with 1X Penicillin/Streptomycin and 10% FBS. All lines were grown at 37°C with 5% CO2.
Drugs and treatments. PI3K, FGFR, CDK4/6 and AKT inhibitors BYL719, JNJ-42756493, PD-0332991 and AZD-5363, respectively, were purchased from TargetMol (San Diego, CA, USA) at a concentration of 10 mM in DMSO. Working dilutions (0.5-20 μM) were prepared in PBS. Cisplatin and docetaxel from Accord Healthcare Limited (Middlesex, UK) were diluted in PBS and used at 5-20 μM.
Supernatant preparation and administration. BJ-hTERT and KS35 cell supernatants were prepared by changing the medium 24 h after cell passaging and leaving new medium on for 72 h. Supernatants were then collected, centrifuged at 300 × g for 5 min and stored at −20°C until use. Before treatment, OPSCC medium was replaced with either conditioned, i.e., BJ-hTERT or KS35 cell supernatant, or control medium. Conditioned medium consisted of ¾ GMEM with ¼ supernatant from BJ-hTERT or KS35 cells. Control medium consisted of ¾ GMEM with ¼ D-MEM for fibroblasts or keratinocyte-SFM supplemented medium for CAFs.
WST-1 viability assay. Viability in 2D cultures was measured 24-72 h after treatment using the WST-1 assay (Roche Diagnostics, Mannheim, Germany) according to the manufacturer’s instructions, as reported earlier (18-20).
Cell confluence assay. Cell confluence/proliferation was assessed using the IncuCyte S3 Live® Cell Analysis System (Essen Bioscience, Welwyn Garden City, UK). Images were taken every 2 h, as described earlier (18-20). Proliferation was determined by calculating the percentage of covered area in each image using the Incucyte software version 2024A.
RealTime-Glo MT cell viability assay. Viability of spheroids was measured 24-72 h after treatment using RealTime-Glo MT Cell viability assay (Promega, Stockholm, Sweden) according to the manufacturer’s instructions.
Spheroid size during treatment. Real-time imaging was done using the Incucyte® SX5 Live-Cell Analysis Instrument from Sartorius (Sartorius, Essen Bioscience, Göttingen, Germany) as described before (20). Spheroid size was determined by calculating the area of the largest object in each image using the Incucyte software version 2024A. Of three repeated experiments, a representative is shown.
Statistical analysis. To verify effects w/wo cell supernatant, a two-way ANOVA followed by Šidák’s correction for multiple comparisons was used. IC50 values were calculated using a non-linear regression model.
Results
Effects of PI3K (BYL719), FGFR (JNJ-42756493), CDK 4/6 (PD-0332991) and AKT (AZD-5363) inhibitors, cisplatin and docetaxel with/without (w/wo) normal fibroblast supernatants on OPSCC cell viability. HPV+ CU-OP-2 and −20 and HPV− CU-OP-17, w/wo BJ-hTERT cell supernatants, were exposed to 0.5-10 μM BYL719, JNJ-42756493, PD-0332991 and AZD-5363, or 5-20 μM cisplatin and docetaxel, and viability was followed for 72 h. All absorbance values were normalized and compared to the respective PBS control.
All CU-OP lines without BJ-hTERT supernatants showed dose-dependent decreases in viability analogous to previous studies with BYL719, JNJ-42756493, and PD-0332991 (18-20). AZD-5363, not tested before, was efficient, with the HPV+ lines being more sensitive (Figure 1, Figure 2, and Figure 3). BJ-hTERT supernatants modulated the responses as shown below, in Figure 1, Figure 2, and Figure 3 for CU-OP-2, 20, and 17 cells, respectively, and Table I.
WST-1 viability assay for CU-OP-2 cells treated with targeted therapies and chemotherapeutics ± BJ-hTERT fibroblast supernatant. Cell viability was measured after 24, 48, and 72 h of treatment with: (A) PI3K inhibitor BYL719; (B) FGFR inhibitor JNJ-42756493; (C) CDK4/6 inhibitor PD-0332991; (D) AKT inhibitor AZD-5363; (E) cisplatin; and (F) docetaxel, each with or without BJ-hTERT supernatant. Ratio of viability (drug ± supernatant) to control is shown for (G) 10 μM JNJ-42756493 and (H) 10 μM PD-0332991.
