Abstract
Background/Aim: Oral squamous cell carcinoma (OSCC) is characterized by early metastasis, clinical resistance and poor prognosis. Recently, we showed that aggressive OSCC cells co-express endothelial cell markers and can form tube-like structures, known as vasculogenic mimicry (VM), a process associated with poor prognosis in head and neck cancers. Given the limited success of current antiangiogenic therapy in treating OSCC, this study sought to explore the efficiency of these drugs in targeting an ex vivo model of VM. Materials and Methods: OSCC cell lines from the tongue and floor of the mouth in addition to human endothelial cells were used. The treatments comprised a set of clinically relevant antiangiogenic drugs: sorafenib, sunitinib, and axitinib, which were administered in different doses. Multiple ex vivo approaches including cell tubulogenesis, proliferation, apoptosis, and migration assays were used. Results: Although these drugs inhibited the formation of endothelial cell capillaries, they showed clear differential effects on OSCC cell-derived VM and cell morphology. Sorafenib inhibited the tubulogenesis of aggressive OSCC cells compared with the limited effect of sunitinib and axitinib. Furthermore, our data consistently demonstrated a preferential efficacy of certain drugs over others. Sorafenib and sunitinib exhibited anti-cancer effects on tumor cell proliferation, apoptosis, and cell migration, compared with the limited effect of axitinib. Conclusion: The antiangiogenic drugs, except sorafenib, had limited effect on VM formation in vitro and exhibited varying anti-cancer effects on OSCC cells. These data support the notion that VM formation may in part explain the development of drug resistance in OSCC cells.
Oral squamous cell carcinoma (OSCC) is the most common cancer of the head and neck area with an estimated annual occurrence of 370,000 cases and 177,000 deaths worldwide (1). OSCC arises in the buccal mucosa, gingiva, hard palate, tongue, and floor of the mouth. There are distinct geographic differences in the incidence and prevalence. However, the incidence rate has been increasing in the Scandinavian countries, including Finland (2). The primary treatment for OSCC is surgery with adjuvant radiotherapy, which can be combined with chemotherapy in advanced cases. In addition, targeted and immune therapies are indicated for patients with locoregionally advanced, metastatic, and recurrent disease (3, 4). Despite the multimodality approach in OSCC management, the 5-year survival rate remains low at approximately 55%, which has mainly been attributed to the early dissemination of the tumor cells (2, 4, 5).
Angiogenesis is a hallmark of cancer that mediates tumor progression and metastasis in solid tumors, including OSCC (6, 7). In addition to conventional angiogenesis, vasculogenic mimicry (VM) has been identified as an additional vascularization pattern, supplying oxygen and nutrients to growing tumors (8). VM is a phenomenon where aggressive tumor cells reorganize and transdifferentiate to form tube-like structures in the absence of endothelial cells (8). The exact mechanisms behind VM formation remain partly unknown. A hypoxic environment and epithelial-to-mesenchymal transition (EMT) are linked with VM (9). Importantly, VM has been associated with advanced cancer stages and poor prognosis in HNSCC (10).
Given the importance of tumor vasculature, antiangiogenic treatment (AAT) has been approved by the US Food and Drug Administration (FDA) to treat a few solid tumors including renal, colorectal, and hepatocellular carcinomas (HCC) (7). AAT, such as the monoclonal antibody bevacizumab and tyrosine kinase inhibitors (TKI; e.g., sorafenib, sunitinib, and axitinib), were designed to target especially the vascular endothelial growth factor (VEGF) pathway (7, 11). However, there are currently no clinically approved antiangiogenic drugs for head and neck squamous cell carcinoma (HNSCC) including OSCC (12). Despite some encouraging results from a number of studies, targeting the angiogenic pathways in HNSCC showed limited success in most clinical trials (11-17).
Recently, VM has been suggested as one of the key mechanisms in resistance to AAT (18, 19). Interestingly, bevacizumab treatment induced VM-formation in an ovarian cancer model, which was associated with more distant metastasis (20). In line with these data, sunitinib treatment failed to inhibit VM formation in a breast cancer model (21). Yet, there is limited knowledge regarding the AAT’s effect on OSCC. Given the limited success of AAT in HNSCC clinical trials, this study aimed to examine whether a set of FDA-approved angiogenesis inhibitors could influence VM formation by OSCC cell lines and impact their proliferation, apoptosis, and migration ex vivo.
