KRT4 suppresses oral squamous cell carcinoma development by reducing ATG4B-mediated autophagy

Head and neck squamous cell carcinoma is the sixth most common tumor worldwide, and half of head and neck squamous cell carcinoma patients are with oral squamous cell carcinoma (OSCC). 300,000 new cases of OSCC were reported annually. Even with multi-modality treatment, the prognosis of OSCC remains unsatisfactory. Thus, it is urgent to discover novel therapeutic targets for OSCC. Some microarray studies have revealed that Keratin 4 (KRT4) is downregulated in OSCC, whereas its role in OSCC development remains unknown. The present study revealed that KRT4 suppressed OSCC progression by inducing cell apoptosis and inhibiting cell invasion. In addition, KRT4 over-expression inhibited autophagy by blocking the interaction of autophagy-related 4B cysteine peptidase (ATG4B) and microtubule-associated protein 1A/1B light chain 3 (LC3) to regulate apoptosis and invasion of OSCC. In conclusion, KRT4 played an important role in OSCC development through regulating ATG4B-mediated autophagy and may be a novel therapeutic drug target of OSCC.


Introduction
Head and neck squamous cell carcinoma is the sixth most prevalent tumor worldwide, and half of those occur in the oral cavity (Haddad and Shin, 2008;Kademani, 2007). OSCC accounts for more than 90% of all oral malignancies and 2-3% of all cancers, with more than 300,000 new cases reported annually (Choi and Myers, 2008;Ferlay et al., 2010). In China, over 14,000 deaths due to OSCC are reported each year (Zheng et al., 2015). With the development of cytobiology and molecular biology, therapies of OSCC have been significantly improved. However, more than 50% of OSCC patients still die under current therapies (Chiou et al., 2008;Lo et al., 2003), as the development of OSCC is a multi-step process, always going through stages of normal mucosa, precancerous lesions, and cancer (Argiris et al., 2008;Hashibe et al., 2009). Thus, it is urgent to elucidate novel mechanisms regulating OSCC development and search for promising biomarkers for targeted therapies of OSCC.
Technological breakthroughs, including laser capture microdissection and microarray technology, have fundamentally altered cancer research, which allows the highthroughput analysis of thousands of genes. Several studies have revealed that KRT4 was downregulated in OSCC (Lallemant et al., 2009;Toruner et al., 2004;Ye et al., 2008). KRT4 belongs to the Type II keratin, which is specifically expressed with family member KRT13 in differentiated layers of the mucosal and esophageal epithelia (Zhang et al., 2018a). Mutations of these genes would lead to White Sponge Nevus (WSN) (Zhang et al., 2018a). A previous study has revealed that KRT4 serves as a tumor marker in esophageal squamous cell carcinoma (Takikita et al., 2011). However, the role of KRT4 in OSCC remains unknown.
Through affinity capture-MS detection, KRT4 is found to interact with ATG4B (Behrends et al., 2010). ATG4B is required for autophagy and serves as a cysteine protease for hydrolyzing LC3I to LC3II or for the deconjugation of LC3II (Nakatogawa et al., 2007;Pengo et al., 2017). Autophagy mediates bulk degradation involved in the constitutive clearance of intracellular components, including damaged cellular organelles and protein aggregations, and plays an important role in maintaining cellular homeostasis (Mizushima et al., 2008;Weidberg et al., 2011). Dysfunction of autophagy leads to many diseases, such as cancer (Levine and Kroemer, 2019). In different contexts, autophagy plays neutral, suppressive, or promoting roles in cancer. For example, autophagy could maintain tumor growth in a nutrient-limited micro-environment (Amaravadi et al., 2016). Numerous studies have indicated that elevated expression of ATG4B enhances tumor growth, migration, and invasion by inducing autophagy in many types of cancer (Akin et al., 2014;Levine, 2007;Tran et al., 2013). To date, the role of ATG4B in OSCC is little known.
The aim of this study was to investigate the role of KRT4 in OSCC development. We found that KRT4 was downregulated in OSCC cells. Overexpression of KRT4 significantly induced OSCC cell apoptosis and suppressed cell invasion of OSCC cells. Moreover, KRT4 regulated apoptosis and invasion in OSCC cells by suppressing autophagy via blocking the interaction of ATG4B and LC3. These results revealed that KRT4 suppressed OSCC development through regulating autophagy.

Materials and Methods
Cell culture Normal oral keratinocytes (NOK) and OSCC cells (HN6) obtained from Cell Bank at the Chinese Academy of Sciences (Shanghai, China) were cultured with DMEM supplemented with 10% fetal bovine serum (FBS), 100 U/mL penicillin, and 100 µg/mL streptomycin in a humidified atmosphere of 5% CO 2 at 37°C.

