Introduction

Exploring the precursor changes of oral mucosal epithelial cells in cancer progression and related prognostic biomarkers have become a hotspot for the early diagnosis and treatment of oral squamous cell carcinoma (OSCC), which aim to improve the diagnostic efficacy based on the specific diagnostic biomarkers in the pre-cancer stage or early stage of the cancerous lesions, and to intervene in the treatment as early as possible, so as to enhance the survival rate of the patients1. In the process of OSCC development, ferroptosis is promoted by induction of tumor cells, it can reduce the physiological effects of tumor resistance and inhibit tumor cell differentiation, invasion and metastasis2. Based on the above reasons, it has become the direction of our in-depth study of the development process of OSCC to explore the mechanism of tumor development process by exploring the pathway of targeted interference of tumor biomarkers associated with ferroptosis.

Cytokeratin is an important cytoskeletal protein that maintains the integrity of epithelial cells and can be used as a common tumor biomarker for OSCC diagnosis3,4. Among the cytokeratin subtypes, CK19, as a typical component of epithelial cells with a protein molecular weight of 40 kDa, is the smallest protein isolated from squamous cell carcinoma cells5. It mainly exists in monolayer epithelial cells and germinal layer cells, and is rarely found in normal oral mucosa and normal skin epidermis6. CK19 is important for predicting the metastatic potential of tumors such as thyroid, cervical and colon cancers and is a useful research tool for tumor diagnosis, treatment and prognosis7. Tanaka S et al. found that CK19 was involved in the invasion and metastasis of OSCC by comparing the clinicopathological features and survival of 100 squamous carcinoma patients, and CK19 may be a novel biomarker for OSCC8.

Previous studies have shown that CK19 can activate the Wnt/β-Catenin signaling pathway and participate in the development of epithelial ovarian cancer9. Wnt/β-Catenin signaling pathway can regulate the expression of glycogen synthase kinase-3 (GSK-3), and GSK-3 co-phosphorylates β-catenin with APC, Fzd and axin, resulting in β-catenin degradation through ubiquitin–proteasome pathway10,11. GSK-3 phosphorylates a set of Ser residues in the Neh6 domain of mouse NF-E2-related factor 2 (Nrf2) to promote its degradation, and Nrf2 is also regulated by GSK-3 activity12,13,14,15. Nrf2 serves as a crucial regulatory factor in ferroptosis, orchestrating the transcription of GPX4 and other pertinent genes to modulate iron homeostasis16,17. In addition, the expression of ACSL4 is also regulated by Nrf2, and some studies have found that TFR1 and ACSL4 can be down-regulated by activating the Nrf2 signaling pathway, thus upregulating GPX18,19. Therefore, we speculated that CK19 might be involved in the OSCC ferroptosis pathway by regulating ACSL4 and GPX4.

As a special programmed cell death regulation pathway, ferroptosis is driven by lipid ROS and regulated by iron ions20. It is significantly different from other types of death regulation mechanisms such as apoptosis, necrosis, autophagy and so on in cell morphology, biochemical and genetic characteristics21. GPX4 and ACLS4 have become the key factors in the regulation of ferroptosis pathway. Some studies have shown that both of them are related to the poor prognosis of OSCC22, but some contrary studies have shown that the expression of GPX4 may not be directly related to the prognosis of OSCC23. Studies have shown that down-regulation of GPX4 can inhibit tumor growth and affect disease progression in OSCC24. Studies have also shown that up-regulation of ACL4 transcription level or stable protein level can promote ferroptosis and inhibit tumor development25. With the in-depth study of the mechanism of ferroptosis, it has been found that promoting ferroptosis of cancer cells can inhibit cancer cell differentiation, migration and reduce drug resistance21. At present, there are few studies on CK19 and ferroptosis in OSCC, and its function and specific mechanism of action are not yet clear, which needs further study.

