The tumour-suppressive function of CLU is explained by its localisation and interaction with HSP60

Abstract

The product of the CLU gene promotes or inhibits tumourigenesis in a context-dependent manner. It has been hypothesised that different CLU isoforms have different and even opposing biological functions, but this theory has not been experimentally validated. Here we show that molecules involved in survival pathways are differentially modulated by the intracellular or secreted forms of CLU. Secreted CLU, which is selectively increased after transformation, activates the survival factor AKT, whereas intracellular CLU inhibits the activity of the oncogenic transcription factor nuclear factor kappa B. Furthermore, intracellular CLU is inactivated by the pro-proliferative and pro-survival activity of the chaperone protein HSP60 in neuroblastoma cells by forming a physical complex. Thus, localisation is key for CLU physiology, explaining the wide range of effects in cell survival and transformation.

Figure 1

Secreted CLU is unable to modulate NF-κB activity. (a) SHSY5Y or (b) LA-N-1 neuroblastoma cells were transfected with a NF-κB luciferase reporter vector, incubated with control or CLU-conditioned medium, as indicated in the bottom of the graphs, and subjected to luciferase assays. Activation of NF-κB by TNF served as a positive control for the assay. Relative luciferase activity is indicated on the y axis. (c) SHSY5Y or (d) LA-N-1 cells were co-transfected with the NF-κB reporter vector and a control (empty MIG), or CLU, expressing plasmid with or without TNF. Error bars indicate standard deviations, and significance is indicated by the P. NS indicates that the difference is not significant

Figure 2

Secreted CLU modulates AKT and is increased relative to intracellular CLU during transformation. (a) Western blot analysis of phosphorylated and total AKT in the indicated neuroblastoma cell lines. Hours of stimulation with control or CLU-conditioned media (CLU-MIG) are indicated on the top of the blots. (b) Western blot analysis of phosphorylated and total ERK in the indicated neuroblastoma cell lines. Hours of stimulation with the control or CLU-conditioned media are indicated on the top of the blots. (c) Western blot analysis of CLU expression in the isogenic cell lines WI38 (primary human fibroblasts) and VA13 (SV40-transformed WI38). (d) Western blot analysis of the expression of CLU in the adrenal glands and tumours resected from MYCN transgenic mice. pCLU, precursor CLU; sCLU, secreted CLU

Figure 3

Identification of HSP60 as a CLU-interacting protein. (a) proteins pulled down by the indicated GST fusion proteins were visualised by colloidal Coomassie staining. A protein of ∼60 kDa interacting with full length and alpha chain CLU (indicated by the arrow) was excised from the gel and subjected to mass spectrometry analysis. (b) Proteins pulled down by the indicated fusion proteins were subjected to western blot analysis with a HSP60 antibody. (c) Co-immunoprecipitation assay. 293 cells were transfected with control or CLU plasmids, lysed and immunoprecipitated with CLU, HSP60 or control antibodies. After transfer to nitrocellulose, immune complexes were subjected to western blot analysis with a HSP60 antibody. Specific binding proteins are indicated by an asterisk

Figure 4

HSP60 and CLU form a complex in neuroblastoma cells. (a) Western blot analysis of LA-N-1 cell lysate showing expression of CLU at different time points after a sub-lethal heat shock. Actin and HSP60 expression were used as a loading control. (b) Reciprocal co-immunoprecipitation of heath shocked LA-N-1 cells with CLU, HSP60 and control antibodies. The native CLU/HSP60 complex is indicated by the arrows. Inp, input

Figure 5

HSP60 positively regulates proliferation and survival of established and primary neuroblastoma cell lines. (a) Western blot analysis to monitor the expression of HSP60 in three SHSY-5Y cell clones obtained after infection with shRNA control (+) or HSP60 (−) targeting viruses. Quantification of HSP60 expression relative to actin is shown in the bottom of the gel. (b) Fluorescence microscopy analysis to visualise GFP expression of the SHSY-5Y clones shown in (a). Pictures were taken the day after infections (t 0) or after 6 days in culture (t 6). (c) Quantification of the experiment shown in (b). (d) Western blot analysis of primary human neuroblastoma cells infected with a control (+) or HSP60 (−) shRNA lentivirus. Relative expression of HSP60 is indicated at the bottom of the gel. (e) Proliferation assay of primary human neuroblastoma cells after ablation of HSP60 expression. (f) PI staining and FACS analysis to assess cell cycle profile and DNA fragmentation in primary human neuroblastoma cells after downregulation of HSP60. Percentages of cells in the different phases of the cell cycle or subG1 DNA (fragmented DNA) are indicated

Figure 6

HSP60 is a negative prognostic marker in human neuroblastoma. (a) Box plot showing the significant association between MYCN amplification and HSP60 expression in 101 neuroblastoma patients. Data were mined through the Oncomine website (http://www.Oncomine.org). Kaplan–Meier survival curves were generated from the Oncogenomics website (http://pob.abcc.ncifcrf.gov/cgi-bin/JK) using the data extracted from the Oberthuer. Black dots indicate outliers. (b) and Seeger (c) affymetrix datasets

Figure 7

Downregulation of CLU rescues apoptosis induced by ablation of HSP60. Cell cycle profiles and DNA fragmentation (indicated by percentages of subG1 DNA, in bold) were assessed by flow cytometry after knockdown of CLU and HSP60 in SHSY-5Ycell clones. The experiment was repeated twice. (a) experiment 1; (b) experiment 2. Quantification of the percentages of cells showing fragmented DNA in the different conditions is shown in the right of each panel

Figure 8

HSP60 signalling is upstream of CLU and NF-κB in neuroblastoma. (a) SHSY5Y or (b) primary human neuroblastoma cells (HNB) were stably infected with scrambled or HSP60 shRNA lentiviruses and subsequently transfected with a NF-κB reporter vector. Luciferase assays show reduced NF-κB activity in the HSP60-interfered cells as compared with the scrambled infected, control cells. Western blot analyses to verify HSP60 downregulation are shown in the right. (c) Analysis of co-expression of HSP60 and NF-κB target genes (boxed) in neuroblastoma patients. The analysis was carried out using tools in the Oncomine website and is based upon the affymetrix datasets indicated in the legend. (d) Western blot analysis showing increased CLU expression in the primary neuroblastoma cells (HNB) after shRNA-induced downregulation of HSP60. (e) Model of the HSP60–CLU axis operating in neuroblastoma cells and the outcome of its manipulation