Cell transfection
Cells were transfected with the KRT4 expression vector, ATG4B expression vector, or ATG4B siRNAs using Lipofectamine 2000 (Invitrogen, Carlsbad, USA), according to the manufacturer's instruction. After 48 h, cells were collected for further experiments.
Western blotting Equal amounts of protein from different groups were resolved by 10% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and subsequently transferred onto Immobilon-P membranes (Millipore, Shanghai, China). The membranes were blocked with 5% nonfat milk at room temperature (RT) for 1 h and subsequently incubated with the following primary antibodies in 1% nonfat milk at 4°C overnight: Rabbit anti-KRT4 (1:500), mouse anti-ATG4B (1:400), rabbit anti-LC3 (1:500). All primary antibodies were obtained from Abcam (Boston, USA). Following primary antibody incubation, the membranes were washed for 45 min with Tris-buffered saline containing 0.1% Tween20 (TBS-T) and then exposed to goat anti-rabbit or mouse secondary antibody. The signals of proteins were detected by the SuperSignal West Femto Maximum Sensitivity Substrate kit (Thermo Scientific, Rockford, USA).
Cell Counting Kit-8 (CCK8) assay The cell viability was detected using CCK8 assay following the manufacturer's protocol. Briefly, cells were seeded in a 96-well culture plate (5 × 10 3 cells/well) for 24 h. Next, 10 μL of CCK-8 reagent (Beyotime, Shanghai, China) was added into each well. Subsequently, cells were incubated at 37°C for 2 h, followed by measuring the absorbance (optical density, OD) at 450 nm on an enzyme immunoassay analyzer (Bio-Rad).
Scratch wound healing assay Cell migration was tested by the scratch wound-healing assay. 5 × 10 5 cells were grown in 35 mm plates and followed by transfection of expression vectors for 2 h. Then a scratch wound was created in two perpendicular directions by scraping the cell monolayer using a pipette tip, and the culture medium was replenished immediately. After 48 h, phase-contrast images of the cells were captured. Then wound width was calculated as the average distance between the sides of the scratches measured by ImageJ (https:// imagej.nih.gov/ij/). Subsequently, the migration rate was identified through decreased wound width dividing the duration of migration.

Transwell invasion assay
Cells were suspended in 200 μL DMEM medium without FBS and seeded in the upper compartment Transwell inserts (Millipore) coated with Matrigel (Corning, Shanghai, China). Meanwhile, the lower chamber was filled with DMEM medium supplemented with 10% FBS. After 24 h incubation at 37°C, noninvasive cells were cleaned by cotton swabs, and the invasive cells in the lower chamber were fixed with 4% paraformaldehyde followed by the staining of 0.5% crystal violet for 10 min at RT. Next, invasive stained cells were photographed and counted by a light microscope (Leica, Heidelberg, Germany).

Immunostaining
Cells were fixed in 4% paraformaldehyde for 20 min at RT, followed by PBS washing three times. Then cells were blocked by 10% donkey serum albumin in PBST (0.1% Triton X-100 in PBS) for 1 h at RT. The primary antibody LC3 antibody (1:200) was diluted in PBS supplementary with 1% donkey serum and incubated with cells at 4°C overnight. Next, cells were washed by PBST five times and incubated with secondary antibody in the dark at RT. Finally, cells were counterstained with DAPI and mounted with an aqueous mounting medium (Sigma-Aldrich). Images were captured by an inverted microscope (Leica).

Statistical analysis
All data in this study were representative of three repeated experiments. The quantitative data were present as means ± standard deviation (SD). Statistical differences were analyzed by the unpaired Student's t-test between two groups or one-way ANOVA followed by Student-Newman-Keuls post hoc test between multiple groups. Moreover, all statistical analyses were carried out via the SPSS 20.0 software (SPSS Inc., Chicago, USA). The value of p < 0.05 indicated statistical significance.

KRT4 overexpression induces apoptosis in OSCC cells
Compared to NOK cells, the KRT4 mRNA and protein levels were significantly downregulated in HN6 cells (Figs. 1A and 1B). Moreover, overexpression of KRT4 did not regulate HN6 cell proliferation, whereas suppressed NOK cell viability, detected by CCK8 assay (Fig. 1C). Furthermore, cell apoptosis was performed by flow cytometric analysis to count the rate of early and late apoptotic cells. Results showed that KRT4 overexpression increased the rate of early and late apoptotic HN6 cells from 9.6% to 18.8%, whereas KRT4 had no effect on NOK cell apoptosis (Fig. 1D). These results suggested that KRT4 enhanced apoptosis in OSCC cells.