In this study, we intended to analyze the expression of CK19 and its effect on the prognosis of HNSC by TCGA, and re-confirm the differential expression of CK19 in the development of OSCC, and further explore the specific mechanism of how CK19 affects the ferroptosis pathway to participate in the regulation of OSCC development process through silencing CK19. At the same time, a nude mouse model of silencing CK19 was constructed to verify whether targeted silencing of CK19 affected the expression of ferroptosis-related genes such as GPX4 and ACSL4, thereby regulating ferroptosis of OSCC.

Materials and methods

Cell culture

Human OSCC cell lines (HSC4) were obtained from Sichuan University State Key Laboratory of Oral Diseases. Cells were cultured in DMEM culture medium (Thermo Fisher Scientific, China) with 1% cyan-stranded biclonal antibody and 10% fetal bovine serum (Solarbio, China) at 37 ℃ and 5% CO2. The solution was changed every 2 days. When cell fusion to 80%, cells were passaged for subsequent experiments.

Experimental animals and specimens

Twelve SPF grade BALB/C male nude mice aged 3–4 weeks, weighing 15-18 g, were purchased from Beijing Huafukang Biotechnology Co., Ltd. (Certificate No.: SCXK (Beijing) 2019-0008) and raised in the Experimental Animal Center of Southwest Medical University. The animal study was reviewed and approved by all procedures were approved by the Southwest Medical University Animal Ethics Committee (Permit Number: 201905-7). All experiments were performed in accordance with relevant named guidelines and regulations. The study is reported in accordance with ARRIVE guidelines. All animals were anesthetized and sedated with sodium pentobarbital. Research investigators choose to euthanize mice by cervical dislocation.

A case of well differentiated squamous cell carcinoma surgically resected and pathologically diagnosed in the Affiliated Hospital of Southwest Medical University in March 2023 was collected. After fresh sampling, the specimens were fixed in 10% neutral formalin for 24 h, and then subjected to routine tissue treatment, paraffin embedding, and continuous 4 um tissue sections. The informed consent of the patient and her guardian has been obtained, and the informed consent has been signed. All experiments were performed in accordance with relevant named guidelines and regulations. The study was reviewed and approved by all procedures were approved by medical Ethics Committee of Affiliated Stomatological Hospital of Southwest Medical University (Permit Number: 20180510001).

TCGA dataset analysis

The correlation between tumor and ferroptosis gene, the correlation between CK19 and cancer or normal tissues, and the expression of CK19 in various types of cancer were obtained from the TCGA dataset. KM survival analysis was performed using v4.0.3 version R software (R foundation for statistical computing, 2020) and survival status maps of head and neck cancer patients and CK19 were drawn. Spearman correlation analysis between CK19 and ferroptosis-related genes was performed using the R software package ggstatsplot.

Screen the optimal infection MOI value and lentivirus

The construction of pHS-ASR-1115 (H5), pHS-ASR-1116 (H6) and pHS-ASR-1117 (H7) silencing lentivirus and pHS-ASR-LW429 (H9) empty lentivirus were completed by Shanghai Synbiotics. Five groups of H5, H6, H7, H9 and HSC-4 (blank control cells without reagents) were designed. Each group added different lentiviruses based on MOI = 40. After 24 h of transfection, observe the cell status and exchange fluids, extract total RNA from each group of cells, and perform RT-qPCR to measure the expression of CKl9 mRNA in each group. Interference sequence of lentivirus was shown in Supplementary Table 1.

