From January 2015 to December 2022, the samples (30 pairs of HNSCC tissues and matched normal tissues) were obtained from The Third Central Hospital of Tianjin. According to clinical diagnosis, 30 cases of HNSCC were determined, and 30 cases of adjacent normal tissue (more than 2 cm away from cancer tissue) were selected as control group. Inclusion criteria: first-episode patients who had not received other treatment before this surgery; the patient did not receive preoperative radiotherapy, chemotherapy, and intervention, as well as all disease-related management; postoperative pathological diagnosis of HNSCC; patients who have not been diagnosed and treated in other hospitals. Exclusion criteria: previous chemoradiotherapy; patients with multiple primary tumors; liver, kidney, and heart dysfunction combined with other serious diseases; secondary HNSCC.

ATCC (USA) provided the human normal oral epithelial keratinocyte (HOK). CAL27 and SAS were purchased from Bohui Biotech (Guangzhou, Cina). CAL33 and SCC25 cells were purchased from Wanwu Biotech (Hefei, China). DMEM (CAS # 9001331600, Beijing Solarbio Science & Technology Co., Ltd., Beijing, China), with 10% fetal bovine serum (FBS, CAS # 10099-141, Gibco, Thermo Fisher Scientific, Waltham, MA, USA), streptomycin (100 mg/mL, CAS # S8290, Beijing Solarbio Science & Technology Co., Ltd., Beijing, China), and penicillin (100 U/mL, CAS # IA0340, Beijing Solarbio Science & Technology Co., Ltd., Beijing, China), was utilized to culture all cells. The cells were cultured in T75 cell culture flasks (CAS # 708013, Wuxi Nesi Life Technology Co., Ltd., Wuxi, China) at 37 °C and 5% CO${}_2$ (v/v) for about 48 h, and then the cells were digested with 0.25% trypsin (CAS # T1320, Beijing Solarbio Science & Technology Co., Ltd., Beijing, China) and passed in a ratio of 1:3. All cell lines were validated by STR profiling and tested negative for mycoplasma. Cells were all cultured in a humidified incubator at 37 °C and 5% CO${}_2$.

One day before transfection studies, six-well plates were used for HNSCC cell culturing (1.2 $\times$ 10${}^6$). The cells were transfected when the cell fusion degree reached approximately 80% on the second day. pcDNA3 vector (NC) was purchased from Youbao Bio (CAS # VT2051, Hunan, China). pcDNA3-DDX60 (DDX60) was constructed by General Biology (Anhui) Co., Ltd. (Anhui, China). Small interfering RNA control (si-NC, 5${}^{\prime }$-UUCUCCGAACGUGUCACGUTT-3${}^{\prime }$) and small interfering RNA-DDX60 (si-DDX60, 5${}^{\prime }$-CCACUUUCCUACGAAUUGUTT-3${}^{\prime }$) were synthesized by General Biology (Anhui) Co., Ltd. (Anhui, China). Lipofectamine 2000 (CAS # 11668019, Invitrogen, Waltham, MA, USA) was employed to transfect the HNSCC cells with the plasmids or siRNAs, following the instructions outlined by the manufacturer. Specific transfection steps were implemented as per the direction specified for Lipofectamine 2000.

Total RNA was isolated from cancer tissues and cells in each group utilizing TRIzol Reagent (CAS # R1100, Beijing Solarbio Science & Technology Co., Ltd., Beijing, China). cDNA was reverse-transcribed with an EasyScript® One-Step gDNA Removal and cDNA Synthesis Super Mix (CAS # AW311-02, TransGen Biotech, Beijing, China). An SYBR Green qPCR Mix Kit (CAS # BL697A, Biosharp, Beijing, China) was employed to perform SQ RT-PCR. The 2${}^{-\mathrm{\Delta }\mathrm{\Delta }\text{Ct}}$ method was applied to calculate the relative expression. $\beta$-actin was used as the internal reference for DDX60, IFI27, Vimentin, E-cadherin and N-cadherin. Supplementary Table 1 summarizes the primers used for SQ RT-PCR.