Materials and Methods
Cell lines and culture conditions. Three OSCC cell lines were selected based on our previous report (22), as follows: Two highly aggressive oral tongue squamous cell carcinoma (OTSCC) cell lines HSC-3 (JCRB 0623; Osaka National Institute of Health Sciences, Osaka, Japan) SAS (JCRB-0260). Third cell line, namely UT-SCC-28 (hereafter SCC-28) was established at the Department of Otorhinolaryngology, Head and Neck Surgery Unit, Turku University Hospital, Finland. The SCC-28 cell line originated from a primary floor of the mouth tumor. All cancer cell lines were cultured in similar conditions using 1:1 DMEM-F12 medium (Gibco/Invitrogen, Tokyo, Japan) supplemented with 10% heat-inactivated fetal bovine serum (Gibco), 50 μg/ml penicillin-streptomycin (15140-122, Thermo Fisher Scientific, Waltham, MA, USA), 50 μg/ml ascorbic acid (A1052, AppliChem, Chicago, IL, USA), 250 ng/ml amphotericin B (A2942, Sigma-Aldrich, St. Louis, MO, USA) and 0.4 μg/ml hydrocortisone (H088, Sigma-Aldrich). Cells were maintained in a 95% humidified incubator of 5% CO2 at 37°C. Human umbilical vein endothelial cells (HUVEC; Thermo Fisher Scientific;) were cultured in 200PRF medium supplemented with a low serum growth supplement (Thermo Fisher Scientific) and maintained in a 95% humidified incubator of 5% CO2 at 37°C.
Antiangiogenic compounds. Three FDA-approved antiangiogenic drugs were used including a multi protein kinase inhibitor, sorafenib (LC Laboratories, S-8502) and two small-molecule receptor tyrosine kinase (RTK) inhibitors: sunitinib (LC Laboratories, S-8803, lot FIMM23820-003) and axitinib (LC Laboratories, AG013736, A-1107, lot FIMM003778-001). These drugs were clinically employed as a monotherapy to treat advanced HNSCC (12). Sorafenib was used at concentrations of 1, 2.5, and 5 μM, whereas sunitinib and axitinib were used at 0.5, 1, and 2.5 μM, as recently reported (23-26).
Tube formation assay. The commercial mouse Engelbreth-Holm-Swarm sarcoma matrix, Matrigel (Corning, Corning, NY, USA) was used as a 3D matrix for the tube formation assay as previously described (22, 27). Matrigel was thawed overnight at 4°C. The next day, 50 μl of Matrigel was dispensed into the bottom of a 96-well plate (Costar, 3596, Corning Incorporated) and incubated for 30 min at 37°C. Cancer cells and HUVECs were detached from a 75 cm2 flask (Sigma-Aldrich) using 1× trypsin-EDTA and then resuspended in serum-free DMEM or 200PRF medium. Cells were counted using Scepter™ 2.0 Cell Counter (Merck, Millipore, Burlington, MA, USA) and seeded on the top of Matrigel at a density of 40×103 cells per well in 50 μl of serum-free media containing the antiangiogenic compounds at different concentrations. In the control wells, no antiangiogenic compound was added. The optimal density of each cell line was selected based on our recent study (22). HUVECs and OSCC cell lines were incubated for 24 and 12 h, respectively. After incubation the wells were washed with phosphate-buffered solution (PBS), fixed using 4% PFA for 20 min at room temperature, and then stored at 4°C for imaging.
