Abstract
Background/Aim: We evaluated the impact of FosL1, a member of the activated protein-1 family, on the pathways leading to regional metastasis of head and neck squamous cell carcinoma (HNSCC). Materials and Methods: We examined the influence of small interfering RNA (siRNA) and short heparin RNA (shRNA) mediated knockdown of FosL1 on cell migration, invasion, and proliferation in vitro as well as on regional metastasis in vivo. The prognostic significance of FosL1 was also analyzed using the Kaplan– Meier plotter using data from an HNSCC patient database. Results: Down-regulation of FosL1 inhibited cell migration, invasion, and proliferation in vitro, decreased the incidence of regional metastases, and prolonged the survival of mice in vivo. We also determined that HNSCC patients with higher expression levels of FosL1 had a significantly shorter survival time than those with low expression of FosL1. Conclusion: FosL1 plays a crucial role in promoting cell migration, invasion, and proliferation in HNSCC.
- FosL1
- AP-1
- Head and neck squamous cell carcinoma (HNSCC)
- cervical lymph node metastasis
- short hairpin RNA (shRNA)
Head and neck squamous cell carcinoma (HNSCC) patients often have advanced disease with cervical lymph node metastases at the time of diagnosis (1), resulting in poor prognosis despite advances in treatment modalities. Furthermore, treatment failure is often associated with locoregional recurrence in HNSCC patients. The presence of cervical lymph node metastasis is reported as one of the reliable prognostic factors for the survival of patients with HNSCC (2). Therefore, it is necessary to elucidate the mechanisms of cervical lymph node metastases to develop new treatments to improve the survival outcomes of patients with HNSCC.
We previously reported that JunB, a member of the AP-1 family, regulates metastatic pathways by promoting cell invasion and migration of HNSCC (3). Thus, AP-1, which forms heterodimers with proteins of the Fos family, the Jun family, musculoaponeurotic fibrosarcoma (Maf), and activating transcription factor (ATF) (4, 5), plays crucial roles in regulating cellular processes such as proliferation, differentiation, death, and survival induced by growth factors, cell-matrix interactions, infections, cytokines, and stresses (6, 7). The Fos family of proteins and c-Jun have also been reported to induce tumorigenesis and metastasis (3, 4, 8-10).
These studies led us to characterize the regional metastatic potential in vivo using 14 different HNSCC cell lines in an orthotopic nude mouse model. We performed an upstream and key node analysis using whole-gene microarray data of these lines to determine key molecules associated with the regional metastatic potential of HNSCC. As we identified FosL1 as one of the key molecules associated with the regional metastatic potential of HNSCC, we examined the impact of FosL1 knockdown on cell migration in vitro, as well as cervical lymph node metastasis in vivo, to test the hypothesis that FosL1 could be the key molecule regulating the pathways associated with neck lymph node metastasis in HNSCC.
Materials and Methods
Cell lines. Information and appropriate growth media for the 25 HNSCC cell lines have been described previously (3, 11). All cells were maintained on plastic plates as adherent monolayer cultures at 37°C and 5% CO2 and authenticated by short tandem repeat genotyping as described previously (11).
Animal maintenance and an orthotopic mouse model of HNSCC. Athymic nude mice, aged 7-8 weeks, were purchased from the animal production facilities of the National Cancer Institute-Frederick Cancer Research and Development Center (Frederick, MD, USA), Oriental Yeast (Tokyo, Japan), and Japan SLC, Inc. (Shizuoka, Japan). The mice were housed and maintained as described previously (12). All animal experiments were conducted in accordance with procedures approved by the Institutional Animal Care Use Committee at Yokohama City University, School of Medicine (approval ID: FA-18-026, Yokohama, Japan).
The baseline tumorigenic and cervical metastatic potential of the orthotopic mouse model of HNSCC was then evaluated as per our previous report (12) and current animal experiments with YCU-T892, YCU-MS861, KCC-T871, HSC-3, KCC-T873, YCU-OR891, YCU-M911, KCC-L871, YCU-M862, and KCC-M871 cells as described previously (13). Tumor formation and the presence of tongue and neck lymph node metastases were evaluated pathologically by light microscopy (Figure 1A and B), and the proportion of mice with cervical lymph node metastasis was calculated as described previously (12). The survival of the mice was also analyzed.
