High SOD1 expression is associated with poor prognosis in nasopharyngeal carcinoma
SOD1 is best known for its role in redox homeostasis and is often dysregulated during cancer development. We found that SOD1 mRNA was significantly increased in head and neck cancer tissues from the Oncomine microarray database (Fig. 1a, https://www.oncomine.org). Additionally, the SOD1 mRNA level was significantly increased in NPC tissues from our institute (SYSUCC) (Fig. 1b). The SOD1 protein level was notably increased in 17 representative tumours compared with normal tissues (Fig. 1c). We also analysed SOD1 expression in 100 human NPC tissue biopsies and 10 normal nasopharyngeal epithelial tissue biopsies. Representative IHC staining confirmed that, compared with that in normal tissues, SOD1 expression in human paraffin-embedded NPC tissues substantially increased (Fig. 1d). The expression profiles of SOD1 were also analysed in NPC cell lines, and compared with those in non-tumourigenic NP69 cells, both the mRNA and protein levels of SOD1 in NPC cells increased (Fig. 1e and f). As expected, compared with that in NP69 cells, the enzymatic activity of SOD1 in both 5-8F cells and CNE2 cells also increased (Fig. 1g). Together, these results suggest that SOD1 expression was significantly increased in human NPC cells and NPC tissues and that it may act as a key regulator of the antioxidant defence system in NPC.
To determine the clinical importance of this finding, we next evaluated the expression of SOD1 in 100 archived NPC tissues (Fig. 1h). Patients with low SOD1 expression had more favourable clinical outcomes, while patients with high SOD1 expression had shorter survival times (Fig. 1i). On the basis of the above observations, we concluded that low SOD1 protein expression predicts a favourable clinical outcome and better survival.
SOD1 knockdown disrupts NPC cell growth in vitro
Since SOD1 is overexpressed in NPC cancer cells, to examine the effect of SOD1 on NPC cells, SOD1 was knocked down using siRNA in 5-8F and CNE2 cells (Fig. 2a). Compared with the control cells, cells with SOD1 knocked down presented with reduced viability and clonogenicity (Fig. 2b-d). The overexpression of SOD1 (Fig. 2e) increased clonogenicity (Fig. 2f) and conferred resistance to apoptosis induced by the ROS inducer phenazine methosulfate (PMS) (Fig. 2h). Conversely, the knockdown of SOD1 increased the number of apoptotic cells (Fig. 2g).
Knocking down SOD1 expression reduces NPC cell growth in vivo
Next, we further investigated whether SOD1 knockdown affects NPC cell growth in cell-based xenografts. Using a recombinant adeno associated virus, we first confirmed SOD1 knockdownin 5-8F and CNE2 cells (Fig. 3a). To evaluate the consequences of SOD1 loss in vivo, we used a nude mouse tumour model and administered 5-8F-shCon or 5-8F-shSOD1 cells and CNE2-shCon or CNE2-shSOD1 cells via subcutaneous injections into the flank. Compared with the vector control-derived tumours, the tumours from 5-8F or CNE2 cells with SOD1 knockdown presented with significant growth inhibition (Fig. 3b-e) and an increased number of apoptotic cells (Fig. 3f). These results suggest that disrupting SOD1 reduces NPC cell growth in vivo.
LCS-1 suppresses NPC cell growth in vitro and in vivo
To further characterize the contribution of SOD1 to NPC cell growth, we treated 5-8F and CNE2 cells with the specific SOD1 inhibitor LCS-1. As shown in Fig. 4a-c, LCS-1 dose- and time-dependently inhibited the proliferation of 5-8F and CNE2 cells. Moreover, compared with the control-treated cells, 5-8F and CNE2 cells treated with LCS-1 had decreased clonogenicity and increased apoptosis (Fig. 4d and e). To further investigate whether LCS-1 affects NPC growth in vivo, CNE2 cells (1 × 106) were injected subcutaneously into the mouse flank. Ten days later, LCS-1 (0.76 mg/kg) was administered every 4 days for 4 weeks. Compared with the control group, the LCS-1-treated group presented a greater reduction in tumour weight and growth (Fig. 4f) but a higher percentage of TUNEL-positive cells (Fig. 4g).
The above data showed that inhibiting SOD1 by LCS-1 treatment drastically reduced NPC cell viability in vitro.