WST-1 viability assay for CU-OP-20 cells treated with targeted therapies and chemotherapeutics ± BJ-hTERT fibroblast supernatant. Cell viability was assessed as in Figure 1 for: (A) BYL719; (B) JNJ-42756493; (C) PD-0332991; (D) AZD-5363; (E) cisplatin; and (F) docetaxel. Viability ratios (± supernatant) are shown for (G) 5 μM JNJ-42756493 and (H) 15 μM docetaxel.
WST-1 viability assay for CU-OP-17 cells treated with targeted therapies and chemotherapeutics ± BJ-hTERT fibroblast supernatant. As in Figures 1–2, viability was measured for: (A) BYL719; (B) JNJ-42756493; (C) PD-0332991; (D) AZD-5363; (E) cisplatin; and (F) docetaxel. Viability ratios (± supernatant) are shown for: (G, H) 5 and 10 μM BYL719; (I, J) 5 and 10 μM JNJ-42756493; and (K) 10 μM PD-0332991.
IC50 values for per time (hours) for CU-OP-2, 20 and CU-OP-17 with and without conditioned media of the normal fibroblast BJ-hTERT cell line.
BYL719: BJ-hTERT cell supernatants ameliorated the inhibition of viability of CU-OP-17 cells treated with 5 and 10 μM BYL719 (Figure 3A, G-H). BJ-hTERT cell supernatant had a similar effect on CU-OP-20 cells treated with 0.5 μM and 1 μM BYL719 (Figure 2A), while no effect was observed on CU-OP-2 cells (Figure 1A).
JNJ-42756493: BJ-hTERT cell supernatants reduced the effects of JNJ-42756493 in all cell lines at various time points (Figure 1B, Figure 2B and Figure 3B). CU-OP-20 and CU-OP-17 cells showed a significantly higher viability with 5 μM JNJ-42756493 in the presence of BJ-HTERT cell supernatants 72 h after treatment (Figure 2B and G and Figure 3B and I). Similarly, BJ-HTERT cell supernatant ameliorated the viability inhibition of 10 μM JNJ-42756493 in CU-OP-2 cells at 24 h (Figure 1B and G) and in CU-OP-17 cells at 48 and 72 h after treatment (Figure 3B and J).
PD-0332991: BJ-hTERT cell supernatants reduced the effects of PD-0332991 on all cells at some time point (Figure 1C, Figure 2C, and Figure 3C). More specifically, BJ-hTERT cell supernatant reduced cell death induced by 10 μM PD-0332991 in CU-OP-2 and CU-OP-17 cells, and a significant reduction was observed after 48 h in CU-OP-2 cells (Figure 1C and H and Figure 3C and K). A similar trend was observed for CU-OP-20 cells after 72 h treatment with 5 μM PD-0332991 (Figure 2C).
AZD-5363: BJ-hTERT cell supernatants did not modulate the response to AZD-5363 in any of the tested cell lines (Figure 1D, Figure 2D, and Figure 3D).
Cisplatin: BJ-hTERT cell supernatants did not modulate the response to cisplatin in any of the tested cell lines (Figure 1E, Figure 2E, and Figure 3E).
Docetaxel: BJ-hTERT cell supernatant significantly reduced death of CU-OP-20 cells treated for 72 h with 15 μM docetaxel (Figure 2F and H) but did not affect docetaxel effect in CU-OP-2 and −17 cells (Figure 1F and Figure 3F).
In summary, the BJ-hTERT cell supernatant modulated the sensitivity of the cell lines to several inhibitors, with the strongest effect observed for the FGFR inhibitor JNJ-42753493. The HPV− CU-OP-17 cell line was the most affected cell line (Figure 1, Figure 2, and Figure 3). This is further reflected in the IC50 values (Table I). In contrast, AZD-5363 and cisplatin were not, and docetaxel rarely affected by the BJ-hTERT cell supernatant (Figure 1, Figure 2, and Figure 3).
Effects of PI3K, FGFR, CDK4/6 and AKT inhibitors, cisplatin and docetaxel w/wo CAF supernatants on HPV+ and HPV− OPSCC cell viability. KS35 supernatants were also tested for potential inhibition of viability induced by the above drugs and all three OPSCC lines. However, the large quantity of growth factors added in the CAF medium led to a high variability in the responses of CU-OP cells, and further experiments were therefore discontinued (data not shown).
