Identification of a structural motif in the tumor-suppressive protein GRIM-19 required for its antitumor activity

Abstract

We have previously isolated GRIM-19, a novel growth suppressor, using a genetic method. GRIM-19 ablates cell growth by inhibiting the transcription factor signal transducer and activator of transcription 3 (STAT3). Up-regulation of STAT3 and growth promotion were observed in a number of human tumors. Although the tumor-suppressive actions of GRIM-19 are known, the structural elements required for its antitumor actions are not understood. Mutational and protein sequence analyses identified a motif in the N terminus of GRIM-19 that exhibited similarity to certain RNA viral proteins. We show that disruption of specific amino acids within this motif cripples the antitumor actions of GRIM-19. These mutants fail to interact with STAT3 efficiently and consequently do not inhibit growth-promoting gene expression. More importantly, we show that a clinically observed mutation in the N terminus of GRIM-19 also weakened its interaction with STAT3 and antitumor action. Together, these studies identify a major role for the N terminus of GRIM-19 in mediating its tumor-suppressive actions.

Figure 1

Identification of human squamous cell carcinoma cell lines with low GRIM-19 expression. A: The indicated cell lysates (∼40 μg) was used for Western blot analyses with the indicated antibodies. The HSC3 cell line has the lowest GRIM-19 level. WB, Western blot. B: Levels of endogenous GRIM19 and STAT3 transcripts were quantified using real-time PCR analysis with gene-specific primers. RNA isolated from two pooled samples of normal buccal tissue from surgical waste was used as a control. Each bar represents the mean ± SE (n = 6). C: Real-time PCR analysis of steady-state STAT3-responsive gene levels in HSC2 and HSC3 cells. Each bar represents the mean ± SE (n = 4). Bar markings are same as in B. D: Cell growth pattern of HSC2 and HSC3 cells (n = 6). Statistical significance was obtained by Student’s t-test: **P < 0.005; ***P < 0.001.

Figure 2

Identification of a growth-regulatory domain in GRIM-19 protein. A: Effect of wild-type GRIM-19 on soft-agar colony formation by HSC3 cells. Photomicrographs of representative fields are shown. Original magnification, ×20. B and D: Quantification of colony formation. Bars indicate the mean ± SE of triplicates in each experiment (n = 3). The Western blots (WB) below these graphs show the expression levels of GRIM-19 mutants and actin in these cells. A portion of the cells used for colony formation assays (in A) were plated in parallel, without soft agar, and selected with G418 for three weeks. An equal quantity of protein (50 μg) from each transfectant was used for Western blot analysis. C: Identification of proteins with GRIM-19-like domains. A homology map is shown: asterisk (∗), colon (:), and period (.) indicate identity, high similarity, and low similarity, respectively. E: Effect of QDMP mutations on cell motility. Assays were performed as described in Materials and Methods. Right panel shows a typical pattern of motility observed. White lines in these photomicrographs show the original boundaries of the scratched area. The graph below the pictures shows quantification of distances moved at the end of four hours after the initial introduction of a scratch into the monolayers. Each bar represents the mean ± SE obtained with quadruplicates in each experiment (n = 4). F: Cellular localization of GRIM-19 mutants in HCS3 cells. Immunofluorescent images of HSC3 cells infected with the indicated GRIM-19 construct (green channel) and mitochondria visualized using MitoTracker (red channel) are shown. The majority of the protein is found in the mitochondrion, but some nuclear fluorescence is also seen. Arrowhead indicates noninfected cell(s). G: Cellular subfractionation of HSC3 cells expressing the indicated GRIM-19 construct. Distribution of GRIM-19 in the cell is not exclusively mitochondrial as evidenced by its presence in nuclear and cytoplasmic fractions. c-Jun, cytochrome c, and β-tubulin were used as markers for nuclear, mitochondrial, and cytoplasmic fractions, respectively. A Myc-tag-specific antibody was used to detect GRIM-19. Statistical significance was obtained by Student’s t-test: *P < 0.01; **P < 0.005; ***P < 0.001. EV, empty vector control; WT, wild-type.

Figure 3

Effect of N-terminal mutations on GRIM-19 activity. A: Quantitative representation of soft agar colony formation by HSC3 cells in the presence of K5N, a tumor-derived mutation. A Myc-tag-specific antibody was used in Western blots (WB) to detect GRIM-19. B: Effect of GRIM-19 mutants on cell growth. Equal numbers of HSC3 cells expressing the indicated mutants were used in a growth assay. Significant growth differences between wild-type GRIM-19 (WT) and empty-vector control (EV) (P < 0.001), D9A (P < 0.005), Q8A (P < 0.005), and K5N (P < 0.01) are seen. Each data point represents the mean ± SE of eight replicates in each case (n = 4). Statistical significance was obtained by Student’s t-test: ***P < 0.001. EV, empty vector control; WT, wild-type.

Figure 4

Effect of GRIM-19 mutants on STAT3-responsive gene expression. A: Real-time PCR analysis with gene-specific primers was used for quantifying gene expression in HSC3 cells stably expressing the indicated mutants. White and black bars represent steady state and interleukin-6-induced mRNA levels. B: The indicated luciferase reporters driven by the promoters of various STAT3-inducible genes, were used for studying the effect of GRIM-19 mutants on gene expression. Transfection efficiency was normalized to the expression of the internal control (β-galactosidase). Each data point represents the mean ± SE of triplicates in each experiment (n = 3). Statistical significance was obtained by Student’s t-test: **P < 0.005; ***P < 0.001. EV, empty vector control; WT, wild-type.

Figure 5

Effect of GRIM-19 mutants on S3C-responsive gene expression, apoptosis, and transformation by v-Src. A: Stat3-responsive gene expression was measured using real-time PCR after infection of 3Y1 cells expressing S3C with lentiviral particles coding for the indicated GRIM-19 mutants. Each bar represents the mean ± SE of six replicates in each experiment (n = 3). B: Effect of Fas ligation on cell death in the presence of S3C and GRIM-19 mutants. The Jo-1 monoclonal antibody (200 ng/ml), which activates the Fas death receptor, was incubated for 16 hours before cells were stained with tetramethylrhodamine B isothiocyanate-labeled annexin V. The percentage of cells stained positive (apoptotic) was quantified after florescence-activated cell sorting and plotted. Each data point represents the mean ± SE of triplicates in each experiment (n = 3). C: Effect of GRIM-19 mutants on v-Src-induced cellular transformation. 3Y1 cells expressing v-Src were infected with lentiviral vectors coding for the indicated mutants. Soft agar colony formation was monitored. Mean ± SE were plotted in each case (n = 5). Myc-tag-specific antibody was used in Western blots to detect GRIM-19. D: Interaction of STAT3 with GRIM-19 mutants in HSC3 cells. Coimmunoprecipitation (IP) and Western blot (WB) analysis were performed after infection with lentiviral particles coding for GRIM-19. Statistical significance was obtained by Student’s t-test: *P < 0.01; **P < 0.005; ***P < 0.001. EV, empty vector control; WT, wild-type.

Figure 6

Effect of GRIM-19 mutants on tumor growth in vivo. HSC3 cell lines expressing the indicated genes were transplanted into athymic nude mice (n = 10) subcutaneously, and tumor growth was monitored as in our earlier studies. Significant growth differences between wild-type GRIM-19 and empty vector control (P < 0.001), D9A (P < 0.005), and K5N (P < 0.005) are seen. Statistical significance was obtained by Student’s t-test.