Real-time cell proliferation and apoptosis assays. OSCC cells were labelled using CellTrace™ Far Red according to the manufacturer’s protocol (Invitrogen, Carlsbad, CA, USA). The labeled cells were seeded into a 96-well plate at a density of 2-3×103 cells per well in 100 μl of medium and incubated overnight. Next day, the medium was replaced with Caspase-3/7 Apoptosis Assay reagent (1:1,000, Essen Bioscience, Ann Arbor, MI, USA, Cat No. 4440) together with the tested drugs at the selected concentrations. Each dose of the tested compounds was used in triplicate and incubated for 48 h in an IncuCyte Zoom incubator (Essen BioScience). Cell proliferation was monitored every two hours using the IncuCyte® Live-Cell Analysis system at objective 10× (4 images per well) and the proliferation rate was quantified with the IncuCyte analysis software.
Scratch-wound assay. On the first day, ImageLock 96-well plate (Essen BioScience) was coated with 50 μl of Myogel (0.5 mg/ml) and incubated at 37°C overnight. Myogel was then aspirated, and cells were seeded at 35×103 cells/well in 100 μl of complete media and incubated overnight. After reaching confluency, the monolayers were scratched using WoundMaker™ 96-pin wound-making tool (Essen, BioScience) and the wells were carefully washed with complete medium and inspected under microscopy. Finally, the media was changed to 100 μl of complete media containing the above-mentioned concentrations of the tested compounds. For live imaging, the well plates were monitored hourly for 24 h using the IncuCyte® Live-Cell Analysis system (Essen BioScience Inc.).
Imaging and statistical analyses. All experiments were repeated independently at least three times, in duplicates or triplicates. The tube formation samples were photographed using the reverse Nikon Digital Sight DS-U3 microscope (Nikon, Tokyo, Japan) with magnifications of 4× and 10×. We used ImageJ software (Wayne Rasband, National Institute of Health, Bethesda, MD, USA) to analyze the images from the tube formation assays. At least three images were obtained from each well representing the tube formation. The ImageJ software was applied to analyze multiple parameters including number of nodes, meshes, segments, total mesh area, total branching length, and total length. Data are given as means±standard error of the mean (SEM). Numeric data were analyzed and visualized using GraphPad Prism version 9.4.1 for Mac OS X (GraphPad Software, San Diego, CA, USA). To assess the statistical significance, one-way ANOVA followed by Dunnett’s multiple comparison test were applied. Statistical significance was set at p<0.05.
Results
AAT differentially influenced OSCC-derived VM and tumor cell morphology. Recently, we showed that certain aggressive OSCC cells express the endothelial cell marker CD31 and can initiate VM networks similar to the endothelial tubulogenesis when cultured on biological hydrogels (22, 24). VM is considered a possible mechanism promoting drug resistance in cancer patients (18, 19). Hence, we examined whether a set of selected FDA-approved antiangiogenic drugs could abrogate such a phenomenon in OSCC. Expectedly, the tested compounds impaired the formation of HUVEC capillaries in a dose-dependent manner (sorafenib 5 μM; p=0.005, axitinib 2.5 μM; p=0.04; Figure 1). Interestingly, while sorafenib dose-dependently inhibited HSC3-derived tubulogenesis (sorafenib 5 μM; p=0.02), both axitinib and sunitinib showed little or no effect on these networks (p>0.05), respectively (Figure 2A). Furthermore, none of the tested drugs had any statistically significant inhibitory effect on neither SAS- nor SCC28-derived tubulogenesis ex vivo (Figure 2B and C). These cells were able to develop fully interconnected networks when treated with either axitinib or sunitinib.
The effect of antiangiogenic drugs on the tubulogenic potential of normal human endothelial cells (HUVEC). The antiangiogenic drugs inhibited the tubulogenesis of HUVEC in a dose-dependent manner (Sorafenib 5.0 μM; p=0.005 and Axitinib 2.5 μM; p=0.4). Scale bar=500 μm. Data are given as means±SEM of at least three independent experiments. *p-value <0.05; **p-value <0.01.
The effect of antiangiogenic drugs on the tubulogenic potential of metastatic and primary oral squamous cell carcinoma (OSCC) cells. Scale bar=500 μm. A) Sorafenib inhibited the tubulogenesis of metastatic HSC-3 cells in a dose-dependent manner (sorafenib 5 μM; p=0.02). In contrast, axitinib and sunitinib showed little or no effect on these OSCC cells, respectively (p>0.05). B) Antiangiogenic drugs had no major effect on the tubulogenesis of SAS cells (p>0.05). C) The pre-irradiated cell line, SCC-28 formed consistent and extensive tube-like structures despite the antiangiogenic treatment (p>0.05). *p-value <0.05; **p-value <0.01; ***p-value <0.001.
