ASCT2 (SLC1A5) is an EGFR-associated protein that can be co-targeted by cetuximab to sensitize cancer cells to ROS-induced apoptosis

Abstract

Therapeutic targeting of ASCT2, a glutamine transporter that plays a major role in glutamine uptake in cancer cells, is challenging because ASCT2 also has a biological role in normal tissues. In this study, we report our novel finding that ASCT2 is physically associated in a molecular complex with epidermal growth factor receptor (EGFR), which is often overexpressed in human head and neck squamous cell carcinoma (HNSCC). Furthermore, we found that ASCT2 can be co-targeted by cetuximab, an EGFR antibody approved for treating metastatic HNSCC. We demonstrated that cetuximab downregulated ASCT2 in an EGFR expression-dependent manner via cetuximab-mediated EGFR endocytosis. Downregulation of ASCT2 by cetuximab led to decreased intracellular uptake of glutamine and subsequently a decreased glutathione level. Cetuximab thereby sensitized HNSCC cells to reactive oxygen species (ROS)-induced apoptosis and, importantly, it is independent of effective inhibition of EGFR downstream signaling by cetuximab. In contrast, knockdown of EGFR by siRNA or inhibition of EGFR kinase with gefitinib, an EGFR kinase inhibitor, failed to sensitize HNSCC cells to ROS-induced apoptosis. Our findings support a novel therapeutic strategy for EGFR-overexpressing and cetuximab-resistant cancers by combining cetuximab with an oxidative therapy.

Keywords: ASCT2; Cetuximab; EGFR; Glutathione; HNSCC; ROS.

Figure 1

ASCT2 is physically associated with EGFR. (A) HN5 cell lysates were subjected to EGFR immunoprecipitation (IP) with cetuximab or a control antibody, or subjected to ASCT2 immunoprecipitation with an anti-ASCT2 rabbit monoclonal antibody (D7C12, Cell Signaling) or a control rabbit anti-mouse IgG antibody, followed by Western blotting (WB) of the immunoprecipitates with an anti-ASCT2 rabbit polyclonal antibody (H-52, Santa Cruz Biotechnology) and with an anti-EGFR mouse monoclonal antibody (F4, Sigma-Aldrich), respectively. HC, heavy chain. (B) ASCT2 co-immunoprecipitation by cetuximab was validated by Western blotting using different ASCT2 (D7C12) and EGFR (D38B1, Cell Signaling) antibodies as indicated. (C) HN5 cells were untransfected or transfected with control siRNA, ASCT2 siRNA, or EGFR siRNA for 72 h before being plated on coverslips for overnight. The cells were then incubated with blocking buffer only or with an ASCT2 antibody (H-52) and an EGFR antibody (F4), alone and together, as indicated, and then subjected to Duolink proximity ligation assay as described in Materials and methods. Scale bars, 25 μm. (D) Lysates from the indicated HNSCC cell lines were subjected to EGFR immunoprecipitation with cetuximab or ASCT2 immunoprecipitation with an ASCT2 antibody (H-52), along with a control IgG immunoprecipitation. The immunoprecipitates were then analyzed by Western blotting with the indicated antibodies.

Figure 2

ROS stimulation leads to a higher level of ASCT2-EGFR association. HN5 cells were untreated or treated with 1 mM H₂O₂ or 10 mM DCA as indicated for overnight (16 h). Cell lysates were subjected to immunoprecipitation (IP) for EGFR by cetuximab or control IgG, followed by Western blotting with antibodies against ASCT2 (H-52) and EGFR (F4), respectively. HC, heavy chain.

Figure 3

Cetuximab downregulates ASCT2 via cetuximab-mediated EGFR endocytosis, leading to inhibition of glutamine uptake and decreased glutathione synthesis. (A) HN5 cells were treated with or without 20 nM cetuximab for 10 h. Fractions of cell membrane (indicated by the presence of E-cadherin, a membrane marker) and cytoplasm (indicated by the presence of β-actin, a cytoplasmic marker) were analyzed for Western blotting with the indicated antibodies. (B) HN5 cells were subjected to knockdown of Rab5 for 72 h. Cell lysates were subjected to Western blotting with the indicated antibodies. (C–E) HN5 cells were untreated or treated with 20 nM cetuximab for the indicated times. Cell lysates were subjected to Western blotting with the indicated antibodies (C). ³H-labeled glutamine uptake was determined as described in Materials and methods (D), and glutathione level was measured using the Cayman glutathione assay kit (E). (F) HN5 cells were subjected to knockdown of Rab5 for 72 h as in (B). Glutathione level was measured using the Cayman glutathione assay kit as in (D). All error bars, S.D.

Figure 4

Cetuximab sensitizes cancer cells to ROS-elevating agent-induced apoptosis in an EGFR-expression-dependent manner. (A) HN5 cells were subjected to knockdown with each of two different EGFR siRNAs or control siRNA for 72 h. During the last 24 h of siRNA transfection, the cells were either untreated or treated with 10 mM DCA, 20 nM cetuximab, or both. Cell lysates were subjected to Western blotting with the indicated antibodies. (B) HN5 and FaDu cells were subjected to knockdown of Rab5 or Rab11 with specific siRNAs or control siRNA for 72 h. During the last 24 h of siRNA transfection, the cells were either untreated or treated with 10 mM DCA, 20 nM cetuximab, or both. Cell lysates were subjected to Western blotting with the indicated antibodies.

Figure 5

Combination of cetuximab with DCA, but not combination of gefitinib with DCA, induces ROS and apoptosis in HN5 cells. (A) and (B) HN5 cells were subjected to knockdown of PDK1 followed by treatment with 0.2 μM gefitinib or 20 nM cetuximab for 24 h. (C) and (D) HN5 cells were co-treated with 10 mM DCA plus 0.2 μM gefitinib or 20 nM cetuximab for 24 h. (E) HN5 cells were left untreated or treated with 20 nM cetuximab or 0.2 μM gefitinib, each alone and in combination with 10 mM DCA as indicated. The cells were stained with Enzo’s ROS detection kit and observed under a microscope. Scale bars, 50 μm.

Figure 6

Cetuximab plus DCA induces apoptosis in cetuximab-resistant HNSCC cells. (A) and (B) HN5 cells and HN5 cells transfected with a constitutively active Ras (H-RasG12V) were subjected to treatment with PDK1 siRNA or control siRNA with or without 20 nM cetuximab as indicated for 24 h. Lysates of HN5 cells treated as indicated were subjected to Western blotting with the indicated antibodies (A) and to ELISA for quantification of apoptosis (B).