The Transwell experiment was conducted following a previously reported method [20]. In brief, for Transwell migration experiments, we resuspended the transfected cells (2 $\times$ 10${}^4$) in serum-free DMEM (200 µL) and inoculated them in each chamber of the upper chamber (CAS # 3470, Corning Incorporated, Corning, NY, USA). For the Transwell invasion experiment, Matrigel (40 µL, 1 mg/mL, CAS # 356234, Beijing Solarbio Science & Technology Co., Ltd., Beijing, China) was added in the upper chamber, and then DMEM without serum (200 µL) was used to resuspend the transfected cells (5 $\times$ 10${}^4$), followed by inoculating them in each chamber of the upper chamber. In the bottom chamber, 24-well plates were filled with DMEM complete medium (600 µL) supplemented with 20% FBS. Then, 4% paraformaldehyde (PFA) was used to fix cells in the bottom chamber for 20 min after 48 h of incubation. Cell staining was performed with 0.1% crystal violet for 5 min. Transwell cells were randomly assigned to three areas for shooting and counting analyses.

RIPA lysate (CAS # P0013B, Beyotime, Shanghai, China) containing the protease inhibitor mixture (CAS # 04693116001, Roche, Mannheim, Germany) was used to extract total proteins. The bicinchoninic acid test (CAS # P0010S, Beyotime, Shanghai, China) was employed to determine the total protein concentration. Thirty micrograms of protein lysates were separated with 6% to 15% SDS-PAGE according to the molecular weight. Subsequently, they were transferred onto the PVDF membrane (CAS # 0000187216, 0.45 µm, Millipore Corp., Burlington, MA, USA). The PVDF membrane was first blocked by 5% BSA (CAS # ST2249-100g, Beyotime, Shanghai, China) and then subjected to incubation in the primary antibody throughout the night at 4 °C. Then, the goat anti-rabbit secondary antibody, labeled with HRP, was introduced to incubate the membrane (1:5000, CAS #pr30011, Proteintech, Wuhan, Hubei, China) for 2 h at room temperature (RT). Subsequently, the Tanon-5200 automatic chemiluminescence imaging analysis system (Tanon, Shanghai, China) was applied to identify the bands. ImageJ (version 1.60; National Institutes of Health, Bethesda, MA, USA) was used to quantify the data obtained using western blotting. Primary antibodies included DDX60 polyclonal antibody (Rabbit, 1:500, CAS #ab139807, Abcam, Boston, MA, USA), N-cadherin polyclonal antibody (Rabbit, 1:3000, CAS # 22018-1-AP, Proteintech, Wuhan, China), E-cadherin polyclonal antibody (Rabbit, 1:5000, CAS # 20874-1-AP, Proteintech, Wuhan, China), Vimentin polyclonal antibody (Rabbit, 1:5000, CAS # 10366-1-AP, Proteintech, Wuhan, China), p65 recombinant antibody (Rabbit, 1:5000, CAS # 80979-1-RR, Proteintech, Wuhan, China), p-p65 monoclonal antibody (Rabbit, 1:1000, CAS # ab76302, Abcam, Cambridge, MA, USA), IFI27 polyclonal antibody (Rabbit, 1:500, CAS # ab171919, Abcam, Cambridge, MA, USA), and $\beta$-Tubulin polyclonal antibody (Rabbit, 1:8000, CAS # 10094-1-AP, Proteintech, Wuhan, China).

The experiments were repeated three times. The format of mean $\pm$ standard error of the mean (SEM) was used to present the collected data. The data were statistically analyzed, and GraphPad Prism 8.0 (GraphPad Software, Inc., San Diego, CA, USA) was applied for plotting. Student’s t-tests were used to understand the differences between the two subgroups. Multiple subgroups were compared via a one-way analysis of variance and Tukey post-hoc test. Based on the distribution patterns of indicators, Pearson or Spearman correlation tests were used to estimate the association between variables. Using the survival package, univariate and multivariate Cox regression models were used to assess the survival significance of variables (p 0.05; statistically significant).