Construct silencing CK19 lentivirus and transfect HSC-4

CK19-shRNA fragment and silencing CK19 lentivirus was provided by Beijing Hesheng Gene Technology Co., Ltd. CK19-shRNA01, 5′-CCAGCGGCTCATGGACATCAA-3′; CK19-shRNA02, 5′-GCGAAGCCAATATGAGGTCAT-3′; CK19-shRNA03, 5′-CGAGAAGCTAACCATGCAGAA-3′. The shRNA DNA sequence was inserted into the lentiviral vector pLV-hU6-shRNA-Puro. Sequencing primer: pHS-ASR-1115, 5′-CGGGTTTATTACAGGGACAGCAG-3′; pHS-ASR-1116, 5′-CGGGTTTATTACAGGGACAGCAG-3′; pHS-ASR-1117, 5′-CGGGTTTATTACAGGGACAGCAG-3′. Lentivirus transfection: 24 h before transfection, HSC4 cells were digested and inoculated in 6-well plates. When the cell fusion rate reached 30%, the lentivirus was transfected into HSC4 cells, and the medium was changed after 12 h of culture, and the culture was continued. The cells were divided into silencing group, empty group and untransfected group.

Real-time q-PCR analysis

Total RNA was extracted from each group of cells, the RNA concentration was determined, and the cDNA was reversely transcribed, and the SYBR method was used for the experiment. The relative expression of mRNA in each group was calculated using the 2−ΔΔCt method. All primers were synthesized by Shanghai Shenggong Co., Ltd. The primer sequence was shown in Supplementary Table 2.

Immunohistochemical analysis

The cells in the silencing group and the control group were prepared, and the cells were required to be evenly distributed on the slides, and 50–60% was appropriate. Rabbit anti-human ACSL4 polyclonal antibody (Abways, China) and rabbit anti-human GPX4 polyclonal antibody (Abways, China) were prepared according to 1:100, and PBS buffer was used as negative control. The slides were fixed with 10% neutral formaldehyde solution for 20 min and incubated with 3% hydrogen peroxide for 10 min. The primary antibody was incubated overnight at 4 ℃ and then incubated with goat anti-rabbit TRITC (Bioss, China) in the dark. The nuclei were stained with DAPI. Finally, the slide was sealed with glycerin, the fluorescence was observed, and the image was analyzed.

Wound healing assay

The control group and the silencing group were prepared with 6 mL of cell suspension at a concentration of about 3–5 × 104 cells/mL, and 2 mL was added to the hole to ensure that the cell suspension was evenly spread on the bottom of the plate. Each group was repeated three times. On the second day, cells were cultured to be more than 90% confluence, the monolayer was scratched the tip of a 100 μL sterile pipette. After washing the scratched cells with PBS (Solarbio, China), serum-free medium was added. After 24 h and 48 h of cell culture in the silence group and the control group, the photographs were observed and recorded under the microscope (OLYMPUS, Japan).

Cell migration assay

0.6 mL of 20% serum medium was added to the lower chamber of the cell chambers, and the upper chamber was inoculated with 0.1 mL of pure DMEM medium re-suspended at a concentration of 3 × 104 cells/mL in the silencing and control groups. After 48 h of incubation, the cells at the bottom of the upper chamber were wiped off. The cell structure was fixed with 4% paraformaldehyde. 700 μL of 1:1000 DAPI solution (Solarbio, China) was added to each well of a 24-well plate and stained with light for 30 min. The DAPI solution was washed off. Five fields of view were randomly selected for counting under a Leica fluorescence microscope (OLYMPUS, Japan), and cells were counted by Image software.

Western blot assay

Cells were lysed with RIPA tissue/cell lysate: PMSF (100:1, Solarbio, China). Protein concentration was determined using a BCA kit (Solarbio, China) at a volume ratio of A:B = 1:50. The samples were electrophoresed and transferred to PVDF membranes. The PVDF membrane was closed in skimmed milk for 1 h, incubated with primary antibody at 4℃ overnight and secondary antibody at room temperature for 1 h.

MDA content assay

MDA levels were detected using an MDA detection kit (Solarbio, China) according to the instructions. MDA content (nmol/104 cell) = (12.9 × (ΔA532–ΔA600)–2.58 × ΔA450) × Vtotal ÷ (400 ÷ Vextract × Vsample) = 0.125 × (12.9 × (ΔA532–ΔA600)–2.58 × ΔA450).

Cellular Fe2+ content assay

Fe2+ levels were measured using the Cellular Ferrous Ion Colourimetric Assay Kit (Elabscience, China) according to the instructions.