Of note, the administration of these antiangiogenic drugs provoked morphological changes in SAS cells, which attained a peculiar thinner “spindle” morphology with more “spike-like” extensions compared to the classical honeycomb-networks in the treatment-free controls (Figure 3A). In contrast, no similar morphological alteration was observed in HSC-3 or SCC-28 cells (Figure 3B).
The effect of antiangiogenic drugs on the morphology and proliferative potential of tumor cells. A) Treatment with antiangiogenic drugs induced morphological changes in the tongue SAS cells characterized by spindle shaped networks and more “spike-like” extensions compared to the non-treated controls. Scale bar=500 μm. B) No similar morphological changes were observed in other tumor cells, such as SCC-28. C-E) Sorafenib significantly inhibited the proliferation of HSC-3 (5 μM; p=0.002), SAS (2.5 and 5 μM; p=0.002, <0.001, respectively), and SCC-28 (5 μM; p=0.01). F) Sunitinib showed inhibitory effects only on the proliferation of HSC-3 cells (2.5 μM; p=0.0277). Data are given as means±SEM of at least three independent experiments. *p-value <0.05; **p-value <0.01; ***p-value <0.001.
Sorafenib suppressed OSCC cell proliferation. Cell proliferation is a key process during tumor angiogenesis and development that precedes cancer cell migration (28). Therefore, we sought to investigate the possible anti-proliferative effect of the tested AATs on OSCC cell lines using the IncuCyte® Live-Cell Analysis system. The multi-kinase inhibitor, sorafenib, showed a good potency and significantly inhibited the proliferation of HSC-3 (5 μM; p=0.002), SAS (2.5 and 5 μM; p=0.002, <0.001, respectively), and SCC-28 (5 μM; p=0.01) (Figure 3C-F). Sunitinib only inhibited the metastatic HSC-3 cell proliferation at the dose of 2.5 μM (p=0.0277) (Figure 3F). Otherwise, sunitinib and axitinib showed no anti-proliferative effect on OSCC cell proliferation (Figure 4).
The effect of antiangiogenic drugs on the proliferative potential of tumor cells. A-B) Sunitinib did not exhibit a significant anti-proliferative effect over SAS or SCC-28 cells (p=0.5). C-E) Axitinib also showed almost no effect on the proliferation of OSCC cells (p=0.5). Data are given as means±SEM of at least three independent experiments.
Sorafenib and sunitinib induced OSCC cell apoptosis. Evidence from preclinical and clinical studies revealed that AAT can induce apoptosis in endothelial and tumor cells alike (29, 30). Thus, we next employed the real-time caspase-3/7 activity assay to explore the pro-apoptotic efficiency of the drugs on OSCC. Interestingly, treatment with sorafenib and sunitinib resulted in an increased apoptosis rate in HSC-3 (5 μM; p=0.03 and p<0.001, respectively; Figure 5A and B). In SAS, higher doses of sorafenib increased cell apoptosis; however, the effect was not statistically significant (p>0.05; Figure 5C), whereas sunitinib significantly induced apoptosis (p=0.02; Figure 5D). Similarly, SCC-28 cells showed higher apoptosis percentage when treated with sorafenib (2.5 and 5 μM; p<0.001 and <0.001, respectively) or sunitinib (1 and 2.5 μM; p=0.03 and <0.001, respectively) (Figure 5E and F). Axitinib had no effect on OSCC cell apoptosis (Figure 5G).