The cervical lymph node metastatic potential of 26 HNSCC cell lines in an orthotopic nude mouse model of HNSCC. A. HE staining of a tongue with HNSCC in an orthotopic nude mouse model of HNSCC (arrow: tumor lesion). Scale bar: 1,000 μm. B. HE staining of cervical lymph node metastasis in an orthotopic nude mouse model of HNSCC (arrow: metastatic lesion). Scale bar: 100 μm. C. The tumorigenicity of 26 HNSCC cell lines in an orthotopic nude mouse model of HNSCC on day 32. D. The cervical metastatic potential of 14 HNSCC cell lines forming tongue tumors at a rate of more than 80% in the orthotopic nude mouse model of HNSCC. Two HNSCC cell lines (FaDu and Detroit562) established more than 70% cervical lymph node metastases, while 2 cell lines (YCU-MS861 and YCU-OR891) did not establish cervical lymph node metastases in any of the inoculated mice. E. The results of principle component analysis (PCA). PCA was performed based on the expression profiles of the samples. The first 3 PCAs for 4 HNSCC cells were plotted. Metastatic lines (black square, FaDu; black triangle, Detroit562), intermediate metastatic lines (gray triangle, HN31; gray rhombus, HN5; gray trapezoid, OSC-19; gray inverted trapezoid, HSC3; gray diamond, KCC-T871; gray pentagon, KCC-T873; gray ellipse, SqCC/Y1; gray hexagon, TR146; gray inverted triangle, UM-SCC17B; gray clover, YCU-T892), non-metastatic lines (white star, YCU-MS861; white hexagram, YCU-OR891).
Microarray analysis and upstream and key node analysis with ExPlain™. First, 14 of the 26 HNSCC lines that showed more than 80% of primary tumor formation were chosen for further studies. Total RNA from these 14 HNSCC cell lines was then extracted for microarray analysis as described previously (14). Whole-genome gene profiling was performed using a SurePrint G3 Human GE 8×60 K Microarray (Agilent Technologies, Santa Clara, CA, USA) as described previously (3) (Gene Expression Omnibus accession number: GSE79637).
Next, genes with an absolute fold change (FC) value >22 (150 genes) and <1.05 (818 genes) between two metastatic lines and two non-metastatic lines were evaluated by upstream and key node analysis using ExPlain™ (15) to investigate the key upstream molecules involved in the regional metastatic pathways in HNSCC, as described previously (3, 16), after the principal component analysis.
Western blotting analysis. The expression of FosL1 and JunB in HNSCC cell lines was analyzed by western blotting as described previously (3). Antibodies were purchased from the following sources and used at the indicated dilutions: FosL1 (1:1,000; Cell Signaling Technologies, Danvers, MA, USA), JunB (1:1,000; Cell Signaling Technologies), and α-tubulin (1:1,000; Cell Signaling Technologies). Using western blotting, protein expression levels of siRNA- or shRNA-mediated knockdown of FosL1 were also compared with those of cells transfected with a negative siRNA/shRNA control.
Bioinformatics analysis using the Kaplan–Meier Plotter. In this study, the Kaplan–Meier Plotter was used to evaluate the impact of FosL1 on the survival of HNSCC (17). This online database includes RNA seq data of pan-cancer, including 500 cases of HNSCC. Briefly, FosL1 was uploaded into this database to examine the overall survival with 60 months as the follow-up threshold and to auto-select the best cutoff including all stage, gender, race, and grade as described previously (18).
siRNA/shRNA-mediated knockdown of FosL1 in HNSCC cells. siRNA-mediated knockdown of FosL1 in Detroit562, HN30, and KCC-T871 HNSCC cells was performed by transient transfection with negative control or two independent siRNAs for FosL1 (siRNA IDs: #1=s15583 and #2=s15584) (Life Technologies, Gaithersburg, MD, USA) using Lipofectamine RNAiMAX (Life Technologies) according to the manufacturer’s instructions.
For the establishment of shRNA-mediated knockdown of FosL1 in HNSCC cells, plasmid vectors containing shRNA targeting FosL1 and nonspecific control (shControl) were purchased (Origene Technologies, Inc., Rockville, MD, USA). The retrovirus Packaging Kit Ampho (Takara Bio Inc, Shiga, Japan) was then used to transfect 293T cells according to the manufacturer’s instructions for the generation of retroviral particles. At 48 h post-transfection, the supernatants containing virus particles were filtered using 0.45 μm filters. FaDu cells were infected with concentrated viral supernatants using 8 μg/ml Polybrene (Sigma-Aldrich, St Louis, MO, USA) for 24 h and the medium was changed to Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal bovine serum (FBS). FaDu cells expressing shRNA were selected with 1 μg/ml puromycin, and single colonies were isolated and then maintained in DMEM with 10% FBS and 1 μg/ml puromycin.