We explored the expression level of DDX60 and verified its regulatory mechanism in HNSCC to investigate the role of DDX60. First, the Gene Expression Profiling Interactive Analysis (GEPIA; http://gepia.cancer-pku.cn/) database was searched to predict the DDX60 mRNA level in pan-cancers. Through analysis, DDX60 was upregulated in several cancers as well as HNSCC (Fig. 1A). In addition, the Encyclopedia of RNA Interactomes (ENCORI; https://starbase.sysu.edu.cn/) database was searched to further verify the mRNA level of DDX60 in HNSCC (Fig. 1B). The survival curve was analyzed via GEPIA and ENCORI, and patients who were suffering from HNSCC and were characterized by high mRNA levels of DDX60 exhibited a lower 10-year overall survival rate than patients with HNSCC who were characterized by a lower expression level of DDX60 (Fig. 1C,D). Thirty pairs of tissues (HNSCC tumor and adjoining normal tissues) were used to study the DDX60 mRNA levels. The results revealed the importance of upregulated DDX60 in HNSCC tissues. The DDX60 mRNA level was increased in HNSCC tissues as opposed to the adjoining non-tumor tissues (Fig. 1E). Moreover, HNSCC tissues also exhibited DDX60 upregulation as opposed to normal tissues (Fig. 1F). To investigate the role of DDX60 in HNSCC, four cell lines originated from tongue were selected: CAL27, SAS, CAL33, and SCC25. We also studied the mRNA and protein levels of DDX60 in human normal oral epithelial keratinocytes (HOK) and these four HNSCC cells. Results showed that both DDX60 mRNA and protein levels were found to be elevated in the four HNSCC cell lines relative to the normal cell line (Fig. 1G,H).

Fig. 1.

DEXD/H box helicase 60 (DDX60) expression was upregulated in HNSCC tissues and cells. (A) Analysis results of the data from the GEPIA database show that DDX60 was upregulated in DLBC, ESCA, GBM, HNSCC, LAML, LGG, PAAD, and THYM. (B) Analysis results of the data from the ENCORI database confirmed that DDX60 was upregulated in HNSCC. (C,D) Survival curves of patients with HNSCC predicted by the GEPIA database (C) and ENCORI database (D) based on the mRNA levels of DDX60. (E) The mRNA levels of DDX60 in 30 pairs of HNSCC and its paraneoplastic tissues were detected by SQ RT-PCR. (F) The DDX60 expression levels in 4 pairs of HNSCC and its paraneoplastic tissues were detected by western blot. (G,H) The DDX60 mRNA and protein levels in human normal oral epithelial keratinocytes (HOK) and four HNSCC cells (CAL27, SAS, CAL33, and SCC25) were detected by SQ RT-PCR (G) and western blot (H). *p 0.05, **p 0.01, ***p 0.001, ****p 0.0001. DLBC, Lymphoid neoplasm diffuse large B-cell lymphoma; ESCA, Esophageal carcinoma; GBM, Glioblastoma; HNSCC, Head and neck squamous cell carcinoma; LAML, Acute myeloid leukemia; LGG, Lower-grade glioma; PAAD, Pancreatic adenocarcinoma; THYM, Thymoma; mRNA, messenger RNA; SQ RT-PCR, Semi-Quantitative Real-Time PCR; GEPIA, Gene Expression Profiling Interactive Analysis.

Tumor initiation, invasion, and metastasis all require the involvement of epithelial-mesenchymal transition (EMT), one of the hallmarks of malignant tumors [21, 22, 23]. Therefore, we initially investigated whether DDX60 affected EMT in HNSCC cells. Transwell migration and invasion assay results revealed that DDX60 overexpression could enhance cell migration and invasion abilities of HNSCC cells, whereas DDX60 expression was silenced, and these processes were inhibited (Fig. 2A–D). Thus, we hypothesize that DDX60 promoted EMT in HNSCC cells.

Fig. 2.

DDX60 enhanced the migratory and invasive capacities of HNSCC cells. (A,B) Effect of DDX60 overexpression and knockdown on migration in HNSCC cells. (C,D) Effect of DDX60 overexpression and knockdown on invasion in HNSCC cells. Bars = 100 µm. **p 0.01, ***p 0.001, ****p 0.0001. NC, negative control.