The effect of antiangiogenic drugs on the apoptosis of metastatic and primary oral squamous cell carcinoma (OSCC) cells. A-B) Sorafenib and sunitinib induced the apoptosis of metastatic HSC-3 cells (5 μM p=0.03; 2.5 μM, p<0.001, respectively). C) Higher doses of sorafenib increased SAS cell apoptosis but the effect did not reach statistical significance (p>0.05). D) Sunitinib induced the apoptosis of SAS cells (2.5 μM; p=0.02). E-F) Similarly, SCC-28 cells underwent significant apoptosis following treatment with sorafenib (2.5 and 5 μM; p<0.001 and p<0.001, respectively) and sunitinib (1 and 2.5 μM; p=0.03 and <0.001, respectively). G) Axitinib did not significantly altered the apoptosis rate in OSCC cells (p≥0.5). Data are given as means±SEM of at least three independent experiments. *p-value <0.05; **p-value <0.01; ***p-value <0.001.
Sunitinib inhibited SAS cell migration. Finally, we aimed to evaluate whether these treatments can interfere with cancer cell migration since it represents an essential prerequisite for tumor angiogenesis, metastasis, and clinical relapse. To this end, we assessed the horizontal OSCC cell migration in precisely created two-dimensional wounds using the WoundMaker™ tool and IncuCyte® Live-Cell Analysis system. In this assay, cell migration efficiency is measured as a function of the scratch-wound closure over time. Of interest, our results showed that only sunitinib significantly inhibited the migration of SAS cells at a dose of 2.5 μM (p=0.01). Sunitinib slightly inhibited SAS migration at 1 μM; however, it was not statistically significant (Figure 6). The other compounds did not seem to influence OSCC cell migration (Figure 7A and B).
Assessment of SAS cell migration using the scratch-wound assay. Sunitinib at the concentration of 2.5 μM impeded the horizontal migration of SAS cells compared with the non-treated controls (p=0.01). Left: representative images of the scratched wounded areas (marked by dotted lines) on confluence monolayers of SAS cells treated with sunitinib at different time points. Right: Real-time analysis of the scratch-wound closure of SAS treated with all compounds using the IncuCyte® Live-Cell Analysis system. Data are given as means±SEM of at least three independent experiments. *p-value <0.05; **p-value <0.01; ***p-value <0.001.
Assessment of HSC-3 and SCC-28 cell migration using the scratch-wound assay. A-B) Right: Real-time analysis of the scratch-wound closure of HSC-3 and SCC-28 treated with all compounds using the IncuCyte® Live-Cell Analysis system. Data are given as means±SEM of at least three independent experiments. A) None of the tested antiangiogenic drugs had a significant influence on the migration of HSC-3 cells (p≥0.05). Left: representative images of the scratched wounded areas (marked by dotted lines) on confluence monolayers of HSC-3 cells treated with sorafenib at different time points. B) The pre-irradiated primary oral squamous cell carcinoma cells, SCC-28, showed a slower migratory rate compared to the other cancer cell lines, which was not however affected by any of the antiangiogenic drugs (p≥0.05). Left: representative images of the scratched wounded areas (marked by dotted lines) on confluence monolayers of SCC-28 cells treated with sorafenib at different time points. *p-value <0.05; **p-value <0.01; ***p-value <0.001.
Discussion
Although AAT has shown promise in controlling neovasculature in certain solid tumors, its therapeutic effect in OSCC has been limited by the low response rates and the development of clinical resistance. Despite the fact that these tumors arise in highly vascularized tissues, there are no FDA-approved AAT for OSCC (12). Currently, drug resistance is the main reason for limiting the efficacy of AAT including TKIs on cancer, which inevitably develops during the course of therapy, deteriorating the patients’ survival (31). It is therefore important to understand what could drive such selective resistance in OSCC. A better understanding of these mechanisms may help improve the development of new target-specific compounds and guide the future preclinical and clinical studies.