Scratch and invasion assays. The scratch assay was performed in Detroit562, HN30, and KCC-T871 cells infected with negative control or FosL1 siRNA as described previously (3). Briefly, after incubation of cells until confluent, a horizontal wound was made and images were captured at 0, 9, 15, and 18 h post-wound for each cell line.
Invasion assays were performed in KCC-T871 and HN30 cells as described previously (3). Briefly, cells in a serum-free medium were plated in the upper chamber and incubated with a medium containing 10% FBS in the bottom of the chamber for 22 h. Invaded cells were counted in all fields of the chamber membranes. The experiments were repeated three times.
Cell viability assay. Cell viability assay was performed in FaDu-shControl and FaDu-shFOSL1 cells using 96-well microplates at 24, 48, and 72 h, as described previously (3).
Orthotopic nude mouse model of HNSCC with FaDu-shControl and FaDu-shFOSL1 cells. The cervical lymph node metastatic potential of FaDu-shControl or FaDu-shFOSL1 cells was examined using an orthotopic mouse model as described above. For this experiment, 15 mice were prepared for each cell line. Mice were euthanized using carbon dioxide asphyxiation when they lost more than 20% of their pre-injection body weight or at 64 days after cell injection.
The animal experiments were performed with 13 mice for the control and 14 mice for the shRNA-mediated FaDu knockdown group. All mice were euthanized at 31 days after cell injection, and the tumor size and the presence of cervical lymph nodes were evaluated pathologically.
Statistical analysis. Survival was analyzed by the Kaplan–Meier method and compared using log-rank tests. The results of the cell viability assay were compared using unpaired t-tests corrected for multiple comparisons with the Holm-Sidak method. The results of migration, invasion assays, and tumor volumes were compared using a paired 2-tailed t-test. The incidences of cervical lymph node metastasis were compared using Fisher’s exact test. GraphPad Prism version 7.04 (GraphPad Software, La Jolla, CA, USA) was used to perform statistical analyses. For all comparisons, p<0.05 was considered statistically significant.
Results
The neck lymph node metastatic potential of 26 HNSCC lines in an orthotopic nude mouse model. The mean tumor volume on day 32 in the orthotopic nude mouse models of the 26 HNSCC lines is shown in Figure 1C. While 18 (69.2%) of the 26 HNSCC cell lines established tumors in the tongue, no tumor formation was observed in 8 HNSCC cell lines.
Among the 18 cell lines, PE/CA-PJ34, MDA1986LN, HN4, and UMSCC1 were excluded due to lower potential for tumor formation with a very small tumor size, which was only visible under the microscope. We then selected two metastatic HNSCC lines (FaDu and Detroit562), which showed cervical lymph node metastases in more than 70% of mice, and two non-metastatic HNSCC lines (YCU-MS861 and YCU-OR891), which did not have any cervical lymph node metastasis in mice (Figure 1D).
Microarray analysis and upstream and key node analysis. Next, we performed a microarray analysis using the above 14 cell lines, following 3 principal component analyses. We observed a clear distinction between the two highly metastatic HNSCC cell lines (FaDu and Detroit562) and the two non-metastatic HNSCC cell lines (YCU-MS861 and YCU-OR891), as shown in Figure 1E. Furthermore, we analyzed the microarray data by calculating FC in expression and performed upstream and key node analysis using ExPlain™ as described previously (3). In this procedure, 150 genes with a more than 22 absolute FC value and 818 genes with a less than 1.05 absolute FC value between the two metastatic and two non-metastatic lines were loaded into ExPlain™.
We identified 160 genes as candidate key factors of pathways associated with cervical lymph node metastasis in HNSCC. A list of the top 17 genes with a score ≥10 according to the ExPlain™ tool is shown in Table I. Additionally, the microarray data were analyzed in conjunction with our previous data on distant metastasis (3), as the regulation of regional metastasis could share some pathways with that of distant metastasis in HNSCC. Similarly, both 138 genes with an absolute FC >2.2 (according to the results of regional metastatic potential of the orthotopic model of HNSCC) and FC >2.0, with p<0.05 (according to the results of the distant metastatic potential of an experimental lung metastatic mouse model), as well as 16 genes with FC <1.13 and 716 genes with FC <1.13 according to the results of regional/distant metastatic potential of animal models, were loaded into ExPlain™. A total of 210 genes were identified as candidate key factors in the regulation of pathways associated with metastasis in HNSCC. A list of the top 23 genes with a score ≥10 according to the ExPlain™ tool is shown in Table II. We observed that these lists included several AP-1 proteins with high scores, suggesting that the AP-1 family could play a crucial role in the regional spread of cervical lymph node metastasis in HNSCC.