KRT4 overexpression suppresses invasion of OSCC cells
To identifying the role of KRT4 in migration and invasion of OSCC cells, scratch wound healing assay and Transwell assay were performed. Results indicated that overexpression of KRT4 did not affect HN6 cell migration ( Fig. 2A). Furthermore, KRT4 overexpression reduced the invasive HN6 cell number from 157 to 93 (Fig. 2B). The above data suggested that KRT4 inhibited the invasion of OSCC cells.

KRT4 protein interacts with ATG4B protein in OSCC cells
The above results had demonstrated the cancer-suppressive role of KRT4 in OSCC. Next, the mechanism of how KRT4 suppressed OSCC development was explored. A previous study has demonstrated that KRT4 protein associates with ATG4B protein in 293T cells (Behrends et al., 2010). Thus, we investigated whether KRT4 protein interacted with ATG4B protein in HN6 cells by Co-IP using a KRT4 antibody. Results revealed that KRT4 protein bound the ATG4B protein in HN6 cells (Fig. 3A). In addition, the expression of ATG4B detected by qPCR and WB was not changed in HN6 cells compared to that in NOK cells (Figs. 3B and 3C). These results together suggested that KRT4 may exert a tumor-suppressive effect through ATG4B.

Silence of ATG4B triggers apoptosis and suppresses invasion in OSCC cells
The role of ATG4B in OSCC remains unclear so far. To identify the effect of ATG4B on OSCC cells, siRNAs were transfected into HN6 cells to knockdown ATG4B. Among the three siRNAs, ATG4B siRNA1 was found to have the most powerful effect (Fig. 4A), which was used for subsequent experiments. Further results revealed that ATG4B silence increased the rate of early and late apoptotic HN6 cells from 4.58% to 13.95%, suggesting that ATG4B induced HN6 cell apoptosis (Fig. 4B). Moreover, ATG4B silencing slightly inhibited HN6 cell migration (Fig. 4C). Besides, the silencing of ATG4B reduced the number of invasive HN6 cells from 203 to 96, significantly suppressed HN6 cell invasion (Fig. 4D). These results suggested that ATG4B silence induced apoptosis and inhibited invasion in OSCC cells, which were consistent with those in KRT4overexpressed HN6 cells.
Furthermore, ATG4B overexpression reversed the effect of KRT4 overexpression on HN6 cell apoptosis (Fig. 5A). Transwell invasion assay also revealed that ATG4B overexpression abolished the effect of KRT4 overexpression on HN6 cell invasion (Fig. 5B). All these data indicated that KRT4 regulated OSCC development by inhibiting the function of ATG4B.

KRT4 overexpression reduces autophagy in OSCC cells
As ATG4B is required for autophagy, we detected the effect of KRT4 on autophagy of HN6 cells. The number of autophagosomes that responded to autophagy levels was detected through LC3 immunostaining kin HN6 cells. Results showed that KRT4 overexpression dramatically decreased the number of LC3 green puncta indicated autophagosome in HN6 cells (Fig. 6A). Moreover, the level of LC3 II required for autophagosome formation was also reduced in KRT4-overexpressed HN6 cells (Fig. 6B). Consistently, ATG4B silence also decreased the number of autophagosomes and reduced LC3II level (Figs. 6C and 6D). All these results suggested that KRT4 suppressed autophagy through ATG4B in OSCC cells. KRT4 suppresses autophagy through inhibiting the interaction of ATG4B and LC3 ATG4B facilitates autophagy through binding with LC3 to hydrolyze LC3I to LC3II or deconjugate LC3II. To further demonstrate the mechanism of how KRT4 suppressed autophagy in HN6 cells, the interaction of ATG4B and LC3 (including LC3I and LC3II) was detected by Co-IP. Compared to the control group (Group 2), KRT4 overexpression (Group 1) reduced the interaction between ATG4B and LC3 (Fig. 7), suggesting that KRT4 prohibited autophagy by disrupting the interaction between ATG4B and LC3.