The VM has been associated with an aggressive clinical course and shorter survival outcomes in HNSCC (10). Importantly, VM has been suggested as a possible mechanism behind AAT resistance as it serves as an additional vascularization route independent from angiogenesis (19, 32). However, there is limited evidence regarding whether AAT has an effect on VM in OSCC. Therefore, we examined the efficacy of three angiogenesis inhibitors, used as monotherapy in clinical trials for HNSCC, on the tubulogenesis of selected OSCC cell lines in vitro. We selected the cell lines based on previous findings demonstrating that highly aggressive cell lines (i.e., HSC-3, SAS, SCC-28) form mature tube-like structures on Matrigel compared to less aggressive cells, which remained as single cells or small clusters (22). Moreover, SCC-28 is derived from pre-irradiated floor of the mouth OSCC, which represents a high-risk site for invasion and metastasis. Importantly, irradiation has been suggested to enhance the metastasis and VM (33, 34). Interestingly, sorafenib was able to inhibit the tubulogenic potential of the metastatic OSCC cells (HSC-3) and normal endothelial cell lines, whereas SAS and SCC-28 formed extensive tube-like structures despite the treatment. In agreement, previous studies have demonstrated that HSC-3 cells exhibit high plasticity and express endothelial cell markers (22, 24). However, sorafenib has been reported to selectively inhibit tube formation in HCC cells but not in breast cancer (35, 36).
Our results demonstrate that sunitinib and axitinib had no inhibitory effect on OSCC cell-derived tubulogenesis. On the contrary, axitinib treatment inhibited the tube formation of non-collagenous 11-domain expressing OSCC cells (37). In accordance with our data, these two drugs showed no favorable effect on the tubulogenesis of tumor cells from breast, nasopharyngeal, and renal cancers (21, 26, 38-40). Interestingly, AAT provoked a phenotypic change in SAS tubulogenesis towards spindle-like morphology with spiky-extensions. Of note, a previous report demonstrated that HNSCC cells with an EMT-expression profile had similar spindle-like morphology (41). While we acknowledge the lack of an in vivo model, the link between AAT resistance and VM has been previously studied in vivo. Using an ovarian cancer mouse model, bevacizumab treatment was associated with higher lung metastases and more VM (20). Furthermore, VM was significantly higher in AAT-treated glioblastoma tumors, while sunitinib failed to inhibit VM in an aggressive breast cancer model (21, 42). Interestingly, Serova et al. showed that in a renal cell carcinoma model, VM was initially inhibited by sunitinib but it was eventually associated with tumor resistance and aggressive phenotype changes both in vitro and in vivo. The study showed that human CD31 expression was higher in sunitinib-resistant tumors, suggesting the implication of phenotypic plasticity of human cancer cells since they were the only human material transplanted into the nude mice (43). Similarly, in a glioblastoma mouse model, sunitinib treatment led to activation of evasive resistance, triggering a more invasive phenotype and promoting distant metastasis (44). Acquired resistance to AAT could be in part mediated by both phenotypic and genotypic plasticity of certain cancer cells. VM is generally considered a VEGF-independent phenomenon, with its receptors being among the main targets of sorafenib, sunitinib, and axitinib (45, 46). Moreover, AAT induces a hypoxic tumor microenvironment, which is widely recognized as a promoter of VM in many cancers (9, 20, 46). However, the mechanisms behind AAT resistance and VM are yet to be elucidated.
Besides endothelial cells, it has been established that the target receptors of AAT including VEGFRs are also present on numerous cancer cell types, including HNSCC. Therefore, the effects of antiangiogenic compounds could be mediated by a direct effect on tumor cells (47-49). The multi-kinase inhibitor sorafenib has shown anti-proliferative effects in several preclinical models, including HNSCC (50-54). While proliferation is a fundamental process in tumor growth and angiogenesis (28), there are limited data regarding the potential anti-proliferative effect of AAT in OSCC specifically. In line with previous results, we demonstrate that sorafenib inhibited the proliferation of OSCC cells in a dose-dependent manner. However, sunitinib decreased the proliferation of only the metastatic OSCC cell line, whereas axitinib showed limited anti-proliferative effects. In accordance with our results, sunitinib had a limited anti-proliferative effect on anaplastic thyroid cancer cells (55) or renal cancer cells (56). Several studies have reported that a sunitinib concentration around 0.01 μM was effective in inhibiting the phosphorylation of VEGFR-2 and PDGFR (56, 57). Therefore, it is suggested that high concentrations (over 5 μM) might not be pharmacologically or clinically relevant (56). Nevertheless, axitinib and sunitinib have been previously shown to inhibit the proliferation of tumor cells in different cancer models, including melanoma (58), esophageal (59), gastric (60), colorectal (61), ovarian (23) and breast cancer (62, 63) in which the tested cell lines often harbored activating mutations of the AAT’s target receptors of or their ligands. Moreover, sunitinib caused significant growth inhibition of nasopharyngeal carcinoma cells that were confirmed to express PDGFR, a target receptor of sunitinib (64). Such varying anti-proliferative effects could partly be explained by differences in cell line origin, phenotype, and the different targeting profile of each compound (19). For instance, the anti-proliferative effect of sorafenib is mediated partly via inhibition of Raf-kinase, whereas axitinib and sunitinib are more selective to VEGFRs and PDGFRs (49, 64-67). Therefore, further analyses on the differences in protein expression of OSCC cell lines would be beneficial in the future.