The list of top 17 upstream key molecules associated with the incidence of neck lymph node metastasis of head and neck squamous cell carcinoma cells (HNSCC) cells.
The list of the top 23 upstream key molecules associated with the incidence of both neck lymph node metastasis and distant metastasis of head and neck squamous cell carcinoma (HNSCC) cells.
Expression of the AP-1 family proteins in HNSCC cells. We then analyzed the expression levels of FosL1 and JunB in metastatic HNSCC cells (Detroit562 and FaDu) and non-metastatic HNSCC cells (YCU-OR891 and YCU-MS861) with or without HGF stimulation (20 ng/ml for 4 h) by western blotting. We observed that HGF dramatically stimulated the expression of FosL1 in both Detroit562 and FaDu cells, and the stimulation was stronger than that in the two non-metastatic HNSCC lines. On the other hand, we observed only a slight effect of HGF-stimulation on the expression of JunB on Detroit562 cells, as shown in Figure 2A.
Expression of AP-1 family proteins in HNSCC cells, analyzed via western blotting. A. The cervical lymph node metastatic HNSCC cell lines (FaDu and Detroit562) showed higher levels of FosL1 expression than the non-metastatic HNSCC cell lines (YCU-MS861 and YCU-OR891), while the metastatic cell lines had JunB expression levels comparable to the non-metastatic cell lines. α-Tubulin (α-Tub) was used as an internal control. B. The survival analysis of HNSCC using the Kaplan–Meier plotter database to evaluate the prognostic significance of FosL1.
Survival analysis of HNSCC through the Kaplan–Meier Plotter Database. To evaluate the prognostic significance of FosL1 in patients with HNSCC, we used the Kaplan–Meier Plotter database, which comprised RNA seq data of pan-cancer, including 500 cases of HNSCC. We observed that patients with high expression of FosL1 had a significantly shorter survival time than those with low expression of FosL1, as shown in Figure 2B (hazard ratio=1.62, 95%confidence interval=1.22–2.14, p=0.0007). Taken together, these results led us to examine the roles of FosL1 in regulating the pathways associated with cervical lymph node metastasis in HNSCC.
siRNA and shRNA knockdown of FosL1 in metastatic HNSCC cells suppresses tumor invasion, migration, and proliferation. To determine whether FosL1 promotes cell invasion, migration, and proliferation in HNSCC cells, we first established an siRNA-mediated knockdown of FosL1 in metastatic Detroit562 cells, KCC-T871 cells with low metastatic potential, and non-metastatic HN30 cells. The role of FosL1 was examined using invasion and scratch assays. The effect of the siRNA-mediated knockdown was confirmed using western blotting, as shown in Figure 3A. The invasion assay revealed a 31.86% reduction in the invasion potential of KCC-T871-siFOSL1#1 cells compared to negative control siRNA (73.50±7.01) with a significant difference (p=0.0199). HN30-siFOSL1#1 cells also showed a 41.76% reduction in invasion potential when compared to the control; however, the difference was not significant (p=0.0626, Figure 3B). The scratch assay revealed significant reductions in the motility of both Detroit562-siFOSL1#1 cells (70.25%±2.57% vs. 50.61%±3.61%, p<0.0001) and Detroit562-siFOSL1#2 cells (57.26%±5.04% vs. 32.21%±3.93%, p<0.0001) compared to the control. We also observed significant reductions in the motility of HN30-siFOSL1#1 cells (92.15%±2.36% vs. 58.21%±7.09%, p<0.0001), HN30-siFOSL1#2 cells (85.63%±2.27% vs. 69.54%±3.07%, p<0.0001), KCC-T871-siFOSL1#1 cells (65.27% ±4.24% vs. 52.25%±2.83%, p=0.0084), and KCC-T871-siFOSL1#2 cells (34.63%±2.08% vs. 13.95%±3.87%, p=0.0010) when compared to the control. Thus, siRNA-mediated knockdown of FosL1 inhibited cell migration ability of HNSCC cells with both high and low metastatic potential, as shown in Figure 3C.
