MS#: ONC-2006-01136 (revised) supplementary data

November 25, 2006

GRIM-19 associates with the serine protease HtrA2 for promoting cell death

Xinrong Ma¶, Sudhakar Kalakonda¶, Srinivasa M. Srinivasula$, Sekhar P. Reddy‡, Leonidas C. Platanias♣, Dhananjaya V. Kalvakolanu*

GreenebaumCancerCenter, University of MarylandSchool of Medicine, Baltimore, MD$Laboratory of Immune Cell Biology, National Cancer Institute, Bethesda, MD

‡Department of Environmental Health Sciences, The Johns Hopkins University Bloomberg School of Public Health, Baltimore, MD, ♣Lurie Cancer Center, Northwestern University School of Medicine, Chicago, IL

Most experiments shown in the main paper were peformed in HeLa, PC3 and 293 cells to demonstrate that the GRIM-19:HtrA2 interactions are not cell type specific.

GRIM-19 interacts with HtrA2: To identify the cellular proteins that interact with GRIM-19, we screened a yeast 2-hybird (Y2H) library prepared from human bone marrow(expressing ~1x106 cDNAs)and identified 16 individual positive colonies from a library in the 1st round. Among these, 4 independent clones were identified later as those coding for the serine protease HtrA2.Fig.S1 shows a typical interaction between GRIM-19 and HtrA2. While all yeast clones carrying individual plasmids grew normally a nutrient-rich growth medium, only the clone that coexpressed both GRIM-19 and HtrA2 thrived on selective medium. These cells turned blue when the medium was supplemented with X-galactoside (data not shown), again indicating that GRIM-19 and HtrA2 proteins physically interact. All 4 clones contained the c-terminal sequences of HtrA2. Two of them contained sequences capable of coding for a HtrA2 peptide with a predicted molecular weight of 10.4 kDa (133 amino acids) and the other for a peptide of 21.5kDa (199 amino acids). The largest one of them contained a ~1.050 kb of HtrA2 cDNA. Peptides coded by the smallest two of these clones harbor a PDZ domain, a signature motif common to the HtrA2 family of proteases, which permits protein-protein interactions [Pallen, 1997 #3018]. Thus, it appears that the PDZ domain of HtrA2 is sufficient for binding to GRIM-19.

Fig. S1:GRIM-19 binds to HtrA2 in Y2H screen. Yeast cells transformed with the indicated expression vectors were grown on plates with media containing (A) or lacking (B) Tryptophan, Leucine and Histidine. Plasmids transformed in each case were -1: TAD; 2: DBD; 3: HtrA2; 4: GRIM-19; 5: GRIM-19+TAD; 6: HtrA2+GRIM-19. This figuretypifies an interaction between GRIM-19 and a protein produced by the shortest HtrA2 clone found in the Y2H screen. (C) A schematic representation of the structures of HtrA2 clones used in the Y2H screen in panels A and B. FL-HtrA2: full length found in mammalian cells. HtrA2 clone: shortest of HtrA2 clones (c-terminal 133 aa) rescued in the Y2H screen. DBD: GAL4-DBD alone: TAD: GAL4-Transactivation domain alone; GAL4-TAD-GRIM-19: a chimerical yeast plasmid expressing human GRIM-19 as fusion protein with the TAD of GAL4.

In vitro and in vivo interactions between GRIM-19 and HtrA2:

To demonstrate the occurrence of IFN/RA-induced HtrA2 and GRIM-19 interactions in normal cells, we immunoprecipitated the endogenous proteins from a non-oncogenic immortalized mammary epithelial cell line hTERT-HME1 (Fig. S2A-C). Equivalent protein extracts from control and IFN/RA treated (for 6h) cells were immunoprecipitated with normal and HtrA2-specific IgGs. The products were subjected to WB analysis with GRIM-19-specific monoclonal antibodies. As shown inFig. S2A, only the HtrA2 specific IgG, but not the normal IgG, was able IP the GRIM-19 protein. More importantly, IFN/RA-treatment significantly enhanced these interactions. Under these conditions, there was no rise GRIM-19 and HtrA2 protein levels (Figs.S2BC). Thus, GRIM-19:HtrA2 interactions occur in normal cells.

In a separate experiment we have determined whether GRIM-19 directly bound to HtrA2 (Fig. S2D&E). In these experiments we incubated a His-tagged recombinant GRIM-19 protein with in vitro translated 35S-labeled N133-HtrA2 protein. For a control, we used the recombinant expression-tag (18kDa) from the bacterial expression vector pET32A. The proteins were then bound to Nickel-NTA agarose and then washed extensively. The bound proteins were eluted, separated on a 10% SDS-PAGE and subjected to fluorography. As shown in Fig.S2D, radiolabeled HtrA2 specifically bound to GRIM-19 but not to the expression tag.