The anti-proliferative effect of cancer therapy can partly be mediated by the induction of apoptosis and thus, we studied whether AAT could affect this process in OSCC cells. By assessing the activity of caspase-3, a central player in apoptosis (68), we showed that both sorafenib and sunitinib induced tumor cell apoptosis in a dose-dependent manner. In this regard, Garten et al. showed that treatment of HCC cell lines with sorafenib dose-dependently induced apoptosis, likely driven by the NAD/SIRT1/AMPK axis (69). Moreover, supporting our results, antiangiogenic compounds have been previously reported to induce apoptosis in several solid tumors, although there are limited data regarding HNSCC or OSCC (23, 62, 63, 65, 70, 71). Once they survive the harsh conditions and apoptosis-inducing signals within the TME, tumor cells can start to migrate and invade surrounding tissues (72). Herein, neither sorafenib nor axitinib had clear effects on tumor cell migration, which contrasts previous reports showing the anti-migratory effect of sorafenib in different cancers (23, 53, 73-76). In hepatocellular carcinoma and osteosarcoma, sorafenib was found to inhibit migration by reducing cancer cell matrix metalloproteinase activity or EMT (73, 76). Intriguingly, sorafenib showed a dual mode of activation by stimulating and inhibiting bladder cancer cell migration at lower and higher doses, respectively (77). Also, in a colorectal cancer model, chronic exposure to AAT led to enhanced migration and invasion capability, indicating that compensatory pathways could also mediate tumor cell migration (78). In line with the previous experimental reports, we demonstrated that sunitinib exhibited anti-migratory effects on OSCC (61, 79, 80). These results indicate that differences in cell line origin or experimental settings may drive such variations in the results.
In conclusion, we examined the potential anti-cancer effects of selected FDA-approved antiangiogenic compounds: sorafenib, sunitinib, and axitinib – all of which have been used as a monotherapy in clinical trials of HNSCC. These drugs differentially affected certain initial processes of tumorigenesis, including proliferation and apoptosis. While sorafenib and sunitinib exhibited anti-proliferative, anti-migratory, and apoptosis-inducing effects on selected OSCC cell lines, axitinib did not significantly alter these processes. We acknowledge the limitations of the present study including the lack of an in vivo model, yet our data support the notion that VM may represent a possible mechanism behind AAT resistance in OSCC cells. Although more mechanistic studies are warranted to understand the differential effects of the tested drugs, our data encourage further research both in the preclinical and clinical setting.
Footnotes
Authors’ Contributions
All Authors contributed to the study’s conception, design, and funding. Roosa Hujanen performed the experiments and collected data. Roosa Hujanen, Kari K. Eklund and Abdelhakim Salem performed data analysis and interpretation. Roosa Hujanen and Abdelhakim Salem wrote the manuscript. Abdelhakim Salem is the corresponding author and main supervisor of this work. All Authors read, commented, and approved the final manuscript.
Conflicts of Interest
The Authors have no conflicts of interest to declare in relation to this study.
Funding
This work was financially supported by the Finnish Dental Society (Apollonia); Finnish Medical Foundation; Minerva Foundation (Selma and Maja-Lisa Selander’s Fund); Päivikki and Sakari Sohlberg foundation; Cancer Society of Finland.
- Received March 26, 2024.
- Revision received May 3, 2024.
- Accepted May 7, 2024.
- Copyright © 2024 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).