Reactive oxygen (ROS) level assay

The cells in the silence group and the control group were inoculated on the plate, and each group was set up with 3 vice holes. After 24 h, the cells in the 6-well plate were observed. When the plate reached 85%, 2 mL of 10 μmol/L DHE probe was added, and 3 μL of 3% H2O2 was added to each well. The cells were cultured in the incubator away from light for 30 min. After incubation, the cells were washed three times with PBS and observed under a laser microscope (OLYMPUS, Japan).

Subcutaneous tumor formation in nude mice

The two kinds of cells transfected with CK19 silencing lentivirus and empty lentivirus in good growth state were digested and resuspended in serum-free medium, 0.2 ml/only (final concentration of 1 × 107) was inoculated subcutaneously in the right upper limb of nude mice according to the group to establish the animal model of tumor-bearing nude mice.

HE staining

Nude mice with loaded tumors were sacrificed and then the transplanted tumors were embedded in conventional paraffin, and the tissue specimens were cut into 4 μm-thick sections. In a fume hood it was first dewaxed then soaked in hematoxylin stain, then fixed in xylene solution, and finally closed with an appropriate amount of neutral gum and observed under a microscope.

TUNEL assay for detection of apoptosis

The stripped transplanted tumor tissue was fixed, embedded in paraffin, and sliced at 4 μm thickness, and then operated according to the instructions of the TUNEL detection kit. The results were immediately photographed under a fluorescence microscope and analyzed by Image J.

Statistical analysis

SPSS25.0 and Graph Pad Prism were used for statistical processing and analysis. All quantitative results were expressed as mean ± standard deviation (Mean ± SD). When continuous variables met normal distribution and variance chi-square test showed that the variance was uniform, the methods used were independent samples t-test or ANOVA. For continuous variables that did not meet the normal distribution, the rank-sum test was used to perform the statistical analysis. For the categorical variables expressed as frequencies (percentages) and analyzed by the chi-square test. P < 0.05 was as statistical significance.

Ethical approval

The animal study was reviewed and approved by all procedures were approved by the Southwest Medical University Animal Ethics Committee (Permit Number: 201905-7). The study was reviewed and approved by all procedures were approved by medical Ethics Committee of Affiliated Stomatological Hospital of Southwest Medical University (Permit Number: 20180510001).

Results

The expression of CK19 may not be related to the survival prognosis of patients with HNSC

The TCGA dataset showed CK19 expression was not significantly different in HNSC, it was significantly down-regulated in KICH, LGG, ACC, DLBC, GBM, LIHC, SKCM, UVM, and SARC, and up-regulated in BLCA, BRCA, CESC, CHOL, COAD, ESCA, OV, and MESO (Fig. 1A). Figure 1B shows the relationship between CK19 expression level and survival prognosis of patients with HNSC, and ROX analysis suggested that the expression level of CK19 had no significant correlation with survival time of patients with HNSC.

Figure 1
figure 1

Results of the TCGA dataset analysis and RT-qPCR (A) Picture of expression of CK19 gene in tumor and normal tissue (B) KM survival curve distribution of survival time and survival status in patients with CK19 and HNSC tumors note: Distribution of CK19 and KM survival curves in the HNSC database of TCGA (p = 0.72), with a sample size of 130 for the high CK19 expression group and 389 for the low to moderate CK19 expression group. (C) The expression of CK19 in H5, H6, H7, H9 and HSC-4 untransfected groups. *P < 0.05, ***P < 0.001.

Full size image

HSC-4 human oral squamous cell line was used as the experimental cell model

Due to the HSC-4 cell line had a high content of CK19, it was selected for subsequent experiments. When MOI = 40, the amount of virus used was the least, so the amount of virus with MOI = 40 was selected for subsequent experiments. As qPCR results showed, compared with other groups, the effect of lentivirus H6 silencing CK19 gene in HSC-4 cells was good (Fig. 1C). HSC-4 cells transfected with lentivirus pHS-ASR-1116 (H6) and empty lentivirus pHS-ASR-LW429 were respectively used as the silencing group and control group for subsequent experiments.

