ATF4 destabilizes RET through nonclassical GRP78 inhibition to enhance chemosensitivity to bortezomib in human osteosarcoma

Rationale: Activating transcription factor 4 (ATF4) is a central regulator of the cellular stress response and reduces tumor burden by controlling the expression of target genes implicated in the induction of apoptosis. Evidence shows ATF4 activation is responsible for proteasome inhibitor bortezomib (BTZ)-induced osteosarcoma (OS) cell death. However, it remains unclear how such suppressive function is impaired during prolonged therapeutic interventions. Methods: Stable cells and in vivo xenograft models were generated to reveal the essential role of ATF4 in cell apoptosis and tumor growth. Fluorescence in situ hybridization (FISH) and immunohistochemistry were employed to detect the expression and significance of ATF4 in the specimens from osteosarcoma patients. Biochemical differences between chemoresistant and chemosensitive cancer cells were determined by proliferation, apoptosis, real-time PCR, immunoblotting and immunofluorescence. Promoter activity was analysed using the luciferase reporter assay. Immunoprecipitation was used to explore the interaction of proteins with other proteins or DNAs. Results: ATF4 significantly inhibited OS tumorigenesis, whereas knockdown of ATF4 prevented the antitumor effects of BTZ. Normal osteoblasts are supposed to preferentially express ATF4, but ATF4 silencing was detected in both OS clinical samples and BTZ-resistant sublines (OS/BTZ). We found that ATF4 downregulation was tightly linked to the aberrant expression of RET, primarily due to RET stabilization in OS/BTZ cells. Loss of RET upregulated ATF4 and potentiated the apoptotic response to BTZ. ATF4 recognized the TK domain of RET by recruiting its transactivated E3 ligase Cbl-c to accelerate RET proteasomal turnover, which in turn prevented BTZ resistance. In contrast, the chaperone GRP78 bound to RET and interfered with ATF4/RET interactions, promoted RET stabilization. Intriguingly, ATF4 repressed GRP78 transcription in OS/BTZ cells via the first ERSE, instead of transactivating GRP78 in wild-type OS via classical CRE element, revealing a dual targeting of RET and GRP78 to overcome chemoresistance. Conclusion: The results uncover a crucial role for ATF4 in blocking the progression and resistance response in RET/GRP78-positive human osteosarcoma.


