Map1lc3b and Sqstm1 Modulated Autophagy for Tumorigenesis and Prognosis in Certain Subsites of Oral Squamous Cell Carcinoma

Abstract

Oral squamous cell carcinoma (OSCC) is one of the most common cancer types worldwide and can be divided into three major subsites: buccal mucosal SCC (BMSCC), tongue SCC (TSCC), and lip SCC (LSCC). The autophagy marker microtubule-associated protein light chain 3B (MAP1LC3B) and adaptor sequestosome 1(SQSTM1) are widely used proteins to evaluate autophagy in tumor tissues. However, the role of MAP1LC3B and SQSTM1 in OSCC is not fully understood, particularly in certain subsites. With a tissue microarray comprised of 498 OSCC patients, including 181 BMSCC, 244 TSCC, and 73 LSCC patients, we found that the expression levels of MAP1LC3B and cytoplasmic SQSTM1 were elevated in the tumor tissues of three subsites compared with those in adjacent normal tissues. MAP1LC3B was associated with a poor prognosis only in TSCC. SQSTM1 was associated with poor differentiation in three subsites, while the association with lymph node invasion was only observed in BMSCC. Interestingly, MAP1LC3B was positively correlated with SQSTM1 in the tumor tissues of BMSCC, whereas it showed no correlation with SQSTM1 in adjacent normal tissue. The coexpression of higher MAP1LC3B and SQSTM1 demonstrated a significantly worse disease-specific survival (DSS) and disease-free survival (DFS) in patients with BMSCC and LSCC, but not TSCC. The knockdown of MAP1LC3B and SQSTM1 reduced autophagy, cell proliferation, invasion and tumorspheres of BMSCC cells. Additionally, silencing both MAP1LC3B and SQSTM1 enhanced the cytotoxic effects of paclitaxel in the tumorspheres of BMSCC cells. Taken together, MAP1LC3B and SQSTM1 might modulate autophagy to facilitate tumorigenesis and chemoresistance in OSCC, particularly in BMSCC.

Keywords: MAP1LC3B; SQSTM1; autophagy; oral cancer; prognosis; subsites; tumorigenesis.

Figure 1

Protein levels of MAP1LC3B puncta and SQSTM1 in OSCC. The MAP1LC3B puncta and cytoplasmic SQSTM1 were stained by immunohistochemistry and categorized into four different levels as follows: 0 = negative staining; 1 = weak; 2 = moderate; 3 = strong.

Figure 2

DSS survival curves for MAP1LC3B and SQSTM1 expression in patients with different subsites of OSCC. DSS survival curves of MAP1LC3 (A–D) and SQSTM1 (E–H) are shown for OSCC (A,E) and three main subsites, BMSCC (B,F), TSCC (C,G) and LSCC (D,H). The cutoff values for high or low expression of MAP1LC3B and SQSTM1 in tumor tissues were based on the receiver operating characteristic (ROC) curve.

Figure 3

DSS survival curves for the coexpression of MAP1LC3B and SQSTM1 in patients with different subsites of OSCC. DSS survival curves for the coexpression of MAP1LC3 and SQSTM1 are shown for OSCC (A) and three main subsites, BMSCC (B), TSCC (C) and LSCC (D). The cutoff values for high or low coexpression of MAP1LC3B and SQSTM1 in tumor tissues were based on the receiver operating characteristic (ROC) curve.

Figure 4

DFS survival curves for MAP1LC3B and SQSTM1 expression in patients with different subsites of OSCC. DFS survival curves of MAP1LC3 (A–D) and SQSTM1 (E–H) are shown for OSCC (A,E) and three main subsites, BMSCC (B,F), TSCC (C,G) and LSCC (D,H). The cutoff values for high or low expression of MAP1LC3B and SQSTM1 on tumor tissues were based on the receiver operating characteristic (ROC) curve.

Figure 5

DFS survival curves for the coexpression of MAP1LC3B and SQSTM1 in patients with different subsites of OSCC. DFS survival curves for the coexpression of MAP1LC3 and SQSTM1 are shown for OSCC (A) and three main subsites, BMSCC (B), TSCC (C) and LSCC (D). The cutoff values for high or low coexpression of MAP1LC3B and SQSTM1 in tumor tissues were based on the receiver operating characteristic (ROC) curve

Figure 6

Effects of silencing MAP1LC3B and SQSTM1 on autophagy and cell proliferation in BMSCC cells. (A) The BMSCC cell lines OC3 and (B) TW2.6 were transfected with 5 nM scrambled siRNA or siRNA against MAP1LC3B or SQSTM1 for 48 h and then were treated with ConA for 4 h. The knockdown efficiency was determined by immunoblotting. (C) Silencing effects on MAP1LC3B flux and (D) SQSTM1 levels were quantified and analyzed.

Figure 7

Effects of silencing MAP1LC3B and SQSTM1 on metastatic characteristics, tumorsphere formation and paclitaxel sensitivity in BMSCC cells. (A) The BMSCC cell lines OC3 and TW2.6 were transfected with 5 nM scrambled siRNA or siRNA against MAP1LC3B or SQSTM1 for 72 h to measure cell proliferation with Cell titer Glo. The cells treated with the autophagy inhibitor ConA for 24 h were used as a control. (B) OC3-I5 cells, the highly invasive strain of OC3, were silenced for 48 h and were seeded in Matrigel-coated Transwell filters to assess the cell invasion of BMSCC cells. (C) The silenced OC3 and (D) TW2.6 cells were cultured in nanoplates to examine tumorsphere formation. The tumorspheres were also treated with or without paclitaxel (PTX, 50 or 100 nM) to determine the effects of genes on drug resistance. The results represented three independent experiments.