Fig. S2:PanelsA-C: IFN/RA-induced interactions between endogenous GRIM-19 and HtrA2 in a non-oncogenic mammary epithelial cell line, hTERT-HME1. (A) Eighty percent confluent cultures were exposed to IFN/RA for 6h where indicated with a plus sign. Total protein (120mg) was used for IP analysis with the indicated antibodies. Con: Normal rabbit IgG and HtrA2: HtrA2-specific IgG. Direct interactions between GRIM-19 and HtrA2. pCDNA 3.0 vector or the same vector carrying the ND133 version of HtrA2 was linearized using XhoI and employed for generating HtrA2 RNA in vitro (Ribomax system, Promega Inc). The product was programmed into a nuclease treated rabbit reticulocyte lysate and translated in vitro in the presence of 35S-labeled methionine and cytsteine amino acid mix (100Ci). The reaction products were incubated with Nickel-NTA agarose pre-loaded with either the pET32A tag (control) or His-tagged GRIM-19. After washing the agarose beads, the bound proteins were denatured in SDS-Gel loading buffer and separated on a 10% SDS-gel. The gel was fixed and fluorographed. In panel E 1/50 of the amount used for binding analysis (panel D) was separated on the gel for in put control.

IFN/RA-induced concurrent release of GRIM-19 and HtrA2 from mitochondrion: Since we observed a physical interaction and a functional interdependence between HtrA2 and GRIM-19, we next determined if GRIM-19 and HtrA2 were concurrently released into the cytoplasm during apoptosis. Our previous studies have shown that IFN/RA induced mitochondrial damage occurs slowly with delayed kinetics. MCF-7cells were enucleated with cytochalasin B, prior to IFN/RA-treatment for preventing GRIM-19 gene induction. Mitochondrial and cytosolic fractions were prepared and subjected to WB analyses with cytochrome c (a mitochondrial marker), GRIM-19, HtrA2 and actin (a cytoplasmic marker) specific antibodies. As shown Fig.S3, mitochondrial levels of GRIM-19, HtrA2 and cytochrome c declined following IFN/RA-treatment. As expected, there was no detectable actin in these fractions indicating their purity. Conversely, IFN/RA stimulated release of cytochrome c, HtrA2 and GRIM-19 into the cytoplasm. Cytochrome c and HtrA2 were not found in the cytoplasm prior to IFN/RA treatment. There was, however, a low baseline level of GRIM-19 present in the cytosol, which increased after IFN/RA treatment. There was no change in the total levels of all 4 proteins. Thus, IFN/RA treatment concurrently releases HtrA2 and GRIM-19 from mitochondrion, under these experimental conditions.

Fig, S3: Co-localization of GRIM-19 and HtrA2 in the mitochondrion and their concurrent release into the cytoplasm. Cytoplasts were prepared and then treated with IFN/RA for various hours. Sub cellular fractions were subjected to WB analysis with various antibodies. Only the relevant portions of the blots are shown.

In vitro enhancement of HtrA2 activity by GRIM-19: We also examined if GRIM-19 can promote the activity of HtrA2 in vitro, purified recombinant proteins were incubated with -casein (as substrate). In order to distinguish cooperative effects on the substrate, a suboptimal level of HtrA2 was used in these experiments (Fig.S4A). Neither the assay buffer nor GRIM-19 alone caused any detectable degradation. As expected, HtrA2 alone promoted the degradation of -casein. In the presence of GRIM-19, the degradation of -casein was strongly enhanced. Equal in put of GRIM-19 or HtrA2 into these reactions was ensured with a Western blot analysis of the reactions with specific antibodies (Fig.S4B&C). Quantification of band intensities from several reactions (n=5) showed that the catalytic activity of HtrA2 (Fig.S4E) is increased ~4.8 fold in presence of GRIM-19 (p> 0.001).


Fig. S4: (A) In vitro enhancement of the degradation of -casein by HtrA2 and GRIM-19 combination. Enzyme activity was determined as described in materials and methods. (B&C) Western blots for HtrA2 and GRIM-19. (D) Quantification of HtrA2 activity. Reaction products of -casein-cleavage assays were separated on SDS-PAGE, bands corresponding to substrate (filled section) and cleaved -casein products (empty section) were quantified using UVP Epichem3 system; converted to a percent of the total input and plotted. Each bar shows the mean of 5 separate samples.