Discussion
KRT4 is known as the inactivator of WSN characterized by benign oral, esophageal, and anal leukoplakia (Chao et al., 2003;Kimura et al., 2013;Zhang et al., 2018a;Zhang et al., 2009). However, the role of KRT4 in OSCC is little known, though its transcription is significantly reduced in OSCC. In the present study, we demonstrated that KRT4 played a cancer-suppressive role in OSCC through inducing cell apoptosis and suppressing cell invasion. This study further revealed that KRT4 suppressed OSCC development by inhibiting ATG4B-mediated autophagy via disrupting the interaction between ATG4B and LC3. This study revealed the cancer-suppressive role of KRT4 for the first time.
Usually, induction of apoptosis leads to the reduction of cell proliferation. However, our results indicated a dramatic induction of apoptosis in HN6 cells caused by KRT4 overexpression, while the proliferation of HN6 cells remained unaffected by KRT4 overexpression. Some previous studies have revealed a similar finding. For example, terazosin and doxazosin trigger prostate apoptosis without modifying cell proliferation in patients with benign prostatic hyperplasia (Chon et al., 1999). In addition, the apoptotic index of emphysematous lungs is significantly higher than that of the control group, whereas the proliferation index is consistent in emphysema patients and healthy people (Calabrese et al., 2005). Therefore, induction of apoptosis may not affect cell proliferation.
In different contexts, autophagy plays neutral, suppressive, or promoting roles in cancer (Amaravadi et al., 2016). Several studies have revealed that deletion of autophagy-related gene Atg5 or Atg7 in mice induces benign tumors, suggesting that autophagy suppresses cancer initiation (Akito et al., 2011;Rosenfeld et al., 2014). In addition, autophagy could promote the growth of advanced cancer, such as deletion of Atg7 reduces oncogenic Krasdriven lung tumor cell proliferation (Guo et al., 2013). Moreover, impaired autophagy caused by monoallelic loss of Becn1 suppresses Palb2-associated mammary tumorigenesis (Huo et al., 2013). Our results indicated that KRT4 suppressed OSCC development by inhibiting ATG4Bmediated autophagy. Thus, ATG4B-mediated autophagy played tumor-promoting roles in OSCC.
This study also revealed that invasion of OSCC was inhibited via the downregulation of autophagy. This finding is in accordance with previous studies. For example, inhibition of ROS-dependent NUPR1-mediated autophagy antagonizes repeated cadmium exposure-induced OSCC invasion (Fan et al., 2019). Moreover, PI3K inhibitors targeted disruption of PI3K/Akt/mTOR signaling pathway suppress autophagy and decrease the invasiveness of OSCC (Aggarwal et al., 2019). In addition, silencing phosphofructokinase-platelet (PFKP) prohibits starvationinduced autophagy and the subsequent epithelialmesenchymal transition (EMT) required to invasion in OSCC (Chen et al., 2018). Interestingly, PI3K/AKT signaling pathway contributes to both apoptosis and invasion, suggesting that PI3K/AKT may be one of the subsequent signaling pathways of KRT4 in OSCC tumor progression.
Further analysis demonstrated that KRT4 inhibited ATG4B-mediated autophagy via disrupting the interaction between ATG4B and LC3. ATG4B induces autophagy through interacting with LC3 in several ways. For example, ATG4B hydrolyses LC3 at the C-terminal region to expose a glycine residue to LC3II required for autophagosome formation (Shu et al., 2010;Li et al., 2011). ATG4B is also required for the deconjugation of membrane-bound MAP1LC3-II from non-autophagosomes or from autophagosomes to support the LC3II supplement (Yu et al., 2012). The interaction between ATG4B and LC3 is critical for facilitating autophagy. A previous study has indicated that phosphorylation of ATG4B at Ser383 and Ser392 decreases the proteolytic activity of ATG4B and leads to delipidating MAP1LC3-II (Yang et al., 2015). Thus, this finding suggested that KRT4 may competitively bind with ATG4B to disturb the interaction of ATG4B and LC3.
To date, therapies against OSCC are limited. Cancer stem cells (CSCs) contribute to the metastasis of OSCC through modulating cytokines and regulating cadherins (Najafi et al., 2019). Thus, CSCs may become a promising therapeutic target for OSCC. Autophagy is involved in modulating stemness, recurrence, and resistance to anticancer therapies  (Yun and Lee, 2018). However, the role of KRT4 in CSCs of OSCC remains unclear. According to the findings of this study, we guessed KRT4 may inhibit autophagy in CSCs of OSCC and subsequently suppress the metastasis of OSCC, which may be a potential therapeutic drug target for OSCC through regulating CSCs.

Conclusions
Summarily, this study revealed that KRT4 suppressed OSCC development through inhibiting ATG4B-mediated autophagy, which uncovered the cancer-suppressive role of KRT4 for the first time. Moreover, the finding of the present study may provide a potential therapeutic drug target for OSCC.
Availability of Data and Materials: The data of this study are available from the corresponding author upon reasonable request.
Author Contribution: The authors confirm contribution to the paper as follows: study conception and design: Xiaoxu Li. Author, Bin Cheng. Author; data collection: Xiaoxu Li. Author, Yun Wang. Author, Juan Fang. Author, Zhi Wang. Author; analysis and interpretation of results: Xiaoxu Li. Author, Yun Wang. Author, Juan Fang. Author, Zhi Wang. Author, Xiaoan TAO. Author, Juan XIA. Author; draft manuscript preparation: Xiaoxu LI. Author, Bin Cheng. Author. All authors reviewed the results and approved the final version of the manuscript.