Silencing CK19 could reduce the migration rate of HSC-4 cells

After 48 h of cell scratch test, the migration rate of the silencing group was lower than that of the control group. After silencing CK19, the migration rate of HSC-4 cells slowed down (Fig. 2A,B). After 24 h of cell migration in the silencing group and the control group, the number of migrated cells in the silencing group was lower than that in the control group. After silencing CK19, the migration rate of HSC-4 cells slowed down (Fig. 2C,D).

Figure 2
figure 2

Wound healing assay and cell migration assay. (A) Observation of cell migration under the microscope, (B) Histogram of cell migration and healing area percentage, (C) Number of migrating cells observed under the microscope, (D) Histogram of the number of migrating cells. Date are shown as the mean ± SD (n ≥ 3). ***P < 0.001.

Full size image

ACSL4 and GPX4 acted as key genes for ferroptosis associated with CK19

The TCGA database found that ferroptosis-related genes GPX4, ACSL4, CDKN1A, HSPA5, EMC2, SLC7A11, LPCAT3, CARS, ALOX15, FANCD2, CICD1, SLC1A5, SAT1, TFRC, and RPL8 were significantly different from normal tissues in various tumors (Fig. 3A). The linear regression analysis showed that ACSL4 was negatively correlated with CK19, while GPX4 was positively correlated with CK19. Therefore, GPX4 and ACSL4 were screened to represent the genes of ferroptosis as subsequent study (Fig. 3B).

Figure 3
figure 3

The relationship between ferroptosis-related genes and CK19 was analyzed by TCGA database. (A) Scatter plots of ferroptosis related genes associated with tumor and normal tissue (B) Correlation regression between CK19 gene and ferroptosis related genes ACSL4 and GPX4.

Full size image

Silencing CK19 significantly increased the expression of ACSL4 and decreased the expression of GPX4

The expression of ACSL4 and GPX4 in HSC-4 cells transfected with silencing CK19 and control groups were analyzed by immunofluorescence staining, and the results showed that ACSL4 and GPX4 were expressed in all groups, and mainly in the cytoplasm (Fig. 4A,B). By counting the fluorescence intensity, it was found that the expression of GPX4 was significantly lower and the expression of ACSL4 was significantly higher in the CK19-shRNA group, compared with control group (Fig. 4C). Furthermore, Western blot showed the similarly results as above (Fig. 4D–G).

Figure 4
figure 4

The expression of ACSL4 and GPX4 in silent and control groups. (A) Cellular immunofluorescence of ACSL4 (inverted microscope,  × 200) bar = 200 µm, (B) Cellular immunofluorescence of GPX4 (inverted microscope,  × 200) bar = 200 µm, (C) Fluorescence intensity statistics (D) Western blot analysis of CK19, ACSL4 and GPX4 (Original blots/gels are presented in Supplementary Fig. 1) (E–G) Gray value statistics. *P < 0.05, **P < 0.01.

Full size image

Silencing CK19 could increase content of MDA, ROS and Fe2+

We measured the ROS, MDA and Fe2+ contents in the CK19-shRNA group and the control group respectively. The results showed that the content of ROS and MDA in the CK19-shRNA group was higher than that in the control group (Fig. 5A–C), indicating that the oxidative stress response of ferroptosis in HSC-4 cells was enhanced after CK19 silencing. And there were higher Fe2+ contents in the CK19-shRNA group (Fig. 5D).

Figure 5
figure 5

The content of MDA, ROS and Fe2+ in silent and control groups. (A) ROS fluorescence in silent and control groups. (× 200) bar = 200 µm (B) Fluorescence intensity statistics. (C) The content of MDA in silent and control groups. (D) The content of Fe2+ in silent and control groups. ***P < 0.001.