EMTs are characterized by the absence of epithelial cell markers (e.g., E-cadherin) and upregulation of mesenchymal cell markers (like vimentin and N-cadherin) [24]. We investigated how DDX60 affected EMT markers, such as Vimentin, E-cadherin, and N-cadherin. SQ RT-PCR results showed that forced overexpression of DDX60 remarkably elevated the mRNA levels of N-cadherin and Vimentin while reducing those of E-cadherin (Fig. 3A–D). Conversely, DDX60 knockdown could reduce the mRNA levels of N-cadherin and Vimentin while increasing those of E-cadherin (Fig. 3A–D). In addition, increased protein levels of N-cadherin and Vimentin but reduced those of E-cadherin were observed after DDX60 overexpression, whereas the opposite results occurred by downregulation of DDX60 in HNSCC cells (Fig. 3E–I).

Fig. 3.

HNSCC cells’ epithelial-mesenchymal transition (EMT) was enhanced by DDX60. (A–D) The mRNA levels of DDX60 (A), N-cadherin (B), Vimentin (C), and E-cadherin (D) in CAL27 and SAS of HNSCC cells were detected by SQ RT-PCR. (E) Western blotting was implemented to determine the expression levels of DDX60, N-cadherin, Vimentin, and E-cadherin expressed by CAL27 and SAS HNSCC cells. (F–I) Quantification of protein levels of DDX60 (F), N-cadherin (G), Vimentin (H), and E-cadherin (I) in (E). *p 0.05; **p 0.01; ***p 0.001; ****p 0.0001.

To further explore the potential mechanism of DDX60 in HNSCC, we performed the correlation analysis using GEPIA. According to the findings, IFI27 (interferon alpha inducible protein 27) mRNA level was considerably elevated in HNSCC tissues in comparison to the normal tissues, and IFI27 was positively corrected with DDX60 (Fig. 4A,B). Patients with HNSCC who had high levels of IFI27 mRNA had a lower 10-year overall survival rate in comparison to those with lower levels (Fig. 4C). Furthermore, the above prediction results were verified by SQ RT-PCR (Fig. 4D,E). Moreover, overexpression or knockdown of DDX60 promoted or inhibited IFI27 at the protein and mRNA levels (Fig. 4F–I).

Fig. 4.

DDX60 was positively corrected with interferon alpha inducible protein 27 (IFI27) in HNSCC cells. (A) The relationship between DDX60 and the mRNA level of IFI27 was established by the GEPIA database. (B) IFI27 mRNA levels in HNSCC tissues and control tissues were predicted utilizing data from the GEPIA database. (C) A correlation between IFI27 mRNA levels and HNSCC patient survival was predicted using the GEPIA database. (D) IFI27 mRNA level was measured by SQ RT-PCR in 30 matched pairs of HNSCC and its paraneoplastic tissues. (E) Analysis of the correlation between DDX60 and IFI27 according to their mRNA levels in 30 pairs of HNSCC and its paraneoplastic tissues detected by SQ RT-PCR. (F) The mRNA levels of DDX60 and IFI27 in CAL27 and SAS of HNSCC cells were detected by SQ RT-PCR. (G) The expression levels of DDX60 and IFI27 in CAL27 and SAS of HNSCC cells were detected by western blot. (H,I) Quantification of DDX60 (H) and IFI27 (I) protein levels in (G). *p 0.05; **p 0.01; ***p 0.001; ****p 0.0001. TPM, transcript per million.

Based on the above results, we ask whether DDX60 regulated the EMT through the IFI27-mediated mechanism. A rescue experiment was implemented to validate the hypothesis, and the results illustrated that silencing IFI27 in HNSCC cells attenuated the DDX60-induced upregulation of N-cadherin and Vimentin mRNA levels and the E-cadherin inhibition (Fig. 5A–E). In addition, a similar phenomenon has been observed at the protein level (Fig. 5F–K). Then, further research was conducted to determine whether DDX60 influenced the migratory and invasive capacities of HNSCC cells via modulation of IFI27. The analysis showed that HNSCC cells’ migratory and invasive capacities were remarkably reduced upon DDX60 knockdown, whereas co-transfection with IFI27 overexpressing plasmid removed this inhibition (Fig. 5L–N). Thus, we hypothesize that these processes can be promoted by DDX60 via the upregulation of the IFI27 in HNSCC cells.

