Suprabasin is hypomethylated and associated with metastasis in salivary adenoid cystic carcinoma

Abstract

Background: Salivary gland adenoid cystic carcinoma (ACC) is a rare cancer, accounting for only 1% of all head and neck malignancies. ACC is well known for perineural invasion and distant metastasis, but its underlying molecular mechanisms of carcinogenesis are still unclear.

Principal findings: Here, we show that a novel oncogenic candidate, suprabasin (SBSN), plays important roles in maintaining the anchorage-independent and anchorage-dependent cell proliferation in ACC by using SBSN shRNA stably transfected ACC cell line clones. SBSN is also important in maintaining the invasive/metastatic capability in ACC by Matrigel invasion assay. More interestingly, SBSN transcription is significantly upregulated by DNA demethylation induced by 5-aza-2'-deoxycytidine plus trichostatin A treatment and the DNA methylation levels of the SBSN CpG island located in the second intron were validated to be significantly hypomethylated in primary ACC samples versus normal salivary gland tissues.

Conclusions/significance: Taken together, these results support SBSN as novel oncogene candidate in ACC, and the methylation changes could be a promising biomarker for ACC.

Figure 1. SBSN is hypomethylated in primary ACC samples, and its expression is induced by CpG island demethylation in SACC83.

A. We first confirmed 5-aza-dC/TSA-induced expression of SBSN in SACC83 mRNA levels by qRT-PCR. B. Graph of actual data using TaqMan qRT-PCR analysis. X-axis, amplification cycle numbers; y-axis, ΔRn values used to plot signal attributable to the 5′ nuclease reaction, which reflects the quantity of amplicon. SBSN was amplified between 25–30 cycles in 5-aza-dC/TSA treatment, whereas it was amplified between 30–35 cycles in mock treatment. 5-aza-dC/TSA- or mock-treated samples were loaded in the same amount as indicated by GAPDH. C. qMSP was conducted in a paraffin-embedded ACC cohort, which consisted of 62 ACC samples and 25 normal salivary gland tissues. Significant hypomethylation in SBSN was shown in ACC versus normal salivary gland tissue (p<0.0001, Student’s t-test). SBSN methylation scores were normalized by β-actin. Error bars indicate the standard deviation.

Figure 2. SBSN is important in maintaining anchorage-independent and anchorage-dependent growth in SACC83.

Scramble, shRNA 1, shRNA 2, and shRNA 3 indicate the control and three types of SBSN shRNAs used to establish stable clones in SACC83. A, Representative photographs of anchorage-independent growth by soft agar assay. The size of colonies indicates the focus formation ability of each type of stable clone. B, The number of colonies counted in each type of stable clone. C, Anchorage-dependent cell proliferation assay. CCK-8 absorbance indicates the amount of cells at time-points 0, 24, and 48 hours. Scramble clones grew faster than SBSN shRNA stable clones at 24 and 48 hours, p<0.01. Statistical comparisons were performed with Student’s t-test. Error bars indicate the standard deviation of triplicate assays. D. SBSN mRNA levels in different stable clones were determined by qRT-PCR. This confirmed that SBSN was silenced by ∼50% in shRNAs clones compared to scramble clones.

Figure 3. SBSN is important in maintaining invasion and metastatic capability in SACC83.

Matrigel invasion assay was performed with scramble control and three types of SBSN shRNA stable clones made from SACC83, as indicated by scramble, shRNA 1, shRNA 2, and shRNA 3. Representative photos of whole transwell membranes are pictured at 4× magnification; the inset pictures were taken at 20× magnification at randomly selected central locations. This metastasis analysis was performed in triplicate.