Effects of PI3K (BYL719), FGFR (JNJ-42756493), CDK) 4/6, (PD-0332991) and AKT (AZD-5363) inhibitors, cisplatin and docetaxel w/wo normal fibroblast supernatants on proliferation/confluence of HPV+ CU-OP-20 and HPV− CU-OP-17 cells. The effect of BJ-hTERT cell supernatant on the growth inhibitory activity of the inhibitors in CU-OP-17 and CU-OP-20 cells (i.e., a sensitive and less sensitive line with regard to viability), was examined by monitoring confluence for 72 h following treatment with 1-10 μM of each inhibitor and 1-20 μM of cisplatin and docetaxel (Figure 4 and Figure 5).
Confluence analysis of CU-OP-17 cells treated with targeted therapies and chemotherapeutics ± BJ-hTERT fibroblast supernatant. Measured by IncuCyte S3, confluence was recorded for: (A) BYL719; (B) JNJ-42756493; (C) PD-0332991; (D) AZD-5363; (E) cisplatin; and (F) docetaxel. Confluence ratios (± supernatant) are shown for: (G) 1 μM BYL719 and (H) 5 μM PD-0332991.
Confluence analysis of CU-OP-20 cells treated with targeted therapies and chemotherapeutics ± BJ-hTERT fibroblast supernatant. Confluence was monitored as in Figure 4 for: (A) BYL719; (B) JNJ-42756493; (C) PD-0332991; (D) AZD-5363; (E) cisplatin; and (F) docetaxel. Confluence ratios (± supernatant) are shown for: (G) 5 μM PD-0332991 and (H, I) 5 and 10 μM AZD-5363.
CU-OP-17. The BJ-hTERT cell supernatant limited the effect of the inhibitors on CU-OP-17 cell growth compared to their effects as single agents (Figure 4). More specifically, only a general decrease in inhibition of growth was observed with 1 μM BYL719, which was significant at 48 and 72 h, whereas 5 μM PD-0332991 exerted a significantly lower inhibition at 48 h (Figure 4A, C and G-H).
CU-OP-20. The BJ-hTERT cell supernatant decreased the growth inhibitory effect of 5 μM PD-0332991 on CU-OP-20 cells after treatment for 48 and 72 h (Figure 5C and G). Unexpectedly, the BJ-hTERT cell supernatant also modulated the response of CU-OP-20 cells to 5 and 10 μM AZD-5363 48 and 72h after treatment (Figure 5D and H-I).
In summary, the effects of the BJ-HTERT cell supernatant on the reduction of confluence induced by the drugs in CU-OP-17 and CU-OP-20 cells, were analogous to those observed for viability with two exceptions. The BJ-hTERT cell supernatant did not reduce the inhibition of proliferation induced by JNJ-42756493 on either cell lines, while it showed some effect on the response to AZD-5363 in CU-OP-20 cells.
Effects on viability and confluence of PI3K (BYL719), FGFR (JNJ-42756493), CDK 4/6, (PD-0332991) and AKT (AZD-5363) inhibitors on HPV+ CU-OP-20 spheroid cultures, w/wo normal fibroblast supernatants. To further model the effects of the supernatant on the inhibitors in an in vivo setting, HPV+ CU-OP-20 cells were selected since they were relatively sensitive in both viability and confluence assays.
Viability. Analogous to 2D CU-OP-20 cells, JNJ-42756493 efficacy was most affected by the BJ-hTERT cell supernatants (Figure 6A-G). Notably, with 10 μM JNJ-42756493, viability remained high throughout the whole experiment in the presence of BJ-hTERT cell supernatant (Figure 6E). A slight modulation of the inhibition of viability was also observed with 20 μM PD-0332991 24 h after treatment (Figure 6C and G).
Viability and morphological effects of targeted inhibitors ± BJ-hTERT supernatant on CU-OP-20 spheroids. Spheroid viability was measured after 24, 48, and 72 h using RealTime-Glo for: (A) BYL719; (B) JNJ-42756493; (C) PD-0332991; and (D) AZD-5363, each ± BJ-hTERT supernatant. Viability ratios (± supernatant) are shown for: (E, F) 10 and 20 μM JNJ-42756493; (G) 20 μM PD-0332991. (H) Morphological changes in spheroids after 72 h treatment with 10 μM of each drug, ± supernatant, compared to PBS control.
Growth. A difference in growth inhibition w/wo the supernatant was observed for 10 μM of BYL719 and AZD-5363, which resulted in less intact spheroids, with a large halo structure, compared to the non-supernatant-treated spheroids (Figure 6H). A larger halo formation was also seen with 10 μM JNJ-42756493 (Figure 6H).