Full size image

Silencing CK19 significantly increased the expression of ACSL4 and decreased the expression of GPX4 and could increase content of MDA and Fe2+ in nude mice

We inoculated nude mice with CK19-shRNA group and control group cells, and found that 12 nude mice all formed tumors after 30 days, and the tumor formation rate was 100%. In the CK19-silenced group, the tumor volume and weight were significantly smaller than that of the control group (Fig. 6A). The tumor tissues of the two groups of nude mice had heterogeneity and obvious tumor nodule structure, the cells were irregularly arranged, and the nuclear-to-cytoplasmic ratio of the cells increased, and the nuclear size was different, showing the characteristics of tumor cells. The control group could see obvious tumor necrosis area (Fig. 6B). Functional experiments were performed to determine the role of silencing CK19 in OSCC tumor cells in vivo. TUNEL staining results showed that silencing CK19 could promote apoptosis of tumor cells in vivo (Fig. 6C). Immunohistochemical staining results showed that the expression of CK19 in OSCC tissue was significantly higher than that in normal oral mucosa (Fig. 6D), and the expression of CK19 in tumor-bearing nude mice was significantly higher than that in normal epithelium of nude mice (Fig. 6E).

Figure 6
figure 6

HE staining, TUNEL staining and immunohistochemical analysis. (A) Control and silent groups stripped of transplanted tumors and statistical analysis of weight volume. ***P < 0.001 (B) HE staining of stripped graft tumor tissue in both groups. Bar = 200 µm (C) TUNEL staining of tumor tissue (× 200) and average fluorescence intensity statistics. Bar = 100 µm, *P < 0.05 (D) The expression of CK19 in oral squamous cell carcinoma and normal oral mucosa and statistical analysis. Bar = 100 µm, **P < 0.01 (E) The expression of CK19 in tumor-bearing nude mice and normal epithelium of nude mice and statistical analysis. Bar = 100 µm, ***P < 0.001.

Full size image

Further, ACSL4 was highly expressed and GPX4 was lower expressed in transplanted tumors in CK19-shRNA mice group (Fig. 7A). Western blot was used to analyze the expression of each protein in CK19 silencing group and control group. CK19, GPX4 protein levels were significantly lower and ACSL4 levels were significantly higher in the CK19-silenced group compared to the control group (Fig. 7B). Detection of Fe2+ levels and MDA content within stripped graft tumor serum showed that silencing CK19 significantly increased Fe2+ levels and MDA content (Fig. 7C,D).

Figure 7
figure 7

The expression of ACSL4 and GPX4 in silent and control groups by immunohistochemical analysis and western blotting. (A) The expression of ACSL4, GPX4 and CK19 in silent and control groups and statistical analysis bar = 100 µm, *P < 0.05 (B) Western blot showed the relative expression of CK19, ACSL4 and GPX4 in silent and control groups and statistical analysis (Original blots/gels are presented in Supplementary Fig. 1). *P < 0.05, **P < 0.01 (C,D) The MDA and Fe2+ content in transplanted tumors. **P < 0.01.

Full size image

Discussion

In this study, the mRNA expression profile of CK19 and the clinical information of HNSC patients were collected from the TCGA database for bioinformatics analysis. We found that CK19 was significantly expressed in pan-cancer tissues, and there was no significant difference in the expression of CK19 between tumor and normal tissues. This is contrary to the conclusion that CK19 is significantly highly expressed in tumor tissues found in previous studies and clinical practice26,27, which needs further experimental verification. Through univariate Cox analysis and KM analysis, the prognostic gene characteristic curve was established. It was found that there was no significant statistical difference in CK19 gene expression difference analysis and survival analysis of HNSC, which may not be directly related to prognosis. However, this experiment showed that the migration ability of HSC4 cells was weakened after silencing CK19 through cell scratch test and migration test. It was also verified by immunohistochemistry that the expression of CK19 in human squamous cell carcinoma tissues and mouse tumor-bearing tissues was higher than that in normal tissues. This result was consistent with the previous study that the expression level of CK19 in the development process of OSCC was up-regulated with the degree of abnormal hyperplasia of oral mucosal epithelium28,29,30. The above shows that the data of HNSC as a sample in the bioinformatics analysis cannot objectively reflect the expression trend of CK19 in OSCC and normal oral mucosa, which may be because OSCC is only a part of HNSC.