Fig. 5.

DDX60 regulated the EMT via the IFI27-mediated mechanism in HNSCC cells. (A–E) After co-transfecting HNSCC cells with NC + si-NC, DDX60 + si-NC, and DDX60 + si-IFI27, the mRNA levels of DDX60 (A), IFI27 (B), N-cadherin (C), Vimentin (D), and E-cadherin (E) were determined via SQ RT-PCR. (F) Western blotting was used to determine DDX60, IFI27, E-cadherin, Vimentin, and N-cadherin expression levels. (G–K) Quantification of protein expression levels of DDX60 (G), IFI27 (H), N-cadherin (I), Vimentin (J), and E-cadherin (K) in (F). (L) After co-transfecting HNSCC cells with si-NC + NC, si-DDX60 + NC, si-DDX60 + IFI27, the migratory and invasive capacities of HNSCC cells were measured using transwell assays. (M) Transwell migration and (N) transwell invasion abilities were quantified from (F). Bars = 100 µm. *p 0.05; **p 0.01; ***p 0.001; ****p 0.0001; ns, no significance.

We delved further into the molecular mechanism via which DDX60 upregulates IFI27 in HNSCC cells. When comparing patients with high and low DDX60 expression, GSEA revealed that the NF-$\kappa$B pathway was significantly enriched in the former group (Fig. 6A), indicating that the NF-$\kappa$B pathway might be activated by DDX60.

Fig. 6.

DDX60 regulated IFI27 by activating the NF-$\kappa$B pathway. (A) GSEA results indicated that DDX60 was significantly and positively correlated with NF-$\kappa$B in the analysis of the TCGA datasets. (B) After overexpression or knockdown of DDX60, the p65 and p-p65 protein levels were detected by western blot. (C,D) Quantification of p65 (C) and p-p65 (D) protein levels in B. (E) After transfection NC+si-NC, DDX60+si-NC, and DDX60+si-p65, western blotting was used to determine the protein levels of DDX60, p65, p-p65, and IFI27. (F–I) Quantification of protein expression levels DDX60 (F), p65 (G), p-p65 (H) and IFI27 (I) in E. (J) Schematic representation of the proposed DDX60 regulation of the IFI27 expression via the NF-$\kappa$B–mediated mechanism. *p 0.05; **p 0.01; ***p 0.001; ns, no significance. GSEA, gene set enrichment analysis; TCGA, The Cancer Genome Atlas.

To further our understanding of how DDX60 works in HNSCC cells, we examined how changed DDX60 expression affected the protein levels of p65 and p-protein levels, two key molecular components in the NF-$\kappa$B pathway. We discovered that the protein level of p65 did not change significantly regardless of DDX60 levels (Fig. 6B–D). However, while p-p65 protein levels rose remarkably during DDX60 overexpression, they dropped remarkably after DDX60 knockdown (Fig. 6B–D). The direct transcription of IFI27 is induced by the NF-$\kappa$B pathway activation, and previous research has shown that the promoter region of IFI27 contains NF-$\kappa$B-p65 binding sites [25]. Therefore, we investigated whether DDX60 could upregulate IFI27 via activating the NF-$\kappa$B pathway. We discovered that DDX60 overexpression could significantly increase the p-p65 protein level. A significant reduction in the p65 and p-p65 protein levels was also observed, while p65 expression levels were also inhibited in DDX60 overexpressed cells (Fig. 6E–H). In addition, reducing p65 protein levels could attenuate the promoting effect of DDX60 overexpression on IFI27 expression (Fig. 6E–I). The results indicated that IFI27 expression levels in HNSCC cells may be upregulated by DDX60 via the NF-$\kappa$B pathway (Fig. 6J).