To summarize, the BJ-hTERT cell supernatant reduced the effects of JNJ-42756493 and PD-0332991 at some time points in 3D CU-OP-20 cell cultures, similar to 2D CU-OP-20 cultures. Moreover, the BJ-hTERT cell supernatant sometimes reduced the effects of BYL719, JNJ-42756493 and AZD-5363 on 3D CU-OP-20 confluence.
Discussion
In this pilot study, combining normal fibroblast BJ-hTERT supernatants with BYL719, JNJ-42756493, AZD-5363 and PD-0332991 reduced the effectiveness of the drugs in reducing cell viability and proliferation in the tested OPSCC cell lines. On the contrary, the effect of cisplatin was unchanged, and only a minimal effect was observed with docetaxel in the HPV+ CU-OP-20 line. Mainly analogous effects with BJ-hTERT cell supernatant were observed on the effects of the above inhibitors on proliferation. Conversely, CAF supernatants induced variable results due to the influence of growth factors added in their media and were therefore excluded from further analysis.
Notably, the FGFR inhibitor (JNJ-42756493) was affected the most, very likely due to the BJ-hTERT fibroblast-secreted fibroblast growth factors (FGFs), which potentially abrogated the effect of this inhibitor (21). Importantly, FGF signaling through FGFRs also activates the PI3K/AKT signaling pathway (27, 28), which could explain why the cell supernatant also affected the inhibition induced by the PI3K and AKT inhibitors. Moreover, it has also been shown in breast cancer that FGFR signaling can lead to CDK4/6 inhibitor resistance, including PD-0332991, but that it could be rescued by FGFR inhibition with JNJ-42756493 (29). The fact that both the PI3K and CDK4/6 inhibitors were influenced should be taken into account as these two drugs are often used in combination in e.g., breast cancer (30).
Cisplatin and docetaxel were not (or mainly not and then very slightly) affected by the BJ-hTERT cell supernatant, which might be expected since they are not involved in the investigated targeted signaling pathways. Of note, while CAFs have been shown to promote head and neck cancer cell proliferation (31), thereby potentially sensitizing the cells to cisplatin and docetaxel that primarily target rapidly dividing cells, this was not observed with the fibroblast supernatant in our study.
This pilot study, however, has its limitations, since the effect of a supernatant from one normal fibroblast type on a small array of targeted therapies and chemotherapeutic drugs was studied on only three OPSCC cell lines. Moreover, apart from the secreted factors from the fibroblasts and CAFs, these cells could exert additional effects that could modulate drug responses, which could only be studied by performing co-culture experiments.
However, this study suggests that the microenvironment should not be neglected when studying targeted therapies. Nevertheless, more investigations are warranted with additional fibroblast/CAF supernatants and cell lines, including co-culture studies, with several drugs to further explore the effects on drug responses in more detail.
Acknowledgements
The Authors thank Professor Arne Östman and Dr Neil Powell for kindly providing us with some of the cell lines.
Footnotes
Authors’ Contributions
Conceptualization, M.B., O.K., and T.D.; methodology, M.B., T.C., O.K., E.W., and T.D.; software, M.B., TC, and O.K.; validation, M.B., T.C., O.K., E.W., and TD.; formal analysis, M.B., T.C., and T.D.; investigation, M.B., T.C., O.K., E.W., and T.D.; resources, M.B., O.K., and T.D.; data curation, M.B., T.C., and T.D.; writing – original draft preparation M.B., T.C., and T.D.; writing – review and editing, M.B., T.C., O.K., E.W., and TD; visualization, M.B., and T.C. supervision, M.B., O.K., and TD.; project administration, M.B., and T.D.; funding acquisition, M.B., O.K., and T.D. All Authors have read and agreed to the submitted version of the manuscript.
Conflicts of Interest
The Authors declare no conflicts of interest in relation to this study.
Funding
This research was funded by the Swedish Cancer Foundation (grant no. 23 2700), the Stockholm Cancer Society (grant no. 241131), the Stockholm City Council and Karolinska Institutet.
Artificial Intelligence (AI) Disclosure
AI tools were not used to generate this manuscript.
- Received July 15, 2025.
- Revision received July 30, 2025.
- Accepted July 31, 2025.
- Copyright © 2025 The Author(s). Published by the International Institute of Anticancer Research.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.