SiRNA knockdown of FosL1 in HNSCC cells suppressed migration. A. The western blotting analysis of Detroit562, HN30, and KCC-T871 cells transfected with negative control or FosL1 siRNAs (#1 and #2). B. SiRNA knockdown of FosL1 in KCC-T871 and HN30 cells inhibited cell invasion. N=12; columns, mean number of invading cells; bars, SEM. C. SiRNA knockdown of FosL1 in Detroit562, HN30, and KCC-T871 cells inhibited cell migration. N=24; columns, % of wound closure; bars, SEM.
To confirm that the knockdown of FosL1 could decrease motility of HNSCC cells, we also established shRNA-mediated knockdown of FosL1 in metastatic FaDu cells (Figure 4A). The scratch assay showed significant reductions in the motility of FaDu-shFOSL1 cells (13.46%±1.91%) compared to the control (42.61±2.48%, p<0.0001, Figure 4B). Also, to examine whether FosL1 promotes HNSCC cell proliferation, we performed cell viability assays with FaDu-shControl and FaDu-shFOSL1 cells. We observed a significant reduction in the proliferation of FaDu-shFOSL1 compared to the control, as shown in Figure 4C. Thus, these results suggest that FosL1 has the potential to promote HNSCC cell invasion, migration, and proliferation.
ShRNA knockdown of FosL1 in HNSCC cells suppressed cell proliferation and cervical lymph node metastases in orthotopic nude mouse model. A. Western blotting analysis of FaDu cells transfected with negative control or FosL1 shRNA. B. ShRNA knockdown of FosL1 in FaDu cells inhibited cell migration. Columns, mean % of wound closure; bars, SEM. C. A significant reduction in cell proliferation ability for FaDu-shFOSL1 compared to control was observed in cell viability assays. N=16. D. Survival time of the control group with FaDu-shControl (N=15) and FosL1 knockdown (KD) group with FaDu-shFOSL1 (N=15). The survival rate in the control group was significantly lower than that in the FosL1 KD group. E. Tumor volume of the control group with FaDu-shControl (N=13) and FosL1 KD group with FaDu-shFOSL1 (N=14). Median tumor volume in the FosL1 KD group (0.00 mm3, 0.00-15.59 mm3) was significantly smaller than that in the control group (39.48 mm3, 0.00-329.36 mm3); Columns, tumor volume; bars, SEM. F. The incidence of cervical lymph node metastases of the control group with FaDu-shControl (N=13) and FosL1 KD group with FaDu-shFOSL1 (N=14). There was a significant reduction in microscopic cervical lymph node metastases in the control group mice (46.15%) compared to the FosL1 KD group (7.14%).
Knockdown of FosL1 in metastatic HNSCC cells reduced the incidence of neck lymph node metastasis in vivo. Lastly, we examined the effect of FosL1 knockdown on tumor growth and regional metastatic potential using an orthotopic nude mouse model to clarify the role of FosL1 in cell migration, invasion, and proliferation in HNSCC in vivo.
We observed that the median survival period of mice injected with FaDu-shFOSL1 cells (the FosL1 KD group, 64 days) was significantly longer than that of mice injected with FaDu-shControl cells (the control group, 51 days, p=0.0081, Figure 4D). We repeated the in vivo experiment to evaluate tumor volume and the incidence of cervical lymph node metastasis at 31 days after cell injection using an orthotopic nude mouse model of HNSCC. We observed that tumors were formed in 13 (86.7%) of the 15 mice in the control group and 14 (93.3%) of the 15 mice in the FosL1 KD group, although the tumors formed in the mice of the FosL1 KD group were small. We observed that the median tumor volume of the FosL1 KD group (3.35 mm3±1.53 mm3) was significantly smaller than that of the control group (79.67 mm3±27.36 mm3, p=0.0078, Figure 4E). Furthermore, we also observed that the rate of cervical lymph node metastases in the FosL1 KD group (7.14%) was significantly smaller than that in the control group (46.15%, p=0.0329, Figure 4F). Overall, we confirmed that knockdown of FosL1 in metastatic HNSCC cells significantly reduced tumor growth, as well as the incidence of cervical lymph node metastasis in vivo.