Figure S2. RET expression promotes osteosarcoma tumorigenesis and BTZ resistance. A,
Venn diagram indicating duplication among the three gene sets regarding cancer (hsa05200), protein processing in endoplasmic reticulum (hsa04141) and ubiquitin mediated proteolysis (hsa04120) in KEGG PATHWAY database. 782 nonredundant genes were obtained. A set of 19 genes were enriched for the three KEGG pathways mentioned above (Combination) and 92 genes (Reference) which were studied or interested by us and were confirmed the interaction in STRING database. B, Relative expression of the 19 genes were determined for all bone-related sarcoma samples (n = 311) in TCGA. C, The interactions of HSPA5, ATF4, CBLC and RET in STRING database. D, The BTZ-resistant subline U-2 OS/BTZ and HOS/BTZ was established. Cells were seeded into 96-well culture plates and incubated for 1-7 days, respectively. MTT assay was performed to measure the cell growth curve. E, Colony formation efficiencies were determined in BTZ-resistant models of U-2 OS and HOS cells treated with BTZ (100 nM). F, Immunofluorescence analysis of GRP78 protein (green) in OS/BTZ cells, in comparison to the parental cells. Nuclei were stained with Hoechst 33342 (blue). Images were taken using the highcontent imaging system. Scale bars, 10 μm. G, Synthetic RET-siRNAs effectively suppressed endogenous RET mRNA in U-2 OS cells. Cells were transfected with 100 nM siRNA targeting RET (#1, 2, 3) or control siRNA (siCon), or cotransfected with ATF4 expression plasmid for 24 h. qRT-PCR quantitative analyses were performed to evaluate the efficiency of RET knockdown.
H, RET colony formation-promoting activity. Control and RET knockdown (siRET#2 and #3) OS or OS/BTZ cells were cultured for 2 weeks with two-day BTZ (100 nM) treatment at the beginning.
Then cells were fixed and stained. I, MTT assay of U-2 OS cells transfected with siRET (#1, 2, 3) or control siRNA in the presence of RTK inhibitor Cabozantinib (CBZ, 5 μM). J, Colony formation efficiencies were determined in RET-overexpressing OS or OS/BTZ cells with BTZ (100 nM) treatment. Error bars represent mean ± SD from three independent experiments (*P < 0.05, **P < 0.01; n.s., non-significant).  days. Error bars represent mean ± SD from three independent experiments (*P < 0.05, **P < 0.01). Figure 4F. Stable U-2 OSATF4 cells transfected with RET or siRET#1 were challenged with either vehicle or BTZ (100 nM), stained with FITC-conjugated Annexin V and PI, and analysed by flow cytometry to evaluate the role of RET in the prevention of apoptosis induction involving ATF4 overexpression. Significance is shown by *P < 0.05 and **P < 0.01. D, Figure 4G. Colony formation assays in OS and OS/BTZ cells, with either ATF4 or RET alterations mediated by the transient transfection of siRNAs or expression vectors, cultured in 100 nM for the first two days of the experiment. The cells were fixed, stained, and photographed after two weeks. Error bars represent mean ± SD from three independent experiments (*P < 0.05, **P < 0.01, ***P < 0.001; n.s., non-significant).    Figure 7H. Colony formation assay was performed using paired OS and OS/BTZ cells transfected with control or siGRP78, GRP78, ATF4 and RET vectors with BTZ (100 nM) treatment for the first two days of the experiment. The cells were fixed, stained, and photographed after 14 days. Error bars represent mean ± SD from three independent experiments (*P < 0.05, **P < 0.01). B, Relative luciferase activity driven by HSPA5promoter in U-2 OS and U-2 OS/BTZ cells. Cells were cotransfected with increasing concentrations of ATF4 plasmid and HSPA5 promoter (-457 to +1)-luciferase reporter plasmid for 24 h, and then the luciferase values were determined. C, ATF4 binds to the HSPA5 promoter. U-2 OS and HOS cells were transfected with ATF4 or vector control. Chromatin immunoprecipitation was performed using either a control IgG antibody or antibody against ATF4. PCR primers were designed to amplify the specific HSPA5 promoter fragment spanning from -457 to +1. Primers for the DDIT3 promoter were used as positive control. D, ATF4 transcriptionally regulates HSPA5 through the DNA binding domain.

Quantification analysis of
Cells were transfected with the shRNA-ATF4, ATF4 or ΔDBD ATF4 expression plasmid and the HSPA5 promoter (-457 to +1)-luciferase reporter plasmid, and luciferase activity was determined 24 h after transfection. E, ATF4-mediated HSPA5 promoter repression in wild-type OS requires the CRE element. U-2 OS cells were transfected with the plasmid encoding the firefly luciferase gene driven by the illustrated promoter together with vector control (white bar) or the plasmid encoding ATF4 (black bar). After 24 h, cells were harvested and measured for luciferase activity.
Data are shown as the mean ± SEM. Statistical significance was determined by Student's t-test. *P < 0.05, **P < 0.01; n.s., non-significant. F, Indirect immunofluorescence detecting the expression of ATF4 in U-2 OS/BTZ cells before subjected to Piperine or Ribociclib for 24 h.
Images were taken using the high-content imaging system. Scale bars, 10 μm. G, OS/BTZ cells transiently transfected with siCon or siATF4 were challenged with piperine or ribociclib for 24 h.
Cell lysates were collected, and the indicated proteins were analysed. Data from representative immunoblots of three independent assays are shown. Figure S8. A, Quantification analysis of Figure 8A. Proliferation of control or piperine and ribociclib-treated OS and OS/BTZ cells was evaluated by colony formation assays. Error bars represent mean ± SD from three independent experiments (*P < 0.05, **P < 0.01). B, Western blotting analysis of the basal protein level of ATF4 with an increasing loading lysis from HOS and U-2 OS cells. C, Western blotting analysis of another 4 stress-related proteins ATF6, ERN1, HSP90B1and MAPK8 (JNK) in the parental and BTZ-resistant U-2 OS cells. D, The half-life of ATF4 protein in U-2 OS and U-2 OS/BTZ cell lines. Cells transfected with FLAG-ATF4 or control vector were treated with CHX (10 μM), and the expression of ATF4 was determined by immunoblotting at the indicated times. Data are representative immunoblots of three independent assays.