SOD1 suppression increases ROS and induces lipid accumulation
Because SOD1 acts as a regulator of anti-oxidative capacity, we posited that SOD1 knockdown may increase superoxide ion and induce cellular apoptosis. As expected, compared with the control cells, cells with SOD1 knocked down (Fig. 5a and b) or subjected to LCS-1 treatment (Fig. 5c-f) exhibited a substantial increase in superoxide levels, and LCS-1-inhibited clonogenicity was rescued by the administration of TEMPO, a superoxide dismutase mimetic (Fig. 5g). Furthermore, eliminating superoxide anion suppression by knocking down SOD1 or treating with LCS-1 caused a marked increase in apoptosis, as detected by the TUNEL assay (Fig. 5h).
These findings demonstrate that SOD1 inhibition increases cellular superoxide and induces apoptosis, which indicated that SOD1 is required for NPC cell detoxification.
Recent studies have found that SOD1 not only functions as an antioxidant enzyme but also plays a key role in cellular metabolism [12,13,14]. The enhanced requirement of FAO for ATP and NADPH is frequently occurs in proliferating cells [15], block lipid oxidation via CPT1 as a therapeutic target for prostate cancer and myeloma [16, 17]. Inhibiting FA transport into the mitochondria is also able to effectively kill chronic lymphocytic leukaemia cells [18] and breast cancer [19]. Reducing fatty acid transport into the mitochondria forced fatty acids to lipid droplets for storage in Clear cell renal cell carcinoma [20]. To determine if SOD1 affects NPC cell mitochondrial metabolism, some of the key enzymes involved in mitochondrial pyruvic acid and FAO metabolism were examined. After 24 h, compared with the control cells, the LCS-1-treated 5-8F cells exhibited lower PDHE1A and CPT1A mRNA levels (Fig. 6a). Though SOD1 inhibition reduced PDHE1A mRNA expression, we observed minimal difference in the proliferation of siPDHE1A cells relative to control cells (Additional file 1: Figure S1).
Then, we further tested the effects of SOD1 on lipid metabolic enzymes expression. Compared with that in the control cells, the CPT1A protein level in the cells treated with LCS-1 or transfected with SOD1 siRNA decreased (Fig. 6b-c). In addition, the overexpression of SOD1 increased CPT1A protein levels (Fig. 6d). In addition, we found that compared with control group in NPC cell, expression of AGPAT1 and DGAT1 at the protein level was upgulated or downregulated in SOD1 overexpression group or SOD1 knockdown group, respectively. The protein level of ATGL was fund remarkably downregulated in SOD1 knockdown group compared with control group in NPC cell (Fig. 6e). As CPT1A is rate-limiting enzyme of the long chain fatty acids transport system controlling entry into the mitochondrion. Recently, it was reported that repress CPT1A enhance fatty acids to lipid droplets for storage [20, 21]. We therefore asked whether SOD1 or CPT1A inhibition induce lipid accumulation. As shown in Fig. 6f and g, inhibiting SOD1 or CPT1A induces lipid accumulation. Further study showed that disruptin CPT1A downregulated DGAT1 and up-regulated ATGL (Fig. 7a), while over-express CPT1A in 5-8F and CNE2 cells reduced protein levels of DGAT1 and up-regulated ATGL protein levels (Fig. 7b-c). CPT1A restoration or TEMPO led to significant decreases lipid accumulation in NPC cells treated with LCS-1(Fig. 7d-e). In addition, etomoxir, a specific CPT1A inhibitor, dose-and time-dependently inhibited the proliferation of 5-8F and CNE2 cells (Fig. 7f-h). To further explore the effect of SOD1 and CPT1A on NPC cells migration and epithelial–mesenchymal transition (EMT). Disrupt SOD1 by LCS-1 or siRNA weaken migration ability of 5-8F and CNE2 cells (Additional file 1: Figure S2) and the expression of twist (mRNA) and vimentin (mRNA and protein) was significantly decreased, coupled with remarkable upregulation of epithelial marker E-cadherin (mRNA and protein) (Additional file 1: Figure S3). However, cell migration was not affected by CPT1A suppression (Additional file 1: Figure S4A-C). Consistently, western blot analysis of the expression of key EMT markers, including E-cadherin and vimentin, showed no significant changes in CPT1A inhibition cells compared with control cells (Additional file 1: Figure S4D-F). These findings suggest that SOD1 promotes the migration of NPC cells independence CPT1A.
In summary, these results suggested that disrupting SOD1 led to the over production of superoxide ion and induced lipid accumulation in NPC cells.