Discussion
In this study, we identified FosL1 as one of the main molecules regulating cervical lymph node metastasis in HNSCC through upstream and key node analyses using whole gene microarray analysis. We then demonstrated that siRNA-and shRNA-mediated knockdown of FosL1 in metastatic HNSCC cells markedly suppressed tumor invasion, migration, and proliferation of HNSCC cells both in vitro and in vivo. These results suggest that FosL1 plays an important role in promoting HNSCC cell invasion, migration, and proliferation, resulting in the regulation of the pathway associated with the regional spread of cervical lymph node metastasis, one of the most powerful prognostic factors in patients with HNSCC.
FosL1 has been reported to promote cell invasion, migration, and metastasis in several solid carcinomas (10, 19-22). Epithelial-to-mesenchymal transition (EMT), which is a well-known crucial phenomenon associated with cancer cell invasion and stem cell-ness, has been reported to be associated with FosL1, with respect to the regulation of metastatic potential (23, 24). Also, miR-34 and miR-130a have been reported to regulate FosL1 to suppress breast cancer invasion and metastasis through the upregulation of ZO-1, a tight junction protein and an epithelial marker (25, 26). Thus, it seems that FosL1 is associated with EMT in the regulation of cancer cell invasion and migration. However, we did not observe any change in cell morphological features by knockdown of FosL1 (data not shown), although we did not perform any experiments under stimulation by TGF-β or other cytokines. Similarly, our previous study indicated that JunB was not associated with EMT, with respect to the promotion of cell invasion and migration of HNSCC cells, suggesting that EMT may not affect the regulation of metastasis of malignancies with huge mutation burdens such as HNSCC.
Recent studies have reported that FosL1 regulates HMGA1 expression during the migration and proliferation of several malignancies (27, 28). Furthermore, FosL1 has also been reported to activate MMP family proteins, which are associated with angiogenesis, and tumor metastasis (29-32). Thus, FosL1 appears to interact with many molecules associated with pathways that regulate tumor metastasis. Furthermore, we observed the prognostic significance of FosL1 expression on the survival of patients with HNSCC via bioinformatics analyses using the cancer genome atlas database. The impact of FosL1 as a prognostic biomarker has been reported in several malignancies (27, 33, 34). Our results suggest that FosL1 has the potential to play a crucial role in regional metastasis of HNSCC and be a reliable prognostic factor for the survival of patients with HNSCC.
The present study has limitations. Our conclusions were based on the results of the metastatic potential of an orthotopic animal model. Although this animal model has the advantage of recapitulating both local tumor growth and pathways of regional spread of cancer (13), to date, there is no perfect animal model that can mimic all features of cancer metastasis, including the tumor microenvironment. The tumor microenvironment, consisting of stromal cells, immune cells, matrix proteins, and soluble factors, has a huge influence on tumor progression and metastasis by providing the necessary stimuli for tumor survival, growth, and invasiveness (35). Further study is mandatory to evaluate the details of the mechanisms underlying FosL1-mediated promotion of tumor progression, metastasis, and microenvironment in HNSCC.
In conclusion, we demonstrated that knockdown of FosL1 reduced tumor migration, invasion, and proliferation in vitro, as well as the incidence of cervical lymph node metastasis of HNSCC with prolonged survival of mice in vivo. Our results suggest that FosL1 could be a key molecule regulating the pathways associated with cervical lymph node metastasis in HNSCC, and thus, pathways involving or interacting with FosL1 might be useful therapeutic targets against regional cervical metastasis of HNSCC.
Acknowledgements
The Authors thank Mari Mitsuka (Department of Biology and Function in Head and Neck, Yokohama City University Graduate School of Medicine, Yokohama, Japan) and Hideaki Mitsui (Department of Pathology, Yokohama City University Graduate School of Medicine, Yokohama, Japan) for their excellent technical assistance.
Footnotes
↵* These Authors contributed equally to this work.
Authors’ Contributions
HH, KS, and DS conceived the study. HH and DS wrote the main manuscript and prepared the figures. HH, KS, HT, TH, KS, YI, SS, KT, TK, and YA were involved with data collection. DS and NO performed the analysis. All Authors discussed the results of the study, made comments on the manuscript, and approved the final manuscript.
Conflicts of Interest
The Authors declare that they have no competing interests in relation to this study.
Funding
This work was supported by JSPS KAKENHI Grant Number 19K09851 (PI: HH) from the Japan Society for the Promotion of Science.
- Received March 2, 2021.
- Revision received May 30, 2021.
- Accepted June 1, 2021.
- Copyright © 2021 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.