The effect of silencing CK19 on tumor growth and apoptosis was further investigated from the perspective of in vivo experiments by establishing a nude mouse xenograft model. The results showed that the final weight and volume of transplanted tumors in the silent CK19 group were significantly lower than those in the control group, and the apoptotic cells in the tumor tissues of the silent CK19 group were significantly increased. These results indicate that after silencing CK19, the tumor grows slowly, the tumor cells are well differentiated, the degree of malignancy is reduced, and the apoptosis of tumor cells can be promoted. As in a study, the release of serum CYFRA 21-1 (a fragment of CK19) in liver cancer cells is closely related to tumor cell apoptosis31.

Ferroptosis is a novel form of programmed cell death that is different from apoptosis, necrosis or autophagy, and is an anti-tumor mechanism that is essential for eliminating malignant cells32. Hassannia B et al. found that ferroptosis-related genes FRGs and pathways may also be involved in the development of OSCC and can be used to develop prognostic markers33. Tumor cells have a greater demand for iron than normal cells, which makes them more susceptible to ferroptosis33. When iron overload occurs in cells, excessive free Fe2+ reacts with H2O2 to produce a large number of toxic products and ROS to promote ferroptosis34. Ferroptosis is controlled by GPX4, and GPX4 inactivation is thought to lead to cell death35. In this study, TCGA database was used to analyze the differences of ferroptosis-related genes between oral cancer and adjacent normal tissues, and a variety of differentially expressed ferroptosis genes were screened out. The correlation analysis between CK19 and candidate differential genes showed that two classical ferroptosis-related genes ASCL4 and GPX4 were significantly correlated with CK19. And studies had shown that CK19 could enhance the invasion, migration and proliferation of cancer cells through the Wnt/β-catenin pathway9. CK19 may regulate the expression of GSK-3 by activating Wnt/β-Catenin signaling pathway, and then affect AMPK/Nrf2/GPx4 pathway, and finally regulate the expression of GPX4 and ACSL411,12,13,16. GPX4 can convert lipid hydroperoxides into fatty alcohols. Inhibition of GPX4 expression leads to lipid peroxidation and can lead to ferroptosis25,36,37. Studies have shown that ionizing radiation can induce the expression of ACSL4, leading to lipid peroxidation and ferroptosis38,39,40. This is consistent with our findings that silencing CK19 promotes ferroptosis and inhibits tumor growth by activating ACSL4 expression and inhibiting GPX4 expression. However, the pathways through which CK19 completes this process are still unclear, and further research is needed.

Ferroptosis can cause dysfunction of the antioxidant system and lead to a decrease in the overall antioxidant capacity of cells41. Mitochondria are damaged with the increase of external stimuli. When mitochondria are severely damaged, they lose the ability to regulate ROS, which further leads to an increase in ROS41. Ferroptosis is also closely related to the content of MDA and Fe2+. Previous studies have also used the content of MDA and Fe2+ as key indicators to explore the regulation of ferroptosis in OSCC42,43,44. We found that the content of Fe2+ and MDA in the silent CK19 group was significantly higher than that in the control group, indicating that silencing CK19 can promote ferroptosis. In the future, we will verify our conclusions by restoring CK19, overexpressing CK19, overexpressing ACSL4 and GPX4 and using other ferroptosis inhibitor and activator.

In summary, silencing CK19 can affect the expression of GPX4 and ACSL4 to regulate ferroptosis and at the same time increase the content of MDA, Fe2+ and ROS levels, thereby activating the regulation of ferroptosis pathway in the development of OSCC.